View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

ECONOMIC REPORT OF
THE PRESIDENT

Together With
THE ANNUAL REPORT
of the
COUNCIL OF ECONOMIC ADVISERS

Transmitted to the Congress
March 2014

economic
re p ort
of the
president

transmitted to the congress
march 2014
together with

the annual report
of the

council of economic advisers
united states government printing office
washington : 2014

-092301-2

90000

For sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; DC area (202) 512-1800
Fax: (202) 512-2104 Mail: Stop IDCC, Washington, DC 20402-0001
I S B N 978-0-16-092301-2

23012

C O N T E N T S
ECONOMIC REPORT OF THE PRESIDENT.....................................................1
ANNUAL REPORT OF THE COUNCIL OF ECONOMIC ADVISERS*........7
CHAPTER 1.	
PROMOTING OPPORTUNITY AND SHARED,
SUSTAINABLE GROWTH................................................... 21
CHAPTER 2.	

THE YEAR IN REVIEW AND THE YEARS AHEAD...... 45

CHAPTER 3.	
THE ECONOMIC IMPACT OF THE AMERICAN
RECOVERY AND REINVESTMENT ACT FIVE
YEARS LATER........................................................................ 91
CHAPTER 4.	
RECENT TRENDS IN HEALTH CARE COSTS, THEIR
IMPACT ON THE ECONOMY, AND THE ROLE OF
THE AFFORDABLE CARE ACT.......................................147
CHAPTER 5.	

FOSTERING PRODUCTIVITY GROWTH.....................179

CHAPTER 6.	
THE WAR ON POVERTY 50 YEARS LATER:
A PROGRESS REPORT.......................................................221
CHAPTER 7.	
EVALUATION AS A TOOL FOR IMPROVING
FEDERAL PROGRAMS.......................................................269
REFERENCES	 ................................................................................................299
APPENDIX A.	
REPORT TO THE PRESIDENT ON THE ACTIVITIES
OF THE COUNCIL OF ECONOMIC ADVISERS
DURING 2013.........................................................................345
APPENDIX B.	
STATISTICAL TABLES RELATING TO INCOME,
EMPLOYMENT, AND PRODUCTION.............................359

____________

*For a detailed table of contents of the Council’s Report, see page 11.

 iii

economic report
of the

president

economic report of the president

To the Congress of the United States:
This year’s Economic Report of the President describes how after 5
years of grit and determined effort, the United States is better-positioned
for the 21st century than any other nation on Earth. We’ve now experienced 4 straight years of economic growth with more than 8 million new
private-sector jobs. Our unemployment rate is the lowest it’s been in more
than 5 years. Our deficits have been cut by more than half. For the first
time in nearly 20 years, we produce more oil at home than we buy from
the rest of the world. The housing market is rebounding, manufacturers
are adding jobs for the first time since the 1990s, and we sell more of what
we make to the rest of the world than ever before.
But in many ways, the trends that have threatened the middle class
for decades have grown even starker. While those at the top are doing better than ever, average wages have barely budged. Inequality has deepened.
Too many Americans are working harder and harder just to get by, and
too many still aren’t working at all. Our job is to reverse those trends. It is
time to restore opportunity for all—the idea that no matter who you are or
how you started out, with hard work and responsibility, you can get ahead.
That’s why this must be a year of action. I’m eager to work with the
Congress to speed up economic growth, strengthen the middle class, and
build new ladders of opportunity into the middle class. But America does
not stand still, and neither will I. Wherever and whenever I can take steps
without legislation to expand opportunity for more American families, I
will. Because opportunity is who we are. And the defining project of our
generation is to restore that promise.
Simply put, this opportunity agenda has four parts. Number one is
more new jobs. Number two is training more Americans with the skills to
fill those jobs. Number three is guaranteeing every child access to a world-

Economic Report of the President  |  3

class education. And number four is making sure hard work pays off for
every American.
With the economy picking up speed, companies say they intend to
hire more people this year. We should make that decision even easier for
them by closing wasteful tax loopholes and lowering tax rates for businesses that create jobs here at home, and use the money we save in the
process to create jobs rebuilding our roads, upgrading our ports, and
unclogging our commutes. We should help America win the race for the
next wave of high-tech manufacturing jobs by connecting businesses and
universities in hubs for innovation. We should do more to boost exports
and fund basic research. We should maintain our commitment to an allof-the-above-energy strategy that is creating jobs and leading to a safer
planet. Finally, we should heed the call of business leaders, labor leaders,
faith leaders, and law enforcement, and fix our broken immigration system. Independent economists say this will grow our economy and shrink
our deficits by almost $1 trillion in the next two decades. We should get it
done this year.
Creating jobs is step one, but in this rapidly-changing economy,
we also must make sure every American has the skills to fill those jobs.
I’ve asked Vice President Biden to lead an across-the-board reform of
America’s training programs to make sure they have one mission: training
Americans with the skills employers need, and matching them to good jobs
that need to be filled right now. That means more on-the-job training, and
more apprenticeships that set a young worker on an upward trajectory for
life. It means connecting companies to community colleges that can help
design training to fill their specific needs.
I’m also convinced we can help Americans return to the workforce
faster by reforming unemployment insurance so that it’s more effective in
today’s economy. But first, the Congress needs to restore the unemployment insurance it let expire at the end of last year, affecting around 2 million workers.
Of course, it’s not enough to train today’s workforce. We also have
to prepare tomorrow’s workforce, by guaranteeing every child access to a
world-class education. Our high school graduation rate is higher than it’s
been in 30 years, and more young people are earning college degrees than
ever before. The problem is we’re still not reaching enough kids, and we’re
not reaching them in time.
That has to change. I am repeating a request I made last year asking you to help States make high-quality preschool available to every four

4  |  Economic Report of the President

year-old. In the meantime, I’m going to pull together a coalition of elected
officials, business leaders, and philanthropists willing to help more kids
access the high-quality early education they need. I’ll also work to redesign
high schools and partner them with colleges and employers that offer the
real-world education and hands-on training that can lead directly to a job
and career, and follow through on my pledge to connect 99 percent of our
students to high-speed broadband over the next 4 years. With the support
of the FCC, we’ve announced a down payment to start connecting more
than 15,000 schools and 20 million students over the next 2 years, without
adding a dime to the deficit, and with the help of some of America’s top
companies, we’re going to make the most of these new connections.
My Administration is also shaking up our system of higher education, so that no middle-class family is priced out of a college education.
We’re offering millions the opportunity to cap their monthly student loan
payments to ten percent of their income, and I will continue to look for
other ways to see how we can help even more Americans who feel trapped
by student loan debt.
But we know our opportunity agenda won’t be complete—and too
many young people entering the workforce today will see the American
Dream as an empty promise—unless we do more to make sure hard
work pays off for every single American. This year, we should do more to
secure a women’s right to equal pay for equal work. We should expand the
Earned Income Tax Credit to help more workers without children make
ends meet, and help more Americans save for retirement through the new
“MyRA” plans my Administration is creating. We should protect taxpayers from ever footing the bill for a housing crisis ever again. And we will
continue the work of making sure every American has access to affordable,
quality health insurance that’s there for them when they need it.
And we should raise a minimum wage that in real terms is worth
less than it was when Ronald Reagan took office. In the year since I first
asked the Congress to raise the minimum wage, six States raised theirs, and
more companies like Costco see paying fair wages as one of the best ways
to reduce turnover, increase productivity, and boost profits. As America’s
chief executive, I agree, which is why I signed an Executive Order requiring
Federal contractors to pay their federally funded employees a fair wage of
at least $10.10 an hour for new contracts. There is a bill in front of both the
House and the Senate that would raise the minimum wage to $10.10 for all
Americans. The Congress should pass that bill and give America a raise.

Economic Report of the President  |  5

I believe this can be a breakthrough year for America. But it falls
to all of us to grow the economy and create new jobs, to strengthen the
middle class, and to build new ladders of opportunity for folks to work
their way into the middle class. So in the coming months, let’s see where
we can make progress together. Let’s continue to make this a year of action.
Together, we can restore an economy that works for everybody, and our
founding vision of opportunity for all.

the white house
march 2014

6  |  Economic Report of the President

the annual report
of the

council of economic advisers

letter of transmittal
Council of Economic Advisers
Washington, D.C., March 10, 2014

Mr. President:
The Council of Economic Advisers herewith submits its 2014
Annual Report in accordance of the Employment Act of 1946 as amended
by the Full Employment and Balanced Growth Act of 1978.
Sincerely yours,

Jason Furman
Chairman

Betsey Stevenson
Member

James H. Stock
Member

 9

C O N T E N T S

CHAPTER 1. PROMOTING OPPORTUNITY AND SHARED,
SUSTAINABLE GROWTH........................................................................... 21
THE ECONOMY FIVE YEARS AFTER THE CRISIS.................................. 23
HOW WE GOT HERE: THE ADMINISTRATION’S RESPONSE TO
THE CRISIS.................................................................................................... 26
SOURCES OF OPPORTUNITY IN 2014 AND BEYOND............................ 28

Cyclical Factors..................................................................................... 29
Structural Trends ................................................................................. 30
Long-Term Fiscal Sustainability......................................................... 33

THE CHALLENGES THAT REMAIN AND THE PRESIDENT’S
PLANS TO ADDRESS THEM........................................................................ 34

Continuing to Restore the Economy to its Full Potential............... 34
Expanding the Economy’s Potential.................................................. 37
Promoting Economic Opportunity..................................................... 40

CONCLUSION............................................................................................... 43

CHAPTER 2. THE YEAR IN REVIEW AND THE YEARS
AHEAD................................................................................................................ 45
KEY EVENTS OF 2013................................................................................... 46

Aggregate Output Growth During the Year..................................... 46
Fiscal Policy .......................................................................................... 47
Monetary Policy ................................................................................... 50
Financial Markets................................................................................. 51
International Developments................................................................ 53

DEVELOPMENTS IN 2013 AND THE NEAR-TERM OUTLOOK............. 56

Consumer Spending.............................................................................. 56
Business Investment.............................................................................. 59
State and Local Governments............................................................. 62
International Trade.............................................................................. 64
Housing Markets................................................................................... 67
11

Energy..................................................................................................... 72
Labor Markets....................................................................................... 76
Wage Growth and Price Inflation...................................................... 81
THE LONG-TERM OUTLOOK.................................................................... 84

The 11-Year Forecast............................................................................ 84
Growth in GDP over the Long Term................................................. 86

CONCLUSION............................................................................................... 89

CHAPTER 3. THE ECONOMIC IMPACT OF THE AMERICAN
RECOVERY AND REINVESTMENT ACT FIVE YEARS
LATER.................................................................................................................. 91
THE 2007-09 RECESSION AND THE EARLY POLICY RESPONSES........ 93

Initial Policy Responses........................................................................ 94

AN OVERVIEW OF THE RECOVERY ACT AND SUBSEQUENT
JOBS MEASURES........................................................................................... 95

The Recovery Act................................................................................... 96
Subsequent Jobs Measures................................................................... 99
Automatic Countercylical Measures................................................ 100
Total Fiscal Response.......................................................................... 103

NEAR-TERM MACROECONOMIC EFFECTS OF THE RECOVERY
ACT AND SUBSEQUENT FISCAL LEGISLATION...................................103

Model-Based Estimates of the Macroeconomic Effects of the
Recovery Act and Subsequent Fiscal Legislation............................ 105
Cross-State Evidence........................................................................... 111
International Comparison................................................................. 114
Benchmarking the Economy’s Performance Since 2009............... 114

EFFECTS OF THE RECOVERY ACT IN PROVIDING RELIEF FOR
INDIVIDUALS.............................................................................................117

Tax Cuts for Families......................................................................... 118
Unemployment Insurance................................................................. 119

THE EFFECT OF THE RECOVERY ACT ON LONG-TERM
GROWTH.....................................................................................................122

Protecting and Expanding Investments in Physical Capital........ 123
Protecting and Expanding Investments in Human Capital......... 126
Investments in Technology and Innovation................................... 128
Fiscal Sustainability and the Recovery Act..................................... 131

CONCLUSION.............................................................................................132

12  |  Annual Report of the Council of Economic Advisers

APPENDIX 1: COMPONENTS OF THE RECOVERY ACT AND
SUBSEQUENT FISCAL MEASURES..........................................................133

Tax Relief ............................................................................................ 134
Aid to Affected Individuals................................................................ 136
State Fiscal Relief................................................................................ 137
Investments.......................................................................................... 138
Subsequent Fiscal Measures.............................................................. 139

APPENDIX 2: FISCAL MULTIPLIERS: THEORY AND EMPIRICAL
EVIDENCE...................................................................................................139

Forward-Looking Models with Rigidities ...................................... 141
Time Series Evidence ......................................................................... 142
Cross-Sectional Multipliers ............................................................... 145

CHAPTER 4. RECENT TRENDS IN HEALTH CARE COSTS,
THEIR IMPACT ON THE ECONOMY, AND THE ROLE OF
THE AFFORDABLE CARE ACT..............................................................147
RECENT TRENDS IN HEALTH CARE COSTS.........................................150
WHAT IS HAPPENING NOW, AND WHAT WILL HAPPEN NEXT?...156

The Role of the 2007-09 Recession................................................... 156
Non-ACA Factors Affecting Health Spending Growth................. 160
The Role of the Affordable Care Act................................................ 162

ECONOMIC BENEFITS OF SLOW HEALTH SPENDING GROWTH....171

Higher Living Standards.................................................................... 171
Lower Deficits...................................................................................... 173
Higher Employment and Economic Growth................................... 176

CONCLUSION.............................................................................................178

CHAPTER 5. FOSTERING PRODUCTIVITY GROWTH.............179
TRENDS IN TOTAL FACTOR PRODUCTIVITY.....................................181

Labor Productivity, Total Factor Productivity, and Multifactor
Productivity.......................................................................................... 181
Postwar U.S. Productivity Growth................................................... 182

PRODUCTIVITY GROWTH AND INEQUALITY GROWTH.................189

Trends in Inequality, Productivity Growth, and
Compensation...................................................................................... 189
Technological Change and Inequality............................................. 190

Contents  | 13

POLICIES TO FOSTER PRODUCTIVITY GROWTH AND TO HELP
ENSURE THAT EVERYONE BENEFITS FROM IT..................................193
TELECOMMUNICATIONS AND PRODUCTIVITY GROWTH.............197

Innovation and Investment............................................................... 197
Four Key Areas for Telecommunications Policy............................ 198
Challenges to Broad Adoption of Telecommunications
Technology........................................................................................... 209

PATENTS......................................................................................................212

Standard-Essential Patents................................................................ 213
Patent Assertion Entities.................................................................... 215

CONCLUSION.............................................................................................217

CHAPTER 6. THE WAR ON POVERTY 50 YEARS LATER:
A PROGRESS REPORT...............................................................................221
MEASURING POVERTY: WHO IS POOR IN AMERICA?.......................223

Measuring Poverty.............................................................................. 223
The Official Poverty Measure........................................................... 223
The Supplemental Poverty Measure................................................ 225
Who is Poor?........................................................................................ 228
Employment......................................................................................... 229
Education Level................................................................................... 230
Children................................................................................................ 230
The Elderly.......................................................................................... 230
Women................................................................................................. 232
Race and Ethnicity.............................................................................. 232
People with Disabilities...................................................................... 233
Rural and Urban Communities........................................................ 233

ASSESSING THE WAR ON POVERTY......................................................234

Context................................................................................................. 234
Correcting the Historical Account of Poverty Since the 1960s .... 240
Measuring the Direct Impact of Antipoverty Efforts..................... 242

THE ROLE OF ANTIPOVERTY PROGRAMS: A CLOSER LOOK...........245

Antipoverty Effects of Specific Programs......................................... 245
The Effects of Antipoverty Programs on Work and Earnings...... 248
Economic Mobility.............................................................................. 252
Intergenerational Returns.................................................................. 255

14  |  Annual Report of the Council of Economic Advisers

THE OBAMA ADMINISTRATION’S RECORD AND AGENDA TO
STRENGTHEN ECONOMIC SECURITY AND INCREASE
OPPORTUNITY...........................................................................................258

Taking Immediate Action During the Economic Crisis................ 258
Expanding Health Care Security ..................................................... 260
Rewarding Work................................................................................. 260
Empowering Every Child with a Quality Education .................... 263
Creating Jobs and Growing Our Economy .................................... 265
Investing in and Rebuilding Hard-Hit Communities................... 265

CONCLUSION.............................................................................................266

CHAPTER 7. EVALUATION AS A TOOL FOR IMPROVING
FEDERAL PROGRAMS...............................................................................269
CONDUCTING RIGOROUS IMPACT EVALUATIONS IN FEDERAL
PROGRAMS..................................................................................................271

Estimation of Causal Effects of a Program or Intervention......... 272
Other Criteria for High-Quality, Successful Impact
Evaluations.......................................................................................... 274
Lower-Cost Ways for Impact Evaluations to Facilitate Real-Time
Learning................................................................................................ 275

IMPACT OF THE EVIDENCE-BASED AGENDA.....................................278

Uses of Evaluation ............................................................................. 278
Building Evidence when Existing Evidence is Limited.................. 283

FURTHERING THE EVIDENCE AGENDA...............................................286

Legislative Support for Evaluation .................................................. 287
Building Evaluation into the Design of Programs ........................ 290
Developing the Capacity to Link to Other Administrative and
Survey Data Sources........................................................................... 294
Facilitating Researcher Access to Federal Data while
Protecting Privacy ......................................................................... 296

CONCLUSION.............................................................................................297

REFERENCES................................................................................... 299

A.	
B.	

APPENDIXES
Report to the President on the Activities of the Council of
Economic Advisers During 2013........................................................345
Statistical Tables Relating to Income, Employment, and
Production..............................................................................................359

Contents  | 15

1.1.	
1.2.	
1.3.	
1.4.	
1.5.	
1.6.	
1.7.	
1.8.	
1.9.	
1.10.	
1.11.	
1.12.	
1.13.	
1.14.	
1.15.	
2.1.	
2.2.	
2.3.	
2.4.	
2.5.	
2.6.	
2.7.	
2.8.	
2.9.	
2.10.	
2.11.	
2.12.	
2.13.	
2.14.	

FIGURES
Monthly Change in Private Nonfarm Payrolls, 2007–2014.............. 22
U.S. Merchandise and Overall Trade Deficits, 2000–2013............... 24
Major Deficit Reduction Episodes Over a Four–Year Period
Since the Demobilization from WWII................................................. 24
Real GDP Per Working–Age Population in 2007–2008 Banking
Crisis Countries, 2007–2013.................................................................. 27
Quarterly Effect of the Recovery Act and Subsequent
Fiscal Measures on Employment, 2009–2012..................................... 27
Change in Poverty Rate from 2007–2010, With and Without
Tax Credits and Benefits........................................................................ 29
Domestic Crude Oil Production and Net Imports, 2000–2013....... 31
Growth in Real Per Capita National Health Spending,
1961–2013................................................................................................. 33
Unemployment Rate by Duration, 1994–2014................................... 35
Growth in Real Average Hourly Earnings for Production and
Nonsupervisory Workers, 2007–2014.................................................. 36
Real Median Family income, 1980–2012............................................. 36
Building Permits for New Residential Units, 1960–2014.................. 38
Growth in Total Factor Productivity, 1953–2012.............................. 39
Share of National Income Earned by Top 1 Percent,
1915–2012................................................................................................. 41
Growth in Productivity and Average Wage, 1947–2013................... 41
Mean GDP Growth, 2007–2013............................................................ 47
Federal Budget Deficit, 1950–2015....................................................... 50
Interest Rates, 2010–2014...................................................................... 52
Treasury Bills Maturing in Late October–Early
November, 2013...................................................................................... 53
Current Account Balance by Country, 2000–2013............................ 55
Cumulative Flows into Mutual and Exchange–Traded Funds
Investing in Emerging Markets, 2010–2014........................................ 56
Household Deleveraging, 1990–2013................................................... 58
Consumption and Wealth Relative to Disposable Personal
Income (DPI), 1952–2013...................................................................... 58
Business Investment and the Acceleration of Business Output,
1965–2013................................................................................................. 62
Real State and Local Government Purchases During Recoveries.... 63
State and Local Pension Fund Liabilities, 1952–2013........................ 64
Trade in Goods and Services, 2007–2013............................................ 65
U.S. Exports Growth, 2009–2013.......................................................... 65
Current Account Balance, 1985–2013.................................................. 66

16  |  Annual Report of the Council of Economic Advisers

2.15.	 Housing Starts, 1960–2013.................................................................... 71
2.16.	 National House Price Indexes, 2000–2013.......................................... 71
2.17.	 Cumulative Over– and Under–Building of Residential and
Manufactured Homes, 1996–2013........................................................ 73
2.18.	 Petroleum Net Imports, 1980–2015..................................................... 74
2.19.	 Monthly Crude Oil Production and Net Imports, 1990–2013......... 74
2.20.	 Wind and Solar Energy Production, 2000–2013................................ 75
2.21.	 U.S. Per Capita Consumption of Gasoline and Real Gasoline
Prices, 2000–2013.................................................................................... 76
2.22.	 Unemployment Rate, 1979–2014.......................................................... 78
2.23.	 Nonfarm Payroll Employment, 2007–2014......................................... 78
2.24.	 Unemployment Rate by Duration, 1990–2014................................... 79
2.25.	 Predicted vs. Actual Manufacturing Payroll Employment,
2000–2014................................................................................................. 80
3.1.	 Recovery Act Programs by Functional Categories............................. 98
3.2.	 Recovery Act and Subsequent Fiscal Measures by Functional
Category..................................................................................................102
3.3.	 Automatic Stabilizers and the Budget Balance, 2009–2013............102
3.4.	 Fiscal Expansion as a Percentage of GDP.........................................103
3.5.	 Estimates of the Effects of the Recovery Act on the Level of
GDP, 2009–2013....................................................................................109
3.6.	 Estimates of the Effects of the Recovery Act on Employment,
2009–2013...............................................................................................109
3.7.	 Quarterly Effect of the Recovery Act and Subsequent Fiscal
Measures on GDP, 2009–2012............................................................110
3.8.	 Quarterly Effect of the Recovery Act and Subsequent Fiscal
Measures on Employment, 2009–2012..............................................110
3.9.	 Change in Nonfarm Employment......................................................113
3.10.	 Disposable Personal Income With and Without the ARRA..........119
3.11.	 Recovery Act Cumulative Public Investment Outlays,
2009–2013...............................................................................................124
3.12.	 Advanced Renewable Electric Power Net Generation,
2000–2012 .............................................................................................129
4.1.	 Growth in Real Per Capita National Health Expenditures,
1961–2013...............................................................................................154
4.2.	 General and Health Care Price Inflation, 1960–2013......................154
4.3.	 Growth in Real Per Enrollee Health Spending by Payer.................157
4.4.	 Medicare 30–Day, All–Condition Hospital Readmission
Rate, 2007–2013.....................................................................................166
4.5.	 Inflation–Adjusted Premiums for Medicare Parts B and D,
2000–2014...............................................................................................173
4.6.	 Recent CBO Projections of Medicare and Medicaid Outlays........176

Contents  | 17

5.1.	
5.2.	

Nonfarm Private Business Productivity Growth, 1949–2012.........185
15–Year Centered Moving Average of Annual Growth Rates for
Labor and Multifactor Productivity, 1956–2005..............................187
5.3.	 Growth in Productivity and Average Wage, 1947–2013.................190
5.4.	 Basic Research Expenditures in the U.S. by Source Funding,
2010.........................................................................................................195
5.5.	 Composition of Total R&D Spending as a Share of GDP,
1953–2011...............................................................................................196
5.6.	 Relative Investment of the Telecommunications Sector, 2011......199
5.7.	 Exclusive and Shared Allocation of Radio Spectrum.......................202
5.8.	 Federal Agencies with Most Spectrum Assignments.......................203
5.9.	 Percentage of Households with Access to Download Speeds
of 6 Megabytes per Second or Greater...............................................209
5.10.	 Patents Issued in the U.S. by Technological Category....................213
6.1.	 Trends in the Official Poverty Measure, 1959–2012........................235
6.2.	 Average Real Household Income by Quintile, 1967–2012.............236
6.3.	 Women’s 50–10 Wage Gap vs Real Minimum Wage,
1973–2012...............................................................................................238
6.4.	 Official vs Anchored Supplemental Poverty Rates,
1967–2012...............................................................................................243
6.5.	 Trends in Market and Post–Tax, Post–Transfer Poverty
1967–2012...............................................................................................244
6.6.	 Trends in Market and Post–Tax, Post–Transfer Deep Poverty,
1967–2012...............................................................................................246
6.7.	 Percentage Point Impact on SPM Child Poverty for
Selected Years.........................................................................................250
6.8.	 Percentage Point Impact on Deep SPM Child Poverty for
Selected Years.........................................................................................250
6.9.	 Real Per Capita Expenditures on Select Programs, 1967–2012.....254
6.10.	 Economic Mobility for Children from First Income Quintile.......256
6.11.	 Recovery Act and Subsequent Extensions: Cumulative
Person–Years Kept from Poverty, 2008–2012..................................261
7.1.	 Outlays for Grants to State and Local Governments,
1992–2012...............................................................................................281
7.2.	 Inventory of Beds for Homeless and Formerly Homeless People,
2007–2012...............................................................................................287
2.1.	
2.2.	

TABLES
Administration Economic Forecast...................................................... 85
Supply–Side Components of Actual and Potential Real GDP
Growth, 1952–2014................................................................................. 87

18  |  Annual Report of the Council of Economic Advisers

3.1.	

Forecasted and Actual Real GDP Growth and
Unemployment Rate............................................................................... 95
3.2.	 An Overview of Recovery Act Fiscal Impact...................................... 99
3.3.	 Recovery Act Programs by Functional Categories............................. 99
3.4.	 Fiscal Support for the Economy Enacted After the
Recovery Act..........................................................................................101
3.5.	 Estimated Output Multipliers for Different Types of Fiscal
Support....................................................................................................108	
3.6.	 Estimates of the Effects of the Recovery Act on the Level
of GDP....................................................................................................112
3.7.	 Tax Relief and Income Support in the Recovery Act and
Subsequent Measures, 2009–2012......................................................118
3.8.	 Recovery Act Long Term Growth Investment by Category...........123
3.9.	 Recovery Act Outlays, Obligations, and Tax Reductions...............135
3.10.	 Recovery Act Fiscal Stimulus by Functional Category....................136
3.11.	 Fiscal Support for the Economy Enacted After the
Recovery Act .........................................................................................140
3.12.	 Summary of Cross–Sectional Fiscal Multiplier Estimates..............146
4.1.	 Real Per Capita NHE Annual Growth Rates by Payer and
Spending Category................................................................................151
4.2.	 Recent Trends in Several Indicators of Health Care Spending
and Price Growth..................................................................................155
5.1.	 Sources of Productivity Improvement, Nonfarm Private
Business, 1948–2012.............................................................................184
5.2.	 Nonfarm Private Business Growth.....................................................186
5.3.	 Average Annual Rates of Change in the Nonfarm Business
Sector.......................................................................................................191
6.1.	 Poverty Rates by Selected Characteristics, 1959–2012....................229
6.2.	 Poverty Rate Reduction from Government Programs, 2012..........248
Box 2–1:	
Box 2–2:	
Box 2–3:	
Box 2–4:	
Box 2–5:	
Box 3–1:	
Box 3–2:	

BOXES
The 2013 Comprehensive Revision to the National Income
and Product Accounts..................................................................... 60
Administration Trade Policy Initiatives....................................... 68
The Climate Action Plan................................................................ 77
Unemployment Duration and Inflation....................................... 82
Immigration Reform and Potential GDP Growth...................... 88
Other Administration Policy Responses to the
Economic Crisis.............................................................................104
The U.S. Recovery in Comparative International and
Historical Context..........................................................................116

Contents  | 19

Box 4–1:	 Two Measures of Growth in Health Care Costs: Spending
and Prices........................................................................................152
Box 4–2:	 How Will the ACA’s Coverage Expansion Affect Total
Spending Growth?.........................................................................163
Box 4–3:	 The Cost Slowdown and ACA Reforms are Reducing
Medicare Beneficiaries’ Out–of–Pocket Costs..........................172
Box 4–4:	 Premiums on the ACA Marketplaces are Lower than
Projected..........................................................................................175
Box 5–1:	 Measuring Multifactor Productivity...........................................183
Box 5–2:	 Does Inequality Affect Productivity?..........................................194
Box 5–3:	 Just–in–Time Manufacturing......................................................200
Box 5–4:	 Spectrum Investment Policies......................................................204
Box 5–5:	 Electronic Health Records............................................................210
Box 5–6:	 The Leahy–Smith America Invents Act.....................................216
Box 5–7:	 Pay–For–Delay Settlements in Pharmaceutical
Patent Cases....................................................................................218
Box 6–1:	 Flaws in the Official Poverty Measure........................................224
Box 6–2:	 A Consumption Poverty Measure...............................................226
Box 6–3:	 Women and Poverty.....................................................................231
Box 6–4:	 Social Programs Serve All Americans.........................................237
Box 6–5:	 Raising the Minimum Wage........................................................262
Box 7–1:	 Impact Evaluations, Process Evaluations, and Performance
Measurement..................................................................................270
Box 7–2:	 Using Behavorial Economics to Inform Potential Program
Improvements................................................................................277
Box 7–3:	 “Rapid Cycle” Evaluations in Center for Medicare
and Medicaid Innovation.............................................................279

20  |  Economic Report of the President

C H A P T E R

1

PROMOTING OPPORTUNITY AND
SHARED, SUSTAINABLE GROWTH

A

s the 2014 Economic Report of the President goes to press, the U.S.
economy stands five years removed from one of the most tumultuous
and challenging periods in its history. The state of acute crisis that emerged
in the months just before President Obama took office was, in some respects,
worse than the initial shock that touched off the Great Depression. The
plunge in stock prices in late 2008 proved similar to what occurred in late
1929, but was compounded by sharper home price declines, ultimately
leading to a drop in overall household wealth that was substantially greater
than the loss in wealth at the outset of the Great Depression (Romer 2009;
see also Greenspan 2013, Almunia et al. 2010). As the recession unfolded,
the economy’s total output contracted more sharply than at any other time
since World War II, and the fallout ultimately cost the country a staggering
8.8 million private-sector jobs.
As of early 2014, however, the economic landscape looks vastly different: total output has grown for 11 consecutive quarters, businesses have
added 8.5 million jobs since February 2010, and a range of analysts are
optimistic that the economy will further strengthen in the years ahead. Yet
despite this progress, many American families are still struggling to join
the middle class or to stay there, as they face the lingering after-effects of
the crisis on top of a long-standing trend of widening inequality that has
caused the rungs on the economic ladder to grow further apart. This fundamental issue—making sure the economy provides opportunity for every
American—is the President’s central economic focus.
To move toward the President’s goal of a broader, more solid foundation for future growth, three key imperatives must be addressed. The first
and most immediate imperative is to continue to restore the economy to
its full potential. Although the recovery from the Great Recession is well
underway, it remains incomplete. The second imperative is to expand the
economy’s potential. In the decades after World War II, rapid productivity
21

Figure 1-1
Monthly Change in Private Nonfarm Payrolls, 2007–2014

Thousands, seasonally adjusted
400
300
200
100
0

Jan-2014

-100
-200

12-month moving
average

-300
-400
-500
-600
-700
-800
-900

2007

2008

2009

2010

2011

Note: Shading denotes recession.
Source: Bureau of Labor Statistics, Current Employment Statistics.

2012

2013

2014

growth propelled the American economy forward and grew the middle
class. The historic surge of women into the labor force further transformed
America’s society and economic capacity. But in the last several decades,
productivity growth has declined relative to the early postwar years, and
looking ahead, America’s workforce is now expected to grow more slowly
as members of the baby-boom generation move into retirement. As a result,
efforts to enhance overall productivity and the skills of American workers,
and to expand the labor force, will be as critical as ever. The third imperative
is to ensure that the economy provides all Americans with greater opportunity to realize their full individual potential and to experience the prosperity
they work to create. A typical family’s inflation-adjusted income barely
budged in the years leading up to the crisis, which placed undue strain on
millions of hardworking households, contributed to instability in the overall
economy, and raised questions about how the American ideals of opportunity and mobility would manifest themselves in the 21st century.
These challenges are substantial, but so is America’s potential. As
discussed in the remainder of this chapter and throughout this report, the
President has set out an ambitious agenda to make progress on all three
imperatives, both by working with Congress and by taking action on his
own where possible. To return the economy to its full potential more
quickly, the President has called for a range of measures including an
22  |  Chapter 1

Opportunity, Growth, and Security initiative, along with steps to pair business tax reform with a major effort to upgrade our Nation’s infrastructure.
To expand the economy’s potential, the President continues to urge the
House of Representatives to follow the Senate’s lead and pass commonsense
immigration reform, which would help attract a new wave of inventors and
entrepreneurs to American soil. The President also wants to build on the
great strides that have already been made in the technology and energy sectors, which are laying the groundwork for a more productive economy in the
years ahead. And to make sure that the economy provides opportunity for
every American, the President’s agenda includes improved job training programs and an increase in the minimum wage, because no one working full
time should have to raise their family in poverty. In addition, the President
has set a goal of preschool for all, since one of the best ways to ensure that
all Americans have a chance to succeed is to invest in their early childhood
development. Implementation of the Affordable Care Act is another critical
step in this direction, because it is helping to provide financial security for
more American families and to slow the growth in health care costs that cut
into workers’ take-home pay.
The President is calling for significant legislative measures in these
areas, but will continue to pursue progress through executive authority
by, for example, streamlining the infrastructure-permitting process, creating new products to improve retirement security, and raising wages for
Federal contractors. The President is also using his influence to work with
businesses, universities, and nonprofits to, for instance, help low-income
students succeed in college and ensure that the long-term unemployed have
a fair shot at a new job. These are just a few of the President’s key efforts in
what he has described as “a year of action,” and the President remains ready
to work with anyone offering constructive ideas to move forward.

The Economy Five Years After The Crisis
The recovery from the Great Recession took another major step
forward in 2013, as detailed in Chapter 2. Businesses added 2.4 million jobs
over the course of the year, or 197,000 a month, marking the third consecutive year that private employment increased by more than 2 million. From
its trough in February 2010 through January 2014, private employment has
risen in 47 consecutive months for a total of 8.5 million jobs added (Figure
1-1). The unemployment rate remains unacceptably high, primarily due to
the continued elevation of long-term unemployment, but it has continued
to recover, reaching a five-year low of 6.6 percent in January. The housing
and automotive sectors—two areas that were especially hard-hit by the

Promoting Opportunity and Shared, Sustainable Growth  |  23

Figure 1-2
U.S. Merchandise and Overall Trade Deficits, 2000–2013

Real merchandise trade deficit, bil. chained 09$

900
800

Overall trade deficit, percent of GDP

8

All other goods (left axis)

Petroleum products (left axis)

7

Overall trade deficit
(right axis)

700

6

600

5

500

4

400

3

300

2

200

1

100
0

2000

2002

2004

2006

2008

2010

2012

0

Note: "All other goods" includes a residual due to chained-dollar price adjustment.
Source: Census Bureau, U.S. International Trade in Goods and Services; Bureau of Economic Analysis,
National Income and Product Accounts; CEA calculations.

Figure 1-3
Major Deficit Reduction Episodes Over a Four-Year Period
Since the Demobilization from WWII

Reduction in the Federal budget deficit, percentage points of GDP
7
6

5.7

5

FY 2013 alone: 2.7

4

3.7

3

3.2

2.7

2
1
0

2009-2013

1996-2000
1992-1996
Fiscal years

1983-1987

Source: Office of Management and Budget; Bureau of Economic Analysis, National Income and Product
Accounts; CEA calculations.

24  |  Chapter 1

crisis—continue to rebound and contribute to growth, as real (inflationadjusted) residential investment rose more than 6 percent over the four
quarters of the year, while motor vehicle production rose 9 percent. And in
October 2013, the amount of crude oil produced domestically exceeded the
amount imported for the first time since 1995, providing further evidence
of a domestic energy boom that is supporting jobs and helping to sustain a
markedly narrower trade deficit relative to the years leading up to the crisis
(Figure 1-2).
The progress seen in 2013 is notable in part because the steep decline
in the Federal budget deficit during the year was a major headwind on macroeconomic performance. Figure 1-3 shows the largest four-year reductions
in the Federal budget deficit since the demobilization from World War II.
Since the end of fiscal year 2009, the deficit has fallen by 5.7 percentage
points of GDP, with nearly half of that reduction—2.7 percentage points
of GDP—coming in FY 2013 alone. Some of the deficit reduction in 2013
was a natural consequence of the gradual improvement in the economy,
and a large portion of it was due to policy decisions, like the spending caps
agreed to in the Budget Control Act of 2011, the increase in tax rates for
top earners at the beginning of 2013, as well as the end of the temporary
payroll tax holiday. The fiscal contraction would, of course, have been much
worse were it not for the permanent extension of tax cuts for middle-income
households. While these factors reflected a balanced approach to making
progress toward fiscal sustainability, they were unnecessarily compounded
with the onset of budget sequestration in March, which the Congressional
Budget Office (CBO 2013a) estimated slowed real GDP growth over the four
quarters of the year by 0.6 percentage point and reduced employment by the
equivalent of roughly 750,000 full-time jobs.
On top of the fiscal headwinds, the economy was also forced to contend with a disruptive government shutdown for 16 days in October, as well
as dangerous brinksmanship over raising the Federal debt limit. The shutdown cost the government more than $2 billion in lost productivity alone as
Federal workers were furloughed for a combined 6.6 million days. In addition, families were unable to travel to national parks, oil and gas drilling permits were delayed, Small Business Administration loans were put on hold,
and licenses to export high-tech products could not be granted, to name just
a few effects. Consumer sentiment, as measured by the Reuters/University
of Michigan Index, fell to its lowest point of the year in October, and fourthquarter GDP growth was constrained by a large negative contribution from
the Federal sector. Moreover, because Congress was slow to act to raise the
debt ceiling, several large investment management firms announced that
they had divested from Treasury securities maturing around the time of the
Promoting Opportunity and Shared, Sustainable Growth  |  25

potential debt-ceiling breach. Ultimately, the episode provided yet another
reminder of the need for policymakers to avoid self-inflicted wounds to the
economy and to stay focused on constructive steps that support growth and
job creation.

How We Got Here: The Administration’s
Response to the Crisis
In considering the recovery over the last five years and the sources of
optimism for the years ahead, it is important not to lose sight of the critical
policy decisions that averted a second Great Depression and made it possible
for the economy to arrive at this point. Recoveries from financial crises tend
to be slower than from recessions caused by other types of shocks because
heavy household debt burdens and tight credit conditions can linger for an
extended period of time. However, the U.S. economy has fared better than
most other developed countries in recent years. As shown in Figure 1-4,
among the 12 countries that experienced a systemic financial crisis in 2007
and 2008, the United States is one of just two in which output per workingage population has returned to pre-crisis levels. The fact that the United
States has been one of the best performing economies in the wake of the
crisis supports the view that the full set of policy interventions in the United
States made a major difference in averting a substantially worse outcome.
Chapter 3 looks back at the American Recovery and Reinvestment Act
of 2009 (the Recovery Act), along with more than a dozen subsequent jobs
measures that the President signed into law, including the payroll tax cut,
extensions of unemployment insurance, and tax cuts for business investment
and hiring. At the macroeconomic level, CEA estimates that the Recovery
Act, by itself, cumulatively saved or created about 6 million job-years, where
a job-year is defined as one full-time job for one year (equivalent to an
average of 1.6 million jobs a year for four years). Adding in the subsequent
jobs measures, the cumulative gain in employment through the end of 2012
grows to 9 million job-years (Figure 1-5). This analysis is broadly similar to
others provided by a variety of sources, including the Congressional Budget
Office, and is generally consistent with a growing academic literature that is
discussed in Chapter 3. In addition to the macroeconomic impact, Chapter
3 also reviews the Recovery Act’s key investments in areas like education,
clean energy, physical infrastructure, and technological infrastructure that
will continue to pay dividends long after the Recovery Act has phased out.
Fiscal measures were only part of the overall policy response. The
President acted decisively to save the American auto industry and the suppliers and economic ecosystem that depend on it. The Administration also
26  |  Chapter 1

Figure 1-4
Real GDP Per Working-Age Population
in 2007–2008 Banking Crisis Countries, 2007–2013

Index, pre-crisis peak =100

105

Germany

100

Netherlands

Spain

95
90

France
Portugal
Iceland

U.K.

85

U.S.

Italy

Ukraine

80

Greece

75
70

Ireland

0

4

8

12

16

Quarters from peak

20

24

28

Note: U.S. as of 2013:Q4, all others as of 2013:Q3 except Iceland (Q2). Working-age population is 16-64
for U.S. and 15-64 for all others. Population for Ukraine is interpolated from annual estimates. Selection
of countries based on Reinhart and Rogoff (forthcoming).
Source: Statistical Office of the European Communities; national sources; CEA calculations.

Millions
3.5

Figure 1-5
Quarterly Effect of the Recovery Act and Subsequent
Fiscal Measures on Employment, 2009–2012
Other Fiscal
Measures

3.0
2.5

2012:Q4

2.0

ARRA

1.5
1.0
0.5
0.0

2009:Q2

2010:Q2

2011:Q2

2012:Q2

Source: Bureau of Economic Analysis, National Income and Product Accounts; Congressional Budget
Office; CEA calculations.

Promoting Opportunity and Shared, Sustainable Growth  |  27

instituted a range of measures that helped families stay in their homes,
facilitated refinancing into lower interest rates, and took steps toward alleviating the housing blight that can threaten neighborhoods and restrain home
values. Furthermore, the Treasury Department promoted financial stability
and instituted a range of programs to help restore the flow of credit for
both large and small businesses. The Federal Reserve undertook important
independent actions as well, which are more fully described in Bernanke
(2012, 2014).
Another positive but less widely appreciated story from the recession
and recovery is the performance of America’s social safety net. Though
employment and incomes declined sharply as the recession unfolded, millions of Americans were kept out of poverty by tax credits and programs
like Social Security, nutrition assistance, and unemployment insurance.
Excluding these measures, the poverty rate would have risen 4.5 percentage
points from 2007 to 2010, but in fact it only rose half a percentage point
(Figure 1-6). The direct effect of the Recovery Act on incomes, not even
counting its impact on jobs and the broader economy, reduced the poverty
rate in 2010 by 1.7 percentage points, equal to 5.3 million people kept out of
poverty. As discussed in Chapter 6, these developments represent the continuation of a longer-running trend in which essentially all of the progress
made in reducing poverty has come as a direct consequence of government
programs. The poverty rate excluding tax credits and public programs actually rose from 1967 to 2012, but when tax credits and programs are included,
the poverty rate was cut by 38 percent (Wimer et al. 2013). Nevertheless,
with 49.7 million Americans still living in poverty as of 2012, far more work
remains to be done. As the Nation recently marked the 50th anniversary of
President Lyndon B. Johnson’s declaration of a war on poverty, Chapter 6
details lessons that have been learned, ways that antipoverty programs can
be strengthened, and other policies that can help take the next step forward
by achieving meaningful reductions in poverty even before tax credits and
government programs kick in.

Sources of Opportunity in 2014 and Beyond
The U.S. economy has made substantial gains over the last five years,
and while many challenges remain, including recent weather-related disruptions and some turbulence in emerging markets, there are a number of
reasons to be optimistic about the economy’s prospects. Cyclical developments like diminished fiscal headwinds and an improvement in household
finances are likely to contribute to a strengthening of the recovery in the
near-term. At the same time, emerging structural trends like the decline in

28  |  Chapter 1

Figure 1-6
Change in Poverty Rate From 2007–2010,
With and Without Tax Credits and Benefits

Change in poverty rate, percentage points (p.p.)
5.0
4.5
4.0

+4.5 p.p.

3.5
+3.0 p.p.

3.0
2.5
2.0

+2.2 p.p.

1.5

Actual:
+0.5 p.p.

1.0
0.5
0.0

Without Tax Credits &
Benefits

Without ARRA, With
Other Tax Credits &
Benefits

With All Tax Credits &
Benefits, Including ARRA

Source: Wimer et al. (2013); CEA calculations.

the rate of health care cost growth, the surge in domestic energy production,
and continued technological progress will support growth on a sustained
basis into the future. As these developments unfold, it will be critical to take
additional steps to ensure that the middle class and those striving to join it
have opportunities to succeed. But these emerging trends will help create
the framework for the sustainable, broad-based growth that the President is
seeking to promote.

Cyclical Factors
Diminished fiscal headwinds. The most predictable reason for optimism about the U.S. economy in 2014 is the waning drag from fiscal policy
and reduced fiscal uncertainty. In December, a bipartisan budget agreement
averted a second round of discretionary sequester cuts that were scheduled
to go into effect in January and also relieved a portion of the cuts that had
already taken place during the preceding year. While Congress could do
substantially more to support job growth and economic opportunity, the
economy is unlikely to face anything like the fiscal consolidation seen at
the Federal level in 2013, with deficit reduction continuing at a much more
gradual pace going forward. As part of the budget deal, Congress also agreed
on discretionary funding levels for the remainder of FY 2014 and all of FY
2015, offering a way to avoid another counterproductive shutdown. Earlier
Promoting Opportunity and Shared, Sustainable Growth  |  29

this year, Congress passed appropriations bills for FY 2014 consistent with
these spending levels and also extended the debt limit into 2015.
As fiscal headwinds ease at the Federal level, State and local governments are also showing encouraging signs. After shedding more than
700,000 jobs from 2009 to 2012, State and local governments added 32,000
jobs in 2013.
Improved household finances. American households saw trillions
of dollars in wealth wiped out as a result of the recession, but recent data
indicate that a large degree of progress has been made in the recovery. As
of the third quarter of 2013, real per-capita household wealth had recouped
over 80 percent of the large decline from its peak, reflecting gains in housing and stock prices, as well as the progress households have made in
deleveraging. Moreover, the household debt service ratio—the estimated
required payments on mortgage and consumer debt as a share of disposable
income—was 9.9 percent in the third quarter of 2013, the lowest since the
data began in 1980, and down from 13 percent in 2007. Further improvements in household finances and expanded access to credit will contribute
to strengthening in consumer spending. Looking over the course of the
recovery, real personal consumption expenditures have grown just 2.2
percent at an annual rate, compared with the 2.9 percent pace during the
2000s expansion period, a fact that partly reflects the lingering after-effects
of the financial crisis. A noticeable pickup in consumer spending—which
comprises more than two-thirds of the U.S. economy—would represent an
important step toward turning the page on the crisis era.
While the aggregate statistics on household wealth paint a picture of
improvement, too many families have not shared in the gains. For instance,
middle-income households have on average a larger fraction of wealth in
their homes relative to equities, and house prices—despite recent improvements—have not recovered as sharply as equities, which represent a larger
fraction of the wealth held by upper-income households. The challenges of
ensuring that more Americans share in the gains from economic growth are
discussed in additional detail later in the chapter.

Structural Trends
Along with these positive near-term developments, this report also
highlights three longer-term, structural trends that have emerged recently
and will support growth on a sustained basis into the future.
Domestic energy boom and changes in energy use. The first major
trend is the dramatic increase in domestic energy production combined with
a shift in the use of energy that represents an important opportunity not
just for the economy, but also for America’s security and climate. Current
30  |  Chapter 1

Figure 1-7
Domestic Crude Oil Production and Net Imports, 2000–2013

Million barrels per day
14

Net
Imports

12
10

Dec-2013

8
6
Domestic
Production

4
2
0
2000

2002

2004

2006

2008

2010

Note: Not seasonally adjusted.
Source: Energy Information Administration, Petroleum Supply Monthly.

2012

projections indicate that the United States became the world’s largest
producer of oil and gas in 2013, exceeding both Russia and Saudi Arabia.
As noted earlier, domestic production of crude oil rose above imports in
October for the first time since 1995 (Figure 1-7), and further increases in
domestic production and reduced oil imports are expected in the coming
years. Moreover, natural gas production continued to rise in 2013 from the
2012 record high and is up more than 20 percent over the past five years.
The power sector has undertaken a shift from coal to natural gas, which was
responsible for 27 percent of our overall energy consumption in 2012, up
from 24 percent in 2008. But the progress is not limited to oil and gas—consistent with the President’s “all of the above” energy strategy, great strides
have also been made in renewables and energy efficiency. Wind and solar
power generation have each more than doubled since the President took
office, while oil consumption has fallen over this time, as stronger fuel economy standards and investments in cutting-edge technologies have led to the
most fuel-efficient light vehicle fleet ever. This broad-based energy boom
supports jobs directly in production and distribution, and also indirectly,
by making the United States more attractive as a location for multi-national
firms in energy-intensive industries like manufacturing.
The President recently announced new steps to further capitalize on
these exciting developments in the energy sector and reduce our dependence

Promoting Opportunity and Shared, Sustainable Growth  |  31

on foreign energy sources while creating new jobs. In the 2014 State of the
Union address, the President announced his intention to forge ahead with
new executive actions that will improve the fuel efficiency of the nation’s
trucking fleet and help states and localities attract investment in new factories powered by natural gas.
Although many of the recent trends in the energy sector are positive,
looking ahead over the coming decades, climate change continues to pose
considerable threats to America’s environment, economy, and national
security. The combined effects of shifting electricity production from coal to
cleaner-burning natural gas, large increases in wind and solar power generation, and ongoing progress in energy efficiency have made a large contribution to reducing national energy-related carbon dioxide emissions by more
than 10 percent since 2005. In his Climate Action Plan, the President set out
the concrete steps that the Administration is taking to address the costs of
climate change through new actions to reduce greenhouse gas emissions and
prepare for the future climate changes that are an inevitable consequence of
past emissions. The President has also recently directed his Administration
to continue working with states, utilities, and other stakeholders to set new
standards on carbon pollution from power plants.
Health care cost slowdown. The second structural trend is the slowdown in the growth of health care costs. The growth rate of real per-capita
health care expenditures from 2010 to 2012 was the lowest since the Center
for Medicare and Medicaid Services data began in the 1960s (Figure 1-8),
and preliminary data and projections indicate that slow growth continued
into 2013. As detailed in Chapter 4, this historic slowdown in health care
cost growth does not appear to be merely an after-effect of the recession.
The slowdown has persisted even as the economic recovery has unfolded,
and it is evident in areas like Medicare as well as in the gap between health
care price inflation and overall inflation, neither of which should be sensitive
to cyclical fluctuations. Chapter 4 also presents evidence that some alreadyimplemented features of the Affordable Care Act, including reductions in
overpayments to Medicare providers and private insurers as well as payment
reforms that incentivize better patient outcomes, are contributing to this
trend. Primarily as a consequence of slower health care cost growth, the
Congressional Budget Office has marked down its forecast of spending on
Medicare and Medicaid in the year 2020 by about 13 percent relative to the
projection it issued in August 2010. Employers and families are also likely
to see significant benefits as health care places less pressure on employers’
compensation costs, and the resulting savings are passed on to workers in
the form of higher wages.

32  |  Chapter 1

Figure 1-8
Growth in Real Per Capita National Health Spending, 1961–2013

Annual percent change
8
7
6
5
4
3
2

2013

1
0
1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Note: Data for 2013 is a projection.
Source: Centers for Medicare and Medicaid Services, National Health Expenditure Accounts; Bureau of
Economic Analysis, National Income and Product Accounts; CEA calculations.

Expansion of innovation. The third emerging trend that presents a
major opportunity for long-term growth is the rapid advance in telecommunications technology, particularly in fast and widely available wired and
wireless broadband networks, and in their capacity to allow mobile devices
to take advantage of cloud computing. The economic potential of these technologies and the broader context of U.S. productivity growth are discussed
in greater detail in Chapter 5. From 2009 to 2012, annual investment in U.S.
wireless networks grew more than 40 percent from $21 billion to $30 billion,
and the United States now leads the world in the availability of advanced 4G
wireless broadband Internet services. This infrastructure is at the center of
a vibrant ecosystem that includes smartphone design, mobile app development, and the deployment of these technologies in sectors like business,
health care, education, public safety, entertainment, and more. All told, the
expansion of advanced telecommunications technology—along with the
slowdown in health care cost growth and the rise in domestic energy production—are major reasons to be upbeat about the U.S. economy’s growth
prospects in the coming years.

Long-Term Fiscal Sustainability
To the extent these structural trends continue to unfold and support
stronger-than-projected economic growth in the years ahead, they will
Promoting Opportunity and Shared, Sustainable Growth  |  33

help move the Federal government closer to fiscal sustainability over the
medium- and long-run. The steep decline in the Federal deficit over the last
several years discussed earlier has been accompanied by similar improvement in the long-term fiscal outlook. One key gauge of the long-term fiscal
outlook is the fiscal gap, which represents the amount of tax increases or
spending cuts as a share of GDP required in the present to stabilize the
debt-to-GDP ratio over the next 75 years. While long-run fiscal projections
are always subject to a wide margin of error, recent estimates of the fiscal
gap are smaller than those issued just a few years ago. These improvements
are thanks in large part to the aforementioned slowdown in health care
cost growth, including the cost-saving measures in the Affordable Care Act
together with other spending restraint and the restoration of higher tax rates
on high-income households.

The Challenges That Remain and the
President’s Plans to Address Them
In the five years since the depths of the Great Recession, the U.S. economy has strengthened considerably. Nevertheless, many of the challenges
left in the wake of the recession linger, as do other challenges that built up
in the decades before the recession. Presently, the first and most immediate
imperative is to support the recovery and continue to restore the economy
to its full potential. But going forward, it will also be critical to find ways to
expand the economy’s potential and to ensure that all Americans have the
opportunity to experience the prosperity that they help create.

Continuing to Restore the Economy to its Full Potential
Despite the 8.5 million private-sector jobs added over the last 47
months and the decline in the unemployment rate to a five-year low, the
economy has not fully healed from the massive blow of the Great Recession,
and helping restore the economy to its full potential is the most immediate
challenge policymakers face. This imperative can in large part be understood
through the prism of the U.S. labor market, which is currently subject to
multiple distinct but closely related challenges. First, given the magnitude of
the job losses stemming from the Great Recession and the need to add jobs
to support a growing working-age population, the economy continues to
exhibit an absolute shortfall in the number of jobs.
Moreover, while the long-term unemployment rate has trended down,
it still remains markedly elevated. As shown in Figure 1-9, the prevalence of
persons unemployed for 26 weeks or less has returned to its pre-recession
average (4.2 percent of the labor force in January 2014, same as the average
34  |  Chapter 1

Figure 1-9
Unemployment Rate by Duration, 1994–2014

Percent of labor force
8
7
6
5

2001–2007
average

4

Unemployed for 26 Weeks
or Less

Jan-2014

3
2

2001–2007
average

1
0
1994

1996

1998

2000

Unemployed for 27 Weeks
and Over

2002

2004

2006

2008

2010

2012

2014

Note: Dotted lines represent average during the December 2001 to December 2007 expansion period as
defined by the National Bureau of Economic Research. Shading denotes recession.
Source: Bureau of Labor Statistics, Current Population Survey; CEA calculations.

from 2001 to 2007), but the long-term unemployment rate is more than
twice what it was during the pre-crisis years (2.3 percent in January 2014
compared with 1.0 percent on average from 2001 to 2007). Reducing longterm unemployment presents a major challenge because these individuals
may face stigmatization from employers or experience skill deterioration.
Even as the economy continues to add jobs, it will also be important
to see a concurrent recovery in the volume of job-to-job mobility. The flow
of workers across firms plays a critical role in the economy because job
mobility enables rising productivity and wages as individuals switch to jobs
for which they are better suited. However, in recent months there have been
fewer than 9 million combined hires and separations a month, compared
with more than 10 million a month on average from 2005 to 2007. The rate
of voluntary separations—a measure of workers’ confidence in labor market
conditions—also remains below pre-recession levels.
As the unemployment rate trends down and worker mobility picks
up, real wages should grow more quickly; but currently, the sluggish real
wage growth seen in recent years represents another serious outstanding
challenge in the labor market. Real wage growth remained positive for most
of 2013, a key sign of the progress being made (Figure 1-10). However, as
discussed in greater depth below, substantially faster real wage increases will

Promoting Opportunity and Shared, Sustainable Growth  |  35

Figure 1-10
Growth in Real Average Hourly Earnings for
Production and Nonsupervisory Workers, 2007–2014

12-month percent change
7
6
5
4
3
2

Jan-2014

1
0
-1
-2
-3
-4

2007

2008

2009

2010

2011

Note: Shading denotes recession.
Source: Bureau of Labor Statistics, Real Earnings.

2012 dollars
70,000

2012

2013

2014

Figure 1-11
Real Median Family Income, 1980–2012

68,000
66,000
64,000
62,000

2012

60,000
58,000
56,000
54,000
52,000
50,000
1980

1985

1990

Note: Shading denotes recession.
Source: Census Bureau, Historical Income Data.

36  |  Chapter 1

1995

2000

2005

2010

be needed to make up for a decades-long trend of average wages failing to
keep pace with productivity gains.
The typical family’s inflation-adjusted income has also been slow to
recover. The Census Bureau reported in September that the median family
earned $62,241 during 2012, little changed in real terms from the preceding
year (Figure 1-11). After increasing cumulatively less than half a percent in
the seven years leading up to the recession, real median family income fell
markedly during the recession and its aftermath and, as of 2012, was still 8
percent below its previous peak. Going forward, this is an important indicator of the way in which the recovery makes progress for the middle class.
To speed the recovery, boost job creation, and tighten labor markets
so as to put upward pressure on median wages, the President has repeatedly
called for investment in America’s infrastructure. This type of investment
would not only help address the 11 percent average unemployment rate in
the construction sector during 2013, but would also foster stronger longrun growth. In the 2014 State of the Union, the President renewed his call
for using the one-time revenue associated with the transition to a reformed
business tax system to finance a major modernization of U.S. transportation
infrastructure. But in the absence of the necessary Congressional action on
this proposal, the President will press ahead in other ways by speeding up
the permitting process for new construction projects.
In addition, the President’s budget includes an Opportunity, Growth,
and Security initiative, which will finance additional discretionary investments in areas such as education, research, infrastructure, and national
security. The $56 billion initiative is evenly split between defense and
non-defense and is fully paid for with mandatory spending reforms and tax
loophole closers—and would both speed the return of the economy to its full
potential and also expand that potential.
In addition to the challenges in the areas of jobs and income, the housing sector represents another key area with scope for further improvement.
As shown in Figure 1-12, construction activity fell so far in the wake of the
recession that, despite notable gains in recent years, the rate of permitting
for new residential sites still remains well below the level suggested by demographic trends and home depreciation. To help unlock this potential, steps
must be taken to bring certainty to the mortgage finance system, and also
to support communities that were particularly hard hit when the housing
bubble burst and are still coping with a legacy of foreclosures and blight.

Expanding the Economy’s Potential
In addition to taking steps that speed the economy’s return to
full potential, the Administration continues to push simultaneously for
Promoting Opportunity and Shared, Sustainable Growth  |  37

Figure 1-12
Building Permits for New Residential Units, 1960–2014

Thousands, seasonally adjusted annual rate
3,000
2,500
2,000
1,500

Estimate of Current
Annual Average
Demand

1,000
500
0
1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

Jan-2014

2010

Note: CEA estimates that approximately 1.6 million new units per year are needed to keep pace with
household formation and home depreciation. Shading denotes recession.
Source: Census Bureau, New Residential Construction; CEA calculations.

measures that will expand that potential. To understand the importance of
this goal, consider that an American worker in 2012 could produce more
than four times as much per hour as his or her counterpart in 1948. About 10
percent of the increase is due to improvements in the composition of labor,
mostly because of greater education, and 38 percent is due to increases in
the amounts of capital that workers have at their disposal. Fully 52 percent
is due to increases in total factor productivity, or what the Bureau of Labor
Statistics calls multifactor productivity, which reflects technological change
as well as the scale of markets and organization of production processes.
The growth of total factor productivity can vary widely year-to-year,
but the longer-term trends can be broadly illustrated by splitting the last
60 years into three periods, as shown in Figure 1-13. First, from the 1950s
through the early 1970s, total factor productivity grew at a relatively rapid
1.8 percent annual rate, fueled in part by public investments like the interstate highway system and the commercialization of innovations from World
War II like the jet engine and synthetic rubber. Then, from the mid 1970s to
the mid 1990s, the rate of total factor productivity growth slowed substantially, to just 0.4 percent a year. The causes of this slowdown have been the
subject of extensive academic debate, with some evidence pointing to the
disruptive effect of higher and volatile oil prices. Finally, from the mid 1990s
through the latest available data for 2012, total factor productivity growth

38  |  Chapter 1

Figure 1-13
Growth in Total Factor Productivity, 1953–2012

Annual percent change
2.50
2.25

15-year centered
moving average

2.00
1.75
1.50
1.25

1953–1973:
1.8 percent per year

1.00

1996–2012:
1.1 percent per year

0.75
0.50
0.25

1974–1995:
0.4 percent per year

0.00
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Note: The dotted lines divide the last 60 years into three periods that broadly reflect three "episodes" in
productivity growth. For the nonfarm business sector.
Source: Bureau of Labor Statistics, Multifactor Productivity; CEA calculations.

picked up to a 1.1 percent-a-year rate, in part reflecting vast improvements
in computer technology and software during this time. Although differences
across these episodes may seem small, over time they compound to enormous differences in output and living standards.
While the growth rate of total factor productivity is always critical to
an economy’s long-run potential, the projected slowdown in the growth of
America’s workforce due to the aging of the population places even greater
emphasis on productivity going forward. Since an economy’s potential
output depends fundamentally on the number of workers and the average
output per worker, slower productivity growth can in theory be offset by
rapid population growth. And during the aforementioned 1974 to 1995
period marked by slower total factor productivity growth, the working-age
(16 to 64) population continued to expand at a solid rate of more than 1
percent a year. However, the Census Bureau projects that over the 20 years
from 2012 to 2032, the working-age population will grow only 0.3 percent a
year, largely a consequence of the aging of the baby boomers into retirement.
Thus, looking ahead, productivity-enhancing investments are likely to be as
critical as ever.
The President has several key proposals that will expand the U.S.
economy’s long-run potential. Comprehensive immigration reform would
counteract the slower growth of America’s workforce by attracting highly
Promoting Opportunity and Shared, Sustainable Growth  |  39

skilled inventors and entrepreneurs to create jobs in the United States.
The Congressional Budget Office (2013b) has affirmed that the bipartisan
bill passed by the Senate would raise the economy’s productivity and total
output.
In addition, while the Opportunity, Growth and Security initiative,
infrastructure investments, and business tax reform proposals discussed earlier will speed the recovery in the near term, these steps will also help make
the economy more productive over the long run. Specifically, the President’s
framework for business tax reform would expand the economy’s potential
by reducing the distortions in the current system that skew investment decisions. By developing a system that is more neutral, corporate decision makers can act for business reasons, not tax reasons, creating an environment in
which capital will flow to the most efficient purposes.
Expanding the economy’s potential also means ensuring that the
Federal budget is fiscally sustainable over the medium- and long-run. Fiscal
sustainability frees up resources for productive investment and reduces
borrowing from abroad that would ultimately require commensurate reductions in future national income. For these reasons, the President’s budget
proposals have repeatedly included additional steps toward long-run fiscal
sustainability, including proposals to promote additional efficiencies in
the health care system and limit tax benefits and loopholes for the highest
earners. The Administration has also placed substantial emphasis on using
evidence and evaluation in the budgeting process, as discussed in Chapter 7.
These practices will help make the Federal Government more efficient and
save taxpayer money for years to come.
Executive actions—including establishing new, innovative manufacturing institutes and expanding the SelectUSA initiative to attract foreign
investment—will also help improve productivity and create well-paying jobs
in the United States. Further, the Administration continues to negotiate new
trade and investment partnerships with Europe and Asia, to help American
consumers, the American businesses that sell their products overseas, and
the workers they employ. The President is asking Congress for the Trade
Promotion Authority he will need to make these agreements a reality.

Promoting Economic Opportunity
The third major challenge the economy faces is the need to ensure
that every American has an opportunity to realize their full potential and
partake in the prosperity they help create. Since the late 1970s, the United
States has seen a major increase in income inequality due to a combination
of technological change, globalization, changes in social norms, and institutional shifts like the erosion of the inflation-adjusted minimum wage and
40  |  Chapter 1

Percent
24

Figure 1-14
Share of National Income Earned by Top 1 Percent, 1915–2012

22

2012

20
18
16
14
12
10
8
6
4
2
0
1915

1925

1935

1945

1955

1965

Note: Excludes capital gains.
Source: September 2013 update to Piketty and Saez (2003).

1975

1985

1995

2005

Figure 1-15
Growth in Productivity and Average Wage, 1947–2013

Index, 1947=100 (log scale)
500

2013

400
300

Real Output
Per Hour

200

Real Average Wage
(output deflator)

Real Average Wage
(CPI deflator)

100
75
1945

1955

1965

1975

1985

1995

2005

Note: Real output per hour is for all workers in the nonfarm business sector. Average wage is for private production and
nonsupervisory workers. Output deflator is the price index for nonfarm business output. CPI deflator is the CPI-W. Data
on wages before 1964 reflect SIC-based industry classifications.
Source: Bureau of Labor Statistics, Productivity and Costs, Current Employment Statistics; CEA calculations.

Promoting Opportunity and Shared, Sustainable Growth  |  41

the decline in union membership. As shown earlier, in 2012, the median U.S.
family’s inflation-adjusted income was less than it was in 1997. In contrast,
separate data derived from administrative tax records show that the top 1
percent of tax units—who earned an average of more than $1 million in
2012—received 19.3 percent of total income (excluding income from capital
gains, which can be highly volatile year-to-year), the largest share since 1928
(Figure 1-14).
These statistics are symptomatic of a troubling disconnect between
the economy’s productivity and ordinary workers’ wages that has emerged
over the last 40 years. As shown in Figure 1-15, real average hourly earnings for production and nonsupervisory employees roughly kept pace with
productivity growth in the nonfarm business sector during the early postwar
years. But starting around the 1970s, a large gap emerged between overall
productivity and an ordinary worker’s take-home pay. Several factors may
be contributing to this gap, including the relatively rapid increase in nonwage compensation like employer-sponsored health insurance, as well as the
likelihood that productivity gains may translate into higher pay differently
across occupations. Moreover, the gap is substantial regardless of whether
one adjusts for inflation using the Consumer Price Index, which is more
indicative of the typical household’s purchasing power, or using the price
index for total nonfarm business output, which allows for a more direct
comparison with the productivity data. Ultimately, the Figure illustrates the
growing concern that too many ordinary workers are being left behind, and
helps explain why the President has said that the basic bargain in America—
that those who work hard will have the chance to get ahead—has frayed.
To address this issue and to restore a greater measure of fairness
and opportunity to the economy, the President has proposed a number of
important measures. In the near term, the most direct step is to raise the
minimum wage, which, after adjusting for inflation, has declined by more
than one-third from its peak in the 1960s and is now worth less than it was
when President Ronald Reagan first took office in 1981. Along with a minimum wage increase, the President has also called for other measures that
would help those striving to join the middle class, including an expansion
of the Earned Income Tax Credit for workers without children. In addition
to these immediate steps, the President has set out a range of ideas to invest
in education and equip workers with the skills they will need to compete
in the global economy for years to come. For instance, the ConnectEd program continues to move forward, putting high-speed Internet in classrooms
across America, and the Administration has also secured over 150 new commitments from universities, businesses, and nonprofits to improve college
opportunity and outcomes for students from low-income families.
42  |  Chapter 1

Along with steps that create jobs, boost incomes, and invest in educational opportunities, the President is pursuing measures to ensure that
families can experience a greater degree of financial security. Three million
young adults under age 26 have gained coverage through their parents’ plans
because of the Affordable Care Act, another 4 million people enrolled in
insurance plans through State and Federal marketplaces as of late February,
and millions more have been determined eligible for State Medicaid programs. Additionally, in the 2014 State of the Union, the President announced
the creation of MyRA, a new safe and easy-to-use savings vehicle that can
help millions Americans start saving for retirement.

Conclusion
The challenges discussed above are substantial, but the President
believes that it is well within our capacity as a Nation to address these
issues and to move toward shared and sustainable growth. The President
has set out an ambitious agenda to support the recovery in the near term,
while building on emerging strengths to expand the economy’s potential
over the long term. In this context, the President also remains focused on
restoring greater measures of fairness and opportunity to our economy, to
strengthen the middle class and give a boost to those striving to join it. This
agenda includes measures to create new well-paying jobs, continue to reduce
dependence on foreign energy sources, equip workers with skills to compete
in the global economy, support those hardest hit by economic change, and
provide families with a greater sense of financial security. These steps, and
the rationale underlying them, are the focus of the pages that follow.

Promoting Opportunity and Shared, Sustainable Growth  |  43

C H A P T E R

2

THE YEAR IN REVIEW AND
THE YEARS AHEAD

T

he economy continued to recover and strengthen in 2013, nearly five
years after the worst of the financial crisis. Building on the progress
of the previous two years, businesses added 2.4 million jobs over the 12
months of 2013: in total, the private sector added 8.5 million jobs during 47
months of consecutive job growth. The unemployment rate fell 1.2 percentage points in 2013, a larger decline than in previous years and more than
was forecast by most private-sector economists. Output growth started the
year slowly, largely because of headwinds from fiscal drag and slow growth
among many of our trading partners that reduced demand for U.S. exports.
Output strengthened, however, in the second half of the year. Overall, real
gross domestic product (GDP) grew 2.5 percent during the four quarters of
the year, up from the 2.0 percent growth during each of the preceding two
years. Growth in consumer spending, homebuilding, and exports supported
aggregate demand growth. Inventory investment was also a positive factor,
partially due to an increase in agricultural production reflecting a plentiful
crop in 2013 following a year of drought in 2012. Federal fiscal policy was a
drag on the economy because of the tightening due to the expiration of the
temporary payroll tax cut and sequester-related spending cuts beginning
in March, and because of the uncertainty caused by a partial government
shutdown in October and the brinksmanship over the debt limit. Inflation
remained low and roughly stable, with the consumer price index (CPI) up
1.5 percent during the 12 months of 2013, and the CPI excluding food and
energy up 1.7 percent over this period, slightly below the year-earlier pace.
Looking ahead, a wide variety of indicators suggest that the economy
is well situated for a pickup in growth in 2014. Following the largest fouryear reduction in the Federal deficit as a share of GDP since the post-WWII
demobilization, Federal fiscal policy will be much less of a drag in 2014 and
thus will likely constrain overall growth by less than during the preceding
years. State and local government spending appears to have turned the
45

corner, with purchases increasing during the second and third quarters of
2013. The mid-February action of Congress to suspend the debt limit until
March 2015 relaxes a situation that had been headed towards unwelcome
uncertainty.
Although the economy still remains challenging for many, households—on average—are in an improved position to increase spending as
they have further reduced their debt burden and seen a substantial increase
in housing and stock-market wealth. More household wealth will facilitate
an increase in spending on consumer durables such as motor vehicles, which
are showing their age and due for replacement. Homebuilding, which grew
rapidly last year, is likely to continue growing on a path up to levels consistent with the demographic forces of the next decade, with mortgage interest
rates still below their pre-recession levels, despite a mid-2013 rise. Business
fixed investment also has potential to accelerate after relatively slow growth
in 2013 as aggregate demand picks up and businesses can take advantage of
their sizeable cash flows.
Nevertheless, several downside risks to economic growth remain in
2014 as unforeseen events both domestically and internationally may pose a
risk to the economy. Recently, for example, severe cold weather and storms
in the United States and a global reduction in asset prices have contributed
to some economic activity falling below trend rates of growth in the last few
months.
The pace of the recovery will depend, in part, on policy choices.
Additional measures that increase aggregate demand would add impetus to
the economy in 2014. In particular, the Budget also includes the Opportunity,
Growth, and Security initiative, which will finance additional discretionary investments in areas such as education, research, infrastructure, and
national security. The $56 billion initiative is evenly split between defense
and non-defense and is fully paid for with mandatory spending reforms and
tax loophole closers. In addition, investments in infrastructure or extending
emergency unemployment benefits would expand demand immediately
while measures like business tax reform would help the economy by increasing certainty.

Key Events of 2013
Aggregate Output Growth During the Year
Growth in aggregate economic activity was fairly steady during 2013,
with quarterly growth rates between 1.8 and 3.0 percent at an annual rate
for the first three quarters of the year, as measured by the average of the
46  |  Chapter 2

Figure 2-1
Mean GDP Growth, 2007–2013

Percent change at an annual rate
6

2013:Q4

4

4.9

2
0
-2

1.8
0.2 0.6
-0.4

1.6

0.9
-1.0

2.2
1.0

4.6

3.7

4.0

3.3

2.8

1.8

1.8
0.4

0.3

2.5

2.8
1.8

2.9
2.4

(GDP only)

-0.2

-1.8

-4
-6

-5.8

-8
-10

-7.9

2007:Q1

2009:Q1

2011:Q1

2013:Q1

Note: Mean real GDP growth is the average of the growth rates of real GDP and real gross domestic income
(GDI). The bullets show mean GDP and the bars show the GDP and GDI growth in each quarter. The
estimate for 2013:Q4 is for GDP only. Shading denotes recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

income- and product-side of real GDP (Figure 2-1).1 During the four quarters of the year, growth was strong in exports (4.9 percent) and in residential
investment (6.6 percent), and moderate in business fixed investment (3.0
percent) and consumer spending (2.1 percent). State and local purchases
edged up slightly following four years of decline, while Federal spending fell
6.2 percent.

Fiscal Policy
Federal fiscal policy evolved through several near- or after-deadline
Congressional actions that made fiscal policy uncertain and created a difficult planning environment for businesses and consumers.
Toward the end of 2012, policy focused on the potential negative
effects of the “fiscal cliff,” a confluence of expiring tax cuts and scheduled
spending declines that were on track to occur simultaneously, which might
have resulted in a sharp fiscal-policy tightening on January 1, 2013. The
Congressional Budget Office (CBO) estimated that these policies, if allowed
to occur, would have lowered real GDP growth by about 2.25 percent during
1 Research shows that an average of the two growth rates is better correlated with a wide
variety of economic indicators than either the product-side measure (which is headlined in the
Commerce Department reports) or the income-side measure alone (Nalewaik 2010, Economic
Report of the President 1997, pp. 72-74).

The Year In Review And The Years Ahead  |  47

the four quarters of 2013, or enough to cause a decline in real GDP. On the
tax side, the 2001 tax cuts, previously extended through 2012, were expiring. Also expiring at the end of 2012 was a 2-percentage point cut in the
Social Security payroll tax that was first instituted for one year in 2011, and
an increase of the threshold for the Alternative Minimum Tax (AMT). On
the spending side, defense and nondefense spending were each scheduled
for across-the-board cuts (sequestration) of $55 billion. Medicare payments
to physicians and emergency unemployment benefits were among other
spending programs scheduled to be cut in January 2013.
On January 1, 2013, Congress passed the  American Taxpayer Relief
Act of 2012  (ATRA). The ATRA addressed the revenue side of the fiscal
cliff by making permanent the middle-class tax cuts, indexing the AMT to
inflation permanently, and raising revenues over 10 years by allowing highincome tax cuts to expire. The ATRA also allowed the temporary payroll
tax cut to lapse. On the spending side, the ATRA extended Emergency
Unemployment Compensation and delayed the Medicare physician cuts
for an additional year, but the Act delayed sequestration only until March
1, 2013.
When Congress failed to reach a budget agreement by March 1, allowing the sequester to go into effect, cuts to discretionary and non-exempt
mandatory programs were distributed over the remaining seven months
of the fiscal year (rather than the full fiscal year in the sequester’s original
design). As a result, many Federal agencies furloughed civil servants, which
reduced Federal compensation by $0.6 billion at an annual rate in the second
quarter and by $5.5 billion at an annual rate during the third quarter of 2013
(a total of $1.5 billion not at an annual rate). The CBO projected that the
sequester would cut 750,000 jobs and reduce growth during the four quarters of 2013 by a 0.6-percentage point.
The debt ceiling had technically been reached on December 31, 2012
when the  Treasury Department commenced “extraordinary measures” to
enable the continued financing of the government through mid-February.
Around the end of February, however, Congress passed and the President
signed a bill that suspended the debt limit though May 18. The next day, on
May 19, the debt ceiling was reinstated at a level that reflected borrowing
during the suspension period, but no more. As a result, the Treasury began
applying extraordinary measures once again and, in late September, the
Treasury announced that these extraordinary measures would be exhausted
by October 17.
Adding to the debt-ceiling stress, more uncertainty arose in early
October when the continuing resolution needed to fund the government
was not extended into the new fiscal year beginning on October 1. As a
48  |  Chapter 2

result, the U.S. Government went into a partial shutdown. About 850,000
Federal civilian employees were initially put on temporary leave, but many
civilian Defense Department employees were recalled during the second
week of the shutdown. An agreement for a continuing resolution to end
the shutdown and extend the debt ceiling was reached on October 16, and
the Federal government returned to normal operations the next day. The
Bureau of Economic Analysis (BEA) has estimated that the shutdown was
directly responsible for a 0.3 percentage point reduction in the annualized GDP growth rate for the fourth quarter, although this estimate does
not incorporate indirect effects that operate through reductions in private
activity dependent on government services, reductions in confidence, or
increases in uncertainty. Confidence in government policy, as measured
by the Thompson Reuters-University of Michigan Survey, fell to a level in
October which was in the bottom 5 percent of the monthly series since it
began in 1978.
The agreement to end the shutdown (the Continuing Appropriations
Act of 2014) funded the government through January 15, 2014, and suspended the debt ceiling until February 7, 2014, after which time it was
suspended again until March 2015. In mid-December, Congress passed
an agreement to provide partial relief from the automatic sequestration of
discretionary spending in FY 2014 and 2015, and offset those increases with
increased pension contributions from new Federal civilian employees, as
well as a variety of higher fees and spending reductions. The bill provided
only an overall discretionary cap and, in January, Congress passed FY 2014
appropriations bills consistent with these spending levels. Notably, the bill
would fully restore cuts to Head Start programs, which provide early childhood education to children from low-income families, partially restore cuts
to medical research and job training programs, and finance new programs
to combat sexual assault in the military.
As a result of this fiscal stringency and continued GDP growth, the
Federal deficit-to-GDP ratio fell 2.7 percentage points to 4.1 percent in FY
2013 and ranks among one of the largest year-over-year declines ever (Figure
2-2). The deficit-to-GDP ratio in FY 2009 was elevated by the steep recession
as well as the fiscal stimulus to combat that recession (See Chapter 3). Since
then, the four-year decline in the deficit-to-GDP ratio of 5.7 percentage
points was the largest since the demobilization at the end of World War II.
Overall fiscal support substantially raised the level of output and employment since 2009, as discussed in Chapter 3. But the reduction in the deficit,
especially in 2013, has acted as a drag on growth rates. One reason for the
fiscal drag was the winding down of various countercyclical fiscal policies
taken during the recession. Fiscal drag is likely to moderate substantially in
The Year In Review And The Years Ahead  |  49

Percent of GDP
12

Figure 2-2
Federal Budget Deficit, 1950–2015
FY 2015

10
8

Actual
(FY 2013)

6
4

FY 2015 Budget
Projections

2
0
-2
-4
-6
1950

1960

1970

1980
1990
Fiscal Year

2000

2010

Source: Bureau of Economic Analysis, National Income and Product Accounts; Office of
Management and Budget.

FY 2014, with a projected further 0.4 percentage point decline to 3.7 percent
in the deficit to GDP ratio under the President’s policies.

Monetary Policy
In 2013, the Federal Open Market Committee (FOMC) continued to
provide substantial policy accommodation. With its usual tool—the federal
funds rate—near its effective lower bound, the Committee employed both
forward guidance for the federal funds rate and additional purchases of
longer-term securities.
The FOMC made clear its intention to keep the target range for
the federal funds rate “exceptionally low” and maintained throughout the
year the forward guidance it issued in December 2012 indicating that the
Committee will maintain the current level of the federal funds rate at least
“as long as the unemployment rate remains above 6.5 percent, inflation
between one and two years ahead is projected to be no more than half a percentage point above the Committee’s 2 percent longer-run goal, and longerterm inflation expectations continue to be well anchored.” Moreover, in its
December 2013 statement, the FOMC added to its forward guidance, stating
that “The Committee now anticipates, based on its assessment of these factors (labor market conditions, inflation, and inflation expectations), that

50  |  Chapter 2

it likely will be appropriate to maintain the current target range [of 0 to ¼
percent] for the federal funds rate well past the time that the unemployment
rate declines below 6-1/2 percent (emphasis added), especially if projected
inflation continues to run below the Committee’s 2 percent longer-run
goal.” This additional information was intended to provide greater clarity
on the Committee’s policy intentions once the unemployment threshold is
crossed.
With regard to asset purchases during 2013, the Federal Reserve
continued expanding its holding of mortgage-backed securities at a rate of
$40 billion a month and longer-term Treasury securities at a pace of $45
billion a month in an attempt to “support a stronger economic recovery and
to help ensure that inflation, over time, is at the rate most consistent with
its dual mandate,” to achieve “maximum employment and price stability.”
In the period leading up to the June FOMC meeting, financial market participants interpreted some Federal Reserve communications as implying an
earlier-than-expected reduction in the pace of purchases. This interpretation
contributed to an increase in market volatility and a marked rise in longerterm Treasury yields over the summer that were only partly reversed in the
fall, as the Federal Reserve continued to purchase assets at an unchanged
pace. However, at its December meeting the FOMC decided to begin reducing the pace of its purchases in January, cutting the monthly increase in its
holdings by $10 billion to $75 billion. In addition, the Committee indicated
that if incoming information broadly supports its expectation of ongoing
improvement in labor market conditions and inflation moving back toward
its longer-run objective, it will likely reduce the pace of asset purchases in
further measured steps at future meetings. This tapering of asset purchases
was continued in January 2014 as announced in the FOMC meeting that
month.

Financial Markets
Financial developments over the course of the year reflected the
evolving economic outlook as well as Federal Reserve communications. In
the spring and the summer, speculation about a possible reduction in the
pace of Federal Reserve asset purchases contributed to a sizeable increase in
longer-term interest rates (Figure 2-3).
Yields on 10-year Treasury notes were 1.7 percent at the start of
May before rising to 2.6 percent in July, and yields continued to rise to
about 2.9 percent just before the September FOMC meeting. In response
to the Committee’s decision to leave the pace of purchases unchanged in
September, the 10-year yield retraced part of the summer increase, dropping to 2.6 percent for the month of October. In addition, Federal Reserve
The Year In Review And The Years Ahead  |  51

Percent
5.0
4.0

Figure 2-3
Interest Rates, 2010–2014

Lower growth and inflation
forecasts around 2011

3.0

June FOMC
Meeting
Week ended
2/21/2014

10-Year
Treasury Yields

2.0
1.0

90-Day Treasury
Yields

0.0
-1.0
2010

2011

Source: Federal Reserve Board, H.15.

2012

2013

2014

communications appeared to lead investors to push back their expectations
for the timing of the first increase in the federal funds rate during the fall.
Toward the end of the year, however, the better-than-expected readings
on payroll employment and on other economic indicators, followed by the
FOMC’s decision to reduce the pace of its asset purchases, boosted longerterm Treasury yields in the final weeks of 2013. The 10-year Treasury yield
closed 2013 at roughly 3 percent. Short-term rates (such as the rate on federal funds, and the 91-day Treasury bill rate) were more stable throughout
the year—remaining under 0.2 percent—although expectations of future
short-term rates fluctuated.
In October, brinksmanship over the debt ceiling—which was expected
to be hit soon after October 17—and the two-week government shutdown
weighed heavily on financial markets. Through September and early
October, several indicators of financial stress reflected market participants’
concerns about the debt limit. As shown in Figure 2-4, yields on specific
Treasury bills maturing around that time increased in anticipation of potential delayed payments.
Moreover, institutional money market funds saw a sizeable $86 billion
of outflows (about 5 percent of assets) in the three-week period that ended
October 16. Fidelity Investments—the nation’s largest manager of money
market mutual funds—declared publicly in early October its decision not
52  |  Chapter 2

Figure 2-4
Treasury Bills Maturing in Late October–Early November, 2013

Yield, percent

0.6

Maturity Date:

10/15

10/17/2013
10/24/2013
10/31/2013
11/7/2013

0.5
0.4
0.3
0.2
0.1

11/1

0.0
-0.1

Apr.

May.

Source: Bloomberg.

Jun.

Jul.

Aug.

Sep.

Oct.

Nov.

to hold U.S. government debt set to mature around the date of the potential
debt ceiling breach. Finally, interest rates on overnight repurchase agreements, or repos, collateralized by Treasury securities, a common source of
funding for financial institutions, spiked in early October. With the resolution of the debt ceiling debate, all such indicators returned to normal levels.
Reflecting the ongoing economic recovery and the improved outlook
over the course of the year, U.S. equity markets remained on a general
upward path despite the increase in interest rates. The Standard and Poor’s
500 rose by 30 percent in 2013, reaching a record high in nominal terms
at year-end. When adjusted for GDP price inflation, however, it remained
below its March 2000 peak. The Standard and Poor’s edged up slightly during the first two months of 2014.

International Developments
The past year also saw the beginnings of a recovery in Europe, with
real GDP edging up between 1.0 and 1.6 percent annual rate in the second,
third, and fourth quarters of 2013. These were the first three consecutive
quarters of positive real GDP growth for the 28-country European Union
since 2011. Concerns about the stability of the European monetary union
(the 17-country “euro area”) that surfaced in 2011 and 2012 have subsided.

The Year In Review And The Years Ahead  |  53

In the euro area, the unemployment rate stabilized at a record high of 12.1
percent from April to September before ticking down to 12.0 percent in
the fourth quarter. Euro area inflation was subdued, declining to only 0.8
percent during the 12 months of 2013 from 2.2 percent a year earlier. The
recent low rate of inflation has fueled concerns about possible deflation.
The European Central Bank policy target for price stability is “below, but
close to, 2 percent.” The Central Bank’s Outright Monetary Transactions
program, first announced in August 2012, has helped bring a measure of
stability to European sovereign debt markets, with Italy and Spain’s 10-year
yields ending the year right around a manageable 4 percent. During the year,
euro area states made substantial progress to centralize and harmonize bank
supervision and regulation at the euro level.
There were notable developments in several European countries as
well. In the runup to the euro area crisis, countries including Greece, Spain
and Portugal saw a large runup in their current account deficits to finance
private and public borrowing that supported consumption and investment. In the wake of the euro area crisis, these countries have adjusted,
largely eliminating their current account deficits through reductions in
unit labor costs and improved price competiveness, as shown in Figure
2-5. Nevertheless, unemployment rates remain particularly high in these
countries.
Japan’s real GDP grew a solid 2.7 percent during the four quarters of
2013 following a 0.4 percent decline during 2012. Japan’s core consumer
price index (that is, excluding food and energy) turned positive, 0.7 percent
during the 12 months of 2013, up from a 0.6 percent decline during 2012.
This follows in the wake of the election of Shinzo Abe in December 2012,
the appointment of a new governor of the Bank of Japan in March, and the
announcement in April that the Bank intended to double the monetary base
by the end of 2014. Under this policy, bond purchases amount to about $80
billion a month (basically, the same pace as the Federal Reserve but in a
smaller economy). Expansionary monetary policy was part of a three-prong
strategy that initially included fiscal stimulus and structural reforms meant
to support positive growth and to keep Japan from slipping back into a
period of deflation.
China’s real GDP grew 7.7 percent during the four quarters of 2013,
slightly below the year-earlier pace, but noticeably slower than 10 percent
and 9 percent growth rates during 2010 and 2011, respectively. Xi Jinping
assumed the presidency in March and presided over the Third Plenary
Session of the Communist Party, which resulted in a raft of economic
reform proposals. China’s interbank lending rates have spiked on several
occasions this year. During these episodes, the People’s Bank of China was
54  |  Chapter 2

Figure 2-5
Current Account Balance by Country, 2000–2013

Percent of GDP, four-quarter moving average
10
8

2013:Q2

Germany

6

Ireland

4
2

Euro Area

0

Italy

-2
-4

Spain

-6

Portugal

-8

-10
-12
-14

Greece

-16

-18
2000:Q1

2003:Q1

Source: Eurostat; national sources.

2006:Q1

2009:Q1

2012:Q1

slow to inject liquidity, which many interpreted as a warning to banks that
have increased off-balance sheet commitments to bypass administrative and
regulatory controls and expand lending.
Among other emerging market economies, the pace of real GDP
growth fell in Indonesia, Malaysia, Mexico, South Africa and Thailand.
But growth also increased in a few, such as Brazil, India, and Turkey. Low
interest rates in the United States since the recession coupled with higher
investment return prospects in emerging market economies prompted an
increase in capital flows toward emerging markets. As interest rates in the
United States began to rise and growth prospects abroad waned, however,
investors started adjusting their portfolios, which in some cases had adverse
effects on emerging-market currencies and interest rates. Foreign mutual
funds withdrew $53 billion from emerging markets between mid-May and
August, leading to sharp drops in a number of currencies and emerging
market equity indexes and causing central banks in several affected countries (India, Indonesia, Turkey, Brazil, and Pakistan) to raise domestic policy
interest rates. Nevertheless, even with the withdrawals, investment holdings
remained well above the levels of just a few years ago as shown in Figure
2-6. In some instances, currencies and bond markets have retraced their
earlier losses, especially as global investors have increasingly differentiated

The Year In Review And The Years Ahead  |  55

Figure 2-6
Cumulative Flows into Mutual and Exchange-Traded Funds Investing
in Emerging Markets, 2010–2014

Billions of dollars, cumulative since January 2010
140
120

Equity Funds

100
80

Bond Funds

60

Week ended
2/26/14

40
20
0
Jan-10

Jan-11

Source: EPFR Global, a subsidiary of Informa plc.

Jan-12

Jan-13

Jan-14

debt by country according to the underlying economic fundamentals of each
country’s economy.

Developments in 2013 and the Near-Term Outlook
Consumer Spending
Real consumer spending grew about 2 percent during each of the past
three years. With consumer spending constituting 68 percent of GDP, that
stability explains much of the stability of the growth of aggregate demand
during those three years. Yet the stability of consumption growth during
2013 results from several offsetting developments. The termination of
the temporary 2-percentage point cut in payroll taxes reduced disposable
income during 2013 by $115 billion relative to 2012. This subtracted about
0.9 percent from disposable income, and held down consumption growth
by about half a percent. Higher taxes on high-income households from the
American Taxpayer Relief Act likely had little impact on spending due to
their smaller aggregate size and the relatively low marginal propensities to
consume for high-income households. Also, by reducing the medium- and

56  |  Chapter 2

long-term budget deficit, the higher tax rates on high-income households
will contribute to stronger and more-sustainable growth over time.
Strong gains in aggregate household net worth—both an increase
in assets and a decline in the debt burden—have supported aggregate
consumer spending. Debt service (that is, required minimum payments on
household debt) has fallen from 13 percent of disposable income at the end
of 2008 to 10 percent by the third quarter of 2013 (the latest data available,
Figure 2-7). Some of the decline in debt service is due to declines in interest
rates on mortgages and consumer credit, but some of the decline is due to
declines in the ratio of household debt to income, a process called deleveraging. Debt has fallen from about 1.3 times annual income in 2008 to 1.1 times
annual income by the third quarter of 2013—with most of the decline in
this ratio due to rising nominal incomes, although nominal debt has edged
down 5 percent. Together, these declines in household debt and debt-service
relative to income show that the household sector as a whole has progressed
in reducing these burdens. Although these figures are relevant for projecting
aggregate output and consumption growth, they do not reflect the change in
debt and debt service for moderate-income and median-income households
who, in many cases, continued to face challenges in 2013.
Overall wealth also grew in 2013, as shown in Figure 2-8. Although
these wealth increases were in all categories of holdings, wealth likely
increased substantially more for high-income households (which have a
larger share of their wealth in equities) than for the typical household (which
has more of their wealth in housing that appreciated more slowly than equities in 2013). As a result, this suggests that wealth inequality continued to
grow as middle-class families faced persistent economic challenges. While
gains in stock-market wealth have been happening since the trough of the
recession in 2009, those increases were particularly sharp during 2013, when
the Wilshire 5000 stock market index increased 31 percent. During the four
quarters of 2013, stock market wealth is estimated to have increased by an
amount equivalent to 39 percent of annual disposable income. Housing
wealth (net of mortgage liability) also increased notably during the year.
Housing prices, as measured by the CoreLogic National House Price Index,
hit bottom around March 2011 and have increased 11 percent during the 12
months of 2013. As a result, net housing wealth is on track to increase by
another 13 percent of annual disposable income in 2013.
The increases in stock market and housing assets point to an increase
in the ratio of net worth to income amounting to 52 percent of annual disposable income. An increase in wealth raises annual consumer spending by
about 3 percent of that increase. As a result, the expansion of wealth alone

The Year In Review And The Years Ahead  |  57

Figure 2-7
Household Deleveraging, 1990–2013

Percent of disposable income
14
2013:Q3

Years of disposable income
1.4
1.3
1.2

13

Debt Service Share of Income
(right axis)

12

1.1
1.0
0.9

11
10

Liabilities-to-Income Ratio
(left axis)

9

0.8

8

0.7

7

0.6
1990:Q1
1995:Q1
2000:Q1
2005:Q1
2010:Q1
Note: Shading denotes recession.
Source: Federal Reserve Board, Financial Accounts of the United States.

6

Figure 2-8
Consumption and Wealth Relative to Disposable
Personal Income (DPI), 1952–2013

Consumption/DPI ratio
1.10
1.05
1.00
0.95
0.90
0.85
0.80

Years of disposable income
2013:Q4 7
6

Total-wealth-to-DPI ratio
(right axis)

Consumption-to-DPI ratio
(left axis)
Net housing
wealth-to-DPI ratio
(right axis)

5
4
Stock market
wealth-to-DPI ratio
(right axis)

3
2
1

0.75
0
1950
1958
1966
1974
1982
1990
1998
2006
2014
Note: Values imputed for 2013:Q4 by CEA. Shading denotes recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts; Federal
Reserve Board, Financial Accounts of the United States; CEA calculations.

58  |  Chapter 2

could support a rise in consumption of 1.7 percent of disposable income, or
more than enough to offset the rise in taxes in 2013.
Looking ahead, consumer spending in 2014 is likely to grow faster
than its 2-percent rate during the past three years. The rise in wealth and
the progress in deleveraging have created a more-stable platform on which
to base the growth of consumer spending. The rapid growth of consumer
durables during 2013 (5.6 percent) is likely to continue or increase further.
The average age of light motor vehicles on the road has risen to 11.4 years
and it appears likely that some pent-up demand remains for motor vehicles
and other durables whose purchases have been delayed during the recession
and the slow recovery.

Business Investment
Business Fixed Investment. Real business fixed investment grew
moderately, 3.0 percent during the four quarters of 2013, down from a 5.0
percent increase during 2012. The slower pace of business investment during 2013 was concentrated in structures and equipment investment, while
investment in intellectual property products grew faster in 2013 than the
year earlier. Investment in nonresidential structures declined 0.2 percent following robust growth of 9.2 percent during 2012. Investment in equipment
slowed to 3.8 percent, following a 4.5 percent increase in 2012. In contrast,
investment in intellectual property products picked up to 4.0 percent during
2013 from 2.9 percent in 2012. (In July 2013, as part of a comprehensive revision to the National Income and Product Accounts, the Bureau of Economic
Analysis revised its classifications for business fixed investment to include
1) Research and Development and 2) Entertainment, Literary, and Artistic
originals in a new category of Intellectual Property Products, which also
includes software investment. See Box 2-1 on the July 2013 benchmark of
the National Income and Product Accounts.)	
Within equipment investment, major components such as information processing equipment and transportation equipment posted less robust
growth in 2013 than in 2012, offsetting stronger growth in industrial equipment investment. Within investment in information processing equipment,
declines were posted in investment in computers and photocopy equipment.
Within transportation equipment, growth was not as fast as 2012 for investment in autos, aircraft, and ships.
Real investment in nonresidential structures edged down 0.2 percent
during the four quarters of 2013, down from growth of 9.3 percent in 2012.
Solid growth in petroleum and natural gas drilling was offset by declines in
the construction of manufacturing structures and power and communication facilities.
The Year In Review And The Years Ahead  |  59

Box 2-1: The 2013 Comprehensive Revision to the
National Income and Product Accounts
In July 2013, the Commerce Department released the results of
the first comprehensive revision to its National Income and Product
Accounts—the raw material underlying the calculation of gross domestic product (GDP)—since 2009. These revisions, which reach back to
1929, include additional source data as well as methodological changes
designed to reflect the evolving nature of the U.S. economy. In particular,
the Bureau of Economic Analysis has expanded its definition of business
investment to include spending on research and development (R&D)
and the creation of original works of art like movies, all of which are now
recorded as intellectual property products. The Commerce Department
also recognized the increase in pension obligations as savings for households and a liability for governments and businesses. In the Federal
Reserve’s Financial Accounts of the United States, the cumulative values
of these liabilities are now recognized as assets of the household sector
and liabilities of governments and businesses.
All told, these and other changes effectively increased the size of the
economy as measured in the first quarter of 2013 by $551 billion dollars
at an annual rate (or 3.4 percent). The changes also held implications for
the path of growth of GDP over time, with the statistical updates from
the new annual source data affecting mainly more recent years, while
Real GDP Before and After the 2013
Comprehensive Revision, 2007‒2013

Index, 2007:Q4 = 100
108

Current
(+8.5% through 2013:Q1)

106
104
102

13:Q4

13:Q1
-4.3%

100
98
96
94

Previously Reported
(+8.1% through 2013:Q1)
-4.7%

92
2007:Q1 2008:Q1 2009:Q1 2010:Q1 2011:Q1 2012:Q1 2013:Q1

Note: Shading denotes recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts.

60  |  Chapter 2

the methodological revisions (such as intellectual property) affected the
entire historical series. Real GDP growth during the 16 quarters following the end of the recession in the second quarter of 2009 was revised
up by an average of 0.1 percentage point to 2.2 percent a year, and the
decline in GDP observed during the recession (starting in the fourth
quarter of 2007) was revised up 0.3 percentage point to -2.9 percent at
an annual rate, making the recession less steep and the recovery stronger
than what was previously reported. The cumulative decline in real GDP
during the recession is now reported at 4.3 percent rather than 4.7
percent, followed by an increase during the expansion of 8.5 percent
through the first quarter of 2013, as opposed to 8.1 percent published
previously (see Figure).
Since the beginning of 2009, the average absolute revision (without
regard to sign) from the advance quarterly estimate of real GDP growth
to the latest data was 1.3 percentage points. The magnitude of these
changes highlights the difficulty in measuring economic performance
real-time.

The pace of growth of business fixed investment is puzzling because
interest rates are low and the internal funds available for investment are
high. Interest rates on corporate Baa bonds were low in both nominal and
real terms. Nominal Baa rates averaged 5 percent during 2013, and adjusted
for expected inflation of about 2 percent, this translates into a real rate of 3
percent, substantially below the 60-year average of about 4.7 percent.
Funds for investment were also easily available from internal sources
such as undistributed profits and depreciation. For the nonfinancial business sector, the sum of these sources, known as cash flow in national income
accounting, was 10.1 percent of GDP in the first three quarters of 2013,
well above the historical average of 8.7 percent. Historically, nonfinancial
corporate investment averages 103 percent of cash flow, with the sector as a
whole borrowing from banks and the public for the rest. In contrast, during
the first three quarters of 2013, investment was only 90 percent of cash flow.
The cash flow that was not available for investment appears to have been
spent on share repurchases, a way of returning funds to shareholders that is
similar to dividends, but more volatile.
With interest rates low and internal funds readily available, the
growth rate of investment might be attributable to low expectations of
output growth. In a relationship known as “the accelerator,” the growth
of investment is related to the change in growth (that is, the acceleration)
of output, as shown in Figure 2-9. For example, when output accelerated
in 2010 (that is, when output growth increased from negative in 2009 to
The Year In Review And The Years Ahead  |  61

Figure 2-9
Business Investment and the Acceleration of
Business Output, 1965–2013

Percent
25

4-quarter change in business fixed
investment (left axis)

20

Percentage points
25
2013:Q4
20

15

15

10

10

5

5

0

0

-5

-5

-10
-15
-20
1965

-10

Acceleration of business output
(right axis)
1970

1975

1980

1985

1990

1995

2000

2005

-15
2010

-20

Note: The accelerator is the 6-quarter percent change in business output (AR) less the 6-quarter ago 6quarter percent change in business output (AR). Shading denotes recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts.

positive in 2010), investment increased very fast so that the capital stock
could service the new level of demand. But when business output growth
settled down to an annual rate of roughly 3 percent during the three years
through 2013, investment did not need to grow so fast, and indeed it has
slowed, as shown in Figure 2-9.
Inventory Investment. Inventory investment made a substantial
contribution to real GDP growth during the four quarters of 2013 when
it accounted for 0.8 percentage point of the 2.5 percent total growth. An
increase in agricultural inventory investment accounts for 0.3 percentage
point of that overall 0.8 percentage-point contribution and reflects the
rebound to a strong harvest following a severe drought in 2012. In the
manufacturing and trade sector, the buildup of inventories through the year
was no faster than sales, so that by December, inventory stocks were at a 1.30
months’ supply, roughly the same level as at year-end 2012.

State and Local Governments
Although State and local governments continued to experience fiscal
pressure in 2013, the four-year contraction in the sector—measured in terms
of both purchases (consumption and investment) and employment—finally
appears to have ended. State and local purchases, which had generally

62  |  Chapter 2

Figure 2-10
Real State and Local Government Purchases During Recoveries

Indexed to 100 at NBER-defined trough
120

Average,
1960–2007

115
110

1991

105

2001

100
95
Current
(2009:Q2 trough)

90
85
80

-24

-20

-16

-12

-8

-4
Trough
4
Quarters from trough

8

12

16

20

24

Note: The 1960‒2007 average excludes the 1980 recession due to overlap with the
1981‒82 recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts; National Bureau of
Economic Research; CEA calculations.

declined for 13 quarters through the first quarter of 2013, ended the year at
a higher level than in the first quarter, marking its first increase over three
quarters since 2009. The cumulative decline in State and local purchases
during this recovery contrasts with the usual experience during recoveries
(Figure 2-10). In a typical recovery, growth in State and local government
bolsters the economic recovery. In contrast, declines in State and local government have been a headwind to private-sector growth and hiring during
the first four years of this recovery.
Similar to the 2013 pickup in spending, State and local employment
has begun to show signs of life, adding 32,000 jobs during the 12 months
of 2013, after shedding almost 700,000 jobs from the end of the recession
through year-end 2012.
Despite these positive signals during 2013, major obstacles to growth
remain: in particular, the burden of unfunded pension obligations of State
and local governments. In its benchmark revision to the National Income
and Product Accounts of the United States in July 2013, the Commerce
Department, in cooperation with the Federal Reserve, began to measure
State and local defined-benefit plans on an accrual basis rather than a cash
basis, thereby tracking funded and unfunded pension liabilities. As can be
seen in Figure 2-11, the size of these liabilities relative to State and local
receipts ballooned immediately after the recession and remains elevated at a
The Year In Review And The Years Ahead  |  63

Figure 2-11
State and Local Pension Fund Liabilities, 1952–2013

Percent of annual receipts
100
80
60

2013:Q3

40
20
0
-20
-40
-60
1950

1960

1970

1980

1990

2000

2010

Note: Shading denotes recession.
Source: Federal Reserve Board, Financial Accounts of the United States.

level that is currently at about 60 percent of a year’s revenue for the sector.
Adding in State and local bond liabilities does not change the shape of the
plot shown in the figure, although they elevate the level of the liabilities-toreceipts ratio to about two hundred percent of a year’s revenue.

International Trade
In 2013, U.S. exports of goods and services to the world averaged
nearly $189 billion a month and imports averaged nearly $229 billion a
month (Figure 2-12). Exports accounted for 13.5 percent of U.S. production
(GDP) in 2013, the same as in 2011 and 2012.
The U.S. trade deficit, the excess of the Nation’s imports over its
exports, averaged nearly $40 billion a month in 2013. Import demand fell
during the recession and, as a result, the trade deficit fell from $66 billion in
July 2008 to $25 billion in May 2009. Exports fell too because of recessionrelated declines in domestic demand abroad (see Figure 2-13), but the recession was not as severe in many parts of the global economy as in the United
States. Since May 2009, growth rates of exports and imports have each been
averaging about 0.8 percent a month.
Figure 2-13 suggests that slower economic growth among our main
trading partners dampens U.S. export growth. In recent years, the top five

64  |  Chapter 2

Billions of dollars
250

Figure 2-12
Trade in Goods and Services, 2007–2013
Dec-13

200

Exports

150

0
0

100

0

50

0

0

0

-50

Balance

-100

0
0

-150

0

(-) Imports

-200
-250
-300

1

0
2007

2008

2009

2010

2011

2012

2013

2014

0

Source: U.S. Census Bureau, Foreign Trade Division.

Figure 2-13
U.S. Exports Growth, 2009–2013

Foreign growth, 4-q percent change
4
3

U.S. export growth, 4-q percent change
20
Actual U.S. Exports Growth
(right axis)
15

2

10

1

5

0

0

-1

Real domestic
demand growth
in trading partners,
weighted by trade
shares from:
(left axis)

-2
-3
-4
2009:Q1

2010: Q1

2011:Q1

China

Japan

European Union

Mexico

-10
-15

Canada
2012: Q1

-5

2013: Q1

-20

Source: IMF; Eurostat; Bureau of Economic Analysis; U.S. Census Bureau, Foreign Tade Divison.

The Year In Review And The Years Ahead  |  65

Percent of GDP
2

Figure 2-14
Current Account Balance, 1985–2013
0

1

0

0
-1

2013:Q3

-2
-3

0
0

-4
-5

0

-6
-7
1985:Q1

1990:Q1

1995:Q1

2000:Q1

2005:Q1

2010:Q1

0

Note: Shading denotes recession.
Source: Bureau of Economic Analysis, National Income and Product Accounts.

destinations for U.S. exports, in order from highest to lowest typically were:
Canada, the European Union, Mexico, China, and Japan. While growth generally slowed in all these trading partners, it actually turned to recession for a
time in our No.2 (European Union) and No. 5 (Japan) export recipients, and
their recoveries look to be gradual. In the European Union, real GDP fell 0.7
percent during the four quarters of 2012, then grew 1.1 percent during 2013,
and is forecasted to grow 1.4 percent during 2014 (European Commission
2013). Japan’s real GDP fell 0.4 percent during the four quarters of 2012, but
grew 2.7 percent during 2013, but is projected to edge up only 0.6 percent in
2014 (OECD 2013).
The trade balance is the major component of the current account balance. Other components of the current account balance include net income
on overseas assets and unilateral transfers such as foreign aid and remittances. The United States has run a current account deficit in all but two
quarters since 1985; however, the trend from 1990 through the mid-2000s
of ever-increasing deficits appears to have reversed. Figure 2-14 shows the
current account balance as a percentage of GDP since 1985. Since peaking
at more than 6 percent of GDP in the fourth quarter of 2005, the current
account balance has fallen as a share of GDP by more than 3 percentage
points. The sharpest decrease occurred during the recession of 2008-09, and
although there have been some periods of increase since then, the current
66  |  Chapter 2

account deficit recently reached a 15-year low in the third quarter of 2013 of
2.3 percent. An important driver of the decrease in current account deficit
in recent years is the increased domestic production of oil and gas, and the
associated reduced demand for imported oil, a shift discussed in more detail
later in the chapter. Removing oil, which depends on prices that are set on
world markets, the U.S. current account deficit is substantially smaller.
The United States has one of the most open and transparent trade and
investment regimes in the world, with a trade weighted applied tariff of 1.3
percent, making it a friendly market for imports and foreign investment. A
prime motivation behind U.S. trade policy initiatives is to ensure that our
accommodative trade and business environment is reciprocated when U.S.
actors have the same opportunities to compete in other markets that foreign
exporters and investors have in the United States. U.S. trade policy also seeks
to level the playing field, including by seeking to raise standards abroad so
they are closer to our own in key areas such as intellectual property, labor,
and environment. Box 2-2 discusses Administration trade policy initiatives.

Housing Markets
Housing activity continued its recovery in 2013 despite headwinds
from mortgage interest rates that rose approximately 1 percentage point
in mid-summer, continued tight credit conditions, and waning investor
demand for foreclosed properties. On the production side, new housing
starts for both single-family and multi-family structures continued their
2012 growth during 2013, despite relatively higher mortgage rates. For 2013
as a whole, starts were roughly 930,000 units, up from 780,000 in 2012, and
up from an all-time low of 554,000 units in 2009 (Figure 2-15).
Demand for housing increased, with new and existing home sales
reaching their highest levels in 2013 since the Great Recession. With the
lowest level of mortgage delinquencies and foreclosure completions in five
years, the composition of sales shifted markedly to non-foreclosure properties as fewer households sold homes under distressed conditions.
Supported by a tight supply of homes for sale, housing prices climbed
further in 2013, according to every major measure of house prices (Figure
2-16). As of November 2013, quality-adjusted house prices—as measured
by the FHFA index—were 7.7 percent higher than their year-ago level and
15.3 percent higher than at their trough in early 2011. Two considerations
provide some context for the brisk growth in house prices in 2013. First,
such behavior appears to be typical following recessions. Even though house
prices bottomed out well after the end of the Great Recession, the recovery
since then has, on net, been at a rate just below the average growth rate in
house prices seen during the aftermath of the eight post-war recessions of
The Year In Review And The Years Ahead  |  67

Box 2-2. Administration Trade Policy Initiatives
The United States has been pursuing the most ambitious trade
agenda in a generation. In the President’s first term, this included
upgrading, passing and implementing market-opening trade agreements
with Korea, Panama, and Colombia. U.S. tariffs on imports from those
countries were generally much lower than were the tariffs on U.S. exports
to those countries at the start of negotiations, and while the United States
did further lower tariff barriers as a result of the agreements, the larger
barriers were removed by U.S. trading partners.
In December 2013, the United States played a leadership role,
working with the 159 countries of the World Trade Organization
(WTO), to conclude a Trade Facilitation Agreement, the first multilateral trade agreement concluded by that body in its 20-year history.
This global agreement will expedite the movement of goods and services across borders and improve customs cooperation among WTO
Members, making it easier to support jobs through trade. Among other
things, the Agreement seeks to reduce documentary requirements,
require transparency in customs regulations and procedures, encourage
countries to accept electronic payments of customs duties and charges,
and ensure the quick release of perishable goods. Streamlined procedures
and enhanced transparency reduce the costs to businesses of exporting
and particularly assist small business for which logistical complexity can
be particularly challenging.1
The United States is currently pursuing two comprehensive, highstandard regional trade agreements that are ambitious in the size of the
overall markets that they seek to affect and in the scope of provisions to
be covered under the agreements. Negotiations are nearing completion
on the Trans-Pacific Partnership Agreement (TPP), which includes
12 nations that rim the Asia-Pacific region. Negotiations between
the United States and the 28-country European Union (EU) for the
Transatlantic Trade and Investment Partnership (T-TIP) are at an earlier
stage.
The Figure in this box demonstrates the importance of the regions
encompassed by these two proposed agreements to U.S. trade. Together,
the partner countries in the TPP and the T-TIP buy around 60 percent
of all U.S. exports and provide about 53 percent of U.S imports. The TPP
and T-TIP therefore seek to build on already robust trading relationships.
1 USTR. 2013a. “Weekly Trade Spotlight: The Benefits of the WTO Trade Facilitation
Agreement to Small Business.” (http://www.ustr.gov/about-us/press-office/blog/2013/
December/Benefits-of-WTO-Trade-Facilitation-Agreement-to-Small-Business).

68  |  Chapter 2

According to the Office of the United States Trade Representative
(USTR 2013a), the TPP “… is the foundation of the Obama
Administration’s economic policy in the Asia-Pacific Region” and
“promotes regional integration by establishing a common set of trade
and investment commitments, and also addresses 21st century issues
like state-owned enterprises, intellectual property rights, regulatory
convergence, and global supply chains.”2
U.S. Good and Services Imports and Exports, 2012
Exports

Imports

Other
45%

TTIP
21%

TPP
35%

Other
39%

T-TIP
22%

TPP
39%

Notes: Services trade data for TPP nations Brunei, Peru, and Vietnam are not available..
Source: Bureau of Economic Analysis; International Trade Commission.

The T-TIP seeks to strengthen trade and investment linkages
between the United States and the European Union and to set a template
for raising standards across the global trading system. It aims to create
new openings for service providers and to make regulations and standards more compatible between the two parties. The T-TIP should also
create new channels of cooperation to address shared interests in global
trade (USTR 2013b).3

2 USTR. 2013b. “Acting Deputy U.S. Trade Representative Wendy Cutler discusses Japan
and the TPP at the Peterson Institute for International Economics.” (“http://www.ustr.gov/
about-us/press-office/blog/2013/November/Cutler-TPP-Japan-PIIE”). The TPP participants
are: Australia, Brunei Darussalam, Canada, Chile, Japan, Malaysia, Mexico, New Zealand,
Peru, Singapore, Vietnam, and the United States.
3 USTR. 2013b. “Ambassador Froman discusses the Transatlantic Trade and Investment
Partnership at the Munich Security Conference.” (“http://www.ustr.gov/about-us/
press-office/blog/2013/November/Froman-Munich-Security-Conference”).

The Year In Review And The Years Ahead  |  69

Upon completion, the TPP and T-TIP agreements, together, will
place the United States at the center of an open trade zone representing
around two thirds of global economic output. The United States is also
in the process of negotiating several other agreements: an International
Services Agreement that would liberalize trade in services among
countries representing nearly 70 percent of the global services market;
another agreement that would further liberalize trade in information
technology products among countries representing 90 percent of that
market; and an agreement that would liberalize trade in environmental
goods among countries representing 86 percent of that market.
The Administration’s trade policy initiatives provide production
and consumption opportunities otherwise not available to the American
economy, and serve the ultimate goals of promoting growth, supporting
higher-paying jobs, and thus strengthening the middle class.

the 20th century. Second, house prices at the end of 2013 appear close to
their long-run relationship with rents, one measure of housing’s fundamental value. During the mid-2000s, house prices increased much more rapidly
than did rents before plummeting. The recent growth in house prices has
left prices broadly in line or perhaps above their long-run relationship with
rents, which suggests that much of these increases have been tied to improving economic fundamentals.
Home sales, construction, and prices generally appear to be on firm
footing in spite of higher mortgage rates, which increased about 100 basis
points to 4.4 percent, on net, after the May-July interest rate rise (discussed
earlier in this chapter) and remained close to that level for the remainder of
2013. Although nominal mortgage rates remain low by historical standards,
all else equal, higher rates raise the cost of financing a home purchase, which
puts downward pressure on housing demand and residential investment.
Also, builders’ capacity for funding new construction falls, albeit sometimes
with a delay, when interest rates rise. Indeed, residential investment, which
grew 15.5 percent during the four quarters of 2012, slowed to a 6.7 percent
rate of growth during 2013. The slowdown is accounted for by diminishing
increases in starts as well as a drop in commissions in the fourth quarter of
2013 due to a decline in sales of existing homes. But for the year as a whole,
new home sales increased 17 percent in 2013, while housing starts rose by a
comparable amount.
Another indication that housing market activity is holding steady:
households remain optimistic about home prices, according to the Reuters/
Michigan Survey of Consumers. Housing affordability remains high and 77

70  |  Chapter 2

Figure 2-15
Housing Starts, 1960–2013

Millions of units at an annual rate
2.5

Total
2.0

1.5
2013:Q4

1.0

One-unit
structures

0.5

0.0
1960:Q1

Multifamily structures
1970:Q1

1980:Q1

1990:Q1

Note: Shading denotes recession.
Source: Census Bureau, New Residential Construction.

2000:Q1

2010:Q1

Figure 2-16
National House Price Indexes, 2000–2013

Index, Jan. 2012 = 100
160
150
140
130

S&P/Case-Shiller
(Dec. 13)

120

Zillow
(Jan. 14)

FHFA
(Dec. 13)

110
100

CoreLogic
(Dec. 13)

90
80
2000

2002

2004

2006

2008

2010

2012

2014

Note: The S&P/Case-Shiller, FHFA, and CoreLogic indexes all adjust for the quality of homes sold but
only cover homes that are bought or sold, whereas Zillow reflects prices for all homes on the market.
Shading denotes recession.
Source: Zillow; CoreLogic; FHFA; S&P/Case-Shiller.

The Year In Review And The Years Ahead  |  71

percent of households report that it is a “good time to buy a house” (Reuters/
Michigan Survey of Consumers).
At a more fundamental level, pent-up demand for housing due to suppressed levels of household formation since 2009 is likely to boost housing
demand and to help absorb the large supply of vacant homes and homes still
in the foreclosure process. During the Great Recession, the number of new
households forming each year dropped to below 1 million a year and has
remained low ever since. As Figure 2-17 shows, during the housing bubble of
the mid-2000s more homes were built than were consistent with the underlying rate of household formation based on demographic trends that would
call for about 1.6 million new housing units a year. This oversupply peaked
in 2007 and—because of low levels of home construction—this oversupply
began to fall. And by 2011, the oversupply turned into an undersupply. The
increase in the stock of homes now lags behind the usual rates of household
formation.
As employment prospects improve, household formation is likely to
pick up. However, the extent to which the increase in the number of households translates into stronger housing demand depends critically on the
easing of credit standards (that might have been over-tightened following
the financial crisis), particularly for first-time homebuyers. In 2013, lending
standards eased somewhat for prime residential mortgages, according to
the Federal Reserve’s Senior Loan Officer Opinion Survey, and this easing
helped support a rise in mortgage purchase originations from the low levels
seen in recent years.

Energy
In 2013, the United States continued to benefit from developments in
the oil and gas sectors, as well as from growth in energy efficiency and the
production and integration of renewable energy. As shown in Figure 2-18,
net petroleum imports have fallen from more than 12 million barrels a day
in 2005 to approximately 6.2 million barrels a day in 2013. Moreover, as
shown in Figure 2-19, beginning in October 2013, domestic crude oil production exceeded crude oil imports for the first time since 1995.
Crude and refined oil products constitute the vast majority of the
country’s energy imports. This reduction in energy imports has multiple
benefits: it has been a major driver of the improvement in the U.S. balance of
trade, it reduces the vulnerability of the U.S. economy to foreign oil supply
disruptions, and it supports American jobs both in energy production and
in manufacturing. The dramatic increase in domestic oil and natural gas
production added about 0.2 percentage point to U.S. GDP growth in both
2012 and 2013.
72  |  Chapter 2

Figure 2-17
Cumulative Over- and Under-Building of Residential
and Manufactured Homes, 1996–2013

Millions of units
3.0
Relative to projected
annual average demand
2.0
for new units based
on demographic trends

Apr. 2007

1.0
0.0
-1.0
"Boom years"
1996–2006

-2.0
-3.0
1996

1998

2000

2002

"Correction years"
2007–2013

Dec. 2013

2004

2006

2008

2010

2012

2014

Source: Census Bureau, New Residential Construction (completions) and Manufactured
Homes Survey (placements); CEA (1998); CEA calculations.

The ongoing trend toward reduced energy imports is driven both
by roughly stable energy demand and increases in domestic energy supply.
Overall, economy-wide energy use has declined 0.8 percent at an annual
rate since 2007. The increase in domestic energy supply reflects major gains
in unconventional oil and natural gas production. The sharp increase in
unconventional domestic gas production has led to a 73 percent drop in the
wholesale (Henry Hub) price of natural gas from a high of $13.42 in October
2005 to $3.68 in October 2013. The United States is now the largest producer
of natural gas in the world, and the 2013 International Energy Outlook
projects that the United States will remain the largest producer through 2030
(U.S. Energy Information Administration 2013). Since 2007, over 50,000
jobs have been created in oil and natural gas extraction alone, with more
than 160,000 jobs being created along the oil and natural gas supply chain.
Low natural gas prices also help manufacturing as discussed below, and have
been an important driver in the reduction of U.S. carbon dioxide emissions
as electricity production has shifted from coal to cleaner-burning natural
gas. Indeed, between 2010 and 2013, the total U.S. carbon dioxide emissions
from energy consumption decreased by 4.3 percent. In addition to providing
cost savings to consumers today, this reduction in greenhouse gas emissions
will benefit future generations.

The Year In Review And The Years Ahead  |  73

Figure 2-18
Petroleum Net Imports, 1980‒2015

Million barrels per day
14

Actual Projected

12
10
8
6
4
2
0
1980

1985

1990

1995

2000

2005

2010

2015

Source: Energy Information Admnistration, Monthly Energy Review, Short-Term Energy
Outlook.

Figure 2-19
Monthly Crude Oil Production and Net Imports, 1990‒2013

Million barrels per day
11

Net Imports

10

Dec-2013

9
8
7
6
Production

5
4
3
1990

1995

2000

2005

Note: This data is not seasonally adjusted.
Source: Energy Information Administration, Petroleum Supply Monthly.

74  |  Chapter 2

2010

2015

Trillion Btu
1,800

Figure 2-20
Wind and Solar Energy Production, 2000‒2013
2013

1,600
1,400

Wind

1,200
1,000
800
600
Solar

400
200
0
2000

2002

2004

2006

2008

2010

2012

2014

Note: Data for 2013 are projections.
Source: Energy Information Adminstration, Monthly Energy Review, Short-Term Energy Outlook.

The other part of the energy supply story, shown in Figure 2-20, is the
dramatic growth in wind and solar electricity production, which have each
more than doubled since President Obama took office. In 2012, a record
13 gigawatts of new wind power capacity was installed, roughly double the
amount of newly installed capacity in 2011. More than 5 gigawatts were
installed in December 2012 alone as firms scrambled to take advantage of
the expiring 2.3 cent per kilowatt-hour production tax credit (Congress later
extended the tax credit for 2013). These 13 gigawatts of new wind capacity
represented the largest share of additions to total U.S. electric generation
capacity in 2012.
In addition to increased domestic supply, energy imports have
declined because of reduced energy demand across all the main energy sectors. As shown in Figure 2-21, gasoline demand per capita rose through the
early 2000s and plateaued in the mid-2000s before dropping substantially
during the recession. As the economy has recovered, however, gasoline
demand per capita has continued to fall. Some of this continued decline in
gasoline demand stems from the relatively high real gasoline prices shown in
Figure 2-21, but that is only a partial explanation. Increasing fuel efficiency
brought about by Federal fuel efficiency standards also played a role; and,
in 2012, the Administration finalized fuel economy standards that, together
with the Administration’s first round of standards, will nearly double the

The Year In Review And The Years Ahead  |  75

Figure 2-21
U.S. Per Capita Consumption of Gasoline and Real
Gasoline Prices, 2000‒2013

Gallons per person per day
1.10
1.05
1.00

Dec-2013

Consumption 12-month moving
average (left axis)

2013 $
4.50
4.00
3.50

0.95

3.00

0.90

2.50

0.85

2.00

0.80

1.50

Retail gasoline price (right-axis)

0.75
0.70
2000

1.00
2002

2004

2006

2008

2010

2012

2014

0.50

Note: Retail gasoline prices deflated using PCE chain price index. Consumption includes
residential, commercial, industrial and transporation sectors.
Source: Energy Information Administration, Monthly Energy Review; Census Bureau; CEA
calculations.

fuel economy of light- duty vehicles to the equivalent of 54.5 miles per gallon by the 2025 model year from 2010 levels. Further, beginning in model
year 2014, medium- and heavy-duty trucks must meet new energy efficiency
standards as well, which will increase their fuel efficiency by 10 to 20 percent
by 2018.
Despite these significant improvements in energy efficiency and
reductions in energy-related carbon dioxide emissions, continued work is
needed to reduce greenhouse gas emissions. In June 2013, the President laid
out his Climate Action Plan (summarized in Box 2-3), which aims to reduce
both greenhouse gas emissions and the impact of climate change on future
generations.

Labor Markets
The major U.S. labor market indicators continued to recover during
2013 even as the unemployment rate remained unacceptably high. As shown
in Figure 2-22, the unemployment rate dropped 1.2 percentage points during the 12 months of 2013, somewhat faster than the average 0.9 percentage
point annual drop during the three preceding years. Similarly, as shown in
Figure 2-23, establishment employment finished its third year of growth at

76  |  Chapter 2

Box 2-3: The Climate Action Plan
In 2009, the President committed the United States to cut greenhouse gas emissions by approximately 17 percent below 2005 levels by
2020. The President’s June 25, 2013 Climate Speech noted that, “Climate
change represents one of our greatest challenges of our time, but it is a
challenge uniquely suited to America’s strengths.” Following that speech,
the President laid out a three-pronged approach to addressing the
challenges of climate change: 1) reduce carbon emissions in the United
States; 2) prepare America for the impacts of climate change; and 3) lead
international efforts to fight climate change and adapt to its impacts.
The United States has already made substantial progress toward
the 2020 emissions reduction goal. In 2012, U.S. carbon emissions
declined to their lowest levels in nearly 20 years while the economy
continued to grow. The Administration has continued to build on this
progress by proposing tough new rules to cut carbon pollution from
new fossil-fuel-fired power plants and by developing new rules to reduce
carbon pollution from existing power plants, as well as by proposing
new energy efficiency standards for appliances, announcing new funding
for advanced fossil-energy projects, and other important actions. These
steps will help to protect the welfare of future generations and will put
America in a position to achieve sustainable economic growth by relying
on the Nation’s clean energy sources.
The Climate Action Plan also lays out steps to ensure that the
country is ready to manage the inevitable and already realized impacts
of climate change. For example, the Administration will lead an effort to
assist State and local governments to make our infrastructure, communities, and natural resources more resilient, including through strengthening our roads, bridges, and shorelines to better protect people’s homes,
businesses and everyday lives from severe weather worsened by climate
change.
Climate change is a global challenge that cannot be solved by any
single country; therefore, it is imperative for the United States to couple
action at home with leadership internationally. America must help forge
a truly global solution to this global challenge by galvanizing international action to significantly reduce emissions (particularly among the
major emitting countries), preparing for climate impacts, and driving
progress through international negotiations.

The Year In Review And The Years Ahead  |  77

Figure 2-22
Unemployment Rate, 1979–2014

Percent
11
10
9
8
7

Jan-2014

6
5
4
3
1979

1983

1987

1991

1995

1999

Note: Shading denotes recession.
Source: Bureau of Labor Statistics, Current Population Survey.

2003

2007

2011

Figure 2-23
Nonfarm Payroll Employment, 2007–2014

12-month change, millions, not seasonally adjusted
4
2

Private

Jan-2014

Total

0
-2
-4
-6
-8
Jan-07

Jan-08

Jan-09

Jan-10

Jan-11

Jan-12

Jan-13

Note: Total excludes temporary decennial Census workers. Shading denotes recession.
Source: Bureau of Labor Statistics, Current Employment Statistics.

78  |  Chapter 2

Jan-14

Figure 2-24
Unemployment Rate by Duration, 1990–2014

Percent of civilian labor force
8

Jan-2014

7
6

Unemployed for
26 Weeks or Less

5
4

2001-07 Average

3
2
2001-07 Average

1
0
Jan-90

Jan-94

Jan-98

Unemployed for
27 Weeks & Over

Jan-02

Note: Shading denotes recession.
Source: Bureau of Labor Statistics, Current Population Survey.

Jan-06

Jan-10

Jan-14

roughly 2.3 million a year (or about 190,000 a month)2. The strength of the
labor market was not matched by the growth of output, with some puzzling
developments in the relation between the unemployment rate and GDP, and
also the relationship between employee-hours and output (productivity).
The current elevation of the unemployment rate is entirely due to
long-term unemployment. In December 2013, the unemployment rate for
workers unemployed 26 weeks or less fell to lower than its average in the
2001-07 period, while the unemployment rate for workers unemployed 27
weeks or more remained higher than at any time prior to the Great Recession.
But the long-term unemployment rate has declined by 1.1 percentage points
in the last two years, a steeper decline than the 0.5 percentage point drop in
the short-term unemployment rate over that period (Figure 2-24).
2 The Department of Labor conducts several labor market surveys. The household survey—
conducted in cooperation with the Bureau of Census—queries 60,000 households every month
with a variety of questions including whether members of that household were working
or looking for a job, and this survey is the source of the unemployment rate, among other
important statistics. The Establishment (or Payroll) survey queries employers about how
many workers they employed, how many hours did they work, and what they were paid.
The Establishment survey is the source of the most quoted figures for job growth. The Job
Openings and Labor Turnover Survey (JOLTS) (a relatively new survey, begun in 2000) also
queries employers about their job openings (vacancies) as well as their hiring, quits, and
layoffs.

The Year In Review And The Years Ahead  |  79

Figure 2-25
Predicted vs. Actual Manufacturing
Payroll Employment, 2000–2014

Millions, seasonally adjusted
18
17
16
15
14
13

Actual
(Jan-2014)

12
11
10
2000

Range of normal cyclical rebound
(2013:Q4)
2002

2004

2006

2008

2010

2012

2014

Note: Gray shading denotes recession. Blue shading denotes the 95% confidence interval of the normal
cyclical rebound.
Source: Bureau of Labor Statistics, Current Employment Statistics; CEA calculations.

Of the 2.3 million increase in payroll employment during the 12
months of 2013, about 4 percent was in manufacturing, 7 percent was in
construction, and 90 percent was in the private service-providing industries.
Within the service-providing industries, the sectors showing the strongest
job growth were professional and business services (29 percent of total
employment growth), retail trade (15 percent) and health care (9 percent of
the total).
Over the course of the recovery, manufacturing has added 622,000
jobs since its trough. Some have pointed to this growth, following a decade
of job losses, as indicating a resurgence in manufacturing, while others
have suggested that this rebound simply reflects the normal cyclical pattern
given the depth of the recession. The Council of Economic Advisers (CEA)
analysis suggests that while the overall recovery did in fact contribute to the
stabilization of manufacturing job losses, the job gains are about 500,000
above and beyond what would be associated with the historical cyclical pattern (Figure 2-25).
Further evidence of the healing of the job market comes from the
number of job vacancies, which increased 6 percent during the 12 months
through November (the latest available at press time). There are now 2.6
unemployed workers for each job vacancy, less than half of the number

80  |  Chapter 2

following the business-cycle trough in 2009, but still in excess of the average
two-to-one ratio from 2001 to 2007.

Wage Growth and Price Inflation
Hourly compensation (including non-wage benefits) increased 2.0
percent during the 12 months of 2013, the fourth consecutive year of growth
at around a 2-percent rate, according the Employment Cost Index. Prices
in the nonfarm business sector increased at a 1.6 percent annual rate during these four years; so from the viewpoint of a typical employer, the real
product hourly compensation increased 0.4 percent at an annual rate. These
four-year growth rates for real hourly compensation were less than the 1.2
percent increase in labor productivity, and as a result, the labor share of
nonfarm business output (and of gross domestic income) declined.
Growth in real wages (that is, take-home wages not including benefits)
of production workers picked up to 0.7 percent in 2013 from a 0.1 percent
decline a year earlier. Nominal wages increased 2.2 percent in 2013 (up from
a year earlier) while prices for wage earners rose 1.5 percent (down from a
year earlier).
Consumer prices excluding food and energy (the core CPI) rose 1.7
percent during the 12 months of 2013, down from 1.9 percent during 2012.
Overall, consumer prices rose just 1.5 percent during the year as food prices
increased only 1.1 percent and energy prices inched up 0.5 percent.
Although inflation edged lower in 2013, the relative stability of inflation during the recession and slow recovery presents a puzzle. During this
period, the unemployment rate has been much higher than its long-term
average, and higher than the rate that is generally considered consistent with
stable inflation. Under these circumstances, conventional economic theory
and historical experience would have expected declining inflation and
perhaps even negative inflation. In contrast, inflation has remained fairly
stable since the business-cycle peak with the 12-month change in core CPI
inflation never falling below 0.6 percent, raising a puzzle of missing disinflation. Standard explanations of the missing disinflation focus on anchored
expectations arising from increased Federal Reserve credibility associated
with targeting an inflation rate of approximately 2 percent (for example,
Fuhrer and Olivei 2010, Stock and Watson 2010, Ball and Mazumder 2011).
In addition to anchored expectations, a second factor behind the lack
of disinflation appears to be the unusually high fraction of the long-term
unemployed in this recovery. Those unemployed for only short durations
search more intensely for a new job (Krueger and Mueller 2011) and are
also potentially more likely to match with a good job, which suggests that the
short-term unemployed put more downward pressure on wages than those
The Year In Review And The Years Ahead  |  81

Box 2-4: Unemployment Duration and Inflation
A standard wage-price Phillips curve relates wage inflation, minus
expected price inflation, to the unemployment rate. A benchmark
specification uses previous-year price inflation as a proxy for expected
price inflation (for example, Gordon 1990). In this specification, 2009-13
represents a cluster of outliers in which wages fell less than would have
been expected based on historical relationships and the very-elevated
unemployment rate. But some research, both older and recent, suggests
that the composition of unemployment by duration can be important,
in particular that the short-term unemployment rate might be a better
measure of wage pressure than the total unemployment rate, perhaps
because employers prefer to hire those who have spent less time since
their last job or because job-search intensity declines with the duration
of unemployment (Layard, Nickell, and Jackman 1991, Blanchard and
Diamond 1994, Krueger and Mueller 2011, Stock 2011, Gordon 2013). In
fact, as is shown in the Figure below, if this wage-price Phillips relation is
expressed in terms of the short-term unemployment rate rather than the
overall unemployment rate, the recovery is no longer an outlier.
Growth in Expected Average Hourly Earnings vs.
Short-Term Unemployment, 1976‒2013

Wage inflation minus preceding year's price inflation (percentage points)

3

2000

2

1999

2

2001
2007

1997

2006

1

2005
2004

0

1996
2002

2008

2003

2013

1978

2010

1977

1979

2011

2009

1995
2012

-1

1988
1989

1994

1987

1

1980

0

1990
1993

1981

1991

-1

1985

1992

1976
1984

-2

1986

-3
-4

3

1998

1976-2008

2

3

4

5

-2

1983

-3

1982

6

Unemployment rate: 14 weeks & less (percent)
Note: Average hourly earnings covers production and nonsupervisory workers.
Source: Bureau of Labor Statistics, Current Population Survey; Bureau of Economic
Analysis, National Income and Product Accounts.

7

-4

A second way to illustrate the lack of disinflation is to consider
dynamic forecasts produced by a standard backwards-looking Phillips
curve, in which the change in core price inflation depends on past core

82  |  Chapter 2

price inflation and a measure of economic slack. Estimating this model
through 2007, then simulating it using the actual unemployment rate
post-2007, but not using prices during that period (a method referred to
as a dynamic simulation), permits judging whether the actual inflation
path accords with what would have been predicted based on historical
experience. As the Figure below shows, when the dynamic simulation
is conducted using the total unemployment rate, the historical relationship would have predicted substantially more disinflation than actually
occurred. In contrast, there is no missing disinflation when the measure
of economic slack is the short-term unemployment rate. The wage-price
Phillips curve in the figure above, and the dynamic price Phillips curve
forecasts in the figure below, suggest that the short-term unemployment
rate might be a better measure of effective economic slack than the longterm unemployment rate.
Core Consumer Price Inflation: Actual vs. Predicted, 2006‒2013

4-quarter change
3.0

2013:Q4

2.5

Predicted using short-term
unemployment rate

2.0
Actual

1.5
1.0

Predicted using official
unemployment rate

0.5
0.0

2006

2007

2008

2009

2010

2011

2012

2013

Note: Dynamic simulation of price-price Phillips Curve. Core consumer price inflation is measured by
the price index for consumer spending excluding food and energy in the National Income Accounts.
Shading denotes recession.
Source: Bureau of Labor Statistics, Current Population Survey; Bureau of Economic Analysis, National
Income and Product Accounts; CEA calculations.

who have been unemployed for more than six months. While the relationship between the overall unemployment rate and inflation in recent years is
puzzling, the relationship between short-term unemployment and inflation
is less so, as discussed in Box 2-4.

The Year In Review And The Years Ahead  |  83

The Long-Term Outlook
The 11-Year Forecast
Although real GDP has grown at a roughly 2 percent rate for each of
the past three years, a foundation is in place for faster growth during 2014,
as most components of demand point to faster growth while the supply side
does not appear constraining. Although fiscal policy has generally increased
the level of output, it has been a drag on GDP growth in the last several years
and especially in 2013. The rate of decline in the deficit-to-GDP ratio will
likely moderate in 2014 under the President’s Budget policy as well as under
current law, as noted earlier in this chapter. Consumer spending likely has
adjusted by now to the expiration of the payroll tax cut, but it probably has
not adjusted to the gains in housing and stock market wealth. End-of-2013
indicators suggest that growth among our European trading partners is
looking up, suggesting stronger exports in 2014 than in 2013. While not
much growth can be expected from real State and local spending, the latest
quarterly data suggest that it will no longer be a substantial drag on overall
growth. As discussed earlier in the chapter, firms appear ready to step up
business investment if consumer spending picks up. Business investment
will grow if everything else does. With the unemployment rate in January
2014 at 6.6 percent and the capacity utilization rate in manufacturing at
about 77 percent, the economy has room to grow.
The Administration’s economic forecast, as finalized on November
21, 2013 is presented in Table 2.1, and is the forecast that underpins the
President’s FY 2015 Budget. The Administration expects real GDP to accelerate from a 2.3 rate during the four quarters of 2013 to 3.3 percent during
2014. (Data released after the forecast was finalized show a slightly fasterthan expected growth rate during 2013, 2.5 percent rather than 2.3 percent.)
These projections, as is standard for the Administration’s budget forecast,
assume enactment of the President’s Budget—including the Opportunity,
Growth and Security initiative.
The forecast assumed that the unemployment rate would fall 0.5
percentage point in the four quarters of 2014. Since the forecast was finalized in November the unemployment rate has fallen from 7.3 percent (as
first published for October) to 6.6 percent in January 2014, considerably
faster than the pace forecasted by the Administration or by the consensus of
private sector forecasters. As a result, the Administration’s budget forecast
of an unemployment rate averaging 6.9 percent in 2014 does not reflect the
latest information and an updated projection would forecast a continued
decline in the unemployment rate over the course of the year. A revised
84  |  Chapter 2

Table 2–1
Administration Economic Forecast
Nominal
GDP

Real
GDP price Consumer
GDP
index
price index
(chain-type) (chain-type) (CPI-U)

Unemployment rate
(percent)

Percent change, Q4-to-Q4

Interest rate, Interest rate,
91-day
10-year
Treasury
Treasury
bills
notes
(percent)
(percent)

Level, calendar year

2012 (actual)

3.8

2.0

1.8

1.9

8.1

0.1

1.8

2013

3.6

2.3

1.3

1.1

7.5

0.1

2.3

2014

5.0

3.3

1.6

1.9

6.9

0.1

3.0

2015

5.2

3.4

1.8

2.1

6.4

0.3

3.5

2016

5.3

3.3

2.0

2.2

6.0

1.2

4.0

2017

5.3

3.2

2.0

2.3

5.6

2.3

4.3

2018

4.7

2.6

2.0

2.3

5.4

3.2

4.6

2019

4.6

2.5

2.0

2.3

5.4

3.6

4.7

2020

4.5

2.4

2.0

2.3

5.4

3.7

4.9

2021

4.4

2.3

2.0

2.3

5.4

3.7

5.0

2022

4.4

2.3

2.0

2.3

5.4

3.7

5.1

2023

4.4

2.3

2.0

2.3

5.4

3.7

5.1

2024

4.4

2.3

2.0

2.3

5.4

3.7

5.1

Note: These forecasts were based on data available as of November 21, 2013, and were used for the FY 2015
Budget. The interest rate on 91-day T-bills is measured on a secondary-market discount basis.
Source: The forecast was done jointly with the Council of Economic Advisers, the Department of Commerce,
(the Bureau of Economic Analysis) and the Department the Treasury, and the Office of Management and Budget.

Administration forecast will be released in the Mid-Session Review of the
Budget over the summer.
Real GDP is projected to grow in the 3.2-to-3.4 percent range during
the four years through 2017, as the economy gradually uses up the slack suggested by the current elevated level of the unemployment rate. By the fourth
quarter of 2017, the unemployment rate is expected to fall to 5.5 percent.
Nominal interest rates are currently low due to the fact that the economy has not fully healed together with monetary policy that has kept rates
low across a wide range of Treasury securities. Consistent with the forward
policy guidance at the time that the forecast was made, interest rates are projected to increase for maturities that extend through periods covering dates
when the unemployment rate is expected to fall below 6.5 percent. Interest
rates are expected to continue to climb as the economy approaches full
employment. After that point, projected real interest rates (that is, nominal
rates less the projected rate of inflation) will be close to their historical average. These interest-rate paths are close to those projected by the consensus
of professional economists.

The Year In Review And The Years Ahead  |  85

Growth in GDP over the Long Term
As discussed earlier, the growth rate of the economy over the long run
is determined by the growth of its supply-side components, demographics,
and technological change. The growth rate that characterizes the long-run
trend in real U.S. GDP—or potential GDP—plays an important role in guiding the Administration’s long-run forecast. Through 2020, potential real
GDP is projected to grow at a 2.4 percent annual rate, before slowing to 2.3
percent during the three-year period 2021–24. These growth rates are slower
than in the past because of the movement of the baby-boom generation into
the retirement years. These growth rates for potential real GDP are based
on the assumption of no change to immigration law. If, however, immigration law were to be revised along the lines of the Border Security, Economic
Opportunity, and Immigration Modernization Act (S.744) that the Senate
approved in June, the growth rate of potential real GDP would be higher,
because of faster growth of the working-age population and increased total
factor productivity growth (Box 2-5). The Budget totals reflect the effects
of immigration reform by incorporating the CBO score directly into the
Budget. This CBO score incorporates both direct policy effects and the
broader economic impact. In order to avoid double counting with this
estimate, the economic forecast does not reflect the effects of immigration
reform.
Table 2-2 shows the Administration’s forecast for the contribution of
each supply-side factor to the growth in potential real GDP: the workingage population, the rate of labor force participation, the employed share of
the labor force, the ratio of nonfarm business employment to household
employment, the length of the workweek, labor productivity, and the ratio
of real GDP to nonfarm output. Each column in Table 2-2 shows the average
annual growth rate for each factor over a specific period of time. The first
column shows the long-run average growth rates between the businesscycle peak of 1953 and the business-cycle peak of 2007, with business-cycle
peaks chosen as end points to remove the substantial fluctuations within
cycles. The second column shows average growth rates between the fourth
quarter of 2007 and the third quarter of 2013, a period that includes the
2007–09 recession and the recovery so far. The third column shows the
Administration’s projection for the entire 11-year forecast period, from the
third quarter of 2013 to the fourth quarter of 2024. And the fourth column
shows average projected growth rates between the fourth quarter of 2020
and the fourth quarter of 2024; that is, the last four years of the forecast
interval when the economy is assumed to settle into steady-state growth.

86  |  Chapter 2

Table 2–2
Supply-Side Components of Actual and Potential Real GDP Growth, 1952–2024
Growth ratea
History,
peak-topeak

Recent
history,
since peak

Forecast

Out-year
forecast

1953:Q2 to
2007:Q4b

Component

2007:Q4 to
2013:Q3

2013:Q3 to
2024:Q4

2020:Q4 to
2024:Q4

1.  Civilian noninstitutional population aged 16+

1.4

1.1

1.0

0.9

2.  Labor force participation rate

0.2

–0.7

–0.2

–0.3

3.  Employed share of the labor force

0.0

–0.5

0.2

0.0

4.  Ratio of nonfarm business employment to
  household employment

0.0

–0.5

0.0

–0.4

5.  Average weekly hours (nonfarm business)

–0.3

0.0

0.0

0.0

2.2

1.7

2.1

2.2

–0.2

–0.1

–0.3

–0.1

8.  Sum: Actual real GDPc

3.3

1.1

2.7

2.3

9.  Memo: Potential real GDP

3.3

2.0

2.3

2.3

6.  Output per hour (productivity, nonfarm business)c
7.  Ratio of real GDP to nonfarm business outputc

a. All contributions are in percentage points at an annual rate, forecast finalized in November 2013. Total may
not add up due to rounding.
b. 1953:Q2 and 2007:Q4 are business-cycle peaks.
c. Real GDP and real nonfarm business output are measured as the average of income- and product-side measures.
Note: Population, labor force, and household employment have been adjusted for discontinuities in the population
series. Nonfarm business employment, and the workweek, come from the Labor Productivity and Costs database
maintained by the Bureau of Labor Statistics. 
Source: Bureau of Labor Statistics, Current Population Survey, Labor Productivity and Costs; Bureau of Economic
Analysis, National Income and Product Accounts; Department of the Treasury; Office of Management and Budget;
CEA calculations.

The population is projected to grow 1.0  percent a year, on average,
over the projection period (line 1, column 3), following the projection
published by the Social Security Administration. Over this same period, the
labor force participation rate is projected to decline 0.2 percent a year (line
2, column 3). This projected moderate decline in the labor force participation rate reflects a balance of opposing influences: a negative demographic
trend partially offset by increasing demand. The entry of the baby-boom
generation into its retirement years is expected to reduce the participation rate trend by about 0.4 percent a year through 2020 and by about 0.3
percent during the 2020-24 period (as can be seen in column 4). During
the next several years, however, rising labor demand due to the continuing
business-cycle recovery is expected to offset some of this downward trend.
Young adults, in particular, have been preparing themselves for labor-force
entry through additional education. The share of young adults aged 16 to
24 enrolled in school between January 2008 and December 2012 rose well
above its trend, enough to account for the entire decline in the labor force
participation rate for this age group over this period. As these young adults

The Year In Review And The Years Ahead  |  87

Box 2-5: Immigration Reform and Potential GDP Growth
Immigration reform would boost real GDP growth during
the 10-year budget window and for the 10 years through 2034 too.
Immigration reform would directly raise the growth of the workingage population. As a result, the labor force would grow faster as well.
According to the Congressional Budget Office (CBO), the labor force
would grow 0.35 percentage point a year faster through 2033 than
without the legislation. The faster growth of the labor force would be the
prime reason supporting an additional 0.3 percent a year of real GDP
growth.
In addition, CBO also assumes that immigration reform would add
to real GDP growth by boosting investment and raising the productivity of labor and capital (known as total factor productivity). Although
immigrants constituted just 12 percent of the population in 2000, they
accounted for 26 percent of the U.S.-based Nobel Prize winners between
1990 and 2000. Immigrants also comprised 25 percent of the founders
of public-venture–backed companies started between 1990 and 2005,
and they received patents at twice the rate of the native-born population.

complete their education, most are expected to enter or reenter the labor
force.
The employed share of the labor force—which is equal to one minus
the unemployment rate—is expected to increase at an average 0.2 percent a
year over the next 11 years. It is expected to be unchanged after 2018 when
the unemployment rate converges to the rate consistent with stable inflation.
The workweek is projected to be roughly flat during the forecast period,
somewhat less of a decline than its long-term historical trend of -0.3 percent.
The workweek is expected to stabilize because some of the demographic
forces pushing it down are largely spent, and because a longer workweek
is projected to compensate for the anticipated decline in the labor force
participation rate.
Labor productivity is projected to increase 2.1  percent a year over
the forecast interval and 2.2 percent in the long run (line 6, columns 3 and
4), roughly the same as the average growth rate from 1953 to 2007 (line 6,
column 1). The elevated rate of long-term unemployment poses some risk
to the projection insofar as the human capital of workers may deteriorate
with prolonged unemployment. That said, higher rates of school enrollment
among young adults in recent years, as noted, should contribute to productivity growth in the coming years.

88  |  Chapter 2

The ratio of real GDP to nonfarm business output is expected to subtract from GDP growth over the projection period (line 7, column 3), consistent with its long-run trend. The nonfarm business sector generally grows
faster than government, households, and nonprofit institutions, where an
accounting convention holds productivity growth to zero.
Summing the growth rates of all of its components, real GDP is projected to rise at an average 2.7 percent a year over the projection period (line
8, column 3), somewhat faster than the 2.3 percent annual growth rate for
potential real GDP (line 9, column 3). Actual GDP is expected to grow faster
than potential GDP primarily because of the projected rise in the employment rate (line 3, column 3) as millions of currently unemployed workers
find jobs.
Real potential GDP (line 9, column 4) is projected to grow more
slowly than the long-term historical growth rate of 3.3 percent a year (line
9, column 1). As discussed earlier, the projected slowdown in real potential GDP growth primarily reflects the lower projected growth rate of the
working-age population and the retirement of the baby-boom cohort. If the
effects of immigration reform were incorporated into this forecast, however,
then it would show a higher real potential GDP growth rate.

Conclusion
As of December 2013, private payroll employment had increased
for 46 months, and more gains are expected during the coming year. The
economy is well situated for a pickup in growth, with households having
made progress in deleveraging and building wealth, with housing demand
gathering momentum, with inflation that is low and stable, and especially
with the four-year period of fiscal consolidation now largely behind us.
This past year’s budget brinksmanship has receded into legislation that
will provide some stability during the coming year. If international economies and markets are stable or improving, that would support exports.
The energy sector has also supported sustainable growth with substantial
increases in domestic energy supply, declines in energy imports, and progress toward reducing carbon dioxide emissions. With these foundations, the
Administration forecast projects an increase in growth during the next few
years. The growth rate over the budget window will be limited, however, by
demographic forces that lower the participation rate, although immigration
reform would both raise the participation rate and raise the growth rate of
the working-age population.

The Year In Review And The Years Ahead  |  89

Even with this growth, however, the economy would remain below
its full potential and the unemployment rate would remain unacceptably
high. Additional sound policies would speed the return of the economy to
its full potential, including policies like investments in infrastructure and
increasing certainty through business tax reform. Conversely, adverse policy
developments in the United States or adverse shocks in the United States or
abroad could impede this favorable scenario.

90  |  Chapter 2

C H A P T E R

3

THE ECONOMIC IMPACT OF
THE AMERICAN RECOVERY
AND REINVESTMENT ACT
FIVE YEARS LATER

O

n February 17, 2009, President Obama signed into law the American
Recovery and Reinvestment Act of 2009, also known as the Recovery
Act, or “ARRA.” At the time, the country was going through the worst
economic and financial crisis since the Great Depression. In the year leading up to the passage of the Act, private employers shed 4.6 million jobs
and another 698,000 were lost that February alone. Trillions of dollars of
household wealth had been wiped out, and the economy’s total output, as
measured by real gross domestic product (GDP), was in the midst of its most
severe downturn since World War II.
The purpose of the Recovery Act was to provide countercyclical fiscal
support for the economy as part of a suite of monetary and fiscal policies
aimed at containing the already-severe recession that, had it spiraled further, could have resulted in a second Great Depression. The Act was also
intended to lay the foundation for a stronger and more resilient economy
in the future.
In the four years following the Recovery Act, the President built on
this initial step, signing into law over a dozen fiscal measures aiming to
speed job creation. These measures, which extended key elements of the
Recovery Act and provided new sources of support, were motivated by a
deepening understanding of the severity of the initial shocks to the economy,
as well as by new challenges that subsequently arose. These additional measures nearly doubled the size and impact of the Recovery Act’s fiscal support
to the economy through the end of 2012.
Nearly half of the jobs measures in the Recovery Act and subsequent
legislation, or $689 billion, were tax cuts—with most of them directed at

91

families. The other half was for investments in critical areas such as rebuilding bridges and roads, supporting teacher jobs, or providing temporary help
for those who found themselves unemployed because of the impact of the
Great Recession.
The economic picture today is much brighter. GDP per capita started
expanding in the third quarter of 2009 and reached its pre-crisis level in
about four years, considerably faster than the historical record suggests is
the typical pace of recovery following a systemic financial crisis.1 Since 2010,
the U.S. economy has also consistently added over 2 million private-sector
jobs a year, bringing the overall unemployment rate down to its lowest level
since October 2008. Job growth has been broad-based across sectors and
has withstood significant headwinds, including more recent fiscal contraction at all levels of government, and concerns stemming from the European
sovereign debt crisis.
As part of the unprecedented accountability and transparency provisions included in the Recovery Act, the Council of Economic Advisers
(CEA) was charged with providing to Congress quarterly reports on the
effects of the Recovery Act on overall economic activity, and on employment
in particular. In this chapter, CEA provides an assessment of the effects of
the Act through the third quarter of 2013, and of subsequent jobs measures
through 2012.
This chapter assesses the role of the Recovery Act and the subsequent
jobs measures in helping to facilitate the economic turnaround since 2009.
It updates previous estimates from the Council of Economic Advisers and
other sources on the Act’s contribution to employment and output growth,
and expands the estimates to reflect the impact of the full set of fiscal measures undertaken. The chapter also considers how many investments contained in the Recovery Act have laid the groundwork for a more productive
economy in the years ahead and will support growth long after the spending
authorized by the Act has fully phased out.
Consistent with the preponderance of evidence from numerous
private-sector, academic, and government analyses, this chapter finds that
the Recovery Act substantially boosted employment and output. CEA
estimates that, by itself, the Recovery Act saved or created an average of 1.6
million jobs a year for four years through the end of 2012 (cumulatively,
equivalent to about 6 million job-years, where a job-year is defined as one
full-time job for one year). In addition, the Recovery Act alone raised the
level of GDP by between 2 and 2.5 percent from late 2009 through mid-2011.
The Recovery Act also helped individuals, businesses, and State and local
governments directly affected by the downturn, and put the economy on a
1 See Reinhart and Rogoff (forthcoming).

92  |  Chapter 3

better trajectory for long-run growth by undertaking targeted investments
in education, energy, and health care, among other areas.
Combining effects of the Recovery Act and additional countercyclical
fiscal legislation that followed, CEA estimates that the cumulative gain in
employment was about 9 million job-years through the end of 2012. The
cumulative boost to GDP from 2009 through 2012 is equivalent to 9.5 percent of fourth quarter 2008 GDP.
While these estimates are substantial, they still understate the full
impact of the Administration’s economic policies in tackling the Great
Recession due to being based only on the effect of fiscal measures. CEA
estimates do not account for the broader set of responses that included policies to stabilize the financial system, rescue the auto industry, and provide
support for the housing sector—in addition to the independent actions
undertaken by the Federal Reserve.

The 2007-09 Recession and the
Early Policy Responses
In the run-up to the 2007-09 recession, the country experienced a
dramatic escalation in home prices starting in the mid-1990s, fueled by
lax mortgage underwriting standards and an abundance of global capital
in search of a safe, dollar-denominated return. This escalation came to an
abrupt halt in 2006. Home prices stopped rising and then started falling,
eventually dropping by 30 percent nationwide and even more in some areas.
Millions of homeowners found themselves “under water”—that is, their
mortgage loan balances exceeded the value of their homes—and many were
unable to make scheduled mortgage payments.
Fallout from the housing crisis quickly spread to the broader economy
through a complex web of opaque financial instruments and questionable
business practices, including excessive leverage and an overreliance on
short-term debt (Financial Crisis Inquiry Report 2011). Investors pulled
back from risky assets and, during one fateful week in September 2008, the
investment bank Lehman Brothers went out of business, a prominent money
market fund “broke the buck” (meaning that depositors could no longer
count on getting their money back in its entirety, an almost unprecedented
event), and the large insurance firm American International Group (AIG)
teetered on the edge of bankruptcy until the U.S. government provided $85
billion in financial support.
This financial turmoil led to sharp declines in real economic activity. From the third quarter of 2007 through the first quarter of 2009, the
economy lost more than $13 trillion in wealth, nearly one-fifth of the total,
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  93

because of rapidly declining stock and house prices. This was much larger
than the initial decline in wealth at the outset of the Great Depression.2
Falling asset prices reduced the value of collateral and further restricted
the availability of credit and, as credit dried up, many small businesses and
even some large, well-known corporations reported trouble meeting basic
expenses such as payroll. Faced with extraordinary uncertainty about the
economic future, businesses stopped hiring, laid off workers, and shelved
investment plans. As housing and financial wealth plummeted and concerns
over job security mounted, consumers cut back on spending. The effect was
immediate and drastic: in the fourth quarter of 2008, personal consumption expenditures fell by nearly 5 percent and private investment shrunk 31
percent at an annual rate.
Most economic forecasters underestimated the magnitude of the toll
these shocks would take on the economy, in large part because the United
States had not gone through a systemic financial crisis since the Great
Depression. Forecasts made at the time were also subject to considerable
uncertainty about the spillovers to the rest of the world, and about how
the economy would respond to other macroeconomic policy interventions
after the federal funds rate had already hit zero. As shown in Table 3-1, in
December 2008, for example, the Blue Chip panel of economic forecasters
projected that real GDP would fall at a 1.4 percent annual rate in the first half
of 2009, less than half the 2.9 percent annualized rate of decline that actually
occurred. Moreover, the Blue Chip panel of forecasters estimated that the
unemployment rate would rise to 7.7 percent in the second quarter of 2009,
well below the actual rate of 9.3 percent. Other indicators showed similarly
large deteriorations relative to forecasts.

Initial Policy Responses
As the economy slid into recession, Congress and the Bush
Administration enacted the Economic Stimulus Act of 2008 in February.
They designed the Act to counteract a short recession by providing temporary support to consumer spending, but it was not sufficient to reverse
the emerging distress and, by design, did not have long-lasting effects. In
fall 2008, as the initially mild recession turned into a full-blown financial
crisis, the U.S. government mounted a coordinated emergency response to
prevent a meltdown of the financial system.3 The Federal Reserve, which had
progressively cut its federal funds target rate several times over the previous
2 See Romer (2011).
3 A comprehensive timeline of the policy actions taken by the U.S. government can be found
on the Federal Reserve Bank of St. Louis website http://timeline.stlouisfed.org/

94  |  Chapter 3

Table 3–1
Forecasted and Actual Real GDP Growth and Unemployment Rate
Real GDP Growtha

Unemployment Rate

Blue Chipb

Survey of
Professional
Forecastersc

Actual

Blue Chip

Survey of
Professional
Forecasters

Actual

2008:Q4

–4.1

–2.9

–8.3

6.7

6.6

6.9

2009:Q1

–2.4

–1.1

–5.4

7.3

7.0

8.3

2009:Q2

–0.4

0.8

–0.4

7.7

7.4

9.3

Note: a. Percent change from prior quarter at an annualized rate.
b. Blue Chip forecasts for both GDP and Unemployment were reported on December 10, 2008.
c. Survey of Professional Forecasters forecasts for both GDP and Unemployment were reported on November 17,
2008.
Source: Blue Chip Economic Indicators; Survey of Professional Forecasters; Bureau of Labor Statistics, Current
Population Survey; Bureau of Economic Analysis, National Income and Product Accounts.

year, lowered the rate still further in December 2008 to near zero, where it
remains to this day.
To prevent runs on banks and other financial institutions, the Treasury
Department established a temporary guarantee program for money market
mutual funds and the Federal Deposit Insurance Corporation expanded its
guarantee on bank deposits and debt. The Bush Administration proposed
and Congress approved the Troubled Asset Relief Program (TARP), providing up to $700 billion to stabilize troubled banks, automakers, insurance
companies, secondary markets for consumer and small business loans, and
the housing sector.4
These early policy responses proved fundamental to rescuing the global
financial system. They helped repair the balance sheets of both financial and
non-financial institutions, restored investor confidence, and restored the
flow of credit to struggling businesses and families. Nevertheless, the economy continued to deteriorate, and aggregate demand remained depressed.
With the traditional tool of monetary policy, the federal funds rate, reaching
its lower bound of zero, conventional countercyclical monetary policy could
go no further, and the Federal Reserve ultimately opted for additional, nonstandard measures.

An Overview of the Recovery Act
and Subsequent Jobs Measures
Amid very real concerns about a substantial and protracted fall in
GDP accompanied by persistent elevated unemployment, the incoming
4 The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) later
reduced that amount to $475 billion. A detailed description of the TARP can be found on the
Treasury website http://www.treasury.gov/initiatives/financial-stability/Pages/default.aspx

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  95

Obama Administration and the 111th U.S. Congress took immediate action.
In December 2008, the President-Elect and the transition team proposed the
overall scope and elements of what they called the American Recovery and
Reinvestment Act. Just days after the President’s inauguration, on January
26, 2009, House Appropriations Committee Chair David Obey introduced
H.R. 1 with the same name on the floor of the U.S. House of Representatives.
The legislation passed the House and Senate soon afterwards. By February
13, both houses of Congress agreed to a compromise measure, which the
President signed into law on February 17, 2009.

The Recovery Act
In early 2008, before the Nation realized the full extent of the economic challenge, fiscal expansion policy was guided by the “3T’s” advocated
by Summers (2007), Sperling (2007), and Elmendorf and Furman (2008):
timely, targeted, and temporary. By the end of 2008, however, it was clear
that the recession had turned into a major financial crisis and that a new
approach was needed, what Former Treasury Secretary Lawrence Summers
called “speedy, substantial, and sustained.”5
Several principles guided the new Administration’s policymaking.
First, the fiscal effort was to be implemented speedily, unlike previous
incoming presidents’ economic programs, which were generally not passed
until they were six months or more into office. Second, it should be substantial, given the very large scope of the economic problem. Finally, it should
be a sustained effort that would not only have significant spend-out over the
first two years, but would continue some temporary support thereafter. The
new approach would require a mix of instruments, with some being faster
to spend-out, such as tax cuts and other temporary assistance that put cash
in the hands of households who immediately needed it. Other components
would be more lagged but have larger cumulative countercyclical impacts
and greater longer-run benefits, such as investments in infrastructure and
innovation. In all cases, however, the measures would end and would not
have long-term impacts on the Federal Government’s primary budget
deficit.6
Goals of the Recovery Act. Overall, this approach was embodied in the
stated goals of the Recovery Act, as written into the legislation:
(1) To preserve and create jobs and promote economic recovery;
(2) To assist those most impacted by the recession;
(3) To provide investments needed to increase economic efficiency by
spurring technological advances in science and health;
5 Speech at the Wall Street Journal CEO Council conference in Washington, DC, Nov 19, 2008.
6 The primary deficit excludes interest payments on the national debt.

96  |  Chapter 3

(4) To invest in transportation, environmental protection, and other
infrastructure that will provide long-term economic benefits;
(5) To stabilize State and local government budgets, in order to minimize and avoid reductions in essential services and counterproductive State
and local tax increases.
Scale of the Recovery Act. At passage, CBO estimated that the
Recovery Act would cost $787 billion, although this estimate would increase
as the full magnitude of the recession became apparent. The most recent
CBO estimates show that the fiscal support from the Recovery Act will
total $832 billion through 2019.7 Of this total, $69 billion was allocated to a
routine set of patches for the Alternative Minimum Tax (AMT). This part
of the Act, a continuation of a long-standing practice, is best thought of as
ongoing fiscal policy, not as a temporary fiscal impulse designed specifically
to counter the effects of an economic recession. Excluding the AMT patch,
the Recovery Act provided a total fiscal impulse of $763 billion.
Composition of the Recovery Act. The initial cost projections of the
Recovery Act showed the law would be fairly evenly distributed across tax
cuts ($212 billion), expansions to mandatory programs such as Medicaid
and unemployment benefits ($296 billion), and discretionary spending
($279 billion) in areas ranging from direct assistance to individuals to
investments in infrastructure, education, job training, energy, and health
information technology. More specifically, Figure 3-1 shows how Recovery
Act policies excluding the AMT patch can be divided into five functional
categories: individual tax cuts, business tax incentives, State fiscal relief, aid
to directly impacted individuals, and public investments.8
Timing of the Recovery Act Spend-Out. The Nation felt the early
effects of the Recovery Act almost immediately, as enhanced Medicaid payments started to flow to states on March 13, 2009 and individual income tax
withholdings were reduced by April 1, 2009. As of the third quarter of 2009,
roughly one-quarter of all spending and tax cuts had occurred, with another
half spread across the four quarters after that, roughly consistent with CBO
projections as of 2009. By September 30, 2013, the Federal Government had
disbursed $805 billion on Recovery Act programs, as shown in Table 3-2.
As shown in Table 3-3, individual tax cuts, aid to States, and aid to
individuals directly affected by the recession were among the first Recovery
7 CBO’s original estimate of the cost of the Recovery Act, $787 billion (CBO 2009b), was
revised to $862 billion (CBO 2010a), then to $814 billion (CBO, 2010b), $821 billion (CBO
2011a), $831 billion (CBO 2012a), $830 billion (CBO 2013a), and most recently to $832 billion
(CBO 2014). The estimates evolved because economic conditions deteriorated more than had
been assumed in earlier projections, resulting in higher-than-expected use of certain assistance
programs.
8 Additional detail on these components of the Recovery Act can be found in Appendix 1.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  97

Figure 3-1
Recovery Act Programs by Functional Categories

Public Investment
Outlays
(37%)

Aid to Directly
Impacted
Individuals
(15%)

Individual Tax Cuts
(25%)
Business Tax
Incentives
(4%)
State Fiscal Relief
(19%)

Note: Percentages may not add to 100 due to rounding. Data does not include AMT Relief.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the
Treasury, Office of Tax Analysis, based on the FY2013 Mid-Session Review.

Act programs to take effect, providing the largest initial boost to spending
in fiscal year 2009. Each of these categories tapered after 2010, with only a
small amount of outlays in 2012 and 2013, while public investment outlays
now constitute the bulk of continuing Recovery Act expenditures.
Accountability, Transparency, and Oversight. In keeping with the
Administration’s commitment to the highest standards of accountability,
transparency, and oversight, the Recovery Act took unprecedented steps to
track and report the use of Federal funds and to prevent waste, fraud, and
abuse. The Act established a Recovery Accountability Transparency Board
comprised of an independent director and 12 agency inspectors general, as
well as a Recovery Implementation Office that reported directly to the Vice
President. Recipients (including vendors, nonprofit organizations, and State
and local governments) were required to report regularly to the Board on
their use of funds and the number of jobs created or saved.9
All of the information received from agencies and recipients has been
posted on a website (www.recovery.gov). Users can sort and display data
on funding in different ways (by category of funding, by agency, by state),
making it easy to obtain and analyze information. The website also offers
the opportunity for the public to report fraud or waste. Reported instances
9 Title XV, Section 1512

98  |  Chapter 3

Table 3–2
An Overview of Recovery Act Fiscal Impact
Billions of Dollars, Fiscal Years
2009

2010

2011

2012

2013

Total
Through
2013

Outlays

110.7

197.1

112.7

56.8

35.0

512.4

Obligations

256.3

196.1

41.2

21.8

18.5

533.8

Tax Reductions

69.8

188.7

37.2

–5.4

1.9

292.2

Sum of Outlays and
Tax Reductions

180.5

385.8

149.9

51.4

37.0

804.6

Note: Items may not add to total due to rounding.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury,
Office of Tax Analysis based on the FY2013 Mid-Session Review.

Table 3–3
Recovery Act Programs by Functional Categories
Billions of Dollars, Fiscal Years
2011

2012

2013

Total
Through
2013

91.3

46.6

0.4

0.4

181.7

69.6

–14.4

0.0

0.0

69.0

23.1

18.2

–5.9

–3.7

–2.9

28.8

State Fiscal Relief

43.8

63.3

26.0

6.0

4.0

143.0

Aid to Directly Impacted
Individuals

31.8

49.5

15.5

8.8

5.9

111.5

Public Investment Outlays

25.1

94.0

82.0

39.9

29.6

270.5

Total

180.5

385.8

149.9

51.4

37.0

804.6

2009

2010

Individual Tax Cuts

42.9

AMT Relief

13.8

Business Tax Incentives

Note: Items may not add to total due to rounding.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury,
Office of Tax Analysis based on the FY2013 Mid-Session Review.

of waste, fraud, and abuse remain low—at less than 1 percent of all grant
awards.

Subsequent Jobs Measures
While the Recovery Act was the first and largest fiscal action undertaken after the financial crisis to create jobs and strengthen the economy,
many subsequent actions extended, expanded, and built on the Recovery Act.
Parts of the Recovery Act were extended to address the continuing needs of
the economy, including Emergency Unemployment Compensation, accelerated depreciation of business investment for tax purposes (that is, “bonus
depreciation”), measures for teacher jobs, and aid to states for Medicaid. In

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  99

other cases, new measures expanded on elements of the Recovery Act, such
as the temporary payroll tax cut in 2011 and 2012, which was nearly 50 percent larger than the Making Work Pay credit it replaced, and an even greater
allowance for businesses to write off the cost of investments when computing their tax liability (that is, “expensing”). The following measures built on
the goals of the Recovery Act and are counted as part of the fiscal impulse
in the analysis that follows: the cash-for-clunkers program enacted in summer 2009, an expanded homebuyer tax credit and business tax incentives in
fall 2009, the HIRE Act tax credit and additional infrastructure investment
incentives in March 2010, a small business tax cut and credit bill in fall
2010, veterans hiring incentives in fall 2011, plus the additional payroll tax
cut extensions and unemployment insurance extensions passed in 2011 and
2012. All told, these subsequent jobs measures, listed in Table 3-4, provided
an additional $674 billion in countercyclical fiscal support through the end
of 2012. This total excludes routine or expected policies such as continuing
the 2001 and 2003 tax cuts, passing so-called “tax extenders” to address
regularly expiring tax provisions, and fixing Medicare’s Sustainable Growth
Rate formula.10
Of the $674 billion in fiscal support following the Recovery Act, 31
percent was accounted for by the payroll tax cut from 2011 to 2012, 24
percent was accounted for by extended unemployment insurance, and the
remainder included a variety of actions such as relief for States and tax
incentives for businesses. Figure 3-2 shows a breakout of the policies of the
Recovery Act and the subsequent jobs legislation.
In addition, the President proposed further measures for the economy
that were not passed by Congress, most notably the American Jobs Act,
which was proposed in September 2011 and would have provided additional
investments—totaling $447 billion—in everything from infrastructure to
teacher jobs to a robust tax credit for small business hiring.11

Automatic Countercylical Measures
In addition to Obama Administration policies, previously enacted
laws have built-in provisions that allow for automatic support when economic conditions worsen. For example, personal income tax payments
decline when income declines, and spending on unemployment insurance picks up as more individuals struggle to find work. These automatic
10 This category includes items like the Research and Experimentation tax credit, the tax
deduction for State and local sales taxes for States without income taxes, and numerous other
tax provisions that have been routinely extended as a group in the past. Going forward, the
President’s budget is proposing that all tax extenders are either made permanent and paid for
or allowed to expire.
11 See http://www.whitehouse.gov/the-press-office/2011/09/08/fact-sheet-american-jobs-act

100  |  Chapter 3

Table 3–4
Fiscal Support for the Economy Enacted After the Recovery Act
Billions of Dollars
2009–12

2009–19

Worker, Homeownership, and Business Assistance Act (HR 3548)

35

24

Supplemental Appropriations Act of 2009 (HR 2346)
(Cash for Clunkers)

3

3

Defense Appropriations Act of 2010 (HR 3326) (Unemployment
Insurance and COBRA)

18

18

Enacted 2009

Enacted 2010
Temporary Extension Act of 2010 (HR 4691)

9

9

Hiring Incentives to Restore Employment Act (HR 2847)

13

15

Continuing Extension Act of 2010 (HR 4851)

16

16

Unemployment Compensation Act of 2010 (HR 4213)

33

34

FAA Safety Improvement Act (HR 1586) (Education Jobs/ FMAP
Extension)

26

12

Small Business Jobs Act (HR 5297)

68

10

Tax Relief, Unemployment Insurance Reauthorization,
and Job Creation Act (HR 4853)

309

237

Enacted 2011
Temporary Payroll Tax Cut Continuation Act (HR 3765)

28

29

VOW to Hire Heroes Act (HR 674)

0

–0

Middle Class Tax Relief and Job Creation Act of 2012 (HR 3630)

98

123

American Taxpayer Relief Act of 2012 (HR 8)

17

178

674

709

Enacted 2012

Total

Note: All measures use prospective CBO cost estimates for 2009–19. Routine tax extenders have been removed
from the cost estimates. Column 1 contains data through the end of calendar year 2012, while Column 2 contains
data through the end of fiscal year 2019.
Source: Congressional Budget Office; Joint Committee on Taxation.

responses—known as “automatic stabilizers”—can help moderate business
cycles (as shown for instance by Auerbach and Feenberg (2000) and Follette
and Lutz (2010)) in addition to alleviating the human costs of economic
downturns.
As has been the case over the last several decades, automatic stabilizers also played a significant role during the most recent recession and
recovery. Although CBO (2014) estimated that most fiscal expansion came
from enacted legislation or discretionary fiscal policy, automatic stabilizers
accounted for about one-quarter of the countercyclical fiscal expansion
that occurred in FY 2009, and a much larger fraction thereafter as shown in
Figure 3-3.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  101

Figure 3-2
Recovery Act and Subsequent Fiscal Measures
by Functional Category

Public Investment
Outlays (18%)

Individual Tax Cuts
(31%)

Aid to Directly
Impacted
Individuals (20%)
Business Tax
Incentives
(19%)

State Fiscal
Relief (12%)

Note: Percentages may not add to 100 due to rounding. Data does not include AMT Relief.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of
the Treasury, Office of Tax Analysis, based on the FY2013 Mid-Session Review; Congressional
Budget Office.

Figure 3-3
Automatic Stabilizers and the Budget Balance, 20092013

Federal Budget Deficit as a Percent of GDP
12
10

Automatic
Stabilizers

8

6

Structural
Component

4
2
0

2009

2010

2011

2012

2013

Source: Bureau of Economic Analysis, National Income and Product Accounts; Congressional Budget
Office, The Budget and Economic Outlook: 2014 to 2024.

102  |  Chapter 3

Percent
6.0

Figure 3-4
Fiscal Expansion as a Percentage of GDP
Recovery Act, subsequent
fiscal measures, and
automatic stabilizers

5.0
4.0
3.0

Recovery Act and subesquent
fiscal measures

2.0
1.0
0.0
2009

Recovery Act

2010

2011

2012

Note: Data is displayed in calendar year terms for all series.
Source: Congressional Budget Office, The Budget and Economic Outlook: 2014 to 2024; Office of
Management and Budget; Bureau of Economic Analysis, National Income and Product Accounts.

Total Fiscal Response
All told, the Great Recession triggered a substantial fiscal response.
Figure 3-4 illustrates the scale of the Recovery Act and of the other major
fiscal measures implemented by the Administration. As noted earlier, fiscal
policy represented only one part of the Administration’s broader economic
strategy to foster recovery and protect households, as described more fully
in Box 3-1.

Near-Term Macroeconomic Effects of the
Recovery Act and Subsequent Fiscal Legislation
This chapter reviews the range of evidence on the effect of the
Recovery Act. In particular, it shows that a wide range of approaches—
including model-based estimates by CEA, CBO and private forecasters,
cross-state evidence and international evidence—all find that the Recovery
Act had a large positive impact on employment and output.
Overall, CEA estimates that the Recovery Act saved or created about
6 million job-years (where a job-year is the equivalent of one full time job
for one year) through 2012 and raised GDP by between 2 and 2.5 percent in
FY 2010 and part of FY 2011. These estimates are consistent with estimates

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  103

Box 3-1: Other Administration Policy
Responses to the Economic Crisis
The Recovery Act was part a comprehensive policy response to
the economic crisis, one that included stabilizing the financial system,
helping responsible homeowners avoid foreclosure, and aiding small
businesses. Highlighted here are some of the other key non-fiscal programs (not counting the important steps taken independently by the
Federal Reserve).
Housing. The Administration took several steps to strengthen
the housing market. The most important initiative, the Making Home
Affordable Program (MHA), provided several ways to help struggling
homeowners avoid foreclosure. A detailed description is available at
www.makinghomeaffordable.gov. The Home Affordable Modification
Program was the cornerstone of the initiative, allowing eligible homeowners to reduce their monthly mortgage payments through loan modifications. Among the many other MHA programs, the Home Affordable
Refinancing Program helped homeowners who, because of plummeting
home prices, were “underwater” on their mortgages or in danger of
becoming so, allowing them to refinance at a lower interest rate. The
MHA also committed funds to help struggling homeowners in hard-hit
areas (under the Hardest Hit Fund).
In addition, the Administration created the Consumer Financial
Protection Bureau to establish safe mortgage standards to protect homebuyers and homeowners, among other purposes. The Administration
also helped negotiate the National Mortgage Servicing Settlement with
the largest mortgage servicers. While the housing market continues to
heal, housing is in much better shape overall than it was just a few years
ago. Home prices are about 15 percent higher than they were at the end
of 2011, and sales of new and existing homes are higher than at the end
of 2011 while the number of seriously delinquent mortgages is now at its
lowest level since 2008.
Auto Industry. Recognizing that a collapse of the auto industry
would have resulted in huge job losses and the devastation of many communities, the Administration, under the Troubled Asset Relief Program
(TARP), provided financial support to auto companies to keep them
afloat. The Administration committed additional assistance to Chrysler
and General Motors, while at the same time working on a comprehensive
restructuring of these companies. Since then, these auto companies
have become profitable again, and auto sales have been trending up
since 2009. The auto industry has added more than 420,000 jobs since
June 2009. In December 2013, the Treasury sold its remaining shares of
General Motors.

104  |  Chapter 3

Financial Industry. TARP and other programs implemented during the height of the financial crisis helped prevent a meltdown of the
global financial system, but did not solve many longer-running, more
structural problems. The Administration pushed for an overhaul of
the financial regulatory system, and its proposals eventually led to the
Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010.
Among its many provisions, the Dodd-Frank Act required stress tests
to assess the health of financial institutions, provided tools for orderly
liquidations of large financial firms, and increased the transparency of
derivatives markets. As a result of these actions, large banks are now
much better capitalized and credit flows have resumed. While some of
the Dodd-Frank Act’s provisions are still being implemented and much
work remains to be done, the financial system has become less vulnerable and families are better protected when making important financial
decisions.
Small Businesses. Small business struggled under the weight
of weak consumer demand and tight credit in the recession, and the
Administration provided support in several ways. Specifically, the
Administration extended the guarantees and the availability of Small
Business Administration loans and created new programs such as the
Small Business Lending Funds and the State Small Business Credit
Initiative. It also helped small businesses indirectly by providing TARP
funds to small and large banks across the country. Bank credit to small
businesses, which had contracted sharply during the recession, has been
expanding since 2011.

made by CBO (2013a) and by independent academic studies, which use a
variety of methodologies and data sources. Adding the estimated effect of
subsequent fiscal policy measures, CEA model finds that the combined effect
of these actions increased GDP above what it otherwise would have been
by more than 2 percent a year for three years, and created or saved about
9 million job-years through 2012. Moreover, research on economic growth
generally finds that these types of benefits have a long-lasting impact on
the economy even after the initial policy impetus has expired. This is even
more true when the policy itself included significant measures for long-term
growth, as described later in this chapter.

Model-Based Estimates of the Macroeconomic Effects of the
Recovery Act and Subsequent Fiscal Legislation
Evaluating effects of fiscal policy in general, and the Recovery Act in
particular, is challenging for several reasons. Appendix 2 describes these

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  105

challenges, and how economists have addressed them, in greater detail. A
key issue is that estimating effects entails comparing what actually happened
with what might have happened (what economists call the “counterfactual”).
However, because counterfactual outcomes are not actually observed, other
methods are needed.
Estimating the Short-run Macroeconomic Effect of Fiscal Policy. A
key concept for estimating the macroeconomic effects of fiscal policy is what
economists call the fiscal multiplier. The fiscal expenditure multiplier is the
change in GDP resulting from a $1 increase in government expenditures,
and the tax multiplier is the change in GDP resulting from a $1 decrease
in taxes. Because a $1 increase in spending or decrease in taxes has ripple
effects in subsequent transactions as it passes through the broader economy,
theory suggests that the fiscal multiplier may be greater than one—a $1
increase in spending or reduction in taxes may support an increase in output
of more than $1.
The standard theory of fiscal policy in a recession holds that when
government demand goes up, firms hire workers and raise production,
which boosts employment, income, and GDP. The initial effect is amplified
as workers spend additional income, and businesses purchase more raw
materials and make investments to meet increased demand. In its most
basic form, the government spending multiplier is the sum of the first-round
direct effect of spending on GDP, the second-round effect with consumption
by those paid for providing goods and services, and the subsequent-round
effects. In this model, the multiplier effect depends on the marginal propensity to consume (MPC)—the fraction of an additional dollar of income that
is spent rather than saved.12 Because the MPC is thought to be large, especially in a recession when individuals face problems borrowing, models can
generate multipliers much higher than one. Tax cuts also increase individual
income, but the multiplier effect on overall output is generally thought to be
slightly less than it is for government expenditures. Because the individual
receiving the tax cut saves part of it, the first-round effect on overall spending is smaller, making the subsequent ripple effects smaller as well. Thus, the
basic multiplier for a tax cut is less than the government spending multiplier;
specifically, the tax multiplier is the spending multiplier times the MPC.
The model is a useful conceptual starting point, but it makes many
simplifications. Appendix 2 to this chapter reviews recent theoretical
research on the effects of fiscal policy, especially in a deep recession. This
research suggests that, in normal times, fiscal multipliers can be small both
because consumers save a substantial fraction of a temporary fiscal measure
and because monetary policy tends to counteract the fiscal measure in an
12 This basic multiplier thus equals 1 + MPC + MPC2 + … = 1/(1 – MPC).

106  |  Chapter 3

attempt to maintain stable inflation. In a severe recession, however, especially when monetary policy is constrained by the fact that interest rates
cannot drop below zero (the zero lower bound), fiscal multipliers can be
much larger. Taking further into account the fact that long-term unemployment can lead to transitions out of the labor force, with a resulting long-term
effect of depressing output and employment (an effect referred to as “hysteresis”), multipliers can be larger yet for fiscal policies that support aggregate
demand and reduce the average duration of unemployment.
Numerical estimates of fiscal multipliers are typically computed using
historical data on fiscal interventions and macroeconomic outcomes, and
Appendix 2 also discusses the recent empirical research on this topic. This
empirical work provides estimates of multipliers for different types of fiscal
interventions (government spending and individual income tax cuts). Once
estimated, the resulting multipliers can be used to estimate the macroeconomic effect of the Recovery Act; that is, to compare what happened under
the Recovery Act with what likely would have happened in its absence.
CEA’s and CBO’s Estimates of the Recovery Act. To estimate the
effect of the Recovery Act on GDP, CEA applied a different fiscal multiplier
to each component, and then aggregated the effects of each component to
arrive at the overall GDP effect. For government spending (corresponding
to public investment outlays and income and support payments) and for
tax cuts, CEA used multipliers derived from the empirical estimates of the
spending and tax multipliers discussed in Appendix 2. For other components
of the Act, such as State and local fiscal relief, CEA used a multiplier equal
to a weighted average of one or both of the tax and spending multipliers.13
The CBO used a similar approach in its quarterly reports on the effects
of the Recovery Act (although their estimates include the impact of AMT
relief and so are not completely comparable to CEA estimates).14 Because of
the range of estimates of multipliers in the economic literature, CBO provided an upper and a lower bound for the fiscal multipliers on the various
components of the Act. As shown in Table 3-5, CEA multipliers are within
the range suggested by CBO (2013a).
The multipliers presented here indicate that the Recovery Act
had a large effect on output. As shown in Figure 3-5, the Recovery Act
quickly raised the level of GDP in the first half of 2009, jump-starting the
13 For State and local fiscal relief, CEA assumed that 60 percent of the transfer is used to avoid
spending reductions, and 30 percent is used to avoid tax increases. One-time tax rebates and
one-time payments to seniors, veterans, and disabled are assumed to have half of the effects of
conventional tax cuts. The effect of business tax incentives is very uncertain. Conservatively,
the multiplier to this component is set equal to 1/12 of the spending multiplier. See CEA
(2009a).
14 CBO’s methodology is described in Reichling and Whalen (2012).

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  107

Table 3–5
Estimated Output Multipliers for Different Types of Fiscal Support
CEA

CBO Low

CBO High

Public Investment Outlaysa

1.5

0.5

2.5

State and Local Fiscal Relief

1.1

0.4

1.8

Income and Support Paymentsb

1.5

0.4

2.1

One-time Payments to Retirees

0.4

0.2

1.0

Tax Cuts to Individuals

0.8

0.3

1.5

Business Tax Incentives

0.1

0.0

0.4

Note: The CEA multipliers show the impact of a permanent change in the component of 1% of GDP after 6 quarters, or, equivalently, the cumulative impact of a one-time change of 1% of GDP over 6 quarters. The CBO multiplers
show the cumulative impact of a one-time change of 1% of GDP over several quarters.
a. Includes transfer payments to state and local government for infrastructure and tax incentives to businesses
directly tied to certain types of spending.
b. Includes such programs as unemployment compensation, COBRA, and SNAP
Source: Congressional Budget Office, Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output from October 2012 Through December 2012; CEA Calculations.

recovery. According to these estimates, the Act raised GDP by 2 to 2.5 percent between the fourth quarter of 2009 and the second quarter of 2011, and
it continued to exert a positive effect even as it was winding down in 2012.
These numbers are almost entirely within the range of those implied by the
CBO analysis, with the exception being a few quarters in late 2012 and early
2013, when CEA estimate is slightly higher.
Using the historical relation between increases in GDP and employment, CEA and CBO also estimated the number of jobs generated by
the Recovery Act. According to CEA model, the Recovery Act increased
employment by more than 2.3 million in 2010 alone, and continued to have
substantial effects into 2012, as demonstrated in Figure 3-6. Cumulating
these gains through the end of FY 2013, the Recovery Act is estimated to
have generated about 6.4 million job-years. These estimates are also within
the range of CBO’s upper- and lower-range estimates of 1.6 to 8.3 million
job-years.
CEA Estimates of the Recovery Act and Subsequent Fiscal Measures
Combined. The combined effect of the Recovery Act and the subsequent
countercyclical fiscal legislation is substantially larger and longer lasting
than the effect of the Recovery Act alone.15 The Recovery Act represents
only about half of total fiscal support for the economy from the beginning
of 2009 through the fourth quarter of 2012. Moreover, as shown in Figures
3-7 and 3-8, the bulk of the effects of the other fiscal measures occurred as
15 CEA’s estimates of the effects of the subsequent fiscal measures are based on CBO’s initial
cost estimates, not actual spending. CEA assigned each of the subsequent fiscal measures to
the same functional categories that were used to analyze the Recovery Act, and then applied
the corresponding multipliers as discussed above. Quarterly costs were interpolated when only
annual cost estimates were available.

108  |  Chapter 3

Percent
5.0
4.5

Figure 3-5
Estimates of the Effects of the Recovery Act
on the Level of GDP, 2009‒2013
CBO: High

4.0
3.5
3.0
2.5

CEA Model

2013:Q3

2.0
1.5
1.0
0.5
0.0
2009:Q2

CBO: Low
2010:Q2

2011:Q2

2012:Q2

2013:Q2

Source: Congressional Budget Office, Estimated Impact of the American Recovery and Reinvestment
Act on Employment and Economic Output in 2013; CEA calculations.

Millions
4.0
3.5
3.0
2.5

Figure 3-6
Estimates of the Effects of the Recovery
Act on Employment, 2009‒2013
CBO: High

CEA Model
2013:Q3

2.0
1.5
1.0
0.5

0.0
2009:Q2

CBO: Low
2010:Q2

2011:Q2

2012:Q2

2013:Q2

Source: Congressional Budget Office, Estimated Impact of the American Recovery and Reinvestment
Act on Employment and Economic Output in 2013; CEA calculations.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  109

Figure 3-7
Quarterly Effect of the Recovery Act and Subsequent
Fiscal Measures on GDP, 2009–2012

Percent of GDP
3.5

2012:Q4

3.0

Other Fiscal
Measures

2.5

ARRA

2.0
1.5
1.0
0.5
0.0

2009:Q2

2010:Q2

2011:Q2

2012:Q2

Source: Bureau of Economic Analysis, National Income and Product Accounts; Congressional Budget
Office; CEA calculations.

Millions
3.5

Figure 3-8
Quarterly Effect of the Recovery Act and Subsequent
Fiscal Measures on Employment, 2009–2012
Other Fiscal
Measures

3.0

2012:Q4

2.5
2.0

ARRA

1.5
1.0
0.5
0.0

2009:Q2

2010:Q2

2011:Q2

2012:Q2

Source: Bureau of Economic Analysis, National Income and Product Accounts; Congressional Budget
Office; CEA calculations.

110  |  Chapter 3

the Recovery Act was phasing down. These other measures thus served to
sustain the recovery as effects of the Recovery Act waned. CEA multiplier
model indicates that by themselves these additional measures increased the
level of GDP by between 1.0 and 1.5 percent each quarter from mid-2011
through the end of calendar year 2012. Altogether, summing up the effects
for all quarters through the end of calendar year 2012, the Recovery Act
and subsequent fiscal measures raised GDP by an average of more than 2.4
percent of GDP annually—totaling a cumulative amount equal to about 9.5
percent of fourth quarter 2008 GDP.
The contribution of all fiscal measures to employment is equally
substantial. Other fiscal measures beyond the Recovery Act are estimated to
have raised employment by 2.8 million job-years, cumulatively, through the
end of calendar year 2012. Adding these jobs to those created or saved by the
Recovery Act, the combined countercyclical fiscal measures created or saved
more than 2.3 million jobs a year through the end of 2012—or 8.8 million
job-years in total over the entire period.
Estimates from Private Forecasters. Private forecasters and domestic
and international institutions have used large-scale macroeconomic models,
mostly to estimate the effects of either the Recovery Act by itself or other
policies in isolation. The models used by these individuals and organizations
generally employ a similar multiplier-type analysis as is found in CEA and
CBO work, although they vary considerably in their structure and underlying assumptions. Although no outside estimates of the total impact of all the
fiscal measures are available, Table 3-6 displays the estimates of the impact
of the Recovery Act offered by several leading private-sector forecasters
before the Act was fully implemented. Despite the differences in the models,
these private-sector forecasters all estimated that the Recovery Act would
raise GDP substantially from 2009 to 2011, including a boost to GDP of
between 2.0 and 3.4 percent in 2010.
Taking a broader view that incorporates fiscal measures in addition to
the Recovery Act, Blinder and Zandi (2010) estimate the effect of the fiscal
policies enacted through 2009 (the Economic Stimulus Act, the Recovery
Act, cash for clunkers, the unemployment insurance benefits extensions of
2009). They find that these policies raised the level of GDP in 2009 by 3.4
percent in the third quarter and by 4.3 percent in the fourth quarter.

Cross-State Evidence
A different approach to estimating the effect of fiscal policy is to use
variation in spending across states. As noted earlier, estimates of the effects
of macroeconomic policy are inherently difficult because the counterfactual
outcome is not observed. One way economists have attempted to address
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  111

Table 3–6
Estimates of the Effects of the Recovery Act on the Level of GDP
Percent
2009

2010

2011

2012

2013

CEA: Model Approach

+1.1

+2.4

+1.8

+0.8

+0.3

CBO: Low

+0.4

+0.7

+0.4

+0.1

+0.1

CBO: High

+1.7

+4.1

+2.3

+0.8

+0.3

Goldman Sachs

+0.9

+2.3

+1.3

—

—

HIS Global Insight

+0.8

+2.2

+1.6

+0.6

—

James Glassman, JP Morgan Chase

+1.4

+3.4

+1.7

0.0

—

Macroeconomic Advisers

+0.7

+2.0

+2.1

+1.1

—

Mark Zandi, Moody’s Economy.com

+1.1

+2.6

+1.7

+0.4

—

Note: Firm estimates were obtained from and confirmed by each firm or forecaster, and collected in CEA’s Ninth
Quarterly Report.
Sources: Congressional Budget Office, Estimated Impact of the American Recovery and Reinvestment Act on
Employment and Economic Output from October 2012 Through December 2012; CEA Ninth Quarterly Report;
CEA Calculations.

this difficulty is to isolate a component of the Act that was implemented in a
random or quasi-random manner across different states, mimicking a randomized controlled trial used for research in other disciplines like medicine.
If some states received more Recovery Act funds than others for reasons
unrelated to their economic needs, then this portion of the funds can allow
for an independent, unbiased evaluation of the effects, much like two groups
of individuals participating in a drug trial that receive different dosages of
the same medicine.
A notable drawback of using State-level data, however, is that this
approach estimates local, not economy-wide, multipliers. These local
multipliers do not incorporate out-of-state spending effects, nor do they
incorporate the general equilibrium and monetary policy feedback effects
that are the focus of much of the theoretical work discussed above and in
the Appendix. 16
One portion of the Act that was distributed independently of states’
immediate economic needs was the additional grants to states for Medicaid.
Under the Act, states received a 6.2 percentage point increase in their
expected Federal reimbursement rate (the Federal Medical Assistance
Percentage).17 This increase was worth more to states that spent more per
capita on Medicaid before the recession (in FY 2007). To the extent that
16 See Nakamura and Steinsson (forthcoming) and Farhi and Werning (2012) for formal
discussion of the relationship between local multipliers and the economy-wide multiplier.
17 In addition, states were “held harmless” from planned reductions in FMAP rates due to
personal income growth prior to the recession and they received an additional increment in
the FMAP linked to local unemployment. The analysis presented here relies only on the 6.2
percent (non-cyclical) increase.

112  |  Chapter 3

Figure 3-9
Change in Nonfarm Employment

Percent of State Population, January-July 2009
1
0.5
0
-0.5
-1

-1.5
-2
-2.5
-3

0

25
50
75
100
125
150
Estimated non-cyclical Recovery Act Medicaid Payouts in 2007, per capita ($)

Note: Size of circle is proportional to 2008 state population.
Source: Centers for Medicare and Medicaid Services, Data Compendium; Bureau of Labor Statistics,
Current Employment Statistics; CEA Calculations.

the relative severity of the recession at the state level was unrelated to its
previous level of per-capita Medicaid spending, this portion of funds might
be thought of “as if” randomly assigned. In other words, the spending
was effectively independent of the strength or weakness of the state-level
economy once the recession hit. As Figure 3-9 shows, states that received
more funds stemming from this non-cyclical part of the formula tended to
exhibit greater employment gains through the first half of 2009 compared
with states receiving less funds.
Refining this approach, Chodorow-Reich et al. (2012) found that
each additional $100,000 of formula-based Medicaid grants generated an
additional 3.8 job-years, which translates into a $26,000 cost per job. Other
academic papers following a similar approach, but assessing broader measures of Recovery Act spending, reached similar conclusions. For example,
Wilson (2012) estimates a cost per job of about $125,000 for all Recovery
Act spending programs other than those implemented by the Department of
Labor (mostly unemployment insurance). Feyrer and Sacerdote (2011) and
Conley and Dupor (2013) also find positive effects of the Act on employment, although the ranges of effects estimated in both papers include magnitudes similar to those discussed above as well as somewhat smaller effects.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  113

International Comparison
The 2008 crisis reverberated worldwide. In addition to seeing sharp
reductions in output and employment, many countries also experienced
large government budget deficits because of countercyclical fiscal policies
and a fall in tax revenues caused by the recession. These changes in budget
deficits across countries can be used to derive an international estimate of
the impact of fiscal policy. The International Monetary Fund’s (IMF) early
estimates using pre-crisis cross-country data suggested expenditure multipliers averaging 0.5, although with substantial variation across countries.18
However, subsequent research by IMF (2012) and Blanchard and Leigh
(2013) reassessed this earlier work and estimated multipliers substantially
above 1.0 during the crisis, consistent with the discussion earlier in this
chapter about recent fiscal multipliers in the United States.
This international evidence also suggests that the structural reductions in government budget deficits (or “fiscal consolidation”) implemented
by many countries has had a large negative impact on economic activity in
the short run, at least when interest rates are low or at the zero lower bound
and when there is already substantial economic slack. Previous research,
summarized in Alesina and Ardagna (2010), hypothesized that fiscal consolidation can sometimes boost GDP because it increases investors’ confidence
and lowers interest rates. But Blanchard and Leigh’s (2013) results, as well
as findings by Perotti (2011) and Guajardo, Leigh, and Pescatori (forthcoming), point instead to significant short-run costs of deficit reductions and
suggest a more gradual strategy of fiscal consolidation, as explained for
instance in Blanchard, Dell’Ariccia, and Mauro (2010).
It is notable that the United States is one of only two of the 12 countries that experienced systemic financial crises in 2007 and 2008 but have
seen real GDP per working-age person return to pre-crisis levels (see Box
3-2). Although this does not provide any specific evidence on the effect of
U.S. fiscal measures, it is consistent with the proposition that the full set of
U.S. policy interventions made a sizable difference in reversing the downward spiral of falling employment and output.

Benchmarking the Economy’s Performance Since 2009
While the bulk of the available evidence indicates that the Recovery
Act and subsequent fiscal legislation helped avert what might have become a
second Great Depression and paved the way for stronger economic growth,
many households continue to struggle with the after-effects of the recession. In addition, from a macroeconomic perspective, the average rate of
18 See for example Ilzetzki, Mendoza, and Vegh (2011).

114  |  Chapter 3

real GDP growth in this recovery (2.4 percent a year) has been slower than
many would have liked. Some critics have argued that this slower growth is
evidence that economic policymaking has gone awry, and that the interventions undertaken since 2008 have had unintended detrimental consequences
on growth. Taylor (forthcoming) argues that fiscal policy not only failed to
help but actually hurt the economy.
As discussed earlier, it is impossible to infer the causal impact of a set
of policies from the observed outcomes because these observed outcomes
do not reveal what would have happened absent the policy interventions.
The research that attempts to answer that counterfactual conclusion using
a variety of different methods has generally come to the conclusion that
the Recovery Act and subsequent measures had a large positive impact on
growth and employment.
In particular, claims based on the recovery are often based in part on
a misleading apples-to-oranges comparison to past growth and also fail to
take into account the key features of the recession and recovery. First, the
economy’s potential growth rate is slower now than it has been in previous
post-World War II recoveries for long-standing reasons unrelated to the
Great Recession or the policies that followed in its wake. This lower rate of
potential growth reflects several key factors: slowing growth in the workingage population as baby boomers move into retirement, a plateauing of
female labor force participation following several decades of transformative
increases, and a slowdown in productivity growth. CBO (2012a) estimates
that slower growth of potential GDP accounts for about two-thirds of the
difference between observed real GDP growth in the current recovery and
growth on average in the preceding postwar recoveries, an estimate that is in
line with other recent studies (see the 2013 Economic Report of the President
for further discussion).
Second, the economy has encountered a long list of additional
headwinds in recent years. This list includes international events like the
European sovereign debt crisis, the tsunami and nuclear accident in Japan,
and the disruption of Libya’s oil supply. It includes extreme weather like
Hurricane Sandy and the 2012 drought that was described by the U.S.
Department of Agriculture as the “most severe and extensive drought in at
least 25 years.”19 It includes fiscal austerity at the state and local level that
intensified as the Recovery Act began to phase out and has cost hundreds
of thousands of additional job losses even during the expansion period.
And it includes measures like the sequester which CBO estimated took 0.6
percentage point off growth in 2013, the 16-day government shutdown
19 http://www.ers.usda.gov/topics/in-the-news/us-drought-2012-farm-and-food-impacts.aspx#.
Uu1MXfldV5A

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  115

Box 3-2: The U.S. Recovery in Comparative
International and Historical Context
The 2007-09 recession was the most severe recession experienced
by the United States since World War II. The so-called Great Recession
lasted 18 months from the business-cycle peak in December 2007 to the
trough in June 2009, nearly twice the 10-month average length of the
previous post-war recessions. The 2007-09 recession was also the sharpest, with a 4.5 percent peak-to-trough drop in real GDP, compared to an
average decline of 2 percent in previous post-war recessions.
Most importantly, the 2007-09 recession was the only post-war
U.S. recession associated with a systemic financial crisis. Severe financial
crises tend to have long-lasting effects that can stymie an economic
rebound in several ways. First, households burdened by high debt
and losses on their assets can be reluctant to increase spending for an
extended period of time, instead choosing to pay down debt and repair
their finances. Business and residential investment can also be slow to
resume, because over-leveraged banks and other financial institutions
reduce the supply of credit as they reestablish the health of their balance
sheets. When both credit supply and demand are suppressed, low interest rates induced by conventional monetary policy have a more limited
impact than they otherwise would.
The U.S. economy has performed better over the past five years
than would be suggested by the historical record of financial crises.
Although the financial shocks that the United States suffered in 2007 and
2008 were similar to, if not larger than, the shocks that set off the Great
Depression, the outcome was strikingly different. In the recent crisis,
GDP per working-age person returned to its pre-crisis level in about
four years, while it took 11 years in the United States during the Great
Depression and, on average, 10 years in the 13 other countries affected

that the Bureau of Economic Analysis (BEA) estimated directly subtracted
0.3 percentage point from growth in the fourth quarter, and dangerous
brinksmanship around the debt limit in 2011 and 2013.
In addition, the unique after-effects of financial crises discussed in
Box 3-2 have also created substantial challenges for faster growth. The Great
Recession was the first downturn brought about by a systemic financial
crisis in nearly 80 years. Macroeconomic data from the Great Depression
is limited, and models based on more readily available post-World War II
data do not contain any comparable benchmark for the shock that hit the
economy in 2008. Many of these models are still being refined to include
more extensive and detailed linkages between the macroeconomy and the
financial sector, based on lessons learned as a result of the crisis.
116  |  Chapter 3

by systemic crises during the 1930s identified by Reinhart and Rogoff
(2009).
The Figure below compares the performance of the U.S. economy
with that of the other economies hit by the recent financial crisis. Of the
12 countries that suffered systemic financial crises in 2007 and 2008, real
GDP per working-age adult has recovered to its pre-crisis levels in only
the United States and Germany.1
Real GDP Per Working Age Population
in 2007–2008 Banking Crisis Countries, 2007–2013
Index, pre-crisis peak =100

105

Germany

100

Netherlands

Spain

95
90

France
Portugal
Iceland

U.K.

85

U.S.

Italy

Ukraine

80

Greece

75
70

Ireland

0

4

8

12

16

Quarters from peak

20

24

28

Note: U.S. as of 2013:Q4, all others as of 2013:Q3 except Iceland (Q2). Working age population is 1664 for U.S. and 15-64 for all others. Population for Ukraine is interpolated from annual estimates.
Selection of countries based on Reinhart and Rogoff (forthcoming).
Source: Statistical Office of the European Communities; national sources; CEA calculations.

1 For a historical account of financial crises in the United States and abroad, see Reinhart
and Rogoff (2009) and Laeven and Valencia (2012).

All these factors must be taken into account in assessing the economy’s performance in recent years—and understanding what would have
happened without the significant policy actions that are described in this
chapter.

Effects of the Recovery Act in
Providing Relief for Individuals
As noted at the outset of this chapter, the Recovery Act had goals
beyond preserving and creating jobs and promoting economic recovery.
This section evaluates the impact of the Recovery Act in helping those
most affected by the recession weather an extraordinarily trying period.
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  117

Table 3–7
Tax Relief and Income Support in the Recovery Act and Subsequent Measures, 2009‒2012
Billions of Dollars
Subsequent
Legislation

Total

112.2

—

112.2

—

206.8

206.8

Recovery Act
Making Work Pay
Payroll Tax Cut
Other tax relief for individuals and families
EITC third child and marriage penalty

6.0

4.3

10.3

Child tax credit refundability

18.7

12.3

31.0

American Opportunity tax credit

17.8

11.3

29.1

Partial exemption of tax on unemployment benefits

6.5

—

6.5

Sales tax deduction for vehicle purchase

1.3

—

1.3

First-time homebuyer tax credit

4.6

12.0

16.6

Emergency Unemployment Compensation
and Extended Benefits

43.2

160.6

203.8

Additional $25 payment

14.1

—

14.1

Unemployment Insurance Modernization

3.5

—

3.5

COBRA

9.2

9.8

19.1

Supplemental Nutrition Assistance Program

37.6

0.4

38.0

$250 Payment to Seniors, Veterans, and the Disabled

13.8

—

13.8

Wounded Warrior Tax Credit

—

0.3

0.3

TANF emergency fund

4.7

—

4.7

293.3

417.8

711.1

Unemployment Insurance

Total

Note: Data consist of cumulative outlays through the end of calendar year 2012. Items may not add to total due to
rounding.
Source: Department of the Treasury, Office of Tax Analysis; Office of Management and Budget, Agency Financial
and Activity Reports; Congressional Budget Office.

The Recovery Act included substantial assistance for middle-class families,
unemployed workers struggling to find a job, and households in poverty or
in danger of slipping into poverty. Many of these measures were extended
or retooled in subsequent legislation. This assistance was partly to help
families maintain their consumption even when income fell and credit dried
up, a phenomenon economists refer to as “consumption smoothing.” But
the support was also motivated by the fact that people would quickly spend
a large fraction of this assistance boosting aggregate demand and, in turn,
job creation. Table 3-7 shows the programs that provided direct assistance
to individuals.

Tax Cuts for Families
The Recovery Act’s income support and individual tax cut provisions
allowed households to maintain their purchasing power through one of the

118  |  Chapter 3

Figure 3-10
Disposable Personal Income With and Without ARRA

Per capita 2009 $, quarterly rate
9,400
9,200

2013:Q1

Disposable Personal
Income Without
ARRA

Actual Personal
Disposable Income

9,000
8,800
8,600
8,400
8,200
2008:Q1

2009:Q1

2010:Q1

2011:Q1

2012:Q1

2013:Q1

Note: Figures deflated using the price index for personal consumption expenditures.
Sources: Bureau of Economic Analysis, Effect of the ARRA on Selected Federal Government
Sector Transactions, National Income and Product Accounts.

worst recessions of the century. The Making Work Pay tax credit in effect
in 2009 and 2010 provided 95 percent of workers with a tax cut worth $400
for a typical single worker and $800 for a typical married couple. Without
these provisions, aggregate real disposable personal income would have been
$354 billion lower than what it actually was in 2009. As shown in Figure
3-10, despite $300 billion in lost private wages and salaries, real disposable
incomes actually grew throughout calendar year 2009 (CEA 2010b), primarily due to tax cuts for families, the largest of which was the Making Work Pay
tax credit in the Recovery Act. The Making Work Pay credit was replaced by
the even-larger payroll tax cut in 2011 and 2012 that provided a tax cut for
all 160 million workers, with $1,000 for the typical worker making $50,000
per year.

Unemployment Insurance
Regular state-based unemployment insurance (UI) programs typically
provide benefits for 26 weeks, but as the average duration of unemployment
rose to record highs in the 2007-09 recession and its wake, additional steps
were needed. The Recovery Act expanded unemployment insurance in several ways. First, the Act provided for a 100 percent Federal contribution to
the Extended Benefits program, which has been in place since 1970 to assist
states that experience especially sharp increases in unemployment but has
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  119

traditionally been jointly financed by Federal and State governments. The
Recovery Act also extended the Emergency Unemployment Compensation
(EUC) program enacted in 2008, which extended the duration of benefits
available under Extended Benefits. It also provided an additional $25 a
week in benefits through the end of 2009 and offered incentive funds for
States that chose to modernize their unemployment insurance systems.
Subsequent to the Recovery Act, Congress passed several more extensions
and expansions of unemployment insurance, and another round of reforms
aimed at assisting people searching for work.
Effects of Unemployment Insurance on Workers. In total, 24 million
U.S. workers have received extended unemployment insurance benefits.
Counting workers’ families, over 70 million people have been supported by
extended UI benefits, including more than 17 million children. Benefits have
helped a broad swath of individuals, including 4.8 million with a bachelor’s
degree or higher. The impact was profound: the Census Bureau estimates
that from 2008 to 2012, unemployment insurance kept over 11 million
people out of poverty.
Beyond providing income support and keeping families out of poverty, unemployment benefits also affect labor markets. As discussed in
the Executive Office of the President report on unemployment insurance
(Council of Economic Advisers and Department of Labor, 2014), elevated
unemployment rates in recent years were driven by declines in the demand
for labor, with only slight reductions in labor supply stemming from unemployment insurance extensions. Moreover, as shown by Chetty (2008),
unemployment benefits can also have a positive effect on labor productivity,
because they give people time to search for a job better suited to their skills.
In addition to supporting incomes, unemployment benefits deter
the long-term unemployed from dropping out of the labor force. After the
extension of the unemployment benefits program in 2008, the long-term
unemployed dropped out of the labor force at a considerably reduced rate,
and Rothstein (2011) suggests that most of the small increase in unemployment rates due to extended benefits can be attributed to this phenomenon.
While job-finding rates for the long-term unemployed remain low, keeping
people in the labor market increases the chance that they will eventually
resume working, which supports the economy’s long-run potential.
Unemployment Insurance Reforms. The Recovery Act also included
the most significant reforms to unemployment insurance in decades through
a $7 billion fund to incentivize states to modernize their UI systems and to
update eligibility rules to reflect the changing labor market. States received
an incentive payment if they implemented some suggested improvements
to their eligibility rules. These suggested improvements included allowing a
120  |  Chapter 3

worker to become eligible based on his or her most recent earnings (rather
than earnings in the previous calendar year) or when quitting a job because
of certain circumstances (compelling family responsibilities, a relocating spouse, domestic violence, or sexual assault). Proposed reforms also
included offering benefits to individuals seeking only part-time work and
providing for a dependent allowance.
Overall, states invested $3.5 billion of Recovery Act funds toward
these modernization efforts. The law prompted 41 states to make nearly
100 reforms to their unemployment insurance programs. Numerous states
expanded eligibility to workers whose job loss was due to compelling family circumstances, with 13 states adding coverage for domestic violence,
14 states adding coverage to care for a sick family member, and 16 states
extending coverage to a relocating spouse.
In February 2012, the President signed into law more reforms to
unemployment insurance—many of which were originally proposed in the
American Jobs Act—including measures to help the long-term unemployed
get back to work. Specifically, the new law created opportunities for states
to test new strategies to help the long-term unemployed find new work.
The Administration also expanded “work-sharing” programs across the
country, which will help prevent layoffs by encouraging struggling employers to reduce hours for workers rather than cut headcount. Additionally,
for the first time, the reforms allowed the long-term unemployed who were
receiving federal benefits to start their own businesses, while also providing
support to states to expand entrepreneurship programs.
Protecting the Most Vulnerable. The Recovery Act and subsequent
legislation also included a range of proposals focused on protecting the
most vulnerable. These measures included expanding the Earned Income
Tax Credit (EITC) and the refundable portion of the child tax credit, both
of which provide an additional reward to work for low-income families.
The Administration also sought to ensure that the Making Work Pay
credit was refundable, so that it benefited not just middle-class families but
moderate-income working families as well. The Recovery Act expanded
the Supplemental Nutrition Assistance Program (SNAP) to help families
through tough times while also providing emergency benefits through
Temporary Assistance to Needy Families (TANF), including subsidies to
encourage hiring of low-income parents. The Recovery Act also ended
or prevented homelessness for over 1.3 million families through the
Homelessness Prevention and Rapid Rehousing Program.
All told, the pre-existing social insurance system combined with
the expansions in the Recovery Act and subsequent extensions were very
effective in preventing a large rise in the poverty rate, despite a substantial
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  121

downturn in the economy. Even though the economy was dealt its most
severe blow since the Great Depression, Wimer et al. (2013) find that from
2007 to 2010, the poverty rate measured to include the effects of antipoverty
policy measures rose just half a percentage point. Excluding these measures,
the poverty rate would have risen 4.5 percentage points—nine times greater
than the actual increase. Chapter 6 further discusses the effects of the
Administration’s policies on reducing poverty.

The Effect of the Recovery Act
on Long-Term Growth
The Recovery Act and subsequent jobs measures also contained a
large number of provisions that were aimed at strengthening long-term
growth. In designing the Act, the Administration believed that it was not
just the quantity of the fiscal support that mattered, but the quality of it as
well. In this sense, the Administration took to heart a lesson that has been
pointed out by many but can be traced back as early as the 19th century to
a French writer and politician named Frederic Bastiat. Bastiat (1848) wrote
of a shopkeeper’s careless son who broke a window in the storefront. When
a crowd of onlookers gathered to inspect the damage, Bastiat took objection
to the discussion that ensued: “But if, on the other hand, you come to the
conclusion, as is too often the case, that it is a good thing to break windows,
that it causes money to circulate, and that the encouragement of industry in
general will be the result of it, you will oblige me to call out, ‘Stop there!’”
For this reason, the Recovery Act was designed not just to provide an
immediate, short-term boost to the economy, but also to make investments
that would enhance the economy’s productivity and overall capacity even
after the direct spending authorized by the Act had phased out. The Act’s
investments in expanding broadband infrastructure and laying the groundwork for high-speed rail, to take two examples, are a far cry from the broken
window in Bastiat’s parable because they do so much more than simply
restore things to how they once were. Rather, these types of investments will
raise the economy’s potential output for years to come, from a rural school
that can now offer its students and teachers high-speed Internet access, to a
business that has a new option to transport its goods more quickly.
As shown in Table 3-8, the Recovery Act included $300 billion of
these types of investments in areas such as clean energy, health information
technology, roads, and worker skills and training. Figure 3-11 suggests that
the timing of these investments was relatively more spread out than some
of the Act’s other measures, consistent with the longer-term focus of these
projects.
122  |  Chapter 3

Table 3–8
Recovery Act Long Term Growth Investment by Category
Billions of Dollars
Estimated Cost (2009–2019)a
Capital
Construction of Transportation Infrastructure

30.0

Environmental Cleanup and Preservation

28.0

Construction of Buildings

23.9

Public Safety and Defense

8.9

Economic Development

14.6

Memo: Business Tax Incentives

11.7

Labor
Pell Grants

17.3

Special Education

12.2

Help for Disadvantaged Children

13.0

Other Human Capital

10.3

Technology
Scientific Research

18.3

Clean Energy

78.5

Health and Health IT

32.0

Broadband

6.9

Other
Total Public Investment

6.7
b

300.6

Note: a. Estimated cost includes appropriations and tax provisions through 2019:Q3.
b. Items may not add to total due to rounding. Total excludes Business Tax Incentives.
Source: Office of Management and Budget; Department of the Treasury, Office of Tax Analysis based on the
FY2011 budget; CEA calculations.

Protecting and Expanding Investments in Physical Capital
The Recovery Act and subsequent jobs measures were designed to
expand both private capital and public capital.
Business Tax Incentives for Private Capital. The theory behind
incentivizing private capital holds that, at a time of systemic financial crisis,
firms might not have access to sufficient capital through financial markets
to invest or might be overly deterred from investing due to uncertainty,
as explained in a report by the Treasury’s Office of Tax Policy (2010a).
To overcome these impediments to private investment, the Recovery Act
and subsequent measures included business tax cuts designed to increase
cash flows—like extended periods for net operating loss carrybacks and
bonus depreciation—that, in effect, constituted an interest-free loan to
businesses. Some economic research (House and Shapiro 2008) has shown
that bonus depreciation policies can noticeably raise investment. At low
interest rates, these measures had a small net present value cost to the
Federal Government, but provided resources to credit-constrained firms to
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  123

Figure 3-11
Recovery Act Cumulative Public Investment Outlays, 2009‒2013

Billions of $
300

2013:Q3

250
200
150
100
50
0
2009:Q1

2010:Q1

2011:Q1

2012:Q1

2013:Q1

Source: Office of Management and Budget; Department of the Treasury, Office of Tax Analysis,
based on the FY2013 Mid-Session Review.

support investment. Building on this approach, in fall 2010 the President
proposed 100 percent expensing for business investment, which, as passed
by Congress in December 2010, became the largest temporary business
investment tax incentive in history.
Transportation and Other Investments in Public Capital. A modern
and effective transportation infrastructure network is both necessary for the
economy to function and a prerequisite for future growth. Numerous studies have found evidence of large private-sector productivity gains from public infrastructure investments, as highlighted in a report by the Department
of Treasury and CEA (2012).20 The early stage of the recovery has been a
particularly opportune time to undertake such investments because of the
high level of underutilized resources in the economy and low construction
costs. The Treasury report also points out that transportation investments
can create middle-class jobs and lower transportation costs, which would
otherwise weigh on household budgets.
The Recovery Act allocated $48 billion to programs administered by
the Department of Transportation, with almost 60 percent for highways
and 37 percent for public transportation and intercity passenger rail. The
magnitude of this aid was substantial. While it is difficult to estimate what
transportation expenditures would have been without the Recovery Act,
20 Many of these studies are summarized in Munnell (1992) and Fernald (1999).

124  |  Chapter 3

total highway spending in 2010 was about $27 billion (or 24 percent) higher
than in 2007. This increase occurred in a period when user revenues (such
as fuel taxes and other fees), the usual source of funding for states for transportation projects, were declining. Moreover, an equal dollar amount of
expenditures was more effective during the recession, because construction
costs for highways (as measured by the National Highway Construction
Cost Index) declined about 20 percent between mid-2008 and mid-2009,
and remained relatively flat through 2011.21
In addition to these direct programs, the Recovery Act also provided
indirect support for transportation projects through Build America Bonds.
The Federal Highway Administration estimated that 26 percent of the total
funds raised by Build America Bonds (or $48 billion) were used by states for
transportation projects. Further, a Recovery Act provision that temporarily
exempted Private Activity Bonds (PABs) from the Alternative Minimum
Tax (AMT) enabled airports across the country to access credit at affordable
rates. The Federal Aviation Administration estimated that 24 U.S. airports
issued $12.7 billion in bonds under the Recovery Act AMT exemption,
realizing $1.06 billion of present value savings ($1.8 billion in gross savings)
through early November 2010.
With these funds, shovels went in on more than 15,000 transportation
projects across the Nation. The Department of Transportation estimates that
these projects will improve nearly 42,000 miles of road, mend or replace
over 2,700 bridges, and provide funds for over 12,220 transit vehicles.
The Recovery Act also made the largest-ever investments in American
high-speed rail, constructing or improving approximately 6,000 miles of
high-performance passenger rail corridors and procurement of 120 nextgeneration rail cars or locomotives.
Finally, the Recovery Act initiated the Transportation Investment
Generating Economic Recovery (TIGER) grant program, which allowed
the Department of Transportation to invest in critical projects that were
difficult to fund through traditional means. The TIGER program included a
competitive process that encouraged innovation and regional collaboration.
The program made extensive use of benefit-cost analysis to evaluate project
applications, and required grant recipients to track the performance of their
projects once launched to ensure that they achieve the promised benefits.
The program also allowed many cities, counties, and other government
entities to access direct Federal funds for the first time. The initial $1.5
billion TIGER program was deemed so successful that it was extended five
additional times and is currently in effect through September 2014.
21 See Transportation Investments in Response to Economic Downturns, Special Report 312,
Transportation Research Board of the National Academies.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  125

The Recovery Act also invested in restoring or otherwise improving
infrastructure to allow Americans to safely and easily access public lands and
waters. Investments included about $1 billion to the National Park Service,
Fish and Wildlife Service, and Forest Service for deferred maintenance of
facilities and trails and for other critical repair and rehabilitation projects.
These projects help support the infrastructure needed to sustain the outdoor
recreation economy and contribute to the enjoyment of public lands.
The Recovery Act included funding for programs administered by
the Environmental Protection Agency (EPA) to protect and promote both
a healthier environment and jobs. These investments have generated substantial environmental benefits, such as cleaning up contaminated land and
putting that land back to economic use, reducing air pollution from diesel
engines, and reducing contaminants in both surface water and drinking
water. EPA’s Brownfields program used $100 million in Recovery Act funds
to leverage additional funds and cleaned up 1,566 acres of properties that
are now ready for reuse, far exceeding the original target of 500 acres. The
Act’s funding led to 30,900 old diesel engines being retrofitted, replaced, or
retired, which has reduced lifetime emissions of carbon dioxide by 840,300
tons and particulate matter by 3,900 tons.22 More than 3,000 water quality
infrastructure projects and Clean Water projects are improving or maintaining sewage treatment infrastructure for over 78 million people nationwide,
as another Act investment. The Recovery Act funds have also enabled 693
drinking water systems, serving over 48 million Americans, to return to
compliance with Safe Drinking Water Act standards.23

Protecting and Expanding Investments in Human Capital
The Recovery Act was also aimed at protecting and expanding human
capital. Saving and creating jobs helps protect human capital, in part, by preventing the loss of skills—including job search skills—that can come from
prolonged periods of unemployment. The evidence shows that protracted
unemployment in Europe in the 1980s and 1990s resulted in sustained loss
of human capital (Blanchard and Summers 1986, Ljungqvist and Sargent
1998). Helping workers better connect with jobs, whether through unemployment insurance reforms or job subsidies in the TANF emergency fund,
has helped protect human capital.
Significant investments and reforms in education were critical to
actually expanding human capital. State and local governments typically
provide more than 90 percent of the funding for elementary and secondary
education and about 40 percent of the funding for public institutions of
22 See Environmental Protection Agency (2013).
23 See http://www.epa.gov/recovery/accomplishments.html

126  |  Chapter 3

higher education in the United States. As the economy slowed in 2008, State
revenues declined, putting pressure on education budgets.
The Recovery Act dramatically increased funding for education through Title I grants to local education agencies (LEAs), School
Improvement Grants, and grants for special education. In addition, the Act
increased student aid and support for post-secondary institutions to invest
in new buildings and research in innovative health and energy technologies.
In response to these grants, recipients reported that more than 800,000
education job-years were saved or created, keeping teachers, principals,
librarians, and counselors as well as university faculty and staff on the job.
States also reported that they used State Fiscal Stabilization Funds
from the Recovery Act to restore sizable shares of K-12 education funding.
For example, the Recovery Act restored 9 percent of K-12 education funding
in California, Indiana, Alabama and Oregon; 12 percent of such funding in
Florida, Wisconsin and South Carolina; and 23 percent of K-12 education
funding in Illinois in fiscal year 2009. In at least 31 states, Recovery Act funds
prevented or lessened tuition increases at public universities, including
universities in Massachusetts, Minnesota, and Virginia. Without this influx
of State Fiscal Stabilization Funds, these states would have endured drastic
cuts in education funding.
The Recovery Act launched the innovative Race to the Top Program
with $4.35 billion. Race to the Top is a competitive grant program designed
to encourage and reward States to implement critical reforms designed to
help close the achievement gap and improve student outcomes, including
better student assessments; better data systems to provide teachers and
parents with information about student progress; new steps to develop and
support effective teachers; and efforts to turn around low-achieving schools.
Encouraged by the incentives included in Race to the Top, states across the
country chose to adopt more rigorous academic standards aligned to higher
expectations for college and career readiness. To date, 19 states, representing
45 percent of all K-12 students, have received Race to the Top funds; and to
compete for funds, 34 states modified state education laws and policies in
ways known to improve education.
The Recovery Act also expanded the Pell Grant program, raising
the maximum grant from $4,731 to $5,550, and it created the American
Opportunity Tax Credit to modify and replace the Hope higher education
credit (this policy was later extended by the Tax Relief, Unemployment
Insurance Reauthorization and Job Creation Act of 2010 and the American
Taxpayer Relief Act of 2012). The passage of the Health Care and Education
Reconciliation Act of 2010 enabled further expansion of the Pell Grant
award. Together, these efforts to expand higher education opportunity
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  127

helped individuals who chose to return to school or remain in school to
bolster their skills in a demanding job market. As a result, the Pell Grant
program offered $36.5 billion in aid to more than 8.8 million undergraduate
students in FY 2010, compared to roughly half as much aid, $18.3 billion, for
6.2 million students in FY 2008 (U.S. Department of Education, 2011). The
largest growth in Pell Grant applications came from students in the lowestincome categories. For tax year 2009, 8.3 million tax returns claimed $14.4
billion in American Opportunity Tax Credits. This level of education credits
(including the lifetime learning credit) was a nearly $10 billion increase from
the prior year (U.S. Department of the Treasury, 2010b).

Investments in Technology and Innovation
Some of the highest returns to investment are in the area of innovation. Often, innovation produces large returns for the economy as a whole,
but since businesses do not capture all these society-wide returns, they
tend to underinvest in innovation. For example, firms may not undertake
research and development even though it would benefit the rest of their
industry, other industries, and their regional economies. In general, looking
across a range of industries, economists have estimated that the divergence
between private and social returns to investment may be as great as two-toone (for instance, Hall et al. 2009). This is especially true when investments
can result in externalities that are not captured by the entity making the
investment. For instance, in the energy sector, there are substantial climate
and national security benefits to cleaner energy that are not fully internalized
in the form of financial rewards for individual firms. The Recovery Act made
a significant impact on innovation—complementing the other measures this
Administration has taken to encourage innovation.
Scientific Research. The Recovery Act provided a one-time supplemental appropriation of over $3 billion for the National Science Foundation
and $1 billion for the National Aeronautics and Space Administration. It also
increased support for the National Institute of Standards and Technology
and provided the Advanced Research Projects Agency-Energy (ARPA-E)
with funding of $400 million. The ARPA-E is charged with researching
transformative energy technologies. So far it has pioneered research into socalled second-generation biofuels, which utilize agricultural and municipal
waste, as well as more efficient batteries, superconducting wires, and vehicles
powered by natural gas.
Clean energy. Clean energy was the focus of more than $90 billion in
government investment and tax incentives in the Recovery Act. The purpose of these investments was to help create new jobs, reduce dependence
on foreign oil, enhance national security, and improve the environment
128  |  Chapter 3

Figure 3-12
Advanced Renewable Electric Power Net Generation, 2000‒2012

Billion kilowatt hours
200
180

Solar/PV

160

Wind

140

Waste
Wood

120

Geothermal

100
80
60
40
20
0

2000

2002

2004

2006

2008

Source: Energy Information Administration, Monthly Energy Review.

2010

2012

by countering climate change. Key targets included energy efficiency (with
programs such as the weatherization assistance program), renewable generation (with investments in wind turbines, solar panels, and other renewable
energy sources), and grid modernization. Many of these clean energy programs were administered though the $38 billion Recovery Act portfolio of
the Department of Energy.
Using the multiplier model described earlier in the chapter, CEA
estimated that clean energy investment created or saved about 650,000 jobyears, directly or indirectly, through 2012.24 These investments have started
to drive changes in energy production, as highlighted, for instance, by Aldy
(2013). Owing in large part to these clean energy incentives and investments,
renewable wind, solar, and geothermal energy have increased their contributions to U.S. energy supply each year since 2008. For instance, as shown in
Figure 3-12, wind electricity net generation nationwide grew by 145 percent
from 2008 to 2012. Solar thermal and photovoltaic electricity net generation
more than quadrupled during the same period. Meanwhile, carbon dioxide
emissions from the electric power sector fell approximately 14 percent over

24 See CEA’s (2010a) second report on the Recovery Act of 2009 for a detailed discussion on
the macroeconomic effects of clean energy investment. The latest estimates are presented in
CEA’s (2010c) fourth quarterly report.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  129

the period, even though total power generation declined by only about 2
percent.
Many of the clean energy provisions of the Recovery Act were
designed to bring in private funds through co-investment. For example,
through Energy Cash Assistance, individuals and businesses that installed
certain types of renewable energy generation received a grant equal to 30
percent of the project’s cost.
Of course, not every investment in clean energy will ultimately result
in a transformative technology. Because funding is often directed to projects
based on ideas that are at the frontier of scientific research, there is a certain
degree of risk involved. But given the grave economic, environmental, and
national security consequences of climate change, these types of investments
must continue. An independent review released in 2012 found that, on the
whole, the Department of Energy loan guarantee programs are expected to
perform well and hold even less risk than initially envisioned by Congress.
Health Care Information Systems. The Health Information
Technology for Economic and Clinical Health (HITECH) Act, enacted as a
part of the Recovery Act, encouraged adoption and use of health information technology. The core of the HITECH Act is a set of financial incentives
to health care providers to adopt and make “meaningful use” of electronic
health records. The HITECH Act also provided $2 billion to the Department
of Health and Human Services to fund activities to encourage the diffusion
of health information technology, such as investing in infrastructure and
disseminating best practices. The Act also made a variety of other changes,
including provisions to facilitate data sharing across health care providers to
support coordinated care and protect patient privacy.
Fully integrated electronic health record systems allow immediate
and complete access to all relevant patient information. These innovations
have the potential to greatly improve coordination of care—for example, by
limiting the unnecessary duplication of tests and procedures—and also to
reduce medical errors, thereby lowering health care costs. Chapter 5 further
explains the benefits of fully integrated electronic health record (EHR)systems and discusses the dramatic increase in the share of medical providers
using electronic health records in recent years.
Broadband. The Recovery Act helped increase access to broadband
and drive its adoption across the country, both directly through grants, and
indirectly through tax incentives such as increased expensing of investment
costs.25 It provided $4.4 billion through the Department of Commerce’s
National Telecommunications and Information Administration to deploy
25 See the Office of Science and Technology Policy and the National Economic Council report
Four years of broadband growth (2013).

130  |  Chapter 3

broadband infrastructure (for instance, laying new fiber-optic cables or
upgrading wireless towers, and connecting key institutions such as schools,
libraries, hospitals, and public safety facilities) and support public computer
centers (establishing new public computer facilities to provide broadband
access to the general public or specific vulnerable and underserved populations). The funding also encouraged sustainable adoption of broadband (for
instance, through digital literacy training and outreach campaigns), led to
the publication of the National Broadband Map (www.broadbandmap.gov),
and supported state broadband leadership and capacity building activities
(through, for example, local broadband planning teams and information
technology assistance provided to small businesses, schools, libraries, and
local governments). The Recovery Act also provided $2.5 billion through
the Department of Agriculture Rural Utilities Service to expand broadband
access in rural areas.
Because of these grants, over 110,000 miles of broadband infrastructure have been added or improved, and high-speed connection has been
made available to about 20,000 community institutions. These projects have
also delivered about 16 million hours of technology training to more than 4
million users.
In part as a result of the Recovery Act and related policies, broadband
access has risen substantially in recent years. Chapter 5 discusses broadband
development in more depth.

Fiscal Sustainability and the Recovery Act
The Recovery Act and subsequent fiscal measures were part of an
overall fiscally responsible economic strategy that cut the deficit in the
medium and long run. Moreover, given the economic context in which the
jobs measures were passed, these measures alone had little if any impact on
long-run fiscal sustainability.
The Recovery Act is entirely temporary—it cost $763 billion over the
first decade (not counting the extension of the AMT patch) and it has no
long-term impact on non-interest outlays or revenues. Overall, assuming
the CBO score, the Act at most added less than 0.1 percentage point to the
75-year fiscal gap.
These estimates are small but may nevertheless overstate the true cost
of the Recovery Act. To the degree that the Act successfully expanded output
and boosted employment, those gains would result in additional revenue
and less spending on countercyclical programs than would otherwise have
occurred. Taking the estimates presented in this chapter for the increase
in GDP over the 2009-12 period—and assuming these increases led to
additional tax revenue at 18 percent of GDP, roughly the recent historical
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  131

average—then the resulting increase in revenue would alone be enough to
offset roughly one-quarter of the Act’s cost.
Moreover, to the degree that effects on output are persistent—a factor that is not captured in the estimates in this chapter but is assumed by
the IMF (2009) and Reifschneider, Wascher, and Wilcox (2013), then the
positive fiscal feedback effects could be even larger. DeLong and Summers
(2012) have shown, for example, that with plausible multipliers and persistence in output effects, it is possible that the additional output associated
with the Recovery Act, and associated additions to revenue and reductions
to debt, could result in a reduced debt-to-GDP ratio by the end of the decade.
These estimates do not reflect the potential benefits for long-term
growth of the productivity-enhancing investments in the Recovery Act. For
example, if an infrastructure project has a total rate of return of 10 percent,
and if overall revenues are about 18 percent of GDP, then it would have a
rate of return to the Federal taxpayer of about 2 percent. Given the Federal
borrowing costs at the time of the Recovery Act, the investment would conceivably pay for itself over time and reduce Federal debt as a share of GDP
as the investment produces returns.
None of these estimates should be taken as conclusive or as a suggestion that official budget scoring should take these feedback effects into
account. When the economy is operating at full employment, and monetary
policy is not constrained by the zero lower bound, many of these macroeconomic feedback effects would be less relevant or not even operative at all.
Moreover, if fiscal policy actions raised the specter of substantially larger
and less sustainable future deficits and debt, that could reduce confidence
and raise interest rates, undermining any beneficial economic feedback.
But in this case, these measures were passed at the same time that the
Administration was also laying out steps for longer-term deficit reduction
and reducing the fiscal gap by passing major deficit-reduction measures,
including the Affordable Care Act and the Budget Control Act.
As a result, given the overall context of highly insufficient aggregate
demand, monetary policy operating at the zero lower bound, and other
measures for medium- and long-term deficit reduction, fiscal measures to
support jobs have the potential for even larger impacts on output and thus
greater associated revenue feedbacks and a much lower long-run fiscal cost,
if they have any long-run fiscal cost at all.

Conclusion
The Recovery Act and subsequent jobs measures were designed to help
propel the economy out of the worst contraction since the Great Depression

132  |  Chapter 3

and to set the stage for stronger future growth. Considerable evidence suggests the Federal Government’s efforts to jump-start the economy were
successful. CEA estimates that the Recovery Act provided an important and
timely boost to GDP in 2009 and 2010, and led to the creation of about 6.4
million additional job-years through 2013—estimates that are in line with
those of CBO and of other forecasting groups. Other fiscal efforts enacted
subsequent to the Recovery Act brought the total to 8.8 million job-years.
The Administration’s actions have been guided by the notion that fiscal support measures would only be needed for a temporary period, and this
view is being borne out. Most temporary measures to support the economy
expired in 2013, most notably the payroll tax cut. Businesses and households
are now in far better shape as a result of several years of deleveraging, and
private-sector growth has led the way since 2010. Although many challenges
linger, and supportive measures like emergency unemployment insurance
remain necessary given the unacceptably high rate of long-term unemployment, the economy has the potential for even stronger growth in 2014.
Public policy, in particular public investment in areas like research,
infrastructure, and innovation, will continue to play an important role in the
economy. The President is proposing additional investments and reforms
in all of these areas. But, in these cases, investments are part of a longerterm, sustained commitment to expanding the productive capacity of the
economy without the same need for immediate countercyclical support.
Overall, the Recovery Act and subsequent measures are one of the
main reasons why the U.S. economy was able to return to record levels of per
working-age population GDP within just over four years of the onset of the
recession and to bring the unemployment rate down by 0.8 percentage point
per year—when many other countries with systemic financial crises have not
seen their GDP per working-age population fully recover or their unemployment rates start a sustained fall. In the longer run, the benefits of all of these
efforts will be more difficult to isolate from other simultaneous changes,
but they will be no less profound in terms of their cumulative impact on the
economic well-being of the Nation.

Appendix 1: Components of the Recovery
Act and Subsequent Fiscal Measures
Table 3-9 reports the actual budgetary impact of the Recovery Act
from its inception through the latest data available (the end of fiscal year
2013).
Table 3-10 reports the budgetary impact classified into the six
broad functional categories shown also in Figure 3-1: individual tax cuts,
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  133

Alternative Minimum Tax relief, business tax incentives, State fiscal relief,
aid to directly impacted individuals, and public investments. The following
sections of this appendix will discuss each of these categories in more detail.

Tax Relief
Within the first three categories of tax cuts, major programs included
the Making Work Pay tax credit, which provided a 6.2 percent credit on earnings up to a maximum value of $400 for individuals and $800 for couples,
phasing out starting at income above $75,000 and $150,000, respectively
(estimated to cost about $116 billion between 2009 and 2011). The credit was
administered through reducing tax withholdings and the Internal Revenue
Service required that companies reduce withholding by April 1, 2009. In
addition, the legislation made $250 one-time payments to seniors, veterans,
and people with disabilities. The Recovery Act included the Making Work
Pay tax credit for 2009 and 2010. In December of 2011 Congress enacted a
2-percentage point reduction of the Social Security payroll tax for 2011 that
was extended through 2012 and expired at the start of 2013.
Additionally, the Recovery Act provided tax credits for families, such
as an expansion of the child tax credit, including making it refundable for
more low-income families (at a total estimated cost of $15 billion), expansions of the earned income tax credit for married couples and families with
more than three children ($5 billion), and the American Opportunity Tax
Credit to help make college more affordable. All of these measures have
since been extended through 2017, and the President’s Budget for 2014
proposes to make them permanent—rendering them among the only items
from the Recovery Act intended to be permanent.
The Recovery Act also raised the exemption amount for the AMT to
$46,700 for individual taxpayers and $70,950 for joint filers, at an estimated
cost of $70 billion. Because this was a widely expected continuation of previous AMT patches, this component of the Recovery Act did not represent a
net new fiscal impetus for the economy and is not included in CEA’s macroeconomic estimates.
For businesses, the legislation provided cost-effective incentives to
expand investment by allowing businesses to immediately deduct half of the
cost of their investments (bonus depreciation) and also to extend the period
over which small firms (except those receiving TARP funds) could claim
losses and expense capital purchases. Businesses buying back or exchanging
their own debt at a discount were also allowed to defer any resulting income.
All of these measures were designed to improve the cash flow for firms that
might be facing credit constraints and to increase incentives to invest. Longrun costs to the Federal Government were limited because the measures
134  |  Chapter 3

Table 3–9
Recovery Act Outlays, Obligations, and Tax Reductions
Through
the end ofa

Outlays

Obligations

Tax Reductions

Sum of Outlays and
Tax Reductionsb

2009:Q1

8.6

30.5

2.4

11.0

2009:Q2

47.7

127.3

35.6

83.3

2009:Q3

54.4

98.5

31.8

86.2

2009:Q4

53.5

57.6

30.2

83.7

2010:Q1

46.7

48.2

64.9

111.6

2010:Q2

46.4

41.7

77.3

123.6

2010:Q3

50.6

48.6

16.4

66.9

2010:Q4

40.7

20.8

8.4

49.1

2011:Q1

25.0

6.2

31.9

56.9

2011:Q2

25.1

5.0

–5.1

20.0

2011:Q3

21.9

9.2

2.1

23.9

2011:Q4

17.7

5.7

2.0

19.6

2012:Q1

14.3

5.2

–4.0

10.4

2012:Q2

12.8

6.5

–3.0

9.8

2012:Q3

12.0

4.4

–0.5

11.6

2012:Q4

11.2

5.8

0.5

11.7

2013:Q1

11.0

6.2

0.7

11.7

2013:Q2

7.2

4.0

0.4

7.7

2013:Q3

5.6

2.5

0.4

5.9

512.4

533.8

292.2

804.6

Total
Through 2013:Q3b

Notes: a. Data on outlays and obligations are for the last day of each calendar quarter.
b. Items may not add to total due to rounding.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury,
Office of Tax Analysis based on the FY2013 Mid-Session Review.

largely advanced tax benefits that companies would receive anyway. The
50 percent bonus depreciation was subsequently extended and expanded to
100 percent expensing, and the net operating loss carryback was extended
to larger firms. In addition, the Recovery Act included incentives for investments in renewables and advanced energy manufacturing, and in areas
undergoing significant distress through State and local government-issued
Recovery Zone Bonds. The Recovery Act also increased funding for the New
Markets Tax Credit and provided incentives to hire unemployed veterans
and disconnected youth.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  135

Table 3–10
Recovery Act Fiscal Stimulus by Functional Category
Through the
end ofa

Individual
Tax Cuts

AMT
Relief

Business
Tax
Incentives

State Fiscal
Relief

Aid to
Directly
Impacted
Individuals

Public
Investment
Outlays

Totalb

2009:Q1

2.3

0.0

0.1

8.5

0.0

0.0

11.0

2009:Q2

26.3

7.8

12.5

19.6

9.6

7.4

83.3

2009:Q3

14.3

6.0

10.5

15.6

22.2

17.6

86.2

2009:Q4

15.8

3.5

9.0

15.5

23.4

16.5

83.7

2010:Q1

43.3

11.4

6.9

16.2

16.1

17.7

111.6

2010:Q2

22.4

47.5

4.9

16.6

5.2

27.0

123.6

2010:Q3

9.8

7.2

–2.6

15.0

4.7

32.8

66.9

2010:Q4

8.6

0.0

–1.5

14.6

4.7

22.6

49.1

2011:Q1

25.5

4.6

–1.5

4.4

3.5

20.4

56.9

2011:Q2

12.2

–19.0

–1.5

4.7

3.3

20.3

20.0

2011:Q3

0.3

0.0

–1.5

2.3

4.1

18.7

23.9

2011:Q4

0.1

0.0

–0.9

1.9

2.4

16.2

19.6

2012:Q1

0.3

0.0

–0.9

1.7

2.2

7.1

10.4

2012:Q2

0.0

0.0

–0.9

1.2

2.2

7.3

9.8

2012:Q3

–0.0

0.0

–0.9

1.2

2.0

9.3

11.6

2012:Q4

0.1

0.0

–0.7

0.9

1.6

9.9

11.7

2013:Q1

0.3

0.0

–0.7

1.3

1.6

9.2

11.7

2013:Q2

0.0

0.0

–0.7

1.2

1.6

5.5

7.7

2013:Q3

–0.0

0.0

–0.7

0.6

1.1

5.0

5.9

Total
Through
2013:Q3b

181.7

69.0

28.8

143.0

111.5

270.5

804.6

Notes: a. Data on outlays and obligations are for the last day of each calendar quarter.
b. Items may not add to total due to rounding.
Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury,
Office of Tax Analysis based on the FY2013 Mid-Session Review.

Aid to Affected Individuals
An expansion in unemployment benefits offered significant aid to individuals.26 Typically, American workers who have lost their jobs are entitled
to 26 weeks of benefits under the unemployment insurance (UI) program,
which tends to replace about half of lost earnings and is paid for entirely by
the states through payroll taxes levied on employers. In June 2008, Congress
created the Emergency Unemployment Compensation (EUC) program,
which provided an additional 13 weeks of federally financed compensation
in all states to eligible individuals who had exhausted their UI benefits. The
26 For a comprehensive discussion of the various employment benefits programs implemented
in recent years, see the Council of Economic Advisers and the Department of Labor report The
economic benefits of extending unemployment insurance (2014).

136  |  Chapter 3

Recovery Act extended and expanded the EUC program to reflect that fact
that with jobs increasingly scarce the optimal balance of unemployment
insurance shifted towards covering people for a longer period of time. It also
provided 100 percent Federal funding of the pre-existing Extended Benefit
(EB) program, which provides an additional 13 or 26 weeks of benefits in
states where unemployment is exceptionally high and rising. (EB costs are
usually borne half by the Federal Government and half by the States.)
The Recovery Act also added $25 a week to benefits and exempted the
first $2,400 in yearly unemployment benefits from taxes. The CBO (2009a)
estimated the total costs of these changes to the unemployment compensation system at $39 billion. In addition, the Federal government offered states
incentives to modernize their UI programs, picking up the cost of new
provisions allowing workers to become eligible based on recent earnings
(rather than those from the previous calendar year) and extending benefits
to part-time job seekers.
The Recovery Act provided a 13 percent increase in SNAP payments
and lifted several restrictions governing the length of time that individuals
could collect food stamps, at an estimated cost of $20 billion. For the first
time, the Federal Government agreed to temporarily pay 65 percent of
health insurance premiums for laid off workers who wanted to continue
with their employer-sponsored health insurance. Other aid to individuals
included funds for job training and improving skills of the hard to employ
and young workers.
The Recovery Act devoted substantially more resources compared
with previous antirecessionary policies to investments in education and
research and development. The legislation increased the Pell Grant maximum by $500 to $5,550, at an estimated cost of $17 billion over ten years.
The Recovery Act also boosted Title I aid and other programs for disadvantaged children ($13 billion) and funds for special education ($12 billion).

State Fiscal Relief
The Recovery Act provided unprecedented support for State and local
governments, which often face budget challenges in a recession because their
revenues rise and fall with the economy, while pressures on spending, especially on programs targeted to the disadvantaged, tend to move in the opposite direction. The result can be budget shortfalls, or gaps between expected
revenues and expenditures. These gaps pose problems for state residents
already affected by the downturn, and for the larger economy because most
State and local governments are generally bound by constitutional or statutory requirements to balance their operating budgets each year. As shown

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  137

by Poterba (1994), states and localities have to raise taxes or cut spending,
precisely when doing so can most harm recovery.
To dampen such counterproductive tax increases or budget cuts, the
Recovery Act boosted Federal Medicaid payments by $87 billion, including a 6.2 percent across-the-board increase in the Federal matching rate,
plus delays of a planned reimbursement cut for some states (based on
income growth before the recession) and an increment of aid linked to local
unemployment conditions. It also established a $53.6 billion State Fiscal
Stabilization Fund to be administered by the Department of Education, but
with some funds available for other “high priority needs” such as public
safety. Unlike most previous increases in Federal grants to States and localities during a recession, these transfers were available for general fiscal relief
or left to local discretion to use, as long as recipients met basic maintenance
of effort or minimal spending requirements.
Beyond direct spending, the Recovery Act made new types of borrowing available for State and local governments. Build America Bonds (BABs)
allowed State and local governments to access non-traditional markets,
including pension funds and international investors who would not normally purchase U.S. municipal bonds because they do not owe U.S. income
taxes and therefore do not benefit from these bonds’ tax-exempt status.
Under BABs, State and local issuers could offer higher taxable interest rates
on bonds and choose to make a Federal income tax credit available to buyers or to take a direct subsidy offsetting 35 percent of their borrowing costs.
State and local governments issued $181 billion of Build America Bonds
before the program expired at the end of 2010. The Treasury Department
has estimated that this action saved issuers $20 billion in present value of
borrowing costs as well as alleviating supply pressures in the tax-exempt
market (Department of the Treasury 2011).

Investments
The Recovery Act made numerous investments in human capital,
clean energy, health information technology, roads, and the skills of U.S.
workers.27 For example, the Recovery Act provided an additional $27.5
billion for highway construction, $18 billion for public transit and inter27 CEA counts as public investment any Recovery Act expenditure or tax program that directly
results in activity that increases the capital stock of the Federal government, State and local
governments, or private firms. We also count provisions that affect the Nation’s human
capital and knowledge capital, areas not measured in the national income accounts but which
economists have identified as crucial to generating long-run economic growth. Note that tax
programs are included if they function similarly to direct spending. In other words, entities
can claim tax benefits only when associated spending occurs (e.g., the Advanced Energy
Manufacturing Tax Credit).

138  |  Chapter 3

city passenger rail, $10 billion for water infrastructure, and $18 billion for
government facilities. It also made available $57 billion for investment in
smarter grid technology, renewable energy, and energy efficiency improvements through a combination of grants, loans, and pilot programs, including
$5 billion to help low-income households weatherize their homes. Scientific
projects from the National Science Foundation, National Institutes of
Health, NASA, the Department of Energy, and others received over $15
billion for scientific facilities, research, and instrumentation. Additionally,
the Recovery Act provided $7 billion to expand broadband Internet access
in underserved areas of the country. The Recovery Act also provided several
investments in health care and health information technology, including an
$18 billion measure to encourage hospitals and physicians to computerize
medical records, $2 billion for Community Health Centers, $1 billion for
fighting preventable chronic diseases, and $1 billion for researching the
effectiveness of various medical treatments. In total, more than $100 billion
of the investments—including some tax incentives—were explicitly targeted
at innovation.28

Subsequent Fiscal Measures
Table 3-11 shows the total fiscal support provided by the
Administration, by fiscal year, with a brief description of the main programs for each measure. (These data were summarized in Table 3-4.) All
measures use prospective CBO cost estimates. These totals only include
measures explicitly designed to address job creation and provide relief and
do not include routine extensions, like so-called “tax extenders” or the fix to
Medicare’s Sustainable Growth Rate formula.

Appendix 2: Fiscal Multipliers:
Theory and Empirical Evidence
Although the multiplier described in the text is simple and intuitive,
it relies on several unrealistic assumptions, and much research in macroeconomic theory over the past four decades has focused on overcoming
those conceptual problems. For example, because deficit spending in a
recession could be offset by higher taxes in a boom, Barro (1974) argued that
forward-looking individuals might save much or all of a tax cut in anticipation of higher taxes later. Although the extreme version of this argument
requires consumers who are unrealistically liquid and prescient, in general
28 Executive Office of the President and Office of the Vice President, The Recovery Act:
Transforming the American Economy through Innovation, August 2010.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  139

140  |  Chapter 3

46

7

4

13

9

0

0

0

0

0

0

0

0

0

0

98

0

0

0

0

0

0

0

0

16

1

0

2

0

1

294

0

0

0

0

158

80

23

25

2

6

1

2

0

–3

253

0

90

–0

27

145

–9

2

0

0

3

0

0

0

–6

113

68

33

1

2

22

–12

0

0

0

1

0

–0

0

–4

13

55

–0

–0

0

–27

–10

–3

0

0

1

0

–0

0

–3

–40

9

–0

–0

0

–21

–22

–5

0

0

0

0

–0

0

–2

–18

15

–0

–1

0

–16

–12

–3

0

0

0

0

–0

0

–1

–0

18

0

–0

0

–12

–4

–2

0

–0

0

0

–0

0

–1

6

17

0

–0

0

–7

–2

–1

0

–0

0

0

0

0

–1

–11

–4

0

–0

0

–5

–1

–0

0

0

0

0

0

0

–0

674

17

98

0

28

309

68

26

33

16

13

9

18

3

35

709

178

123

–0

29

237

10

12

34

16

15

9

18

3

24

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2009–12 2009–19

Description

Extended UI through 2013 • Extension of expanded
AOTC, EITC, and CTC • Extension of small business tax cut and bonus depreciation

Payroll tax and UI through end of 2012

Returning Heroes and Wounded Warrior tax credits

Payroll tax and UI through February 2012

Payroll tax cut for 2011 • Extended UI through 2011
• Extension of expanded AOTC, EITC, and CTC

Small business lending fund • Small business tax cut
and bonus depreciation for all businesses

Education Jobs Fund • Extension of FMAP relief

Extended UI 6 months • Extended first time homebuyers tax credit

Extended UI/COBRA 2 months

Hiring tax credit • Subsidized bonds for school
construction and renewable energy

Extended UI/COBRA 1 month

Extended UI/COBRA 2 months

Cash for Clunkers

Expanded weeks in UI (by 20 weeks) • Extended
first time homebuyers tax credt

Note: All measures use prospective CBO cost estimates for 2009–19. Routine tax extenders have been removed from the cost estimates. Columns for individual years contain data in fiscal year
terms. The column for 2009–2012 total contains data through the end of calendar year 2012, while the column for 2009–2019 contains data through the end of fiscal year 2019.
Source: Congressional Budget Office; Joint Committee on Taxation.

Total

American Taxpayer Relief Act of
2012 (HR 8)

Middle Class Tax Relief and Job
Creation Act of 2012 (HR 3630)

VOW to Hire Heroes Act (HR 674)

Tax Relief, Unemployment Insurance Reauthorization, and Job
Creation Act (HR 4853)
Temporary Payroll Tax Cut Continuation Act (HR 3765)

Small Business Jobs Act (HR 5297)

Worker, Homeownership, and Business Assistance Act (HR 3548)
Supplemental Appropriations Act of
2009 (HR 2346)
Defense Appropriations Act of 2010
(HR 3326)
Temporary Extension Act of 2010
(HR 4691)
Hiring Incentives to Restore Employment Act (HR 2847)
Continuing Extension Act of 2010
(HR 4851)
Unemployment Compensation Act
of 2010 (HR 4213)
FAA Safety Improvement Act (HR
1586)

Table 3–11
Fiscal Support for the Economy Enacted After the Recovery Act

introducing forward-looking behavior by consumers and firms planning for
the future changes the dynamics and magnitude of Keynesian multipliers.

Forward-Looking Models with Rigidities
Many modern macroeconomic models combine forward-looking
behavior with some form of slow-moving prices or wages, sometimes
called “New Keynesian” models. In normal times, when monetary policy is
unconstrained and interest rates can vary, these models tend to imply fiscal
expenditure multipliers that are positive but smaller than one, as shown for
instance by Cogan et al. (2010) and Coenen et al. (2012), in part because of
increases in the interest rate from monetary policy which partially offsets the
fiscal expansion.
The onset of low interest rates has spurred considerable interest in
how these models perform when monetary policy is constrained by the
zero lower bound, that is, when the nominal federal funds rate falls to
zero, as in the recent recession. For instance, Eggertson (2001), Christiano,
Eichenbaum, and Rebelo (2011) and Woodford (2011) have shown that
when nominal interest rates are near zero, government spending can be
particularly effective and generate spending multipliers that are greater
than one; at the zero lower bound, expansionary fiscal policies can increase
inflation expectations and thereby reduce real interest rates, which spurs
investment and consumption, and monetary policy does not counteract
fiscal policy. Coenen et al. (2012) simulate the effect of the Recovery Act
spending in some forward-looking models with rigidities, both conventional
models (such as Smets and Wouters (2007)) and models augmented by the
zero lower-bound effects. Their results show that the standard models imply
a notable increase in output for several years, but with multipliers smaller
than one, while the models augmented by zero-lower-bound effects imply
multipliers that are much larger than one over the first few years.
The 2007-09 recession was unusual both because the Federal Reserve
was at the zero lower bound and because of its severity. This severity raises
the specter of high unemployment and—because the path to recovery from
a deep shock is long—unusually long spells of unemployment. Long-term
unemployment can lead to deterioration of skills and to stigmatization,
which makes finding employment even more difficult. For these and other
reasons, the longer the spell of unemployment, the less likely is an individual to find a job in any given month, and the more likely he or she is to
remain unemployed or stop looking for a job altogether. This can lead to
a vicious circle: persistent slack demand means many people out of work
and long spells of unemployment, which in turn reduces the chances of the
unemployed finding a job, which perpetuates slack and further lengthens
The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  141

spells. Because the resulting unemployment dynamics depend on the path
of unemployment, not just on its current level, this phenomenon is often
referred to as “hysteresis” in the rate of unemployment.
The potential for hysteresis in unemployment—the economy getting
stuck at high rates of unemployment for an extended period—provides a
further argument for activist fiscal policy, and models that build in hysteresis effects can have large and sustained multipliers (see for example Phelps
1972, Blanchard and Summers 1986, Ball 2009, and DeLong and Summers
2012). Reifschneider, Wascher, and Wilcox (2013) stress the relevance
of these channels to the current recovery. Their research shows that the
financial crisis damaged the productive capacity of the economy, by causing a steep decline in capital accumulation, lower productivity growth, and
structural damages to the labor market, and a large portion of this damage to
the productive capacity stemmed from weak demand. These results suggest
that under such conditions fiscal policy can continue to have a meaningful
effect on output with a substantial lag.
This recent work has moved far beyond the basic multiplier. It shows
that fiscal and monetary policy can influence each other in substantial ways.
While fiscal multipliers might be less than the basic model suggests in mild
recessions and when monetary policy is unconstrained, they can be large
when monetary policy is at the zero lower bound. In addition, fiscal expansion in a deep recession can have additional long-term benefits, and therefore high multipliers, by shortening spells of unemployment, minimizing the
erosion of human capital, and increasing future productivity.

Time Series Evidence
Evaluations of fiscal effects using the structural models described
above reflect the economic theory used to construct the models. The reliability of the resulting estimates therefore depends on the reliability of the
underlying macroeconomic theory. A complementary approach to evaluating the effects of fiscal policy is instead to use models that rely less on economic theory and more on historical empirical evidence.
The main challenge to credibly implementing this data-driven
approach is using just enough theory, or finding enough independent variation in the data, to estimate the causal effect of fiscal policy on the economy:
simply noting that two variables move together does not establish causality. For example, if Congress passed countercyclical fiscal policy whenever
a recession loomed, a figure plotting the countercyclical policy variable
and GDP growth would show that countercyclical policy occurred at the
beginning of recessions. An analyst might conclude, incorrectly, that this
policy caused the recession, when in fact the policy was itself caused by the
142  |  Chapter 3

recognized onset of the recession. Analysis based on this hypothetical figure
suffers from two central problems in the estimation of causal effects from
observational (as opposed to experimental) data: simultaneous causality (the
looming recession spurred Congressional action and the fiscal policy potentially affected the course of the economy) and the presence of other omitted,
confounding factors (perhaps the Federal Reserve moved countercyclically
and it was those actions, not Congress’s, that muted the recession). The latter problem of omitted variables can be partially addressed using multiple
regression methods, but the problem of simultaneous causality requires
other approaches, and relying on simple plots or multiple regression can
lead to misleading results.29 Because such plots or regressions are uninformative, a vast literature developed over the past four decades uses more
sophisticated methods to estimate causal effects in general and the effect of
fiscal policy in particular.30
Evaluation of the effects of fiscal policy in general, and the Recovery
Act in particular, faces several additional challenges. First, the effect of activist fiscal policy must be disentangled from the automatic stabilizers built into
the tax and safety net system. Second, the effect of fiscal policy unfolds over
time, so there is not a single causal effect but rather a sequence of dynamic
causal effects, including long-lasting effects of investment on productivity
that could last for many years. Third, different fiscal policy instruments
(expenditures, taxes, transfers) will in general have different effects. Fourth,
as discussed above, theory suggests that the effect of fiscal policy could
depend on the economic environment, and in particular could depend both
on the severity of the recession and on the reaction of monetary policy.
A vast body of empirical literature now employs time-series data to
estimate the macroeconomic effect of fiscal policy. Broadly speaking, this
29 The multiple regression analysis in Taylor (2011), which estimates the effect of fiscal
policy in the 2000s, addresses in part the problem of omitted variables but not the problem
of simultaneous causality. Taylor measures the direct impact of fiscal policy on income by
the component of disposable income due to countercyclical fiscal policy from 2009-based on
the 2001, 2008, and Recovery Act fiscal programs. The level of quarterly consumption is then
regressed on contemporaneous values of on personal income, the fiscal policy measure, wealth,
and oil prices. Thus this regression controls for the separate effects of oil price movements,
in case they co-move with fiscal policy. As observed in the text, however, the sign and
magnitude of the coefficient on fiscal policy is ambiguous a-priori because of simultaneous
causality: it could be positive, zero, or negative. As it turns out, the coefficient is positive but
small, a finding that is consistent with fiscal policy having a large positive effect which, in
the regression, is offset by the fact that Congress passed it in a recession, or with fiscal policy
having little effect. Because of simultaneous causality, this regression analysis, like its graphical
equivalent, sheds little light on the question of the effect of fiscal policy.
30The field of the econometric estimation of causal effects has seen tremendous advances in
both methods and applications; for a review see Angrist and Pischke (2010), Sims (2010), and
Stock (2010). For additional methodological discussion of simultaneous causality, see Stock
and Watson (2010, Chapters 9 and 12).

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  143

literature uses two different approaches to isolate (to “identify”) the effect
of fiscal policy. The first is to impose a minimal amount of structure on
an otherwise unrestricted time series model, typically a so-called structural
vector autoregression. In an influential contribution, Blanchard and Perotti
(2002) assume that, because of implementation lags and limitations on the
information considered by policymakers, fiscal policy does not respond
immediately to other economic shocks. Under this assumption, any unpredicted movements in the fiscal variable (that is, movements that differ from
what standard fiscal policy would have suggested) are unrelated to contemporaneous economic shocks, so the effect of fiscal policy can be estimated
by tracing out the effect of those unpredicted movements on output and
employment. Using this approach, Blanchard and Perotti (2002) estimate
the government spending multiplier on GDP to be in the range 0.9 to 1.2.
Ramey (2011b) reviews the large body of research that uses structural vector
autoregressions to build on this approach to identifying the effects of fiscal
shocks. The common theme of this work is using a component of fiscal
policy—in Blanchard-Perotti (2002), the unpredictable component—which
is “as-if random” in the sense that it is unrelated to other economic shocks.
A second approach to identifying the effect of fiscal policy is to exploit
external information, such as institutional or historical knowledge, to find
changes in fiscal policy that are in effect random (that is, independent of
macroeconomic conditions), which can in turn be used to trace out the fiscal effect. Because this information falls outside the time series model being
estimated, this approach is called the method of external instruments. In
this vein, Ramey and Shapiro (1998) and Ramey (2011a) use expenditures
on wars and military buildups, arguing that they are determined by international and political, not economic, considerations. These authors estimate
GDP multipliers in the range of 0.6 and 1.2. Romer and Romer (2010),
instead, use narrative evidence from Presidential and Congressional records
and similar documents to identify tax changes that were not implemented
in response to current or forecasted economic conditions. They find that the
identified tax cuts have a sustained and large effect on output, with multipliers as high as 3. Mertens and Ravn (2012) use Romer and Romer’s (2010)
narrative to distinguish between the effects of anticipated and unanticipated
tax changes and, surprisingly and in contrast to Ramey (2011a), find little
difference in the two effects. Additional recent contributions include Favero
and Giavazzi (2012) and Mertens and Ravn (2013). Estimates of fiscal policy
effects obtained using this so-called method of external instruments are
reviewed in Ramey (2011b) and Stock and Watson (2012).
The foregoing time series estimates are predicated upon fiscal multipliers having the same size in booms and in recessions. Recent work by
144  |  Chapter 3

Auerbach and Gorodnichenko (2012) suggests that while the spending
multiplier can be relatively small during expansions, it can be much greater
than one during recessions. These results are consistent with conventional
models in recessions, but with neoclassical ones in booms, and suggest that
multipliers obtained also using fiscal policy changes that happen in booms
(such as the military buildup used by Ramey and Shapiro (1998) and Ramey
(2011a) to identify fiscal shocks) could underestimate the effect of the policies in recessions.
Finally, a different approach is to use consumer-level microeconomic
data on specific policy events, as highlighted by Parker (2011). For instance,
looking at the 2001 and 2008 tax rebates, which reached recipients in different months, Johnson, Parker, and Souleles (2006) and Parker, Souleles,
Johnson, and McClelland (2011) show that a sizable fraction of the rebate
was spent, especially by lower income or liquidity constrained households.
Their results indicate that income transfers can be an effective way to raise
consumption in the short run. This approach has the advantage of directly
estimating consumption effects, although it does not capture the full
dynamic, indirect response of the economy to the fiscal shock.

Cross-Sectional Multipliers
In addition to the works on the Recovery Act cited in the crossstate evidence section, recent work has also exploited other sources of
cross-sectional variation in government spending to estimate the size of
the fiscal multiplier. For instance, looking at the effects of windfall returns
on pension funds, Shoag (2013) estimated a local output multiplier of 2.1.
Suarez, Serrato, and Wingender (2011) reached a similar conclusion using
on changes in Federal transfers due to the decennial census. Nakamura and
Steinsson (2011) detected a 1.5 local multiplier based on regional differences
in Federal defense spending.
Table 3-12 summarizes the fiscal multipliers implied by the economic
literature on state-level effects of fiscal policies.

The Economic Impact of the American Recovery and Reinvestment Act Five Years Later  |  145

Table 3–12
Summary of Cross-Sectional Fiscal Multiplier Estimates
Study

Source of Variation

Regional Multiplier

Cost per Job

Chodorow-Reich et al.
(2011)

Formulaic spending in American
Recovery and Reinvestment Act
of 2009

2.1

$26,000

Wilson (2011)

Formulaic spending in American
Recovery and Reinvestment Act
of 2009

—

$125,000

Suarez Serrato and
Wingender (2011)

Impact of decennial census on
Federal transfers

1.9

$30,000

Shoag (2010)

Windfall returns on pension
investments

2.1

$35,000

Nakamura and Steinsson
(2011)

Regional distribution of changes
in defense spending

1.5

—

Source: Romer (2012).

146 | 146

C H A P T E R

4

RECENT TRENDS IN HEALTH
CARE COSTS, THEIR IMPACT ON
THE ECONOMY, AND THE ROLE
OF THE AFFORDABLE CARE ACT

D

ramatic progress is being made in addressing one of the enduring
problems of the U.S. health care system: the fact that millions of
Americans lack access to quality, affordable health insurance. Since January
1, the Affordable Care Act (ACA) has extended coverage to millions of
Americans, and the Congressional Budget Office (CBO) estimates that, by
2016, the ACA will reduce the number of people without health insurance
by 25 million (CBO 2014). If all states elect to take up the ACA’s Medicaid
expansion, the ACA will reduce the number of people without health insurance even further.
But the U.S. health care system also faces another enduring challenge:
decades of rapid growth in health care spending. While much of this historical increase reflects the development of new treatments that have greatly
improved health and well-being (Cutler 2004), most agree that the system
suffers from serious inefficiencies that hike costs and reduce the quality
of care that patients receive. Another key goal of the ACA was to begin
wringing these inefficiencies out of the health care system, simultaneously
reducing the growth of health care spending—and its burden on families,
employers, and State and Federal budgets—and increasing the quality of the
care delivered.
This chapter analyzes recent trends in U.S. health care costs and
documents a dramatic slowdown in recent years. According to the final data
on national health expenditures, real per-capita health spending grew at an
average annual rate of just 1.1 percent from 2010 to 2012. Preliminary data
as well as projections by the Office of the Actuary at the Centers for Medicare
and Medicaid Services (CMS) imply this slow growth continued in 2013, and

147

the CMS projections show real per-capita health spending growth averaging
just 1.2 percent over the three years since 2010. These spending growth rates
are the lowest rates on record for any two- and three-year periods, and less
than one-third the long-term historical average of 4.6 percent that stretches
back to 1960. Moreover, they have occurred at a time when the aging of the
population would have been expected to modestly increase the growth rate
of health care spending.
The historically slow growth in health costs has appeared not only in
health care spending, but also in the prices paid for health care goods and
services. Measured using personal consumption expenditure price indices,
health care inflation is currently running at around 1 percent on a year-overyear basis, a level not seen since 1963. Health care inflation measured using
the consumer price index (CPI) for medical care is at levels not seen since
1972. Health care inflation measured relative to general price inflation is also
unusually low in historical terms.
An important question is what has caused these trends and whether
they are likely to persist in the years ahead. Although the slowdown is not yet
fully understood, the evidence available to date supports several conclusions
about its causes and the role of the Affordable Care Act.
The 2007-09 recession and its aftermath have likely played some role
in the recent slowdown in health costs, and this portion of the slowdown is
likely to fade as the economic recovery continues. However, several pieces
of evidence imply that the slowdown in health care cost growth is more than
just an artifact of the 2007-09 recession: something has changed. The fact
that the health cost slowdown has persisted even as the economy is recovering; the fact that it is reflected in health care prices, not just utilization;
and, the fact that it has also shown up in Medicare, which is more insulated
from economic trends, all imply that the current slowdown is the result of
more than just the recession and its aftermath. Rather, much of the slowdown appears to reflect “structural” changes in the U.S. health care system,
suggesting that at least part of this trend—although it is uncertain how
much—is likely to persist. This conclusion is consistent with a substantial
body of recent research that seeks to quantify the recession’s contribution to
the slowdown and has found that the recession alone cannot explain recent
trends.
While various non-recession factors unrelated to the ACA appear to
be contributing to the recent slow growth in spending—including a longterm decline in the development of new prescription drugs and a long-term
increase in cost-sharing in employer sponsored plans—the ACA is also
playing a meaningful role. For example, by curtailing excessive Medicare
payments to private insurers and medical providers, the law has contributed
148  |  Chapter 4

to the recent slow growth in health care prices and spending, reducing health
care price inflation by an estimated 0.2 percentage points each year since
2010.
The ACA’s measures to reduce costs and improve quality by improving the payment incentives faced by medical providers also appear to be
beginning to bear fruit. For example, hospital readmission rates have turned
sharply lower since the ACA began penalizing hospitals that readmit a larger
number of patients soon after discharge. Similarly, the ACA has substantially
increased health care provider participation in payment models designed to
promote high-quality, integrated care. These are hopeful signs and provide
reason to believe that, as the ACA’s payment reforms continue to take effect
over the coming years, they will make an important contribution to extending the recent slowdown.
An emerging literature also suggests that the ACA’s payment reforms,
which operate primarily through Medicare (and, to a lesser extent, through
Medicaid), may generate “spillover” benefits throughout the health system.
This literature finds that when Medicare reduces payments to medical providers, private payers tend to follow suit, and also finds that the same is true
for changes to the structure of how Medicare pays providers. Some recent
evidence also suggests that changes in payment structures by one insurer
may benefit patients covered by other insurers, even if those other insurers
do not adopt the new payment structures. One possibility is that changes
by one insurer induce changes in providers’ “practice styles” that affect all
patients that providers see. This evidence suggests that the ACA’s reforms to
the Medicare payment system may be, in economic terms, “public goods.”
The presence of spillover benefits would imply that the contribution
of the ACA to the recent slowdown in health costs growth is considerably
larger than previously understood. As noted above, ACA provisions that
curb excessive Medicare payments to private insurers and medical providers
have directly reduced health care price inflation by an estimated 0.2 percent
a year since 2010. A calculation accounting for spillovers raises this estimate
to 0.5 percent a year—a substantial share of the recent slowdown in health
care price inflation.
This chapter concludes with a consideration of the economic benefits
of a sustained slowdown in health care costs. Over the long run, slower
growth in health care spending that is achieved without compromising the
quality of care will raise living standards. These gains may be substantial. If
even just one-third of the recent slowdown in spending can be sustained,
health care spending a decade from now will be about $1,200 per person
lower than if growth returned to its 2000-07 trend, the lion’s share of which

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 149
 
the Affordable Care Act

will accrue to workers as higher wages and to Federal and State governments
as lower costs.
Recent Congressional Budget Office estimates offer a concrete illustration of the potential for improvements in the Federal fiscal outlook. Since
August 2010, CBO has reduced its projections of combined Medicare and
Medicaid spending in 2020 by $168 billion and 0.5 percent of gross domestic
product (GDP). The $168 billion reduction represents a 13 percent reduction in previously projected spending on these programs and primarily
reflects the recent slow growth in health care spending. These revisions are,
however, distinct from the deficit reduction directly attributable to the ACA,
which CBO estimates will be substantial. Due in large part to the ACA’s
role in slowing the growth of health care spending, CBO estimates that the
provisions of the ACA will directly reduce deficits by about $100 billion over
the coming decade and by an average of 0.5 percent of GDP a year over the
following decade.
Slower growth in long-term health spending also reduces employers’
compensation costs in the short run, increasing firms’ incentives to hire
additional workers. This chapter surveys the available evidence on the likely
effects on employment to conclude that short-run employment gains could
be substantial, although the magnitude of these gains is quite uncertain.
This chapter proceeds as follows. The first section quantifies the
recent slowdown in health care costs. The second section discusses possible
factors behind the slowdown in costs, and also discusses the effects of the
ACA on quality of care so far and in the future. The final section discusses
the slowdown’s potential economic benefits.

Recent Trends in Health Care Costs
To document the historically slow growth in health care costs seen
in recent years, this section uses the National Health Expenditure (NHE)
Accounts, which were recently updated by the Office of the Actuary at the
Centers for Medicare and Medicaid Services (CMS) to incorporate data
through 2012 (Martin et al. 2014). These data permit a detailed and comprehensive look at recent trends in the Nation’s health care spending.
The analysis is extended through 2013 using the most recent NHE
projections, which were published by CMS in September 2013 (Cuckler et
al. 2013) and reflect Medicare and Medicaid spending data and macroeconomic data available through June 2013 (CMS Office of the Actuary 2013).1
1 The final health spending growth rate for 2012, as reported by CMS in January 2014, came
in approximately 0.2 percentage points below what CMS had projected in September 2013. To
account for this lower base in 2012, this analysis uses CMS’ projections of the 2013 growth rate
of health spending, not the level of health spending.

150  |  Chapter 4

Table 4–1

Real Per–Capita NHE Annual Growth Rates by Payer and Spending Category

Category

Average annual growth
from 2010 through...

Historical average
annual growth

2012
Total national health expenditures

2013

1960–
2010

2000–
2007

2007–
2010

1.1

1.2

4.6

4.0

1.9

Major payers (per enrollee)
Private insurance

0.9

1.2

N/A

5.2

4.1

Medicare

–0.3

–0.4

N/A

5.5

2.4

–1.6

–0.1

N/A

0.4

0.3

Medicaid

Major categories of spending
Hospital care

1.6

1.6

4.5

4.0

3.2

Physician and clinical services

1.7

1.6

4.6

3.2

1.7

Prescription drugs

–1.1

–1.3

4.6

6.3

0.5

Home health and skilled nursing care

0.9

1.2

6.6

3.0

2.9

Note: Inflation adjustments were made using the GDP deflator. Per-enrollee growth figures are not available for
the 1960-2010 period because Medicare and Medicaid did not exist in 1960 and because CMS does not provide
enrollment by insurance type for years before 1987.
Source: Centers for Medicare and Medicaid Services, National Health Expenditure Accounts and National Health
Expenditure Projections; Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations..

NHE “tracking” estimates constructed by the Altarum Institute using data
on health spending from the Bureau of Economic Analysis imply that final
estimates of NHE for 2013 will come in very close to the CMS projection
(Altarum 2014).
Table 4-1 summarizes recent trends in spending growth, and Figure
4-1 depicts these trends graphically. From 2010 through 2012, the last year
for which final data are available, real per capita national health expenditures
grew at an annual rate of just 1.1 percent. The CMS projections show slow
growth continuing through 2013, with the annual average real per-capita
growth averaging just 1.2 percent. These slow growth rates since 2010 are
less than one-third of the long-term historical average growth rate of 4.6
percent and substantially below the average growth rates recorded from
2000-07 and over the three years immediately prior to 2010.2 These growth
rates since 2010 are, in fact, the lowest on record; from 1960, the first year the
NHE data are available, through the present, no other two- and three-year
periods saw lower growth rates.
The slow growth is reflected in all three payer categories depicted in
Figure 4-3, which appears on page 157. Real per enrollee spending growth
2 The periods 2000-07 and 2007-10 were chosen as comparison periods in order to facilitate the
discussion in the next section of the role of the 2007-09 recession in driving recent trends.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 151
 
the Affordable Care Act

Box 4-1: Two Measures of Growth in Health
Care Costs: Spending and Prices
This report examines two different measures of growth in health
care costs: growth in the prices of health care goods and services and
growth in total spending on health care goods and services. These two
types of data are useful for answering different questions.
The growth in health care prices tells us how the amount of money
needed to purchase a given amount of health care—a bypass surgery, a
doctor’s visit, or a tablet of aspirin—is changing over time. By contrast,
the growth in health care spending captures not only changes in the
prices of health care good or services (like the price of a doctor’s visit),
but also changes in the quantity of health care goods and services consumed (like the number of doctor’s visits made).
In theory, increases in health care prices (above general price
growth) are unambiguously bad for consumers since they reduce the
amount of health care a consumer can buy with a given number of (real)
dollars. By contrast, increases in health care spending can be good or
bad. If spending rises because consumers are receiving more care and
that care improves health, then spending increases are a good thing. If,
on the other hand, spending rises because the price of care is rising or
because consumers are receiving additional care that does not improve
health, then higher spending is a bad thing. Concern about the long-term
growth in health care spending reflects a belief that much of that growth
reflects higher prices or increased use of low-value care.
In practice, measuring changes in health care prices is more challenging than in the idealized discussion presented above. In light of the
rapid technological change that has been seen in the health care sector,
comparing goods and services over time can be difficult. For example,
an appendectomy done in 1990 and an appendectomy done in 2010
might be treated as the “same item” in a health care price index, but
it is likely that the 2010 version of the procedure reflects substantial
improvements in surgical technique relative to its 1990 counterpart,
improvements in quality that may be important for health outcomes and
of great value to patients. As a result, simply knowing that the price of an
appendectomy has risen from 1990 to 2010 is not enough to determine
whether someone in need of an appendectomy was better off in 1990 or
in 2010; one must somehow account for the fact that the 2010 patient is
effectively purchasing a greater quantity of “improved health” than the
1990 patient.
Cutler et al. (1998) document that these measurement challenges
are a substantial problem in practice. Focusing on care for heart attack
patients, the authors show that mortality outcomes for these patients

152  |  Chapter 4

have improved dramatically in ways not accounted for in major price
indices. As a result, these indices dramatically overstate the extent to
which rising medical prices are making people worse off over time.
As a final note, to the degree that statistical agencies have gotten
better at measuring quality improvements over time, long-term comparisons of health care price inflation can be misleading. Indeed, it is
possible that some of the long-term decline in health care price inflation
depicted in Figure 4-2 results from methodological improvements of this
kind. However, methodological improvements of this kind are unlikely
to play a substantial role over short time periods, and they likely play
little or no role in explaining the sharp declines in health care price inflation over the last few years.

in private insurance over the 2010-13 period is less than one-quarter its level
from 2000-07 and less than one-third its level from 2007-10. The change in
Medicare spending growth has been similarly dramatic, with real growth
in per beneficiary Medicare costs essentially ceasing over this period. In
Medicaid, the already slow growth in real per beneficiary costs seen in recent
years has continued and turned slightly negative from 2010 to 2013.
The slowdown is similarly broad-based when looking across spending
categories. Real per capita growth in spending on hospital care—the largest
single category of spending, accounting for almost one-third of total spending—is growing at less than half the long-term historical average rate and
more than 1 percentage point slower than the most recent historical period.
Prescription drugs have seen particularly sharp reductions in growth, with
spending actually shrinking in real per capita terms at a 1.3 percent annual
rate over the last three years. Physician and clinical services and home health
and skilled nursing care show similarly slow growth rates in a historical
context.
Panel A of Table 4-2 documents a similar slowdown in the growth
of prices paid for health care goods and services, which is also depicted
in Figure 4-2. Health care inflation, whether measured using the personal
consumption expenditure (PCE) price indices or the CPI for medical care,
is running at half or less the rate seen historically, and below the rates seen
over the last decade. Indeed, year-over-year inflation as measured using PCE
data is currently running at around 1 percent, a level last seen in 1963. The
recent behavior of the CPI for medical care is similar, with recent months’
year-over-year inflation rates reaching low levels not seen since 1972.
It is important to note that this slow growth in prices for health care
goods and services is not simply a reflection of the fact that the prices of all
goods and services have grown slowly in recent years. Panel B demonstrates
Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 153
 
the Affordable Care Act

Figure 4-1
Growth in Real Per Capita National Health Expenditures, 1961–2013
Annual percent change
8
7
6
5
4
3
2013

2
1
0
1960

1970

1980

1990

2000

2010

Note: Data for 2013 is a projection.
Source: Centers for Medicare and Medicaid Services, National Health Expenditure Accounts;
Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

Figure 4-2
General and Health Care Price Inflation, 1960–2013

12-month percent change
14
12

PCE for Health Care Goods and Services

10
8
6

Dec-2013

4
2
PCE for All Goods and Services

0
-2
1960

1970

1980

1990

2000

2010

Source: Bureau of Economic Analysis, National Income and Product Accounts; CEA
calculations.

154  |  Chapter 4

Table 4–2
Recent Trends in Several Indicators of Health Care Spending and Price Growth
Annual
Fre- Available growth,
quency through ACA–
present

Category

Historical average
annual growth
1960–
ACA

2000–
2007

2007–
ACA

Panel A: Health care inflation
PCE prices for health care goods & services

Monthly Dec–13

1.7

5.4

3.3

2.8

CPI for medical care

Monthly Dec–13

2.9

5.9

4.3

3.5

Panel B: Health care inflation relative to general price inflation
PCE prices for health care goods & services

Monthly Dec–13

0.1

1.7

1.0

1.2

CPI for medical care

Monthly Dec–13

0.8

1.7

1.6

1.8

Panel C: Employer premiums for family coverage (adjusted for inflation)
KFF/HRET survey

Annual

2013

4.1

N/A

6.8

3.0

MEPS–IC

Annual

2012

3.7

N/A

6.4

3.4

Panel D: PCE spending on health care goods & services (adjusted for inflation and population)
PCE spending for health care goods & services Monthly Dec–13

2.2

4.7

3.9

1.4

Note: For monthly data, end points for periods starting or ending in a listed year are treated as occurring in July
of that year. Time periods listed as starting or ending with the ACA start with March 2010 for monthly series and
2010 for annual series. PCE stands for personal consumption expenditures. PCE for health care goods and services
includes the following categories of spending: health care, pharmaceutical and other medical products, therapeutic
appliances and equipment, and net health insurance. Price indices for these categories are combined to construct a
Fisher index for the aggregate, and it is growth in this index that is reported in Panel A and Panel B. In Panel D, the
PCE spending data are adjusted for inflation using the general PCE deflator and BEA’s population series. CPI stands
for consumer price index. Employer premium growth is adjusted for inflation using the GDP deflator. Because
MEPS-IC data are not available for 2007, the figures shown for that series reflect average growth rates for the period
2000-2006 and 2006-2010.
Source: Bureau of Economic Analysis, National Income and Product Accounts; Bureau of Labor Statistics, Consumer Price Index; Kaiser Family Foundation, Employer Health Benefits Survey; Agency for Healthcare Research
and Quality, Medical Expenditure Panel Survey, Insurance Component; CEA calculations.

that health care inflation relative to general price inflation has also been
unusually low over the last few years.
Panel C of Table 4-2 examines trends in employer premiums, as
documented in two major surveys of employers. In both surveys, premium
growth rates are more than 2.5 percentage points below the 2000-07 trend.
Panel D tracks real per capita consumption spending for health care goods
and services, based on data from the Bureau of Economic Analysis. By this
measure, spending growth is running at about half the rate seen in the first
portion of the last decade, and even farther below its longer-term historical average. While these series do suggest that growth may have increased
slightly since 2010, they are consistent with the other available data in
showing that current growth rates are very low, whether measured against
short-term or long-term historical experience. In addition, premium growth
Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 155
 
the Affordable Care Act

in particular may not exactly track underlying cost trends on a year-to-year
basis because premiums must be set before actual costs for the year are
known. Over the long-run, however, slower growth in health costs will likely
be fully reflected in the premiums individuals and employers pay.

What is Happening Now, and
What Will Happen Next?
A natural—and important—question becomes: What is driving the
recent slow growth in health care costs? The answer to this question can
shed light on whether the current slow growth will last, and what policies
could help make that occur. Indeed, slowdowns can be temporary; the earlyand mid-1990s also saw several years of slow growth in health care costs, but
costs accelerated once again in the late 1990s and early 2000s.
While final conclusions about the causes of the recent slow growth
and its persistence await additional data and analysis, some conclusions
are possible with the data currently available. Most importantly, the recent
slow growth does not appear to be the result of idiosyncratic factors affecting a single category of spending or a particular payer. As documented in
Table 4-1, the slowdown has affected all major payers and each of the major
categories of spending. The search for explanations must, therefore, look for
factors affecting behavior system-wide. The first part of this section examines the role of the 2007-09 recession, the second part discusses potential
non-ACA, non-recession explanations for the recent slow growth, and the
third part considers the role of the Affordable Care Act, both to date and in
the future.

The Role of the 2007-09 Recession
Some have identified the 2007-09 recession and its aftermath as a
potential driver of system-wide changes. For example, job losses may have
caused reductions in insurance coverage that curtailed access to health care,
or the accompanying falls in families’ disposable incomes could have forced
households to prioritize other needs over medical care. Alternatively, disruptions in financial markets could have depleted providers’ cash reserves
or reduced their ability to borrow in order to invest in new equipment or
facilities, leading to lower utilization in subsequent years.3 If the recession
3 The NHE data do show a very sharp reduction in investment in equipment and structures
in the health care sector over 2009 and 2010 of about 12 percent in real per capita terms. It
is worth noting, however, that this contraction followed two years of very strong investment
growth. Moreover, even as financial conditions have normalized, investment has remained
subdued, suggesting that providers do not view themselves as having incurred a substantial
investment deficit, nor suggesting an imminent investment-driven rebound in health care cost
growth.

156  |  Chapter 4

Figure 4-3
Growth in Real Per Enrollee Health Spending by Payer

Average annual percent change
6
5.2
5
4

4.1

2000-2007
2007-2010

3
2
1

5.5

2010-2013

2.4
1.2

0.4

0
-1

-0.4
Private Insurance

Medicare

0.3
-0.1
Medicaid

Notes: Figures for 2013 are projections.
Source: Centers for Medicare and Medicaid Services; Bureau of Economic Analysis, National Income
and Product Accounts; CEA calculations.

were the primary driver of the current slow growth in health spending, then
growth would likely return to its earlier rapid rate as the economic recovery
continues.
Three features of the recent slow growth in health care costs are
inconsistent with the theory that this slow growth results only from the
recession, suggesting that a substantial portion of the recent slowdown is
“structural” and likely to persist. First, and most simply, the slowdown has
now persisted well beyond the end of the recession. The Great Recession
began in December 2007 and concluded by June 2009. Since that time, the
economy has recorded four years of steady growth. Yet, as shown in Table
4-1 and Figure 4-3, health spending growth has remained subdued relative
to the years during and immediately following the recession. While the
economy may affect health spending with a lag, if the recession were the
primary force driving the slowdown, more substantial acceleration would
likely be visible by now.
Second, as documented in Table 4-1 and Figure 4-3, the slowdown
has affected Medicare in addition to the private sector, a fact highlighted
in a recent analysis by CBO economists (Levine and Buntin 2013). Because
seniors are generally more insulated from a weak labor market, this fact
undermines the notion that the slowdown results primarily from economic
disruptions attributable to the recession. In addition, Levine and Buntin
Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 157
 
the Affordable Care Act

find that even those seniors who did experience relatively larger economic
disruptions during the recession did not spend less on health care. Levine
and Buntin also document, using State-level data, that Medicare spending
growth has historically risen when unemployment rises—the opposite of
the pattern required for the economic downturn to explain the slowdown
in cost growth.4
Third, the recent behavior of health care inflation is difficult to square
with the theory that the slowdown is primarily a result of the recession. As
documented in Table 4-2 and Figure 4-2, health care inflation has decelerated sharply of late, even when measured relative to inflation in the broader
economy. While there are a variety of plausible mechanisms by which the
recession could reduce the quantity of health care services people demand,
and thus reduce total spending, it is difficult to explain why a recession
should cause a reduction in the growth rate of health care prices relative to
price growth in other sectors of the economy.
Many recent studies have also attempted to directly quantify the
role of the recession in driving recent slow growth in health care spending.
These analyses, which use a variety of methods, have generally concluded
that, while the recession likely has depressed health care spending growth in
recent years, health spending is low in historical terms even after accounting
for the recession, and a substantial fraction of the slowdown likely reflects
structural changes that are likely to persist. The remainder of this section
provides a review of this growing literature.
Chandra, Holmes, and Skinner (2013) provide one approach to evaluating the role of the recession. They survey the available micro-econometric
estimates of the effect of income on the demand for health care. Virtually
all such estimates in the existing literature are small, with the largest credible estimates of the income elasticity being the 0.7 estimate provided by
Acemoglu et al. (2013). Applying this upper-bound estimate to the observed
slowdown in GDP growth, they show that the slow economic growth in
recent years explains less than half of the recent slow growth in health
spending. Although they express some uncertainty about the future outlook
for health spending, they nevertheless project that a substantial fraction of
the slowdown will persist, due in part to the potential of payment reforms
included in the Affordable Care Act.
Ryu et al. (2013) take another approach. They examine the role of
two specific mechanisms by which the recession could have affected health
4 The 2013 Economic Report of the President undertakes a related analysis (CEA, 2013). The
report analyzes changes in state-level unemployment from 2007-09 to state-level health
spending growth over that period. While that analysis finds that unemployment is associated
with lower health spending growth, the effect is small and cannot explain a substantial fraction
of the recent downturn in health spending.

158  |  Chapter 4

care cost growth: by reducing insurance coverage via job loss and by causing
firms to offer their employees leaner health plans that require greater costsharing. Focusing on the period 2009-11, they find that recent reductions in
spending growth are, if anything, larger among employed individuals, and
that increases in cost-sharing can account for only one-fifth of the slowdown. On the basis of their results, they advise a “cautious optimism that the
slowdown in health spending may persist.”
Another set of studies evaluates the effect of the recession by estimating the historical relationship between economic growth and health spending growth and using this estimated relationship to simulate how health
spending would have evolved had the recession not occurred. Econometric
time series analyses like these have the important advantage that, by virtue
of their nationwide, aggregate approach, they can capture the effects of a
wide variety of potential mechanisms connecting economic growth to health
spending growth. But the nationwide, aggregate nature of these analyses
is also a weakness; it can be difficult to plausibly control for important
confounding factors, and the paucity of data (only about 50 years of data,
or about 50 total data points, are available) can make these analyses sensitive to seemingly innocuous changes in methodology, as demonstrated by
Chandra, Holmes, and Skinner. The current literature does not, unfortunately, provide persuasive evidence on which econometric specifications are
likely to provide the most reliable results.
Cutler and Sahni (2013) estimate a model relating current health
spending growth to a five-year average of economic growth. Based on
their results, they estimate that spending growth in 2011 and 2012 would
have been on the low end of the historical range even accounting for the
recession, and that more than half of the slowdown over the longer period
2003-12 is due to factors other than the recession. They conclude that “fundamental changes” are underway in the health sector, changes that are not
attributable to the recession alone.
A contrary perspective comes from an analysis from the Kaiser Family
Foundation and the Altarum Institute (KFF and Altarum 2013). They estimate a model relating current health spending growth to economic growth
the current year and each of the prior five years and general price inflation
in the current year and each of the prior two years. On the basis of their
estimated model, the authors conclude that most of the slowdown in health
care spending from the 2001-03 period to the 2008-12 period is attributable
to the macroeconomic factors, although they still attribute 23 percent of the
slowdown to non-macroeconomic factors.
It is important to note, however, that the authors’ calculation applies
to the slowdown in nominal health spending growth over this period, while
Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 159
 
the Affordable Care Act

the slowdown in real (that is, inflation-adjusted) health spending growth is
of greater economic interest. Because inflation was, on average, lower during
the 2008-12 period than during the 2001-03 period, the authors’ approach
overstates the role of macroeconomic factors in explaining the slowdown
in real spending growth. In addition, the authors’ model, by virtue of its
relative complexity, is particularly subject to the shortcomings of the time
series approach described above. Indeed, the model estimated by KFF and
Altarum has one particularly unusual feature: the effect of reduced economic
growth on health spending actually peaks four years later. While not impossible, such lags seem implausibly long.

Non-ACA Factors Affecting Health Spending Growth
As discussed above, the recession does not provide a full, or even
necessarily a major, explanation for the recent slow growth in health spending. While additional factors may be identified in the future, two non-ACA
factors have received substantial attention to date—although it is important
to note that at least one non-ACA factor is modestly increasing health
spending growth.
The long-term trend toward increased patient cost-sharing is one factor that can plausibly explain why slow growth has affected many different
categories of spending at the same time (Cutler and Sahni 2013; Ryu et al.
2013; Chandra, Holmes, and Skinner 2013). The Kaiser Family Foundation/
Health Research and Educational Trust Employer Health Benefits Survey
indicates that recent increases in cost-sharing in employer plans have been
substantial; the typical deductible in an employer plan has increased from
$584 in 2006 to $1135 in 2013, a 70 percent increase after adjusting for inflation (Kaiser Family Foundation 2013a).
Some research suggests that the observed increase in cost-sharing is
having an effect. As noted above, Ryu et al. (2013) examine the importance
of increased cost-sharing in the employer context and conclude that it can
account for 20 percent of the reduction in growth over the 2009-11 period.
Chandra, Holmes, and Skinner (2013) evaluate the role of increased costsharing using estimates from the literature of how utilization responds to
cost-sharing. They conclude that cost-sharing may have played a larger role,
although the precision of their estimates is limited by the poor quality of the
available data on recent changes in cost-sharing and the current incomplete
understanding of how cost-sharing affects utilization.
While it seems possible and perhaps likely that increased cost-sharing
is playing a role, it cannot be the whole story. As discussed in detail above,
the slowdown in Medicare fee-for-service spending has been even more
dramatic than the slowdown in the private sector, and there have been no
160  |  Chapter 4

substantial changes to the core Medicare benefit design in recent years that
parallel the changes seen in the private sector.
The striking slowdown in prescription drug spending, documented in
Table 4-1, also factors into the slow growth trend. Various sources attribute
this sharp drop in prescription drug spending to the expiration of patent
protection for many important drugs. Due to a slowdown in the invention
of new drugs that stretches back more than a decade, the drugs that have
come off patent in recent years are not being replaced by more-recently patented drugs. As a result, the share of prescriptions accounted for by generic
drugs—which typically cost much less—has increased sharply, substantially
reducing costs (Aitken, Berdnt, and Cutler 2009; Cutler and Sahni 2013;
IMS 2013). However, these changes in the prescription drug market are
probably only making a modest contribution to aggregate trends in health
spending since prescription drugs account for less than 10 percent of total
health spending.
There is, however, at least one easily identified factor working against
the recent slowdown: the aging of the U.S. population. In recent decades,
population aging has made a small positive contribution to the growth of
U.S. health spending; White (2007) estimates that over the period 19702002, population aging added about 0.3 percentage points to annual growth.
The contribution of population aging to health care spending growth
appears to have increased by a small amount in recent years. Using data on
the age distribution from the U.S. Census Bureau, data on spending by age
reported by Yamamato (2013), and a methodology similar to that used by
Yamamato, the CEA estimates that population aging added about 0.5 percent to annual growth in health care spending over the 2000-07 and 2007-10
periods and added about 0.8 percent to growth over the 2010-13 period.5
These demographic headwinds mean that the slowdown in the growth of

5 As Yamamato notes, this methodology assumes that spending does not change
discontinuously at age 65 when individuals transition to Medicare. It also does not account
for differences in coverage mix by age in the under-65 population. It does not appear that
accounting for these factors would meaningfully alter the results, but further research in this
area would be worthwhile.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 161
 
the Affordable Care Act

health care costs for an individual of any particular age is actually slightly
larger than shown in Table 4-1.6

The Role of the Affordable Care Act
The evidence discussed above shows that the recession is not the sole
cause of the recent slow growth in health spending, and that the other factors identified to date cannot explain the magnitude or broad scope of the
slowdown. What, then, is the Affordable Care Act’s role in driving changes
in the Nation’s health care system? To be sure, the ACA is not the sole
cause of the slowdown. Health care spending growth had slowed somewhat
even before the ACA was passed (as shown in Table 4-1), the recession and
other changes in the health system have certainly made contributions (as
discussed above), and many of the ACA’s reforms have yet to take full effect.
Nevertheless, the ACA’s reforms aimed at driving out waste and
improving quality are contributing in a meaningful way to recent slow
growth in health costs—including by building on pre-existing trends in
delivery system reform and initiating new ones—and are likely to make
larger contributions in the future. Recent economic research also provides
support for the premise that implementing reforms in Medicare can reduce
the cost and improve the quality of care system-wide. This research supports
the idea that the ACA will play an important role in slowing health care

6 The effect of changing demographics on per beneficiary spending by particular payers may
differ from the effect on the overall population. The average age of Medicare beneficiaries is
currently falling as the youngest baby boomers reach age 65. Consistent with that, Levine and
Buntin (2013) estimate that changes in the age mix of Medicare beneficiaries had no effect
on per beneficiary growth in Medicare spending over the 2000-07 period, but subtracted 0.2
percentage points over the 2007-10 period. Calculations like those in the main text suggest
that these changes in age mix subtracted somewhat more from growth, on the order of 0.4
percentage points, over the 2010-13 period. This represents a modest, but not trivial, share of
the overall slowdown in Medicare spending growth.
In addition, changes in beneficiary mix (that are not primarily attributable to the aging of the
population) appear to have had a larger effect on recent trends in per beneficiary Medicaid
spending. Over the period 2000-10, Medicaid enrollment among children, parents, and
pregnant women increased substantially more rapidly than did enrollment among elderly and
disabled individuals (Kaiser Family Foundation, 2013b). The resulting change in enrollment
mix lowered per beneficiary costs since non-elderly, non-disabled beneficiaries generally use
less health care. Holahan and McMorrow (2012) estimate that this change in enrollment mix
subtracted 1.5 percentage points from the annual growth of Medicaid spending per beneficiary
spending over the 2000-10 period. However, enrollment data reported by the Kaiser Family
Foundation suggest that, if anything, changes in enrollment mix have actually increased per
beneficiary costs since 2010. Thus, adjusting for enrollment mix would make the slowdown in
per beneficiary Medicaid costs over the 2010-13 period more dramatic than shown in Table
4-1.

162  |  Chapter 4

Box 4-2: How Will the ACA’s Coverage Expansion
Affect Total Spending Growth?
As the Affordable Care Act’s coverage expansion takes effect, total
national health care spending will likely grow at an elevated rate for a
few years, reflecting the cost of covering an additional 25 million people
(Cuckler et al. 2013; CBO 2014). This one-time increase in costs is more
than justified by the benefits of bringing quality, affordable health insurance coverage to millions of Americans who lack this protection today.
So the additional cost is neither a surprise, nor a cause for concern.
These increases in total national health expenditures are also
not directly relevant for most individuals and employers, for whom
what matters is how much they are paying in premiums or other costs.
When a previously uninsured person purchases coverage through the
Marketplace or receives it through Medicaid, that does increase total
national health expenditures, but it has no direct effect on costs for
someone who previously had coverage through their employer or the
individual market.
Moreover, one-time changes of this kind will tell us nothing about
the underlying trend in health spending, and it is this underlying trend
that, as discussed in Section 3, will shape Americans’ living standards
over the long run. In addition, the ACA’s Medicare reforms are slated to
continue to phase in over years beyond 2014, and the ACA’s mechanisms
for generating new innovative reforms aimed at reducing costs and
improving quality are just beginning to generate results. As a result,
the savings from these and other aspects of the ACA are likely to grow
substantially in the years ahead. This is an important reason why the
Congressional Budget Office estimates that the extent to which the ACA
will reduce the deficit grows dramatically over time (CBO 2012b).
It is also worth noting that the projected increase in growth over
the next few years is not particularly large. Even after accounting for
transient effects attributable to the ACA’s coverage expansion, CMS
projects that annual real per capita growth in national health expenditures will never exceed 3.4 percent over the next decade. As shown in
Table 4-1, these rates are below the average growth rate recorded over
the 2000-07 period and far below the longer-term historical average.

cost growth over the long term, but also suggests that its provider payment
reforms may be having a larger-than-anticipated impact today.
Reductions in excessive Medicare payments to providers and health
plans. The ACA has already had one easily quantifiable effect on the nation’s
health care spending: reducing excessive payments previously identified by
independent experts (for example, MedPAC (2009)). The original CBO cost

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 163
 
the Affordable Care Act

estimate for the ACA found that its reforms to Medicare would save $17
billion in fiscal year 2013, attributable primarily to reductions in payments
to private insurers that provide coverage through Medicare Advantage and
adjustments in annual updates to Medicare provider payment rates (CBO
2010a).7 Estimated savings of $17 billion constitute about 0.6 percent of
national health expenditures in 2013. Spread out over the three years from
2010 to 2013, this implies that the ACA alone accounts for a 0.2 percentage point reduction in the growth of national health expenditures over this
period, making a meaningful contribution to explaining the slow growth in
health spending observed over these three years. The analysis by Cutler and
Sahni (2013) reaches similar conclusions. These reductions will continue to
phase in over the years ahead and continue to reduce the growth of Medicare
spending.
Deployment of new payment models. The ACA also includes many
reforms intended to identify and promote payment models that encourage
efficient care delivery, reduce care fragmentation, and reward physicians,
hospitals, and others that invest in providing high-quality care rather than
just a high quantity of care.
The ACA made direct changes in Medicare payment systems aimed
at achieving these goals, including creating the readmissions reduction and
shared savings programs discussed in detail below and various “value-based”
purchasing initiatives that tie provider reimbursement to measures of the
quality of the care received by patients. The ACA also provided additional
financial assistance to states through Medicaid to establish health homes to
improve care management for patients with chronic conditions.
In addition, the ACA created the Center for Medicare and Medicaid
Innovation (the “Innovation Center”) to experiment with diverse new payment approaches, including bundled payments, various accountable care
models, and multi-payer initiatives, each of which will be touched on later
in this section. To date, more than 50,000 health care providers from across
every state are participating in an Innovation Center initiative. The Secretary
of Health and Human Services has the authority to take successful pilots to
scale.
Finally, through the Patient-Centered Outcomes Research Institute,
the ACA is funding efforts to identify which treatments work—and for
which patients—and to identify strategies for translating that evidence into
practice. By giving providers the information they need to provide efficient,
7 This chapter also cites a CBO estimate of the budgetary effect of repealing the ACA from July
2012, which suggests that repeal would increase Medicare spending in FY 2013 by $4 billion,
a much smaller sum than the $17 billion cited here. However, as discussed in the CBO letter,
because it would have been too late to unwind some ACA provisions for FY 2013 and due to
other effects, this estimate does not reflect the full effect of the ACA in that year.

164  |  Chapter 4

high-quality care, this research initiative directly complements the ACA’s
efforts to change provider incentives.
The full benefits of the initiatives described above will only be realized
in the years to come. However, the next two subsections discuss a pair of
payment reforms—the ACA’s incentives to reduce hospital readmissions
and its deployment of accountable care payment models—that are already
beginning to show results.
Incentives to reduce hospital readmissions. The ACA made important changes in how Medicare’s hospital payment system treats hospital
readmissions—cases in which a patient returns to the hospital soon after
being discharged. Historically, nearly one-in-five Medicare patients were
readmitted within 30 days of discharge, and it is commonly believed that
many of these readmissions result from low-quality care during the initial admission or poor planning for how the patient will obtain care after
discharge. Prior to the ACA reform, hospitals faced no financial incentive
to invest in activities aimed at reducing readmissions, and could actually
be made financially worse off by doing so since they lose payment for the
avoided readmissions. This misalignment of incentives likely both increased
costs and reduced quality.
The ACA aims to correct these incentives by penalizing hospitals with
high readmission rates for patients with a specified set of diagnoses. Many of
the rules governing these penalties were finalized in August 2011. The penalties took effect at the start of FY 2013 (October 2012), but because penalties
for a given fiscal year are based on hospitals’ readmission rates in prior years,
hospitals’ incentives to begin reducing readmissions began as soon as the
rules were finalized (or earlier, to the extent that hospitals anticipated the
structure of the payment rules).8 The number of conditions included in the
program and the maximum penalty amount will grow over time.
Figure 4-4 provides evidence that this readmission policy has begun
changing patterns of care. After having been flat for several years, overall
30-day hospital readmission rates for Medicare patients turned sharply
lower soon after the program rules were finalized, and, as of July 2013, were
more than one percentage points below their average level from 2007-11.
From January 2012 through August 2013, this reduction corresponded to
130,000 avoided readmissions (CMS, 2013a). The sharp change in trend—
and its timing—implies that the readmissions program played an important
role in causing these changes, although other efforts to reduce readmissions
were underway during this period as well. Among those other activities were
efforts by the Department of Health and Human Services efforts to actively
8 Under current program rules, a hospital’s penalties in a given fiscal year are based on its
readmission rate during the three-year period that ended five quarters earlier.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 165
 
the Affordable Care Act

Percent of patients
20

Figure 4-4
Medicare 30-Day, All-Condition Hospital
Readmission Rate, 2007–2013

2007-2011 Average

19

Jul-2013

18

17
2007

2008

2009

2010

2011

2012

2013

Notes: Recent months are based on preliminary data. The dotted blue lines depict the range in which
the final estimates are likely to fall.
Source: Centers for Medicare and Medicaid Services, Office of Information Products and Data
Analytics.

engage hospitals and community-based organizations in improving discharge processes through the Partnership for Patients and the CommunityBased Care Transitions Program (Gerhardt et al. 2013).
Accountable care payment models. Another important ongoing
ACA reform is the creation of “accountable care” payment models through
the Medicare Shared Savings Program and the Innovation Center. These
programs seek to realign provider incentives to encourage provision of
efficient, high-quality care. Under fee-for-service payment systems, providers delivering more efficient care often end up financially worse off because
lower service volume translates into lower payments from Medicare. In
addition, since provider payments were based on service volume, the preACA payment system gave providers no direct financial incentive to deliver
high-quality care. Prevailing fee-for-service payment systems also pay each
provider separately without regard to how services furnished by that provider fit into the patient’s broader plan of care, and thus create no incentive
for efficient coordination of care across providers.
Under these accountable care programs, a provider or group of providers can seek designation as an Accountable Care Organization (ACO).
ACOs are eligible to share in the savings created when they reduce the cost
of caring for patients assigned to them, which encourages providers to be
efficient in the use of additional services. In addition, because the ACOs earn
166  |  Chapter 4

shared savings based on the total costs of a patient’s care across all providers
and not merely the costs for any particular visit or procedure, ACOs have
incentives to invest in care coordination and avoid duplication. Perhaps
most important, ACOs must achieve designated benchmarks for the quality
of care received by their patients in order to be eligible for shared savings,
which provides strong incentives to ensure that patients receive high-quality
care.
Today, more than 360 organizations serving 5.3 million Medicare
beneficiaries have adopted the ACO model, and the number of beneficiaries
covered will likely grow in the years ahead. A preliminary evaluation of the
Pioneer ACO program (the Innovation Center ACO program for large and
advanced systems) found that costs for beneficiaries aligned with Pioneer
ACOs grew more slowly from 2011 to 2012 than costs for similar beneficiaries not aligned with ACOs (L&M Policy Research, 2013). The annual cost
savings for each enrollee aligned with a Pioneer ACO in 2012, the first year
of the program, were estimated to be at least $150, more than 1 percent of
average Medicare spending per beneficiary in that year. In addition, overall,
the ACOs performed better than fee-for-service benchmarks on all quality
measures for which comparable data are available (CMS 2013b). Academic
research on similar private models also suggests that these payment models
can achieve their intended purpose of reducing costs while improving quality (Song et al. 2012).
The Innovation Center is experimenting with related payment models through its Bundled Payment for Care Improvement Initiative, which
got underway in 2013.9 Under these models, Medicare will make a single
“bundled” payment for all services provided during an “episode” of care
connected to a hospital stay, rather than paying separately for each service
provided during that episode. In the model using the most comprehensive
bundle definition, this payment will cover the hospital stay, physician
services provided during the stay, and post-hospital care. The Innovation
Center is also testing models with narrower bundles covering only services
provided during the hospital stay or only services provided after the hospital
stay. Although the details vary across payment models, the bundled payment will then be allocated across the participating providers according to

9 These models build on several earlier Medicare demonstration projects, with the most similar
being the Acute Care Episode (ACE) demonstration, a much smaller demonstration that
concluded in 2013.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 167
 
the Affordable Care Act

agreements among the providers themselves.10 The models are being tested
for a set of common types of hospital care episodes that account for a significant fraction of all hospital stays.
Much like the accountable care payment models, these bundled payment models encourage providers to be more efficient because providers
receive no additional payment for providing additional services (if the
service is included in the bundle). Similarly, because all providers involved
in an episode of hospital care are jointly accountable for the total cost of the
care episode, the bundled payment structure gives providers strong incentives to coordinate their activities, with attendant benefits for efficiency and
quality of care. Because of this scope for increased efficiency, Medicare can
(and does under the models being tested) set the bundled payment amount
below the total amount it would pay under the existing fee-for-service payment systems. The efficiency gains from these sources could be substantial.
CBO recently estimated that if a bundled payment model that covered services provided during and after the hospital stay and used a 5 percent savings
target were phased in nationwide starting in 2017, the savings to Medicare
would total $47 billion over 10 years (CBO 2013c).
Recent research on cross-payer “spillovers.” In evaluating the direct
effects of the ACA’s Medicare and Medicaid reforms so far and considering their likely effects going forward, an important question is how these
reforms will influence the rest of the health care system. Recent empirical
work in economics and health policy strengthens the premise that reforms
to public-sector health programs that reduce waste and improve quality will
have spillover benefits for the private sector.11
10 The bundled payment is administered in different ways under the different models. In
the model covering only services during the hospital stay, the bundled payment is paid
“prospectively” to a single entity (e.g. the hospital), which is then responsible for paying the
other providers involved in episode. In the other models, Medicare continues to pay providers
according to the existing fee-for-service rules. If total fee-for-service payments are below the
bundled payment amount, Medicare pays the excess to a designated provider, which distributes
that excess among the other involved providers. If total fee-for-service payments are above the
bundled payment amount, the reverse occurs. In principle, the two structures change provider
incentives in similar ways.
11 This growing literature is contrary to the traditional view in some health policy circles, which
held that efforts to achieve savings in Medicare (or Medicaid) cause medical care providers
to increase the prices they charge to private insurers in order to recover the lost revenue,
and, thus, reforms in Medicare simply “shift” costs to the private sector rather than reducing
them. The empirical support for this view was always inconsistent, and, as argued by Dranove
(1988) and Morrissey (1994), this view has important conceptual shortcomings. In particular,
for hospitals to be able to increase the prices charged to private payers after a reduction in
Medicare payment rates, they must have been willingly charging a price below what the market
would bear prior to the reduction in Medicare rates. For a comprehensive overview of this
literature, particularly the older literature, see Frakt (2011; 2013).

168  |  Chapter 4

In particular, various recent studies suggest that efforts by Medicare
to reduce excessive payments for particular services are likely to generate
corresponding savings for private insurers and their enrollees. Clemens and
Gottlieb (2013) study how the prices that private insurers pay to physicians
change when Medicare changes its prices, exploiting a natural experiment
created by regional differences in the effect of earlier reforms to the way
Medicare pays physicians. They find that when Medicare reduces the price
it pays for services, private insurers are able to reduce the amount they pay
for care by similar amounts.
White (2013) and White and Wu (2013) undertake a similar analysis
focused on Medicare payment to hospitals that exploits natural experiments
created by cross-hospital differences in the effect of earlier Medicare payment changes. White (2013) finds that when Medicare reduces its payment
rates, private payers reduce their payment rates by approximately 77 percent
of that amount. White and Wu (2013) find that for each dollar of Medicare
savings, private insurers realize additional savings of 55 cents.
The implications of these estimates are striking. For example, the $17
billion in Medicare savings estimated to have been achieved in FY 2013 as
a result of reducing excessive Medicare payments. Using the same logic
applied previously, these estimated savings correspond to a 0.2 percentage
point reduction in the average growth of health care prices over the period
2010-13. If just half of these price reductions spilled over to the rest of the
health care system to the extent estimated by White (2013), then the implied
reduction in health care inflation economy-wide due to these Medicare
changes would be about 0.5 percent.12 In this scenario, the ACA would be
playing a significant role in driving the observed slow growth in health care
prices—representing about half of the recent slowdown in health care inflation relative to general price inflation.13
Potentially even more important, the work by Clemens and Gottlieb
provides evidence that the benefits of the ACA’s improvements to the
structure of public-sector payment systems may be realized system-wide,
not just among enrollees of those programs. Again focusing on Medicare
12 The reductions in excessive payments to Medicare Advantage plans are less likely to
“spill over” to general private-sector payment rates (although to the extent they lead
MA-participating insurers to negotiate lower provider payment rates, such spillovers could
occur under certain models of spillovers). Since the Medicare Advantage reductions account
for about half of the estimated $17 billion in payment reductions in 2013, the calculation in the
text assumes that only half of this reduction would spill over.
13 Of course, effect on total spending may be smaller or larger to the extent that these price
changes induce changes in volume. Indeed, the estimates of White and Wu, referenced above,
as well as estimates reported by He and Mellor (2012) suggest that volume changes will
generally work to offset these price spillovers. However, even under the estimates of White and
Wu (2013), the savings to private insurers as a result of Medicare changes would be substantial.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 169
 
the Affordable Care Act

payment for physician services, they show that Medicare payment changes
that increase payment for some services and reduce payment for others tend
to be matched by private insurers. Clemens and Gottlieb’s results provide
empirical support for the widely believed notion that Medicare’s payment
structure serves as the “starting point” in negotiations between providers and private insurers in many circumstances, in which case changes in
Medicare will reasonably quickly get picked up in the private sector as well.
This evidence is consistent with historical experience. Medicare introduced
“prospective” payment for inpatient services in the 1980s, under which all
care during an inpatient admission was covered via a single payment determined based on the patient’s diagnosis; virtually all private insurers pay
hospitals using this type of system today.
Some recent evidence suggests that spillover benefits from the ACA’s
public-sector payment reforms may occur even if private payers do not
directly adopt these payment models. McWilliams et al. (2013) study the
Alternative Quality Contract (AQC)—a contract similar to accountable care
payment structures currently being deployed by CMS–that Blue Cross Blue
Shield of Massachusetts has been experimenting with since 2009. Research
cited above (Song et al. 2012) finds that the AQC reduces costs and improves
quality for patients whose care is directly subject to the contract. The
research by McWilliams et al. finds, however, that patients associated with
AQC-participating providers whose care was not subject to the contract (in
this case, Medicare patients) also experienced improvements. In this case,
the cost savings amounted to 3.4 percent, on average, and was accompanied
by improvement on some quality measures. The results may arise because
providers adopt a single “practice style” for all their patients, so that when
incentives from one induce a provider to change its approach in ways that
improve efficiency or quality, all patients seen by that provider benefit.
Taken together, the evidence of cross-payer spillovers reviewed above
suggests that not only are reforms to the structure of the public-sector payment systems helpful in reducing costs and improving quality system-wide,
but that the public sector may be essential to fully realizing the potential
for improvement. In economic terms, the presence of spillovers means that
payment system reforms are “public goods”—investments that generate
benefits for many people other than the purchaser and for which the purchaser cannot capture all the resulting benefits (Clemens and Gottlieb 2013).
Because no individual investor captures the full benefits of investment in
public goods, the private market generates too few of them. As with other
public goods, one solution to the underinvestment is for the government to
invest directly, in this case by implementing reforms through Medicare and
Medicaid.
170  |  Chapter 4

Recognizing the importance of other payers’ decisions in determining providers’ response to new payment arrangements, CMS has launched
demonstration projects that actively engage multiple payers. Incorporating
multiple payers into reform efforts at the outset may increase the possibility that the payment models that emerge can easily cross payer boundaries,
once proven. These initiatives also recognize that engaging private payers in
reform efforts is important for Medicare and Medicaid beneficiaries themselves, in light of the evidence described above that spillovers can run in both
directions: from Medicare and Medicaid to the private sector, and vice versa.
Two multi-payer initiatives merit special mention. Through the
Comprehensive Primary Care Initiative, CMS has enlisted public and private payers in eight states to join with Medicare to invest in primary care
practices, with the potential for shared savings after two years. Another
promising effort is the State Innovation Models Initiative, which provides
grants to states that wish to make statewide, multi-payer changes to provider
payment systems. With support from this program, Oregon has embarked
upon an effort to move its Medicaid beneficiaries, State employees, and
individuals who have purchased coverage through the state’s ACA-created
health insurance marketplace into ACO-like payment models. Arkansas has
undertaken an initiative involving public and private payers aimed at ensuring that half of Arkansans have access to a patient-centered medical home by
2016, and expanding its existing system of episode-based payment.

Economic Benefits of Slow
Health Spending Growth
Slower growth in health care costs has the potential to bring three
important economic benefits: higher living standards; lower deficits, potentially generating faster economic growth; and, at least in the short run,
higher employment. This section of the report considers the implications of
slower growth in health care costs across these variables.

Higher Living Standards
All else equal, when the health sector consumes less of the Nation’s
output, more resources are available for meeting other needs. As a result,
reductions in health care spending that stem from improving efficiency or
eliminating low-value care have the potential to improve living standards.
Because of the large share of the Nation’s resources devoted to health care,
even relatively modest reductions can have very large effects on economic
well-being.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 171
 
the Affordable Care Act

Box 4-3: The Cost Slowdown and ACA Reforms are
Reducing Medicare Beneficiaries’ Out-of-Pocket Costs
As discussed in the text, reductions in Medicare spending growth
have substantial benefits for the Federal budget. Lower growth also has
substantial benefits for Medicare beneficiaries, both because it reduces
their cost-sharing obligations and because many pay a premium to
enroll in Medicare Parts B and D, and premiums are set to cover a specified fraction of the government’s cost of providing that coverage. Due
in large part to the broader trends discussed in this chapter, the base
Medicare Part D premium is down 5 percent in inflation-adjusted terms
relative to 2010 (Figure 4-5). Similarly, the standard Medicare Part B
premium for 2014 is essentially unchanged in inflation-adjusted terms
relative to 2009. (The standard Medicare Part B premium is down 11
percent in inflation-adjusted terms relative to 2010. However, it is more
meaningful to compare to 2009; for technical reasons, many beneficiaries
paid the 2009 premium in 2010 and 2011, and, for these same reasons,
the standard Part B premium is anomalously high in those years (SSA
2013).)
At the same time as Medicare premiums have remained flat, features of the ACA are directly reducing out-of-pocket costs for Medicare
enrollees. Under the ACA, Medicare beneficiaries receive a wide range
of preventive services without cost-sharing requirements. CMS estimates that 34 million Medicare beneficiaries received at least one such
service during 2012 (CMS 2013c). Through a combination of discounts
on brand-name drugs and additional coverage, the ACA is also closing
the “donut hole” in Medicare Part D—a range of drug spending over
which beneficiaries enrolled in the “standard” Medicare Part D plans
were previously required to cover the full cost of their medications.
CMS estimates that 3.5 million Medicare beneficiaries who reached the
coverage gap realized average savings of $706 on brand-name drugs in
2012, while 2.8 million Medicare beneficiaries realized savings of nearly
$40 per person on generic drugs (CMS 2013c).

These benefits accrue to families through two primary channels. First,
standard economics implies that, in the long run, reductions in the cost of
providing benefits such as health insurance are passed through to workers
in the form of higher wages since employers must compete for workers
(Summers 1989). This theoretical prediction has received empirical support (Gruber and Krueger 1991; Gruber 1994; Baicker and Chandra 2006).
Second, as discussed in detail below, lower health care costs have significant
benefits for the Federal budget, which ultimately permit lower taxes or
increased investment in other valued public services.
172  |  Chapter 4

2014 $
140
120

Figure 4-5
Inflation-Adjusted Premiums for Medicare
Parts B and D, 2000–2014
2014
Medicare Part B Standard Premium

100
80
60
Medicare Part D Base Premium

40
20
0
2000

2002

2004

2006

2008

2010

2012

2014

Source: Centers for Medicare and Medicaid Services; Bureau of Economic Analysis, National
Income and Product Accounts; CEA calculations.

One straightforward way of illustrating the magnitude of the potential
impacts is to consider the effect of continuing the slow growth of the last
few years. To that end, recall from Table 4-1 that national health expenditures have grown at a 1.2 percent real per capita annual rate from 2010-13,
whereas health spending grew at a 4.0 percent rate from 2000-07. Suppose
that even just one-third of that slowdown continued, so that instead of
returning to the recent historical rate of 4.0 percent, real per capita health
care costs instead grew at a 3.1 percent rate, similar to the rate projected in
the recent work by Chandra, Holmes and Skinner (2013). Under this illustrative scenario, the savings after a decade would amount to about $1,200 per
person. As discussed above, these savings would materialize primarily in the
form of higher wages and lower State and Federal costs.

Lower Deficits
In 2013, the Federal Government devoted 22 percent of the U.S.
budget, or 4.6 percent of GDP, to Medicare and Medicaid. For this reason,
the future path of health care costs has major implications for the long-term
budget outlook.
Over the last three years, CBO has made a series of downward revisions to its forecast of future spending on Medicare and Medicaid (CBO

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 173
 
the Affordable Care Act

2010a; 2011; 2012c; 2013a; 2014), which are depicted in Figure 4-6. From
the projections CBO published in August 2010 to its most recent set of
projections in February 2014, CBO has reduced its estimate of Medicare and
Medicaid spending in 2020 (the latest year covered by all of the projections
examined here) by $168 billion and 0.5 percent of GDP.14 This $168 billion
represents a 13 percent reduction in spending relative to CBO’s earlier projection of spending on these programs.
These reductions primarily reflect lower projections of future growth
in health care costs.15 To that point, in a recent presentation, CBO Director
Douglas Elmendorf commented: “The slowdown in health care cost growth
has been sufficiently broad and persistent to persuade us to make significant
14 In July 2013, the Bureau of Economic Analysis (BEA) released comprehensive revisions
to the National Income and Product Accounts that increased BEA’s estimate of GDP in
recent years by more than 3 percent. CBO projections of GDP released before and after these
revisions are, therefore, not directly comparable.
The figures reported in the text and displayed in Figure 4-6 account for this issue in the
following manner. For May 2013, CBO released two sets of GDP projections, one before and
one after the BEA revisions; the figures shown use the GDP projections released after the BEA
revisions. For earlier CBO baselines, CEA adjusted CBO’s projections of GDP upward by the
ratio of CBO’s post- and pre-revision May 2013 GDP projections. Without these adjustments,
the reduction in projected Medicare and Medicaid spending as a share of GDP in 2020 from
CBO’s August 2010 baseline to its February 2014 baseline would be 0.6 percent, rather than the
0.5 percent reported in the text, and the decline in Medicare and Medicaid spending shown in
Figure 4-6 would be larger.
15 Several factors other than recent slow growth in health care costs have affected CBO’s
projections of Medicare and Medicaid spending over this period. These factors work in
different directions. First, CBO has revised its general economic projections in ways that, on
net, increase projected future Medicare and Medicaid spending by around $25 billion. Second,
CBO estimates issued after June 2012 incorporate the Supreme Court decision in NFIB v.
Sebelius. CBO materials indicate that this ruling reduced projected Medicaid spending in
2020 by roughly $30 billion as of July 2012, although this figure has likely fluctuated as CBO
has changed its assumptions about how many states will adopt the Medicaid expansion. For
more detailed information, see CBO’s analysis of the budgetary effects of the Supreme Court
decision (CBO 2012c) and CBO’s March 2012 baseline (CBO 2012a). Third, projections issued
in August 2011 and later incorporate the effects of sequestration under the Budget Control Act,
which CBO estimated in May 2013 would reduce Medicare spending by $11 billion in 2020
(CBO 2013a).
CBO itself has cited somewhat larger figures when discussing the extent to which it has revised
down its projections in response to slower health care cost growth. For example, CBO recently
reported that slower growth in health costs has led it to revise down its estimate of Medicare
spending in 2020 by $109 billion since March 2010 (CBO 2014), whereas the comparable figure
based on the approach in the text is $87 billion. CBO’s figure is larger because it excludes
the changes due to updated economic projections discussed above, because it considers a
slightly different time period, and because its figure appears to apply to gross, rather than
net, Medicare spending. On the other hand, CBO’s figure excludes the effect of sequestration,
which partially offsets these differences. The estimates presented in the text were chosen over
the estimates presented by CBO to simplify exposition and presentation.

174  |  Chapter 4

Box 4-4: Premiums on the ACA Marketplaces
are Lower than Projected
The Congressional Budget Office recently reported that actual
2014 premiums on the ACA Marketplaces are about 15 percent below
its earlier estimates (CBO 2014). This has two important benefits. First,
lower premiums will mean lower costs for many families, including those
with incomes too high to qualify for premium tax credits and those that
wish to purchase more comprehensive coverage than that offered by the
second-lowest cost silver plan. Second, lower premiums will result in
lower Federal costs for premium tax credits and cost-sharing assistance.
While CBO states that it has not yet decided whether to mark down
its premium estimates for years beyond 2014, estimates by Spiro and
Gruber (2013) suggest that such a revision would result in Federal savings of more than $100 billion over ten years.
While it is not yet fully understood why premiums on the ACA
Marketplaces are lower than expected, this may be another benefit of
the recent slow growth in health care spending. The Marketplaces may
also have proved better than expected at encouraging insurers to compete on price (Spiro and Gruber 2013). A related possibility is that the
Marketplaces attracted greater-than-expected participation by insurers;
premiums appear to be substantially lower in areas with more participating insurers (ASPE 2013).

downward revisions to our projections of Federal health care spending”
(Elmendorf 2013).
For comparison, in CBO’s most recent long-term budget outlook,
CBO projected that the current law 25-year fiscal gap—a measure of the
annual fiscal adjustment required to stabilize the debt as a share of the
economy over the next 25 years—is just 0.9 percent of GDP (CBO 2013b).
Without these recent improvements in the outlook for Federal health spending, the Nation’s medium-run fiscal problem would therefore be about half
again as large.
It is important to note that the reductions in projected Medicare and
Medicaid spending described above are separate from the deficit reduction
that CBO estimates will occur as a direct result of the ACA. The most recent
CBO estimates indicate that the ACA will reduce the deficit by about $100
billion over the decade 2013-22, and that it will reduce the deficit, on average, by about 0.5 percent of GDP in the subsequent decade (CBO 2012b).
CBO notes that these deficit-reducing effects are likely to continue to grow
in following decades.

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 175
 
the Affordable Care Act

Figure 4-6
Recent CBO Projections of Medicare and Medicaid Outlays

Percent of GDP
6.0

Aug-12

5.8

Aug-10

5.5

Aug-11

Feb-14
May-13

5.3
5.0
4.8
4.5
4.3
4.0
2014

2016

2018

2020

2022

2024

Notes: Medicare outlays reflect spending net of offsetting receipts. Medicaid spending reflects
Federal spending only.
Sources: Congressional Budget Office, Budget and Economic Outlook; CEA calculations.

Higher Employment and Economic Growth
Slower growth in health care costs reduces the growth of the health
insurance premiums paid by employers. As discussed above, in the long
run, because employers must compete for workers, reductions in the cost
of health care are likely to be passed through to workers in the form of
higher wages. Thus, over the long run, changes in the growth rate of health
care costs are unlikely to substantially affect employer’s hiring costs and
decisions.16
In the short run, however, the picture may differ. Wage setting is
subject to various “rigidities” that mean that lower health insurance costs
may not be fully passed through in the short and medium run, potentially
reducing employer costs and spurring hiring (Sommers 2005). Rigidities
of this kind may be particularly important in the aftermath of the 2007-09
recession, as abnormally low inflation has increased the importance of constraints on the adjustment of nominal wages (Daly et al. 2012).
16 Faster growth in health insurance costs could reduce employment through another
mechanism. In particular, if workers do not value the additional health spending, then the
combination of more expensive health insurance and lower wages could make employment less
attractive over time, inducing them to reduce their labor supply. Because evidence suggests that
workers’ labor supply is only modestly responsive to the returns to work, these effects are likely
to be modest in size.

176  |  Chapter 4

There is relatively little empirical literature on the effect of slower
growth in employer health insurance premiums on employment, and there
is no consensus among economists about the likely size of these effects.
There are, however, at least two empirical studies suggesting that these
effects could be substantial.
Baicker and Chandra (2006) use variation in employer health insurance costs resulting from within-state changes in medical malpractice costs
over time to estimate the effect of higher health insurance premiums on
employment. They find that a 10 percent reduction in insurance premiums
increases the share of working-age individuals who are employed by 1.2
percentage points. This estimate suggests that the recent slowdown in the
growth of health insurance premiums could have had a substantial positive
effect on employment.
Sood, Ghosh, and Escarce (2009) take an alternative approach to
quantifying the effect of faster premium growth on employment. Specifically,
they examine whether industries that provide insurance to a large share of
their employees experience relatively lower employment growth during
periods when health costs are growing particularly rapidly. They find that,
for an industry that provides health insurance to all of its workers, increasing
health insurance premiums by 1 percent reduces the industry’s employment
by 1.6 percent relative to an industry that insures none of its workers.
Translating the Sood, Ghosh, and Escarce estimates into effects on
aggregate employment is difficult because their results could arise either
because higher health insurance costs reduce employment overall or because
they cause a reallocation of employment from high-coverage industries
to low-coverage industries. Cutler and Sood (2010) make one set of plausible assumptions about the importance of these two types of employment
changes, and given their estimates of the effect of the ACA on the path of
health care costs, find that the ACA will increase job growth by 250,000 to
400,000 a year by the second half of this decade.
In the longer run, lower deficits due to the ACA and the slowdown in
health costs also have the potential to improve economic growth. Reductions
in long-term deficits increase national saving, which increases capital accumulation and reduces foreign borrowing, and thereby increase national
income and living standards over time. As discussed in detail in a 2009 CEA
report on the potential benefits of health care reform for the economy, this
means that even modest sustained reductions in health care cost growth can
generate substantial economic benefits (CEA 2009).

Recent Trends in Health Care Costs, Their Impact on the Economy, and the Role of | 177
 
the Affordable Care Act

Conclusion
The evidence is clear that recent trends in health care spending and
price growth reflect, at least in part, ongoing structural changes in the
health care sector. The slowdown may be raising employment today and, if
continued, will substantially raise living standards in the years ahead. The
evidence also suggests that the Affordable Care Act is already contributing to lower spending and price growth, and that these effects will grow in
the years ahead, bringing lower-cost, higher-quality care to Medicare and
Medicaid beneficiaries and to the health system as a whole. But realizing
these benefits will require additional action, including continuing aggressive
implementation of the ACA’s reforms, taking full advantage of the ACA’s
mechanisms for developing and deploying innovative new payment models,
and pressing forward with new efforts that build on the ACA’s approach to
reducing health spending system-wide, such as the reform proposals in the
President’s recent budgets.

178  |  Chapter 4

C H A P T E R

5

FOSTERING PRODUCTIVITY
GROWTH

I

n 1870, a family farmer planting corn in Iowa would have expected to
grow 35 bushels an acre. Today, that settler’s descendant can grow nearly
180 bushels an acre and uses sophisticated equipment to work many times
the acreage of his or her forbearer. Because of higher yields and the use of
time-saving machinery, the quantity of corn produced by an hour of farm
labor has risen from an estimated 0.64 bushel in 1870 to more than 60
bushels in 2013. This 90-fold increase in labor productivity—that is, bushels
of corn (real output) an hour—corresponds to an annual rate of increase of
3.2 percent compounded over 143 years. In 1870, a bushel of corn sold for
approximately $0.80, about two days of earnings for a typical manufacturing
worker; today, that bushel sells for approximately $4.30, or 12 minutes worth
of average earnings.1
This extraordinary increase in corn output, fall in the real price of
corn, and the resulting improvement in physical well-being, did not come
about because we are stronger, harder-working, or tougher today than the
early settlers who first plowed the prairies. Rather, through a combination
of invention, more advanced equipment, and better education, the Iowa
farmer today uses more productive strains of corn and sophisticated farming
methods to get more output an acre. Today’s farmer harnesses more capital
equipment, such as advanced planters and combines, to plant more acres,
and has the know-how to operate this sophisticated equipment.
Technological advances such as corn hybridization, fertilizer technology, disease resistance, and mechanical planting and harvesting have
resulted from decades of research and development. While the government
has supported some of this research and its dissemination—for example,
through basic biological research and land-grant universities—much of
this research occurred in the private sector. However, the government has
1 Sources: Parker and Klein (1966), 1870 Census of Manufacturers, Iowa State University
Extension Service (2013), Bureau of Labor Statistics, USDA Economic Research Service.

179

facilitated this private-sector technological innovation by providing the
infrastructure to transport and sell increasing quantities of the products and
a regulatory and legal environment, such as the U.S. patent system, which
clarifies and enforces rights to inventions (more generally, to intellectual
property) so that the private sector can reap the rewards of research. These
property rights create incentives for innovators, while also allowing others
to build on their inventions. The improvements in productivity made possible by technological progress have appeared not just in agriculture, but also
throughout the U.S. economy.
The framework of government support for technological innovation
is facing new challenges that stem from an ever-changing scientific and legal
landscape. Many of these challenges center on the best way to support and
encourage development of intellectual property which now encompasses
improvements, not just to tractor design, but also technological changes
to the software that optimizes its performance. Farmers can now use the
Internet to do market research, purchase inputs, make direct sales, and
participate in online crop and livestock auctions. Other challenges involve
issues surrounding the allocation of the electromagnetic spectrum in a way
that supports the efficient development of new wireless and communications technologies that will improve productivity and connectivity—for the
farmer in the combine’s cab as well as for millions of other consumers and
businesses—while weighing national security and other concerns. These
challenges also include striking the appropriate balance between the need
for the government to support fundamental research, which can have large
positive externalities that will not be realized by any individual private actor,
and the importance of private-sector innovation in driving technology
forward.
Another set of challenges relates to how the gains from innovation are
shared. In the decades following World War II, productivity improvements
translated relatively automatically into compensation increases for families
across the income spectrum. But starting in the 1970s, inequality began its
relentless rise and productivity growth became increasingly disconnected
from compensation growth for typical families. The trends in inequality
are related to the trends in productivity, as well as to other broad economic
trends. Some of the technological changes over the past three decades,
especially those related to information technology, have raised the relative reward to skills obtained through advanced academic study. Thus, the
slowing growth of educational attainment both potentially slows innovation
and increases inequality by raising the returns to the most highly educated
workers. Although expanding the size of markets through globalization can

180  |  Chapter 5

help increase the productivity of the economy, it can also create challenges
for inequality.
This chapter begins with a review of the history of productivity growth
since World War II, emerging inequality trends, and the government’s role
in fostering productivity growth. It then focuses on two important current
issues in more detail: wired and wireless broadband infrastructure and the
efficient allocation of the electromagnetic spectrum; and, new challenges
to the U.S. patent system posed by standard-essential patents and patentassertion entities.

Trends in Total Factor Productivity
The most commonly used measure of productivity is labor productivity—that is, real output per hour worked. Over the long run, improvements
in labor productivity translate into growth of output, wages, and income.
Labor productivity can grow for multiple reasons: more capital per worker
(increased capital intensity), increased labor skills (a more experienced
workforce, more and better education and training), and technological
advances that improve the quality and productivity for a given level of
capital and labor skills (inventions, technological progress, process improvements, and other factors).
Because of the importance of technological progress in enhancing
long-run growth, economists also use another measure of productivity called
total factor productivity, or TFP, which proxies for the effect of technological progress. From 1948 to 2012, labor productivity growth in the private
nonfarm business sector has averaged 2.2 percent per year, and total factor
productivity growth, as measured by the series on multifactor productivity
produced by the Bureau of Labor Statistics (BLS), has averaged 1.1 percent
per year. This growth of productivity has not been constant, however, and
can usefully be thought of as occurring in three episodes: a period of fast
productivity growth through the early 1970s, a period of slow productivity
growth through the mid-1990s, and a period of somewhat faster productivity growth since then, but still not as fast as in the 1950s and 1960s.

Labor Productivity, Total Factor Productivity, and Multifactor
Productivity
The growth rate of labor productivity equals the growth rate of output, minus the growth rate of labor input (worker hours), thus yielding the
growth rate of output per worker hour. In contrast, the growth rate of TFP
is the growth rate of output, minus the growth rate of output that would
be expected solely from the growth rate of the inputs to production. The
Fostering Productivity Growth  |  181

resulting gap between the actual growth rate of output and the growth rate
arising solely because of the growth of inputs is also known as the Solow
residual, and is a measure of how well those inputs are combined. Thus,
the growth rate of TFP tracks a broadly defined concept of technological
change that encompasses scientific innovation and invention, managerial
innovations, effects of reorganization of the production process, and other
efficiency improvements that do not accrue uniquely to a single measured
input.
The concept of total factor productivity is appealing because it estimates the contribution of technological developments to economic growth,
and because it can be applied at the level of an industry as well as to the
overall economy. In practice, measuring TFP poses several challenges. First,
TFP is not observed directly and instead must be estimated using measured
inputs and estimates of how the inputs contribute to output. Second, the
inputs discussed so far have been capital and labor, but other inputs to production also include, in particular, energy, materials, and business services.
Third, for a given level of other inputs, output can increase by hiring bettertrained or higher-skilled workers; so for the purpose of measuring TFP, the
desired concept of labor input captures changes in both the quantity and
quality of labor input. Because labor quality is not observed, proxies such as
age and education must be used. Both academics and the U.S. Government
have tackled these and other measurement challenges, and have developed
estimates of the growth of TFP. This chapter uses an estimate of TFP produced by the Bureau of Labor Statistics called multifactor productivity, or
MFP, which is described in Box 5-1.2

Postwar U.S. Productivity Growth
According to the BLS measure of labor productivity shown in Table
5-1, an American worker could produce more than four times as much
output per hour in 2012 as in 1948.3 Because MFP takes into account the
2 One of the many other challenges in estimating total factor productivity is that the intensity
of utilization of inputs varies over the business cycle. For example, because hiring and training
workers is expensive, firms might retain some workers in a mild downturn, so that fluctuations
in output are greater than fluctuations in employment (a relationship which, when recast in
terms of the unemployment rate, is known as Okun’s Law). The BLS MFP series does not
adjust for changes in factor utilization, which can produce cyclical fluctuations in MFP. Basu,
Fernald, and Kimball (2006) provide an approach to adjusting for such cyclical variation, and
a quarterly TFP series produced using their method is currently maintained by the Federal
Reserve Bank of San Francisco (Fernald 2012).
3 This discussion of postwar productivity performance cites statistics for nonfarm private
businesses. Recall the earlier discussion about how productivity growth in farming allowed
fewer resources to be devoted to it. By 1947, farming accounted for less than a nine percent
share of GDP. Today that share is about one percent.

182  |  Chapter 5

Box 5-1: Measuring Multifactor Productivity
The Bureau of Labor Statistics (BLS) publishes annual data on
multifactor productivity, covering the private business sector, the private
nonfarm business sector, the manufacturing sector, and 18 industries
within the manufacturing sector.
Private business-sector output is a chain-type, annual-weighted
(Fisher-Ideal) index constructed after excluding general government,
nonprofit institutions, private households (including owner-occupied
housing), and government enterprises from gross domestic product
(GDP). The input measure is an aggregation of two inputs, labor and
capital. Labor input is obtained by chained Tornqvist aggregation of
the hours worked in private business by all persons, classified by age,
education, and gender with weights determined by each group’s share of
total labor compensation. Capital inputs are measured based on the flow
of services derived from physical assets. For each of 60 industries in the
private-business sector, quantities of each capital asset are aggregated
into a Tornqvist index, using estimated rental prices. Current-dollar capital costs are found by multiplying the rental price for each asset by the
asset’s constant-dollar stock, adjusting for capital composition effects.
Finally, the combined input (labor and capital) measure is constructed
via another Tornqvist index, taking as weights each inputs’ share of total
costs derived from the National Income and Product Accounts.
Manufacturing is treated somewhat differently. The output measure, known as sectoral output, is the value of production shipped to
purchasers outside the domestic industry, either to satisfy final demand
or to use as an input in other industries. Because additional inputs to
manufacturing can be tracked, the input measures available include
not just capital and labor, but also energy, non-energy materials, and
purchased business services input. Intra-industry purchases are removed
to avoid double counting. The resulting aggregate input is referred to
by the acronym KLEMS—capital (K), labor (L), energy (E), materials (M) and services (S). Given these inputs and outputs, multifactor
productivity is computed for 18 3-digit and 86 4-digit North American
Industry Classification System (NAICS) manufacturing industries and
for the manufacturing sector as a whole using the Tornqvist aggregation
methods described above for the private business-sector manufacturing.

growth of capital and other factors, labor productivity growth generally
exceeds MFP growth. For example, even absent technological change, labor
can be more productive simply by using more capital; that is, by increasing
the capital-labor ratio or so-called capital deepening. Mathematically, the
growth rate of labor productivity is the sum of the MFP growth rate, the
Fostering Productivity Growth  |  183

Table 5–1
Sources of Productivity Improvement, Nonfarm Private Business, 1948–2012
Improvement
(multiple)

Contribution to
Labor Productivity Growth
(percent)

Composition of Labor

1.15

10

Capital

1.74

38

MFP

2.10

52

Labor Productivity

4.21

100

Source

Source: Bureau of Labor Statistics, Productivity and Costs, Multifactor Productivity.

contribution of changes in labor quality (as measured by changes in the
composition of the workforce), and the contribution of the growth in the
amount of capital per worker.4 The final column of Table 5-1 gives this
decomposition, showing that 10 percent of the growth in labor productivity
is due to improvements in the composition of labor (primarily greater educational attainment), 38 percent is due to increases in the amount of capital
the worker has at his or her disposal, and 52 percent is due to increases in
broad technological progress as measured by MFP.
The growth rates of the BLS measures of labor productivity and
multifactor productivity have varied over time and are shown in Figure 5-1.
Over the past 60 years, labor productivity has on average grown just over 1
percentage point faster than MFP: from 1953-2012, labor productivity grew
at an annual rate of 2.2 percent per year, and MFP grew at an annual average
rate of 1.1 percent per year.
As can be seen in Figure 5-1, both labor productivity and MFP are
quite volatile from year to year. One reason for this volatility is measurement
error in the estimation of both series; indeed, proper measurement of the
inputs and outputs is a daunting task and for this reason alone not too much
should be read into the growth of productivity in any one year. Another
reason is that these series, and the gap between them, varies cyclically. For
example, MFP growth fell—in fact, took on negative values—during the
recessions that started in 1969, 1980-81, 1990, and 2007. These negative values do not mean that, during recessions, firms make negative technological
4 Suppose aggregate production can be represented by the Cobb-Douglas production function,
Y = ALαK1-α, where Y is real output, L is labor input measured in labor-quality units, K is
capital, and A summarizes the contribution of technology to production, that is, A is TFP, and
α is a constant. Then output per worker-hour (H) is Y/H = A(L/H)α(K/H)1-α. Thus the annual
growth of output per worker, that is, the growth of labor productivity, is the sum of the growth
of A, that is, the growth of TFP, plus α times the growth of L/H, that is the growth of labor
quality per worker-hour, plus 1-α times the growth of K/H, that is, the growth of the capitallabor ratio. By using Tornqvist aggregation, the BLS MFP measure allows shares (α) to change
over time and does not require an aggregate Cobb-Douglas production function.

184  |  Chapter 5

Figure 5-1
Nonfarm Private Business Productivity Growth, 1949 –2012

Annual percent change
8
6

Labor Productivity

2012

4
2
0
-2
-4

-6
1950

Multifactor Productivity
1960

1970

1980

1990

2000

2010

Source: Bureau of Labor Statistics, Productivity and Costs, Multifactor Productivity.

progress or collectively forget about the innovations they have produced
over the preceding years. Rather, such declines in MFP could come about
from changes in relative prices, so that existing methods of production are
no longer the optimal way to combine inputs to produce output. Negative
MFP growth can also arise from variation in the utilization rates of capital
and labor over the business cycle.
From the perspective of policies to foster long-term economic growth,
these annual and cyclical fluctuations are less relevant than long-term trends
in the growth rates of productivity. Figure 5-2 shows a centered 15-year
moving average of the growth rates of labor productivity and MFP; and,
Table 5-2 summarizes the compound annual growth rates of these series
over 10- and 20-year periods ending in 2012, as well as the 60-year period
from 1953-2012.
Table 5-2 and Figure 5-2 tell a similar story, which has two parts.
First, over the long run the gap between labor productivity growth and MFP
growth has fluctuated in a small range, with a difference of between 1.0 and
1.3 percentage points in decadal averages. Moreover, there is no noticeable trend in this gap: the mean difference in the growth rates of these two
productivity measures over 2003-12 is within 0.2 percentage point of the
mean difference over 1953-62. The stability of the difference between these

Fostering Productivity Growth  |  185

Table 5–2
Nonfarm Private Business Productivity Growth
10–year Average annual rates of change
Period

Multifactor Productivity

Labor Productivity

Difference

1953–1962

1.5

2.6

1.1

1963–1972

1.9

2.8

1.0

1973–1982

–0.1

1.1

1.2

1983–1992

1.1

2.2

1.1

1993–2002

1.1

2.4

1.3

2003–2012

0.9

1.9

1.0

20–year Average annual rates of change
Period

Multifactor Productivity

Labor Productivity

Difference

1953–1972

1.7

2.7

1.0

1973–1992

0.5

1.6

1.1

1993–2012

1.0

2.1

1.1

60–year Average annual rates of change
1953–2012

1.1

2.2

1.1

Source: Bureau of Labor Statistics, Productivity and Costs, Multifactor Productivity.

measures underscores the role of broad technological change—as measured
by MFP—as a key driver of long-term growth of output per worker.
Second, over the past 60 years the long-term mean growth rates of
labor productivity and MFP have varied substantially, in what appear to be
three episodes. The first episode, the 1950s through early 1970s, experienced
high growth of MFP (and of labor productivity), with MFP growth averaging
1.7 percent per year from 1953 through 1972. The second episode, the late
1970s through early 1990s, experienced much lower MFP growth, averaging
0.5 percent per year. The third episode, from the mid-1990s through the
present, experienced an intermediate level of MFP growth of 1.0 percent per
year.
Because productivity is the key to raising output per person, a great
deal of academic research has focused on understanding why productivity
growth varies over time. Research points to several factors that contributed
to the productivity slowdown of the 1970s. A major culprit seems to be the
sharp rise in energy prices during the 1970s that made less energy-intensive
technologies more attractive, thus changing the optimal way to combine
inputs and reducing MFP growth (Jorgenson 1988, Nordhaus 2004). One
lesson learned from this period is how important energy cost fluctuations
are in determining the growth of potential output.

186  |  Chapter 5

Figure 5-2
15-Year Centered Moving Average of Annual Growth Rates for
Labor and Multifactor Productivity, 1956 –2005

Percent
3.5
3.0

2005

Labor Productivity

2.5
2.0
1.5
1.0
0.5
0.0

Multifactor Productivity
1950

1960

1970

1980

1990

2000

Source: Bureau of Labor Statistics, Productivity and Costs, Multifactor Productivity;
CEA Calculations.

2010

Another explanation is due to rapid changes in the labor force in the
1970s, primarily shifting the workforce to newer, less-experienced workers.
The Baby Boom generation (the cohort born between 1946 and 1964) came
of age in the 1970s and 1980s, lowering the overall work experience of the
economy. This was a period of rapid entry of women into the workforce for
the first time, a shift that also temporarily reduced the overall level of workforce experience in the economy (Feyrer 2007, 2011). Moreover, the rapid
entry of these new workers into the workforce outpaced investment, slowing
the growth of the capital-labor ratio.
Another possible part of the story is that productivity growth in the
1950s and 1960s was temporarily spurred by large public investments such
as the interstate highway system and the commercialization of military innovations from World War II like the jet engine and synthetic rubber.
The productivity rebound of the 1990s and 2000s is widely attributable to the information technology (IT) revolution. For the nine years from
1996 to 2005, MFP grew at 1.6 percent per year, a rate not seen in a nine-year
period since the mid-1960s. Although many of the basic technologies that
facilitated this growth, like the personal computer and the software to run
it, were invented in the 1970s and 1980s; improvements in speed, breadth
of applications, and the ability of firms to exploit this technology stretched
through the ensuing decades. The BLS MFP measure suggests that much

Fostering Productivity Growth  |  187

of the productivity improvement resulted from technological and process
improvements, a position supported, for example, by Basu, Fernald, Oulton,
and Srinivasan (2004). Alternatively, Jorgenson (2001) and Jorgenson and
Ho (2012) emphasize the importance of the accumulation of physical IT
capital. Oliner, Sichel, and Stiroh (2007) provide a detailed review of the
literature on the 1990s productivity boom.
A key current question is what the rate of productivity growth will be
going forward—will the U.S. economy maintain the pace of recent decades,
will new innovations accelerate the pace of productivity growth, or will
productivity growth revert to the slower rates before the recent boom? MFP
growth fell sharply in the recession, grew sharply in the early stages of the
recovery, and has averaged 1 percent for 2011 and 2012. These large cyclical
swings make it difficult to assess whether there has been a recent change in
the rate of technological progress, relative to the late 1990s and early 2000s.
The academic literature reaches mixed findings concerning whether the IT
productivity boom was temporary.5 This literature also requires qualification because it predates the substantial data revisions to historical GDP and
productivity that were released in the summer of 2013, which substantially
revised upwards estimated productivity growth in some years in the 2000s.
Some contributions to this debate look further into the future. While
some economists predict labor productivity growth could decline in coming
decades because the scope for future transformative general-purpose inventions is limited (Gordon 2012), others argue that IT is in fact a general-purpose invention and, at least in the medium run, presents an ongoing stream
of opportunities for workplace reorganization and efficiency gains, as well
as spin-off technologies and improvements.6 Bernanke (2012) argued that
making these improvements often requires more than just purchasing hardware and software, and realizing potential productivity gains can require
changes within and between organizations and thus take a considerable time
to be fully realized.7
Ultimately, it is very hard to predict future growth rates in innovation,
and there is no economic reason that these growth rates should be constant
over time. Moreover, the past four decades have seen substantial changes in
5 The findings in the literature on recent productivity growth trends tend to depend on the
statistical approaches used to discern different productivity regimes. Authors that adopt
discrete breaks or regime shifts, including Kahn and Rich (2011) and Fernald (2012), tend to
conclude that the productivity growth boom has passed, whereas Oliner, Sichel, and Stiroh
(2007), who use methods in which productivity growth evolves more slowly, find less of a
slowdown.
6 An example of such now-possible workplace reorganization is telecommuting; see, for
example, Bloom, Liang, Roberts and Ying (2013); Noonan and Glass (2012), Bailey and
Kurland (2002); and Busch, Nash, and Bell (2011).
7 These issues are argued in the February 2013 TED debate between Gordon and Brynjolfsson.

188  |  Chapter 5

the extent to which productivity gains translate into higher incomes across
the board, the topic of the next section.

Productivity Growth and Inequality Growth
Productivity improvements provide more output that has the potential to benefit society broadly. Through the early 1970s, productivity gains
led to increases in labor compensation. Since then, however, productivity
growth has not translated into commensurate growth in labor compensation, and income inequality has increased markedly.

Trends in Inequality, Productivity Growth, and Compensation
Real output per hour was 99 percent higher by the end of 1972 than in
1947, while real average hourly earnings (GDP deflator) grew by 73 percent.
Figure 5-3 shows that since the early 1970s, the paths of labor productivity
and average hourly earnings diverged more widely. As a result, by the end of
September 2013 real output per hour was 107 percent higher than at the end
of 1972, but average hourly earnings had only grown 31 percent.8
Table 5-3 examines the real output per hour and average hourly
earnings for private production and nonsupervisory workers by decade.
From 1953 to 1962, productivity growth exceeded the average annual rate of
change in hourly earnings by only 0.4 percentage point. In the next decade,
the difference in growth had ticked up to 0.6 percentage point. However,
from 1973 through 2012 labor productivity grew 1.4 percentage points faster
than earnings.
Since the 1970s, these trends generally have been worse for lowerincome households than for higher-income households (DiNardo, Fortin,
Lemieux 1996; Piketty and Saez 2003; Lemieux 2008; CEA 2012; Haskel,
Lawrence, Leamer, and Slaughter 2012).9 In particular, the income growth
in the top percentile of the income distribution has been much stronger
than other percentiles. For example, the Congressional Budget Office (CBO
2011) reports that from 1979 to 2007, real before-tax income at the median
of the household income distribution increased by about 19 percent, while
8 An alternative series from BLS measures real total hourly compensation (CPI deflator) for all
nonfarm workers. This measure includes benefits as well as earnings. Since 1972, total hourly
compensation has increased more than hourly earnings, but still only by 46 percent. BLS
decompositions of compensation into real wage and benefit shares have been available since
1991. Since then, real wages grew 7 percent and benefits grew 22 percent, with the strongest
benefit growth in the magnitude of employer contribution to health insurance.
9 Figure 6-2 of this report suggests that the relative slow growth in income of the lower
quintiles may have subsided some recently, particularly during the Great Recession and its
near-term recovery. It is too soon to tell whether this has any implication for longer-term
trends.

Fostering Productivity Growth  |  189

Figure 5-3
Growth in Productivity and Average Wage, 1947–2013

Index, 1947=100 (log scale)
500

2013

400
300

Real Output
Per Hour

200

Real Average Wage
(output deflator)

Real Average Wage
(CPI deflator)

100
75
1945

1955

1965

1975

1985

1995

2005

Note: Real output per hour is for all workers in the nonfarm business sector. Average wage is for private production and
nonsupervisory workers. Output deflator is the price index for nonfarm business output. CPI deflator is the CPI-W.
Data on wages before 1964 reflects SIC-based industry classifications.
Source: Bureau of Labor Statistics, Productivity and Costs, Current Employment Statistics; CEA calculations.

incomes for the top 1 percent of households have increased by around 200
percent.10

Technological Change and Inequality
The lesson from Figure 5-3 is that productivity growth is important
for wage growth, but that does not mean that it automatically leads to wage
growth. One possibility is that the sources of labor productivity and MFP
growth since the early 1970s are qualitatively different than earlier, and that
these different sources of growth drove the trends in inequality over the last
40 years. In the early 1990s, a broad consensus emerged among economists
that an increase in the demand for skill relative to the supply of educated
labor was the primary driver of the sharp rise in inequality in the 1980s
(Bound and Johnson 1995; Katz and Murphy 1992; and Juhn, Murphy,
and Pierce 1993). It soon became accepted that “skill-biased technological change” (SBTC) was the most important cause of increased inequality
(Berman, Bound, and Griliches 1994; Krueger 1993). The crux of the argument is that, as computer technology became increasingly less expensive,
relative demand increased for workers with complementary skills. This
explanation has remained popular among economists with few modifications to the basic argument until recently (for example, Acemoglu 2002).
10 The CBO notes that it chose 1979 and 2007 as points of comparison because there are
cyclical fluctuations in inequality measures and both years are business cycle peaks.

190  |  Chapter 5

Table 5–3
Average Annual Rates of Change in the Nonfarm Business Sector

Period

Real
Output per Hour
of all Workers

Average
Hourly Earnings
for Private Production
and Nonsupervisory
Workers

Difference
(p.p.)

1953–1962

2.5

2.1

0.4

1963–1972

2.7

2.1

0.6

1973–1982

1.1

–0.4

1.4

1983–1992

2.2

0.4

1.8

1993–2002

2.3

1.8

0.5

2003–2012

2.1

1

1.1

Note: Both series are deflated by the price index for output in the nonfarm business sector. Data on earnings before
1964 reflect SIC–based industry classifications.
Sources: Bureau of Labor Statistics, Productivity and Costs, Current Employment Statistics; CEA Calculations.

While this hypothesis has remained influential, there are reasons to
question the primary role of technology in causing the inequality changes
that emerged in the 1980s. For example, many other industrialized nations,
such as Germany and Japan, experienced similar technology shocks in
the 1980s, but saw little or no increase in wage inequality. This led some
economists to expand the framework for explaining inequality to acknowledge the importance of wage-setting institutions in mediating technology
shocks (Freeman and Katz 1995). This critique gained more force with other
researchers finding that changes in institutions—especially the decline in the
real value of the minimum wage and labor unions—could account for much
of the rise in inequality in the 1980s, at least in the bottom of the distribution
(Lee 1999 and DiNardo, Fortin, and Lemieux 1996). An additional challenge to the skill-biased technological change hypothesis is that the timing
of changes in inequality do not line up well with the nature of technological
change across decades. Inequality in the bottom of the distribution rose in
the 1980s, but has been flat or declining since then. However, much of the
widespread business adoption of IT, including the Internet, occurred in the
1990s, and those innovations were at least as significant as the changes in the
1980s (Card and DiNardo 2002). In fact, inequality in the top of the distribution did continue to rise, but after rising sharply in the 1980s, inequality at
the bottom of the distribution has been flat or declining since.
Goldin and Katz (2008) focus on changes in the growth of the supply
of skills rather than on episodic increases in technological change. Using the
ratio of college to non-college workers as a measure of the relative supply
of skills, they show this relative skill supply grew by 3.9 percent from 1960
to 1980. But in 1980, as confirmed by Heckman and LaFontaine (2010)
and others, this increase slowed as high school graduation rates stopped
Fostering Productivity Growth  |  191

improving and college completion rates slowed. Goldin and Katz (2008)
show that a constant increase in the demand for relative skill, combined
with the post-1980 slowdown in the supply of relative skill, explains the time
path of the logarithm of the college wage premium, which is one measured
aspect of wage inequality.11 The nature of rising wage inequality started to
change around the early 1990s, becoming increasingly concentrated in the
top end of the wage distribution. The ratio of the 90th to the 50th percentile
of the wage distribution continued to grow at roughly the same rate it had
since the early 1980s, whereas inequality at the bottom (the 50-10 ratio)
declined somewhat after the late 1980s. Piketty and Saez (2006) find that
income gains have increasingly concentrated in the top 10 percent and top 1
percent since the 1980s. The result has been a “polarization” or a “hollowing
out” of the wage distribution, with relative wage growth in the bottom and
especially the top of the wage distribution relative to the middle (Goos and
Manning 2007; Autor 2010; Acemoglu and Autor 2011; Lemieux 2006).
Autor and coauthors refine the earlier skill-biased technological
change literature and argue that the changes in inequality are driven by
technological change that substitutes for some tasks but not others (Autor,
Levy, and Murnane 2003; Autor, Katz, and Kearney 2006; Acemoglu and
Autor 2011). In particular, this new research argues that computer technologies complement non-routine cognitive tasks, which tend to be highly paid;
substitute for routine tasks, which tend to be in occupations with wages in
the middle of the distribution; and have little effect on manual tasks that
tend to be associated with lower wages. This technological explanation
for polarization has been controversial, however, and Mishel, Shierholz,
and Schmitt (2013) suggest that the theory does not explain the timing of
changes in polarization, and more generally that occupational employment
and wage trends do not explain a large part of the trends in wages or inequality over time. Moreover, one of the most striking changes in inequality over
the past three decades—the sharp growth of incomes at the very top of the
distribution—is unlikely to be related to technological changes or to a relative demand for skill (Alvaredo, Atkinson, Piketty, and Saez 2013).
This discussion has focused on whether increases in productivity translate into increases in earnings or lead to increasing inequality. A
related, less-understood question is whether increasing inequality might
11 This theory is based on evidence from before 2008. The U.S. economy has long had
some skills shortage, which tended to turn up in the form of wage differentials rather than
unemployment. It does not account for the large shock in aggregate demand that characterized
the Great Recession, or the shock-driven unemployment rates from which the economy is still
recovering.

192  |  Chapter 5

directly dampen productivity growth, and this question is addressed further
in Box 5-2.

Policies to Foster Productivity Growth and to
Help Ensure That Everyone Benefits from it
The benefits of technological progress do not accrue only to those who
develop new processes and inventions; they also spill over to the population at large. For this reason, the U.S. Government has a role in supporting
and enabling technological development. This government role includes:
directly funding or providing incentives for research and development
(R&D); providing an institutional, legal, and regulatory environment that
protects competition, defines and supports intellectual property rights,
and thereby encourages private innovation; and developing human capital
through education, especially in scientific and technological fields. In addition, the government has a role in ensuring that everyone benefits from
those technological advances.
Investments in R&D often have “spillover” effects; that is, a part of
the returns to the investment accrue to parties other than the investor. As a
result, investments that are worth making for society at large might not be
profitable for any one firm, leaving aggregate R&D investment below the
socially optimal level (for example, Nelson 1959). This tendency toward
underinvestment creates a role for research that is performed or funded by
the government as well as by nonprofit organizations such as universities.
These positive spillovers can be particularly large for basic scientific
research. Discoveries made through basic research are often of great social
value because of their broad applicability, but are of little value to any individual private firm, which would likely have few, if any, profitable applications for them. The empirical analyses of Jones and Williams (1998) and
Bloom et al. (2012) suggest that the optimal level of R&D investment is two
to four times the actual level. Akcigit et al. (2013) also find underinvestment
in basic research (although, contrary to the bulk of the literature, they find
overinvestment in applied research), and suggest policies that are specifically
targeted at basic research.
Consistent with the presence of large spillover benefits, most basic
research in the United States is funded by the government and other
nonprofit entities. As Figure 5-4 shows, over half comes from government
sources, and less than one-quarter comes from private industry. However,
expenditures on basic research are only a fraction of total R&D expenditures,
as seen in Figure 5-5, and the private-sector share of funding for applied
research and development is much higher than it is for basic research.
Fostering Productivity Growth  |  193

Box 5-2: Does Inequality Affect Productivity?
Although conventional economic models do not include the equality of the income distribution as a determinant of economic output,
some recent research has focused on whether increasing income inequality might reduce the growth rate of productivity. There are at least three
channels that could produce this link and, in each, an underlying source
of income inequality potentially leads to slower productivity. The first
channel is through disparities in access to, and the quality of, publicly
funded secondary education: inequality in educational quality leads to
disparities in skills, so an increase in labor hours might not increase labor
quality, slowing labor productivity growth. For example, Goldin and
Katz (2008) argue that in the 19th and early 20th centuries, greater access
to education in the United States than in Europe resulted in the United
States having higher rates of labor productivity growth. In the United
States today, the relevant channel is not likely related to access to public
schools, but more likely geographic disparity in resources available to
students at those schools.
A second channel is that greater income inequality creates disparities in the ability to pay for privately funded education, especially prekindergarten and college.1 This channel too is relevant because of the
increasing expense of post-secondary education.
A third channel, discussed by Acemoglu and Robinson (2011), is
that sufficiently powerful and entrenched elites have an incentive to use
resources to protect their interest rather than encourage growth. The
relevance of Acemoglu and Robinson’s examples of extractive societies
drawn from world history—ancient Rome, the Mayans, slave-dependent
economies in the early Americas, and so forth—to the United States
today is less clear than that of the other channels.
There have been some attempts to use cross-country differences as
sources of variation for econometric studies of the link from inequality
to productivity growth. Those attempts, however, confront a variety
of data availability and measurement issues, including comparable
measures of inequality (Fields 2001) and insufficient variables to avoid
spurious effects being loaded onto the inequality measure (Banerjee and
Duflo 2003). In any event, the question of whether the increases in U.S.
inequality over the past two decades have dampened, or could dampen,
productivity growth remains an important source of concern.
1 Except for programs like Head Start, pre-kindergarten education is privately financed.
Heckman, Pinto and Savelyev (2012) contribute and list literature demonstrating the
importance of early childhood intervention to subsequent schooling and other life
outcomes. At the college level, nearly all students pay at least some of their educational
expenses.

194  |  Chapter 5

Figure 5-4
Basic Research Expenditures in the U.S.
by Source Funding, 2010

Other
government
3%

Other
nonprofit
11%

U&C
10%
Federal
53%
Industry
23%

Source: National Science Foundation, National Patterns of R&D Resources: 2010-11 update.

	 In addition to direct funding of R&D, the government also provides financial incentives for private R&D investment through tax policy.
Government can also facilitate private R&D investment and technological
progress by providing an institutional, legal, and regulatory framework that
clarifies and enforces intellectual property rights and thereby ensures that
innovators reap enough financial rewards from their innovations to provide
sufficient incentive to engage in a closer-to-optimal level of R&D.12
One important type of intellectual property right is patents. A patent grants the inventor a temporary exclusive right over the invention.
Exercising that right results in high prices and profits for investments that
are successfully commercialized, and those profits provide an incentive to
invent. However, the exercise of the exclusive right will also raise prices on
inventions that would have been created even with weaker patent protection
or with none at all, and these higher prices harm consumers. Moreover,
because patented inventions are sometimes used as inputs in creating
additional innovations, the higher prices created by patents (as well as the
associated legal and administrative burdens, such as negotiating licenses)
could slow down subsequent innovation. As discussed further below, a
central economic challenge of patent policy is to strike the right balance
12 Research in development economics suggests that a key factor in the economic performance
of a country is its “institutions,” such as rule of law and clear property rights (Hall and Jones
1999, Rodrik et al., 2004).

Fostering Productivity Growth  |  195

Percent
3.5

Figure 5-5
Composition of Total R&D Spending
as a Share of GDP, 1953–2011

3.0
2.5
2.0

Development

1.5
1.0
Applied

0.5
0.0
1953

Basic
1963

1973

1983

1993

2003

Source: National Science Foundation, National Patterns of R&D Resources: 2010-11 update.

between providing an economic incentive to invent and the potential harm
from the exercise of patent rights. At a minimum, it is important to ensure
that patents are not wrongly issued, but rather are only issued for inventions
that are non-obvious, useful, and inventive.
The government can also lay the groundwork for greater creativity and invention by supporting the development of human capital.
Investments and improvements in education and training, particularly in
the Science, Technology, Engineering, and Mathematics (STEM) fields,
foster the innovation workforce of the future.13 The productivity of these
workers can be enhanced by investment in “innovation clusters,” which are
dense concentrations of firms and of highly skilled personnel, usually close
to a major research university, whose mutual proximity can further promote
innovation (see Greenstone, Hornbeck, and Moretti 2008).
Immigration reform is another human capital policy that has the
potential to increase the pace of innovation. Studies have found that foreignnationals living in the United States authored or co-authored over 25 percent of U.S. patent applications in 2006, and that over 75 percent of patents
awarded to the top 10 patent-producing American universities in 2011 had at
13 As discussed in Delgado et al. (2012), one determinant of a country’s economic performance
is its science and innovation infrastructure. The authors include in this category a number
of elements that can be influenced by supportive government policy, such as the quality of
scientific research institutions and the quality of math and science education.

196  |  Chapter 5

least one foreign-born inventor. Moreover, the innovation benefits of immigration are not confined to the immigration of innovators. Immigration of
low-skilled workers, as well as immigration of high-skilled workers who are
not innovators, can spur innovation indirectly by increasing specialization.
When more non-innovators are present to specialize more completely in
their occupations, they enable innovators to specialize more completely in
theirs. The Congressional Budget Office (2013) projects that the additional
immigration resulting from the Border Security, Economic Opportunity,
and Immigration Modernization Act, as passed by the Senate, would raise
total factor productivity by roughly 0.7 percent in 2023 and by roughly 1.0
percent in 2033 as a result of increased innovation and task specialization.
Finally, the government has an important role in ensuring that access
to the technologies that catalyze productivity growth, and to the technologies and products that are the fruits of that productivity growth, are broadly
available throughout American society. Sharing these benefits increases
welfare directly, and also ensures that the broad population maintains the
technological skills needed in the workplace and for the education of current
and future generations.
This chapter now turns from a general discussion of the role of government policy in achieving technological progress to a focus on two key
current areas that are important for productivity growth and that are also
a focus of the Administration’s policies: telecommunications and patent
reform.

Telecommunications and Productivity Growth
The telecommunications industry is an important one for fostering
productivity growth. Improved telecommunications infrastructure, particularly fast and widely accessible wired and wireless broadband networks, is
a critical factor in enabling important technological advances in business,
health care, education, public safety, entertainment, and more. Government
policies have an important role to play in facilitating and catalyzing these
improvements, as discussed below. In this chapter, telecommunications
policy is discussed in particular detail, in part due to its importance, and in
part because it serves as a good illustration of more general economic and
policy principles.

Innovation and Investment
The telecommunications sector is a major success story in the U.S.
economy. A recent White House (2013) report, Four Years of Broadband
Growth, documents many of the striking facts, including:
Fostering Productivity Growth  |  197

•	 Just two of the largest U.S. telecommunications companies account
for greater combined stateside investment than the top five oil/gas companies, and nearly four times more than the big three auto companies combined, as seen in Figure 5-6.
•	Between 2009 and 2012, annual investment in U.S. wireless networks grew more than 40 percent, from $21 billion to $30 billion. During
that period, investment in European wireless networks remained flat, and
wireless investment in Asia (including China) rose only 4 percent. The
report projected that U.S. wireless network investment would increase further in 2013, to $35 billion.
•	 The United States leads the world in the availability of advanced 4G
wireless broadband Internet services such as LTE; nearly half of the global
subscriber base for 4G LTE is in the United States.
•	 The United States ranks among the top countries in the world in the
amount of currently licensed spectrum available for mobile broadband.
This infrastructure is at the center of a vibrant ecosystem that includes
smartphone design, mobile applications development, and the use of these
technologies to effect broader changes in the economy and society—all of it
centered in the United States. The mobile applications industry is forecast
to generate more than $25 billion in revenue in 2013, rising to $74 billion
in 2017, with nearly 2 million applications available for download at the
two largest mobile app stores. Improved telecommunications has also contributed to changes in the way that business is organized, and in ways that
may lead to further improvements in productivity. An example of this is
discussed in Box 5-3.

Four Key Areas for Telecommunications Policy
The U.S. Government can support innovation and investment in telecommunications through the same general policies discussed above: direct
government investment in research and development; catalyzing private
innovation through policies such as reforming and extending the Research
and Experimentation Tax Credit; catalyzing technological infrastructure
investment in areas like broadband; and ensuring that everyone benefits
from broadband technologies.
Government Investments in Research and Development. As discussed
above, spillover benefits to research and development, especially for basic
science and technology, creates a role for direct government investment.
Perhaps the most famous government investment in telecommunications
technology was the Defense Advanced Research Projects Agency (DARPA)
development of the Internet. But DARPA has provided other important
defense-based public research contributions as well. These contributions
198  |  Chapter 5

Figure 5-6
Relative Investment of the Telecommunications Sector, 2011
Big Three
Auto
Companies
($9.2 billion)
Top Five Oil/Gas
Companies
($32.8 billion)

Top Two
Telecommunications
Companies
($36.2 billion)

Source: Progressive Policy Institute.

include the radio and, more recently, Global Positioning Systems, which
today are central to a huge number of consumer applications.
Today, the Department of Defense (DOD) continues to play an important role in telecommunications research, particularly in helping to develop
ideas and technologies for sharing of electromagnetic spectrum frequency
bands between different users, including between government and private
users. This, as discussed further below, has been identified as important for
efficient spectrum management in the future (PCAST 2012). For example,
DOD has solicited innovative research proposals aimed at efficient and reliable sharing of spectrum between radar and communications systems. All
told, $100 million in Federal investments are being targeted toward spectrum sharing and advanced communications through the National Science
Foundation (NSF), DARPA, and the Commerce Department.
Catalyzing Private Investment. Reforming, expanding, and making
permanent the Research and Experimentation Tax Credit would increase
investment in telecommunications technology, accelerating innovation.
Immigration reform would accelerate innovation as well. Reforming the
patent system is also important in this industry, especially for technology
deployed in smartphones, which are complex devices that embody thousands of patents. The increasing frequency of patent disputes in this area
suggests that there may be increasing costs to navigating the appropriate

Fostering Productivity Growth  |  199

Box 5-3: Just-in-Time Manufacturing
The just-in-time (JIT) approach to manufacturing aims to maximize profits by dramatically reducing inventories and their costs. By
minimizing the time that inventory is held, the system allows a fixed
amount of inventory space to be used more productively; that is, by
processing more goods through the fixed spaced during a fixed amount
of time. Agrawal (2010) delineates channels through which the JIT
approach can reduce costs, improve quality and customer service, maintain flexibility, and promote logistical efficiency. Many of these channels
now rely on improved information and telecommunications technology.
Since JIT requires precise coordination between demand and supply,
the contemporaneous tracking of each is essential. On the supply side,
Zhang et al. (2012) argue that radio frequency identification technology can provide firms with precise, accurate, real-time information on
materials as they pass through the manufacturing process. But technology that makes JIT feasible is only one requirement. Other studies (Hur,
Jeong and Suh 2009, Tayal 2012, Fairris and Brenner 2001, Agrawal
2010, Sim and Koh 2003) show that organizational experimentation,
innovation, and learning in using the technology can also be necessary
to realizing productivity gains.

licenses. If these costs are high enough to adversely affect the introduction of
new products, then patent reform is particularly important for the telecommunications industry.
Catalyzing Technological Infrastructure Investment. The Federal
Government funded the country’s first investment in telecommunications
infrastructure, a telegraph line from Washington D.C. to Baltimore built in
the 1840s. But since then, appropriately, the vast majority of technological
infrastructure investment has been private. Over the course of decades, an
extraordinary expansion of telecommunications infrastructure made basic
telephone service available to nearly every resident of the country, far sooner
than in most other countries, which is a remarkable achievement given the
large size and relatively low population density of the United States.
Public policy encouraged these investments. Many private carriers, as
regulated monopolies, were permitted to charge high rates for long-distance
calls, business service, and the telephones themselves. A portion of the
resulting funds were required to be used to subsidize basic local phone service, particularly in rural and other areas that are costly to serve due to low
population density and geographic factors. The Telecommunications Act of
1996 sought to reform and improve upon telecommunications regulation

200  |  Chapter 5

by enabling greater competition, particularly in local and long-distance
telephone service, and by rationalizing, and making explicit, the subsidy
system supporting service in high-cost areas. Substantial additional private
investment followed.
In recent years, the U.S. Government has further facilitated private
telecommunications investment through favorable tax policy. In 2010, the
President proposed and signed into law the largest temporary investment
incentive in history—100-percent expensing—that, together with the bonus
depreciation that preceded and followed it, played a critical role in increasing and accelerating investment, including the substantial increases in
both wired and wireless investment in the telecommunications sector. For
example, two major companies in a joint statement said that, “despite the
downturn in the economy, the cable communications sector has been able
to continue steady investment and to retain jobs as a result of policies like
100-percent expensing.”
Catalyzing investment in mobile broadband infrastructure is especially important given the rapidly growing usage and the scarcity of electromagnetic spectrum for carrying wireless broadband traffic. In 2010,
Federal Communication Commission experts predicted that the needs for
broadband capacity will overwhelm available spectrum (the “spectrum
crunch”). If allowed to happen, this would result in higher prices for mobile
broadband services, as well as reduced growth in broadband-based innovation, services, and employment. The scarcity of broadband capacity can be
alleviated through increased investment (denser transmission infrastructure
means more traffic on a given spectrum frequency), fuller deployment of
spectrum already licensed to wireless carriers, spectrum license consolidation, technological advancement, and improvements in spectrum policy.
One important initiative is to seek to reallocate public spectrum when
it has a more valuable private use. The Federal Government is a major user
of spectrum, as Figure 5-7 shows. Most of this usage involves national security and law enforcement functions, as shown in Figure 5-8. Federal use of
spectrum is valuable, but it is not costless. As an economic matter, if a particular spectrum band would produce a larger net social surplus in private
hands than in public hands, then it should be reallocated, and vice-versa.
That is, the Federal Government can alleviate spectrum scarcity by having government agencies vacate certain spectrum bands entirely, or share
them with private users, when this can be achieved without compromising
the agencies’ vital missions (which in many cases involve safety-of-life and
national security) and when the associated costs of relocating government
operations out of those bands are justified by the social value that will be
unlocked as a result of the reallocation to the private sector. The vacated
Fostering Productivity Growth  |  201

Megahertz
1,600

Figure 5-7
Exclusive and Shared Allocation of Radio Spectrum
1,434

1,400
1,200

1,058

1,000
800

629

600

359

400
200
0

Exclusive Federal Exclusive Private Dominant Federal
(Shared)

Shared

Source: National Telecommunications and Information Administration (2009).

spectrum could then be auctioned off to commercial users or, if appropriate, made widely available on an unlicensed basis (more on this below). The
Federal user would be relocated to alternative spectrum that could be used
more intensively and economically, particularly if additional resources were
made available for investment in newer equipment.
In addition to economizing on its aggregate spectrum usage, the government can further alleviate spectrum scarcity by rationalizing spectrum
allocation. There are some spectrum bands that, above and beyond the
properties that make them valuable in general (for example, strong propagation through buildings and in rural areas), are particularly valuable for
commercial applications, such as if they are complementary to other commercial spectrum bands. In those cases, value can be unlocked by having the
government relocate from those bands to other bands that do not have that
property—again, under the condition that this can be done without compromising vital missions and that the relocation costs are not prohibitively high.
Box 5-4 describes several spectrum investment policies that have been
undertaken or proposed by the Administration.
There is also substantial scope to reallocate some spectrum currently
licensed to private entities to a more valuable use in wireless broadband.
Some incumbent firms, such as over-the-air broadcast television stations,
hold rights to spectrum that are much more valuable as wireless broadband

202  |  Chapter 5

Figure 5-8
Federal Agencies with Most Spectrum Assignments
Coast
Guard
Homeland
Security
Justice

Defense

Federal Aviation
Administration
Other Federal
Agencies

Note: "Other Federal Agencies" includes Interior, Agriculture, Energy, Commerce, and the other
remaining 48 agencies and departments with spectrum frequency assignments.
Source: National Telecommunications and Information Adminstration, Government Master file
(2010); Government Accountability Office Analysis.

spectrum. The 2010 National Broadband Plan introduced the idea of
“incentive auctions” as a tool to help meet the nation’s spectrum needs by
giving those rights-holders a share of the auction proceeds if they relinquish
their rights. In the Spectrum Act of 2012, Congress authorized the Federal
Communications Commission to conduct incentive auctions and directed
that the FCC use this innovative tool for an incentive auction of broadcast
television spectrum. In September 2012, the FCC adopted a Notice of
Proposed Rulemaking in order to develop a rulemaking record that will
enable the Commission to meet the challenges presented by the Spectrum
Act’s unique grant of authority. The magnitude of potential gains to social
surplus are enormous when broadcasters with access to new, more-efficient
transmission technologies that use less spectrum, or with a small and shrinking base of over-the-air viewers and annual revenue in the low millions of
dollars, will have an incentive to relinquish spectrum that, when reconfigured for commercial broadband use, will be sold for hundreds of millions of
dollars to companies that will use it to improve services for a vastly greater
number of broadband customers.
Some spectrum can be used effectively without being licensed at all,
but rather made available for anyone to use on an unlicensed basis. Just
as some roads seldom experience traffic jams, in some instances certain
spectrum bands do not become highly congested even when access is free.

Fostering Productivity Growth  |  203

Box 5-4: Spectrum Investment Policies
In 2010, the President issued a Presidential Memorandum
called “Unleashing the Wireless Broadband Revolution,” which
directed the Secretary of Commerce, working through the National
Telecommunications and Information Administration (NTIA), to collaborate with the Federal Communications Commission (FCC) to make
available a total of 500 megahertz (MHz) of Federal and nonfederal spectrum over the next 10 years, suitable for both mobile and fixed wireless
broadband use, nearly doubling the amount of spectrum available for
such purposes. The Secretary of Commerce has been facilitating discussions between agencies and nonfederal entities that have produced an
unprecedented level of information-sharing and collaboration to identify
opportunities for agencies to relinquish or share spectrum, currently
focusing on the 1695-1710 MHz band, the 1755-1850 MHz band, the
3550-3650 MHz band, and the 5350-5470 and 5850-5925 MHz bands.
The President’s fiscal year 2015 budget would invest $7.5 million
to monitor spectrum use by Federal agencies in high-priority markets to
identify opportunities for repurposing spectrum through auctions, while
protecting Federal missions. This budget proposal builds on the Middle
Class Tax Relief and Job Creation Act of 2012 by proposing to authorize
use of a spectrum license user fee for licenses not currently awarded
via auctions (for example, international satellite licenses), to promote
efficient utilization of spectrum. This fee would raise nearly $5 billion
over the next 10 years, and would continue to encourage more efficient
allocation and use of spectrum.

Unlicensed spectrum plays an important role in the broadband ecosystem,
enabling Wi-Fi, Bluetooth, “smart homes,” and more, which operate on
unlicensed spectrum using devices whose power is low enough that interference among numerous devices sharing the spectrum is not a major concern.
It also helps to alleviate scarcity in licensed spectrum bands. This is because
a great deal of mobile usage is not the “on-the-go/in transit” mobile usage
that must be transmitted on a carrier’s licensed mobile network, but rather
is so-called “nomadic” usage (for example, at home, office, or other fixed
location), that is amenable to carriage mostly by a wired broadband connection and then wirelessly completed using a nearby unlicensed Wi-Fi router.
For this reason, the licensed carriers are investing heavily in the deployment
and use of Wi-Fi networks. The value of this unlicensed spectrum has been
estimated at $16 billion to $37 billion per year.
In February, the FCC proposed to make available up to 195 megahertz of additional spectrum in the 5-gigahertz band for unlicensed wireless

204  |  Chapter 5

devices, a 35 percent increase. This band was selected for unlicensed use
in part because the presence of incumbent users of this band, including
automobile makers that have been developing short-range communications
capabilities that could greatly improve traffic safety and efficiency, make it
a poor candidate for being vacated and auctioned off for licensed use. To
unlock the value of the band for unlicensed use, the FCC has also proposed
to create a more flexible regulatory environment, and to streamline existing
rules and equipment authorization procedures for devices throughout this
band. Currently ongoing is the process of identifying regulatory changes
that strike the best balance between unlocking the value of this spectrum
for unlicensed use on the one hand, and avoiding harmful interference with
incumbent users on the other.
Clearing Federal Government spectrum for exclusive licensed use,
and making it available for shared unlicensed use, remain viable solutions
in the near term. However, given the dramatic spectrum challenge and the
fact that much of the lowest-hanging fruit for reallocation has already been
picked, it is also important to focus on newer and more innovative ideas.
These ideas include new advances in the sharing of spectrum between different users, particularly between government and private users. Innovation
in spectrum sharing is both promising and necessary, as there are some
spectrum bands that the government cannot vacate entirely, but that nevertheless have unused capacity, and that with appropriate processes and
procedures in place could be shared, accommodating some valuable private
usage without compromising mission-critical functions.
The President’s Council of Advisors on Science and Technology
(PCAST) released a report estimating that “in the best circumstances, the
amount of effective capacity that can be obtained from a given band of
spectrum can be increased thousands of times over current usage through
dynamic sharing techniques that make optimal use of frequency, geography,
time and certain other physical properties of the specific new radio systems
(PCAST 2012).”
The 2010 Presidential Memorandum that set the Administration’s
spectrum goal contemplated the sharing of Federal Government spectrum
as one means of achieving that goal. More recently, in June 2013, another
Presidential Memorandum established a Spectrum Policy Team in the
Executive Office of the President, which was charged with the mandate
to “monitor and support advances in spectrum sharing policies and technologies.” That Memorandum also contains measures to facilitate research,
development, testing, and evaluation of technologies to enhance spectrum
sharing and other spectrum-related efficiencies.

Fostering Productivity Growth  |  205

To stimulate investment in more advanced forms of spectrum sharing, the Defense Advanced Research Projects Agency (DARPA) is soliciting
innovative research proposals aimed at efficient and reliable sharing of
spectrum between radar and communications systems. Consistent with its
history of promoting groundbreaking technological breakthroughs for both
military and commercial use, DARPA is seeking “innovative approaches
that enable revolutionary advances” in spectrum sharing, specifically in the
spectrum bands that are most amenable to broadband and communications
services. The program may fund multi-year projects designed either to significantly modify existing radar and communications systems or to unveil
new system architectures redesigned from the ground up.
By itself, making additional Federal spectrum available for commercial use, whether on an exclusive or a shared basis, is unlikely to be sufficient
to keep up with the exploding demand for bandwidth. The ambitious goal of
freeing up 500 MHz of spectrum would nearly double the amount of wireless spectrum available for mobile broadband over the course of a decade,
but even that may not be enough to keep up with spectrum usage growth.
Therefore, it is important to do everything from increasing investments in
wired broadband networks that can offload some of the demand (often making the last connection wireless, but through Wi-Fi rather than cellular), to
increasing the density of wireless cells, to encouraging technological innovations for using spectrum more efficiently.
The Administration is trying to help with these efforts in a variety of
other ways, including the June 2012 Executive Order issued by the President
specifying a number of steps that will ease and facilitate carriers’ access to
Federal land and buildings for purposes of deploying broadband infrastructure, including cell towers.
Ensuring Everyone Benefits. It is important to ensure broad participation in the benefits of broadband telecommunications technologies, because
broad participation allows more people to use those benefits to develop their
talents, which lead to higher economic growth and higher living standards in
the future. One element of broad participation is ensuring that technology
and its products are affordable. To that end, vigorous antitrust enforcement
is critical to ensure that that prices are not inflated and choices not limited
by lack of competition. This has been a focus of the law enforcement agencies, and is also important as a policy consideration going forward.
The Obama Administration has made critical investments in expanding broadband to underserved communities. The American Recovery and
Reinvestment Act of 2009 included $6.9 billion in funding to upgrade the
nation’s broadband infrastructure, with $4.4 billion administered by the
Department of Commerce’s National Telecommunications and Information
206  |  Chapter 5

Administration, and $2.5 billion by the Department of Agriculture’s Rural
Utilities Service. Of these funds, a total of $4.4 billion (as of the end of May
2013) went to fund more than more than 325 broadband projects through
the Broadband Technology Opportunities Program and the Broadband
Initiatives Program. The Federal Communications Commission has also
played an important role in expanding broadband deployment in unserved
and underserved areas through Universal Service Reform and the establishment of a $4.5 billion annual Connect America Fund, which reallocates
funds previously used to support voice service.
Education researchers have long believed that technology holds the
potential to profoundly impact the classroom experience, from allowing
students to interact with course content in new and personalized ways to
helping teachers understand what lessons and techniques are most effective.
By making the ever-expanding collection of educational resources available
on the Internet accessible to teachers and students in classrooms, technologically equipped schools enhance learning by gaining access to those
resources, rather than being limited to resources that are physically at hand.
Although more high quality research on the effectiveness of online
educational tools is still needed, these tools do show promise. A metaanalysis of experimental or quasi-experimental studies of the effects of
online education conducted by the Department of Education in 2010 found
that students who receive instruction that combines online and face-to-face
elements performed better than students who received either exclusively
online, or exclusively face-to-face, instruction. Other factors such as instruction time or curriculum may contribute to this positive effect, but the metaanalysis suggests that further research on designing, implementing, and
evaluating these blended approaches may be worthwhile.
Instruction methods that incorporate computers have also shown
promise in mathematics education. Barrow, Markman, and Rouse (2009)
found that students who were randomly assigned to participate in and complete computerized math lessons at their own pace scored 0.17 to 0.25 standard deviations higher on mathematics achievement tests than students who
received traditional instruction. Computer-aided mathematics instruction
has been shown to have similar effects in other contexts. In an experimental
study, Banerjee et al. (2007) find that playing educational math games on
computers for two hours a week improved the math scores of impoverished
elementary school students in India by 0.47 standard deviations. In another
experiment, Carillo, Onofa, and Ponce (2010) find poor Ecuadoran elementary school students who used adaptive math and language software for
three hours a week improved their math scores by 0.30 standard deviations.

Fostering Productivity Growth  |  207

Because current uses of technology can enhance learning and the
potential of future developments is untold, it is critically important that all
students have access to 21st century classrooms. The ConnectED program,
announced by President Obama in June 2013, takes important steps to
ensure that the benefits of improvements in educational technology will
be made widely available. While initiatives like the FCC’s E-Rate program,
established under the Telecommunications Act of 1996, have helped bring
Internet access to almost every school in the nation, many schools do not
have access to the fast broadband speeds enjoyed by most businesses and
households. Further, the E-Rate program was designed primarily to bring
Internet access to the school, with less priority on ensuring that access was
available throughout the school, such as via Wi-Fi technology. As a result,
62 percent of school districts say their bandwidth needs will outstrip their
connections within the next 12 months, and 99 percent say that this will
happen within three years.
ConnectED will bring high-speed broadband and wireless Internet
access to 99 percent of America’s students in their school classrooms and
libraries within five years. To make the most of this enhanced connectivity,
ConnectED will refocus existing professional development funds to train
teachers to take full advantage of these resources in order to improve student
learning. Finally, by equipping schools with the broadband Internet access
necessary to make use of high-tech educational devices, ConnectED will
deepen the market for such devices, as well as the digital educational content
with which they interact, spurring private-sector innovation in this area.
The President has called on the FCC to modernize the E-Rate program, and has also called on the expertise of the NTIA, in order to deliver
this connectivity and meet the goal of connecting 99 percent of America’s
students to the digital age within five years through next-generation broadband and high-speed wireless in their schools and libraries. Answering that
call, the FCC announced in February 2014 that it would invest $2 billion to
connect 20 million students over the next two years, representing a crucial
down-payment on reaching the President’s goal. The initiative, however, is
not just about infrastructure. The President announced in February over
$750 million in private sector commitments to help fill out this vision of
a connected classroom through the digital devices, content and learning
software, home wireless access, and teacher training necessary to make the
best possible use of this infrastructure. By leveraging all these resources, we
are making substantial progress toward a world-class education for every
student that does not depend on their family’s income or on the zip code in
which they were born.

208  |  Chapter 5

Finally, it is crucial that the benefits of broadband technology growth
be consistent with privacy and security. Also, the free expression of ideas
must be protected, so technological development must proceed in a way that
is consistent with an open Internet.

Challenges to Broad Adoption of Telecommunications Technology
Broad adoption of telecommunications technologies faces several
challenges. For example, these technologies are unevenly adopted across
different education and income levels. Home broadband adoption is more
than twice as high for college graduates as for high school dropouts. Overall,
30 percent of Americans do not use broadband at home, and many of these
non-users are in lower-income households. Rural areas also lag in adoption. As illustrated in Figure 5-9, nearly all urban residents have access to
6 megabits per second downloads, but only 82 percent of residents in rural
communities can access those speeds, and the disparity becomes even larger
at faster speeds.
One reason some households do not adopt broadband is cost: unlike
the sharp price declines seen for technological hardware, such as computers,
the prices consumers pay for Internet access have remained steady or risen.
But while broadband prices have not fallen sharply, the speeds that are available at a given price today are often considerably faster than the available
speeds at the same price several years ago, which means that value for money

Fostering Productivity Growth  |  209

Box 5-5: Electronic Health Records
Technological advances in Health Information Technology, especially Electronic Health Records (EHR), hold the promise of improving
patient care and lowering health care costs. Patients are often treated
by multiple providers for the same condition or for related conditions.
Because the correct treatment by one provider often depends upon what
other providers are doing, effective coordination of care between providers can improve health outcomes. Effective coordination also helps to
control costs, as it avoids both the costs of treating follow-on problems
resulting from uncoordinated care and the unnecessary duplication of
tests and procedures.
Some ways of improving care coordination among health care
providers involves changing the way they are paid for their services. The
Affordable Care Act of 2010 (ACA) included a variety of such reforms
that are currently in various stages of implementation, many of which
are discussed in Chapter 4. But other ways involve the application of
better technology, notably EHR systems. As the name would suggest,
these systems enable the creation of a permanent, sharable record of
every aspect of a patient’s care, including test results, past treatments,
and providers’ notes. In a fully integrated EHR system, each provider
has immediate and complete access to all relevant patient information,
which has the potential to greatly improve coordination of care and also
to reduce medical errors.
EHR systems have additional functionality as well, such as providing automatic alerts when treatments are inconsistent with each other or
when a scheduled test has been missed. The systems can also be used to
improve quality more broadly by allowing hospitals and other providers
to keep better track of outcomes and to identify problem areas.
EHR adoption has been promoted by Administration policy.
The Health Information Technology for Economic and Clinical
Health (HITECH) Act, enacted as part of the American Recovery and
Reinvestment Act of 2009, encouraged adoption and use of health information technology, including EHR systems.
Key programs established by the HITECH Act were the Medicare
and Medicaid EHR programs. These programs provide financial incentives to hospitals and health care professionals to adopt EHR systems,
and require that they demonstrate “meaningful use” of the systems. The
meaningful use criteria, which become increasingly rigorous over time,
require providers to demonstrate that they are using EHR systems to
capture patient health information, assist in clinical decision making,
track quality of care, and securely exchange patient information across
health care settings to facilitate coordinated care.

210  |  Chapter 5

Providers who adopt and demonstrate meaningful use of EHR
systems by 2014 (for Medicare) and by 2016 (for Medicaid) are eligible
for bonus payments from those programs. The Medicare program, but
not the Medicaid program, also includes a payment reduction to providers that do not adopt and demonstrate meaningful use of EHR systems.
Medicare providers who have not demonstrated meaningful use by 2015
are subject to penalties that grow over time; for example, for physicians,
penalties start at 1 percent in 2015 and grow to 3 percent or more in
subsequent years. The Congressional Budget Office estimated that the
Medicaid EHR program would award bonuses of $12.7 billion through
2019, while the Medicare EHR program would make bonus payments
net of penalties of $20 billion over that period (CBO 2009).
The HITECH Act also provided $2 billion to the Department of
Health and Human Services to fund activities to encourage the diffusion
of health information technology, such as investing in infrastructure
and disseminating best practices. The Act also made a variety of other
changes, including provisions to facilitate data sharing across health care
providers to support coordinated care and protect patient privacy.
The share of medical providers using EHRs has risen dramatically
in recent years. Data from the National Ambulatory Medical Care Survey
show that the share of office-based physicians using an advanced EHR
system (which are generally more sophisticated than those required meet
the early-stage “meaningful use” criteria) rose from 17 percent in 2008 to
40 percent in 2012 (Hsiao and Hing 2014), and data from the American
Hospital Association’s annual survey of hospitals show that the share
of hospitals that had adopted such a system rose from 9 percent to 44
percent over the same period (Charles et al., 2013). Consistent with this
rapid progress for advanced systems, the Department of Health and
Human Services has estimated that, as of the end of April 2013, over half
of eligible physicians and more than 80 percent of eligible hospitals have
adopted an EHR system and met the criteria for meaningful use (HHS
2013).

has improved. Further, while international comparisons are difficult (due to
variations in factors like taxes, government subsidies, geography, population
density, and product bundles), the United States compares favorably in a
number of respects, including entry-level pricing for slower but still useful
broadband speeds.
A surprisingly large number of households cite a different factor for
their decision not to subscribe to home Internet service: a perceived lack of
relevance to their day-to-day lives. Private- and public-sector broadband
adoption programs address this by focusing on educating non-subscribers
Fostering Productivity Growth  |  211

about the array of services and support mechanisms that are available online,
like job listings and training, educational tools, health care services, and
government resources.
While this chapter has focused heavily on telecommunications technology, there are many other areas where technological advances promise
large social and economic benefits, and where public policy can play an
important role. One important example, discussed in Box 5-5, is Electronic
Health Records and related technologies.

Patents
The rights that prospective innovators have to the economic returns
on their innovations are known as intellectual property (IP) rights, of
which one major category is patents, which apply to inventions. Patents are
granted on inventions in many different areas of technology, as shown in
Figure 5-10. The basic economic logic behind patent protection is simple:
successful inventions are valuable to society, as they lead to new and better
products. But attempting to invent is costly and risky. If successful inventions could be easily imitated by competitors, then prospective inventors
may be in a position where they lose if their invention fails, but gain little or
nothing if it succeeds. This diminishes the incentive to expend resources and
effort on inventing, and the reduced rate of invention is harmful to society.
To prevent this problem, patent protection allows inventors to enjoy a temporary exclusive right to their invention. The super-competitive pricing that
results from this exclusivity provides an incentive to invest. Another benefit
of patent protection is that patents are published, so they can be licensed
and put to other socially valuable purposes other than those of the inventor.
But patent protection can also harm consumers: for inventions that would
have been created with weaker patent protection or even with no protection
at all, patents simply lead to higher prices for the same inventions, not to
additional inventions. The economically optimal strength of patent protection (for example, how many years a patent should run) is the one that best
balances the benefit from accelerated invention with the harm from higher
prices.
There are some additional effects of patent protection that also
deserve mention. One effect is that some inventions are complementary to
each other, meaning that the availability of one makes it easier to develop
others. In those cases, the higher prices resulting from patent protection,
as well as the related legal and administrative burdens (such as negotiating
licenses), raise the cost of, and hence reduce the rate of, subsequent innovation. This effect is relevant for determining the economically optimal patent

212  |  Chapter 5

Figure 5-10
Patents Issued in the U.S. by Technological Category, 2003‒2012

Count
140,000

Electrical and Electronics
Drugs and Medical
Computers and Communications
Unknown

120,000

Chemical
Mechanical
Other

100,000
80,000
60,000
40,000
20,000
0

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

Note: "Unknown" indicates that a patent could not be assigned to a single technological category
with a high enough probability.
Source: United States Patent and Trademark Office.

strength. Another effect is that, in some cases, patent rights can be used to
harm rival firms or to extract license fees that do not correspond to the value
of the patent. As discussed below, it is important to curb such behaviors by
developing sound policies related to patent examination and enforcement.
It is also important to ensure that patents are not wrongly issued, but rather
are only issued for inventions that are non-obvious, useful, and inventive.
The chapter now turns to two specific patent issues that have been the
subject of recent policy scrutiny: how to support standard setting by appropriate use of standard-essential patents, and the activities of Patent Assertion
Entities and the effects of those activities.

Standard-Essential Patents
We take for granted that we can drive our car up to a gas pump and
have the hose fit the car’s nozzle. Similarly, that smartphones created by different manufacturers will communicate with each other. These are examples
of interoperability that result from the standardization of certain product
features. An interesting problem arises when an industry seeks to adopt an
interoperability standard and the available choices for the standard may
include patented inventions.
The nature of the economic problem is to develop a mechanism that
determines when standardization would make market participants better off
Fostering Productivity Growth  |  213

and, in such cases, provides parties with incentives to invent and propose
the invention as a standard, while ensuring that all parties will later find
it in their interest to implement the standard. The central premise of the
economic theory of patents is that granting limited exclusive rights through
the issuing of patents provides an incentive for private investment in invention; absent such rights, the entity making the investment may not receive
sufficient returns to make the investment worthwhile. These exclusive rights
are not meant to preclude similar technologies from being developed and
marketed. In principle, some degree of competition in consumer markets
bounds the power conferred by these exclusive rights. But that bound is
removed if a patented technology becomes the standard and is used in all
products sold in the market. As a result, the patent holder may seek to charge
higher prices than originally agreed on during the standard-setting process
and to use the patent to inefficiently restrict access to the technology. Such
behavior may delay implementation of the technology, as others who may
adapt the technology exit the market or seek other ways around it. The
sought-after standard then fails to become standard (for example, Gilbert
2011).
	 Because industry actors are most likely to understand the substantial complexities of new technologies and the potential products and markets
for their dissemination, there is value in having the standards set voluntarily
by industry-based standards-developing organizations (DOJ 2013). These
organizations provide a place for industry actors to propose their patented
technologies as part of a standard, and to reach consensus on the technologies incorporated into the standard (or to decide on no standard). After a
decision is taken, a chosen patent becomes known as standard essential.
Actual implementation of the agreed-on standard as an observed standard
follows when all implementers and potential implementers pay an agreed
justified price (reasonable, or both fair and reasonable) for the technology,
and their access to the patented technology cannot be improperly restricted
(there is not discrimination). By proposing a patent for use in the standard,
the patent owner is giving up the power to charge higher per-unit prices for
use of the technology, but enjoys returns from the diffusion of the technology more widely across more units.
	 Because the notion behind standard-developing organizations is
voluntary collaboration, there is no guarantee that a standard will be produced. A standard-essential patent holder can refrain from committing to
licensing on reasonable and non-discriminatory (RAND) terms.14 In such
cases, the declared standard is less likely to be the implemented standard and
14 Sometimes the licensing commitment is to fair, reasonable and non-discriminatory terms.

214  |  Chapter 5

market forces may be suggesting that a standard is not needed or may be best
determined over time and in the marketplace directly (Farrell et al. 2007).
When voluntary agreement does produce a standard, there are
instances when parties to the agreement do not feel that others are living up
to the agreement. In such instances, when patent holders have committed
to license on RAND terms, judicial and enforcement procedures should aim
to reproduce the intent of the agreement; that is, to ensure that the patent
holder receives a RAND royalty. Otherwise, judicial and enforcement procedures can tip the balance of power in favor of one party or the other, leading
either to excessive market power in the hands of the patent holder or to nonpayment of reasonable royalties by implementers, and to greater incentives
against establishing a standard in the first place (Lemley and Shapiro 2005).

Patent Assertion Entities
In recent years, organizations known as Patent Assertion Entities
(PAEs) have become common. PAEs brought 24 percent of all patent lawsuits in 2011, and over the 2007-11 period they brought approximately onefifth of all patent lawsuits, covering about one-third of all defendants (GAO
2013). These PAEs purchase rights to patents belonging to other firms, and
then assert them against firms or individuals who are using the patented
technology. Some of this activity is valuable: incentives to invent are stronger
if inventors know they can later sell their patent to, or merely engage the
services of, a PAE that can assert it more effectively than they could do themselves. Also, in some cases, it may be efficient for PAEs to act as intermediaries by obtaining the rights to patents held by disparate inventors in order to
decrease the transaction cost of negotiating licenses. However, many industry observers believe that PAEs often do not assert patents in good faith, but
rather assert them simply in order to extract nuisance payments from firms
looking to avoid costly and risky litigation. In some cases, these patents are
valid but of low value, meaning that absent the high cost of litigation they
would only command very low licensing fees. In other cases, the patents are
invalid (or not infringed), and absent the high litigation costs they would not
command any license fees at all (Scott Morton & Shapiro 2013).
	 This issue is particularly pronounced in smartphones and other
consumer electronics devices (Chien 2012). Many of these products contain
technology based on thousands of patents, and as shown in Figure 5-10
above, the number of patents issued in the “Electrical and Electronics” category has been increasing over the past decade. The large number of patents
embodied in these products, and their complexity, often makes manufacturers subject to patent-infringement claims, with the associated threat of costly
and risky litigation, from owners of low-value valid patents or even from
Fostering Productivity Growth  |  215

Box 5-6: The Leahy-Smith America Invents Act
The Leahy-Smith America Invents Act of 2011 took some important steps to update the U.S. patent system. The Act changed the
system to give priority to the first inventor to file for a patent on a given
invention, moving away from the “first-to-invent” priority system and
bringing the United States in line with every other industrialized nation.
This change eliminated the need for long, expensive administrative
proceedings to resolve disputes among inventors who filed for patents
on the same invention over who invented it first.
The Act also helps ensure that inventors have the opportunity to
share their work early on by maintaining a form of the one-year “prior
art” grace period that had been a feature of the previous system. The grace
period excludes from the previous state of knowledge, against which the
originality of a patent application is judged, any disclosures of details of
the invention made within the year preceding the application date by the
inventor, or by third parties who learned them from the inventor. The
grace period allows inventors to publish their work, prepare application
materials, or seek to raise funds to support their application without fear
that those activities will later be a detriment to that application.
The Act also increases protections from patent infringement lawsuits for innovators who develop and deploy new products or methods
but choose not to patent them, a common practice in the high-tech
industry, by expanding the “prior user rights” infringement defense.
Formerly applicable only to business practices patents, this defense—
which exempts from liability users who can demonstrate that they
independently developed and used the patented product or method that
they are accused of infringing upon, and did so more than a year prior
to the date the patent was filed—is now applicable to all types of patents.

owners of invalid patents. It is therefore an important public policy goal to
find ways to reduce the cost of defending patent lawsuits, and also to reduce
the number of invalid patents, either by reducing the number of invalid
patents that are granted, or by making it easier for them to be challenged.
One important step toward resolving these patent-related problems,
which disrupt the appropriate economic incentives to invent, has been taken
in the form of the Leahy-Smith America Invents Act of 2011, discussed further in Box 5-6. The key provisions of the AIA, which went into full effect in
2012, are helping to improve the patent system for innovators by offering a
fast-track option for patent processing, taking important steps to reduce the
current patent backlog, and increasing the ability of Americans to protect
their intellectual property abroad.

216  |  Chapter 5

Several provisions of the America Invents Act may help address
some of the problematic behavior of PAEs by developing at the Patent and
Trademark Office new programs to create alternatives to litigation over
patent validity, new methods for post-grant review of issued patents, and
major steps to increase patent quality through clarifying and tightening
standards. Yet challenges remain, notably the asymmetry between the cost
to a PAE of bringing a patent lawsuit and the cost to a target firm of defending one, which enables PAEs to bring weak cases in the hopes of extracting
a settlement.
In June 2013, the President issued a set of five executive actions
and seven legislative recommendations to address these challenges. These
included measures to make it more difficult for overly broad claims to
receive patents in the first place, as well as to make it easier to challenge
weak patents once they have been granted. The President’s priorities also
include measures to require greater clarity in patent applications regarding
the precise nature of the claimed invention, as well as the identity of the patent holder. Other measures include ways to make it more difficult for patent
holders to sue end-users (as opposed to manufacturers) of products that
contain patented technology, and to provide judges with more discretion to
award attorney fees and other costs to the prevailing party in patent lawsuits.
Congress has taken up these issues. In December 2013, the House
of Representatives passed a bipartisan bill containing many of the
Administration’s priority items. A related bill is currently under consideration in the Senate Judiciary Committee.
Another important policy issue related to patents is the phenomenon
of “pay-for-delay” settlements of patent lawsuits in the pharmaceutical
industry. This is discussed in Box 5-7 below.	

Conclusion
Productivity growth allows a given set of scarce resources to yield
more output and a higher aggregate standard of living. When private actors
face incentives that lead them to optimal investments in growth-enhancing
technologies, government policy should be to not interfere. But at other
times, a light touch from government is needed to align incentives or to act
in place of incentives that are missing: in the form of conducting of its own
research; or of subsidization of private research; or through appropriate
intellectual property rights laws, regulation, and enforcement. Government
also has a role to play in ensuring that all citizens benefit from productivity
advances that can increase living standards—a step that can form a virtuous
cycle that also increases productivity growth itself by tapping more of the
potential of our citizens.
Fostering Productivity Growth  |  217

Box 5-7: Pay-For-Delay Settlements in Pharmaceutical Patent Cases
Out-of-court settlements of lawsuits are usually socially beneficial,
as they allow disputes to be resolved without a costly trial. There are circumstances, however, where settlement of a patent lawsuit can be used as
a means of extending market power, rather than as a means of efficiently
resolving the dispute. In recent years, this has been a significant issue for
certain cases involving pharmaceutical patents. In these cases, an incumbent seller of a branded drug files a patent infringement suit against one
or more companies seeking Food and Drug Administration approval to
sell a generic version of that drug. The patent at issue is often not the one
covering the drug’s active ingredient (for which assessing infringement
is usually less complicated), but is instead a secondary patent, such as
one covering a particular formulation of the drug (Hemphill and Sampat
2011). The generic entrant will deny the infringement, claiming that the
patent is invalid, that its product does not infringe, or both. A patent
lawsuit results, and is settled through an agreement specifying a date on
which the generic entrant may begin selling its product, which is some
time after the date of the settlement and before the patent expires. The
settlement will also specify a payment from the branded incumbent to
the generic entrant. Absent the settlement, the case would have gone
to trial; had the incumbent won, the generic entrant would have been
barred from entry until the end of the incumbent’s patent term, and had
the entrant won, it could have entered immediately and sold its product,
assuming it had received FDA approval.
The willingness of the incumbent to agree to such a settlement
may seem puzzling, as the payment appears to go the “wrong” way, from
the alleged infringer to the infringed. But the ability to enter into such
settlements can benefit the incumbent by enabling it to “purchase” later
generic entry than would otherwise occur. In other words, settlements
of patent disputes can be used as a vehicle for extending market power.
What drives these settlements is the fundamental economic
principle that the profits of a single seller of a product are greater than
the combined profits of two or more sellers, because a single seller has
greater market power and so can extract a higher price from consumers.
A settlement that delays generic entry of a drug therefore increases the
aggregate profits on that drug. These extra profits create an incentive
for a deal in which entry is delayed; both parties will accept such a deal
as long as the extra profits are divided in such a way that each party is
better off than it would be absent the deal (i.e., better off than by letting
the patent lawsuit proceed to trial). For this reason, these settlements are
often called “pay-for-delay” settlements.
Pay-for-delay settlements undermine existing laws (most notably
the Hatch-Waxman Act) that encourage the development of generic
218  |  Chapter 5

drugs. When generic drugs enter a market, they are offered at a much
lower price than is the branded drug, and they typically capture a large
market share. For these reasons, generic entry results in considerable
savings to consumers and to the health care system. The delay of generic
entry due to pay-for-delay settlements greatly reduces those savings.
The ability of incumbent patent holders to enter into pay-fordelay settlements, and to thereby maintain their patent protection for a
longer period, might be viewed as increasing the value of pharmaceutical
patents, and hence increasing the incentive to invest in discovering new
drugs. However, the value of any increased innovation arising from these
settlements may be relatively small. The most socially valuable drug patents are often those covering new molecular entities. These patents are
relatively unlikely to be successfully challenged, which means the generic
entrant has little prospect of victory at trial or of a lucrative pay-for-delay
settlement. As a result, banning such settlements may not significantly
affect the incentive to invest in inventing new molecular entities. Instead,
pay-for-delay settlements often involve patents on incremental improvements to existing drugs, often ones that make the drug just different
enough that a prescription for the new version cannot be filled with an
existing generic equivalent of the old version. The ability to enter into
pay-for-delay settlements does encourage this type of innovation, but the
social benefits are likely to be comparatively small in many cases.
Pay-for-delay settlements have been the subject of a considerable
amount of litigation, culminating in a 2013 Supreme Court decision
in FTC v. Actavis, involving a drug called AndroGel. The Court ruled
that “pay for delay” settlements are not presumptively unlawful, but
are also not immune from antitrust scrutiny, partially resolving earlier
conflicting rulings by lower courts (see FTC v. Actavis 2013). The Court
did not establish a concrete rule regarding how such settlements should
be treated, however, so substantial uncertainty remains about how these
lawsuits will be adjudicated in practice.
The Administration has proposed legislation that gives the Federal
Trade Commission explicit authority to stop companies from entering
into pay-for-delay agreements. For the reasons described above, such
authority would likely generate billions of dollars in savings for consumers, and also for the Federal Government through lower pharmaceutical
prices paid by Medicare, Medicaid, the Department of Defense, and the
Veterans Administration (see CBO 2011 and FTC 2010).

Fostering Productivity Growth  |  219

C H A P T E R

6

THE WAR ON POVERTY 50 YEARS
LATER: A PROGRESS REPORT

P

resident Lyndon B. Johnson declared an “unconditional war on poverty
in America” on January  8, 1964, and within a few years oversaw the
creation of an array of programs “aimed not only to relieve the symptom of
poverty, but to cure it and, above all, to prevent it.” In the 1964 Economic
Report of the President, President Johnson’s Council of Economic Advisers
outlined the many key points of attack: “maintaining high employment,
accelerating economic growth, fighting discrimination, improving regional
economies, rehabilitating urban and rural communities, improving labor
markets, expanding educational opportunities, enlarging job opportunities
for youth, improving the Nation’s health, promoting adult education and
training, and assisting the aged and disabled.” The report ended with the
declaration that, “It is time to renew our faith in the worth and capacity of
all human beings; to recognize that, whatever their past history or present
condition, all kinds of Americans can contribute to their country; and to
allow Government to assume its responsibility for action and leadership in
promoting the general welfare.”
The War on Poverty ushered in a new era of Federal Government
leadership in providing income and nutrition support, access to education, skills training, health insurance and a myriad of other services to
low-income Americans. During President Johnson’s term, Congress passed
more than a dozen major pieces of legislation that provided such foundational elements of our current social welfare system as the Civil Rights
Act, the Economic Opportunity Act, the Food Stamp Act, Elementary and
Secondary Education Act, the Manpower Act, Medicare, Medicaid, the
Higher Education Act and the Child Nutrition Act. Since then, many of
these programs have been reformed and updated, ensuring that the modern
safety net assists families when they need it most, while also keeping them
connected to the labor force.

221

Optimism was high at the outset of the War on Poverty. One of its
architects predicted poverty would be eradicated within a generation and
saw “on the horizon a society of abundance, free of much of the misery and
degradation that have been the age-old fate of man (Council of Economic
Advisers 1964).” The 50th anniversary of President Johnson’s bold declaration provides an opportunity to assess our achievements in reducing
poverty and to evaluate the record of the poverty-fighting programs created
or enhanced in the wake of his declaration. There is no question that the
material conditions of those in poverty have improved: the percentage of the
poor with indoor plumbing has risen from 58 percent in 1960 to 99 percent
in 2011;1 infant mortality in counties with the highest levels of deprivation
has fallen from 23.2 per 1,000 in 1969 to 9.1 per 1,000 in 2000;2 and today
all American children in poverty have access to affordable health insurance,
as do poor adults in states that have taken up the Affordable Care Act’s
Medicaid expansion. But to what extent have we reduced the proportion
of Americans living in poverty? Indeed, the poverty rate has declined considerably since the beginning of the War on Poverty, but how much of this
improvement is due to the efforts of government? And what are the lessons
this contains for future policy?
This chapter answers these questions by first confronting the challenges in measuring poverty, and highlighting the limitations of the official
poverty measure for tracking progress in the War on Poverty. Using new
historical estimates of poverty based on modern measurement methods, this
chapter presents a more accurate picture of the changes in poverty over the
past five decades, and estimates the contribution of the safety net to these
changes. While the Council of Economic Advisers (CEA) reviewed research
on the effects of antipoverty programs on work and earnings, health, food
security, educational attainment, and other valued outcomes, this chapter
focuses primarily on their impacts on poverty and economic mobility.3
Finally, it discusses the role that President Obama has played in reducing
material hardship among low-income Americans by expanding access to
affordable health insurance and tax credits for working families, and by his
proposals to help ensure that no parent who works full time will have to raise
his or her children in poverty.
1 CEA calculations using 1960 Census and 2011 American Community Survey data.
2 These figures come from Singh and Kogan (2007). The analysis compares birth outcomes
in counties in the top and bottom quintiles of a socioeconomic deprivation index based on
Census information on education, occupation, wealth, income distribution, unemployment,
poverty, and housing quality. The data are broken out into 5-year periods—the statistics cited
compare 1969-74 to 1995-2000.
3 See Bailey and Danziger (2013) for an authoritative assessment of the various components of
the War on Poverty from economists’ perspectives.

222  |  Chapter 6

Measuring Poverty: Who is Poor in America?
Michael Harrington’s influential 1962 book The Other America
depicted the poor as inhabiting an “invisible land,” a world described in
the 1964 Economic Report of the President as “scarcely recognizable, and
rarely recognized, by the majority of their fellow Americans.” One early
achievement of Johnson’s War on Poverty was to cast light on the problem
of poverty by developing an official poverty measure that has been released
by the government in each year since August 1969.4 While reasonable at the
time, this measure has turned out to be ill-suited to capturing the progress
subsequently made in the War. As a result, modern poverty measures tell a
different story of who is poor, and especially how this has changed over time.

Measuring Poverty
Measuring poverty is not a simple task; even defining it is controversial. Just starting with a commonsense definition of the poor—“those whose
basic needs exceed their means to satisfy them”—requires difficult conceptual choices regarding what constitutes basic needs and what resources
should be counted in figuring a family’s means. There are no generally
accepted standards of minimum needs for most necessary consumption
items such as housing, clothing, and transportation. Moreover, our ideas
about minimum needs may change over time. For example, even some
middle-income households did not have hot and cold running water indoors
in 1963. Today, over 99 percent of all households have complete indoor
plumbing.

The Official Poverty Measure
Mollie Orshansky, an economist in the Social Security Administration,
developed the official poverty thresholds between 1963 and 1964 (Fisher
1992). At the time, the U.S. Department of Agriculture had a set of food
plans derived using data from the 1955 Household Food Consumption
Survey, the lowest cost of which was deemed adequate for “temporary or
emergency use when funds are low.” Because families in this survey spent
about one-third of their incomes, on average, on food, Orshansky set the
poverty threshold at three times the dollar cost of this “economy food plan,”
with adjustments for family size, composition, and whether the family lived
on a farm.
These income thresholds that were first used as the poverty thresholds
for the 1963 calendar year have served as the basis for the official poverty
4 A similar poverty measure was adopted internally by the Office of Economic Opportunity in
1965.

The War On Poverty 50 Years Later: A Progress Report  |  223

Box 6-1: Flaws In The Official Poverty Measure
The official poverty measure (OPM) has several flaws that distort
our understanding of both the level of poverty and how it has changed
over time. Perhaps the most significant problem with the OPM is
its measure of family resources, based on pre-tax income plus cash
transfers (like cash welfare, social security, or unemployment insurance
payments), but not taxes, tax credits, or non-cash transfers. As such
it inhabits a measurement limbo between “market poverty” (based on
pre-tax, pre-transfer resources) and “post-tax, post-transfer poverty”
reflecting well-being after taking into account the impact of policies
directed at the poor.
Several other shortcomings are more technical. First, the dollar
value defining the cost of basic needs, or the poverty threshold, was set
in the 1960s and has been updated each year since using an index of food
prices at first, but then the Consumer Price Index (CPI) after 1969. The
use of the CPI is one source of the problems with the official measure
that limits its usefulness for historical comparison. Methodological
advances in measuring prices (for example, rental equivalence treatment
of housing costs, quality adjustments for some large purchases, and geometric averaging for similar goods) have shown that the CPI overstated
inflation substantially prior to the early 1980s, leading to an inflated
estimate of the cost of basic needs and thus higher measured poverty
over time. Revising the OPM measure using the CPI-U-RS (a historical
series estimated consistently using modern methods) results in a fall in
poverty from 1966 to 2012 that is 3 percentage points greater than that
depicted by the official measure. Another flaw with the OPM thresholds
is that they do not accurately reflect geographic variation in costs of living or economies associated with family size and structure.
All current income-based poverty measures, including both the
OPM and the Supplemental Poverty Measure (SPM), suffer from large
underreporting of both incomes and benefits. For example, Meyer, Mok,
and Sullivan (2009) show that in 1984, March CPS respondents reported
only 75 percent of Aid to Families with Dependent Children/Temporary
Assistance for Needy Families (AFDC/TANF) dollars, and this fell to 49
percent in 2004. For SNAP benefits, which are accounted for in the SPM
but not the OPM, 71 percent of the value was reported in 1984 compared
to 57 percent in 2004. Underreporting will tend to increase measured
poverty, so increases in underreporting over time understate the decline
in the poverty rate during this period. The underreporting also means
that the estimated effects of government programs on poverty, as
described below, are likely to be conservative lower-bound estimates of
the true effects.

224  |  Chapter 6

thresholds ever since. These dollar amounts have been adjusted for inflation
to hold the real value of the income needed to be above poverty the same
over time. There have been minor tweaks to the methodology involving
which price index is used to adjust for inflation, and how adjustments are
made for family structure and farm status.
In defining the family resources to be compared to the poverty line,
Orshansky created the thresholds to be applied to after-tax money income—
the income concept used in the 1955 Household Food Consumption Survey.
However, she was forced to use pre-tax money income (including cash
transfer payments) due to limitations in the Current Population Survey
(CPS) data, the only source of nationally representative information on
income. At the time, this was an adequate approximation of disposable
income, as few low-income families had any federal income tax liability
or credits owed, and in-kind transfers were not a quantitatively important
feature of the safety net.

The Supplemental Poverty Measure
While Orshansky’s measure provided a reasonable depiction of
poverty in the 1960s, it has not aged well.5 Today, for example, the value of
the two largest non-health programs directing aid to the poor—the Earned
Income Tax Credit (EITC) and Supplemental Nutrition Assistance Program
(SNAP)—are entirely ignored by the official measure, making it impossible
to assess the success of these tools in fighting poverty. Over the past five
decades, researchers have pointed to many flaws in the official measure
(Box 6-1), leading to the development of alternative measures of poverty
with more comprehensive measures of both family needs and resources.
The Census Bureau created a Supplemental Poverty Measure (SPM), which
departs dramatically from the official measure in its methodology for calculating both the poverty thresholds and family resources.6 This measure, first
published in 2011, calculates poverty thresholds using recent expenditures
by families at the 33rd percentile of the expenditure distribution on an array

5 It should be noted that Orshansky herself noted many flaws with her measure, and believed it
understated poverty. She argued it measured income inadequacy rather than adequacy, stating
“if it is not possible to state unequivocally ‘how much is enough,’ it should be possible to assert
with confidence how much, on an average, is too little” (Orshansky 1965).
6 While the term “family” is used here, the SPM differs in its definition of a “family unit” when
assuming the unit of individuals over which resources are shared. Most importantly, the SPM
includes all related individuals living at the same address, but also cohabiting individuals and
co-residing children in their care.

The War On Poverty 50 Years Later: A Progress Report  |  225

Box 6-2: A Consumption Poverty Measure
Consumption-based poverty measures the amount households
spend relative to a threshold on goods and services, and estimates
“service flows” from large, infrequent purchases like housing and automobiles. Meyer and Sullivan argue in a series of papers (2003, 2012a,
2012b, 2013) that consumption provides a better measure of resources
for low-income households since it reflects accumulated assets, expected
future income, access to credit, assistance from family and friends,
non-market income, and the insurance value of government programs.
They also argue that consumption data are more accurately reported
relative to some safety net benefits included in the SPM, especially for
households in or near poverty.
The figure below shows the trend in Meyer and Sullivan’s (2013)
measure of consumption poverty, updated through 2012 and normalized
to have the same level as the SPM measure in 2012. Although the underlying methodologies differ, the trends in consumption poverty and SPM
income poverty have been remarkably similar over time. For example,
the declines in poverty between 1972 and 2012 shown by each measure
are nearly identical.
It is difficult to identify the effects of particular government
programs on consumption poverty because it is harder to identify and
remove the consumption derived from particular government programs
than it is with income under the SPM. However, the fact that the two
measures are similar suggests that underreporting of benefits has
little impact at the margin of determining whether someone is poor.
Additionally, the similarities in the measure suggest that spending
through savings, or borrowing from friends and family, is rarely able to
keep someone out of poverty.
Percent
30

25

Trends in Consumption and SPM Poverty, 1961–2012

Supplemental
Poverty
Measure

20

Consumption
Poverty

15

10
1961

1971

1981

1991

2001

Source: Wimer et al.(2013); updated data from Meyer and Sullivan (2012).

226  |  Chapter 6

2011

of necessary items, including food, shelter, clothing, and utilities.7 The dollar
amount is calculated separately for families depending on whether they own
or rent their home and whether they have a mortgage, and then increased
by 20 percent to allow for other necessary expenses. Further adjustments are
made based on differences in family size and structure, and, unlike in the
official measure, the threshold is adjusted for geographic variation in living
costs (Short 2013).
The Supplemental Poverty Measure also uses a more accurate measure of disposable income that accounts for both a greater number of income
sources and a wider array of necessary expenditures. Unlike the official
measure, the SPM uses a post-tax, post-transfer concept of resources that
adds to family earnings all cash transfers and the cash-equivalent of in-kind
transfers such as food assistance (for example, SNAP or free lunch) minus
net tax liabilities, which can be negative for families receiving refundable tax
credits like the EITC or CTC. Necessary expenditures on work and childcare are then subtracted from resources.
The Supplemental Poverty Measure also subtracts medical-out-ofpocket expenses from families’ resources since those funds are not available to meet other needs. The SPM can thus be thought of as a measure of
deprivation with respect to non-health care goods and services.8 However, it
does not provide an accurate picture of the benefits of health care. Instead,
the SPM values health insurance only insofar as it reduces households’ outof-pocket medical costs and thus frees up resources for other uses. It misses
benefits that may arise because insurance improves access to health care
and may therefore improve health outcomes, or reduces stress caused by
exposure to financial risk. As a result, the measured trend in SPM poverty
may understate progress in decreasing economic hardship since the War on
Poverty began by ignoring these benefits of increased access to insurance.
One important feature of the Supplemental Poverty Measure design
is that the definition of minimum needs is adjusted each year based on
recent data on family expenditures on necessities rather than adjusting a
fixed bundle only for inflation. By considering families’ expenditures on an
array of necessary items, including food, shelter, clothing, and utilities—and
7 More accurately, the thresholds are based on average expenditures on food, clothing, shelter,
and utilities between the 30th and 36th percentiles of that distribution, multiplied by 1.2 to
account for other necessary expenses and adjusted for geographic differences in cost of living
and family size and structure.
8 Korenman and Remler (2013) argue that the SPM’s treatment of medical-out-of-pocket
expenses actually does a poor job of capturing even deprivation of non-health care goods and
services. They argue that households that are able to spend a large amount on health care
are frequently those with substantial savings or other resources to draw on, and they present
evidence that households with high out-of-pocket expenses frequently score lower on “direct”
measures of hardship, like food insecurity.

The War On Poverty 50 Years Later: A Progress Report  |  227

then setting poverty rates based on how much families at the 33rd percentile
spend—the SPM adjusts poverty thresholds as societies’ spending patterns
on these necessities shifts.9 This type of threshold is “quasi-relative,” in
the sense that thresholds will tend to rise with income, but are not directly
tied to income as in a purely relative definition of poverty, such as setting a
threshold at half the median household income.
Alternatively, it is possible to create an “anchored” version of the
Supplemental Poverty Measure, which is the focus in this chapter. The
anchored version, like the official measure, fixes poverty thresholds based
on expenditures on necessary items in a given year and then adjusts only for
inflation in each year. This version allows for the use of the more comprehensive definition of resources in measuring poverty over time, while setting
a fixed assessment of what constitutes basic needs spending on food, shelter,
clothing, and utilities. The anchored measure is also more consistent with
the vision of the War on Poverty architects, who believed that poverty can
be eradicated. Eliminating poverty defined with a relative measure may be
nearly impossible, as the threshold rises apace with incomes.10

Who is Poor?
In an attempt to “provide some understanding of the enemy” for
Johnson’s War, the 1964 Economic Report of the President presented tables
depicting the “topography of poverty.” As Table 1 shows, some of the landmarks have changed since 1960 while many remain the same. For 2012, the
Table presents the poverty rates measured with both the official poverty
measure and the Supplemental Poverty Measure.11 The official measure is
displayed for comparing the relative poverty of various groups in the two
time periods, but should not be used for comparing changes in the levels of
poverty between the two time periods due to the flaws in the official measure
discussed above. The next section will provide trend data using a consistent
9 So, for example, if families across the income spectrum spend more on, say, housing because
preferences for or the ability to pay for space or bathrooms change then what is considered
necessary for minimum housing will change.
10 The SPM is a hybrid, “quasi-relative” measure such that when spending on necessities
increases, the threshold defining who is poor also increases. It is unlikely to rise at the same
rate as income, as with relative poverty measures, but will adjust more slowly since spending
on necessities grows more slowly than income. Eliminating poverty under this quasi-relative
definition is possible, depending on how a country’s spending on necessities evolves as its
income increases.
11 This Table uses the official Supplemental Poverty Measure statistics published by the Census
Bureau (Short 2013), whereas for historical comparisons below we rely on historical estimates
produced by Wimer et al. (2013), described below. The series from Wimer et al. anchors its
poverty measure at the 2012 SPM thresholds, so their estimated poverty rates for 2012 are very
similar to those in Table 1.

228  |  Chapter 6

Table 6-1

Poverty Rates by Selected Characteristics, 1959 and 2012
1959

2012

Official
Poverty Measure

Official
Poverty Measure

Supplemental
Poverty Measure

24.3

15.1

16.0

Head worked last year

17.8

10.0

10.5

Head did not work last year

55.7

27.4

29.2

Head married

18.9

7.9

10.2

Head single female

47.4

29.1

28.9

Less than high school (age 25-64)

25.3

33.9

35.8

High school (age 25-64)

10.2

15.6

17.5

College (adults 25-64)

6.7

4.5

5.9

Younger than 18

26.8

22.3

18.0

65 years and older

39.9

9.1

14.8

Female

24.9

16.4

16.7

African American

57.8

27.3

25.8

Hispanic

40.5

25.8

27.8

Asian

N/A

11.8

16.7

Native Alaskans/American Indians

N/A

34.2

30.3

White

19.5

9.8

10.7

Immigrant

23.0

19.3

25.4

Disabled (age 18-64)

N/A

28.4

26.5

Lives outside a metropolitan area

32.7

17.9

13.9

All People
Household Characteristics

Individual Characteristics

Note: Calculations based on characteristics of household heads exclude people living in group quarters.
Source: Census Bureau; CEA calculations.

measure. Since historical estimates of the SPM are available only starting
in 1967, Table 1 shows only official poverty rates for 1959 using the 1960
Census.

Employment
Unsurprisingly, unemployment is one of the strongest predictors of
poverty. In 1959, 55.7 percent of individuals in households where the head
was out of work for a full year were poor—three times the rate of individuals in households where the head worked at least one week during the year.
While this rate has declined to 29.2 percent, individuals in households where

The War On Poverty 50 Years Later: A Progress Report  |  229

the head was out of work for a full-year were still three times as likely to be
poor as those in households where the head worked.
However, even full-time employment is not enough to keep all
families out of poverty today. A person working full-time, full-year in 2013
being paid the minimum wage earns $14,500 for the year. These earnings
alone leave such workers below the poverty threshold if they have even one
child. While the EITC, SNAP, and other benefits will help pull a family of
two above the poverty line, for a larger family—such as one with three children—full-time, full-year minimum wage work combined with government
assistance is unlikely to be enough to lift that family out of poverty.

Education Level
Education’s role in poverty prevention has become more important
over time: in 1959, high-school dropouts were 3.8 times more likely to be
poor than college graduates; but in 2012, they were 6.1 times more likely
to be poor (based on the SPM measure). The growth in the poverty gap by
education is driven by growth in earnings inequality, which has led to much
greater earnings for college graduates than for those with less education.

Children
As in 1959, the child poverty rate today is higher than the poverty
rate of the overall population, although the SPM shows that children are
disproportionately helped by our poverty-fighting programs. Once taxes
and in-kind transfers are taken into account, the gap between child and nonchild poverty falls. According to the official measure, the child poverty rate
in 2012 was 22.3 percent—nearly 48 percent higher than the overall rate. But
the official rate ignores the contributions of the most important antipoverty
programs for children: the EITC and other refundable tax credits and SNAP.
Including the value of these resources, the SPM estimates that 18 percent of
children are poor—a rate that is 12.5 percent (2 percentage points) higher
than the overall poverty rate.

The Elderly
One of the most heralded successes of the War on Poverty is the
large reduction in elderly poverty rates. In 1959, poverty rates were highest among the elderly with 39.9 percent of people 65 and older living in
poverty (based on OPM). Today, poverty rates of those 65 and older are
below the national average. Using the SPM measure shows elderly poverty
at 14.8 percent, which is more than 50 percent higher than when taken with
the OPM measure. The reason for this difference is that the SPM subtracts

230  |  Chapter 6

Box 6-3: Women and Poverty
While women continue to have a higher poverty rate than men, the
gap has decreased over time as poverty has fallen more for women than
for men. The following chart shows that the gap between working-age
women and men has decreased from 4.7 percent to 1.7 percent from
1967 to 2012.
The decline in the poverty rate among women has been tempered
by an increase in the number of single mothers, who have higher rates
of poverty. The share of working-age women who are single mothers
rose from 11.6 to 16.1 percent between 1967 and 2012. Had the poverty
rates of demographic groups (based on marriage and motherhood
only) remained as they were in 1967, this change would have increased
poverty rates for working-age women by 2.1 percentage points. In fact,
poverty rates of women in all marriage and motherhood groups fell due
to increased work, rising education, and smaller families (Cancian and
Reed 2009), as well as to the increased impact of the safety net described
in this chapter.
The effect of government transfer and social insurance programs
on poverty is slightly larger for women than for men. These programs
reduced the 2012 poverty rate by 8.1 percentage points for working-age
women compared to 6.4 percentage points for working-age men. This
gender difference has been fairly stable over time, indicating that growth
in these programs does not explain the narrowing poverty gap shown
above. Rather, the closing of the gender poverty gap appears to be due to
increases in women’s education and employment rates relative to men.
SPM Poverty Rates for Working Age Adults by Gender, 1967–2012

Percent
25

20

2012
Women

15

10

Men
Difference (Women-Men)

5

0
1967

1977

Source: Wimer et al (2013).

1987

1997

2007

The War On Poverty 50 Years Later: A Progress Report  |  231

expenditures on medical expenses from a family’s resources, and the elderly
tend to have much higher medical expenses. Indeed, if the out-of-pocket
medical costs were not subtracted, the measured elderly poverty rate would
be only 8.4 percent—lower than their official poverty rate of 9.1 percent.12
In the absence of Medicare and Medicaid, out-of-pocket medical expenses
of the elderly would almost certainly cause their poverty to be much higher
than the SPM poverty rate of 14.8 percent.

Women
Women are more likely than men to be in poverty, with a 2012 poverty rate of 16.7 percent compared to 15.3 percent among men. This gap
largely reflects higher poverty among single women, both those age 18-64
(22.9 percent compared to 20.2 percent among single men) and those age 65
or older (21.2 percent compared to 16.1 percent among single men).
Childcare responsibilities help explain the poverty gap for single
working-age women. Almost one-third (31.0 percent) of single women age
18-64 lived with their children in 2012, among whom the poverty rate was
27.5 percent. Just over one-third (35.2 percent) of single mothers age 18-64
were employed full-time, full-year in 2012, compared to 44.6 percent of
single working-age males and 43.3 percent of single working-age women
without children at home. Increased support for childcare for young children has been shown to positively affect mothers’ employment hours and
earnings (Connelly and Kimmel 2003; Misra, Budig, and Boeckmann 2011).
The high poverty rate among older single women relative to men reflects a
combination of lower Social Security benefits due to lower lifetime earnings;
lower rates of pension coverage; and greater longevity that increases the
chances of outliving their private savings (Anzick and Weaver 2001, SSA
2012).

Race and Ethnicity
Poverty rates have fallen for all racial and ethnic groups over time
and gaps by race have shrunk slightly. However, troubling gaps still remain.
In 1959, nearly three-fifths of African Americans were in poverty, which
was nearly three times the poverty rate of Whites. The fraction of African
Americans in poverty has fallen by more than half since then; yet at 25.8
percent, the SPM poverty rate for African Americans is still more than
double the rate of 10.7 percent for Whites. Today, the SPM poverty rate
among Hispanics is 27.8 percent, similar to that among African Americans.
12 Accounting for out-of-pocket medical expenses raises measured poverty for nonelderly adults
and children by 2.9 and 3.1 percentage points, respectively.

232  |  Chapter 6

However, this reflects smaller declines in poverty among Hispanics over the
past 50 years. Among both African Americans and Hispanics, the official
and supplemental poverty rates tell similar stories. However, the SPM reveals
that Asian Americans have a slightly higher poverty rate than the national
average—at 16.7 percent—while their OPM poverty rate is lower than the
national average. The higher SPM rate for Asian Americans reflects, in part,
the fact that they tend to live in high-cost metropolitan areas (for example,
Los Angeles and New York City) and the SPM poverty thresholds are higher
in such places due to its geographic adjustments for cost of living. Finally,
while measures of poverty among American Indians and Alaska Natives are
not available in the earlier period, currently they have the highest rates of
poverty of any race and ethnicity group at 30.3 percent in 2012.

People with Disabilities
Over one-fourth of working-age people with disabilities are estimated
to live in poverty, using both the OPM (28.4 percent) and SPM (26.5 percent) measures. This largely reflects their low employment rates. The effective poverty rate of people with disabilities may be understated due to the
extra costs that often accompany disability, such as for home and vehicle
renovations, assistive equipment, personal assistance, and other items and
services that may not be covered by insurance or government programs
(Sen 2009: 258, She and Livermore 2007, Fremstad 2009, Schur, Kruse, and
Blanck 2013: 32-33).

Rural and Urban Communities
The official poverty measure overestimates rural poverty since rural
communities tend to have lower costs of living than urban areas and the
official measure does not take geographic cost-of-living differences into
account. However, the OPM has revealed that significant poverty persists in
rural communities throughout the country today. The Economic Research
Service of the U.S. Department of Agriculture estimates that 85 percent of
persistent poverty counties—counties that have been in high poverty (over
20 percent based on the OPM) for at least 30 years—are in rural areas.13 The
gap between poverty rates outside and within metropolitan areas, though,
has narrowed since 1959 when poverty rates outside metropolitan areas
were more than double the rates within these areas. In fact, the adjustments

13 See http://www.ers.usda.gov/topics/rural-economy-population/rural-poverty-well-being/
geography-of-poverty.aspx#.UurSXhBdXA0

The War On Poverty 50 Years Later: A Progress Report  |  233

for different housing costs across geographic areas in the SPM show that
poverty rates are higher in metropolitan areas than in rural areas today.14

Assessing the War on Poverty
This section presents new historical estimates of the poverty rate from
1967 to 2012 based on the Supplemental Poverty Measure and shows that
substantial progress has been made in reducing poverty since President
Johnson began major policy initiatives as part of his fight against poverty. In
the past 45 years, the poverty rate fell from 25.8 to 16.0 percent—a reduction of over one-third. CEA documents that much of this decline was due
to the increased poverty-reducing effects of the safety net expansion set in
motion during the Johnson Administration. Based on a measure of pre-tax,
pre-transfer income, the poverty rate would be about as high today as in
1967: over 28 percent. These analyses show that safety net programs lifted 45
million people from poverty in 2012 and, between 1968 and 2012, prevented
1.2 billion “person years” from living below the poverty line. This section
first reviews changes in the economy that provide context for understanding
the lack of growth in market incomes in the bottom part of the distribution
over this time period. The section then presents estimates of poverty trends
since 1967, measured using modern methods.

Context
The architects of the War on Poverty were confident they would live
to see poverty eradicated. Looking at the data at their disposal at the time, it
is easy to see how an analyst might have believed the end of poverty was on
the horizon. Figure 6-1 shows the trend in the official poverty measure—the
only consistently available measure tracking poverty until recently—from
1959 through 2012. Based on the trend in poverty observed between 1959
and 1968, one would have indeed forecast—extrapolating linearly—that
poverty would be eradicated by 1980. Poverty fell by a remarkably consistent
rate of about 1.15 percentage points a year over that 10-year period, but the
official poverty measure stopped declining afterwards, reaching its lowest
point in 1973.
As previously noted, the OPM is a measure of cash income that does
not include non-cash benefits or tax credits. While it does not accurately
capture the trend in poverty over time, it is nonetheless worth considering
why the improvement in this measure of cash income slowed so abruptly in
14 This comparison may be affected by significant differences among geographic areas in costs
other than housing, such as transportation. In addition, there may be differences in housing
quality that are not captured by the differences in housing costs.

234  |  Chapter 6

Percent
25

Figure 6-1
Trends in the Official Poverty Measure, 1959–2012

20
2012
15
10

Official Poverty Measure

5
0
1957

1967

1977

1987

1997

2007

Source: Bureau of Labor Statistics, Current Population Survey, Annual Social and Economic Supplement;
CEA calculations.

the early 1970s. The first, and clearest, answer is that Social Security expansions during the 1960s brought the rate of poverty among the elderly down
rapidly before leveling in the 1970s (Engelhardt and Gruber 2006). In 1959,
39.9 percent of those 65 and older were in poverty, but by 1974 that fraction
had fallen to 14.6 percent (based on the OPM). Over the next 38 years, the
elderly poverty rate fell further to 9.1 percent in 2012. The deceleration in
poverty reduction is less pronounced for nonelderly adults and children. In
fact, using the SPM measure that accounts for expansions of the EITC and
non-cash transfers, children’s poverty had greater declines in the 1990s than
in the 1960s or 1970s.
Growth in inequality has also helped put the brakes on improvements
in cash income for most households. Economic growth is an important
determinant of poverty (Blank 2000) as long as the gains are shared with
those in the bottom of the income distribution. When growth fails to benefit
the bottom, it cannot play a role in eradicating poverty. As such, the distribution of income can have a profound impact on the level of poverty. While
the real economy grew at an annual rate of about 2.1 percent during the
1970s and 1980s, since 1980 economic growth has not produced the “rising
tide” heralded by President Kennedy, as rising inequality left incomes at
the bottom relatively unchanged (DiNardo, Fortin, Lemieux 1996, Piketty
and Saez 2003, Lemieux 2008). As shown in Figure 6-2, incomes in the top

The War On Poverty 50 Years Later: A Progress Report  |  235

Figure 6-2
Average Real Household Income by Quintile, 1967–2012
Index (1973=100)
160

2012

Fifth Quintile
140
Fourth Quintile
120

Third Quintile
Second Quintile

100

80
1967

First Quintile

1977

1987

1997

2007

Source: Census Bureau.

20 percent of the income distribution rose dramatically until the 2000s and
are about 50 percent higher today than in 1973. By contrast, real household
incomes in the bottom 60 percent of the income distribution stagnated until
the mid-1990s expansion, and today are little changed from the business
cycle peak in 1973.
A large group of poverty scholars have pointed to this rise in inequality as a leading explanation for the lack of progress in reducing poverty
since 1980 (for example, Blank 1993; Gottschalk and Danziger 1995, 2003;
Hoynes, Page, and Stevens 2006).
The failure of the minimum wage to keep up with inflation is an
important reason why inequality increased in the 1980s (DiNardo, Fortin,
and Lemieux 1996, Lee 1999), and progress in the fight against poverty
has slowed. President Johnson extended both the level and scope of the
minimum wage, with its peak reached in real terms in 1968. Since then, the
minimum wage has risen and fallen, but today its level of $7.25 an hour is
the same in real terms as in 1950. At this level, even factoring in the subsidy
provided by the EITC, a single parent of two kids working full time would
still have income near the poverty line.
Several studies have documented a tight link between the value of the
minimum wage and measures of wage inequality in the bottom part of the
income distribution (Lee 1999, DiNardo, Fortin, and Lemieux 1996). For
236  |  Chapter 6

Box 6-4: Social Programs Serve All Americans
While the safety net provides crucial support for families in poverty, far more Americans benefit from the safety net than are poor in
any given year. Of course, all Americans benefit from Social Security and
Medicare support for the aged, shielding both them and their families
from low income in retirement and the costs of adverse health shocks.
And many people benefit from social insurance programs that are not
means-tested. For example, nearly half of all Americans will benefit from
unemployment insurance at some point over a 20-year period.
But even programs targeting primarily low-income families serve
a very high fraction of Americans at some point in their lives. A recent
study using administrative tax records from 1989 to 2006 found that over
50 percent of tax-filers with children benefited from the EITC at some
point over the 19-year period (Dowd and Horowitz 2011). Moreover,
CEA analysis of the National Longitudinal Study of Youth 1979 finds
that of all individuals aged 14 to 22 in 1979, over the 32-year period from
1978 to 2010:
•	 29.6 percent benefitted from SNAP;
•	 34.2 percent received support from SNAP, AFDC/TANF, or SSI;
and
•	 69.2 percent received income from SNAP, AFDC/TANF, SSI, or
UI.
Looking at a broader array of programs, a large fraction of the population benefits in any given year as well. According to the 2013 Annual
Demographic and Economic Supplement of the Current Population
Survey, nearly half (47.5 percent) of all households received support
from either refundable tax credits, SNAP, Unemployment Insurance,
SSI, housing assistance, school lunch, TANF, Women, Infants, and
Children (WIC), Medicaid or Disability Insurance.
An important feature of the safety net that is often overlooked: for
most programs the majority of beneficiaries receive assistance for only
a short period when their earnings drop for some reason, and then they
bounce out again. Research has shown, for example, that 61 percent of
all EITC recipients claimed the credit for two years or less (Dowd and
Horowitz 2011); and, half of all new SNAP participants in the mid-2000s
left the program within 10 months.1

1 See: http://www.fns.usda.gov/sites/default/files/BuildingHealthyAmerica.pdf

The War On Poverty 50 Years Later: A Progress Report  |  237

Figure 6-3
Women's 50-10 Wage Gap vs Real Minimum Wage, 1973–2012

Index (1973=100)
140

120

Index (1973=100), inverted
60

80

Minimum Wage
(right)
2012

100

80

120

50-10 Gap
(left)

60
1967

1977

100

1987

Source: CEA update of analysis in Lemieux (2008).

1997

2007

140

example, as shown in Figure 6-3, changes in the ratio of the median wage
to the 10th percentile of the wage distribution (the “50-10 wage gap”)—a
measure of inequality in the bottom of the wage distribution—for women
correlate very closely with changes in the real value of the minimum wage.15
The best research suggests that increases in the minimum wage do not result
in job losses large enough to undermine the goal of raising incomes for
the poor (Dube, Lester, Reich 2010). Moreover, a recent meta-analysis by
Doucouliagos and Stanley (2009) covering over 1,000 estimates of minimum
wage effects finds “no evidence of a meaningful adverse employment effect.”
Finally, a recent analysis and review of the literature by Dube (2013) finds
consistent evidence across studies that a 10 percent increase in the minimum
wage decreases the poverty rate by about 2.4 percent.
Another important factor in rising inequality and slow wage growth
among low- and middle-income workers has been the decline in unionization. The percent of U.S. workers represented by unions has nearly halved
from 23.3 percent in 1983 to 12.4 percent in 2013.16 This decline has contributed to inequality because unions reduce inequality by raising the wages
15 The figure updates an analysis in Lemieux (2008), who graciously shared data.
16 Retrieved from http://www.bls.gov/cps/cpslutabs.htm, January 30, 2014.

238  |  Chapter 6

of low- and middle-income workers and compressing the returns to skill
(DiNardo, Fortin, and Lemieux 1996; DiNardo and Lemieux 1997).
Many observers have implicated various demographic changes—most
prominently, increased immigration and a decline in two-parent families—
as additional factors behind the lack of progress in market incomes in the
lower part of the distribution. The recent literature on immigration rejects
the claim that competition from immigrants has had a meaningfully adverse
effect on the wages or poverty rates of native workers (Peri 2013). Because
the country-of-origin composition of immigrants has increasingly shifted
toward poorer countries, however, immigration has had a mechanical effect
on poverty rates. Card and Raphael (2013) estimate that changes in the
population shares and the country of origin of the foreign born increased
overall poverty rates (based on OPM) by 3.7 percentage points between
1970 and 2009. While some immigrant households that may have seen their
incomes rise as a result of coming to the United States, many still fall below
the poverty line here. But other analysts focusing on different time periods
generally find much smaller compositional effects. For example, Hoynes,
Page, and Stevens (2006) find that increased immigration accounts for only
a 0.1 percentage point increase in poverty between 1979 and 1999.
Another dramatic change since the 1960s is a large increase in the
number of people living in single-female headed households. As shown in
Table 1, individuals in such households typically have double the poverty
rates of the national average, so this change also tends to increase the poverty
rate. Using decomposition techniques, Hoynes, Page, and Stevens (2006)
show that changes in family structure alone accounted for a 3.7 percentage
point increase in the (OPM) poverty rate between 1967 and 2003.17 In fact,
women’s poverty rates declined over this time period due to their educational attainment, labor force participation, higher earnings, and they had
fewer children (Reed and Cancian 2001). Moreover, changes in family structure can both cause and be caused by changes in economic circumstances.
The last three decades have also seen a historic rise in incarceration,
which has led to greater poverty. The fraction of the population in prison
rose from 221 per 100,000 in 1980 to 762 per 100,000 in 2008 (Western and
Pettit 2010).18 In the short term, imprisonment removes wage earners from
the family, which reduces their family’s income and increases the probability
of their children growing up in poverty. For example, Johnson (2008) finds
17 Like all decompositions, the reference period matters. If the decomposition is performed
using 2009 poverty rates rather than those in 1970, the predicted increase is 2.8 percentage
points.
18 Research suggests the increase is driven primarily by increased sentencing severity rather
than increases in criminal activity (Caplow and Simon 1999, Nicholson-Crotty and Meier
2003).

The War On Poverty 50 Years Later: A Progress Report  |  239

that child poverty increases by 8.5 percentage points and family income falls
by an average of $8,700 while a father is in prison.
There are also long-term negative impacts on earnings from incarceration that lead to higher rates of poverty among those with a criminal record.
Offenders’ wages are lower by between 3 and 16 percent after incarceration
(Raphael 2007; Western 2002) and employment and labor force participation are also negatively affected. Research shows that each additional percentage point of imprisonment of African American men is associated with
a reduction in employment or labor force participation of young African
American men of approximately 1.0 to 1.5 percentage points. This relationship implies that the increases in incarceration over the last three decades
have reduced employment and labor force participation among young
African American males by 3 to 5 percentage points (Holzer 2007). Holzer
also notes that, while the magnitude of the effect of incarceration on White
and Latino offenders is less clear, most studies find that their experience with
employment and labor force participation after incarceration is similar to
that of African American men.
Together, the factors described above created headwinds in the fight
on poverty. Evaluating the precise impact of these factors is beyond the
scope of this report, but it is likely that their combined influence was to exert
modest upward pressure on poverty rates. As shown below, the fact that
“market poverty” stayed relatively constant over this time period suggests
that improvements in education or other factors may have offset the adverse
effects of these demographic and other changes. Previous studies based on
the OPM support this notion. For example, Mishel et al. (2013) suggest
that the impact of increases in education in reducing poverty were slightly
greater than the adverse impact of changing demographics.

Correcting the Historical Account of Poverty Since the 1960s
The official poverty measure introduced by President Johnson’s
administration ignores, by design, the most important antipoverty programs
introduced during and after the War on Poverty. In particular, resources
from nutrition assistance, tax credits for working families, and access to
health insurance are not considered when computing whether a family is
poor by the traditional metric.
The Census Bureau has published the Supplemental Poverty Measure
only as far back as 2009. But recent research by poverty scholars on alternate
measures of poverty all find that the official poverty rate displayed in Figure
6-1 dramatically understates the decline in poverty since the 1960s (Fox
et al. 2013, U.S. Census Bureau 2013, Meyer and Sullivan 2013, Sherman
2013). Work by Wimer, Fox, Garfinkel, Kaushal, and Waldfogel (2013) is
240  |  Chapter 6

particularly valuable since they estimate poverty rates from 1967 to 2012
following the SPM methodology for computing family resources. They also
measure poverty using an “anchored” measure that uses a fixed poverty
threshold based on expenditures on necessary items in 2012, adjusted only
for inflation in each year (with inflation measured using a historically consistent series, the CPI-U-RS).19
Figure 6-4 shows a striking fact: poverty has declined by 38 percent
since 1967, according to the anchored SPM measure. And, unlike the OPM,
it continued to fall after the early 1970s. The figure shows the evolution of
the poverty rate using the anchored SPM measure of poverty from 1967
to 2012, compared with the official poverty rate reproduced from Figure
6-1. Using the more accurate SPM measure of family resources changes
the historical account of poverty in the United States significantly: between
1967 and 2012, poverty rates fell by 9.8 percentage points—from 25.8 to 16.0
percent. The trend in the SPM depicted in Figure 6-4 is very similar to that of
an alternative measure of poverty based on consumption data, which Meyer
and Sullivan (2013) argue is a better measure of material hardship (Box 6-2).
Figure 6-4 shows that the fraction of Americans in poverty fell
smoothly between 1967 and 1979 to a low of 17.4 percent, a period where the
discrepancy with the trend shown by the official measure is driven primarily
by more accurate accounting for inflation in the 1970s. After rising steeply
during the double-dip recession of the early 1980s, poverty rates fell slightly
until rising again with the early 1990s recession.
In contrast to the depiction of the OPM, the steepest declines in the
fraction of people in poverty occurred during the economic expansion of
the 1990s. During that period, poverty fell from 20.7 percent in 1993 to
14.6 percent in 2000, the lowest poverty rate observed since 1967. As shown
in Figure 6-2, economic growth in the 1990s provided a strong boost to
low-income households as earnings grew even in the bottom one-fifth of
incomes, in contrast to the experience of any other decade since the 1960s.
Dramatic increases in the value of the EITC leveraged this upswing in the
19 The “anchored poverty” measure allows progress to be measured against a constant
definition of living standards. Using the SPM methodology of updating poverty thresholds
each year to reflect rising expenditures on food, clothing, shelter, and utilities shows less of a
decline in poverty since the real value of the poverty thresholds rise over time (Fox et al. 2013).
Data constraints prevent Wimer et al. (2013) from following the SPM methodology exactly.
The most important discrepancy from the Census procedure is that Wimer et al. do not adjust
the poverty thresholds for geographic differences in living costs. It is worth noting that an
alternative measure anchors poverty thresholds based on necessary expenditures in 1967, and
adjusts for inflation each year afterwards. Both measures show similar declines in poverty, but
using expenditures in 2012 gives a higher level of poverty in every period since increased real
spending on necessities over time has led to a higher SPM poverty threshold.

The War On Poverty 50 Years Later: A Progress Report  |  241

labor market to further encourage work and channel even more resources to
low-income working families.
One last, and remarkable, fact shown in Figure 6-4 is that the poverty
rate ticked up only slightly during the Great Recession after remaining
steady for most of the 2000s. Despite the largest rise in unemployment since
the Great Depression, the poverty rate rose by only 0.5 percentage points
overall between 2007 and 2010. As discussed below, this shows how effective
the safety net and its expansion through the 2009 American Recovery and
Reinvestment Act (the Recovery Act) have been. Since much of the credit for
this is due to expansions in SNAP and tax credits, the official poverty rate
fails to capture this crucial success.
The poverty rates for children and for working-age adults follow a
similar pattern to the overall trend shown in Figure 6-4. The fraction of children living in poverty declined from 29.4 percent in 1967 to 18.7 percent in
2012; for working-age adults, the poverty rate fell from 19.8 to 15.1 percent
over the same period. For the elderly, the trend in poverty is one of near
continuous decline. Poverty fell quite rapidly up until the early 1980s, driven
by large growth in per capita Social Security payments, from 46.5 percent in
1967 to 20.7 percent in 1984. Elderly poverty fell further during the 1990s
expansion to a low of 16.5 percent in 1999, and then ticked up slightly in
the 2000s. Driven in part by the Recovery Act stimulus payments, elderly
poverty declined from 17.6 percent to 15.2 percent between 2007 and 2009,
and it remained at that level in 2012.

Measuring the Direct Impact of Antipoverty Efforts
The fact that poverty rates have fallen overall, even as household
incomes in the bottom of the distribution have stagnated since the 1980s,
suggests a substantial direct role that policies have played in improving the
well-being of the poor. Wimer et al. (2013) estimate the magnitude of this
impact by constructing “counterfactual” poverty measures that simulate the
fraction of the population that would have been poor in the absence of all
government transfers, including the overall tax system.20 In other words,
they estimate the fraction of families that would have incomes below the
poverty line if the value of all cash, in-kind, and tax transfers they received
(or paid if the family owed taxes on net) were not counted. Comparing the
difference between this measure of “market poverty” and the SPM poverty
20 While this is a “static” exercise in that it assumes that individual earnings themselves are not
affected by the existence of safety net programs, this chapter later reviews research on the effects
of programs on employment and earnings and finds that such effects are generally small where
they exist, and not large enough to meaningfully alter the conclusions from this simplification
(Ben-Shalom, Moffitt, and Scholz 2010).

242  |  Chapter 6

Figure 6-4
Official vs Anchored Supplemental Poverty Rates, 1967–2012

Percent
30
25

Supplemental Poverty Measure (Anchored 2012)

20

2012

15
10

Official Poverty Measure

5
0
1967

1977

Source: Census Bureau; Wimer et al (2013).

1987

1997

2007

rate provides a measure of the reduction in poverty accounted for by government transfers. Figure 6-5 shows the results of this analysis. The height of
the overall shaded region indicates the poverty rate counting only market
income, while the height of the region shaded in black is the SPM poverty
rate shown in Figure 6-4. The difference, shaded in green, represents the
percentage of the population lifted from poverty by the safety net and the
net effect of the tax system.21
In part because of the rising inequality in earnings described above,
market poverty increased over the past 45 years by 1.7 percentage points
from 27.0 percent in 1967 to 28.7 percent in 2012. In contrast, poverty rates
that are measured including taxes and transfers—these taxes and transfers are the green-shaded region in Figure 6-5—fell through most of this
period. Government transfers reduced poverty by 1.2 percentage points in
1967. This impact grew to about 7.4 percentage points by 1975 due to the
21 There are two counterfactuals estimated by Wimer et al. that are used in this report to
discuss the impact of the safety net. For most estimates of the impact of the safety net we use
a counterfactual poverty rate that strips away (“zeroes out”) all cash and in-kind transfers, as
well as refundable tax credits but continues to subtract any “normal” tax liability from family
resources. In this section, we define “market poverty” similarly, only this measure additionally
zeroes out all tax liabilities. Since poor families near the poverty line tend to have positive tax
liabilities, market poverty rates are slightly—about 1.8 percentage points in 2012—lower (since
we assume families can keep the taxes they in fact must pay).

The War On Poverty 50 Years Later: A Progress Report  |  243

Percent
30

Figure 6-5
Trends in Market and Post-Tax, Post-Transfer
Poverty, 1967–2012

25
Market SPM
(Pre-Tax and Pre-Transfer)

20
15
10

Supplemental Poverty Measure
(Post-Tax and Post-Transfer)

5
0
1967

1977

Source: Wimer et al (2013).

1987

1997

2007

expansion of the safety net spurred by the War on Poverty, and hovered
around that level until the Great Recession when it increased to 12.7 percentage points in 2012.
Despite an increase in “market poverty” of 4.5 percentage points
between 2007 and 2010, the actual SPM poverty rate rose only 0.5 percentage
points due to the safety net. In fact, the 2009 Recovery Act reduced poverty
by 1.7 percentage points in 2010 through its extensions to the safety net as
discussed below. Overall, for the entire 45-year period, the poverty decline
of 9.8 percentage points is almost entirely accounted for by the increased
effectiveness of the safety net. 22
Figure 6-6 shows a similar analysis of the effect of the safety net on
trends in deep poverty and highlights two important features of the safety
net often overlooked. First, the safety net improves the well-being of many
more individuals than is reflected in the standard accounting of how many
individuals are lifted from poverty: in 2012, about one in twenty (5.3 percent) Americans lived in deep poverty, yet without government transfers the
number would be closer to one in five (18.8 percent).
Second, the safety net almost entirely eliminates cyclical swings in the
prevalence of deep poverty. Figure 6-6 shows that despite large increases in
22 As described earlier, the underreporting of government income and benefits means that this
is likely to be a conservative lower-bound estimate of the effect of the safety net.

244  |  Chapter 6

deep market poverty driven by the business cycle, there is little if any rise in
actual deep poverty due to the supports provided by the safety net.
Again, this phenomenon is especially visible during the Great
Recession, when the prevalence of deep poverty ticked upward by only 0.2
percentage points despite an increase in market deep poverty of 3.3 percentage points. This corresponds to more than 9.5 million men, women, and
children prevented from living below half the poverty line during the Great
Recession. Over the 45 years shown in Figure 6-6, deep “market poverty”
actually rises from 14.9 to 18.8 percent. Despite that increase, the fraction of
those in deep poverty fell from 8.2 to 5.3 percent.

The Role of Antipoverty Programs: A Closer Look
This section presents further detail on antipoverty effects of specific
components of the safety net. In particular, it highlights the impact that
different groups of programs—cash transfers, in-kind transfers, and tax
credits—have on the poverty rates of different age groups. It also shows how
the relative importance of programs for nonelderly adults and children has
changed since the start of the War on Poverty.
The section then refutes the concern critics have raised that the existence of safety net programs may undermine growth in market incomes as
well as our efforts to fight poverty. The social safety net has increasingly
been designed to reward and facilitate work increasing participation rates—
in many cases, requiring work. Even where programs are not explicitly
designed to require work, the highest-quality studies suggest that adverse
earnings effects of safety net programs are nonexistent or very small, in part
due to reforms over the past two decades that, for example, have phased out
benefits gradually with increases in earnings to minimize disincentives to
work.
Finally, this section presents findings on the level of economic mobility of individuals born into poverty in the United States, and the results of
recent research showing the potentially large returns to social spending in
terms of long-term outcomes of children in families receiving support.

Antipoverty Effects of Specific Programs
This section illustrates the role that various antipoverty programs
have had in improving the well-being of different populations, and how the
relative impacts of these programs have evolved since the War on Poverty
began. It is based on an effectively “static” analysis that zeroes out income
derived from various public programs to ask what effect this would have on

The War On Poverty 50 Years Later: A Progress Report  |  245

Figure 6-6
Trends in Market and Post-Tax, Post-Transfer
Deep Poverty, 1967–2012

Percent
30
25
20
15

Market SPM
(Pre-Tax and Pre-Transfer)

10
5
0
1967

Supplemental Poverty Measure
(Post-Tax and Post-Transfer)
1977

Source: Wimer et al (2013).

1987

1997

2007

poverty. The next subsection considers some of the broader impacts that
these programs have on employment and earnings.
Table 2 shows the impact of various safety net programs on overall
poverty, and for three separate age groups: children, adults age 19 to 64, and
elderly adults age 65 and over. The safety net program with the single greatest impact is Social Security, which provides income to the elderly, people
with disabilities, and surviving spouses and children, reducing the overall
poverty rate by 8.6 percentage points in 2012. The program’s impacts on
elderly poverty are profound: without Social Security income, the poverty
rate of the elderly would be 54.7 percent, rather than its rate of 14.8 percent
in 2012. On the other end of the age spectrum, refundable tax credits like the
EITC and the Child Tax Credit (CTC) have large impacts on child poverty—
reducing the fraction of children in poverty by 6.7 percentage points.
Tax credits also reduce the poverty rates of nonelderly adults by 2.3
percentage points. The Supplemental Nutrition Assistance Program also has
a dramatic effect on poverty, reducing child poverty by 3.0 percentage points
and overall poverty rates by 1.6 percentage points.
Finally, unemployment insurance (UI) reduced poverty by 0.8 percent
overall in 2012. This effect, as with the effects for other programs, was less
than at the height of the Great Recession when more people were without

246  |  Chapter 6

work: in 2010, for example, unemployment insurance reduced poverty by
1.5 percentage points overall (Short 2012).
As noted, all of these estimates ignore the incentives to alter work
behavior created by government programs and are not definitive causal
estimates of the impact of the different programs on poverty. For example,
Social Security may affect market incomes by changing retirement and savings incentives. Similarly, the estimates do not take into account the role
that UI plays in keeping people attached to the labor force, or that the EITC
plays in incentivizing additional hours of work and participation in the labor
force. The importance of these considerations is discussed below.
Even programs with a small impact on overall poverty rates may be
very effective in reducing poverty for certain populations or in alleviating
hardship without lifting individuals out of poverty. For example, SSI reduces
poverty rates by 1.1 percentage point overall, but this represents a large poverty reduction concentrated among a relatively small number of low-income
recipients who are elderly or have a disability. TANF and General Assistance
(state programs which typically provide limited aid to very poor individuals
not qualifying for other aid) have only a small impact on the overall poverty
rate at 0.2 percentage points, as TANF benefits are generally insufficient to
bring people above the poverty level. However, by raising the incomes of
those in poverty these programs have a much greater impact on reducing
deep poverty.
Based on their historical estimates of SPM poverty rates, Wimer et al.
(2013) conduct similar analyses to those in Table 2 for each year since 1967
for different groups of safety net programs. Their results shed light on how
the safety net has changed over the past 50 years.
In aggregate, the antipoverty effects of three types of federal aid programs—support through cash programs like Social Security, SSI, and TANF;
in-kind support like SNAP and housing assistance; and tax credits like the
EITC and CTC—all increase over time, driving down overall poverty rates.
The steady increase in the effect of each type of program masks some differences across the populations served by each program. For the elderly, for
example, the trend is dominated by the growing real value of Social Security
payments steadily driving down the elderly poverty rate.
For children, however, there is a shift in importance of the safety net’s
different components. Figure 6-7 shows the effect of eliminating various
components of the safety net resources on the SPM poverty rate in three
years corresponding to recession-driven peaks of the poverty rate. In the
early 1970s recession, AFDC, and food stamps (now known as Supplemental
Nutrition Assistance Program or SNAP) to a lesser extent, played the most
important role in alleviating childhood poverty; and the EITC had not yet
The War On Poverty 50 Years Later: A Progress Report  |  247

Table 6–2
Poverty Rate Reduction from Government Programs, 2012
All People

Children

Nonelderly
Adults

65 Years
and Older

Social Security

8.56

1.97

4.08

39.86

Refundable Tax Credits

3.02

6.66

2.25

0.20

SNAP

1.62

3.01

1.27

0.76

Unemployment Insurance

0.79

0.82

0.88

0.31

SSI

1.07

0.84

1.12

1.21

Housing subsidies

0.91

1.39

0.66

1.12

School lunch

0.38

0.91

0.25

0.03

TANF/General Assistance

0.21

0.46

0.14

0.05

WIC

0.13

0.29

0.09

0.00

311,116

74,046

193,514

43,245

Population (Thousands)

Note: Data are presented as percentage points.
Source: Census Bureau.

been introduced. Both in-kind transfers and the EITC had a small impact on
child poverty in the early 1990s recession, but cash transfer programs still
outweighed the poverty-reducing impact of both programs put together.
During the Great Recession, however, we see that in-kind aid, cashassistance, and tax credits all played similarly important roles in reducing
poverty for families with children. This shift reflects both the large structural
change away from cash welfare assistance during the 1990s, and expansion
of both SNAP and tax credits through the Recovery Act.
Figure 6-8 is similar to Figure 6-7, only showing the impact of various transfer programs on deep poverty, or the fraction of individuals with
incomes below 50 percent of the poverty threshold. This figure shows that
for deep poverty, in-kind transfers have become the most important safety
net program over time, and tax credits are less effective due to the paucity of
work among families with very low resources.

The Effects of Antipoverty Programs on Work and Earnings
In his remarks before signing the cornerstone legislation of the War
on Poverty, the Economic Opportunity Act, President Johnson declared:
“Our American answer to poverty is not to make the poor more secure in
their poverty but to reach down and to help them lift themselves out of the
ruts of poverty and move with the large majority along the high road of hope
and prosperity. The days of the dole in our country are numbered.” He went
on to describe the need to provide the poor with the means to lift themselves
248  |  Chapter 6

out of poverty through job training and employment services, among other
strategies. In the past 20 years, this emphasis on promoting work through
antipoverty programs intensified and we made dramatic changes to cash
welfare and other programs that shifted the focus toward requiring or
rewarding work.
The most important shift began with the introduction of the EITC in
1975. The EITC was expanded multiple times in the 1980s, 1990s, and 2000s
as the safety net shifted increasingly toward work-based support. The EITC
was expanded most recently in the 2009 Recovery Act, with those improvements extended in 2010 and 2013. Since 1996, the EITC has accounted for
more support for low-income households than traditional cash welfare. In
2012, the EITC and the partially refundable CTC totaled $90 billion annually, more than four times the expenditures on the Temporary Assistance for
Needy Families program.
Because the EITC is a supplement to labor market earnings, the
credit provides strong incentives for those with otherwise low labor force
attachment to increase their hours of work. However, because the credit is
reduced as earnings continue to rise, it may also provide those with earnings
above the phase-out threshold—above $22,870 for a married couple with
three children—an incentive to reduce their earnings. Several studies have
addressed these incentive effects, and reach similar conclusions: the EITC is
associated with increased labor force participation, especially among single
mothers, but it does not appear to substantially alter the hours or earnings
of those already working (Eissa and Liebman 1996, Liebman 1998, Meyer
and Rosenbaum 2001, Hotz and Scholz 2003, and Eissa and Hoynes 2005).23
Taken together, these results suggest EITC expansions played an important
role in increasing labor force participation among single mothers, without
adversely affecting hours worked by those already working.
Meanwhile, the 1996 welfare reform law replaced AFDC with TANF
and significantly strengthened work requirements in the cash assistance program. The Personal Responsibility and Work Opportunity Reconciliation
Act of 1996 ended the entitlement to cash assistance and beneficiaries were
generally required to work or participate in “work activities” to receive assistance. Moreover, the implicit tax rate on benefits in response to increased
earnings—the benefit reduction rate—was reduced dramatically in many
states. Matsudaira and Blank (2013) show that these changes increased the
return to work, with potential income gains of over $1,842 (in 2000 dollars)
for welfare recipients working 30 hours a week. At the same time, it should
23 For example, Meyer and Rosenbaum (2001) suggest that EITC expansion accounts for about
60 percent of the roughly 10 percentage point rise in single mothers’ employment rates relative
to single mothers without children between 1984 and 1996.

The War On Poverty 50 Years Later: A Progress Report  |  249

Figure 6-7
Percentage Point Impact on SPM Child
Poverty for Selected Years

Percentage Points
30
1971

1992

Base Rate

2012

25
20
15
10
5
0

In-Kind

Note: EITC began in 1975.
Source: Wimer et al (2013).

Cash-Transfers

EITC

SPM

Figure 6-8
Percentage Point Impact on Deep SPM Child
Poverty for Selected Years

Percentage Points
30

1971

1992

2012

25
20
15
Base Rate

10
5
0

In-Kind

Note: EITC began in 1975.
Source: Wimer et al (2013).

250  |  Chapter 6

Cash-Transfers

EITC

SPM

be noted that researchers have found these reforms may have increased
hardship among groups with high barriers to employment, highlighting
the continued importance of programs that serve the most disadvantaged
(Blank 2007; Danziger, Turner, and Seefeldt 2006).
While today’s safety net has been reformed to promote work, it is
important to note that careful research has shown that most assistance
programs have only small, if any, disincentive effects on work. This suggests
that the “static” estimates of the antipoverty effects of the programs shown
above largely capture the actual full impact of these programs including any
employment disincentives they may have—and may understate the povertyreducing impacts of programs that effectively reward and facilitate work
and thus increase market earnings. For example, examining labor supply
behavior of individuals in the Oregon Health Insurance Experiment, Baicker
et al. (2013) find that Medicaid recipients are not less likely to be employed
nor do they earn less than they otherwise would have. Similarly, Hoynes and
Schanzenbach (2012) study the initial rollout of SNAP (then food stamps)
and find only small effects on labor supply. While there are no studies of
the employment effects of the TANF program with its work requirements,
prior studies of AFDC and the Negative Income Tax experiments of the
1970’s suggest only modest disincentive effects of a program without an
emphasis on work. Burtless (1986) found that a dollar of benefits reduces
work earnings by 20 cents (total income is increased by 80 cents), but other
evidence suggests the disincentive effects may have been even smaller (SRI
International 1983).
Evaluating the weight of the evidence for all programs, Ben-Shalom,
Moffitt, and Sholz (2011) conclude that the work disincentive effects
of antipoverty programs have “basically, zero” effect on overall poverty
rates. Going program by program, they conclude the behavioral effects of
TANF are likely zero, and that the work disincentives induced by disability insurance, Medicare, and Unemployment Insurance might reduce the
estimated “static” antipoverty effects of those programs by one-eighth or
less. Although housing assistance provides significant benefits to some of
the poorest households including the homeless, its effects on labor supply
among those of working-age and free of disabilities are relatively modest.
Shroder (2010) reports that the net negative impact of rental assistance on
labor supply appears to vary among subgroups, may change over time, and
seems rather small relative to the amounts paid out in subsidy. Jacob and
Ludwig (2012) find that receipt of housing vouchers in Chicago during the
welfare reform era reduced employment by 3.6 percentage points among
able-bodied working-age individuals and reduced earnings an average 19
cents for each dollar of subsidies. Carlson et al. (2011) find employment
The War On Poverty 50 Years Later: A Progress Report  |  251

effects in the same range for Wisconsin voucher recipients. Ben-Shalom,
Moffitt, and Scholz estimate that the poverty rate among housing assistance
recipients is 66.0 percent, and use the Jacob and Ludwig data to estimate
that housing assistance lowers the poverty rate among recipients by 8.2
percentage points rather than the static estimate of 14.9 points. Finally, a
recent study by Chetty, Friedman, and Saez (2012) of the EITC on the earnings distribution also finds that “the impacts of the EITC … come primarily
through its mechanical effects.” They do find, however, that behavioral
responses reinforce the effects of the safety net on deep poverty, so that the
overall impact of the EITC might be somewhat greater than implied by the
“static” estimates because of the increased work induced by the program.24

Economic Mobility
When economic mobility is high, individuals and families can lift
themselves out of poverty by taking advantage of opportunities to improve
their economic well-being. When economic mobility is low, it is difficult
to change one’s economic status and people may become stuck in poverty.
Mobility can be measured either as relative mobility—the likelihood of moving up or down the income distribution—or absolute mobility, which is the
likelihood of improving economic well-being in general without necessarily
moving up in the income distribution (reflected in the saying, “a rising tide
lifts all boats”). When there is low relative mobility, there are few opportunities for poor people to improve their standing in society by moving up the
economic spectrum, and children from poor families are likely to continue
to have low economic and social status as they become adults, even if their
material well-being improves. When there is low absolute mobility, poor
people and their children find it hard to escape the economic and material
hardships of poverty.
While economic mobility in the United States allows some people to
escape poverty, many do not. About half of the poorest individuals remain
in the lowest one-fifth of the income distribution after 10 and 20 years, and
no more than one-fourth make it to one of the top three income quintiles
(Acs and Zimmerman 2008; Auten, Gee, and Turner 2013). Those who
were in poor families as children are estimated to have 20 to 40 percent
lower earnings as adults compared to those who did not grow up in poverty
(Mayer 1997; Corcoran and Adams 1997; Corcoran 2001; Duncan et al.
24 Unlike the earlier literature, they also find evidence of a slight downward adjustment
of earnings for workers near 200 percent of the poverty line. This is consistent with the
predictions of static labor supply theory since that level of earnings is in the phase-out
region of the credit. This negative impact is small, however, and Chetty, Friedman, and Saez
emphasize it is dominated by the (also small) positive impacts at low earnings levels.

252  |  Chapter 6

2012). Among the children of low-earning fathers, about two-fifths of sons
and one-fourth of daughters remained in the lowest earnings quintile when
they became their father’s age (Jäntti et al. 2006). Consistent with this, more
than one-third of children who grew up in the lowest family income quintile
were in the lowest quintile when they became adults (Isaacs 2008). Among
children born in 1971 with parents in the lowest income quintile, 8.4 percent
made it to the top quintile by age 26, compared to 9.0 percent of those born
in 1986, indicating that the chance of moving up the income distribution has
remained low and fairly stable over the past few decades (Chetty, Hendren,
Kline, Saez, and Turner 2014).25 More generally, studies find that about
one-third to slightly less than one-half of parents’ incomes are reflected in
their children’s incomes later in life (Chetty, Hendren, Kline, and Saez 2014,
Black and Devereux 2011, Lee and Solon 2009), indicating that parents heavily influence children’s economic fortunes.
The results of these studies imply strong lingering effects from growing up in poverty. The influence of poverty on future economic prospects
stem not just from one’s own family status, but from growing up in highpoverty neighborhoods that often have lower-quality schools, lower-paying
jobs, higher crime rates, and other conditions that can create disadvantages
(Sharkey 2009). Among children whose families were in the top three
quintiles of family income, growing up in a high-poverty neighborhood
raises the likelihood of downward mobility (falling at least one quintile)
by 52 percent (Sharkey 2009). A comparison across geographic regions
in the United States indicates that economic mobility is higher in areas
with a larger middle class, less residential segregation between low-income
and middle-income individuals, higher social capital, and lower rates of
teen birth, crime, divorce, and children raised by single parents (Chetty,
Hendren, Kline, and Saez 2014).
The above results mostly reflect relative mobility (moving up or down
the economic spectrum). While absolute mobility (improving economic
well-being without moving up the economic spectrum) has generally risen
in the United States, the pace has slowed in recent decades. Comparing
cohorts of men in their 30’s, median personal income went up 5 percent
from 1964 to 1994, but down 12 percent from 1974 to 2004 (Sawhill and
Morton 2007). Counting all family income, income went up 32 percent
between 1964 and 1994, and only 9 percent between 1974 and 2004. So the

25 The stability of intergenerational mobility is also shown by comparing the correlation of
parent and child income ranks over this period (Chetty, Hendren, Kline, Saez, and Turner
2014).

The War On Poverty 50 Years Later: A Progress Report  |  253

2012 $
300

Figure 6-9
Real Per Capita Expenditures on Select Programs, 1967–2012

250

EITC and
Refundable
CTC

200
150
100

2012

EITC

AFDC/TANF

50
0
1967

1977

1987

Source: Office of Management and Budget; CEA calculations.

1997

2007

recent improvement in family income, which reflects the rising employment
of women, was only one-third as large as in the earlier period.26
Despite the popular view that the United States is the land of opportunity, this Nation appears to have lower economic mobility than many other
developed countries. Measuring mobility by the strength of the dependence
of children’s incomes on parents’ incomes, the United States has similar
mobility as the United Kingdom and Italy, but lower mobility than other
European countries as well as Japan, Australia, and Canada (Solon 2002;
Jäntti et al. 2006; Corak 2006, 2011).
Can anything be done to increase mobility? A study of data on
families over time (including comparisons of twins and other siblings)
found that genetics and shared upbringing play a statistically significant, but
quantitatively minor role, in explaining adult earnings differences, indicating that environmental factors other than upbringing are largely responsible
(Björklund, Jäntti, and Solon 2005). The large variation in mobility across
countries and across regions (Chetty, Hendren, Kline, and Saez 2014) is further evidence for the fact that institutions and other potentially changeable
factors can have a large impact on mobility.
26 Using a different method to assess absolute mobility, about half of individuals moved out of
the bottom income quintile in the 1984-94 and 1994-2004 periods when the quintile thresholds
were fixed at the beginning of the period (Acs and Zimmerman, 2008).

254  |  Chapter 6

Education appears to be one of the key factors that drive economic
mobility. As shown in Figure 6-10, among families in the bottom fifth of
the income distribution, almost half (45 percent) of the children who did
not obtain college degrees remained in the poorest fifth as adults, while
only one-sixth (16 percent) of the children who obtained college degrees
remained in the poorest fifth (Isaacs, Sawhill, and Haskins 2008: 95).
The importance of education is also shown by the findings that economic mobility is higher in countries that have a greater public expenditure
on education (Ichino, Karabarbounis, and Moretti 2009) and areas of the
United States that have a higher-quality K-12 education system (Chetty,
Hendren, Kline, and Saez 2014), and is improved for children in the poorest one-third of families when states increase spending on elementary and
secondary education (Mayer and Lopoo 2008).

Intergenerational Returns
While much of the literature on mobility presented above is correlational, a handful of well-crafted studies that track the long-run outcomes of
children exposed to safety net programs highlight the potential for investments in these programs to generate large returns.
Early childhood education has been found by many researchers to
have dramatic, super-normal returns in terms of more favorable adult
outcomes. The Head Start program created early in the War on Poverty has
been heavily researched and the combined results show that it can “rightfully
be considered a success for much of the past fifty years” (Gibbs, Ludwig, and
Miller 2013: 61). Studies following children over time, and accounting for
the influence of family background by comparing siblings, found that Head
Start participants were more likely to complete high school and attend college (Garces, Thomas, and Currie 2002), and scored higher on a summary
index of young adult outcomes that also included crime, teen parenthood,
health status, and idleness (Deming 2009). The latter study found that Head
Start closed one-third of the gap in the summary outcome index between
children in families at the median and bottom quartiles of family income.
Using a regression discontinuity research design that compared access to
Head Start across counties, Ludwig and Miller (2007) found positive impacts
of Head Start on schooling attainment, the likelihood of attending college,
and mortality rates from causes that could be affected by Head Start. Gibbs,
Ludwig, and Miller (2013) suggest the combination of the benefits due to
Head Start might produce a benefit-cost ratio in excess of seven.
Randomized experiments studying the Perry Preschool Project,
Abecedarian Project, Chicago Child-Parent Centers, Early Training Project,
and Project CARE programs largely confirm these findings. A variety of
The War On Poverty 50 Years Later: A Progress Report  |  255

Percent
50

Figure 6-10
Economic Mobility for Children from First Income Quintile

Without College Degree

40

With College Degree

30
20
10
0

First

Second
Third
Fourth
Income Quintile as an Adult

Fifth

Note: Childhood income is the average family income from 1967-71. Adult Income is the average
from 1995, 1996, 1998, 2000, and 2002.
Source: Isaacs, Sawhill, and Haskins (2008).

well-done analyses find that the adult benefits for children—especially girls—
who participated included higher educational attainment, employment, and
earnings along with other benefits (Schweinhart et al. 2005; Anderson 2008;
Campbell et al. 2008; Heckman and Masterov 2007; Heckman et al. 2010;
Heckman et al. 2011). Heckman et al. (2009) find that the returns to one
preschool program (Perry) exceed the returns to equities.
In the past decade, researchers have identified long-run linkages
between early childhood (including exposure in-utero) health interventions
and long-term outcomes. For example, Almond, Chay, and Greenstone
(2006) document that the Johnson Administration used the threat of withheld Federal funds for the newly introduced Medicare program to force
hospitals to comply with the Civil Rights Act mandate to desegregate,
resulting in dramatic improvements in infant health and large declines in
the black-white gap in infant mortality in the 1960s. Chay, Guryan, and
Mazumder (2009) show these improvements in access to health care and
health soon after birth had echoes in the form of large student achievement
gains for black teenagers in the 1980’s, contributing to the decline in the
black-white test score gap. Their results suggest that improved health care
access and better early childhood health improve test scores by between 0.7
and 1 standard deviations—a very large impact that implies large increases
in lifetime earnings. For example, studying a different intervention, Chetty,
256  |  Chapter 6

Friedman, and Rockoff (2011) find the financial value (the net present value
at age 12 of the discounted increase in lifetime earnings) of a standard deviation increase in test scores to be $46,190 per grade.
While it may not be surprising that human capital interventions have
long-run returns, recent studies have also found intergenerational effects
on child outcomes from tax or near-cash transfers to their parents. That is,
recent evidence suggests that government transfers that ameliorate child
poverty by increasing family income have lasting, long-run benefits in terms
of better child outcomes. For example, Hoynes, Schanzenbach, and Almond
(2013) study the initial rollout of the Food Stamp (now SNAP) program
between its initial pilot in 1961 and 1975. While Food Stamps are distributed
as vouchers for food purchases, since their amount is generally less than
households spend on food, the vouchers likely affect family behavior in the
same manner as increased cash income (Hoynes and Schanzenbach 2009).
Hoynes, Schanzenbach, and Almond find that exposure to Food Stamps led
to improvements in adult health (reductions in the incidence of high blood
pressure and obesity) and, for women, increased economic self-sufficiency.
Similarly, Dahl and Lochner (2012) and Chetty, Friedman, and Rockoff
(2011) find that family receipt of additional income from refundable tax
credits improves the achievement test scores of children in the family.
Chetty, Friedman, and Rockoff estimate that the implied increase in adult
earnings due to improved achievement as a child is on the same order of
magnitude, and probably greater, than the value of the tax expenditures.
The results discussed above highlight the crucial fact that government
expenditures on the safety net have a strong economic justification. Not
only do they help to propel struggling adults back onto their feet and protect
them and their families from hardship, they improve opportunity and the
adult outcomes of their children. As such, the poverty-reducing impact of
these programs constitutes an important investment opportunity. To give a
sense of the magnitude of this opportunity, Holzer et al. (2008) estimate the
cost of childhood poverty at about $500 billion (in 2007 dollars) or about
4 percent of gross domestic product (GDP) annually in terms of foregone
earnings, increased costs of crime, and higher health expenditures and lower
health.
While the Holzer et al. study is correlational (though it attempts to
correct for hereditary components of the intergenerational income correlation), the concern over bias in this estimate is overwhelmed by its
magnitude. Based on Census Bureau estimates, the total poverty gap—the
shortfall between family resources and the SPM poverty thresholds—among
all families with children is about $59.8 billion in 2012, or 0.37 percent of
GDP. Even if the Holzer et al. estimate was double the “true” causal effect of
The War On Poverty 50 Years Later: A Progress Report  |  257

eliminating child poverty, the benefit would exceed the added costs five-fold.
These numbers create a powerful case for renewed efforts to fight poverty.

The Obama Administration’s Record
and Agenda to Strengthen Economic
Security and Increase Opportunity
The programs created during the War on Poverty and refined since
have provided crucial support to Americans in need. But challenges clearly
remain. In 2012, 49.7 million Americans, including 13.4 million children,
lived below the poverty line—an unacceptable number in the richest nation
in the world. There is clear evidence that antipoverty programs work, and we
must redouble our efforts to enhance and strengthen our safety net.
At the same time, we must realize that while antipoverty programs
are doing heroic work to lift struggling families from poverty, there is broad
consensus among economists that a strong national recovery in the short
run, and stronger economic growth in the long run, are necessary to sustain
progress in the fight against poverty. Indeed, as our social safety net reinforces the economic benefits of work by supplementing wages for low earners, a strong labor market with jobs available for all is an essential partner in
the fight against poverty. Given the growing economic inequality in the past
few decades, we must strive for balanced growth that benefits all Americans.
To do so, we must ensure that an economic expansion encompasses everyone, and commit to giving all Americans opportunities for lifelong learning
and skills development to ensure a broad base of human capital that will be
rewarded by good wages.
This section documents the Obama Administration’s record in continuing the fight to expand opportunity and reduce poverty, and discusses
the Administration’s proposals to strengthen the safety net and to improve
human capital and increase labor market earnings. These comprise a key
part of the broader economic strategy to further the recovery and increase
growth—all of which combat poverty.

Taking Immediate Action During the Economic Crisis
As the economy was sliding into the Great Recession, the Administration
took action to strengthen the safety net and prevent millions of Americans
from falling into poverty. The Recovery Act instituted a number of temporary antipoverty measures, including the creation of the Making Work Pay
tax credit worth up to $800 for a married couple, a $250 Economic Recovery
Payment for Social Security and SSI recipients; unemployment changes

258  |  Chapter 6

including an additional $25 a week (for up to 26 weeks) to regular UI beneficiaries, increased federal funding for the Extended Benefits program, incentives for states to modernize their UI system to reach part-time workers and
recent workforce entrants, and reauthorized Emergency Unemployment
Compensation; an increase in SNAP benefits; and an expansion of the
Community Services Block Grant (CSBG). The Recovery Act also delivered
nearly $100 billion to States, school districts, postsecondary institutions, and
students to help address budget shortfalls and meet the educational needs of
all. This total included $10 billion for the Title I Grants to Local Educational
Agencies program, a flagship program of the War on Poverty’s Elementary
and Secondary Education Act of 1965, which currently serves more than 23
million students in high-poverty schools, helping ensure access to a highquality public education.
In addition, the Recovery Act included expansions in tax credits
that the Administration intends to make permanent. The EITC expansion
increased the credit for families with three or more children, reflecting the
fact that they have greater needs and higher poverty rates than families with
two children. It also reduced the penalty that some low-income families
faced when getting married. Together, these two provisions benefit about
6 million households a year by an average of $500. Further, the partial
refundability of the CTC for working families was expanded, benefiting
12 million families by an average of $800 each. These changes were subsequently extended through 2017 and the President is proposing to make them
permanent.
The impact of these emergency measures on poverty was dramatic.
The Recovery Act played a large role in keeping Americans out of poverty
during the recession, as shown in Figure 11.27 In total, between 4.0 and 5.5
million people a year were kept out of poverty by these programs from 2009
to 2012. Without the Recovery Act, the Supplemental Poverty Rate would
have been 1.8 percentage points higher in 2009 and 1.7 percentage points
higher in 2010. Over the four years between 2009 and 2012, CEA estimates
that 19.2 million person years were kept from poverty as a result of the
expansions created by the Recovery Act alone (Figure 6-11). This calculation
is conservative, in that it does not account for the impact of those expansions on employment through increased aggregate demand, and does not
attempt to measure the impact of other components of the Recovery Act
such as increased funding for Pell Grants and paying for COBRA for the
unemployed.
27 Estimates of the effects of the Recovery Act on poverty have been updated relative to our
January 2014 report to reflect minor methodological improvements.

The War On Poverty 50 Years Later: A Progress Report  |  259

Expanding Health Care Security
The Affordable Care Act (ACA) ensures that all Americans have
access to quality, affordable health insurance and provides financial incentives to states to expand their Medicaid programs to adults who are in or
near poverty. As of January 2014, 26 states have adopted the Medicaid
expansion. For moderate-income Americans, the ACA provides tax credits
for the purchase of insurance through marketplaces and cost-sharing reductions. The Congressional Budget Office (CBO) estimates that, by 2016, these
measures will increase the number of Americans with health insurance by
25 million (CBO 2013).
Americans of all income levels are already benefiting from insurance
market reforms that ensure access to preventive services and no lifetime
coverage limits. Further, Americans buying insurance are no longer forced
to pay a higher premium because of their gender or health status, and can be
confident that their insurance provides adequate financial protection. The
ACA has also begun reforming the way the United States pays for health care
to promote more efficiency and quality in the health system, the early results
of which are discussed in a recent CEA report (CEA 2013) and in Chapter
4 of this report.
Nutrition programs like SNAP are vital to the economic livelihood of
many families and communities, especially in a recessionary period. Every
time a family uses SNAP benefits to put healthy food on the table, the benefits extend widely beyond those individuals. In fact, the U.S. Department
of Agriculture’s Economic Research Service estimates that an additional $1
billion in SNAP benefits supports an additional 8,900 to 17,900 full-timeequivalent jobs (Hanson 2010).

Rewarding Work
Work that pays enough to support a family is the most central antipoverty measure. In 2013, the Obama Administration finalized rules to
extend minimum wage and overtime protections to nearly 2 million directcare workers who provide care assistance to elderly people and people with
illnesses, injuries, or disabilities. This will ensure our nation’s health aides,
personal care aides, and certified nursing assistants receive the same basic
protections already provided to most U.S. workers, while improving the
quality and stability of care for those who rely on them.
Minimum wage and overtime protections have been a bulwark of
protection against poverty, but the minimum wage has not kept pace with
inflation. Today, a worker trying to support a family on the minimum wage
is still living in poverty. That is why the President signed an Executive Order

260  |  Chapter 6

Figure 6-11
Recovery Act and Subsequent Extensions: Cumulative
Person-Years Kept from Poverty, 2008–2012

Millions of Person-Years
21

19.2

18

15.2

15
12

10.8

9
5.5

6
3
0

2008

2009

2010

2011

2012

Source: Bureau of Labor Statistics, Current Population Survey, Annual Social and Economic
Supplement; CEA Calculations.

in 2014 to raise the minimum wage for workers on new and replacement
Federal service and construction contracts to $10.10 an hour. This step will
to ensure that no worker who provides services to the Federal government
will raise their families in poverty, and will make Federal procurement more
economical and efficient. An extensive body of research suggests that giving
a raise to lower-income workers reduces turnover and raises morale, and
can thus lower costs and improve productivity. To help insure that work is
rewarded for millions more Americans, the President supports the HarkinMiller bill to increase the minimum wage to $10.10 by 2016 (Box 6-5).
To further enhance economic security and incentivize work, the
President has proposed to double the childless EITC to be worth up to
$1,000 and lower the age threshold from 25 to 21 to help more lower-income
young people, while continuing protections to ensure that it does not
benefit, for example, typical full-time students. A small EITC for childless
households was established in 1993, but its maximum value is expected to
be only $503 in 2015 and is fully phased out for individuals making over
$14,790 ($20,290 for married couples). This leaves a childless adult with
wages equal to the poverty line with a federal tax burden (including income
and payroll taxes) of $1,966 after receiving an EITC of $173, driving them
deeper into poverty and making childless workers the sole demographic

The War On Poverty 50 Years Later: A Progress Report  |  261

Box 6-5: Raising the Minimum Wage
In 2013, the federal minimum wage was at the same inflationadjusted level as it was in 1950. A full-time worker earning $7.25 per
hour would not be able to keep a family of four out of poverty, even
with the help of the EITC and CTC. Instead of allowing the value of the
minimum wage to continue to erode while incomes at the top of the
distribution soar, it is time to make the minimum wage a wage people
can live on.
Raising the minimum wage to $10.10 per hour would help a large,
diverse group of workers, and indexing it to inflation would ensure that
its real value does not deteriorate over time, as it has after past increases.
Over 28 million people earning wages near the minimum wage would
be affected by such an increase, 46 percent of whom have household
incomes below $35,000. Full-time workers earning exactly the minimum
wage would see their earnings increase by $5,700, enough to move a
family of four from 17 percent below the poverty line to 5 percent above
it, once tax credit assistance is included. CEA estimates that raising the
minimum wage to $10.10 by 2016 would lift roughly 2 million workers
whose wages are currently near the minimum wage, and members of
their families, out of poverty, while alleviating poverty for about 10 million more.
Opponents of increasing the minimum wage argue that the beneficiaries are largely middle-class teenagers, and those most in need of
assistance are kept out of jobs by high wages. The available evidence does
not support those claims. Among those workers who would be affected
by increasing the minimum wage to $10.10, 92 percent are more than 18
years old. A large literature has considered the effects of minimum wages
on employment, and the best evidence suggests there is little to no effect
(Doucouliagos and Stanley 2009). While a higher minimum wage could
increase compensation costs for employers, they could also reap benefits,
including lower employee turnover rates and, by extension, lower costs
of hiring and training new workers, as well as increased demand for their
goods and services among low-wage workers.

driven deeper into poverty by the federal tax system. Under the President’s
proposal, a household at the poverty line would see its EITC expand to $848,
more than eliminating its income taxes—although it would still pay net taxes
on earnings when including payroll taxes.

262  |  Chapter 6

Empowering Every Child with a Quality Education
To prepare Americans for the jobs of the future, we have to strengthen
our investments in the Nation’s educational system. The Administration has
invested in coordinated state systems of early learning and proposed new
policies building on evidence of how to create a foundation for success in the
formative early years of life. High-quality early learning and development
programs can help level the playing field for children from lower-income
families on vocabulary, social and emotional development, and academic
skills while helping students to stay on track and stay engaged in the early
elementary grades. These programs generate a significant return on investment for society through a reduced need for spending on other services,
such as remedial education, grade repetition, and special education, as well
as increased later productivity and earnings. The Administration’s comprehensive early-learning agenda invests in early childhood education, care,
and development for our nation’s youngest learners. In partnership with the
states, the Preschool for All initiative would provide high-quality preschool
for four-year olds from low- and moderate-income families, while encouraging states to serve additional four-year olds from middle-income families.
The Administration has also proposed investments in high-quality early
learning for infants and toddlers through new Early Head Start-Child Care
Partnerships, as well as an extension and expansion of evidence-based voluntary home visiting programs that allow nurses, social workers, educators,
and other professionals to connect pregnant women and vulnerable families
with young children to tools that positively impact the child’s health, development, and ability to learn.
Over the past 50 years, improvements in education at all levels have
produced large returns for many Americans and played a key role in the
economic mobility of children born to poor families. Because economic
progress and educational achievement are inextricably linked, educating
every American student to graduate from high school prepared for college
and for a career is a national imperative. The President has articulated a goal
for America to once again lead the world in college completion by the year
2020, and the Administration’s education efforts aim toward this overarching objective. To provide a high-quality education to all American children,
the Administration, in partnership with states, has advanced reforms of
the Nation’s K-12 education system to support higher standards that will
prepare students to succeed in college and the workplace; efforts to recruit,
prepare, develop, and advance effective teachers and principals; efforts to
eliminate discrimination on the basis of race, color, national origin, gender,
and disability in public school; efforts to ensure the use of data in the classroom; and a national effort to turn around our chronically lowest-achieving
The War On Poverty 50 Years Later: A Progress Report  |  263

schools. The School Improvement Grants (SIG) program has invested over
$5 billion in 1500 of the nation’s lowest performing schools. The Promise
Neighborhood Program, established by this Administration in 2010, has
funded 58 neighborhoods across the country to design comprehensive projects that include a continuum of services and designed to combat the effects
of poverty and improve education and life outcomes, from birth through
college to career. The Administration has also put forward proposals to
redesign the Nation’s high schools to better engage students and to connect
99 percent of students to high-speed broadband and digital learning tools
within the next five years. Continued investments and reforms are needed to
ensure that all students have access to a high-quality education that prepares
them for college and a career, and for success in today’s global economy.
With the average earnings of college graduates at a level twice as high
as that of workers with only a high school diploma, higher education is now
the clearest pathway into the middle class. Our Nation suffers from a college
attainment gap, as high school graduates from the wealthiest families are
almost certain to continue on to higher education, while just over half of
high school graduates in the poorest one-quarter of families attend college.
This gap has increased over the past several decades. And while more than
half of college students graduate within six years, the completion rate for
low-income students is around 25 percent. To achieve the President’s goal
for college completion, ensure that America’s students and workers receive
the education and training needed for the jobs of today and tomorrow, and
provide greater security for the middle class, the Administration is working
to make college more accessible, affordable, and attainable for all American
families. Under President Obama, Pell Grant funding was increased to serve
over 3 million additional low-income students and the average grant was
increased by more than $900. The Administration also created the American
Opportunity Tax Credit to ease college costs for over 9 million families, and
championed comprehensive reform of student loans that will save taxpayers
$68 billion over the next decade. Finally, the Administration has launched
an array of policies to contain college costs, and make it easier for students to
manage their student debt through income-based repayment reforms, which
limit student loan payments to a percentage of their income so that young
workers will not be swamped by debt payments as they are establishing their
households and careers.
Given the critical importance of education in expanding skills and
opportunities, the Administration implemented several policies designed
to increase college access and affordability for low-income students.
Recognizing that the opportunity to acquire the skills to get and keep a good
job starts early and through education, the President is also proposing to
264  |  Chapter 6

modernize America’s high schools for real-world learning. The goal is to
provide challenging, relevant experiences, and reward schools that develop
new partnerships with colleges and employers, and that create classes that
focus on technology, science, engineering, and other skills today’s employers
are demanding to fill jobs now and in the future.

Creating Jobs and Growing Our Economy
Building on the evidence that well-designed training programs
can improve employment and earnings (Andersson et al. 2013), this
Administration has proposed investing in subsidized employment and
training opportunities for adults who are low-income or long-term unemployed. In 2009 and 2010, 372,000 low-income youth were placed into summer and year-round employment, and supported job opportunities were
created for about 260,000 low-income individuals. In addition, the President
continues to build public-private partnerships to provide opportunities for
low-income youth.
The Administration is using all available tools to help people who
have lost their jobs to find new work or to train for new careers in growth
fields that will provide better jobs and paths to viable careers. This includes
supporting training opportunities that lead directly to a job, and making
sure our unemployment system promotes re-employment through wideranging reforms to the unemployment insurance program, some of which
were adopted in the Middle Class Tax Relief and Jobs Creation Act of 2012,
and continued investment in reemployment services, which have proven
effective in speeding the return to work.
The President has proposed to build on these successes by further
investing in creating job and work-based training opportunities for the
long-term unemployed and youth seeking skills and wanting to get into the
workplace.
This Administration has already invested $1.5 billion in community
college-business partnerships in all 50 states to build capacity and develop
curricula to train workers for jobs in growing industries. President Obama
has proposed to build on these successes with further investments that will
transform community college education and support Americans in getting
training to enter skilled jobs.

Investing in and Rebuilding Hard-Hit Communities
Living in a high-poverty area presents various challenges, including
crime, limited access to quality education, and scarcity of good jobs. Since
these issues often interact with each other and compound the problems they

The War On Poverty 50 Years Later: A Progress Report  |  265

create individually, it is very difficult for people, particularly children, to
overcome the disadvantages created by and associated with poverty.
A child’s zip code should never determine his or her destiny. To help
provide ladders of opportunity so that every child has a chance to succeed,
the Administration is working with State and local governments to focus
public and private resources on transforming areas of high need into communities of opportunity.
The Administration’s Promise Zones initiative focuses existing government resources on competitively selected communities and leverages
private investment to create jobs, improve public safety, increase educational
opportunities, and provide affordable housing. The Administration will
designate 20 communities over the next several years with this intense and
layered approach to community revitalization.
That approach includes working with local leadership, and bringing
to bear the resources of the President’s signature revitalization initiatives
from the Department of Education, the Department of Housing and Urban
Development, the Department of Agriculture, and the Department of
Justice, to ensure that federal programs and resources support the efforts to
turn around 20 of the highest poverty urban, rural and tribal communities
across the country.
The Promise Zones initiative will build on existing programs, including HUD’s Choice Neighborhoods and the Department of Education’s
Promise Neighborhoods grant programs. The Administration has invested
$244 million in Choice and $157 million in Promise since 2010. For every
federal dollar spent, Choice Neighborhoods has attracted eight dollars of
private and other investment and has developed nearly 100,000 units of
mixed-income housing in 260 communities, ensuring that low-income
residents can afford to continue living in their communities. Promise
Neighborhoods grants are supporting approximately 50 communities representing more than 700 schools. To help leverage and sustain grant work,
1,000 national, state, and community organizations have signed-on to partner with a Promise Neighborhood site. By expanding these programs, the
Administration continues to support local efforts to transform low-income
urban, rural, and tribal communities across the country.

Conclusion
The War on Poverty represented a dramatic shift in the Federal
Government’s priorities for helping those who are left behind in a growing
economy. It set in motion a series of changes that transformed our social
safety net and improved the well-being and economic outcomes of countless

266  |  Chapter 6

low-income Americans and their children. The architects of the War on
Poverty believed that the combined effect of government policies attacking poverty on many fronts—providing income when earnings are low,
providing access to health insurance, insuring that people have shelter and a
minimal food budget, and providing access to education at all ages—would
dramatically raise employment and earnings and reduce material hardship.
In the years since 1964, this optimism and belief in the capacity of government to improve the lives of less fortunate Americans has at times given way
to the cynical belief that the safety net is ineffective, or even exacerbates the
problems of the poor by reducing the incentives for those able to work to
do so.
The most important lesson from the War on Poverty is that government programs and policies can lift people from poverty; indeed they have
for the past 50 years. Poverty rates fell from 25.8 percent in 1967 to 16 percent
in 2012—a decline of nearly 40 percent. In 2012 alone, the combined effect
of all federal tax, cash and in-kind aid programs was to lift approximately
14.5 percent of the population—over 45 million people—out of poverty.
But another lesson is that we cannot afford to simply embrace any
program that purports to achieve this goal or attempt to freeze them in time.
Instead, our antipoverty efforts have benefited from enormous changes in
public policy since the 1960s, informed by a wealth of research on both successful and failed programs that provide important insights into what does
and does not work in fighting poverty. Our safety net has evolved to put
more emphasis on rewarding and supporting work, such as by providing
greater support to working families through the EITC and refundable CTC,
while also making it easier for them to get help from programs like SNAP
and Medicaid. In 1967, we spent $19 billion in today’s dollars on what was
then called AFDC and nothing on the EITC. Today the EITC and partially
refundable CTC are 3.8 times the size of the TANF program.28 Meanwhile,
the Affordable Care Act advances the goal of providing quality affordable
health care to all Americans, with financial incentives to states to expand
their Medicaid programs to adults who are in or near poverty and generous
tax credits for moderate-income households. Our safety net remains imperfect, but these reforms and improvements represent important progress—
and they also help many families work and raise the rewards to that work.
Nearly 50 million Americans still live in poverty, however, including
13.4 million children, and so there remains a need to do more to help the
poor. The 1964 Economic Report of the President estimated that the total
shortfall in income necessary to bring all poor families up to the poverty
28 Based on CEA calculations using data from http://www.whitehouse.gov/sites/default/files/
omb/budget/fy2014/assets/hist.pdf.

The War On Poverty 50 Years Later: A Progress Report  |  267

level was $11 billion (about $71 billion in today’s dollars), or about 1.6
percent of the country’s annual GDP. Our nation has grown much richer
since, and today the total shortfall below poverty is only 0.6 percent of GDP.
Continuing to make progress in closing that shortfall will require not just
defending the programs that have helped reduce poverty but also continuing the efforts to strengthen the economy, increase growth and ensure that
growth is reflected in broad-based wage gains so that families can lift themselves out of poverty.

268  |  Chapter 6

C H A P T E R

7

EVALUATION AS A TOOL FOR
IMPROVING FEDERAL PROGRAMS

S

ince taking office, President Obama has emphasized the need to determine what works and what does not in government, and to use those
answers to inform Federal policy and budget decisions. The President’s 21st
Century Management Agenda, submitted to Congress with the fiscal year
2010 Budget, set bold goals for building a more efficient, more effective
government that contributes to economic growth and strengthens the foundations for economic prosperity (OMB 2009a). Today, evaluating Federal
programs and interventions to understand their impact, and developing the
infrastructure within agencies to support a sustained level of high-quality
evaluations, remains an Administration priority. By rigorously testing which
programs and interventions are most effective at achieving important goals,
the government can improve its programs, scaling up the approaches that
work best and modifying or discontinuing those that are less effective.
This Administration has supported the use of rigorous, high-quality
“impact” evaluations to measure changes in a variety of outcomes targeted
by Federal programs, ranging from earnings to health to electricity usage.
Many factors affect whether Federal programs achieve their goals, and
identifying impacts specifically attributable to programs is challenging.
An impact evaluation is a particular type of program evaluation, and aims
to measure the causal effect of a program or intervention on important
program outcomes. This chapter focuses on impact evaluations. “Process”
evaluations (another type of program evaluation) and performance measurement also contribute to building evidence about how well programs are
working, but differ in important ways from impact evaluations (Box 7-1).
Building on the efforts of previous administrations, the Obama
Administration is working to reform the Federal Government’s approach to
improving program performance. In addition to emphasizing transparency
and accountability in tracking progress toward agencies’ priority goals, this
new approach also aims to complement and to draw on the Administration’s
269

Box 7-1: Impact Evaluations, Process Evaluations,
and Performance Measurement
Program managers use many approaches to assess how programs
operate and how well they work. Impact evaluations aim to identify
the causal effects of a program or intervention on some outcome or
outcomes of interest. Impact evaluations are distinct from other types of
program evaluation and performance measurement. For example:
•	 Process evaluations analyze the effectiveness of how programs
deliver services relative to program design, professional standards, or
regulatory requirements. For instance, a process evaluation might focus
on whether a program is reaching the target number of participants or
whether caseworkers are consistently following a specified protocol for
providing services. Process evaluations help ensure that programs are
running as intended, but in general these evaluations do not directly
examine whether programs are achieving their outcome goals (GAO
2011).
•	 Performance measurement is a broader category that encompasses “the ongoing monitoring and reporting of program accomplishments, particularly progress toward pre-established goals” (GAO 2011).
Typically, performance measures provide a descriptive picture of how
a program is functioning and how participants are faring on various
“intermediate” outcomes, but do not attempt to rigorously identify the
causal effects of the program. For instance, performance measures for
a job training program might capture how many individuals are served,
what fraction complete the training, and what fraction are employed a
year later. But these measures will not answer the question of how much
higher these individuals’ employment rates are as a result of having completed the training. Nonetheless, performance measures serve as important indicators of program accomplishments and can help establish that
a program is producing apparently promising (or troubling) outcomes.
While process evaluations and performance measurement are useful at all stages of a program’s maturity, they can be particularly useful
for providing evidence about how programs are working in the early
years of a program’s history when impacts on program outcomes may
not be detectable and rigorous, high-quality impact evaluations are not
possible. A logic model—a tool that depicts the intended links between
program investments and outcomes and helps to ensure program
activities will achieve desired outcomes—can facilitate agency efforts to
develop high-quality “intermediate” indicators of impacts as well as an
understanding of alternative causal channels that can affect important
program outcomes.

270  |  Chapter 7

program evaluation efforts. For example, the Administration this year is
establishing strategic reviews within agencies to strengthen the use of evidence in strategic and budget decisions.
This chapter provides an overview of the implementation and use of
impact evaluation in Federal programs, with a special focus on the lessons
learned so far in this Administration. It begins with a discussion of some
challenges inherent in conducting rigorous impact evaluations in government programs. The chapter then focuses on Administration efforts to build
and to use evidence, including actions taken on lessons learned from completed evaluations, launching new evaluations in areas where not enough
is known, and creating a culture of evidence-building in Federal programs,
especially grant programs. The final section identifies opportunities for
further progress: for example, through increasing legislative support and
removing legislative barriers, embedding evaluation into routine program
operations, and using existing program data to measure outcomes and
impacts.

Conducting Rigorous Impact
Evaluations in Federal Programs
Science, business, and government routinely confront the problem of
ascertaining the effect of a program, policy, or initiative. Is a newly developed
drug effective in treating the condition for which it was developed? Does a
new marketing strategy boost sales? Does a preschool program improve
participants’ outcomes, such as success in elementary school? Despite the
different settings, these questions all focus on measuring the effect of an
intervention or program on one or more outcomes of interest.
One basic approach to answering questions like these is to look at outcomes before and after the “treatment”—for instance, before and after taking
a drug, before and after a new marketing strategy is rolled out, or before
and after participation in an education program. Another straightforward
approach is to compare outcomes for program participants with outcomes
for non-participants. In complex policy environments, however, these
simple approaches will often give the wrong answers. Take, for example, a
job training program designed to help unemployed workers get jobs. The
data may show that program participants were much more likely to be
employed a year after the training program than before they entered the program. But if the unemployment rate has fallen substantially over the course
of the program, then the gains may be due to the improving economy, not
to the training program. Similarly, a government program offering start-up
assistance to new businesses may appear to boost success rates. But if capable
Evaluation as a Tool for Improving Federal Programs  |  271

entrepreneurs are more likely than less capable ones to participate in the
program, then self-selection of program participants, not the program itself,
may be driving those better outcomes.
A strong impact evaluation needs a strategy for constructing more
valid comparisons—specifically, for identifying “treatment” and “control”
groups for which differences in outcomes can reasonably be attributed to
the program or intervention rather than to some other factor. Impact evaluations conducted using rigorous, high-quality methods provide the greatest
confidence that observed changes in outcomes targeted by the program are
indeed attributable to the program or intervention. It is well recognized
within Congress and other branches of government (for example, GAO
2012, National Research Council 2009), in the private sector (Manzi 2012),
in non-governmental research organizations (Coalition for Evidence-Based
Policy 2012, Walker et al. 2006), and in academia (for example, Imbens
2010; Angrist and Krueger 1999; Burtless 1995) that evaluations measuring
impacts on outcomes using random assignment provide the most definitive
evidence of program effectiveness.
Although the classic impact evaluation design entails random assignment of recipients into treatment and control groups as part of the experiment, the goal of constructing valid comparisons sometimes can be achieved
by taking advantage of natural variation that produces as-if randomness, an
approach referred to as a quasi-experiment. Quasi-experiments can sometimes be much less expensive than traditional large-scale random assignment experiments, and are discussed further below.

Estimation of Causal Effects of a Program or Intervention
The starting point for estimating the causal effect of a program or
intervention is being precise about what constitutes a causal effect. Consider
a treatment delivered at the individual level: either the individual received
the treatment, or did not. The difference between the potential outcome if
the individual received the treatment and the potential outcome if the individual did not is the effect of the treatment on the individual.1 The challenge
of estimating this treatment effect stems from the fact that any given individual either receives the treatment or does not (for example, a child either
does or does not attend preschool). Thus, for any given person only one of
two potential outcomes can be observed. The fact that we cannot directly
observe the counterfactual outcome (for example, the earnings a person who
1 No two individuals are the same, so in general the effect of a program or intervention differs
from one individual to the next. For example, the effect of the preschool program could
depend on the child’s learning opportunities at home. Impact evaluation typically focuses on
estimating an average causal effect, which is the average of the individual-level causal effects.

272  |  Chapter 7

went to preschool would have had if they had, in fact, not gone to preschool)
implies that we cannot directly measure the causal effect. This problem of
observing only one of the potential outcomes for any given individual is the
fundamental problem of causal inference (Holland 1986).
Randomization provides a solution to the problem of not observing the counterfactual outcome. If individuals are randomly assigned to
treatment and control groups, then on average the individuals in the two
groups are likely to be the same in terms of other characteristics that might
affect outcomes. As a result, one can safely assume that ex-post differences
between the groups are the result of the treatment. To take the preschool
example, simply comparing test scores of all U.S. elementary school children
who had attended preschool to all U.S. elementary school children who
had not would not provide confidence that higher test scores for the first
group were an effect of preschool. The scores might reflect differences in
family background, elementary school resources, or other important factors
between the two groups. On the other hand, if a group of three-year olds
are randomly assigned to attend or not attend preschool, and the preschool
group has higher test scores in third grade, we can attribute the test score
gains to attending preschool because the two groups would not be systematically different along other dimensions that might impact learning.
In most cases, simple comparisons of treated and untreated individuals without random assignment will not produce valid comparisons
because treatment status will be correlated with other important factors.
For example, if potential preschool enrollees were initially screened so that
those with the least learning opportunities outside school were placed in
the program, then we might find that the treatment group (enrollees) has
worse outcomes than the control group. However, the reason for this finding is that enrollees are more disadvantaged than non-enrollees. The variation between treatment and control groups affects ultimate outcomes both
through the treatment and the differences in learning opportunities outside
school. Thus, any comparison of outcomes between treatment and control
groups would measure the combined effects of both the treatment and those
differences in learning opportunities.
Because randomized experiments can be expensive or infeasible,
researchers have also developed methods to use as-if random variation in
what is known as a quasi-experiment. The necessary condition for a highquality quasi-experimental design is that people are assigned to a treatment
or control group in a way that mimics randomness. This can be done by
forming treatment and control groups whose individuals have similar
observable characteristics, and exploiting some rule that governs assignment

Evaluation as a Tool for Improving Federal Programs  |  273

to the treatment and control groups in a way that is plausibly unrelated to
the outcome of interest.
One example of a quasi-experimental design that lends itself to
estimating impacts of programs or interventions is when eligibility is determined based on one or more variables in a way that individuals who (just)
qualify for the program are very much like those who (just) do not. If so, and
if both eligible and ineligible applicants are tracked, then a method called
regression discontinuity design can be used to compare the outcomes for
individuals on the two sides of the threshold, controlling for other observable differences between the two groups.
Another example of quasi-experimental design is when a program
varies across units for reasons unrelated to the program outcomes. Rothstein
(2011), for example, exploits the fact that, due to different business cycle patterns combined with policy variation created by expirations and renewals of
the Emergency Unemployment Compensation (EUC) program during the
Great Recession, the number of available weeks of benefits available to jobseekers varied dramatically from month to month in differing ways across
states. After controlling for local economic conditions, the haphazard nature
of the changes in EUC benefit levels across states enabled estimation of EUC
benefits on job-finding rates.
Describing the whole range of quasi-experimental approaches is
beyond the scope of this chapter.2 Quasi-experiments require stronger
assumptions than randomized experiments and the debate around those
assumptions makes it harder for quasi-experiments to be convincing, especially to non-experts. However, if the quasi-experimental variation used is
plausibly unrelated to the outcomes of interest except through the treatment,
quasi-experimental evidence can be convincing, with some methods and
applications being nearly as compelling as randomized trials and others
leaving more room for doubt.

Other Criteria for High-Quality, Successful Impact Evaluations
A strong impact evaluation also needs to address questions that are
actionable and relate to outcomes that matter. In some cases, the actionable
information might identify if a program is or is not effective. In other cases,
the actionable information might identify which interventions are best at
achieving important program outcomes, so that programs can be improved
2 For more extensive introductions to impact evaluation (both randomized experiments
and quasi-experiments), see Angrist and Pischke (2008, ch. 1) and Stock and Watson (2010,
ch. 13). Shadish, Cook, and Campbell (2002) and Berk and Rossi (1998) provide more
advanced textbook treatments, and Imbens and Wooldridge (2009) provide a survey of recent
methodological developments in the field.

274  |  Chapter 7

by adopting successful interventions more broadly. However, if there are
legal or other impediments to expanding an evaluated small-scale intervention, then learning that the intervention works does not directly lead to an
action that can improve a program at a national level. In such cases, it may
be better to allocate scarce evaluation resources to testing more modest
interventions or ways to run the program more effectively.
For the second of these criteria—outcomes that matter—the longterm goals of a program must be considered. For a preschool program, the
number of students enrolled is an easy-to-measure intermediate outcome.
However, preschool enrollment may or may not be related to ultimate outcomes, such as high school graduation rates, employment rates, or income.
It is also important to consider program size and stage of development, as
programs or interventions must be sufficiently mature, and treatment and
control groups sufficiently large, to obtain credible estimates of impact.
Other issues must also be addressed to conduct policy-relevant impact
evaluations in government programs. At the most practical level, rigorous
evaluation requires adequate funding, staff expertise, and often cooperation
across different parts of an agency (or across multiple agencies). Rigorous
evaluation also requires support from top agency management and program
managers. Further, many Federal programs have multiple goals, which can
make it hard to take action on evaluation findings when the results support
some goals but not others.
An important part of evaluating a program is remaining open to the
findings, regardless of the outcome, to inform the best course of action to
improve outcomes going forward. Findings of positive impacts provide
important feedback that may indicate whether additional investment is
warranted. Findings of no impact, either for all participants and program
goals or for important subsets of the participants and program goals, also
send valuable signals that modifications—including reallocating program
funding to other strategies that could better achieve outcomes—are needed.

Lower-Cost Ways for Impact Evaluations to Facilitate Real-Time
Learning
Large-scale random-assignment studies of social programs have
been very influential, but also can be quite expensive, and their expense has
been a major impediment to wide-scale adoption of learning and program
improvement through randomization. For this reason, researchers have
focused on lower-cost methods for learning about program effectiveness.
One lower-cost method is to build randomization into the design of
the program, so that data on program performance can be tracked and evaluated on an ongoing basis. This strategy has been pioneered as a management
Evaluation as a Tool for Improving Federal Programs  |  275

tool in the private sector for ongoing product and process improvement.
Indeed, some companies run thousands of randomized studies annually: by
2000, Capital One was running 60,000 studies annually using randomization methods, as they experimented with different strategies to determine
what works. Google has also run randomized experiments in the tens of
thousands in some years (Manzi 2012).
In the public sector, Federal agencies are also finding ways to conduct high-quality evaluation strategies at lower cost, including ways that
employ the lessons learned from behavioral economics (Box 7-2). The U.
S. Department of Agriculture’s Food and Nutrition Service is conducting
a range of rigorous demonstration projects to further develop the evidence
base of effective strategies for programs that address food insecurity and
improve nutrition among children; one such project implements low-cost
environmental changes in lunchrooms to encourage students to make
healthier food choices. One demonstration found that merely placing fruit
in a colorful bowl in a convenient part of the lunch line can lead to an
increase in fruit sales of up to 102 percent (Wansink, Just, and Smith 2011).
Funding research for these simple, evidence-based interventions allows
for the development of effective strategies to strengthen the nutrition and
hunger safety net for the more than 30 million children fed by the National
School Lunch Program.
Utilizing existing data and independent programmatic changes to
measure outcomes is another strategy that agencies are using to minimize
evaluation costs. For example, the Department of Justice’s National Institute
of Justice conducts impact evaluations of interventions that can help inform
the approximately 18,000 local law enforcement agencies that do not individually have the resources to test interventions on their own. Hawaii’s
Opportunity Probation with Enforcement (HOPE) program was established
as a demonstration pilot for drug-involved probationers in Hawaii. The
pilot tested the efficacy of “swift and certain” sanctions against probationers who fail to meet the conditions of their probation. The randomized
controlled experiment found that after one year, probationers who received
very frequent drug testing (every other day) and—if they failed the drug
test—an immediate court date and a modest but certain sanction (a night
in jail), were 72 percent less likely to use drugs, 61 percent less likely to skip
meetings with their supervisory officers, 55 percent less likely to be arrested
for a new crime, and 53 percent less likely to have their probation revoked.
These reductions led to HOPE participants being sentenced to an average
of 48 fewer days in prison than those in the control group who received the
traditional delayed but more severe sentence (National Institute of Justice).
Because of the high costs associated with servicing inmates in prison,
276  |  Chapter 7

Box 7-2: Using Behavioral Economics to Inform
Potential Program Improvements
Increasingly, agencies are using insights from behavioral science
to implement low-cost evaluations that can be used to improve program
design. Utilizing randomized experiments or other rigorous evaluation
designs, these studies examine aspects of program operations that can be
re-designed to help people take better advantage of available programs
and services—such as by simplifying application processes or highlighting the availability of student financial aid. Recently, the White House
Office of Science and Technology Policy assembled a cross-agency
team of behavioral science and evaluation experts—the U.S. Social and
Behavioral Sciences Team—to help agencies identify promising opportunities for embedding behavioral insights into program designs and to
provide the necessary technical tools to rigorously evaluate impact. Such
low-cost, real-time experiments can help Federal programs operate more
effectively and efficiently.

any intervention that reduces prison time can generate large savings. By
making use of available administrative data, evaluations employing quasiexperimental and randomized controlled trial designs were implemented
at a cost of only $150,000 and $230,000, respectively.3 A follow-up analysis
is examining the long-term impacts of the intervention; this model is also
being piloted in four other locations.
The often-lengthy time between implementation and results of a
rigorous evaluation can also discourage its use, but agencies are looking for
ways to speed up the evaluation process to gain actionable insights more
quickly. For example, the Center for Medicare and Medicaid Innovation (the
“Innovation Center”), which was created by the Affordable Care Act of 2010,
is using an innovative “Rapid Cycle” approach and high-quality evaluation
methods to develop and test innovative payment and service delivery models designed to reduce expenditures while preserving or enhancing quality
of care for Medicare, Medicaid and Children’s Health Insurance Program
beneficiaries. By giving more rapid feedback to health providers, as Box 7-3
shows, the Rapid Cycle approach provides actionable information, allows
for more frequent course corrections, and supports continuous quality
improvement (Shrank 2013).

3 Cost estimates supplied by the Department of Justice’s National Institute of Justice.

Evaluation as a Tool for Improving Federal Programs  |  277

Impact of the Evidence-Based Agenda
From its first months, the Administration embedded a strong evaluation focus into many new initiatives to learn what strategies work best and
to scale up approaches backed by strong evidence. During the formulation
of the FY 2011 Budget in fall 2009, the Office of Management and Budget
(OMB) invited agencies to submit new evaluation proposals for building
rigorous evidence and also encouraged agencies to demonstrate that new
program initiatives were based on credible evidence of success or to include
plans to collect evidence where none exists. The Administration has maintained its emphasis on using and building evidence in every subsequent
budget (OMB 2010, 2011, 2012, 2013a, 2013b).

Uses of Evaluation
Agencies have used impact evaluations to inform policy and program
decisions in a wide variety of ways.
Making the Unemployment Insurance System More Effective.
Unemployment insurance (UI) provides an important safety net for
workers who become unemployed. Occasionally, concerns are raised that
UI payments could reduce an unemployed worker’s incentive to find
employment. While the evidence suggests that any such effects are small
(Council of Economic Advisers and the Department of Labor 2013), the
Federal Reemployment and Eligibility Assessment (REA) initiative started
providing funds in 2005 to states and sought to reduce UI duration by
combining in-person UI eligibility reviews with (1) labor market information, (2) developing a reemployment plan, and (3) offering a referral to
reemployment services. The Department of Labor funded research using a
randomized design that showed the REA initiative was effective in reducing
the duration of UI (Benus et al. 2008). However, these studies focused on
measuring reduced duration on UI and associated costs and not on other
outcomes, such as return to employment or increased wages. These studies were followed by another randomized controlled trial which showed
that the REAs were also effective at reducing joblessness when eligibility
assessments were personalized and more closely integrated with the delivery
of reemployment services (Poe-Yamagata et al. 2011). Consequently, the
Administration proposed in the American Jobs Act to create a requirement
that all Emergency Unemployment Compensation claimants receive both an
REA and reemployment services; this was enacted in the Middle Class Tax
Relief and Job Creation Act of 2012.4 Evidence from rigorous evaluations
is playing a role in making the REA initiative more effective and getting
4 Public Law 112-96.

278  |  Chapter 7

Box 7-3: “Rapid Cycle” Evaluations in the Center
for Medicare and Medicaid Innovation
The Center for Medicare and Medicaid Innovation (the “Innovation
Center”) has invested in building information systems and institutional
capacity that permit a “Rapid Cycle” approach to testing a variety of new
models. For instance, evaluators can gather real-time information and
provide performance data to providers who adopted the same model,
allowing these providers to understand and track their own performance,
and to compare their performance with that of other providers. In the
early stages of model implementation, the Innovation Center is applying
this Rapid Cycle approach primarily to process evaluations (see Box 7-1),
turning later to assessing the models’ overall impacts (including impacts
on important subgroups, where feasible) once there is reasonable assurance that the model has been in operation sufficiently long to detect
impacts. When experimental conditions cannot be met due to logistical
or other constraints, quasi-experimental methods are used, with multiple
comparison groups for each treatment group identified where possible to
provide robustness checks of the findings (Shrank 2013).

unemployed Americans back to work faster, and the Administration has
sought to expand it to cover more workers. A modest increase in funding
for REAs was included in the recently enacted Consolidated Appropriations
Act of 2014.
Simplifying Applications for Student Aid. In many cases, actionable
evidence on what works comes from field-generated, grant-funded research
rather than from Federal program evaluations. In 2008, with support from
the Department of Education’s Federal Student Aid Office, Institute of
Education Sciences, and other funders, university-based researchers worked
with H&R Block to set up an experiment providing randomly selected
low-income tax filers in North Carolina and Ohio with pre-populated Free
Application for Federal Student Aid (FAFSA) forms and FAFSA assistance
for themselves or their children, as well as with information about student
aid. This relatively low-cost intervention had a surprisingly large effect on
college enrollment outcomes. For example, college enrollment rates for high
school seniors and recent high school graduates who received this help rose
by about 25 percent—from 34 to 42 percent. Moreover, these gains persisted
over time: three years after the intervention, treatment group students were
8 percentage points more likely to have been enrolled in college for at least
two consecutive years (Bettinger et al. 2012).
The study’s findings helped spur many important policy changes. Most
notably, students and their families now have the option to pre-populate
Evaluation as a Tool for Improving Federal Programs  |  279

the FAFSA with the income information they have already provided the
Internal Revenue Service on their tax returns, similar to the arrangement
the researchers tested with H&R Block. This simplifies FAFSA completion
for students, lowers the risk of errors, and as such should increase access
to college among socioeconomically disadvantaged students and should
lead to gains in college enrollment. As a complement, the Department of
Education has also simplified the FAFSA application to make it easier to
complete for all applicants, but especially for low-income students. In 2012,
the Department of Education awarded a second grant to the research team
to test the effects of FAFSA simplification at scale. The evaluation will use
an experimental design and involve 9,000 tax-filing sites across the United
States.
Institutionalizing Evidence-Based Decision-making in Grant
Programs. In many programs, funds are distributed to states and local
entities through competitive and formula grants. Grants to State and local
governments have constituted roughly one-third of total outlays over the
last 20 years, so increasing the use of evidence in informing policy in grantsupported programs could improve outcomes for a significant portion of
outlays (Figure 7-1). Many effective program structures treat evaluation as
an essential element in the decision-making framework, while also building
in opportunities to scale up approaches that work and scale back or eliminate
approaches that do not. As stated by then-OMB Director Peter Orszag, new
initiatives should ideally have “evaluation standards built into their DNA”
(OMB 2009b). The Administration has experimented with several models
that embed both evidence building and evidence-based decisionmaking into
the “DNA” of grant programs.
During this Administration, several initiatives have adopted a “tiered
evidence” approach that embeds evidence-based decision making into program structure. Tiered evidence programs tie grant funding to the evidence
base behind proposed interventions. In these programs, interventions that
provide better evidence of success move to higher tiers and become eligible
to receive more funding for expanded implementation and evaluation. The
built-in mechanism for scaling up interventions that work helps prevent the
troubling problem of not investing in programs with proven high returns.
A successful example of a three-tier approach is the Investing in
Innovation program at the Department of Education. This program provides seed development grants of up to $3 million for high-potential and
relatively untested interventions, validation grants of up to $12 million for
interventions based on only a moderate amount of evidence, and scale-up
grants of up to $20 million for potentially high-impact, transformative
education interventions. Evidence of effectiveness is an “entry requirement”
280  |  Chapter 7

Figure 7-1
Outlays for Grants to State and Local Governments, 1992‒2012

Percent of total
50
45

FY 2012

40
35
30
25
20
15
10
5
0
1990

1995

2000
Fiscal Year

2005

2010

Note: Total excludes outlays for defense, interest and social security.
Source: Office of Management and Budget, FY 2014 Budget, Tables 3.1 and 12.2.

for validation and scale-up grants, and all grantees are expected to conduct
an evaluation that will add to the evidence base on effectiveness. For the
scale-up and validation grants, the grantee must make the data from their
evaluations available to third-party researchers, consistent with applicable
privacy requirements (Department of Education 2013a).
Similarly, the Maternal, Infant, and Early Childhood Home Visiting
Program in the Department of Health and Human Services (HHS) was an
early Administration initiative that uses a two-tiered evidence structure.
Implemented in 2010 as part of the Affordable Care Act, this voluntary home
visiting program uses trained professionals and paraprofessionals to provide
support to vulnerable pregnant women and parents of young children to
improve health, development, and well-being outcomes for at-risk children
and their families. The Act required that at least 75 percent of the home
visiting program funds be spent on proven, evidence-based approaches and
allowed for the remainder to be spent on promising approaches as long as
they are rigorously evaluated. Currently, 14 home visiting models meet the
HHS criteria of “evidence-based approaches,” and have been evaluated with
a mix of randomized experiments and quasi-experiments using multiple
measures of key outcomes (Paulsell et al. 2013). While the Act funded the
home visiting program through 2014, the Administration has proposed to
continue funding and expand the availability of voluntary evidence-based

Evaluation as a Tool for Improving Federal Programs  |  281

home visiting programs to reach additional families in need as part of a
continuum of early childhood interventions.
In addition to tiered evidence structures, agencies have begun using
other designs in competitive grant programs that encourage the use of
evidence-based practices. One such design is the “Pay for Success” approach.
In this performance-based model, philanthropic and private funding is leveraged and the government provides payment only after targeted outcomes
are achieved. In 2012 and 2013, the Administration started supporting
programs that use a Pay for Success model to fund preventive services,
and which had outcomes that could be measured with credible evaluation
methodologies. The first Pay for Success awards were for projects to prevent
prison recidivism.5 The Consolidated Appropriations Act of 2014 authorized up to $21.5 million for Pay for Success projects.
Even in more traditionally structured grant programs where funding is provided upfront, agencies are embedding more rigorous evaluation
requirements into funding requirements. For example, upfront grants in the
Department of Labor’s Workforce Innovation Fund, first issued in 2011,
fund promising but untested employment and training service and administrative strategies. These grants also fund well-tested ideas being adapted to
new contexts as a way to significantly increase evidence about interventions
that generate long-term improvements in public workforce system performance, such as reduced duration of unemployment. Grantees are required
to conduct rigorous evaluations, and a national evaluation coordinator
works with grantee evaluators to ensure consistent and high-quality evaluations (Department of Labor 2011).
Ending or Reducing Funding for Interventions or Programs. The
Administration’s commitment to evidence-based evaluation means terminating or reducing funding for a program when a body of evidence
consistently shows that the program is not achieving its stated goals, helping
to reduce the use of taxpayer dollars on ineffective programs. The FY 2012
Budget took this approach with the Mentoring Children of Prisoners (MCP)
program run by the Department of Health and Human Services. Rigorous
evaluations show that high-quality mentoring relationships lasting for at
least 12 months can have positive impacts on youth, while relationships
that do not last more than three months can actually have harmful effects
on youth (Grossman and Rhodes 2002). According to the MCP program
5 For example, the Department of Labor allocated nearly $24 million in Workforce Innovation
Fund grants to pilot Pay for Success grants to increase employment and reduce recidivism
among formerly incarcerated individuals (United States Interagency Council on Homelessness,
2013a). DOL required the grantees to employ rigorous evaluation methods in gauging impacts
on outcomes, which was defined in the grant solicitation as an experimental or credible quasiexperimental evaluation design.

282  |  Chapter 7

performance data, fewer than half of program participants each year were
in matches that lasted at least 12 months and, in 2008 alone, as many as 27
percent of matches that ended prematurely ended within three months. An
evaluation of one MCP-funded program suggested that premature terminations were the result of program performance and were independent of the
demographics of the participants (Schlafer et al. 2009).
Interpreting the MCP performance data in light of the evidence
from impact evaluations of other mentoring programs, the Administration
concluded that the MCP was not as effective as it should be. As a result, the
Administration proposed to reduce funding for the MCP, noting that other
competitive grant programs could serve the youth targeted by the MCP,
and that some of those programs, such as Promise Neighborhoods, utilize
evidence-based practices. Congress ultimately eliminated funding for the
program in the Continuing Appropriations Act of 2011.
Even Start, originally designed to improve family literacy in disadvantaged populations, was another program not meeting its stated goals that the
Administration took steps to replace. While the literacy levels of Even Start
children and parents improved, multiple national randomized experiments
showed that parents and children in control groups who did not participate
in Even Start (one-third of whom received other early childhood education
or adult education services) had comparable improvements (see for example
St. Pierre et al. 2003). The President’s FY 2012 Budget proposed, and
Congress approved, the elimination of separate funding for Even Start. The
Administration has proposed incorporating it and other narrowly focused
literacy programs into the newly created literacy component of the Effective
Teaching and Learning program that would support competitive grants to
states for high-quality, evidence-based literacy programs.

Building Evidence when Existing Evidence is Limited
In many of the examples highlighted above, evidence existed on what
programs or interventions were most effective, and the key challenge facing
policymakers was to act on that evidence. However, not enough is known
about what works in many other important areas, and so the first step in
evidence-based policymaking is to invest in developing evidence.
Reducing Electricity Use. Experts have long suggested time-varying
pricing (more costly at times of peak demand) as a way of increasing the
efficiency of electricity use, including reducing electricity demand. Such
time-varying pricing could increase efficiency, defer investments in expensive new power plants, and reduce pollution. However, most electricity
delivery systems have not invested in the in-home technologies necessary to
allow residential consumers to respond to time-varying prices. In addition,
Evaluation as a Tool for Improving Federal Programs  |  283

regulators have been hesitant to approve varying rates, and private companies have been reluctant to invest in modernizing their systems without
knowing whether time-varying pricing will significantly impact consumer
behavior. In recent years, the Federal Government, in partnership with
states and utilities, invested in evaluating the impact of time-varying pricing
on consumer behavior so that this information would be available to utilities,
regulators, and states. These consumer behavior studies were implemented
with American Recovery and Reinvestment Act funds and use randomized
controlled experimental methods. Deciding which type of pricing strategy to
use falls within State jurisdiction, rather than Federal, so these studies will
allow State and local public utilities to make more informed decisions on
pricing models (Cappers et al 2013). While these studies are still ongoing,
two utilities and their regulators have decided to implement time-varying
rates across their service territories based on the results observed to date.
Such efforts can serve as an impetus to get more public utilities to adopt
time-varying pricing.
Improving Health Care Delivery. In another example, the Affordable
Care Act made a number of major investments in understanding how
to improve quality and reduce cost in health care delivery, in addition to
expanding access to affordable health insurance coverage. As described earlier in this chapter, the Center for Medicare and Medicaid Innovation (the
“Innovation Center”), created by the Act, is using high-quality evaluation
approaches to test innovative payment and service delivery models designed
to reduce expenditures while preserving or enhancing quality of care for
Medicare, Medicaid and Children’s Health Insurance Program beneficiaries.
Several ongoing Innovation Center payment reform initiatives—and early
results from those initiatives—are discussed in Chapter 4. The Innovation
Center will use the results of such model evaluations and actuarial data
to identify best practices and determine which successful models could be
implemented more broadly.
Better Outcomes for Youth with Disabilities. The Administration
is also testing many different approaches aimed at youth with disabilities.
The Promoting Readiness of Minors in Supplemental Security Income
(PROMISE) is a joint initiative of the departments of Education, Health and
Human Services, Labor, and the Social Security Administration. PROMISE
aims to improve the education and employment outcomes for youth with disabilities who receive Supplemental Security Income (SSI) and their families,
by improving coordination of services such as those available through the
Individuals with Disabilities Education Act, the Vocational Rehabilitation
State Grants program, Medicaid health and home and community based services, Job Corps, Temporary Assistance for Needy Families, and Workforce
284  |  Chapter 7

Investment Act programs. The PROMISE program allows grantees (states or
consortia of states) to design their own intervention models to serve youth
and their families for three years with a two-year extension option, provided
they include a minimum set of services. Grantees may also apply for waivers
of funding restrictions or rules in individual programs that they believe will
constrain their ability to achieve outcomes. Grantees agree to enroll a large
number of youth (around 2,000) who are eligible to be served by a PROMISE
intervention, and to allow random assignment to be used to assign half of
eligible youth to the treatment group and the remaining youth to a control
group that receives the services that child SSI recipients normally receive.
The first grants were awarded in September 2013. To evaluate whether
PROMISE can help child SSI recipients achieve better outcomes, a national
evaluation will be conducted of all grantees to analyze intervention impacts
on educational attainment, employment credentials and outcomes, and
whether the interventions reduce long-term reliance on public benefits, and
SSI payments in particular (Social Security Administration 2013).	
Improving Outcomes For At-Risk Youth. The Administration also is
working to identify approaches that help at-risk youth. The National Guard
Youth Challenge (ChalleNGe) program, which has been rigorously evaluated, is designed to provide opportunities for adolescents who have dropped
out of school but demonstrate a willingness to turn their lives around.
Using random assignment, Millensky et al. (2011) found significant benefits
to program participation in addition to higher earnings, as ChalleNGe
graduates were more likely than the control group to have obtained a high
school diploma or GED, to have earned college credits, and to be working three years after completing the program. Participation was projected
to increase discounted lifetime earnings by over $40,000 (in 2010 dollars)
(Perez-Arce et al. 2012). After considering education costs to the student
and other non-earnings benefits, the ChalleNGe program was estimated to
generate $2.66 for every dollar of program cost (Perez-Arce et al. 2012). The
Administration now plans to test the application of the ChalleNGe model
to adjudicated youth, through the Department of Labor’s Reintegration of
Ex-Offenders program.
Reducing Homelessness. Sharply reducing homelessness is a key focus
of the Administration.6 Although once considered an intractable problem,
a broad body of research (including rigorous evaluations) documented that
6 Spurred in part by the Homeless Emergency Assistance and Rapid Transition to Housing
Act of 2009, the Obama Administration released Opening Doors: The Federal Strategic Plan
to End Homelessness in 2010. The plan establishes ambitious goals to end veterans’ and
chronic homelessness as well as homelessness among youth and families. The U.S. Interagency
Council on Homelessness serves to coordinate action by 19 member agencies (United States
Interagency Council on Homelessness, 2013b).

Evaluation as a Tool for Improving Federal Programs  |  285

there are models that effectively serve individuals experiencing chronic
homelessness. The Department of Housing and Urban Development (HUD)
has invested heavily in promoting these evidence-based approaches, and has
re-oriented the Homelessness Assistance Grant Program away from such
traditional approaches as transitional housing and toward more-effective
permanent supportive housing (Figure 7-2). Because research on interventions that are effective for homeless families does not yet exist at the same
level of rigor as for homeless individuals (Culhane et al. 2007), HUD has
undertaken an experimental study of family homelessness called the Family
Options Study. This study will compare several combinations of housing
assistance and services in a multi-site experiment to determine which interventions work best to promote housing stability, family preservation, child
well-being, adult well-being, and self-sufficiency. In addition to usual care,
defined as remaining in emergency shelter and accessing whatever resources
that would normally be available to families in shelter, three interventions
are being studied: 1) subsidy only (a voucher primarily), 2) transitional
housing, 3) rapid re-housing.7

Furthering the Evidence Agenda
Relative to when President Obama first took office in January 2009,
agencies are doing more to build actionable evidence to answer important
program and policy questions. These efforts span a wide range of agencies
and programs. While largely focused on improving the performance of
programs that provide direct services to individuals and account for roughly
65 percent of total Federal outlays (OMB 2014), many agencies, including
the Department of Commerce, the Small Business Administration, the
Department of Agriculture, and the Treasury Department are also pursuing
ways to incorporate impact evaluations in programs that provide assistance
to businesses.
Instilling a culture of evidence-based decision making within agencies, and building the foundations that enable rigorous evaluations to guide
new investments and drive policy, is neither quick nor easy. Evaluations of
particular interventions or entire programs should not be isolated exercises
that occur on an ad hoc basis, but rather planned in advance. Challenges
always accompany efforts to enact significant change, but addressing several
key elements can greatly facilitate agency efforts to improve the collection
and use of evidence. While not a comprehensive list, these issues represent
7A summary of the study, as well as the Interim Report (which documents the study design
process of randomization, and characteristics of the study population) can be found here:
http://www.huduser.org/portal/family_options_study.html

286  |  Chapter 7

Figure 7-2
Inventory of Beds for Homeless and Formerly
Homeless People, 2007‒2012

Number of Year-Round Beds (Thousands)
280
270

2012

Permanent Supportive Housing

260
250
240

Emergency Shelters

230
220
210

Transitional Housing

200
190
180
2007

2008

2009

2010

2011

2012

Source: Department of Housing and Urban Development, Homeless Data Exchange, Housing
Inventory Count.

several major areas that provide either useful opportunities, or serve as barriers, in agency efforts to build and advance an evidence-based agenda.

Legislative Support for Evaluation
Authorizing legislation and appropriations bills can direct how agencies should use program funds for a wide variety of activities. Legislation can
encourage stronger and more cost-effective evaluation in many ways. One
is through language that recognizes the importance of conducting rigorous
evaluations. Another is by making sure already-collected program data are
made available for such statistical and analytical purposes.
Legislative Support for Rigorous Evaluations. Two ways that legislation can support rigorous evaluation is through set-asides and support for
evaluation of demonstration programs. In recent years, with support of
top management within agencies and the Administration, several agencies have had set-asides for evaluation specified in program legislation and
appropriations. For example, the Consolidated Appropriations Act of 2012
first enabled the Secretary of Labor to reserve up to 0.5 percent of specific
Department of Labor (DOL) appropriations for evaluations. Also, a set-aside
of 5 percent of competitive grant funds in the Teacher Incentive Fund allows
the Department of Education to conduct a rigorous national evaluation
of the program and to share with grantees the results of current, rigorous
Evaluation as a Tool for Improving Federal Programs  |  287

research to help facilitate ongoing improvement. Additionally, authority to
set-aside a percentage of program funds for evaluation is specified in some
HHS and HUD programs, including several that received additional funding
through the Affordable Care Act of 2010 and the American Recovery and
Reinvestment Act of 2009.8
Many programs are funded through annual appropriations, which generally require obligation of those funds within a given fiscal year. However,
the DOL set-aside for evaluation in the Consolidated Appropriations Act
extends the deadline by which the DOL must obligate transferred evaluation funds to two years. Because designing rigorous evaluations takes time,
a window beyond the standard one-year for obligating evaluation funds can
in some cases enable agencies to plan and execute more thorough, higherquality evaluations.
Legislation that specifies funding for demonstration pilots also provides important support for developing an evidence base. The legislatively
authorized demonstrations being conducted by the Innovation Center are
recent examples that illustrate the value of legislative support for evidence
building.9 As another example, the Department of Health and Human
Services’ child welfare waiver authority allows states to design and demonstrate a wide range of approaches for reforming child welfare and improving
outcomes, including decreased first-time entries and re-entries into foster
care, and improvements in various aspects of child developmental, behavioral, and social functioning. States are required to conduct rigorous impact
evaluations as well as process evaluations as part of their waiver agreements.10 In addition, the Administration is proposing to restore demonstration authority for the Disability Insurance program, while also providing
new authority for the Social Security Administration and partner agencies
to test early-intervention strategies that would help people with disabilities
remain in the workforce.
Legislation can also encourage stronger evaluations through explicit
language requiring grantees to participate in evaluations and by requiring use of proven interventions. The Healthy, Hunger-Free Kids Act, for
8 While set-asides within programs are useful, some have noted that department-wide setasides may have advantages over program-level set-asides by providing agencies with more
flexibility over maximizing the return to evaluation investments. Also, set-asides will be used
most effectively when agencies have a demonstrated capacity to manage evaluation funds.
9 Prior to passage of the ACA, existing demonstration payment waiver authority allowed HHS
to conduct Medicare demonstrations of the impacts of new service delivery methods and new
payment approaches. However, due to statutory restrictions these demonstrations tended to
be relatively small. The ACA provided the Secretary with more flexible authority for testing
payment and delivery system innovations, and expanding them based on evidence. This work
is conducted under the auspices of the CMMI.
10 Child and Family Services Improvement and Innovation Act, Title II, Sec. 201, P.L. 112-34.

288  |  Chapter 7

example, included a nondiscretionary provision that requires State and local
grant recipients in a number of nutrition assistance programs, including
the National School Lunch Program, the Special Supplemental Nutrition
Program for Women, Infants and Children, and other programs authorized
in the National School Lunch Act and the Child Nutrition Act to cooperate
in evaluations conducted by or on behalf of the Department of Agriculture.11
This Act also reformed the structure of the nutrition education provided
through the Supplemental Nutrition Assistance Program, one of the
Nation’s main anchors of the social safety net that provides nutrition assistance to eligible low-income individuals and families. It established a new
and improved Nutrition Education and Obesity Prevention Grant Program
that requires a greater emphasis on evidence-based, outcome-driven interventions, with a focus on preventing obesity and coordinating with other
programs for maximum impact and cost-effectiveness.
Legislative Support for Access to Data for Statistical Purposes,
Including Evaluations. Existing laws can be explicit or implicit regarding
whether information collected as part of administering programs can be
used for statistical purposes integral to evaluation. Explicit and supportive
laws can save significant time and effort in negotiating agreements to provide data for evaluations and can facilitate more and better analysis. For
example, the Social Security Act explicitly states that one of the agency’s
datasets can be used for statistical and research activities conducted by
Federal and State agencies.
Some legislation provides the agency head with broad authority to
determine appropriate uses of program data. Given that the statistical uses
of data in program evaluation often inform the context, policies, and operations of the same programs authorized by a given statute, agencies sometimes determine that their general statutory authority can grant sufficient
authorization to provide administrative data to other Federal agencies for
statistical purposes. For example, the Social Security Administration provides certain datasets for statistical and research purposes as described in its
implementing regulations.
Multiple legitimate goals must be balanced when determining appropriate access to data, including reducing the burden of data collection on
individuals and institutions and protecting personal privacy. Even so, careful crafting of legislative language can achieve those aims while still making
data available for Federal researchers to rigorously evaluate and to improve
11 U.S. Department of Agriculture, Food and Nutrition Service Final Rule, Cooperation
in USDA Studies and Evaluations, and Full Use of Federal Funds in Nutrition Assistance
Programs Nondiscretionary Provisions of the Healthy, Hunger-Free Kids Act of 2010, Public
Law 111–296. Federal Register Vol 76, No. 125, June 29, 2011.

Evaluation as a Tool for Improving Federal Programs  |  289

government programs. Key considerations include: avoiding vague or
unclear authority to determine appropriate uses of program data; avoiding
narrowly written statutory language that only allows access to data for narrowly defined programmatic reasons; or restricting a Federal agency’s ability
to collect data from grantees.
The information needs in programs managed at the State level could
theoretically be addressed through non-Federal data  systems, but this is
not always possible. States or other grantees may not voluntarily develop
comprehensive data systems in ways that are comparable across states, or
have the capacity or incentive to make data available to researchers. When
no feasible solutions exist to alleviate these issues, legislation may be warranted to authorize creation of Federal datasets accessible to researchers, or
to establish requirements for State-held datasets that enable data exchange
and comparability across states and to ensure access by researchers.

Building Evaluation into the Design of Programs
Many of the examples described earlier demonstrate the ways in
which agencies are designing programs to facilitate evaluation. But agencies can still do more to embed rigorous evaluation designs into both new
programs and existing programs.
New programs. The benefits of adopting evidence-based program
designs, like the tiered evidence structure in the Investing in Innovation
program and in HHS’ home visiting program, include the ability to guide
competitive grant funds to the strategies with a strong evidence base, while
also requiring grantees to conduct evaluations where no evidence is yet
available. Even without such a program structure, agencies implementing
new programs over the past five years have increasingly required grantees
to collect data and develop administrative data systems that can improve
comparability and facilitate evaluation in addition to meeting program
operating needs. For example, the Department of Education’s Promise
Neighborhoods initiative implemented in 2010 requires grantees to collect
and track outcome data in an individual-level longitudinal data system to
facilitate rigorous evaluation. This initiative aims to improve educational
and developmental outcomes of children and youth in distressed communities.12 To assist grantees in collecting high-quality and comparable data,
the Education Department is providing grantees with extensive guidance on
data collection and reporting (Comey et al. 2013).
12 This program is based on the Harlem Children’s Zone model, which was found to increase
earnings for students, decrease the probability of committing crimes and decrease health
disability probabilities—with the potential for providing large public benefits (Dobbie and
Fryer 2011).

290  |  Chapter 7

Other benefits of considering evaluation needs in the design of new
programs include creating opportunities to save time and money by identifying evaluation data up front, minimizing burden on program respondents,
and avoiding the loss of information all together that cannot be created too
long after the fact. When not considered in the earliest stages of program
design, a typical alternative to collecting the information needed for evaluation is to conduct surveys, which requires identifying expertise to design and
test the surveys, gaining approval for their use, and then administering them
to collect data, often long after the fact. Surveys add to the time and cost
to build evidence, due to the time and skill involved in developing survey
instruments that will yield high-quality data, the requirements for obtaining
needed approvals, and the actual implementation of the survey.13 Careful
planning can help limit the need for evaluation-related surveys to data that
cannot be obtained in any other way, such as information on post-program
choices, earnings, or jobs necessary for identifying longer-term impacts of a
program or intervention.
One of the most important ways the design of a program can facilitate evidence building is through careful consideration of how treatment
and control groups can be established to facilitate impact evaluation. As
discussed earlier, randomly assigning potential program participants to
treatment and control groups enables the most credible impact evaluations.
Several mechanisms exist for creating good comparison groups that allow
for experimental or quasi-experimental techniques to be employed to produce high-quality estimates of program effectiveness.
Several options for enrolling potential participants in a program or
intervention, presented in order of the rigor of evaluation they might support, are as follows:
1. Random assignment by lottery when capacity is limited. In many
instances, due to limited funds or other constraints, a program or intervention cannot serve every person or entity that is eligible to apply. In such
cases, rather than “first-come, first-serve” or other nonrandom devices,
implementing a lottery to select which applicants may participate in a
program or intervention generates a low-cost randomized experiment. This
13 For example, the Paperwork Reduction Act (PRA), first enacted in 1980 and amended
in 1995 (44 U.S.C., Chapter 35), requires Federal agencies to obtain OMB approval when
an agency plans to collect information from ten or more persons using identical reporting,
recordkeeping, or disclosure requirements. Among the PRA’s goals are ensuring the greatest
possible public benefit from and maximizing the utility of information created, collected,
maintained, used, shared and disseminated by or for the Federal government and minimizing
the burden for persons resulting from the collection of information by or for the Federal
government. As a further example, some data collections are subject to review by Institutional
Review Boards, in order to protect the rights of the human subjects of such research, a
requirement under (42 USC 289) under 45 CFR 46.

Evaluation as a Tool for Improving Federal Programs  |  291

strategy has been used recently to determine the impact of Medicaid access
(Baicker et al. 2013), charter school attendance (Abdulkadiroglu et al. 2011),
and small business entrepreneurship training (Benus et al. 2009). Note that
the losers of the lottery need to be followed to track their outcome data.
2. Assignment based on a continuous “need score.” A common objection to random assignment is that resources should be targeted to those with
the greatest need, or those most likely to benefit. In this situation, program
assignment might lend itself to a strong evaluation if it incorporates some
sort of explicit, continuous, ranking of applicant need (or likely benefit),
and bases program eligibility on some cutoff in need. For example, Ludwig
and Miller (2007) study the effect of participating in Head Start on mortality rates for children by exploiting the fact that the Office of Economic
Opportunity provided technical assistance to the 300 poorest counties in
1965. This created lasting differences in Head Start funding rates for counties with poverty levels just below and just above the poverty rate of the
300th poorest county. With this type of assignment rule, a regression discontinuity design can be used to study the impact of the program. The logic
of the design is that individuals with “scores” just above and just below the
threshold—in Ludwig and Miller, living in a county with a poverty rate just
above or below the poverty rate of the 300th poorest county’s rate—are likely
to be similar to each other in ways that affect their outcomes, except that
those just below receive the treatment (in this case, participation in Head
Start). This design can deliver estimates of the effect of the program that are
similar to randomized experiments.14
3. Staging the rollout of a large program. If a program will be introduced that will ultimately serve many participants spread across different
geographic areas, or schools, or other natural groupings, staggering the
rollout across time and space, with the rollout sequence chosen randomly,
makes it much easier to evaluate. For example, suppose a mentoring program aimed at increasing college attendance will be introduced in a group of
schools and the government hopes to learn about the effect of the program
by estimating the change in college enrollment among students at the school
before and after the program is introduced. If the program is introduced
in only one school district, then any other changes that the school district
introduced around the same time might affect the change in outcomes and
bias the conclusion. Similarly, if the program is introduced in many different schools but all in the same year, then any other changes in policy, the
economic climate, or other macro-economic conditions may be confounded
14 On the other hand, the estimates from an RDD strictly pertain only to the types of
participants “near” the cutoff. To the extent the impact varies across participants with different
levels of need, this can be a limitation.

292  |  Chapter 7

with the treatment effect and thus may “bias” the estimate of the treatment
effect. Staggering the rollout of the program over time and space, using randomization and possibly further matching treatment and control units based
on observable characteristics, helps to control for these potential biases, and
thus allows for better estimates of a program’s impact. This strategy has been
used by Rothstein (2010) to study the effect of extended unemployment
benefits.
The three strategies above create experiments or quasi-experiments
that lend themselves to high-quality impact evaluations. In the absence of
such devices, evaluators need to acknowledge the differences that do exist
between program (or intervention) participants and non-participants and to
use statistical techniques like multivariate regression and matching to control for these differences. Since these strategies all attempt to compare the
outcomes of program participants and non-participants with similar characteristics, the success of the evaluation will be determined by the availability
of good information on the characteristics of the population that are most
predictive of the outcome under study, as well as the reasons why individuals
choose to participate in a program. However, for these strategies to work, the
variation between the treatment and control groups after using statistical or
matching techniques to control for differences between these groups must
be plausibly unrelated to the outcomes of interest, except through the effect
of the treatment. There needs to be some part of that variation between the
treatment and control groups that operates like randomization.
Existing Programs. Designing programs to facilitate evaluation may
be relatively simpler in new programs than in existing programs, due to
program manager reluctance in the latter to trying new strategies, concerns
about equity among participants if the control group receives no services,
and other reasons. But experiences at several agencies demonstrate these
barriers can be overcome. Lotteries for oversubscribed programs are as
applicable in longstanding programs as in new ones (see for example the
Jacob and Ludwig (2012) study on impacts of housing vouchers). However,
increasing opportunities for evidence-based decisionmaking in programs
that allocate funds to states on the basis of formulas remain especially challenging, because evaluations and evidence-based funding allocations are
not a requirement of States receiving the funds. Waiver authorities or other
mechanisms to incentivize evaluations in these programs are only available
in a few instances. A control that could prompt State and local grant recipients to do evaluations in these types of programs is a legislated requirement
that a certain portion of funds be set aside for evidence-based grants or
models of delivery. For example, in the Senate Appropriations Bill for FY
2014, the Substance Abuse and Mental Health Services Administration
Evaluation as a Tool for Improving Federal Programs  |  293

mental health block grant programs included language defining a 5 percent
set-aside for evidence-based grants.15
There is still work to be done to embed evaluation and evidence-based
decision making into more programs. Agencies can focus evaluation efforts
in those programs that can help ensure that the agency’s most critical program and policy questions are addressed.16

Developing the Capacity to Link to Other Administrative and
Survey Data Sources
Increasingly, agencies are seeking opportunities to improve their
evaluation approaches by supplementing their administrative program data
with other available government data, where appropriate and while ensuring strong privacy protections. Using pre-existing data collected for other
reasons, while maintaining strong privacy protections, provides a number
of benefits. Several challenges arise when doing so, and the Administration
is taking steps to address these challenges.
Benefits of using existing data resources. Using pre-existing administrative data collected for other reasons, while maintaining strong privacy
protections, can help agencies answer important policy questions that could
not otherwise easily be addressed with a single program database or survey.
Administrative data provides the most complete and accurate source of
information on program participation and can provide more accurate data
on earnings, test scores, and other outcomes of interest. Indeed, the benefits
of using pre-existing administrative data for evaluation and other statistical
purposes have been widely acknowledged for some time. Data from multiple
sources have been used in a number of impact evaluations, primarily to
identify the characteristics of treatment and control groups, identify outcome variables which indicate the impacts of treatments, reduce study costs
and reduce the burden on study participants by avoiding the need to collect
the data via another survey (Coalition for Evidence-Based Policy 2012;
Finkelstein et al. 2012; Bettinger et al. 2009; Jacob and Ludwig 2012). Linked
datasets are also facilitating current evidence-building efforts in various
agencies, such as in the Department of Health and Human Service’s Office
of Child Support Enforcement, which is currently implementing a child
support-led employment services demonstration project with a random
assignment impact evaluation (where treatment consists of extra services
under the program, and the control group receives regular services that are
available) and a cost-benefit analysis. The planned evaluation will draw on
15 S. 1284, Report No. 113–71.
16 Recognizing that agencies operate with scarce resources, OMB has encouraged agencies to
adopt such a focus (OMB 2013b).

294  |  Chapter 7

unemployment insurance wage and benefit records, as well as State administrative data on benefits in the Supplemental Nutrition Assistance Program
and other public assistance programs, criminal justice system data and other
data to more cost-effectively and accurately determine the effectiveness and
the true costs and benefits of the program. As another example, HUD and
HHS are pilot-testing links between HUD administrative data and HHS
Medicare/Medicaid data, to build evidence on opportunities to improve the
health of Medicare/Medicaid beneficiaries in HUD-assisted housing as well
as the impact of housing assistance on health.
Challenges and Solutions. Nevertheless, accessing administrative
data for these statistical uses is challenging. These data are collected to
facilitate day-to-day program operations, including developing performance
measures. Unless evaluation needs are considered in the database design
stage, however, the meaningfulness of administrative data for conducting
rigorous evaluation may be limited. Also, data definitions can vary dramatically across datasets; especially with State-level data, the definitions often
vary across states and even counties. Aside from definitional differences,
the quality of programmatic data—its completeness and accuracy—can
vary dramatically across datasets. Significant data-quality gaps or errors can
compromise analysis. It can also be costly to negotiate access to data on a
state-by-state basis.
One key practical challenge is that agencies, in an attempt to be
privacy sensitive, may not include in program databases unique identifiers
for program applicants and participants. Such unique identifiers facilitate
linking to data provided by subjects through other programs or even for
the same program over time. Linking datasets through name and address
matching or matching on other less unique variables can introduce bias and
render the linked data unusable for rigorous analysis. While some agencies
have an established history of allowing use of data (including identifying
information) for statistical purposes, in many cases access to such data is
not readily available due to real or perceived legal, policy, or operational
barriers.17 In some cases, extensive negotiations with the agency responsible
for the data are needed to gain access to the data for use in evaluation studies; sometimes the efforts are not successful even after months or years of
negotiations.
17 One legal barrier is that when a program’s authorizing statute is silent about whether access
to data can be provided for statistical purposes (which includes evaluation), agencies need
to make a determination about allowable uses. In such cases, agencies may conservatively
interpret the lack of an express authority as a prohibition on providing data to another agency.
However, as discussed in OMB memorandum M-14-06, agencies may be able to provide the
data under their general statutory authority (OMB 2014).

Evaluation as a Tool for Improving Federal Programs  |  295

To help address these barriers, the OMB recently issued guidance
to assist both program and statistical agencies (and statistical components
within agencies) in increasing the opportunities to use administrative data
for statistical purposes, which includes evaluation.18 In part, this guidance
requires government departments to engage both program and statistical
agencies in identifying administrative datasets of potential value for statistical purposes; communicating the importance to staff of promoting the use of
administrative data for statistical purposes; and identifying several datasets
with the most value for statistical purposes but which are not currently being
provided, along with descriptions of critical barriers that appear to preclude
providing access for statistical purposes. The guidance also offers tools to
help agencies in these tasks, including guidance in understanding relevant
legal requirements, a tool to facilitate more efficient interagency agreements,
and a tool to assess administrative data quality developed under the auspices
of the Federal Committee on Statistical Methodology. Departments must
also report to the OMB on their efforts to foster collaboration and increase
access to administrative data for statistical purposes.

Facilitating Researcher Access to Federal Data while Protecting
Privacy
Some agencies have developed ways for researchers to access Federal
data for statistical purposes in secure research environments that preserve
the confidentiality of individual records. The Census Bureau and National
Center for Health Statistics operate secure research data centers, in which
qualified researchers with approved projects can use micro-data files for
statistical research. The Retirement Research Consortium is a key tool that
the Social Security Administration uses to facilitate policy-relevant research
on retirement and Social Security. The consortium comprises three competitively selected research centers based at the University of Michigan,
Boston College, and the National Bureau of Economic Research. The centers
perform valuable research and evaluation of retirement policy, disseminate
results, provide training and education awards, and facilitate the use of
SSA’s administrative data by outside researchers. Nonetheless, due to confidentiality restrictions, uneven interpretations of laws governing privacy of
data provided to the government, and other reasons, many data sets remain
18 Statistical purposes is defined in footnote 2 of the OMB memorandum M-14-06 (OMB
2014): [it] refers to “the description, estimation, or analysis of the characteristics of groups,
without identifying the individuals or organizations that comprise such groups,” (PL-107-347,
Title V—Confidential Information Protection and Statistical Efficiency Act (CIPSEA), Section
502 (9)(A)). Statistical purposes exclude “any administrative, regulatory, law enforcement,
adjudicatory, or other purpose that affects the rights, privileges, or benefits of a particular
identifiable respondent” (PL-107-347, Title V—CIPSEA, Section 502 (5)(A)).

296  |  Chapter 7

hard for researchers to access for statistical uses, and opportunities to link to
researcher-collected data remain limited.
This Administration is committed to improving opportunities for
researcher access in ways that fully maintain privacy protections of Federal
program participants. HHS’ Centers for Medicare and Medicaid Services’
(CMS) Virtual Research Data Center is an innovative example of ways agencies are working to improve access to Federal agencies for their own use and
for their grantees carrying out federally sponsored research activities. In
late 2013, the Virtual Research Data Center began providing users with a
dedicated workspace where they can upload external files and use them with
CMS data to run analyses and download aggregate statistical files to their
workstations. This model is a more-efficient, less-expensive, more-flexible
and more-secure way for researchers to access a variety of Medicare and
Medicaid program data, relative to the existing approach that entails cutting,
encrypting, and shipping large quantities of information.

Conclusion
Whatever the findings, rigorous evaluations provide critical and credible feedback about whether the current design of a program is effective
or whether program modifications are needed so that important program
goals are met. Indeed, in some fields—including business and medicine—the
vast majority of randomized controlled trials used to evaluate the efficacy
of interventions and strategies find no positive effects of interventions
(Coalition for Evidence-Based Policy 2013). Rigorous impact evaluations
serve as important learning tools to guide management decisions about
program investments. The Administration continues to support the use
of these tools, broadly and often, to facilitate continuous improvement in
government programs as well as to identify best practices and effective new
approaches that can be shared with organizations delivering services funded
with Federal dollars.
Over the last five years, Federal agencies have increasingly used rigorous impact evaluations to inform program decisions, including how to
improve programs. Agencies are trying new approaches when the evidence
indicates existing strategies are not yielding sufficiently positive impacts
on important outcomes. They are restructuring programs to increase their
effectiveness when evidence shows new strategies produce better results, and
are developing evidence where an insufficient evidence base exists. And they
are scaling up approaches that work, improving public policy and people’s
lives. As part of this effort, agencies are improving the collection and comparability of data to provide new opportunities for evaluation. They are

Evaluation as a Tool for Improving Federal Programs  |  297

also using cutting-edge technology to improve data access to other Federal
agencies and to outside researchers while protecting privacy—strategies
that can enable evaluations to be done more rapidly and at lower cost. The
Administration continues to support these efforts to affect change. By using
rigorous evaluation strategies to identify what works, and by taking steps to
make needed modifications, agencies and taxpayers will have the greatest
confidence that scarce resources are being used as efficiently as possible in
meeting priority goals.

298  |  Chapter 7

REFERENCES
Chapter 1
Almunia, Miguel, Agustin Benetrix, Barry Eichengreen, Kevin H. O’Rourke
and Gisela Rua. 2010. “From Great Depression to Great Credit
Crisis: similarities, differences and lessons.” Economic Policy 25, no.
62: 219-265.
Bernanke, Ben S. 2012. “Monetary Policy since the Onset of the Crisis.”
Speech at the Federal Reserve Bank of Kansas City Economic
Symposium. (http://www.federalreserve.gov/newsevents/speech/
bernanke20120831a.htm).
____. 2014. “The Federal Reserve: Looking Back, Looking Forward.”
Speech at the Annual Meeting of the American Economic Association.
(http://www.federalreserve.gov/newsevents/speech/
bernanke20140103a.htm).
(CBO) Congressional Budget Office. 2013a. “Automatic Reductions in
Government Spending—aka Sequestration.” (http://www.cbo.gov/
publication/43961).
_____. 2013b. “The Economic Impact of S. 744, the Border Security,
Economic Opportunity, and Immigration Modernization Act.”
(http://www.cbo.gov/publication/44346).
Greenspan, Alan. 2013. The Map and the Territory: Risk, Human Nature,
and the Future of Forecasting. The Penguin Press.
Piketty, Thomas and Emmanuel Saez. 2003. “Income Inequality in the
United States 1913–1998.” Quarterly Journal of Economics 118, no.
1: 1–39.

299

_____. 2013. Data update to “Income Inequality in the United States 1913–
1998.” September (http://elsa.berkeley.edu/~saez/TabFig2012prel.
xls).
Reinhart, Carmen M., and Kenneth S. Rogoff. Forthcoming. “Recovery from
Financial Crises: Evidence from 100 Episodes.” American Economic
Review.
Romer, Christina D. 2009. “Back from the Brink.” Speech at the Federal
Reserve Bank of Chicago. (http://www.whitehouse.gov/assets/documents/Back_from_the_Brink2.pdf).
Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and Jane
Waldfogel. 2013.
“Trends in Poverty with an Anchored Supplemental Poverty Measure.”
Working Paper. New York: Columbia Population Research Center.
(http://socialwork.columbia.edu/sites/default/files/file_manager/
pdfs/News/Anchored%20SPM.December7.pdf).

Chapter 2
Ball, Laurence, and Sandeep Mazumder. 2011. “Inflation Dynamics and the
Great Recession.” Brookings Papers on Economic Activity 42, no. 1:
337-405.
Bank of Japan, Financial Markets Department. 2013. Outline of Outright
Purchases of Japanese Government Bonds. Financial Market Department 1-3. April.
Blanchard, Oliver J., and Peter A. Diamond. 1994. “Ranking, Unemployment Duration, and Wages.” Review of Economic Studies 61, no. 3:
417-34.
Board of Governors of the Federal Reserve System. 2009. “FRB/US Equation
Documentation.”
____.

2012. “Press Release: December 12, 2012.” (http://www.federalreserve.
gov/newsevents/press/monetary/20121212a.htm).

. 2013a. “The October 2013 Senior Loan Officer Opinion Survey on
Banking Lending Practices.”

____

____.

2013b. “Press Release: June 19, 2013.” (http://www.federalreserve.gov/
newsevents/press/monetary/20130619a.htm).

____.

2013c. “Press Release: December 18, 2013.” (http://www.federalreserve.
gov/newsevents/press/monetary/20131218a.htm).

300 |  References

____. 2014. “Press Release: January 29, 2014.” (http://www.federalreserve.
gov/newsevents/press/monetary/20140129a.htm).
(BEA) Bureau of Economic Analysis. 2013. “Comprehensive Revision: 1929
Through First Quarter 2013.” U.S. Department of Commerce.
____. 2014. “Technical Note: Gross Domestic Product Fourth Quarter of
2013 (Advance Estimate).” U.S. Department of Commerce. January.
(CBO) Congressional Budget Office. 2012. “Economic Effects of Policies
Contributing to Fiscal Tightening in 2013.” November.
____.

2013a. “Automatic Reductions in Government Spending aka Sequestration.” February.

____.

2013b. “The Economic Impact of S. 744, the Border Security, Economic
Opportunity, and Immigration Modernization Act.” June.

(CEA) Council of Economic Advisers. 1997. Economic Report of the President. February.
____.

2013. “Comprehensive GDP Revision and Advance Estimate for the
Second Quarter of 2013.” July (http://www.whitehouse.gov/
blog/2013/07/31/comprehensive-gdp-revision-and-advance-estimate-second-quarter-2013).

Database of State Incentives for Renewables and Efficiency. 2013. “Renewable Electricity Production Tax Credit (PTC).” October (http://
dsireusa.org/incentives/incentive.cfm?Incentive_Code=US13F).
(DOT) Department of the Treasury. 2012. “Description of the Extraordinary
Measures.”
December
(http://web.archive.org/
web/20131019163723/http:/www.treasury.gov/connect/blog/Documents/Sec%20Geithner%20LETTER%2012-26-2012%20Debt%20
Limit.pdf).
(EIA) Energy Information Administration. 2013. “Industrial Sector Natural
Gas Use Rising.” June (http://www.eia.gov/todayinenergy/detail.
cfm?id=11771).
____.

2013b. “International Energy Outlook.” April (http://www.eia.gov/forecasts/ieo/pdf/ieorefngtab_1.pdf).

____.

2013c. “Oil and Gas Industry Employment Growing Much Faster than
Total Private Sector Employment.” August (http://www.eia.gov/
todayinenergy/detail.cfm?id=12451).

(EPA) Environmental Protection Agency. 2011. “Paving the Way Toward
Cleaner, More Efficient Trucks.” August.

References  | 301

European Central Bank. 2012. “Press Release: Technical Features of Outright
Monetary Transactions.” September (http://www.ecb.europa.eu/
press/pr/date/2012/html/pr120906_1.en.html).
. 2013. “Monetary Policy.” (http://www.ecb.europa.eu/mopo/html/index.
en.html).

____

European Commission. 2013. “European Economic Forecast: Autumn
2013.” November.
Executive Office of the President. 2013. “The President’s Climate Action
Plan.” The White House. June.
Fuhrer, Jeffrey C., and Giovanni P. Olivei. 2010. “The Role of Expectations
and Output in the Inflation Process: An Empirical Assessment.”
Federal Reserve Bank of Boston Public Policy Brief, no. 10-2: 1-39.
Gordon, Robert. 1990. “U.S. Inflation, Labor’s Share, and the Natural Rate of
Unemployment.” Working Papers 2585. Cambridge, MA: National
Bureau of Economic Research.
____.

2013. “The Phillips Curve is Alive and Well: Inflation and the NAIRU
During the Slow Recovery.” Working Paper 19390. Cambridge,
MA: National Bureau of Economic Research.

Layard, Richard, Stephen Nickells, and Richard Jackman. 1991. Unemployment: Macroeconomic Performance and the Labor Market. New
York: Oxford University Press.
Krueger, Alan B., and Andreas I. Mueller. 2011. “Job Search and Job
Finding in a Period of Mass Unemployment: Evidence from HighFrequency Longitudinal Data.” Working Papers 1295. Princeton,
NJ: Center for Economic Policy Studies.
Nalewaik, Jeremy J. 2010. “The Income- and Expenditure-Side Measures of
U.S. Output Growth.” Brookings Papers on Economic Activity 41,
no. 1: 71-127.
(OMB) Office of Management and Budget. 2013. “Impacts and Costs
of the Government Shutdown.” (http://www.whitehouse.gov/
blog/2013/11/07/impacts-and-costs-government-shutdown).
Office of the Press Secretary. 2012. “Obama Administration Finalizes Historic
54.5 MPG Fuel Efficiency Standards.” The White House. (http://
www.whitehouse.gov/the-press-office/2012/08/28/obama-administration-finalizes-historic-545-mpg-fuel-efficiency-standard).
(USTR) Office of the United States Trade Representative. 2013a. “Acting
Deputy U.S. Trade Representative Endy Cutler Discussed Japan
302 |  References

and the TPP at the Peterson Institute for International Economics.”
(http://www.ustr.gov/about-us/press-office/blog/2013/November/
Cutler-TPP-Japan-PIIE).
2013b. “Ambassador Froman Discusses the Transatlantic Trade and
Investment Partnership at the Munich Security Conference.”
(http://www.ustr.gov/about-us/press-office/blog/2013/November/
Froman-Munich-Security-Conference).

____.

2013c. “Statement of the Ministers and Heads of Delegation for the
Trans-Pacific Partnership Countries.” (http://www.ustr.gov/tpp).

____.

2013d. “Weekly Trade Spotlight: The Benefits of the WTO Trade Facilitation Agreement to Small Business.” (http://www.ustr.gov/about-us/
press-office/blog/2013/December/Benefits-of-WTO-Trade-Facilitation-Agreement-to-Small-Business).

____.

2013e. “White House Fact Sheet: Transatlantic Trade and Investment
Partnership (T-TIP).” (http://www.ustr.gov/about-us/press-office/
fact-sheets/2013/june/wh-ttip).

____.

(OECD) Organization for Economic Cooperation and Development. 2013.
“General Assessment of the Macroeconomic Situation.” OECD
Economic Outlook, no. 2: 9-66.
Polk. 2013. “Polk Finds Average Age of Light Vehicles Continues to Rise.”
(https://www.polk.com/company/news/polk_finds_average_age_
of_light_vehicles_continues_to_rise). 	
Stock, James H. 2011. “Discussion of Ball and Mazumder, ‘Inflation
Dynamics and the Great Recession.” Brookings Papers on Economic
Activity 42, no. 1: 387-402.
Stock, James H., and Mark W. Watson. 2010. “Modeling Inflation After the
Crisis.” Working Papers 16488. Cambridge, MA: National Bureau
of Economic Research.
Zindler, Ethan. 2013. “Burst of Construction in December Delivers Record
Year for US Wind.” Bloomberg New Energy Finance. January 18.

Chapter 3
Aldy, Joseph E. 2013. “A Preliminary Assessment of the American Recovery
and Reinvestment Act’s Clean Energy Package.” Review of Environmental Economics and Policy 7(1): 136-155.
Alesina, Alberto,and Silvia Ardagna. 2010. “Large Changes in Fiscal Policy:
Taxes versus Spending.” Tax Policy and the Economy 24: 35-68.
References  | 303

Angrist, Joshua D., and Jörn-Steffen Pischke. 2010. “The Credibility Revolution in Empirical Economics: How Better Research Design Is
Taking the Con out of Econometrics.” Journal of Economic Perspectives 24, no. 2: 3-30.
Auerbach, Alan J., and Daniel Feenberg. 2000. “The Significance of Federal
Taxes as Automatic Stabilizers.” Journal of Economic Perspectives
14(3): 37-56.
Auerbach, Alan, and Yuriy Gorodnichenko. 2012. “Measuring the Output
Responses to Fiscal Policy.” American Economic Journal: Economic
Policy 4(2): 1-27.
Barro, Robert J. 1974. “Are Government Bonds Net Wealth?” Journal of
Political Economy 82(6): 1095-1117.
Ball, Laurence M. 2009. “Hysteresis in Unemployment: Old and New
Evidence.” Working Paper 14818. Cambridge, MA: National Bureau
of Economic Research.
Bastiat, Frédéric.  1848. “Selected Essays on Political Economy.” Seymour
Cain, trans. 1995. Library of Economics and Liberty. 3 February
2014.
Blanchard, Olivier, Giovanni Dell’Ariccia, and Paolo Mauro. 2010.
“Rethinking Macroeconomic Policy.” IMF Staff Position Note.
Washington: International Monetary Fund.
Blanchard, Olivier, and Daniel Leigh. 2013. “Growth Forecast Errors and
Fiscal Multipliers.” IMF Working Paper.
Blanchard, Olivier, and Roberto, Perotti. 2002. “An Empirical Characterization of the Dynamic Effects of Changes in Government Spending
and Taxes on Output.” Quarterly Journal of Economics.
Blanchard, Olivier, and Lawrence H. Summers. 1986. “Hysteresis and
the European Unemployment Problem.” NBER Macroeconomics
Annual, 1, 15-90. Cambridge, MA: National Bureau of Economic
Research.
Blinder, Alan S., and Mark Zandi. 2010. “How the Great Recession was
Brought to an End.” Princeton University and Moody’s Analytics.
Chetty, Raj. 2008. “Moral Hazard vs Liquidity and Optimal Unemployment Insurance.” Working Paper 13967. Cambridge, MA: National
Bureau of Economic Research.
Chodorow-Reich, Gabriel, Laura Feiveson, Zachary Liscow, and William
Gui Woolston. 2012. “Does State Fiscal Relief During Recessions
304 |  References

Increase Employment? Evidence from the American Recovery and
Reinvestment Act.” American Economic Journal: Economic Policy 4,
no. 3: 118-145.
Christiano, Lawrence, Martin Eichenbaum, and Sergio Rebelo. 2011. “When
Is the Government Spending Multiplier Too Large?” Journal of
Political Economy 119(1): 78-121.
Coenen, Gunter, et al. 2012. “Effects of Fiscal Stimulus in Structural Models.”
American Economic Journal: Macroeconomics 4(1): 22-68.
Cogan, John F., et al. 2010 “New Keynesian versus old Keynesian government spending multipliers.” Journal of Economic Dynamics and
Control 34(3): 281-295.
(CBO) Congressional Budget Office. 2009a. “Health Information Technology for Economic and Clinical Health Act.”
___. 2009b. “Cost Estimate: H.R. 1, American Recovery and Reinvestment
Act of 2009.”
___. 2009c. “Cost Estimate: H.R. 3548, Worker, Homeownership, and Business Assistance Act of 2009.”
___. 2010a. “The Budget and Economic Outlook: Fiscal Years 2010 to 2020.”
___. 2010b. “The Budget and Economic Outlook: An Update.”
___. 2010c. “Cost Estimate: H.R. 4691, the Temporary Extension Act of
2010, As Introduced on February 25, 2010.”
___. 2010d. “Cost Estimate: Budgetary Effects of Hiring Incentives to
Restore Employment Act, as Introduced by Senator Reid on
February 11, 2010.”
___. 2010e. “Cost Estimate: Amendment No. 3721 to H.R. 4851, the
Continuing Extension Act, 2010, as Proposed by Senator Baucus.”
___. 2010f. “Cost Estimate: Budgetary Effects of Senate Amendment 4425,
the Unemployment Compensation Extension Act of 2010.”
___. 2010g. “Cost Estimate: CBO Estimate of Changes in Revenues and
Direct Spending for Senate Amendment 4594 in the Nature of a
Substitute to H.R. 5297, the Small Business Jobs and Credit Act of
2010.”
___. 2010h. “Cost Estimate: CBO Estimate of Changes in Revenues and
Direct Spending for S.A. 4753, an amendment to H.R. 4852, the
Tax Relief, Unemployment Insurance Reauthorization, and Job
Creation Act of 2010.”

References  | 305

___. 2010i. “Cost Estimate: Budgetary Effects of Senate Amendment
4575, containing proposals related to education, state fiscal relief,
the Supplemental Nutrition Assistance Program, rescissions, and
revenue offsets.”
___. 2011a. “The Budget and Economic Outlook: Fiscal Years 2011 to 2021.”
___. 2011b. “Cost Estimate: Budgetary Effects of Senate Amendment 927 to
H.R. 674, as proposed by Senator Reid for Senator Tester.”
___. 2011c. “Cost Estimate: H.R. 5297, the Small Business Jobs Act of 2010.”
___. 2011d. “Cost Estimate: Budgetary Effects of the Temporary Payroll
Tax Cut Continuation Act of 2011, as Posted on the Website of the
House Committee on Rules on December 22, 2011.”
___. 2012a. “The Budget and Economic Outlook: Fiscal Years 2012 to 2022.”
___. 2012b. “Cost Estimate: Budgetary effects of the Conference Agreement
for H.R. 3630, the Middle Class Tax Relief and Job Creation Act of
2012, as Posted on the Web Site of the House Committee on Rules
on February 16, 2012.”
___. 2012c. “Cost Estimate: Budgetary effects of the Conference Agreement
for H.R. 3630, the Middle Class Tax Relief and Job Creation Act of
2012, as Posted on the Web Site of the House Committee on Rules
on February 16, 2012.”
___. 2013a. “Cost Estimate: Estimate of Budgetary Effects of H.R. 8, the
American Taxpayer Relief Act of 2012, as passed by the Senate on
January 1, 2013.”
___. 2014a. “The Budget and Economic Outlook: 2014 to 2024.”
___. 2014b. “Estimated Impact of the American Recovery and Reinvestment
Act on Employment and Economic Output in 2013.”
Conley, Timothy, and Bill Dupor. 2013. “The American Recovery and
Reinvestment Act: Public Sector Jobs Saved, Private Sector Jobs
Forestalled.” Journal of Monetary Economics.
(CEA) Council of Economic Advisers. 2009a. “Estimates of Job Creation
from the American Recovery and Reinvestment Act of 2009.”
___. 2009b. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” First Quarterly Report. September.
___. 2010a. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” Second Quarterly Report. January.

306 |  References

___. 2010b. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” Third Quarterly Report. April.
___. 2010c. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” Fourth Quarterly Report. July.
___. 2011. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” Sixth Quarterly Report. March.
___. 2013a. “The Economic Impact of the American Recovery and Reinvestment Act of 2009.” Ninth Quarterly Report. February.
___. 2013b. “Economic Report of the President.” March.
Council of Economic Advisers and the Department of Labor. 2014.
“The Economic Benefits of Extending Unemployment Insurance.”
January.
Delong, J. Bradford, and Lawrence H. Summers. 2012. “Fiscal Policy in
a Depressed Economy.” Brookings Papers on Economic Activity.
Spring.
Department of Education. 2011. “FY 2012 Department of Education Justifications of Appropriation Estimates to the Congress: Student
Financial Assistance” in President’s FY 2012 Budget Request for the
U.S. Department of Education.
Department of the Treasury. 2010a. “The Case for Temporary 100 Percent
Expensing: Encouraging Business to Expand now by Lowering the
Cost of Investment.”
___. 2010b. “The American Opportunity Tax Credit.”
___. 2011. “Treasury Analysis of Build America Bonds Issuance and
Savings.”
Department of the Treasury and Council of Economic Advisers. 2012. “A
New Economic Analysis of Infrastructure Investment.” March.
Eggertson, Gauti B. 2001. “Real Government Spending in a Liquidity Trap.”
New York Federal Reserve.
Elmendorf, Douglas W., and Jason Furman. 2008. “If, When, How: A Primer
on Fiscal Stimulus.” Washington: Brookings Institution.
(EPA) Environmental Protection Agency. 2013. “American Recovery and
Reinvestment Act Quarterly Performance Report – FY2013 Quarter
4 Cumulative Results as of September 30, 2013.”

References  | 307

Executive Office of the President and Office of the Vice President. 2010.
“The Recovery Act: Transforming the American Economy through
Innovation”.The White House.
Farhi, Emmanuel, and Ivan Werning. 2012. “Fiscal Multipliers: Liquidity
Traps and Currency Unions.” Working Paper 18321. Cambridge,
MA: National Bureau of Economic Research.
Favero, Carlo, and Francesco Giavazzi. 2012. “Measuring Tax Multipliers: The Narrative Method in Fiscal VARs.” American Economic
Journal: Economic Policy 4(2): 69–94.
Fernald, John G. 1999. “Roads to Prosperity? Assessing the Link Between
Public Capital and Productivity,” The American Economic Review,
Vol. 89, No. 3:619-638.
Feyrer, James, and Bruce Sacerdote. 2011. “Did the Stimulus Stimulate?
Real Time Estimates of the Effects of the American Recovery and
Reinvestment Act.” Working Paper 16759. National Bureau of
Economic Research.
Financial Crisis Inquiry Commission. 2011. The Financial Crisis Inquiry
Report. New York: PublicAffairs.
Follette, Glenn, and Byron Lutz. 2010. “Fiscal Policy in the United States:
Automatic Stabilizers, Discretionary Fiscal Policy Actions, and the
Economy.” Finance and Economics Discussion Series. Washington:
Federal Reserve Board.
Guajardo, Jaime, Daniel Leigh, and Andrea Pescatori. Forthcoming. “Expansionary Austerity? New International Evidence.” Journal of the
European Economic Association.
Hall, Bronwyn H., Jacques Mairesse, Pierre Mohnen. 2009. “Measuring the
Returns to R&D.” Working Paper 15622. Cambridge, MA: National
Bureau of Economic Research.
House, Christopher L., and Matthew D. Shapiro. 2008. “Temporary
Investment Tax Incentives: Theory with Evidence from Bonus
Depreciation.” American Economic Review 98(3): 737-768.
Ilzetzki, Ethan, Enrique G. Mendoza, and Carlos A. Vegh. 2011. “How Big
(Small?) are Fiscal Multipliers?” IMF Working Paper.
(IMF) International Monetary Fund. 2009 “What’s the Damage? Medium
Term Dynamics after Financial Crises.” World Economic Outlook
October 2009: Sustaining the Recovery Chapter 4, p. 121-151.

308 |  References

___. 2012. World Economic Outlook April 2012: Growth Resuming, Dangers
Remain.”
Johnson, David S., Jonathan A. Parker, and Nicholas S. Souleles. 2006.
“Household Expenditure and the Income Tax Rebates of 2001.”
American Economic Review 96(5): 1589-1610.
Joint Committee on Taxation. 2009. “Estimated Budget Effects of the
Revenue Provisions Contained in the Conference Agreement for
H.R.1, The ‘American Recovery and Reinvestment Tax Act of
2009.’”
___. 2010a. “Estimated Budget Effects of the Revenue Provisions Contained
in Senate Amendment #4594 to H.R. 5297, The ‘Small Business
Jobs Act of 2010,’ Scheduled for Consideration by the United States
Senate on September 16, 2010.”
___. 2010b. “Estimated Revenue Effects of the House Amendment to the
Senate Amendment to H.R. 4853, The ‘Middle Class Tax Relief Act
of 2010,’ Scheduled for Consideration by the House of Representatives on December 2, 2010.”
___. 2012. “Estimated Revenue Effects of H.R. 8, The ‘Job Protection and
Recession Prevention Act of 2012.’”
Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises Database:
An Update.”Working Paper WP/12/163. Washington: International
Monetary Fund.
Ljungqvist, Lars, and Thomas J. Sargent. 1998. “The European Unemployment Dilemma.” Journal of Political Economy 106(3): 514-550.
Mertens, Karel, and Morten O. Ravn. 2012. “Empirical Evidence on
the Aggregate Effects of Anticipated and Unanticipated US Tax
Policy Shocks.” American Economic Journal: Economic Policy 4(2):
145–181.
Mertens, Karel, and Morten Ravn. 2013. “The Dynamic Effects of Personal
and Corporate Income Tax Changes in the United States.” American Economic Review. June.
Munnell, Alicia H, 1992. “Infrastructure Investment and Economic Growth,”
Journal of Economic Perspectives, vol. 6(4), pages 189-98, Fall.
Pittsburg, PA: American Economic Association.
Nakamura, Emi, and Jón Steinsson. 2011. “Fiscal Stimulus in a Monetary
Union: Evidence from U.S. Regions.” Unpublished paper, New
York: Columbia University.

References  | 309

Office of Science and Technology Policy, and The National Economic
Council. 2013. “Four Years of Broadband Growth.” The White
House.
Parker, Jonathan A. 2011. “On Measuring the Effects of Fiscal Policy in
Recessions.” Working Paper 17240. Cambridge, MA: National
Bureau of Economic Research.
Parker, Jonathan A., Nicholas S. Souleles, David S. Johnson, and Robert
McClelland. 2011. “Consumer Spending and the Economic Stimulus Payments of 2008.” Working Paper 16684. Cambridge, MA:
National Bureau of Economic Research.
Perotti, Roberto. 2011. “The Austerity Myth: Gain Without Pain?” Working
Paper 17571. Cambridge, MA: National Bureau of Economic
Research.
Phelps, Edmund. 1972. “Inflation policy and unemployment theory.” New
York: WW Norton and Company.
Poterba, James M. 1994. “State Responses to Fiscal Crises: The Effects of
Budgetary Institutions and Politics.” Journal of Political Economy
102(4): 799-821.
Ramey, Valerie A. 2011a. “Identifying Government Spending Shocks: It’s
All in the Timing” Quarterly Journal of Economics 126, no. 1: 1-50.
___. 2011b. “Can Government Purchases Stimulate the Economy?” Journal
of Economic Literature 49, no. 3: 673-85.
Ramey, Valerie A, and Matthew Shapiro. 1998. “Costly Capital Reallocation and the Effects of Government Spending.” Carnegie-Rochester
Conference on Public Policy 48: 145-94.
Reichling, Felix, and Charles Whalen. 2012. “Assessing the Short-Term
Effects on Output Changes in Federal Fiscal Policies.” Working
Paper 2012-08. Congressional Budget Office.
Reinhart, Carmen M., and Kenneth S. Rogoff. 2009. This time is different:
Eight Centuries of Financial Folly. Princeton, NJ: Princeton University Press.
___. Forthcoming. “Recovery from Financial Crises: Evidence from 100
Episodes.” American Economic Review.
Reifschneider, Dave, William L. Wascher, and David Wilcox. 2013. “Aggregate Supply in the United States: Recent Developments and
Implications for the Conduct of Monetary Policy.” 14th Jacques

310 |  References

Polak Annual Research Conference. Washington: International
Monetary Fund.
Romer, Christina D. 2011. “Back from the Brink.” In The International
Financial Crisis: Have the Rules of Finance Changed? Edited by
Asli Demirgüç-Kunt, Douglas D. Evanoff, George G. Kaufman, pp.
15-31. World Scientific Publishing Company.
___. 2012. “Fiscal Policy In the Crisis: Lessons and Policy Implications.”
University of California-Berkeley, Department of Economics.
Romer, Christina D., and David H. Romer. 2010. “The Macroeconomic
Effects of Tax Changes: Estimates Based on a New Measure of Fiscal
Shocks.” American Economic Review 100(3): 763-801.
Rothstein, Jesse. 2011. “Unemployment Insurance and Job Search in the
Great Recession.” Brookings Papers on Economic Activity, Fall.
Shoag, Daniel. 2013. “Using state pension shocks to estimate fiscal multipliers since the Great Recession.” The American Economic Review
103(3): 121-124.
Sims, Christopher A. 2010. “But Economics Is Not an Experimental Science.”
Journal of Economic Perspectives 24, no. 2: 59-68.
Smets, Frank, and Rafael Wouters. 2007. “Shocks and Frictions in US Business Cycles: a Bayesian DSGE Approach.” American Economic
Review 97(3): 586-606.
Sperling, Gene. 2007. “Ways to Get Economic Stimulus Right This Time.”
Bloomberg.com. December 17.
Stock, James H. 2010. “The Other Transformation in Econometric Practice:
Robust Tools for Inference.” Journal of Economic Perspectives 24,
no. 2: 83-94.
Stock, James H. and Mark W. Watson. 2012. “Disentangling the Channels
of the 2007–09 Recession,” Brookings Papers on Economic Activity.
Spring.
Stock, James H., and Mark W. Watson. 2010. Introduction to Econometrics.
3rd ed. Boston, MA: Addison-Wesley.
Suarez Serrato, Juan Carlos, and Philippe Wingender. 2011. “Estimating the
Incidence of Government Spending.”
Summers, Lawrence H. 2007. “The State of the US Economy.” Presentation
at Brookings Institution Forum on December 19th, 2007.

References  | 311

___. 2008. Speech at the Wall Street Journal-CEO Council Conference.
Washington, D.C., November 19.
Taylor, John B. 2011. “An Empirical analysis of the Revival of Fiscal Activism
in the 2000s.” Journal of Economic Literature 49(3): 686:702.
___. Forthcoming. “The Role of Policy in the Great Recession and the Weak
Recovery.” American Economic Review, Papers and Proceedings.
Transportation Research Board of the National Academies. 2013. “Transportation Investments in Response to Economic Downturns, Special
Report 312.”
Wilson, Daniel J. 2012. “Fiscal Spending Jobs Multipliers: Evidence from
the 2009 American Recovery and Reinvestment Act.” American
Economic Journal: Economic Policy.
Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and
Jane Waldfogel. 2013. “Trends in Poverty with an Anchored
Supplemental Poverty Measure.” Working Paper 1- 25. New York:
Columbia Population Research Center.
Woodford, Michael. 2011. “Simple Analytics of the Government Expenditure Multiplier.” American Economic Journal: Macroeconomics (3),
p. 1-35.

Chapter 4
Acemoglu, Daron, Amy Finkelstein, and Matthew J. Notowidigdo. 2013.
“Income and Health Spending: Evidence from Oil Price Shocks.”
Review of Economics and Statistics 95, no. 4: 1079-1095.
Aitken, Murray, Ernst R. Berndt, and David M. Cutler. 2009. “Prescription
Drug Spending Trends in the United States: Looking Beyond the
Turning Point.” Health Affairs 28, no. 1: 151-160.
Altarum Institute. 2014. “Insights from Monthly National Health Expenditures Estimates through November 2013.” (http://altarum.org/
sites/default/files/uploaded-related-files/CSHS-Spending-Brief_
January%202014.pdf).
ASPE (Assistant Secretary for Planning and Evaluation, Office of). 2013.
“Health Insurance Marketplace Premiums for 2014.” Health and
Human Services. (http://aspe.hhs.gov/health/reports/2013/MarketplacePremiums/ib_marketplace_premiums.cfm).

312 |  References

Baicker, Katherine and Amitabh Chandra. 2006. “The Labor Market Effects
of Rising Health Insurance Premiums.” Journal of Labor Economics
24, no. 3: 609-634.
Chandra, Amitabh, Jonathan Holmes, and Jonathan Skinner. 2013. “Is
This Time Different? The Slowdown in Healthcare Spending.”
Brookings Panel on Economic Activity. Washington: Brookings
Institution. (http://www.brookings.edu/~/media/Projects/BPEA/
Fall%202013/2013b%20chandra%20healthcare%20spending.pdf).
Centers for Medicare and Medicaid Services Office of the Actuary. 2013.
“Projections of National Health Expenditures: Methodology and
Model Specification.”
(CMS) Centers for Medicare and Medicaid Services. 2013a. “New Data
Shows Affordable Care Act Reforms Are Leading to Lower Hospital
Readmission Rates for Medicare Beneficiaries.” (http://blog.cms.
gov/2013/12/06/new-data-shows-affordable-care-act-reforms-areleading-to-lower-hospital-readmission-rates-for-medicare-beneficiaries/).
______

.. 2013b. “Pioneer Accountable Care Organizations succeed in
improving care, lowering costs.” News Release. (http://www.cms.
gov/Newsroom/MediaReleaseDatabase/Press-Releases/2013-PressReleases-Items/2013-07-16.html).

. 2013c. “The Affordable Care Act: A Stronger Medicare Program.”
(http://www.cms.gov/apps/files/Medicarereport2012.pdf).

______

Clemens, Jeffrey and Joshua D. Gottlieb. 2013. “Bargaining in the Shadow
of a Giant: Medicare’s Influence on Private Payment Systems.”
Working Paper 19503. Cambridge, MA: National Bureau of
Economic Research.
(CBO) Congressional Budget Office. 2010a. “Cost Estimate of H.R. 4872,
Reconciliation Act of 2010 (Final Health Care Legislation).” (http://
www.cbo.gov/publication/21351).
______. 2010b. “The Budget and Economic Outlook: An Update.” (http://
www.cbo.gov/publication/21670).
______. 2011. “The Budget and Economic Outlook: An Update.” (http://
www.cbo.gov/publication/41586).
______. 2012a. “Updated Budget Projections: Fiscal Years 2012 to 2022.”
(http://www.cbo.gov/publication/43119).

References  | 313

______. 2012b. “Letter to the Honorable John Boehner providing an estimate for H.R. 6079, the Repeal of Obamacare Act.” (http://www.
cbo.gov/publication/43471).
______. 2012c. “Estimates for the Insurance Coverage Provisions of the
Affordable Care Act Updated for the Recent Supreme Court
Decision.”
(http://www.cbo.gov/sites/default/files/cbofiles/
attachments/43472-07-24-2012-CoverageEstimates.pdf).
______. 2012d. “An Update to the Budget and Economic Outlook: Fiscal
Years 2012 to 2022.” (http://www.cbo.gov/publication/43539).
______. 2013a. “Updated Budget Projections: Fiscal Years 2013 to 2023.”
(http://www.cbo.gov/publication/44172).
______. 2013b. “The 2013 Long-Term Budget Outlook.” (http://www.cbo.
gov/publication/44521).
______. 2013c. “Health-Related Options for Reducing the Deficit: 2014
to 2023.” (http://www.cbo.gov/sites/default/files/cbofiles/
attachments/44906-HealthOptions.pdf).
______. 2014. “The Budget and Economic Outlook: Fiscal Years 2014 to
2024.” (http://www.cbo.gov/publication/45010).
(CEA) Council of Economic Advisers. 2009. “The Economic Case for
Health Care Reform.” (http://www.whitehouse.gov/assets/documents/CEA_Health_Care_Report.pdf).
______. 2013. Economic Report of the President. (http://www.whitehouse.gov/
administration/eop/cea/economic-report-of-the-President/2013).
Cuckler, Gigi A., et al. 2013. “National Health Expenditure Projections,
2012-22: Slow Growth Until Coverage Expands And Economy
Improves.” Health Affairs 32 (September): 1-12.
Cutler, David. 2004. Your Money or Your Life: Strong Medicine for America’s
Health Care System. New York: Oxford University.
Cutler, David and Neeraj Sood. 2010. “New Jobs Through Better Health
Care: Health Care Reform Could Boost Employment by 250,000
to 400,000 a Year this Decade.” Washington: Center for American
Progress. (http://www.americanprogress.org/issues/2010/01/pdf/
health_care_jobs.pdf).
Cutler, David and Nikhil R. Sahni. 2013. “If Slow Rate of Health Care
Spending Growth Persists, Projections May Be Off by $770 Billion.”
Health Affairs 32 (September): 841-850.

314 |  References

Daly, Mary, Bart Hobijn, and Brian Lucking. 2012. “Why Has Wage
Growth Stayed Strong?” Federal Reserve Board of San Francisco. (http://www.frbsf.org/economic-research/publications/
economic-letter/2012/april/strong-wage-growth/).
Dranove, David. 1988. “Pricing by Non-Profit Institutions.” Journal of
Health Economics 7 (March): 47-57.
Elmendorf, Douglas. 2013. “The Slowdown in Health Care Spending.”
Presentation to the Brookings Panel on Economic Activity. Washington: Brookings Institution. September 19 (http://www.cbo.gov/
publication/44596).
Frakt, Austin B. 2011. “How Much Do Hospitals Cost Shift? A Review of the
Evidence.” The Milbank Quarterly 89, no. 1: 90-130.
______. 2013. “The End of Hospital Cost Shifting and the Quest for Hospital
Productivity.” Health Services Research 49 (September): 1-10.
Gerhardt, Geoffrey, et al. 2013. “Medicare Readmission Rates Showed
Meaningful Decline in 2012.” Medicare & Medicaid Research
Review 3, no. 2: E1-E12.
Gruber, Jonathan and Alan B. Krueger. 1991. “The Incidence of Mandated
Employer-Provided Insurance: Lessons From Workers’ Compensation Insurance.” In Tax Policy and the Economy, Volume 5, edited
by David Bradford, pp. 111-144. Cambridge, MA: MIT Press.
Gruber, Jonathan. 1994. “The Incidence of Mandated Maternity Benefits.”
American Economic Review 84, no. 3: 622-641.
He, Daifeng and Jennifer M. Mellor. 2012. “Hospital volume responses to
Medicare’s Outpatient Prospective Payment System: Evidence from
Florida.” Journal of Health Economics 31: 730-743.
Holahan, John and Stacey McMorrow. 2012. “Medicare, Medicaid, and the
Deficit Debate.” Washington: Urban Institute. (http://www.urban.
org/UploadedPDF/412544-Medicare-Medicaid-and-the-DeficitDebate.pdf).
IMS Institute for Healthcare Informatics. 2013. “Declining Medicine Use
and Costs: For Better or Worse?” (http://static.correofarmaceutico.
com/docs/2013/05/20/usareport.pdf).
Kaiser

Family Foundation. 2013a. “2013 Employer Health
Benefits
Survey.”
(http://kff.org/private-insurance/
report/2013-employer-health-benefits).

References  | 315

______. 2013b. “Medicaid Enrollment: June 2012 Data Snapshot.” (http://
kaiserfamilyfoundation.files.wordpress.com/2013/08/805006-medicaid-enrollment.pdf).
Kaiser Family Foundation and Altarum Institute. 2013. “Assessing the Effects
of the Economy on the Recent Slowdown in Health Spending.”
(http://kff.org/health-costs/issue-brief/assessing-the-effects-of-theeconomy-on-the-recent-slowdown-in-health-spending-2).
Levine, Michael and Melinda Buntin. 2013. “Why Has Growth in Spending
for Fee-for-Service Medicare Slowed?” Working Paper 2013-06.
Congressional Budget Office.
Martin, Anne B., et al. 2014. “National Health Spending in 2012: Rate of
Health Spending Growth Remained Low for the Fourth Consecutive Year.” Health Affairs 33, no. 1: 1-11.
McClellan, Mark, et al. 1998. “Are Medical Prices Declining? Evidence from
Heart Attack Treatments.” The Quarterly Journal of Economics 113,
no. 4: 991-1024.
McWilliams, J. Michael, Bruce Landon, Michael E. Chernew. 2013. “Changes
in Health Care Spending and Quality for Medicare Beneficiaries
Associated With a Commercial ACO Contract.” Journal of the
American Medical Association 310, no. 8: 829-836.
MedPAC (Medicare Payment Advisory Commission). 2009. “Improving
Incentives in the Medicare Program.” (http://www.medpac.gov/
documents/jun09_entirereport.pdf).
Morrissey, Michael A. 1994. Cost Shifting in Health Care: Separating
Evidence from Rhetoric. Washington: American Enterprise Institute.
Ryu, Alexander J., et al. 2013. “The Slowdown in Health Care Spending in
2009-11 Reflected Factors Other than the Weak Economy and Thus
May Persist.” Health Affairs 32, no. 5: 835-840.
Sisko, Andrea, et al. 2009. “Health Spending Projections Through 2018:
Recession Effects Add Uncertainty to the Outlook.” Health Affairs
28, no. 2: 346-357.
Sood, Neeraj, Arkadipta Ghosh and Jose J. Escarce. 2009. “Costs, Use
and Outcomes: Employer-Sponsored Insurance, Health Care Cost
Growth, and the Economic Performance of U.S. Industries.” Health
Services Research 44, no. 5: 1449-1464.
Sommers, Benjamin D. 2005. “Who Really Pays for Health Insurance?
The Incidence of Employer-Provided Health Insurance with Sticky

316 |  References

Nominal Wages,” International Journal of Health Care Finance and
Economics 5, no. 1: 89-118.
Song, Zirui, et al. 2012. “The ‘Alternative Quality Contract,’ Based on a
Global Budget, Lowered Medical Spending and Improved Quality.”
Health Affairs 31, no.8: 1885-1894.
Spiro, Topher and Jonathan Gruber. 2013. “The Affordable Care Act’s
Lower-Than-Projected Premiums Will Save $190 billion.” Washington: Center for American Progress. October 21 (http://www.
americanprogress.org/issues/healthcare/report/2013/10/23/77537/
the-affordable-care-acts-lower-than-projected-premiums-willsave-190-billion/).
(SSA) Social Security Administration. 2013. “Annual Statistical Supplement,
2013.” (http://www.ssa.gov/policy/docs/statcomps/supplement/).
Summers, Lawrence H. 1989. “Some Simple Economics of Mandated Benefits.” The American Economic Review 79, no. 2: 177-183.
White, Chapin. 2007. “Health Care Spending Growth: How Different is the
United States from the Rest of the OECD.” Health Affairs, 26 no. 1:
154-161.
______. 2013. “Contrary To Cost-Shift Theory, Lower Medicare Hospital
Payment rates For Inpatient Care Lead To Lower Private Payment
Rates.” Health Affairs 32, no. 5: 935-943.
White, Chapin and Vivian Yaling Wu. 2013. “How Do Hospitals Cope
with Sustained Slow Growth in Medicare Prices?” Health Services
Research 49: 1-21.
Yamamato, Dale H. 2013. “Health Care Costs From Birth to Death.” Health
Care Cost Institute Independent Report Series. (http://www.healthcostinstitute.org/files/Age-Curve-Study_0.pdf).

Chapter 5
Acemoglu, Daron. 2002. “Directed Technical Change.” The Review of
Economic Studies 69, no. 4: 781-809.
Acemoglu, Daron, and David Autor. 2011 “Skill, Tasks and Technologies:
Implications for Employment Earnings.” In The Handbook of Labor
Economics, edited by Orley Ashenfelter and David Card, vol. 4b.
Amsterdam: Elsevier.
Acemoglu, Daron and James Robinson. 2012. Why Nations Fail: The Origins
of Power, Prosperity and Poverty. New York: Crown Publishers.
References  | 317

Acemoglu, Daron and David Autor. 2012. “What Does Human Capital Do?
A Review of Goldin and Katz’s Race between Education and Technology.” Journal of Economic Literature, 50(2): 426–463.
Agrawal, N. 2010. “Review on just in time techniques in manufacturing
systems.” In Advances in Production Engineering & Management 5,
no.2: 101-110
Akcigit, Ufuk, Douglas Hanley, and Nicolas Serrano-Velarde. 2013. “Back to
Basics: Basic Research Spillovers, Innovation Policy and Growth.”
Working Paper 19473. Cambridge, MA.: National Bureau of
Economic Research.
Alvaredo, Facundo, Anthony B. Atkinson, Thomas Piketty, and Emmanuel
Saez. 2013. “The Top 1 Percent in International and Historical
Perspective.” Journal of Economic Perspectives, 27(3): 3-20.
Autor, David. 2010. “The Polarization of Job Opportunities in the U.S. Labor
Market, Implications for Employment and Earnings.” Washington:
Center for American Progress and The Hamilton Project.
Autor, David, Lawrence F. Katz, and Melissa S. Kearney. 2006. “The Polarization of the U.S. Labor Market.” Working Paper 11986. Cambridge,
MA: National Bureau of Economic Research.
Autor, David, Frank Levy, and Richard J. Murnane. 2003. “The Skill Content
of Recent Technological Change: an Empirical Exploration.” Quarterly Journal of Economics 116 (4): 1279-1333.
Ayres, Ian, and Peter Cramton. 1996. “Deficit Reduction Through Diversity:
How Affirmative Action at the FCC Increased Auction Competition,” Stanford Law Review 48: 761-815.
Bailey, Diane E., and Nancy B. Kurland. 1999. “The Advantages and
Challenges of Working Here, There, Anywhere, and Anytime.”
Organizational Dynamics (Autumn): 1-16.
Bailey, Diane E., and Nancy B. Kurland. 2002. “A Review of Telework
Research: Findings, New Directions, and Lessons for the Study
of Modern Work.” Journal of Organizational Behavior 23, no. 4:
383-400.
Baily, Martin N., and Robert J. Gordon. 1988. “The Productivity Slowdown,
Measurement Issues, and the Explosion of Computer Power,”
Brookings Papers on Economic Activity, vol. 19(2): 347-432.
Bakia, Marianne, Karla Jones, Barbara Means, Robert Murphy, and Yukie
Toyama. 2010. “Evaluation of Evidence-Based Practices in Online

318 |  References

Learning: A Meta-Analysis and Review of Online Learning Studies.”
U.S. Department of Education Office of Planning, Evaluation, and
Policy Development Policy and Program Studies Service.
Banerjee, Abhijit V., and Esther Duflo. 2003. “Inequality and Growth: What
Can the Data Say?” Journal of Economic Growth 8, no. 3: 267-299.
Banerjee, Abhijit, Shawn Cole, Esther Duflo, and Leigh Linden. 2007.
“Remedying Education: Evidence from Two Randomized Experiments in India.” The Quarterly Journal of Economics, 1235-1264.
Barrow, Lisa, Lisa Markman, and Cecilia Rouse. 2009. “Technology’s Edge:
The Educational Benefits of Computer-Aided Instruction.” The
American Economic Journal: Economic Policy, (1), 52-74.
Basu, Fernald, Oulton, and Srinivasan. 2004. “The Case of the Missing
Productivity Growth, or Does Information Technology Explain
Why Productivity Accelerated in the United States but Not in
the United Kingdom?” Working Paper 10010. Cambridge, MA:
National Bureau of Economic Research.
Basu, Susanto, John Fernald, and Miles Kimball. 2006. “Are Technology
Improvements Contractionary?” American Economic Review. 96(5),
1418-48.
Berman, Eli, John Bound, and Zvi Griliches. 1994. “Changes in the
Demand for Skilled Labor within U.S. Manufacturing: Evidence
from the Annual Survey of Manufactures.” The Quarterly Journal of
Economics 109(2), 367–397.
Bernanke, Ben. 2013. “Economic Progress for the Long Run.” Speech at Bard
College. May 18.
Bernstein, Jared. 2013. “The Impact of Inequality on Growth,” Washington:
Center for American Progress.
Bloom, Kretschmer, and Van Reenen. 2006. “Work-Life Balance, Management Practices and Productivity.” Centre for Economic Performance
(January): 1-45.
Bloom, Nicholas, Mark Schankerman, and John Van Reenen. 2012. “Identifying Technology Spillovers and Product Market Rivalry.” Centre
for Economic Performance (December): 1-81.
Bloom, Liang, Roberts, and Ying. 2013. “Does Working from Home Work?
Evidence from a Chinese Experiment.” Working Paper. London:
Centre for Economic Performance.

References  | 319

BLS (Bureau of Labor Statistics). “Private Nonfarm Business Sector: Multifactor Productivity”
_____. “Private Nonfarm Business Sector: Output per Hour of all Persons”
_____. “Private Nonfarm Business Sector: Sources of Productivity Growth.”
_____. “Private Nonfarm Business Sector: Real Output Per Hour.”
_____. “Private Nonfarm Business Sector: Real Compensation Per Hour.”
Bound, John and George Johnson. 1995. “What are the Causes of Rising
Wage Inequality in the United States?” Economic Policy Review 1,
no. 1: 9-17.
Bureau of Labor Statistics, Office of Productivity and Technology. 2013. Net
Multifactor Productivity and Cost, 1948-2012.
Busch, Emily, Jenna Nash, and Bradford S. Bell. 2011. “Remote Work: An
Examination of Current Trends and Emerging Issues.” Center for
Advanced Human Resource Studies, Cornell University (Spring):
1-12.
Card, David, and John DiNardo. 2002. “Skill Biased Technological Change
and Rising Wage Inequality: Some Problems and Puzzles.” Journal
of Labor Economics 20(4), 733–783.
Carew, Diana, G. and Michael Mandel. 2013. Progressive Policy Institute.
“U.S. Investment heroes of 2013: The Companies Betting on America’s Future”.
Carillo, Ponce, Mercedes Onofa, and Juan Ponce. 2010. “Information Technology and Student Achievement: Evidence from a Randomized
Experiment in Ecuador.” Washington: Inter-American Development Bank.
CBO (Congressional Budget Office). 2011a. “Trends in the Distribution of
Household Income between 1979 and 2007.”
_____. 2011b. “S. 27, Preserve Access to Affordable Generics Act.”
_____. 2013. The Economic Impact of S. 744, the Border Security, Economic
Opportunity, and Immigration Modernization Act. Government
Printing Office.
Charles, Dustin, et al. 2013. “Adoption of Electric Health Record Systems
among U.S. Non-federal Acute Care hospitals: 2008-2012.” ONC
Data Brief no. 9. The Office of the National Coordinator for Health
Information Technology, U.S. Department of Health and Human
Resources.

320 |  References

Chien, Colleen V. 2012. “Reforming Software Patents.” Legal Studies
Research Paper Series. Santa Clara, CA: Santa Clara University
School of Law.
Coase, Ronald H. 1959. “The Federal Communications Commission,”
Journal of Law and Economics 2: 1-40.
Council of Economic Advisers, the National Economic Council, and the
Office of Science & Technology Policy. 2013. “Patent Assertion and
U.S. Innovation.” The White House.
CEA (Council of Economic Advisers). _____. 2010. “Work-Life Balance and
the Economics of Workplace Flexibility.”(March).
_____. 2012 “The Economic Benefits of New Spectrum for Wireless
Broadband.”
Cramton, Peter and Jesse A. Schwartz. 2000. “Collusive Bidding: Lessons
from the FCC Spectrum Auctions,” Journal of Regulatory Economics
17: 229-252.
_____. 2002. “Collusive Bidding in the FCC Spectrum Auctions,” Contributions to Economic Analysis & Policy 1: 1-18.
Deardorff, Alan V. 1991. “Welfare Effects of Global Patent Protection.”
Economica 59 (May): 35-51.
Delgado, Mercedes, Christian Ketels, Michael E. Porter, and Scott Stern.
2012. “The Determinants of National Competiveness.” Working
Paper 18249. Cambridge, MA.: National Bureau of Economic
Research (July).
Department of Justice and United States Patent and Trademark Office. 2013.
Policy Statement on Remedies for Standard-Essential Patents Subject
to Voluntary FRAND Commitments. Government Printing Office.
DiNardo, John, Nicole M Fortin, and Thomas Lemieux. 1996. “Labor
Market Institutions and the Distribution of Wages, 1973-1992: A
Semiparametric Approach.” Econometrica 64(5), 1001–1044.
Evenson, Robert, and Sunil Kanwar. 2003. “Does intellectual property spur
technological change?” Oxford Economic Papers 55: 235-264.
Fairris, David, and Mark Brenner. 2001. “Workplace Transformation and
the Rise in Cumulative Trauma Disorders: Is There a Connection?”
Journal of Labor Research XXII, no.1 (Winter): 15-28.
Falvey, Rod, Neil Foster, and David Greenaway. 2006. “Intellectual Property
Rights and Economic Growth.” Review of Development Economics
10(4): 700-719.
References  | 321

Farrell, Joseph, John Hayes, Carl Shapiro, and Theresa Sullivan. 2007. “Standard Setting, Patents, and Hold-Up.” Antitrust Law Journal 74, no.
3: 603-670.
Federal Trade Commission v. Actavis, Inc. et al. 570 U.S. __ (2013).
FTC (Federal Trade Commission). 2010. “Pay-for-delay: How drug company
pay-offs cost consumers billions: An FTC staff study.”.
Fernald, John. 2012. “Productivity and potential output before, during, and
after the Great Recession.” Working Paper 2012-18. Federal Reserve
Bank of San Francisco.
Feyrer, James. 2007. “Demographics and Productivity.” The Review of
Economics and Statistics 89, 100-109.
_____. 2011. “The U.S. Productivity Slowdown, the Baby Boom, and
Management Quality.” Journal of Population Economics 24, 267-284.
Fields, Gary. 2001. “Distribution and Development, A New Look at the
Developing World.” Journal of Development Economics, Vol. 70:
238-243.
Foellmi, R. and Zweimuller, J. 2006. “Income Distribution and DemandInduced Innovations.”
Review of Economic Studies 73(4): 941-960.
Freeman, Richard B., and Lawrence F. Katz, eds.  1995. Differences and
Changes in Wage Structures. Chicago, IL: University of Chicago
Press.
Furman, Jason. 2013. “Remarks at AEI’s Center on Internet, Communications and Technology Policy.” Washington, D.C., September 17.
Galor, Oded. 2011.“Inequality, Human Capital Formation and the Process
of Development,” Working Paper 17058. Cambrige: MA: National
Bureau of Economic Research.
Galor, Oded and Omer Moav. 2000. “Ability-Biased Technological Transition, Wage Inequality, and Economic Growth,” Quarterly Journal of
Economics, 115, 469-497.
_____. 2004. “From Physical to Human Capital Accumulation: Inequality
and the Process of Development.” Review of Economic Studies, 71,
1001-1026.
Galor, Oded and Daniel Tsiddon. 1997. “The Distribution of Human
Capital, Technological Progress, and Economic Growth.” Journal of
Economic Growth, 2, 93-124

322 |  References

GAO (Government Accountability Office). 2001. Spectrum Management:
NTIA Planning and Processes Need Strengthening to Promote the
Efficient Use of Spectrum by Federal Agencies. GAO-11-352.
_____. 2012. Electronic Health Records: Number and Characteristics of
Providers Awarded Medicaid Incentive Payments for 2011. Government Printing Office.
_____. 2013a. Assessing Factors That Affect Patent Infringement Litigation
Could Help Improve Patent Quality. GAO 13-465. Government
Printing Office.
_____. 2013b. Electronic Health Records: Number and Characteristics of
Providers Awarded Medicare Incentive Payments for 2011-2012.
Government Printing Office.
Gilbert, Richard J. 2010-2011. “Deal or No Deal? Licensing Negotiations in
Standard-Setting Organizations.” HeinOnline 77 855-888.
Goos, Martin and Alan M. Manning, 2007. “Lousy and Lovely Jobs: the
rising polarization of work in Britain. Review of Economics and
Statistics. 89(1): 118-133.
Goldin, Claudia, and Lawrence F. Katz. 2008. “The Race between Education
and Technology.” Cambridge, MA: The Belknap Press of Harvard
University Press.
Gordon, Robert J. 2012. “Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds.” Working Paper 18315.
Cambridge, MA.: National Bureau of Economic Research.
Greenstone, M., R. Hornbeck, and E. Moretti. 2010. “Identifying Agglomeration Spillovers: Evidence from Winners and Losers of Large Plant
Openings,” Journal of Political Economy, 118, 536-598.
Hall, Robert E., and Charles I. Jones. 1999. “Why Do Some Countries
Produce So Much More Output Per Worker Than Others?” The
Quarterly Journal of Economics: 83-116.
Haskel, Jonathan, Robert Z. Lawrence, Edward E. Leamer, and Matthew J.
Slaughter. 2012. “Globalization and U.S. Wages: Modifying Classic
Theory to Explain Recent Facts.” Journal of Economic Perspectives
26, no. 2: 119-40.
Heckman, James J. and Paul A. LaFontaine. 2010. “The American High
School Graduation Rate: Trends and Levels.” The Review of
Economics and Statistics 92, no. 2: 244-262.

References  | 323

Heckman, James J., Rodrigo Pinto, and Peter A. Savelyev. 2012. “Understanding the Mechanisms Through Which an Influential Early
Childhood Program Boosted Adult Outcomes.” Working Paper
18581, Cambridge, MA.:National Bureau of Economic Research.
Hemphill, C. Scott, and Bhaven N. Sampat. 2011. “When Do Generics
Challenge Drug Patents?.” Journal of Empirical Legal Studies 8.4:
613-649.
HHS (Department of Health and Human Services). 2013. “Doctors and
hospitals’ use of health IT more than doubles since 2012.” May 22.
Holl, Adelheid, Rafael Pardo, and Ruth Rama. 2010. “Inside and outside the
factory: Just-in time manufacturing systems, subcontracting and
geographic proximity.” Regional Studies 44, no. 5: 519-533.
_____. 2011. “Spatial Patterns of Adoption of Just-in-Time Manufacturing.”
Working Paper, No. 1. Spanish National Research Council.
Hong, Jun, George Huang, Pingyu Jiang, Ting Qu, Yingfeng Zahang, and
Guanghui Zhou. 2010. “RFID-enabled real-time manufacturing
information tracking infrastructure for extended enterprises.”
Journal of Intelligent Manufacturing 23 (November): 2357-2366.
Hsiao, Chun-Ju and Esther Hing. 2014. “Use and Characteristics of Electric
Health Record Systems Among Office-based Physician Practices:
United States, 2001-2013.” NCHS Data Brief No. 143. U.S. Department of Health and Human Services.
Hur, Seung Min, Suho Jeong, and Suk-Hwan Suh. 2009. “An experimental
approach to RFID system performance prediction model.” International Journal of Computer Integrated Manufacturing 22, no. 7
(July): 686-697.
Iowa State University Extension Service. 2013. Corn Production, Harvest
and Yield.
JEC (U.S. Congress Joint Economic Committee). 2010a. The Pivotal Role
of Government Investment in Basic Research. Government Printing
Office.
_____. 2010b. Challenges and Opportunities for Job Creation in the Aftermath of the Great Recession. Cong. 2 sess. Government Printing
Office.
Jones, Charles J., and John C. Williams. 1998. “Measuring the Social
Return to R&D.” The Quarterly Journal of Economics (November):
1119-1135.

324 |  References

Jorgenson, Dale. 2001. “Information Technology and the US Economy.”
The American Economic Review. Vol 91 (No. 1). 1-32. March.
Jorgenson, Dale, Mun Ho, and Jon Samuels. 2012. “Information Technology
and U.S. Productivity Growth.” Industrial Productivity in Europe,
35-64. Northampton MA: Edward Elgar.
Jorgenson, Dale W. 1988. “Productivity and Postwar U.S. Economic
Growth.” Journal of Economic Perspectives.” 2(4): 23-41.
Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce. 1993. “Wage
Inequality and the Rise in Returns to Skill.” Journal of Political
Economy 101(3), 410–442.
Kahn, James and Robert Rich. 2011. “The Productivity Slowdown Reaffirmed.” Federal Reserve Bank of New York.
Katz, Lawrence F. and Kevin M. Murphy. 1992. “Changed in Relative Wages,
1963-1987: Supply and Demand Factors.” The Quarterly Journal of
Economics 107, no. 1: 35-78.
Klein, Judith L.V., and William N. Parker. 1966. “Productivity Growth
in Grain Production in the United States, 1840-60 and 1900-10.”
Working Paper 523-582. Cambridge, MA: National Bureau of
Economic Research.
Krueger, Alan B. 1993. “How Computers Have Changed the Wage Structure: Evidence from Microdata, 1984-1989.” The Quarterly Journal
of Economics 108, no. 1: 33-60.
Kwerel, Evan R. and Gregory L. Rosston. 2000. “An Insiders’ View of FCC
Spectrum Auctions,” Journal of Regulatory Economics 17: 253-289.
Lee, David. 1999. “Wage Inequality in the United States during the 1980s:
Rising Dispersion or Falling Minimum Wage?” Quarterly Journal
of Economics. 114(3), 977–1023.
Lemieux, Thomas. 2008. “The Changing Nature of Wage Inequality.”
Journal of Population Economics 21, no. 1: 21-48.
_____. 2006. “Post-Secondary Education and Increasing Wage Inequality,”
American Economic Review 96(2), 195–99.
Lemley, Mark A., and Carl Shapiro. 2005. “Probabilistic Patents.” Journal of
Economic Perspectives 19, no.2 (Spring): 75-98.
Mateyka, Peter J., and Melanie A. Rapino. 2012. “Home-Based Workers in
the United States: 2010.” Household Economic Studies, U.S. Census
Bureau (October): 1-32.

References  | 325

Mishel, Lawrence, Heidi Shierholz, and John Schmitt. 2013. “Don’t Blame
the Robots: Assessing the Job Polarization Explanation of Growing
Wage Inequality.” Washington: Economic Policy Institute.
National Science Foundation. 2013. National Patterns of R&D Resources:
2010-11 Update.
National Telecommunications and information Administration. 2013.
National Broadband Map. “US Broadband Availability Data.”
Nelson, Richard, R. 1959. “The Simple Economics of Basic Scientific
Research.” Journal of Political Economy 297. 727-736.
Noonan, Mary C. and Jennifer L. Glass. 2012. “The hard truth about telecommuting.” Monthly Labor Review, Statistics (June): 38-45. U.S.
Bureau of Labor Statistics.
Nordhaus, William. 2004. “Retrospective on the Postwar Productivity Slowdown.” Cowles Foundation Discussion Paper No. 494. Cambridge,
MA: National Bureau of Economic Research.
Office of Science and Technology Policy and the National Economic Council.
2013. “Four Years of Broadband Growth.” The White House. (June).
Oliner, Sichel. And Stiroh, 2007. “Explaining a Productive Decade,” Finance
and Economics Discussion Series 2007-63, Board of Governors of
the Federal Reserve System.
Pham, Duc Truong, and Paulette Pham. 2009. “FIT Manufacturing: Linking
manufacturing, marketing and product innovation strategies to
achieve long term economic sustainability.” Wales, UK: Cardiff
University.
Piketty, Thomas and Emmanuel Saez. 2003. “Income Inequality in the
United States, 1919 1998.” The Quarterly Journal of Economics118,
no. 1: 1-39.
_____. 2006. “The Evolution of Top Incomes: A Historical and International
Perspective.” Working Paper 11955. Cambridge, Mass: national
Bureau of Economic Research (January).
Presidential Memorandum. 2010. “Unleashing the Broadband Wireless
Revolution.” June 10.
President’s Council of Advisors on Science and Technology. 2012. Report to
the President: Realizing the Full Potential of Government-Held Spectrum to Spur Economic Growth. The White House.
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. “Institutions Rule: The Primacy of Institutions Over Geography and
326 |  References

Integration in Economic Development.” Journal of Economic
Growth 9: 131-165.
Scott Morton, Fiona, and Carl Shapiro. 2013. “Strategic Patent Acquisitions.” Working Paper. University of California at Berkeley. (July).
Shackleton, Robert. 2013. “Total Factor Productivity Growth in Historical
Perspective.” Working Paper 2013-01. Congressional Budget Office
Sim, Khim Ling, and Hian Chye Koh. 2003. “An Empirical Examination
of Management Control Systems in Just-In-Time Manufacturing.”
The Review of Business Information Systems 7, no.3: 71-82.
Stalk, George. 1989. “Time-the next source of competitive advantage.” The
McKinsey Quarterly (Spring): 28-50.
Tayal, S.P. 2012. “Just In Time Manufacturing.” International Journal of
Applied Engineering Research 7, no.11.
USDA. 2013. Farm Computer Usage and Ownership. Government Printing
Office.
U.S. House of Representatives. Committee on the Judiciary. 2011. America
Invents Act. Report 112-98. Part 1 Cong.1 sess. Government
Printing Office.
U.S. Patent and Trademark Office. 2013. US Patents by Technological
Category.
Zhang, Yingfeng, Pingyu Jiang, George Huang, Ting Qu, Guangui Zhou, Jun
Hong. 2012. “RFID-enabled real-time manufacturing information
tracking infrastructure for extended enterprises.”

Chapter 6
Acs, Gregory and Seth Zimmerman. 2008. “U.S. Intragenerational Economic
Mobility from 1984-2004: Trends and Implications.” Washington:
The Pew Charitable Trusts.
Almond, Douglas, Kenneth Y. Chay, and Michael Greenstone. 2006. “Civil
Rights, the War on Poverty, and Black-White Convergence in
Infant Mortality in the Rural South and Mississippi.” MIT Department of Economics Working Paper 07-04.
Anderson, Michael L. 2008. “Multiple Inference and Gender Differences
in the Effects of Early Intervention: A Reevaluation of the Abecedarian, Perry Preschool, and Early Training Projects.” Journal of the
American Statistical Association 103, no. 484: 1481-1495.

References  | 327

Andersson, Fredrik, Harry J. Holzer, Julia I. Lane, David Rosenblum and
Jeffrey Smith. 2013. “Does Federally-Funded Job Training Work?
Nonexperimental Estimates of WIA Training Impacts Using Longitudinal Data on Workers and Firms,” Working Paper 19446.
Cambridge, MA: National Bureau of Economic Research.
Anzick, Michael A. and David A. Weaver. 2001. “Reducing Poverty Among
Elderly Women.” Working Paper Series Number 87. Office of
Research, Evaluation, and Statistics, Social Security Administration.
Auten, Gerald, Geoffrey Gee, and Nicholas Turner. 2013. “Income Inequality,
Mobility, and Turnover at the Top in the US, 1987–2010.” American
Economic Review 103 (May, Papers and Proceedings, 2012): 168–172.
Baicker, Katherine, Sarah L. Taubman, Heidi L. Allen, Mira Bernstein,
Jonathan H. Gruber, Joseph P. Newhouse, Eric C. Schneider, Bill
J. Wright, Alan M. Zaslavsky, and Amy N. Finkelstein. 2013. “The
Oregon Experiment—Effects of Medicaid on Clinical Outcomes.”
The New England Journal of Medicine 368: 1713-1722.
Bailey, Martha J. and Sheldon Danziger, eds. 2013. Legacies of the War on
Poverty. New York: Russell Sage Foundation.
Ben-Shalom, Yonatan, Robert A. Moffitt, and John Karl Scholz. 2011.
“An Assessment of Anti-Poverty Programs in the United States.”
Working Paper 17042. Cambridge, MA: National Bureau of
Economic Research.
Björklund, Anders, Markus Jäntti, and Gary Solon. 2005. “Influences of
Nature and Nurture on Earnings Variation: A Report on a Study
of Various Sibling Types in Sweden.” In Unequal Chances: Family
Background and Economic Success, edited by Samuel Bowles,
Herbert Gintis, and Melissa Osborne Groves, pp. 145-164. Princeton University Press.
Black, Sandra E. and Paul J. Devereux. 2011. “Recent Developments in Intergenerational Mobility.” In Handbook of Labor Economics, Volume
4B, edited by Orley Ashenfelter and David Card, pp. 1487–1541.
Amsterdam: North Holland Publishing Co.
Blank, Rebecca M. 1993. “Public Sector Growth and Labor Market Flexibility: The United States vs. the United Kingdom.” Working Paper
4338. Cambridge, MA: National Bureau of Economic Research.
_____. 2000. “Fighting Poverty: Lessons from Recent U.S. History.” Journal
of Economic Perspectives 14, no. 2: 3-19.

328 |  References

_____. 2007. “Improving the Safety Net for Single Mothers Who Face
Serious Barriers to Work.” Future of Children 17, no. 2: 183-197.
Burtless, Gary. 1986. “Social Security, Unanticipated Benefit Increases, and
the Timing of Retirement.” Review of Economic Studies 53, no. 5:
781-805.
Campbell, Frances A., Barbara H. Wasik, Elizabeth Pungello, Margaret
Burchinal, Oscar Barbarin, Kirsten Kainz, Joseph J. Sparling, and
Craig T. Ramey. 2008. “Young Adult Outcomes of the Abecedarian
and CARE Early Childhood Educational Interventions.” Early
Childhood Research Quarterly 23, no. 4: 452–466.
Cancian, Maria and Deborah Reed. 2009. “Family structure, childbearing,
and parental employment: Implications for the level and trend in
poverty.” Focus 26, no.2: 21-26.
Caplow, Theodore and Jonathan Simon. 1999. “Understanding Prison
Policy and Population Trends.” Crime and Justice 26: 63-120.
Card, David E. and Steven Raphael. 2013. Immigration, Poverty, and Socioeconomic Inequality. New York: Russell Sage.
Carlson, Deven, Robert Haveman, Thomas Kaplan, and Barbara Wolfe.
2011. “The Benefits and Costs of the Section 8 Housing Subsidy
Program: A Framework and Estimates of First-Year Effects.”
Journal of Policy Analysis and Management 30, 2: 233–255.
Census Bureau. 2013. “Poverty – Experimental Measures.” http://www.
census.gov/hhes/povmeas/data/nas/tables/index.html
Chay, Kenneth Y., Jonathan Guryan, and Bhashkar Mazumder. 2009.
“Birth Cohort and the Black-White Achievement Gap: The Roles
of Access and Health Soon After Birth.” Working Paper No. 15078.
Cambridge, MA, National Bureau of Economic Research.
Chetty, Raj. 2008. “Moral Hazard versus Liquidity and Optimal Unemployment Insurance.” Journal of Political Economy 116, no. 2: 173-234.
Chetty, Raj, John N. Friedman, and Jonah E. Rockoff. 2011. “New Evidence
on the Long-Term Impacts of Tax Credits.” Statistics of Income
Paper Series. Internal Revenue Service.
Chetty, Raj, John N. Friedman, and Emmanuel Saez. 2012. “Using Differences in Knowledge Across Neighborhoods to Uncover the Impacts
of the EITC on Earnings.” Working Paper 18232. Cambridge, MA:
National Bureau of Economic Research.

References  | 329

Chetty, Raj, John N. Friedman, Soren Leth-Peterson, Torben Heien Nielson,
and Tore Olsen. 2013. “Subsidies vs. Nudges: Which Policies
Increase Saving the Most?” Issue Brief 13-3. Center for Retirement
Research at Boston College.
Chetty, Raj, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez. 2014.
“Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States.” Working Paper 19843.
Cambridge, MA: National Bureau of Economic Research.
Chetty, Raj, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez, and
Nick Turner. 2014. “Is the United States Still a Land of Opportunity? Recent Trends in Intergenerational Mobility.” Working Paper
19844. Cambridge, MA: National Bureau of Economic Research.
(CBO) Congressional Budget Office. 2013. “Growth in Means-Tested
Programs and Tax Credits for Low-Income Households.” http://
www.cbo.gov/sites/default/files/cbofiles/attachments/43934Means-TestedPrograms.pdf.
Connelly, Rachel and Jean Kimmel. 2003. “The Effect of Child Care Costs
on the Employment and Welfare Recipiency of Single Mothers.”
Southern Economic Journal 69, no. 3: 498-519.
Corak, Miles. 2006. “Do Poor Children Become Poor Adults? Lessons from
a Cross Country Comparison of Generational Earnings Mobility.”
IZA Discussion Paper No. 1993.
_____. 2011. “Inequality from generation to generation: the United States in
Comparison.” Graduate School of Public and International Affairs,
University of Ottawa.
Corcoran, Mary. 2001. “Mobility, Persistence, and the Consequences of
Child Poverty for Children: Child and Adult Outcomes.” In Understanding Poverty, edited by Sheldon H. Danziger and Robert H.
Haveman, pp. 127-140. Cambridge, MA: Harvard University Press.
Corcoran, Mary and Terry Adams. 1997. “Race, Sex, and the Intergenerational Transmission of Poverty.” In Consequences of Growing
Up Poor, edited by Greg J. Duncan and Jeanne Brooks-Gunn, pp.
461-517. New York: Russell Sage Foundation.
(CEA) Council of Economic Advisers. 1964. Economic Report of the
President.
_____. 2013. “Trends in Health Care and Cost Growth and the Role of the
Affordable Care Act.”

330 |  References

Dahl, Gordon B. and Lance Lochner. 2012. “The Impact of Family Income
on Child Achievement: Evidence from the Earned Income Tax
Credit.” American Economic Review 102, no. 5: 1927-1956.
Danziger, Sheldon and Peter Gottschalk. 1995. America Unequal. Cambridge,
MA: Harvard University Press.
Danziger, Sheldon H., Lesley J. Turner, and Kristin S. Seefeldt. 2006.
“Failing the Transition from Welfare to Work: Women Chronically
Disconnected from Employment and Cash Welfare.” Social Science
Quarterly 87, no. 2: 227-249.
Deming, David. 2009. “Early Childhood Intervention and Life-Cycle Skill
Development: Evidence from Head Start.” American Economic
Journal: Applied Economics 1, no. 3: 111-134.
DiNardo, John, Nicole M. Fortin, and Thomas Lemieux. 1996. “Labor
Market Institutions and the Distribution of Wages, 1973-1992: A
Semiparametric Approach.” Econometrica 64, no. 5: 1001-1044.
DiNardo, John, and Thomas Lemieux. 1997. “Diverging Male Wage
Inequality in the United States and Canada, 1981-1988: Do Institutions Explain the Difference?” Industrial and Labor Relations
Review 50, no. 4: 629-651.
Doucouliagos, Hirstos and T.D. Stanley. 2009. “Publication Selection Bias
in Minimum-Wage Research? A Meta-Regression Analysis.” British
Journal of Industrial Relations 47, no. 2: 406-428.
Dowd, Tim and John B. Horowitz. 2011. “Income Mobility and the Earned
Income Tax Credit: Short-Term Safety Net or Long-Term Income
Support.” Public Finance Review 39, no. 5: 619-652.
Dube, Arindrajit, T. William Lester, and Michael Reich. 2010. “Minimum
Wage Effects Across State Borders: Estimates Using Contiguous
Counties.” The Review of Economics and Statistics 92, no. 4: 945-964.
Dube, Arindrajit. 2013. “Minimum Wages and the Distribution of Family
Incomes.” University of Massachusetts, Amherst Working Paper.
https://dl.dropboxusercontent.com/u/15038936/Dube_MinimumWagesFamilyIncomes.pdf.
Duncan, Greg J., Katherine Magnuson, Ariel Kalil, Kathleen Ziol-Guest.
2012. “The Importance of Early Childhood Poverty.” Social Indicators Research 108, no. 1: pp 87-98.

References  | 331

Eissa, Nada and Hilary W. Hoynes. 2005. “Behavioral Responses to Taxes:
Lessons from the EITC and Labor Supply.” Working Paper 11729.
Cambridge, MA: National Bureau of Economic Research.
Eissa, Nada and Jeffrey B. Liebman. 1996. “Labor Supply Response to the
Earned Income Tax Credit.” The Quarterly Journal of Economics
111, no. 2: 605-637.
Engelhardt, Gary V., and Jonathan Gruber. 2006. “Social Security and the
Evolution of Elderly Poverty.” In Public Policy and the Income
Distribution, edited by Alan J. Auerbach, David E. Card, and John
M. Quigley, pp. 259-287. New York: Russell Sage Foundation.
Fisher, Gordon M. 1992. “The Development and History of the Poverty
Thresholds.” Social Security Bulletin 55, no. 4: 3-14.
Fox, Liana, Irwin Garfinkel, Neeraj Kaushal, Jane Waldfogel, and Christopher Wimer. 2013. “Waging War on Poverty: Historical Trends in
Poverty Using the Supplemental Poverty Measure.” Working Paper
13-01. New York: Columbia Population Research Center.
Fremstad, Shawn. 2009. “Half in Ten: Why Taking Disability into Account
is Essential to Reducing Income Poverty and Expanding Economic
Inclusion.” Reports and Issue Briefs 2009-30. Washington: Center
for Economic and Policy Research.
Garces, Eliana, Duncan Thomas, and Janet Currie. 2002. “Longer-Term
Effects of Head Start.” American Economic Review 92, no. 4:
999–1012.
Gibbs, Chloe, Jens Ludwig, and Douglas L. Miller. 2013. “Head Start Origins
and Impacts.” In Legacies of the War on Poverty, edited by Martha
Bailey and Sheldon Danziger, pp. 39-65. New York: Russell Sage
Foundation Press.
Gottschalk, Peter and Sheldon Danziger. 2003. “Wage Inequality, Earnings
Inequality and Poverty in the U.S. Over the Last Quarter of the
Twentieth Century.” Working Papers in Economics 560. Boston
College Department of Economics.
Hanson, Kenneth. 2010. “The Food Assistance National Income-Output
Multiplier (FANIOM) Model and Stimulus Effects of SNAP.”
Economic Research Report No. 103. U.S. Department of Agriculture Economic Research Service.
Harrington, Michael. 1962. The Other America. New York: Simon and
Schuster.

Heckman, James J. and Dimitriy V. Masterov. 2007. “The Productivity
Argument for Investing in Young Children.” Applied Economic
Perspectives and Policy 29, no. 3: 446-493.
Heckman, James J., Seong Hyeok Moon, Rodrigo Pinto, Peter A. Savelyev,
and Adam Yavitz. 2009. “The Rate of Return to the High/Scope
Perry Preschool Program.” Working Paper No. 15471. Cambridge,
MA: National Bureau of Economic Research.
_____. 2010. “A New Cost-Benefit and Rate of Return Analysis for the
Perry Preschool Program: A Summary.” Working Paper No. 16180.
Cambridge, Mass.: National Bureau of Economic Research.
Heckman, James J., Rodrigo Pinto, Azeem M. Shaikh, and Adam Yavitz.
2011. “Inference with Imperfect Randomization: the Case of the
Perry Preschool Program.” Working Paper No. 16935. Cambridge,
MA: National Bureau of Economic Research.
Holzer, Harry J. 2007. “Collateral Costs: The Effects of Incarceration on the
Employment and Earnings of Young Workers.” IZA Discussion
Paper No. 3118.
Holzer, Harry J., Diane Whitmore Schanzenbach, Greg J. Duncan, and Jens
Ludwig. 2008. “The Economic Costs of Childhood Poverty in the
United States.” Journal of Children and Poverty 14, no. 1: 41-61.
Hotz, V. Joseph and John Karl Scholz. 2003. “The Earned Income Tax
Credit.” In Means-Tested Transfer Programs in the U.S., edited by
Robert A. Moffitt, pp. 141-198. Chicago: University of Chicago
Press.
Hoynes, Hilary W., Marianne E. Page, and Ann Huff Stevens. 2006. “Poverty
in America: Trends and Explanations.” Journal of Economic Perspectives 92, no. 3: 748-765.
Hoynes, Hilary W. and Diane W. Schanzenbach. 2009. “Consumption
Reponses to In-Kind Transfers: Evidence from the Introduction of
the Food Stamp Program.” American Economic Journal: Applied
Economics 1, no. 4: 109-139.
_____. 2012. “Work incentives and the Food Stamp Program.” Journal of
Public Economics 96, no. 1: 151-162.
Hoynes, Hilary W., Diane W. Schanzenbach, and Douglas Almond. 2013.
“Long Run Impacts of Childhood Access to the Safety Net.” Working
Paper 18535. Cambridge, MA: National Bureau of Economic
Research.

References  | 333

Ichino, Andrea, Loukas Karabarbounis, and Enrico Moretti. 2009. “The
Political Economy of Intergenerational Income Mobility.” IZA
Discussion Paper No. 4767.
Isaacs, Julia B. 2008. “Economic Mobility of Families across Generations.”
In Getting Ahead or Losing Ground: Economic Mobility in America,
edited by Julia Isaacs, Isabel Sawhill and Ron Haskins. Washington:
The Pew Charitable Trusts.
Isaacs, Julia B., Isabel Sawhill, and Ron Haskins. 2008. Getting Ahead or
Losing Ground: Economic Mobility in America. Washington: The
Pew Charitable Trusts.
Jacob, Brian A. and Jens Ludwig. 2012. “The Effects of Housing Assistance
on Labor Supply: Evidence from a Voucher Lottery.” American
Economic Review 102, no. 1: 272-304.
Jäntti, Markus, Bernt Bratsberg, Knut Røed, Oddbjørn Raaum, Robin
Naylor, Eva Österbacka, Anders Björklund, and Tor Eriksson. 2006.
“American Exceptionalism in a New Light: A Comparison of Intergenerational Earnings Mobility in the Nordic Countries, the United
Kingdom and the United States.” IZA Discussion Paper No. 1938.
Johnson, Rucker C. 2008. “Ever-increasing Levels of Parental Incarceration
and the Consequences for Children.” In Do Prisons Make us Safer?,
edited by S. Raphael and M. Stoll, 177-206. New York: Russell Sage
Foundation.
Korenman, Sanders and Dahlia Remler. 2013. “Rethinking Elderly Poverty:
Time for a Health Inclusive Poverty Measure?” Working Paper
18900. Cambridge, MA: National Bureau of Economic Research.
Lee, Chul-In and Gary Solon, 2009. “Trends in Intergenerational Income
Mobility,” The Review of Economics and Statistics, 91, no. 4: 766-772
Lee, David S. 1999. “Wage Inequality in the United States During the 1980s:
Rising Dispersion or Falling Minimum Wage?” Quarterly Journal of
Economics 114, no. 3: 977-1023.
Lemieux, Thomas. 2008. “The Changing Nature of Wage Inequality.”
Journal of Population Economics 21, no. 1: 21-48.
Liebman, Jeffrey B. 1998. “The Impact of the Earned Income Tax Credit
on Incentives and Income Distribution.” In Tax Policy and the
Economy, Volume 12, edited by James M. Poterba, pp. 83-120.
Cambridge, MA: National Bureau of Economic Research.

334 |  References

Ludwig, Jens and Douglas Miller. 2007. “Does Head Start Improve Children’s Life Chances? Evidence from a Regression Discontinuity
Design.” Quarterly Journal of Economics 122, no. 1: 159-208.
Matsudaira, Jordan D. and Rebecca M. Blank. 2013. “The Impact of Earnings Disregards on the Behavior of Low-Income Families.” Journal
of Policy Analysis and Management 33, no. 1: 7-35.
Mayer, Susan E. and Leonard M. Lopoo. 2008. “Government Spending and
Intergenerational Mobility.” Journal of Public Economics 92, no. 1-2:
139-158.
Mayer, Susan E. 1997. What Money Can’t Buy: Family Income and Children’s Life Chances. Cambridge, MA: Harvard University Press.
Meyer, Bruce D., Wallace K. C. Mok, and James X. Sullivan. 2009. “The
Underreporting of Transfers in Household Surveys: Its Nature and
Consequences.” Working Paper 15181. Cambridge, MA: National
Bureau of Economic Research.
Meyer, Bruce D. and Dan T. Rosenbaum. 2001. “Welfare, the Earned
Income Tax Credit, and the Labor Supply of Single Mothers.” Quarterly Journal of Economics 116, no. 3: 1063-1114.
Meyer, Bruce D. and James X. Sullivan. 2003. “Measuring the Well-Being
of the Poor Using Income and Consumption.” Journal of Human
Resources 38, Supplement: 1180-1220.
_____. 2012a. “Winning the War: Poverty from the Great Society to the
Great Recession.” Brookings Papers on Economic Activity 45, no. 2:
133-200.
_____. 2012b. “Identifying the Disadvantaged: Official Poverty, Consumption Poverty, and the New Supplemental Poverty Measure.” Journal
of Economic Perspectives 26, no. 3: 111-136.
_____. 2013. “Winning the War: Poverty from the Great Society to the
Great Recession.” Working Paper 18718. Cambridge, MA: National
Bureau of Economic Research.
Mishel, Lawrence, Josh Bivens, Elise Gould, and Heidi Shierholz. The State
of Working America, 12th Edition. A forthcoming Economic Policy
Institute book. Ithaca, NY: Cornell University Press.
Misra, Joya, Michelle Budig, and Irene Boeckmann. 2011. “Work-Family
Policies and the Effects of Children on Women’s Employment
Hours and Wages.” Community, Work and Family 14, no. 2:
139-157.

References  | 335

Nicholson-Crotty, Sean and Kenneth J. Meier. 2003. “Crime and Punishment: The Politics of Federal Criminal Justice Sanctions.” Political
Research Quarterly 56, no. 2: 119-126.
Orshansky, Mollie. 1965. “Counting the Poor: Another Look at the Poverty
Profile.” Social Security Bulletin 28, no. 1: 3-29.
Peri, Giovanni. 2013. “Immigrant Workers, Native Poverty and Labor
Market Competition.” Policy Brief, Center for Poverty Research 1,
no. 3.
Peters, Alan H. and Peter S. Fisher. 2002. “State Enterprise Zones: Have
They Worked?” Kalamazoo, MI: W.E. Upjohn Institute for Employment Research Press.
Piketty, Thomas and Emmanuel Saez. 2003. “Income Inequality in the
United States, 1913-1998.” Quarterly Journal of Economics 118, no.
1: 1-39.
Raphael, Steven. 2007. “Early Incarceration Spells and the Transition to
Adulthood.” In The Price of Independence: The Economics of Early
Adulthood, edited by Sheldon Danziger and Cecilia Elena Rouse, pp.
278-306. New York: Russell Sage Foundation.
Reed, Deborah and Maria Cancian. 2001. “Sources of Inequality: Measuring
the Contributions of Income Sources to Rising Family Income
Inequality.” Review of Income and Wealth 47, no. 3: 321-333.
Sawhill, Isabel V. and John E. Morton. 2007. “Economic Mobility: Is
the American Dream Alive and Well?” Washington: Economic
Mobility Project, Pew Charitable Trusts.
Schur, Lisa A., Douglas L. Kruse, and Peter Blanck. 2013. People with Disabilities: Sidelined or Mainstreamed? Cambridge, England: Cambridge
University Press.
Schweinhart, Lawrence J., Jeanne Montie, Zongping Xiang, W. Steven
Barnett, Clive R. Belfield, and Milagros Nores. 2005. Lifetime Effects:
The High/Scope Perry Preschool Study Through Age 40. Monographs
of the High/Scope Educational Research Foundation. Ypsilanti, MI:
High/Scope Press.
Sen, Amartya. 2009. The Idea of Justice. London: Allen Lane.
Singh, Gopal K. and Michael D. Kogan. 2007. “Persistent Socioeconomic
Disparities in Infant, Neonatal, and Postneonatal Mortality Rates in
the United States, 1969-2001.” Pediatrics 119, no. 4: 928-939.

336 |  References

Sharkey, Patrick. 2009. “Neighborhoods and the Black-White Mobility Gap.”
Washington: Economic Mobility Project, Pew Charitable Trusts.
She, Peiyun and Gina A. Livermore. 2007. “Material Hardship, Poverty, and
Disability Among Working‐Age Adults.” Social Science Quarterly
88, no. 4: 970-989.
Sherman, Arloc. 2013. “Official Poverty Measure Masks Gains Made Over
Last 50 Years.” Washington: Center on Budget and Policy Priorities.
http://www.cbpp.org/files/9-13-13pov.pdf.
Short, Kathleen. 2012. “The Research Supplemental Poverty Measure: 2011”
Current Population Reports.
_____. 2013. “The Research Supplemental Poverty Measure: 2012” Current
Population Reports.
Shroder, Mark. 2010. “Housing Subsidies and Work Incentives.” MPRA
Paper 26019, University Library of Munich, Germany.
Solon, Gary. 2002. “Cross-Country Differences in Intergenerational Earnings Mobility.” Journal of Economic Perspectives 16, no. 3: 59– 66.
SRI International. 1983. “Final Report of the Seattle/Denver Income Maintenance Experiment. Vol 1: Design and Results.” Menlo Park, CA.
(SSA) Social Security Administration. 2012. Income of the Population 55 or
Older, 2010. SSA Publication No. 13-11871.
Western, Bruce. 2002. “The Impact of Incarceration on Wage Mobility and
Inequality.”
American Sociological Review 67: 526-46.
Western, Bruce and Becky Pettit. 2010. “Incarceration and Social Inequality.”
Daedalus 139: 8-19.
Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and
Jane Waldfogel. 2013. “Trends in Poverty with an Anchored
Supplemental Poverty Measure.” Working Paper 1-25. New York:
Columbia Population Research Center.

Chapter 7
Abdulkadiroglu, Atila, Joshua D Angrist, Susan M Dynarski, Thomas J
Kane, Parag A Pathak. 2011. “Accountability and Flexibility in
Public Schools: Evidence from Boston’s Charters and Pilots.” The
Quarterly Journal of Economics 126, no. 2: 699-748.

References  | 337

Angrist, Joshua D., and Jörn-Steffen Pischke. 2008. Mostly Harmless Econometrics: An Empiricist’s Companion. Princeton, NJ: Princeton
University Press.
Angrist, Joshua. D., and Alan B. Krueger. 1999. “Empirical strategies in labor
economics.” In Handbook of Labor Economics, edited by O. Ashenfelter and D. Card, pp.1277-1366. Vol 3, Elsevier.
Baicker, Katherine, Amy Finkelstein, Jae Song, and Sarah Taubman. 2013.
“The Impact of Medicaid on Labor Force Activity and Program
Participation: Evidence from the Oregon Health Insurance Experiment.” Working Paper. Cambridge, MA: National Bureau of
Economic Research (October).
Benus, J. T. Shen, S. Zhang, M. Chan and B. Hansen. 2009. “Growing
America Through Entrepreneurship: Final Evaluation of Project
GATE.” Columbia, MD: IMPAQ International LLC.
Benus, J., Poe-Yamagata, E., Wang, Y., & Blass, E. 2008. “Reemployment and
Eligibility Assessment (REA) Study FY 2005 Initiative.” Columbia,
MD: IMPAQ International LLC.
Berk, Richard A., and Peter H. Rossi. 1998. Thinking about Program Evaluation 2. 2nd ed. London: Sage Publications, Inc.
Bettinger, E.P., B.T. Long, P. Oreopoulos and L. Sanbonmatsu. 2009. “The
Role Of Simplification and Information In College Decisions:
Results From The H&R Block FAFSA Experiment.” Working Paper
15361. Cambridge, MA: National Bureau of Economic Research
(September).
Burtless, G. 1995. (The case for randomized field trials in economic and
policy research.) The Journal of Economic Perspectives 9, no. 2
(Spring): 63-84.
Cappers, Peter, Charles Goldman, and Annika Todd. 2013. “Smart Grid
Investment Grant Consumer Behavior Study Analysis: Summary
of Utility Studies.” Environmental Energy Technologies Division,
Lawrence Berkeley National Laboratory.
Card, David, Carlos Dobkin, and Nicole Maestas. 2009. “Does Medicare
Save Lives?” The Quarterly Journal of Economics 124, no. 2: 597-636.
Coalition for Evidence-Based Policy. 2012. “Rigorous Program Evaluations
on a Budget: How Low-Cost Randomized Controlled Trials are
Possible in Many Areas of Social Policy.” Washington, DC (March).

338 |  References

_____. 2013. “Practical Evaluation Strategies for Building a Body of ProvenEffective Social Programs.” Washington, DC (October).
Comey, Jennifer, Kaitlin Franks, Lesley Freiman, Christopher Hayes, Reed
Jordan, Peter A. Tatian, and Mary K. Winkler. 2013. “Measuring
Performance: A Guidance Document for Promise Neighborhoods
on Collecting Data and Reporting Results.” Washington, DC: The
Urban Institute (February).
Council Of Economic Advisers And Department Of Labor. 2013. “The
Economic Benefits of Extending Unemployment Insurance”
(December).
Culhane, D. P., S. Metraux, J.M. Park, M.A. Schretzman, M. A., and J.
Valente. 2007. “Testing a Typology of Family Homelessness Based
on Patterns of Public Shelter Utilization in Four U.S. Jurisdictions:
Implications for Policy and Program Planning.”  Housing Policy
Debate, 18(1), 1-28.
Department of Education. 2012. Education Improvement Programs: Fiscal
Year 2012 Budget Request. http://www2.ed.gov/about/overview/
budget/budget12/justifications/d-eip.pdf
_____. 2013a. Investing in Innovation Fund (i3) Program Guidance and
Frequently Asked Questions (FAQs). http://www2.ed.gov/programs/
innovation/faq.html
_____. 2013b. Promise Neighborhoods: Purpose. http://www2.ed.gov/
programs/promiseneighborhoods/index.html
_____. 2013c. PROMISE: Promoting Readiness of Minors in Supplemental
Security Income. http://www2.ed.gov/about/inits/ed/promise/index.
html#about
Department of Labor. 2011. Employment and Training Administration
Notice of Availability of Funds and Solicitation for Grant Applications for Workforce Innovation Fund Grants. SGA/DFA PY-11-05
at http://www.doleta.gov/workforce_innovation/applicant_information.cfm
_____. 2013. Congressional Budget Justification: Employment and Training
Administration. http://www.dol.gov/dol/budget/2013/PDF/CBJ2013-V1-05.pdf
Departments of Labor, Health and Human Services, and Education, and
Related Agencies Appropriations Act, 2006, Title V, Part D (P.L.
109-149). http://www2.ed.gov/programs/teacherincentive/legislation.html)
References  | 339

Dobbie, Will and Roland G. Fryer. 2011. “Are High Quality Schools Enough
to Increase Achievement Among the Poor? Evidence from the
Harlem Children’s Zone.” American Economic Journal: Applied
Economics 3 (July):158-187.
Finkelstein, Amy, Sarah Taubman,  Bill Wright, Mira Bernstein, Jonathan
Gruber, Joseph P. Newhouse, Heidi Allen, Katherine Baicker, and
the Oregon Health Study Group. 2012. “The Oregon Health Insurance Experiment: Evidence from the First Year.” Quarterly Journal
of Economics 127(3): 1057-1106 (August).
General Accounting Office. 2011. Performance Measurement and Evaluation: Definitions and Relationships. GAO-11-646SP, May.
_____. 2012. “Designing Evaluations: 2012 Revision.” U.S. Government
Accountability Office, GAO-12-208G, January.
Grossman, J.B., and J.E. Rhodes. 2002. “The test of time: Predictors and
effects of duration in youth mentoring programs.” American
Journal of Community Psychology (30), 199-219.
Jacob, Brian A., Jens Ludwig. 2012. “The Effects of Housing Assistance
On Labor Supply: Evidence From A Voucher Lottery.” American
Economic Review 102(1): 272–304.
Holland, Paul W. 1986. “Statistics and Causal Inference.” Journal of the
American Statistical Association 81, no. 396: 945-960.
Imbens, Guido W. and Jeffrey M. Woolridge. 2009. “Recent Developments
in the Econometrics of Program Evaluation.” Journal of Economic
Literature 47, no. 1: 5-86.
Imbens, Guido W. 2010. “Better LATE Than Nothing: Some Comments
on Deaton (2009) and Heckman and Urzua (2009).” Journal of
Economic Literature 48 (June 2010): 399-423.
Ludwig, J. and D. L. Miller. 2007. “Does Head Start improve children’s
life chances? Evidence From a Regression Discontinuity Design.”
Quarterly Journal of Economics 122 no. 1: 159-208.
Manzi, Jim. 2012. Uncontrolled: The Surprising Payoff of Trial-and-Error for
Business, Politics, and Society. New York: Basic Books.
Millensky, Megan, et al. 2011. “Staying on Course: Three-Year Results of the
National Guard Youth ChalleNGe Evaluation. MDRC. http://www.
mdrc.org/sites/default/files/full_510.pdf
National Institute of Justice. 2012. “Swift and Certain” Sanctions in Probation Are Highly Effective: Evaluation of the Hope Program. http://
340 |  References

www.nij.gov/topics/corrections/community/drug-offenders/
hawaii-hope.htm
National Research Council and Institute of Medicine. 2009.  Preventing
Mental, Emotional, and Behavioral Disorders Among Young People:
Progress and Possibilities. Committee on Prevention of Mental
Disorders and Substance Abuse Among Children, Youth and
Young Adults: Research Advances and Promising Interventions.
Mary Ellen O’Connell, Thomas Boat, and Kenneth E. Warner,
Editors. Board on Children, Youth, and Families, Division of
Behavioral and Social Sciences and Education. Washington, DC:
The National Academies Press. Online at http://www.nap.edu/
catalog.php?record_id=12480.
Office of Management and Budget. 2009a. Analytical Perspectives: Budget of
the U.S. Government, February. At http://www.gpo.gov/fdsys/pkg/
BUDGET-2010-PER/pdf/BUDGET-2010-PER.pdf
_____. 2009b. Building Rigorous Evidence to Drive Policy. OMB Blogpost,
http://www.whitehouse.gov/omb/blog/09/06/08/
BuildingRigorousEvidencetoDrivePolicy
_____. 2009c. Increased Emphasis on Program Evaluations. Memorandum
M-10-01, October 7, 2009.
_____. 2010. Analytical Perspectives: Budget of the U.S. Government, at http://
www.gpo.gov/fdsys/pkg/BUDGET-2011-PER/pdf/BUDGET2011-PER.pdf
_____. 2011. Analytical Perspectives: Budget of the U.S. Government,
February 14, 2011 at http://www.gpo.gov/fdsys/pkg/BUDGET2012-PER/pdf/BUDGET-2012-PER.pdf
_____. 2012. Analytical Perspectives: Budget of the U.S. Government,
February 13, 2012 at http://www.gpo.gov/fdsys/pkg/BUDGET2013-PER/pdf/BUDGET-2013-PER.pdf
_____. 2013a. Creating a 21st Century Government, Budget of the United
States Government, Fiscal Year 2014 http://www.whitehouse.gov/
sites/default/files/omb/budget/fy2014/assets/21st_century.pdf
_____. 2013b. Next Steps in the Evidence and Innovation Agenda. Memorandum M-13-17, July 26, 2013.
_____. 2014. Guidance for Providing and Using Administrative Data for
Statistical Purposes. Memorandum M-14-06, February 14, 2014.

References  | 341

Paulsell, Diane, S. Avellar, E. Sama Martin, P. Del Grosso. 2010. “Home
visiting Evidence of Effectiveness Review: Executive Summary.”
Washington, DC:Mathematica Policy Research.
Perez-Arce, Francisco, et al. 2012. “A Cost-Benefit Analysis of the National
Guard Youth ChalleNGe Program.” RAND Corporation. http://
www.rand.org/pubs/technical_reports/TR1193.html#key-findings
Poe-Yamagata, E., Benus, J., Bill, N., Carrington, H., Michaelides, M., &
Shen, T. 2011. “Impact of the Reemployment and Eligibility Assessment (REA) Initiative.” Columbia, MD: IMPAQ International LLC.
Rivkin, Steven G., Eric A. Hanushek and John F. Kain. 2005. “Teachers,
Schools, and Academic Achievement.” Econometrica, vol. 73 (2)
(March): 417-458.
Rothstein, Jesse. 2011. “Unemployment Insurance and Job Search in the
Great Recession.” Brookings Papers on Economic Activity Fall 2011:
143-213.
Shadish, William R., Thomas D. Cook, and Donald Thomas Campbell.
2002. Experimental and Quasi-Experimental Designs for Generalized
Causal Inference. Boston, MA: Houghton Mifflin.
Schlafer, Rebecca J., Poehlmann, Julie, Coffino, Brianna, Henneman, Ashley
(2009). “Mentoring Children with Incarcerated Parents: Implications for Research, Practice, and Policy.” Family Relations (58),
507-519.
Schochet, Peter Z., John Burghardt, and Sheena McConnell. 2008. “Does Job
Corps Work? Impact Findings from the National Job Corps Study.”
American Economic Review 98:5, 1864-1886. http://www.jstor.org/
stable/29730155?seq=2
Shadish, William R., Thomas D. Cook, and Donald Thomas Campbell.
2002. Experimental and Quasi-Experimental Designs for Generalized
Causal Inference. Boston, MA: Houghton Mifflin.
Shrank, W. 2013. “The Center for Medicare and Medicaid Innovation’s
Blueprint for Rapid-Cycle Evaluation of New Care and Payment
Models.” Health Affairs 32(4):807-812.
Social Security Administration. 2013. Solicitation: PROMISE Evaluation
Statement of Work. SSA-RFP-13-0018, July 16, 2013. https://www.
fedconnect.net/FedConnect/PublicPages/PublicSearch/Public_
Opportunities.aspx

342 |  References

Stock, James H., and Mark Watson. 2010. Introduction to Econometrics. 3rd
ed. Boston, MA: Addison-Wesley.
St.Pierre, R., A. Ricciuti, F. Tao, C. Creps, J. Swartz, W. Lee, A. Parsad and
T. Rimdzius. 2003. “Third National Even Start Evaluation: Program
Impacts and Implications for Improvement.” Abt Associates, Inc.
and U.S. Department of Education.
United States Interagency Council on Homelessness. 2013a. “US Labor
Department awards nearly $24 million in Pay for Success grants.”
September 26, 2013, available at http://usich.gov/media_center/
news/us_labor_department_awards_nearly_24_million_in_pay_
for_success_grants (accessed November 24, 2013).
United States Interagency Council on Homelessness, 2013b. “Opening
Doors,” available at http://usich.gov/opening_doors/ (accessed
November 24, 2013).
Walker, Robert, Hoggart, Lesley, and Hamilton Gayle. 2006. “Making
Random Assignment Happen: Evidence from the UK Employment
Retention and Advancement (ERA) Demonstration.” Policy Studies
Institute (PSI), the Office for National Statistics (ONS), the Institute for Fiscal Studies (IFS) and MDRC for the UK Department for
Work and Pensions.
Wansink, Brian, David Just, and Laura Smith. 2011. “Move the Fruit: Putting
Fruit in New Bowls and New Places Doubles Lunchroom Sales.”
Journal of Nutrition Education and Behavior 43:4.

References  | 343

A P P E N D I X

A

REPORT TO THE PRESIDENT
ON THE ACTIVITIES OF THE
COUNCIL OF ECONOMIC
ADVISERS DURING 2013

letter of transmittal
Council of Economic Advisers
Washington, D.C., December 31, 2013
Mr. President:
The Council of Economic Advisers submits this report on its activities
during calendar year 2013 in accordance with the requirements of the
Congress, as set forth in section 10(d) of the Employment Act of 1946 as
amended by the Full Employment and Balanced Growth Act of 1978.
Sincerely yours,
	
	
	

Jason Furman, Chairman
Betsey Stevenson, Member
James H. Stock, Member

Activities of the Council of Economic Advisers During 2013  |  347

Council Members and Their Dates of Service
Name 	

Position 	

Oath of office date 	 Separation date

Edwin G. Nourse	
Leon H. Keyserling	
	
	
John D. Clark	
	
Roy Blough	
Robert C. Turner	
Arthur F. Burns	
Neil H. Jacoby	
Walter W. Stewart	
Raymond J. Saulnier	
	
Joseph S. Davis	
Paul W. McCracken	
Karl Brandt	
Henry C. Wallich	
Walter W. Heller	
James Tobin	
Kermit Gordon	
Gardner Ackley	
	
John P. Lewis	
Otto Eckstein	
Arthur M. Okun	
	
James S. Duesenberry	
Merton J. Peck	
Warren L. Smith	
Paul W. McCracken	
Hendrik S. Houthakker	
Herbert Stein	
	
Ezra Solomon	
Marina v.N. Whitman	
Gary L. Seevers	
William J. Fellner	
Alan Greenspan 	
Paul W. MacAvoy	
Burton G. Malkiel	
Charles L. Schultze	
William D. Nordhaus	
Lyle E. Gramley	
George C. Eads	
Stephen M. Goldfeld	
Murray L. Weidenbaum	

Chairman	
Vice Chairman	
Acting Chairman	
Chairman	
Member	
Vice Chairman	
Member	
Member	
Chairman	
Member	
Member	
Member	
Chairman	
Member	
Member	
Member	
Member	
Chairman	
Member	
Member	
Member	
Chairman	
Member	
Member	
Member	
Chairman	
Member	
Member	
Member	
Chairman	
Member	
Member	
Chairman	
Member	
Member	
Member	
Member	
Chairman 	
Member	
Member	
Chairman	
Member	
Member	
Member	
Member	
Chairman	

August 9, 1946	
August 9, 1946
November 2, 1949
May 10, 1950	
August 9, 1946
May 10, 1950	
June 29, 1950	
September 8, 1952	
March 19, 1953	
September 15, 1953	
December 2, 1953	
April 4, 1955
December 3, 1956	
May 2, 1955	
December 3, 1956	
November 1, 1958	
May 7, 1959	
January 29, 1961	
January 29, 1961	
January 29, 1961	
August 3, 1962
November 16, 1964	
May 17, 1963	
September 2, 1964	
November 16, 1964
February 15, 1968	
February 2, 1966	
February 15, 1968	
July 1, 1968	
February 4, 1969	
February 4, 1969	
February 4, 1969
January 1, 1972	
September 9, 1971	
March 13, 1972	
July 23, 1973	
October 31, 1973	
September 4, 1974	
June 13, 1975	
July 22, 1975	
January 22, 1977	
March 18, 1977	
March 18, 1977	
June 6, 1979	
August 20, 1980	
February 27, 1981	

348  |  Appendix A

November 1, 1949
January 20, 1953
February 11, 1953
August 20, 1952
January 20, 1953
December 1, 1956
February 9, 1955
April 29, 1955
January 20, 1961
October 31, 1958
January 31, 1959
January 20, 1961
January 20, 1961
November 15, 1964
July 31, 1962
December 27, 1962
February 15, 1968
August 31, 1964
February 1, 1966
January 20, 1969
June 30, 1968
January 20, 1969
January 20, 1969
December 31, 1971
July 15, 1971
August 31, 1974
March 26, 1973
August 15, 1973
April 15, 1975
February 25, 1975
January 20, 1977
November 15, 1976
January 20, 1977
January 20, 1981
February 4, 1979
May 27, 1980
January 20, 1981
January 20, 1981
August 25, 1982

Council Members and Their Dates of Service
Name 	

Position 	

Oath of office date 	 Separation date

William A. Niskanen	
Jerry L. Jordan	
Martin Feldstein	
William Poole	
Beryl W. Sprinkel	
Thomas Gale Moore	
Michael L. Mussa	
Michael J. Boskin	
John B. Taylor	
Richard L. Schmalensee	
David F. Bradford	
Paul Wonnacott	
Laura D’Andrea Tyson	
Alan S. Blinder	
Joseph E. Stiglitz	
	
Martin N. Baily 	
Alicia H. Munnell	
Janet L. Yellen	
Jeffrey A. Frankel	
Rebecca M. Blank	
Martin N. Baily	
Robert Z. Lawrence	
Kathryn L. Shaw	
R. Glenn Hubbard	
Mark B. McClellan	
Randall S. Kroszner	
N. Gregory Mankiw	
Kristin J. Forbes	
Harvey S. Rosen	
	
Ben S. Bernanke	
Katherine Baicker	
Matthew J. Slaughter	
Edward P. Lazear	
Donald B. Marron	
Christina D. Romer	
Austan D. Goolsbee	
	
Cecilia Elena Rouse	
Katharine G. Abraham	
Carl Shapiro	
Alan B. Krueger	
James H. Stock	
Jason Furman	
Betsey Stevenson	

Member	
Member	
Chairman	
Member	
Chairman	
Member	
Member	
Chairman	
Member	
Member	
Member	
Member	
Chair	
Member	
Member	
Chairman	
Member	
Member	
Chair	
Member	
Member	
Chairman	
Member	
Member	
Chairman	
Member	
Member	
Chairman	
Member	
Member	
Chairman	
Chairman	
Member	
Member	
Chairman	
Member	
Chair	
Member	
Chairman	
Member	
Member	
Member	
Chairman	
Member	
Chairman	
Member	

June 12, 1981	
July 14, 1981	
October 14, 1982	
December 10, 1982	
April 18, 1985	
July 1, 1985	
August 18, 1986	
February 2, 1989	
June 9, 1989	
October 3, 1989	
November 13, 1991	
November 13, 1991	
February 5, 1993	
July 27, 1993	
July 27, 1993
June 28, 1995	
June 30, 1995	
January 29, 1996	
February 18, 1997	
April 23, 1997	
October 22, 1998	
August 12, 1999	
August 12, 1999	
May 31, 2000	
May 11, 2001	
July 25, 2001	
November 30, 2001	
May 29, 2003	
November 21, 2003	
November 21, 2003	
February 23, 2005	
June 21, 2005	
November 18, 2005	
November 18, 2005	
February 27, 2006	
July 17, 2008	
January 29, 2009	
March 11, 2009	
September 10, 2010	
March 11, 2009	
April 19, 2011	
April 19, 2011	
November 7, 2011	
February 7, 2013	
August 4, 2013	
August 6, 2013

March 30, 1985
July 31, 1982
July 10, 1984
January 20, 1985
January 20, 1989
May 1, 1989
September 19, 1988
January 12, 1993
August 2, 1991
June 21, 1991
January 20, 1993
January 20, 1993
April 22, 1995
June 26, 1994
February 10, 1997
August 30, 1996
August 1, 1997
August 3, 1999
March 2, 1999
July 9, 1999
January 19, 2001
January 12, 2001
January 19, 2001
February 28, 2003
November 13, 2002
July 1, 2003
February 18, 2005
June 3, 2005
June 10, 2005
January 31, 2006
July 11, 2007
March 1, 2007
January 20, 2009
January 20, 2009
September 3, 2010
August 5, 2011
February 28, 2011
April 19, 2013
May 4, 2012
August 2, 2013

Activities of the Council of Economic Advisers During 2013  |  349

Report to the President
on the Activities of the
Council of Economic Advisers
During 2013
The Council of Economic Advisers was established by the Employment
Act of 1946 to provide the President with objective economic analysis and
advice on the development and implementation of a wide range of domestic
and international economic policy issues. The Council is governed by a
Chairman and two Members. The Chairman is appointed by the President
and confirmed by the United States Senate. The Members are appointed by
the President.

The Chairman of the Council
Jason Furman was confirmed by the U.S. Senate on August 1, 2013.
Prior to this role, Furman served as Assistant to the President for Economic
Policy and the Principal Deputy Director of the National Economic Council.
From 2007 to 2008 Furman was a Senior Fellow in Economic Studies
and Director of the Hamilton Project at the Brookings Institute. Previously,
he served as a Staff Economist at the Council of Economic Advisers, a
Special Assistant to the President for Economic Policy at the National
Economic Council under President Clinton and Senior Adviser to the Chief
Economist and Senior Vice President of the World Bank. Furman was the
Economic Policy Director for Obama for America. Furman has also served
as Visiting Scholar at NYU’s Wagner Graduate School of Public Service, a
visiting lecturer at Yale and Columbia Universities, and a Senior Fellow at
the Center on Budget and Policy Priorities.
Alan B. Krueger resigned as Chairman on August 2, 2013 to return
to Princeton University, where he is the Bendheim Professor of Economics
and Public Affairs.

Activities of the Council of Economic Advisers During 2013  |  351

The Members of the Council
Betsey Stevenson was appointed by the President on August 6, 2013.
She is on leave from the University of Michigan’s Gerald R. Ford School
of Public Policy and the Economics Department where she is an Associate
Professor of Public Policy and Economics. She served as the Chief Economist
of the US Department of Labor from 2010 to 2011.
James H. Stock was appointed by the President on February 7, 2013.
He served as Chief Economist of the Council of Economic Advisers from
September 12, 2012 until then. Dr. Stock is on leave from Harvard University,
where he is the Harold Hitchings Burbank Professor of Political Economy.
Dr. Stock served as the Chair of the Harvard University Department of
Economics from 2006 to 2009.
Katharine G. Abraham resigned as Member of the Council on April 19,
2013 to return to the University of Maryland, where she is a professor in the
Joint Program in Survey Methodology and faculty associate in the Maryland
Population Research Center.

Areas of Activities
A central function of the Council is to advise the President on all
economic issues and developments. In the past year, as in the four previous
years, advising the President on policies to spur economic growth and job
creation, and evaluating the effects of the policies on the economy, have been
a priority.
The Council works closely with various government agencies,
including the National Economic Council, the Office of Management and
Budget, White House senior staff, and other officials and engages in discussions on numerous policy matters. In the area of international economic
policy, the Council coordinates with other units of the White House, the
Treasury Department, the State Department, the Commerce Department,
and the Federal Reserve on matters related to the global financial system.
Among the specific economic policy areas that received attention in
2013 were: housing policies; health care cost growth and the Affordable Care
Act; individual and corporate taxation; college affordability and ratings;
regional development; the economic cost of carbon pollution; renewable fuel
standards; energy policy; intellectual property and innovation; infrastructure investment; regulatory measures; trade policies; poverty and income
inequality; unemployment insurance and the minimum wage; labor force
participation; job training; and foreign direct investment. The Council

352  |  Appendix A

also worked on several issues related to the quality of the data available for
assessing economic conditions.
The Council prepares for the President, the Vice President, and the
White House senior staff a daily economic briefing memo analyzing current
economic developments and almost-daily memos on key economic data
releases. Chairman Furman also presents a monthly briefing on the state
of the economy and the Council’s energy analysis to senior White House
officials.
The Council, the Department of Treasury, and the Office of
Management and Budget—the Administration’s economic “troika”—
are responsible for producing the economic forecasts that underlie the
Administration’s budget proposals. The Council initiates the forecasting
process twice each year, consulting with a wide variety of outside sources,
including leading private sector forecasters and other government agencies.
The Council was an active participant in the trade policy process,
participating in the Trade Policy Staff Committee and the Trade Policy
Review Group. The Council provided analysis and opinions on a range of
trade-related issues involving the enforcement of existing trade agreements,
reviews of current U.S. trade policies, and consideration of future policies. The Council also participated on the Trade Promotion Coordinating
Committee, helping to examine the ways in which exports may support
economic growth in the years to come. In the area of investment and security, the Council participated on the Committee on Foreign Investment in
the United States (CFIUS), reviewing individual cases before the committee.
The Council is a leading participant in the Organisation for Economic
Co-operation and Development (OECD), an important forum for economic
cooperation among high-income industrial economies. The Council coordinated and oversaw the OECD’s review of the U.S. economy. Chairman
Furman is chairman of the OECD’s Economic Policy Committee, and
Council Members and staff participate actively in working-party meetings
on macroeconomic policy and coordination and contribute to the OECD’s
research agenda.
The Council issued a wide range of reports in 2013 and early 2014.
In June, the Council released a report assessing the economic benefits of
reforming US’s “broken immigration system.” In October, the Council
released a report analyzing the negative impact that the government shutdown and debt limit brinksmanship had on the economy. Also in October,
the Council released a report describing why the US provides such an attractive investment climate for firms across the globe and how this is beneficial
to the US. In November, the Council released a report analyzing the recent
health care cost trends in addition to the contributions the Affordable Care
Activities of the Council of Economic Advisers During 2013  |  353

Act has had on reducing health care cost growth. In December, the Council
worked with the Department of Labor to study the benefits of extending
unemployment insurance. In January 2014, the Council released a progress
report on the “War on Poverty” that Lyndon B. Johnson declared 50 years
prior. In February 2014, the Council transmitted the final report on the
fifth Anniversary of the Recovery Act to Congress. All of the aforementioned reports can be found on the Council’s website, and some of them are
incorporated into this annual report as well. (http://www.whitehouse.gov/
administration/eop/cea/factsheets-reports.)
The Council continued its efforts to improve the public’s understanding of economic developments and of the Administration’s economic
policies through briefings with the economic and financial press, speeches,
discussions with outside economists, and regular updates on major data
releases and postings of CEA’s Reports on the White House and CEA blogs.
The Chairman and Members also regularly met to exchange views on the
economy with the Chairman and Members of the Board of Governors of the
Federal Reserve System.

Public Information
The Council’s annual Economic Report of the President is an important vehicle for presenting the Administration’s domestic and international
economic policies. It is available for purchase through the Government
Printing Office, and is viewable on the Internet at www.gpo.gov/erp.
The Council frequently prepared reports and blog posts in 2013, and
the Chairman and Members gave numerous public speeches. The reports,
posts and texts of speeches are available at the Council’s website, www.
whitehouse.gov/cea. Finally, the Council published the monthly Economic
Indicators, which is available online at www.gpo.gov/economicindicators.

The Staff of the Council of Economic Advisers
The staff of the Council consists of the senior staff, senior economists,
economists, staff economists, research economists, a research assistant, and
the administrative and support staff. The staff at the end of 2013 was:

Senior Staff
Jessica Schumer 	������������������������������������Chief of Staff
Steven N. Braun	������������������������������������Director of Macroeconomic
Forecasting
Alexander G. Krulic	������������������������������General Counsel

354  |  Appendix A

Adrienne Pilot 	��������������������������������������Director of Statistical Office
Archana Snyder	������������������������������������Director of Finance and
Administration

Senior Economists
David J. Balan	����������������������������������������Industrial Organization, Technology,
Health
Marco Cagetti	����������������������������������������Macroeconomics
Jane K. Dokko	����������������������������������������Housing
Matthew Fiedler	������������������������������������Health
Tracy M. Gordon	����������������������������������Tax, Budget
Douglas Kruse	���������������������������������������Labor, Disability
Jordan D. Matsudaira 	��������������������������Labor, Education
Cynthia J. Nickerson	����������������������������Agriculture, Environment, Evaluation
Ronald J. Shadbegian	����������������������������Energy, Environment
Kenneth A. Swinnerton	������������������������International

Staff Economists
Zachary Y. Brown 	��������������������������������Labor, Health Housing
John Coglianese	������������������������������������Labor, Public Finance,
Macroeconomics
Kevin Rinz	����������������������������������������������Labor, Education

Research Economists
Philip K. Lambrakos 	����������������������������Macroeconomics
Cordaye T. Ogletree	������������������������������Energy, Environment and
International
Krista Ruffini 	����������������������������������������Health
Rudy Telles Jr. 	��������������������������������������International, Technology

Research Assistants
Brendan Mochoruk	������������������������������Tax, Budget
Jenny Shen	����������������������������������������������Energy, Environment
David N. Wasser 	����������������������������������Labor, Immigration, Education

Activities of the Council of Economic Advisers During 2013  |  355

Statistical Office
The Statistical Office gathers, administers, and produces statistical
information for the Council. Duties include preparing the statistical appendix
to the Economic Report of the President and the monthly publication Economic
Indicators. The staff also creates background materials for economic analysis
and verifies statistical content in Presidential memoranda. The Office serves
as the Council’s liaison to the statistical community.
Brian A. Amorosi 	��������������������������������Statistical Analyst
Wenfan Chen	����������������������������������������Economic Statistician

Office of the Chairman and Members
Andrea Taverna	������������������������������������Special Assistant to the Chairman
Natasha Lawrence 	��������������������������������Special Assistant to the Members
Matthew L. Aks 	������������������������������������Special Assistant to the Chairman
and Research Economist

Administrative Office
The Administrative Office provides general support for the Council’s
activities. This includes financial management, human resource management, travel, operations of facilities, security, information technology, and
telecommunications management support.
Doris T. Searles	��������������������������������������Administrative and Information
Management Specialist
Anna Y. Lee	��������������������������������������������Financial Systems Analyst

Interns
Student interns provide invaluable help with research projects, day-today operations, and fact-checking. Interns during the year were: Katherine
Allsop, Brian Bendett, Rachel Burdick, Katherine Carpenter, Benjamin Clark,
Brian Collopy, Christopher Gum, Thomas Hedin, Ashwin Kambhampati,
Michael Kennedy, Samsun Knight, Katelyn Lamson, Catherine Mahoney,
Brennan Mange, David McCarthy, Elliot Melaney, J Mintzmyer, Ivan
Mogensen, Benjamin Murray, Andrew Olenski, Sarah Orzell, Patrick
Rooney, Chase Ross, Michelle Saipe, Julian Sarafin, Leah Soffer, Courtney
Spetko, Benjamin Sprung-Keyser, Mattie Toma, Kate Tomlinson, William
Weber, Katherine Wen, Kayla Wilding, and Andrew Winslow.

356  |  Appendix A

Departures in 2013
In August, David P. Vandivier left his position as Chief of Staff.
The senior economists who resigned in 2013 (with the institutions
to which they returned after leaving the Council in parentheses) were:
Bevin Ashenmiller (Occidental University), Benjamin H. Harris (Brookings
Institution), Susan Helper (Case Western University), Justin Joffrion
(U.S. Air Force Academy), Chinhui Juhn (University of Houston), Paul
Lengermann (Federal Reserve Board), Emily Y. Lin (U.S. Department of
the Treasury), Rodney D. Ludema (Georgetown University), James M.
Williamson (U.S. Department of Agriculture), and Wesley Yin (UCLA).
The economists who departed in 2013 were David Cho (Princeton
University) and Judd N.L. Cramer (Princeton University). David served the
CEA for more than two years and was a recipient of the Robert M. Solow
Award for Distinguished Service.
The staff economists who departed in 2013 were Nicholas Li, Ben
Meiselman, Nicholas Tilipman, Lee Tucker, and Jeffrey Y. Zhang.
The research economists who departed in 2013 were Carys
Golesworthy, Dina Grossman, and Spencer Smith.
Petra S. Starke resigned from her position as General Counsel.
Michael Bourgeois resigned from his position as Special Assistant to the
Chairman. Emily C. Berret resigned from her position as Special Assistant
to the Members. Sarah A. Murray resigned from her position as Economic
Statistician. Thomas F. Hunt resigned from his position as Staff Assistant.

Activities of the Council of Economic Advisers During 2013  |  357

A P P E N D I X

B

STATISTICAL TABLES RELATING
TO INCOME, EMPLOYMENT,
AND PRODUCTION

C O N T E N T S
Page

GDP, INCOME, PRICES, AND SELECTED INDICATORS
B–1. 	 Percent changes in real gross domestic product, 1965–2013��������������������������������  366
B–2.	

Gross domestic product, 1999–2013������������������������������������������������������������������������  368

B–3. 	 Quantity and price indexes for gross domestic product, and percent changes,
1965–2013�������������������������������������������������������������������������������������������������������������������  370
B–4. 	 Growth rates in real gross domestic product by area and country, 1995–2014��  371
B–5. 	 Real exports and imports of goods and services, 1999–2013�������������������������������  372
B–6. 	 Corporate profits by industry, 1965–2013���������������������������������������������������������������  373
B–7. 	 Real farm income, 1950–2014�����������������������������������������������������������������������������������  374
B–8. 	 New private housing units started, authorized, and completed and houses
sold, 1970–2014����������������������������������������������������������������������������������������������������������  375
B–9. 	 Median money income (in 2012 dollars) and poverty status of families and
people, by race, 2003-2012����������������������������������������������������������������������������������������  376
B–10. 	 Changes in consumer price indexes, 1945–2013����������������������������������������������������  377

LABOR MARKET INDICATORS
B–11. 	 Civilian population and labor force, 1929–2014����������������������������������������������������  378
B–12. 	 Civilian unemployment rate, 1970–2014�����������������������������������������������������������������  380
B–13. 	 Unemployment by duration and reason, 1970–2014���������������������������������������������  381
B–14. 	 Employees on nonagricultural payrolls, by major industry, 1970–2014�������������  382
B–15. 	 Hours and earnings in private nonagricultural industries, 1970–2014 ��������������  384
B–16. 	 Productivity and related data, business and nonfarm business sectors,
1965–2013�������������������������������������������������������������������������������������������������������������������  385

INTEREST RATES, MONEY STOCK, AND GOVERNMENT FINANCE
B–17. 	 Bond yields and interest rates, 1942–2014��������������������������������������������������������������  386
B–18.	 Money stock and debt measures, 1974–2014����������������������������������������������������������  388
B–19. 	 Federal receipts, outlays, surplus or deficit, and debt, fiscal years, 1947–2015��  389

Contents  | 361

INTEREST RATES, MONEY STOCK, AND GOVERNMENT FINANCE
—Continued
B–20. 	 Federal receipts, outlays, surplus or deficit, and debt, as percent of gross
domestic product, fiscal years 1942–2015���������������������������������������������������������������  390
B–21. 	 Federal receipts and outlays, by major category, and surplus or deficit, fiscal
years 1947–2015���������������������������������������������������������������������������������������������������������  391
B–22. 	 Federal receipts, outlays, surplus or deficit, and debt, fiscal years 2010–2015���  392
B–23. 	 Federal and State and local government current receipts and expenditures,
national income and product accounts (NIPA), 1965–2013��������������������������������  393
B–24. 	 State and local government revenues and expenditures, selected fiscal years,
1954–2011�������������������������������������������������������������������������������������������������������������������  394
B–25. 	 U.S. Treasury securities outstanding by kind of obligation, 1976–2014��������������  395
B–26. 	 Estimated ownership of U.S. Treasury securities, 2000–2013�������������������������������  396

SOURCES����������������������������������������������������������������������������������������������������������������������������������   397

362  |  Appendix B

General Notes
Detail in these tables may not add to totals due to rounding.
Unless otherwise noted, all dollar figures are in current dollars.
Because of the formula used for calculating real gross domestic product
(GDP), the chained (2009) dollar estimates for the detailed components do
not add to the chained-dollar value of GDP or to any intermediate aggregate.
The Department of Commerce (Bureau of Economic Analysis) no longer
publishes chained-dollar estimates prior to 1999, except for selected series.
Symbols used:
p Preliminary.
	
	
... Not available (also, not applicable).
Data in these tables reflect revisions made by source agencies through
February 28, 2014. In particular, tables containing national income and
product accounts (NIPA) estimates reflect revisions released by the
Department of Commerce in July 2013 and tables containing estimates from
the current employment statistics (CES) survey reflect revisions released by
the Department of Labor in February 2014.
Excel versions of these tables are available at www.gpo.gov/erp.

Notes on Streamlining
With extensive economic data now available online, the 2014 statistical
appendix has been streamlined. Tables that have been retained (some in
modified form) are listed on the following page, matched to their 2013 table
numbers.
Data presented in the past remain available to the public through their
source agencies. For each table in last year’s statistical appendix, the Sources
section starting on page 397 lists source agency (or agencies), website, and
data program for the data featured, along with selected contact information.

General Notes  | 363

2013- to -2014 TABLE NUMBER MATCH
2013	2014	Title
B-1	 B-2	 Gross domestic product
B-2	 B-2	 Real gross domestic product
B-3	 B-3	 Quantity and price indexes for gross domestic product
B-4	 B-1	 Percent changes in real gross domestic product 
B-25	 B-5	 Real exports and imports of goods and services 
B-33	 B-9	 Median money income and poverty status of families and
people, by race
B-35	 B-11	 Civilian population and labor force
B-42	 B-12	 Civilian unemployment rate 
B-44	 B-13	 Unemployment by duration and reason 
B-46	 B-14	 Employees on nonagricultural payrolls by major industry 
B-47	 B-15	 Hours and earnings in private nonagricultural industries 
B-49	 B-16	 Productivity and related data, business and nonfarm
business sectors 
B-56	 B-8	 New private housing units started, authorized, and
completed and houses sold 
B-60	 B-10	 Consumer price indexes for major expenditure classes 
B-63	 B-10	 Changes in special consumer price indexes 
B-69	 B-18	 Money stock and debt measures
B-73	 B-17	 Bond yields and interest rates
B-78	 B-19	 Federal receipts, outlays, surplus or deficit, and debt
B-79	 B-20	 Federal receipts, outlays, surplus or deficit, and debt, as
percent of GDP
B-80	 B-21	 Federal receipts and outlays, by major category, and
surplus or deficit
B-81	 B-22	 Federal receipts, outlays, surplus or deficit, and debt
B-82	 B-23	 Federal and State and local government current receipts
and expenditures
B-86	 B-24	 State and local government revenues and expenditures
B-87	 B-25	 U.S. Treasury securities outstanding by kind of obligation
B-89	 B-26	 Estimated ownership of U.S. Treasury securities
B-91	 B-6	 Corporate profits by industry
B-97	 B-7	 Real farm income
B-112	 B-4	 Growth rates in real gross domestic product, by area and
country

364  |  Appendix B

Historical data may be subject to revision, so source agencies should be
consulted for data no longer shown on these pages. Early data that remain
static and are not available on a source agency website, however, may be
found in previous issues of the Economic Report of the President at www.gpo.
gov/erp and fraser.stlouisfed.org.
Statistical agencies and data aggregators also offer tools allowing users to
download, graph, map, and program data themselves. The Federal Reserve
Bank of St. Louis, a notable aggregator of economic data, features an online
database at its Federal Reserve Economic Data (FRED) site comprising
more than 154,000 economic time series from 59 national, international,
public, and private data sources. In addition to mobile apps and other data
tools, FRED provides application programming interfaces (APIs) to allow
developers to create applications or programs that directly utilize its website
content. For more information, see www.research.stlouisfed.org/fred2.

General Notes  | 365

GDP, Income, Prices, and Selected Indicators
Table B–1. Percent changes in real gross domestic product, 1965–2013
[Percent change from preceding period; quarterly data at seasonally adjusted annual rates]
Personal consumption
expenditures

Year or quarter

1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

Gross
domestic
product

6.5
6.6
2.7
4.9
3.1
.2
3.3
5.2
5.6
–.5
–.2
5.4
4.6
5.6
3.2
–.2
2.6
–1.9
4.6
7.3
4.2
3.5
3.5
4.2
3.7
1.9
–.1
3.6
2.7
4.0
2.7
3.8
4.5
4.4
4.8
4.1
1.0
1.8
2.8
3.8
3.4
2.7
1.8
–.3
–2.8
2.5
1.8
2.8
1.9
1.6
3.9
2.8
2.8
–1.3
3.2
1.4
4.9
3.7
1.2
2.8
.1
1.1
2.5
4.1
2.4

Fixed investment
Nonresidential
Total

6.3
5.7
3.0
5.7
3.7
2.4
3.8
6.1
5.0
–.8
2.3
5.6
4.2
4.4
2.4
–.3
1.5
1.4
5.7
5.3
5.3
4.2
3.4
4.2
2.9
2.1
.2
3.7
3.5
3.9
3.0
3.5
3.8
5.3
5.5
5.1
2.5
2.5
3.1
3.8
3.5
3.0
2.2
–.4
–1.6
2.0
2.5
2.2
2.0
2.1
3.3
2.8
4.3
2.1
1.5
2.1
2.4
2.9
1.9
1.7
1.7
2.3
1.8
2.0
2.6

See next page for continuation of table.

366  |  Appendix B

Gross private domestic investment

Goods

7.1
6.3
2.0
6.2
3.1
.8
4.2
6.5
5.2
–3.6
.7
7.0
4.3
4.1
1.6
–2.5
1.2
.7
6.4
7.2
5.3
5.6
1.8
3.7
2.5
.6
–2.0
3.2
4.2
5.3
3.0
4.5
4.8
6.7
7.9
5.2
3.0
3.9
4.8
5.1
4.1
3.6
2.7
–2.5
–3.0
3.4
3.4
3.3
3.6
4.0
5.2
3.8
7.6
2.7
.2
1.2
5.0
4.6
2.2
3.7
3.7
3.7
3.1
4.5
3.2

Services

5.5
4.9
4.1
5.3
4.4
3.9
3.5
5.7
4.7
1.9
3.8
4.3
4.1
4.6
3.1
1.6
1.7
2.0
5.2
3.9
5.3
3.2
4.5
4.5
3.2
3.0
1.6
4.0
3.1
3.1
3.0
2.9
3.2
4.6
4.1
5.0
2.2
1.8
2.2
3.2
3.2
2.7
2.0
.8
–.8
1.2
2.1
1.6
1.1
1.2
2.4
2.2
2.6
1.8
2.1
2.5
1.1
2.1
1.7
.7
.6
1.5
1.2
.7
2.2

Total

13.8
9.0
–3.5
6.0
5.6
–6.1
10.3
11.3
10.9
–6.6
–16.2
19.1
14.3
11.6
3.5
–10.1
8.8
–13.0
9.3
27.3
–.1
.2
2.8
2.5
4.0
–2.6
–6.6
7.3
8.0
11.9
3.2
8.8
11.4
9.5
8.4
6.5
–6.1
–.6
4.1
8.8
6.4
2.1
–3.1
–9.4
–21.6
12.9
4.9
9.5
5.5
13.6
22.3
13.7
–3.5
–7.5
14.2
2.5
31.9
10.5
–1.6
6.5
–2.4
4.7
9.2
17.2
4.5

Total

10.4
6.2
–.9
7.0
5.9
–2.1
6.9
11.4
8.6
–5.6
–9.8
9.8
13.6
11.6
5.8
–5.9
2.7
–6.7
7.5
16.2
5.5
1.8
.6
3.3
3.2
–1.4
–5.1
5.5
7.7
8.2
6.1
8.9
8.6
10.2
8.8
6.9
–1.6
–3.5
4.0
6.7
6.8
2.0
–2.0
–6.8
–16.7
1.5
6.2
8.3
4.5
.8
13.6
–.4
8.5
–.5
8.6
14.8
10.0
8.6
4.7
2.7
11.6
–1.5
6.5
5.9
3.8

Total
16.7
12.3
–.3
4.8
7.0
–.9
.0
8.7
13.2
.8
–9.0
5.7
10.8
13.8
10.0
.0
6.1
–3.6
–.4
16.7
6.6
–1.7
.1
5.0
5.7
1.1
–3.9
2.9
7.5
7.9
9.7
9.1
10.8
10.8
9.7
9.1
–2.4
–6.9
1.9
5.2
7.0
7.1
5.9
–.7
–15.6
2.5
7.6
7.3
2.8
4.2
11.4
8.3
8.6
–.9
9.9
16.7
9.5
5.8
4.5
.3
9.8
–4.6
4.7
4.8
7.3

Structures
15.9
6.8
–2.5
1.4
5.4
.3
–1.6
3.1
8.2
–2.2
–10.5
2.4
4.1
14.4
12.7
5.9
8.0
–1.6
–10.8
13.9
7.1
–11.0
–2.9
.7
2.0
1.5
–11.1
–6.0
–.3
1.8
6.4
5.7
7.3
5.1
.1
7.8
–1.5
–17.7
–3.9
–.4
1.7
7.2
12.7
6.1
–18.9
–16.4
2.1
12.7
1.4
–25.0
11.8
–5.8
7.7
–29.8
33.7
28.4
14.4
7.0
6.9
5.9
17.6
–25.7
17.6
13.4
.2

Equipment
18.2
15.5
–1.0
6.1
8.3
–1.8
.8
12.7
18.5
2.1
–10.5
6.1
15.5
15.1
8.2
–4.4
3.7
–7.6
4.6
19.4
5.5
1.1
.4
6.6
5.3
–2.1
–4.6
5.9
12.7
12.3
12.1
9.5
11.1
13.1
12.5
9.7
–4.3
–5.4
3.2
7.7
9.6
8.6
3.2
–6.9
–22.9
15.9
12.7
7.6
3.1
31.2
23.3
18.0
11.8
12.0
4.3
20.3
10.2
8.3
5.3
–3.9
8.9
1.6
3.3
.2
10.6

Intellectual
Property
Products
12.7
13.2
7.8
7.5
5.4
–.1
.4
7.0
5.0
2.9
.9
10.9
6.6
7.1
11.7
5.0
10.9
6.2
7.9
13.7
9.0
7.0
3.9
7.1
11.7
8.4
6.4
6.0
4.2
4.0
7.3
11.3
13.0
10.8
12.4
8.9
.5
–.5
3.8
5.1
6.5
4.5
4.8
3.0
–1.4
1.9
4.4
3.4
3.4
–1.6
–2.0
6.1
5.0
3.7
4.9
5.3
5.5
1.3
1.8
2.8
5.7
3.7
–1.5
5.8
8.0

Residential

–2.6
–8.4
–2.6
13.5
3.1
–5.2
26.6
17.4
–.6
–19.6
–12.1
22.1
20.5
6.7
–3.7
–20.9
–8.2
–18.1
42.0
14.8
2.3
12.4
2.0
–.9
–3.2
–8.5
–8.9
13.8
8.2
9.0
–3.4
8.2
2.4
8.6
6.3
.7
.9
6.1
9.1
10.0
6.6
–7.6
–18.8
–24.0
–21.2
–2.5
.5
12.9
12.1
–12.2
23.2
–30.7
7.9
1.7
2.7
6.1
12.2
23.0
5.7
14.1
19.8
12.5
14.2
10.3
–8.7

Change
in
private
inventories
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������

Table B–1. Percent changes in real gross domestic product, 1965–2013—Continued
[Percent change from preceding period; quarterly data at seasonally adjusted annual rates]
Net exports of
goods and services
Year or quarter

1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

Net
exports
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������

Government consumption expenditures
and gross investment
Federal

Exports
2.8
6.9
2.3
7.8
4.8
10.8
1.7
7.5
18.9
7.9
–.6
4.4
2.4
10.5
9.9
10.8
1.2
–7.6
–2.6
8.2
3.3
7.7
10.9
16.2
11.6
8.8
6.6
6.9
3.3
8.8
10.3
8.2
11.9
2.3
4.6
8.4
–5.7
–1.9
1.6
9.4
6.0
8.9
8.9
5.7
–9.1
11.5
7.1
3.5
2.7
6.4
9.5
10.9
12.4
3.8
4.9
7.0
2.7
4.2
3.8
.4
1.1
–1.3
8.0
3.9
9.4

Imports
10.6
14.9
7.3
14.9
5.7
4.3
5.3
11.2
4.6
–2.3
–11.1
19.6
10.9
8.7
1.7
–6.6
2.6
–1.3
12.6
24.3
6.5
8.5
5.9
3.9
4.4
3.6
–.1
7.0
8.6
11.9
8.0
8.7
13.5
11.7
11.4
12.8
–2.9
3.4
4.3
11.0
6.1
6.1
2.3
–2.6
–13.7
12.8
4.9
2.2
1.4
11.9
20.2
14.5
.9
2.8
.7
4.9
5.9
.7
2.5
.5
–3.1
.6
6.9
2.4
1.5

Total
3.2
8.7
7.9
3.4
.2
–2.0
–1.8
–.5
–.3
2.3
2.2
.5
1.2
2.9
1.9
1.9
1.0
1.8
3.8
3.6
6.8
5.4
3.0
1.3
2.9
3.2
1.2
.5
–.8
.1
.5
1.0
1.9
2.1
3.4
1.9
3.8
4.4
2.2
1.6
.6
1.5
1.6
2.8
3.2
.1
–3.2
–1.0
–2.3
–2.9
2.9
–.3
–4.1
–7.5
–1.3
–2.5
–1.5
–1.4
.3
3.5
–6.5
–4.2
–.4
.4
–5.6

Total
0.8
10.7
10.1
1.5
–2.4
–6.1
–6.4
–3.1
–3.6
.7
.5
.2
2.2
2.5
2.3
4.4
4.5
3.7
6.5
3.3
7.9
5.9
3.8
–1.3
1.7
2.1
.0
–1.5
–3.5
–3.5
–2.6
–1.2
–.8
–.9
2.0
.3
3.9
7.2
6.8
4.5
1.7
2.5
1.7
6.8
5.7
4.4
–2.6
–1.4
–5.2
3.8
8.5
3.7
–2.7
–10.5
1.8
–3.4
–3.1
–2.5
–.2
8.9
–13.9
–8.4
–1.6
–1.5
–12.8

National
defense

Nondefense

–1.3
12.9
12.5
1.6
–4.1
–8.2
–10.2
–6.9
–5.1
–1.0
–1.0
–.5
1.0
.8
2.7
3.9
6.2
7.2
7.3
5.2
8.8
6.9
5.1
–.2
–.2
.3
–1.0
–4.5
–5.1
–4.9
–4.0
–1.6
–2.7
–2.1
1.5
–.9
3.5
7.0
8.5
6.0
2.0
2.0
2.5
7.5
5.4
3.2
–2.3
–3.2
–7.0
–1.8
6.4
7.6
–3.5
–14.2
6.8
2.4
–10.2
–6.7
–1.0
12.5
–21.6
–11.2
–.6
–.5
–14.4

State
and
local

7.9
3.6
1.9
1.3
3.9
1.0
5.6
7.2
.2
4.6
3.9
1.6
4.7
6.0
1.7
5.4
1.0
–3.6
4.7
–1.4
5.7
3.1
.2
–4.3
7.2
7.3
2.4
5.9
.0
–.8
.0
–.5
2.8
1.3
2.7
2.3
4.7
7.4
4.1
2.0
1.3
3.5
.3
5.5
6.2
6.4
–3.0
1.8
–1.9
14.8
12.3
–2.8
–1.2
–3.5
–6.5
–13.1
11.3
5.4
1.2
2.8
1.0
–3.6
–3.1
–3.1
–10.1

6.6
6.2
5.0
6.0
3.5
2.9
3.1
2.2
2.8
3.7
3.6
.8
.4
3.3
1.5
–.2
–2.0
.1
1.3
3.8
5.7
5.0
2.2
3.9
4.0
4.1
2.2
2.1
1.2
2.8
2.7
2.4
3.6
3.8
4.2
2.8
3.7
2.9
–.4
–.1
.0
.9
1.5
.3
1.6
–2.7
–3.6
–.7
–.2
–7.1
–.8
–3.1
–5.0
–5.4
–3.4
–1.9
–.4
–.6
.6
–.2
–1.0
–1.3
.4
1.7
–.5

Final
Gross
Gross
Gross
sales of domestic domestic national
domestic
pur2 product 3
1 income
product chases
5.9
6.1
3.3
5.1
3.2
.9
2.7
5.2
5.2
–.3
1.0
4.0
4.4
5.5
3.6
.6
1.5
–.6
4.3
5.4
5.4
3.8
3.1
4.4
3.5
2.1
.2
3.3
2.7
3.4
3.2
3.8
4.0
4.5
4.9
4.2
1.9
1.2
2.8
3.4
3.4
2.6
2.0
.2
–2.0
1.0
2.0
2.6
1.7
.0
2.8
.9
4.5
–.3
2.4
3.0
2.1
3.4
2.2
2.2
2.2
.2
2.1
2.5
2.3

6.9
6.4
6.5
6.9
6.0
6.5
3.0
3.0
2.7
5.2
5.0
4.9
3.2
3.3
3.1
–.1
–.1
.2
3.5
3.0
3.3
5.5
5.5
5.3
4.8
5.8
5.9
–1.2
–.6
–.4
–1.1
–.5
–.4
6.5
5.1
5.5
5.3
4.8
4.7
5.5
5.5
5.5
2.5
2.4
3.5
–1.9
–.1
–.3
2.7
3.0
2.4
–1.3
–1.0
–1.8
5.9
3.3
4.6
8.7
7.8
7.1
4.5
4.0
3.9
3.7
3.0
3.3
3.2
4.3
3.4
3.3
5.1
4.3
3.1
2.5
3.7
1.5
1.5
2.0
–.7
.0
–.2
3.6
3.3
3.5
3.3
2.2
2.7
4.4
4.4
3.9
2.6
3.4
2.8
3.9
4.3
3.8
4.7
5.1
4.4
5.5
5.3
4.4
5.7
4.5
4.9
4.8
4.7
4.2
1.1
1.1
1.1
2.3
1.4
1.7
3.1
2.2
2.9
4.2
3.7
3.9
3.5
3.6
3.3
2.6
4.0
2.4
1.1
.1
2.2
–1.3
–.8
.0
–3.8
–2.6
–3.0
2.9
2.7
2.8
1.7
2.5
2.1
2.6
2.5
2.7
1.7 ��������������� ����������������
2.5
.5
1.7
5.5
2.8
3.9
3.5
5.2
2.6
1.4
1.6
3.2
–1.3
2.0
–.5
2.6
2.3
3.1
1.2
2.2
1.9
5.3
2.6
4.8
3.1
5.4
3.0
1.1
–.6
1.4
2.7
.9
2.4
–.5
4.9
.3
1.4
2.4
.6
2.5
3.2
2.7
3.9
1.8
4.4
1.4 ��������������� ����������������

1 Gross domestic product (GDP) less exports of goods and services plus imports of goods and services.
2 Gross domestic income is deflated by the implicit price deflator for GDP.
3 GDP plus net income receipts from rest of the world.

Note: Percent changes based on unrounded GDP quantity indexes.
Source: Department of Commerce (Bureau of Economic Analysis).

GDP, Income, Prices, and Selected Indicators  | 367

Table B–2. Gross domestic product, 1999–2013
[Quarterly data at seasonally adjusted annual rates]
Personal consumption
expenditures

Year or quarter

Gross
domestic
product

Gross private domestic investment
Fixed investment
Nonresidential

Total

Goods

Services

Total

Total

Total

Structures

Equipment

Intellectual
Property
Products

Residential

Change
in
private
inventories

Billions of dollars
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

9,665.7
10,289.7
10,625.3
10,980.2
11,512.2
12,277.0
13,095.4
13,857.9
14,480.3
14,720.3
14,417.9
14,958.3
15,533.8
16,244.6
16,797.5
14,672.5
14,879.2
15,049.8
15,231.7
15,242.9
15,461.9
15,611.8
15,818.7
16,041.6
16,160.4
16,356.0
16,420.3
16,535.3
16,661.0
16,912.9
17,080.7

6,316.9
6,801.6
7,106.9
7,385.3
7,764.4
8,257.8
8,790.3
9,297.5
9,744.4
10,005.5
9,842.9
10,201.9
10,711.8
11,149.6
11,496.2
10,042.3
10,134.7
10,234.3
10,396.3
10,527.1
10,662.6
10,778.6
10,878.9
11,019.1
11,100.2
11,193.6
11,285.5
11,379.2
11,427.1
11,537.7
11,640.7

2,286.8
2,452.9
2,525.2
2,598.6
2,721.6
2,900.3
3,080.3
3,235.8
3,361.6
3,375.7
3,198.4
3,362.8
3,602.7
3,769.7
3,886.6
3,304.9
3,325.6
3,362.4
3,458.4
3,532.2
3,588.2
3,622.3
3,668.2
3,729.3
3,738.4
3,784.9
3,826.1
3,851.8
3,848.5
3,912.8
3,933.2

4,030.1
4,348.8
4,581.6
4,786.7
5,042.8
5,357.5
5,710.1
6,061.7
6,382.9
6,629.8
6,644.5
6,839.1
7,109.1
7,379.9
7,609.6
6,737.4
6,809.1
6,871.9
6,937.9
6,995.0
7,074.4
7,156.3
7,210.7
7,289.7
7,361.8
7,408.7
7,459.4
7,527.4
7,578.6
7,624.8
7,707.6

1,884.2
2,033.8
1,928.6
1,925.0
2,027.9
2,276.7
2,527.1
2,680.6
2,643.7
2,424.8
1,878.1
2,100.8
2,232.1
2,475.2
2,673.7
1,989.5
2,092.7
2,164.6
2,156.5
2,120.4
2,199.9
2,222.2
2,385.7
2,453.6
2,454.0
2,493.3
2,499.9
2,555.1
2,621.0
2,738.0
2,780.5

1,823.4
1,979.2
1,966.9
1,906.5
2,008.7
2,212.8
2,467.5
2,613.7
2,609.3
2,456.8
2,025.7
2,039.3
2,195.6
2,409.1
2,565.7
1,977.5
2,042.6
2,043.0
2,094.1
2,098.9
2,154.1
2,235.7
2,293.8
2,350.7
2,387.1
2,411.7
2,486.9
2,491.7
2,543.8
2,593.2
2,634.2

1,361.6
1,493.8
1,453.9
1,348.9
1,371.7
1,463.1
1,611.5
1,776.3
1,920.6
1,941.0
1,633.4
1,658.2
1,809.9
1,970.0
2,049.0
1,594.4
1,641.8
1,677.4
1,719.3
1,721.8
1,773.1
1,848.9
1,895.7
1,932.3
1,961.4
1,968.0
2,018.2
2,001.4
2,030.6
2,060.5
2,103.3

283.9
318.1
329.7
282.9
281.8
301.8
345.6
415.6
496.9
552.4
438.2
362.0
380.6
437.3
457.1
352.4
364.5
361.1
370.1
340.8
370.1
397.5
413.9
422.0
431.3
438.3
457.8
429.1
452.6
470.7
475.9

713.6
766.1
711.5
659.6
669.0
719.2
790.7
856.1
885.8
825.1
644.3
731.8
832.7
907.6
939.4
682.7
719.0
751.2
774.4
798.0
809.9
849.8
873.0
895.4
907.9
902.2
925.0
928.0
934.6
935.8
959.1

364.0
409.5
412.6
406.4
420.9
442.1
475.1
504.6
537.9
563.4
550.9
564.3
596.6
625.0
652.5
559.2
558.3
565.1
574.8
582.9
593.1
601.6
608.8
614.9
622.2
627.5
635.4
644.3
643.5
654.1
668.2

461.8
485.4
513.0
557.6
636.9
749.7
856.1
837.4
688.7
515.9
392.2
381.1
385.8
439.2
516.8
383.1
400.8
365.6
374.7
377.1
381.0
386.8
398.1
418.4
425.7
443.7
468.8
490.3
513.2
532.6
531.0

60.8
54.5
–38.3
18.5
19.3
63.9
59.6
67.0
34.5
–32.0
–147.6
61.5
36.4
66.1
107.9
12.1
50.1
121.5
62.4
21.5
45.8
–13.5
91.9
102.9
66.8
81.6
13.0
63.4
77.2
144.8
146.3

494.9
533.5
525.4
432.5
415.8
414.1
421.2
451.5
509.0
540.2
438.2
366.3
374.1
421.6
427.4
359.7
369.8
364.4
371.2
339.8
365.3
388.9
402.2
409.0
416.0
422.0
439.4
407.9
424.8
438.4
438.6

662.4
726.9
695.7
658.0
679.0
731.2
801.6
870.8
898.3
836.1
644.3
746.7
841.7
905.9
934.2
697.7
735.2
766.2
787.8
810.6
819.2
858.0
879.1
896.9
908.5
899.5
918.8
922.5
929.9
930.4
954.0

391.1
426.1
428.0
425.9
442.2
464.9
495.0
517.5
542.4
558.8
550.9
561.3
586.1
605.8
626.3
557.6
554.7
563.0
570.0
575.2
582.0
589.6
597.6
599.6
602.3
606.4
614.9
620.6
618.3
627.0
639.2

633.8
637.9
643.7
682.7
744.5
818.9
872.6
806.6
654.8
497.7
392.2
382.4
384.3
433.7
486.4
383.0
403.5
368.1
375.1
376.7
379.2
384.9
396.2
417.2
423.0
437.3
457.5
471.2
487.1
499.2
487.9

75.5
66.2
–46.2
22.5
22.6
71.4
64.3
71.6
35.5
–33.7
–147.6
58.2
33.6
57.6
83.0
9.8
48.8
116.2
58.1
22.0
42.9
–11.0
80.6
89.2
56.8
77.2
7.3
42.2
56.6
115.7
117.4

Billions of chained (2009) dollars
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

12,071.4
12,565.2
12,684.4
12,909.7
13,270.0
13,774.0
14,235.6
14,615.2
14,876.8
14,833.6
14,417.9
14,779.4
15,052.4
15,470.7
15,759.0
14,597.7
14,738.0
14,839.3
14,942.4
14,894.0
15,011.3
15,062.1
15,242.1
15,381.6
15,427.7
15,534.0
15,539.6
15,583.9
15,679.7
15,839.3
15,932.9

7,788.1
8,182.1
8,387.5
8,600.4
8,866.2
9,205.6
9,527.8
9,814.9
10,035.5
9,999.2
9,842.9
10,035.9
10,291.3
10,517.6
10,723.0
9,915.4
9,995.3
10,063.7
10,169.0
10,221.3
10,258.9
10,311.9
10,373.1
10,447.8
10,496.8
10,541.0
10,584.8
10,644.0
10,691.9
10,744.2
10,812.1

See next page for continuation of table.

368  |  Appendix B

2,460.9
2,588.3
2,666.6
2,770.2
2,904.5
3,051.9
3,177.2
3,292.5
3,381.8
3,297.8
3,198.4
3,308.7
3,419.9
3,534.1
3,660.1
3,247.0
3,288.0
3,319.1
3,380.5
3,402.8
3,404.6
3,415.2
3,457.0
3,495.8
3,514.7
3,546.7
3,579.2
3,611.9
3,639.6
3,680.0
3,708.8

5,344.8
5,611.6
5,736.3
5,840.0
5,965.6
6,154.1
6,349.4
6,519.8
6,650.4
6,700.6
6,644.5
6,727.2
6,871.1
6,982.7
7,062.3
6,668.3
6,707.2
6,744.6
6,788.5
6,818.2
6,854.1
6,896.6
6,915.5
6,951.2
6,981.4
6,993.4
7,004.7
7,031.1
7,051.5
7,063.6
7,102.8

2,231.4
2,375.5
2,231.4
2,218.2
2,308.7
2,511.3
2,672.6
2,730.0
2,644.1
2,396.0
1,878.1
2,120.4
2,224.6
2,436.0
2,569.6
2,012.9
2,116.9
2,185.7
2,166.1
2,124.3
2,196.1
2,209.9
2,368.2
2,427.8
2,418.0
2,456.5
2,441.8
2,470.1
2,524.9
2,627.2
2,656.2

2,165.9
2,316.2
2,280.0
2,201.1
2,289.5
2,443.9
2,611.0
2,662.5
2,609.6
2,432.6
2,025.7
2,056.2
2,184.6
2,365.3
2,472.5
1,997.9
2,062.8
2,060.8
2,103.1
2,100.7
2,144.4
2,219.8
2,273.4
2,320.8
2,347.9
2,363.5
2,429.1
2,420.0
2,458.4
2,494.0
2,517.5

1,510.1
1,647.7
1,608.4
1,498.0
1,526.1
1,605.4
1,717.4
1,839.6
1,948.4
1,934.4
1,633.4
1,673.8
1,800.5
1,931.8
1,986.3
1,615.0
1,659.3
1,692.8
1,728.1
1,724.1
1,765.3
1,835.0
1,877.3
1,903.8
1,925.0
1,926.4
1,971.9
1,949.0
1,971.3
1,994.7
2,030.1

Table B–2. Gross domestic product, 1999–2013—Continued
[Quarterly data at seasonally adjusted annual rates]
Net exports of
goods and services
Year or quarter

Government consumption expenditures
and gross investment
Federal

Net
exports

Exports

Imports

Total

Total

National
defense

Nondefense

State
and
local

Final
Gross
Gross
Gross
sales of domestic domestic national
domestic
pur2 product 3
1 income
product chases

Billions of dollars
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

–261.4
–380.1
–369.0
–425.0
–500.9
–614.8
–715.7
–762.4
–709.8
–713.2
–392.2
–518.5
–568.7
–547.2
–497.3
–495.1
–529.7
–543.8
–505.3
–554.7
–572.2
–553.7
–594.4
–590.8
–557.9
–524.4
–515.8
–523.1
–509.0
–500.2
–456.8

989.2
1,094.3
1,028.8
1,004.7
1,043.4
1,183.1
1,310.4
1,478.5
1,665.7
1,843.1
1,583.8
1,843.5
2,101.2
2,195.9
2,259.8
1,746.4
1,807.0
1,860.3
1,960.4
2,029.5
2,095.5
2,143.4
2,136.2
2,173.4
2,197.4
2,199.2
2,213.7
2,214.2
2,238.9
2,265.8
2,320.1

1,250.6
1,474.4
1,397.8
1,429.7
1,544.3
1,797.9
2,026.1
2,240.9
2,375.5
2,556.4
1,976.0
2,362.0
2,669.9
2,743.1
2,757.0
2,241.4
2,336.7
2,404.0
2,465.7
2,584.1
2,667.7
2,697.1
2,730.7
2,764.2
2,755.3
2,723.5
2,729.5
2,737.3
2,747.9
2,766.0
2,776.9

1,726.0
1,834.4
1,958.8
2,094.9
2,220.8
2,357.4
2,493.7
2,642.2
2,801.9
3,003.2
3,089.1
3,174.0
3,158.7
3,167.0
3,124.9
3,135.7
3,181.5
3,194.7
3,184.2
3,150.0
3,171.7
3,164.6
3,148.5
3,159.7
3,164.1
3,193.5
3,150.7
3,124.1
3,121.9
3,137.5
3,116.2

610.4
632.4
669.2
740.6
824.8
892.4
946.3
1,002.0
1,049.8
1,155.6
1,217.7
1,303.9
1,304.1
1,295.7
1,245.9
1,269.2
1,304.6
1,321.6
1,320.1
1,297.4
1,315.4
1,308.5
1,294.9
1,291.8
1,293.8
1,322.1
1,275.2
1,255.0
1,252.6
1,251.2
1,224.8

1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

–382.3
–482.7
–504.2
–584.9
–641.6
–731.9
–777.1
–786.2
–703.6
–546.9
–392.2
–462.6
–445.9
–430.8
–412.3
–413.6
–474.3
–504.9
–457.5
–456.5
–438.3
–433.9
–454.7
–439.2
–435.3
–436.5
–412.1
–422.3
–424.4
–419.8
–382.8

1,174.1
1,272.4
1,200.5
1,178.1
1,197.2
1,309.3
1,388.4
1,512.4
1,647.3
1,741.8
1,583.8
1,765.6
1,890.5
1,957.4
2,010.0
1,700.4
1,739.3
1,784.9
1,837.7
1,854.7
1,876.9
1,908.9
1,921.7
1,941.4
1,959.8
1,961.6
1,967.0
1,960.5
1,998.4
2,017.6
2,063.5

1,556.4
1,755.1
1,704.7
1,763.0
1,838.8
2,041.2
2,165.5
2,298.6
2,350.9
2,288.7
1,976.0
2,228.1
2,336.4
2,388.2
2,422.3
2,113.9
2,213.6
2,289.8
2,295.2
2,311.3
2,315.2
2,342.8
2,376.4
2,380.6
2,395.1
2,398.0
2,379.1
2,382.7
2,422.9
2,437.3
2,446.2

2,451.7
2,498.2
2,592.4
2,705.8
2,764.3
2,808.2
2,826.2
2,869.3
2,914.4
2,994.8
3,089.1
3,091.4
2,992.3
2,963.1
2,896.3
3,084.3
3,106.2
3,103.5
3,071.5
3,012.0
3,002.4
2,983.2
2,971.7
2,961.3
2,963.5
2,988.8
2,938.8
2,907.4
2,904.5
2,907.4
2,866.2

815.3
817.7
849.8
910.8
973.0
1,017.1
1,034.8
1,060.9
1,078.7
1,152.3
1,217.7
1,270.7
1,237.9
1,220.3
1,157.4
1,247.8
1,273.4
1,285.0
1,276.4
1,241.6
1,247.0
1,236.4
1,226.7
1,219.1
1,218.5
1,244.6
1,198.9
1,172.8
1,168.2
1,163.9
1,124.7

382.7
391.7
412.7
456.8
519.9
570.2
608.3
642.4
678.7
754.1
788.3
832.8
835.8
817.1
770.8
811.9
829.3
846.3
843.5
822.0
844.2
851.6
825.6
816.3
816.7
841.9
793.7
775.8
776.3
777.3
753.7

227.7
240.7
256.5
283.8
304.9
322.1
338.1
359.6
371.0
401.5
429.4
471.1
468.2
478.6
475.1
457.3
475.2
475.3
476.6
475.4
471.2
456.9
469.3
475.5
477.1
480.2
481.5
479.2
476.3
473.9
471.1

1,115.6
1,202.0
1,289.5
1,354.3
1,396.0
1,465.0
1,547.4
1,640.2
1,752.2
1,847.6
1,871.4
1,870.2
1,854.7
1,871.3
1,879.0
1,866.5
1,876.9
1,873.1
1,864.2
1,852.6
1,856.3
1,856.1
1,853.6
1,867.9
1,870.3
1,871.4
1,875.4
1,869.1
1,869.3
1,886.3
1,891.4

9,604.9
10,235.2
10,663.5
10,961.7
11,493.0
12,213.2
13,035.8
13,790.9
14,445.9
14,752.3
14,565.5
14,896.7
15,497.4
16,178.5
16,689.6
14,660.4
14,829.0
14,928.2
15,169.3
15,221.4
15,416.2
15,625.3
15,726.8
15,938.7
16,093.6
16,274.4
16,407.3
16,471.9
16,583.8
16,768.1
16,934.4

9,927.1
10,669.8
10,994.3
11,405.2
12,013.2
12,891.8
13,811.1
14,620.3
15,190.1
15,433.5
14,810.1
15,476.7
16,102.6
16,791.8
17,294.8
15,167.5
15,408.9
15,593.5
15,737.0
15,797.6
16,034.1
16,165.5
16,413.1
16,632.4
16,718.3
16,880.4
16,936.1
17,058.4
17,170.0
17,413.2
17,537.5

9,698.1
9,692.8
10,384.3 10,326.8
10,736.8 10,677.1
11,050.3 11,028.8
11,524.3 11,580.3
12,283.5 12,367.1
13,129.2 13,189.0
14,073.2 13,926.3
14,460.1 14,606.8
14,621.2 14,893.2
14,345.7 14,565.1
14,915.2 15,164.2
15,587.5 15,794.6
16,261.6 16,497.4
�������������� ����������������
14,627.4 14,875.9
14,793.7 15,084.3
15,050.5 15,249.5
15,189.0 15,447.2
15,326.2 15,491.2
15,513.6 15,712.1
15,694.9 15,884.0
15,815.3 16,091.0
16,104.6 16,289.6
16,150.3 16,419.2
16,269.6 16,603.7
16,522.0 16,677.3
16,690.9 16,772.7
16,847.8 16,907.9
17,004.6 17,175.9
�������������� ����������������

12,000.3
12,500.4
12,731.7
12,889.9
13,247.9
13,702.7
14,170.1
14,543.6
14,839.2
14,868.9
14,565.5
14,717.7
15,014.4
15,403.2
15,665.8
14,584.3
14,686.3
14,718.3
14,881.8
14,871.9
14,961.8
15,072.7
15,151.3
15,278.9
15,360.8
15,444.9
15,528.3
15,536.4
15,616.2
15,711.1
15,799.4

12,474.6
13,069.5
13,213.5
13,520.1
13,937.1
14,529.1
15,036.2
15,424.8
15,600.8
15,392.0
14,810.1
15,244.5
15,501.1
15,902.3
16,170.4
15,011.5
15,215.4
15,348.5
15,402.5
15,354.0
15,451.6
15,498.4
15,700.5
15,822.4
15,864.4
15,971.4
15,950.8
16,005.8
16,104.1
16,258.5
16,313.1

12,111.9 12,108.9
12,680.6 12,614.3
12,817.6 12,750.2
12,992.1 12,970.8
13,283.9 13,352.2
13,781.3 13,879.0
14,272.3 14,340.8
14,842.3 14,690.9
14,856.1 15,009.7
14,733.8 15,009.0
14,345.7 14,565.1
14,736.7 14,966.5
15,104.3 15,286.7
15,487.0 15,693.1
�������������� ����������������
14,552.8 14,782.7
14,653.4 14,925.1
14,840.1 15,020.5
14,900.5 15,137.8
14,975.4 15,119.2
15,061.5 15,235.6
15,142.2 15,306.4
15,238.8 15,485.7
15,441.9 15,600.2
15,418.0 15,656.2
15,451.9 15,751.1
15,636.0 15,764.8
15,730.6 15,789.7
15,855.4 15,893.9
15,925.2 16,067.4
�������������� ����������������

Billions of chained (2009) dollars
516.9
512.3
530.0
567.3
615.4
652.7
665.5
678.8
695.6
748.1
788.3
813.5
794.6
769.1
715.0
798.6
811.0
825.9
818.6
787.8
800.8
805.6
784.2
770.7
768.8
791.8
745.0
723.1
722.0
721.2
693.6

298.5
305.4
319.7
343.3
357.5
364.5
369.4
382.1
383.1
404.2
429.4
457.1
443.3
451.2
442.4
449.2
462.4
459.1
457.7
453.7
446.2
430.8
442.5
448.3
449.7
452.8
453.9
449.8
446.2
442.7
431.1

1,643.6
1,689.1
1,751.5
1,802.4
1,795.3
1,792.8
1,792.3
1,808.8
1,836.1
1,842.4
1,871.4
1,820.8
1,754.5
1,742.8
1,738.6
1,836.5
1,832.8
1,818.5
1,795.2
1,770.5
1,755.5
1,746.9
1,745.0
1,742.2
1,745.0
1,744.3
1,739.8
1,734.3
1,736.0
1,743.2
1,741.1

1 Gross domestic product (GDP) less exports of goods and services plus imports of goods and services.
2 For chained dollar measures, gross domestic income is deflated by the implicit price deflator for GDP.
3 GDP plus net income receipts from rest of the world.

Source: Department of Commerce (Bureau of Economic Analysis).

GDP, Income, Prices, and Selected Indicators  | 369

Table B–3. Quantity and price indexes for gross domestic product, and percent changes,
1965–2013
[Quarterly data are seasonally adjusted]

Percent change from preceding period 1

Index numbers, 2009=100
Gross domestic product (GDP)
Year or quarter

1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

Personal consumption expenditures
(PCE)

Real GDP
(chaintype
quantity
index)

GDP
chaintype
price
index

GDP
implicit
price
deflator

PCE
chaintype
price
index

27.555
29.373
30.179
31.660
32.653
32.721
33.798
35.572
37.580
37.385
37.311
39.321
41.133
43.421
44.800
44.690
45.850
44.974
47.057
50.473
52.613
54.460
56.346
58.715
60.875
62.044
61.998
64.202
65.965
68.628
70.493
73.169
76.453
79.855
83.725
87.149
87.977
89.539
92.038
95.534
98.735
101.368
103.182
102.883
100.000
102.507
104.400
107.302
109.301
101.247
102.220
102.923
103.638
103.302
104.115
104.468
105.716
106.683
107.003
107.741
107.780
108.087
108.751
109.859
110.508

18.744
19.270
19.830
20.673
21.692
22.835
23.996
25.038
26.399
28.763
31.435
33.161
35.213
37.685
40.795
44.485
48.663
51.630
53.664
55.570
57.347
58.510
59.941
62.042
64.455
66.848
69.063
70.639
72.322
73.859
75.402
76.776
78.097
78.944
80.071
81.894
83.767
85.055
86.754
89.130
91.989
94.816
97.338
99.208
100.000
101.215
103.203
105.008
106.487
100.509
100.972
101.432
101.948
102.354
103.024
103.651
103.782
104.296
104.751
105.345
105.640
105.994
106.165
106.685
107.103

18.720
19.246
19.805
20.647
21.663
22.805
23.964
25.005
26.366
28.734
31.395
33.119
35.173
37.643
40.750
44.425
48.572
51.586
53.623
55.525
57.302
58.458
59.949
62.048
64.460
66.845
69.069
70.644
72.325
73.865
75.406
76.783
78.096
78.944
80.071
81.891
83.766
85.054
86.754
89.132
91.991
94.818
97.335
99.236
100.000
101.211
103.199
105.002
106.590
100.513
100.958
101.418
101.936
102.343
103.002
103.650
103.783
104.291
104.750
105.292
105.667
106.105
106.259
106.778
107.204

18.680
19.155
19.637
20.402
21.326
22.325
23.274
24.070
25.367
28.008
30.347
32.012
34.091
36.479
39.713
43.977
47.907
50.552
52.728
54.723
56.660
57.886
59.649
61.973
64.640
67.439
69.651
71.493
73.277
74.802
76.354
77.980
79.326
79.934
81.109
83.128
84.731
85.872
87.573
89.703
92.260
94.728
97.099
100.063
100.000
101.654
104.086
106.009
107.210
101.282
101.398
101.698
102.239
102.996
103.938
104.529
104.880
105.471
105.750
106.193
106.622
106.909
106.878
107.387
107.666

1 Quarterly percent changes are at annual rates.
Source: Department of Commerce (Bureau of Economic Analysis).

370  |  Appendix B

Gross domestic product (GDP)

Gross
domestic
PCE
purchases Real GDP
less
price
(chainfood and
index
type
energy
quantity
price
index)
index
19.325
19.761
20.367
21.240
22.237
23.281
24.377
25.164
26.125
28.196
30.557
32.414
34.494
36.801
39.478
43.092
46.856
49.880
52.465
54.644
56.897
58.849
60.717
63.288
65.868
68.491
70.885
73.019
75.006
76.679
78.323
79.799
81.194
82.198
83.290
84.744
86.277
87.749
89.048
90.751
92.710
94.785
96.829
98.824
100.000
101.287
102.743
104.632
105.935
100.911
101.179
101.427
101.632
101.959
102.522
103.039
103.452
104.010
104.482
104.849
105.187
105.542
105.711
106.077
106.410

18.321
18.829
19.346
20.163
21.149
22.287
23.449
24.498
25.888
28.510
31.116
32.821
34.977
37.459
40.729
44.962
49.087
51.875
53.696
55.482
57.150
58.345
59.985
62.091
64.515
67.039
69.111
70.719
72.323
73.835
75.420
76.728
77.851
78.358
79.578
81.641
83.206
84.359
86.196
88.729
91.850
94.782
97.370
100.243
100.000
101.528
103.884
105.599
106.852
101.036
101.285
101.609
102.183
102.900
103.792
104.307
104.538
105.124
105.383
105.742
106.150
106.467
106.526
107.010
107.406

6.5
6.6
2.7
4.9
3.1
.2
3.3
5.2
5.6
–.5
–.2
5.4
4.6
5.6
3.2
–.2
2.6
–1.9
4.6
7.3
4.2
3.5
3.5
4.2
3.7
1.9
–.1
3.6
2.7
4.0
2.7
3.8
4.5
4.4
4.8
4.1
1.0
1.8
2.8
3.8
3.4
2.7
1.8
–.3
–2.8
2.5
1.8
2.8
1.9
1.6
3.9
2.8
2.8
–1.3
3.2
1.4
4.9
3.7
1.2
2.8
.1
1.1
2.5
4.1
2.4

GDP
chaintype
price
index
1.8
2.8
2.9
4.3
4.9
5.3
5.1
4.3
5.4
9.0
9.3
5.5
6.2
7.0
8.3
9.0
9.4
6.1
3.9
3.6
3.2
2.0
2.4
3.5
3.9
3.7
3.3
2.3
2.4
2.1
2.1
1.8
1.7
1.1
1.4
2.3
2.3
1.5
2.0
2.7
3.2
3.1
2.7
1.9
.8
1.2
2.0
1.7
1.4
1.3
1.9
1.8
2.1
1.6
2.6
2.5
.5
2.0
1.8
2.3
1.1
1.3
.6
2.0
1.6

GDP
implicit
price
deflator
1.8
2.8
2.9
4.3
4.9
5.3
5.1
4.3
5.4
9.0
9.3
5.5
6.2
7.0
8.3
9.0
9.3
6.2
3.9
3.5
3.2
2.0
2.6
3.5
3.9
3.7
3.3
2.3
2.4
2.1
2.1
1.8
1.7
1.1
1.4
2.3
2.3
1.5
2.0
2.7
3.2
3.1
2.7
2.0
.8
1.2
2.0
1.7
1.5
1.4
1.8
1.8
2.1
1.6
2.6
2.5
.5
2.0
1.8
2.1
1.4
1.7
.6
2.0
1.6

Personal consumption expenditures
(PCE)
PCE
chaintype
price
index
1.4
2.5
2.5
3.9
4.5
4.7
4.3
3.4
5.4
10.4
8.4
5.5
6.5
7.0
8.9
10.7
8.9
5.5
4.3
3.8
3.5
2.2
3.0
3.9
4.3
4.3
3.3
2.6
2.5
2.1
2.1
2.1
1.7
.8
1.5
2.5
1.9
1.3
2.0
2.4
2.9
2.7
2.5
3.1
–.1
1.7
2.4
1.8
1.1
1.4
.5
1.2
2.1
3.0
3.7
2.3
1.3
2.3
1.1
1.7
1.6
1.1
–.1
1.9
1.0

Gross
domestic
PCE
purchases
less
price
food and
index
energy
price
index
1.3
2.3
3.1
4.3
4.7
4.7
4.7
3.2
3.8
7.9
8.4
6.1
6.4
6.7
7.3
9.2
8.7
6.5
5.2
4.2
4.1
3.4
3.2
4.2
4.1
4.0
3.5
3.0
2.7
2.2
2.1
1.9
1.7
1.2
1.3
1.7
1.8
1.7
1.5
1.9
2.2
2.2
2.2
2.1
1.2
1.3
1.4
1.8
1.2
1.0
1.1
1.0
.8
1.3
2.2
2.0
1.6
2.2
1.8
1.4
1.3
1.4
.6
1.4
1.3

1.7
2.8
2.7
4.2
4.9
5.4
5.2
4.5
5.7
10.1
9.1
5.5
6.6
7.1
8.7
10.4
9.2
5.7
3.5
3.3
3.0
2.1
2.8
3.5
3.9
3.9
3.1
2.3
2.3
2.1
2.1
1.7
1.5
.7
1.6
2.6
1.9
1.4
2.2
2.9
3.5
3.2
2.7
3.0
–.2
1.5
2.3
1.7
1.2
1.8
1.0
1.3
2.3
2.8
3.5
2.0
.9
2.3
1.0
1.4
1.6
1.2
.2
1.8
1.5

Table B–4. Growth rates in real gross domestic product by area and country, 1995–2014
[Percent change]

Area and country

World ��������������������������������������������������������������������������������������������
Advanced economies ������������������������������������������������������������
Of which:
United States ������������������������������������������������������������������
Euro area 2 ����������������������������������������������������������������������
Germany �������������������������������������������������������������������
France �����������������������������������������������������������������������
Italy ���������������������������������������������������������������������������
Spain �������������������������������������������������������������������������
Japan ������������������������������������������������������������������������������
United Kingdom ��������������������������������������������������������������
Canada ����������������������������������������������������������������������������
Other advanced economies ��������������������������������������������
Emerging market and developing economies �����������������������
Regional groups:
Central and eastern Europe ��������������������������������������������
Commonwealth of Independent States 3 �����������������������
Russia �����������������������������������������������������������������������
Excluding Russia �������������������������������������������������������
Developing Asia �������������������������������������������������������������
China �������������������������������������������������������������������������
India 4 ������������������������������������������������������������������������
ASEAN-5 5 ����������������������������������������������������������������
Latin America and the Caribbean �����������������������������������
Brazil �������������������������������������������������������������������������
Mexico ����������������������������������������������������������������������
Middle East, North Africa, Afghanistan, and Pakistan ��
Sub-Saharan Africa ��������������������������������������������������������
South Africa ��������������������������������������������������������������

1995–
2004
annual
average

2005

2006

2007

2008

2009

2010

2011

3.6
2.8

4.7
2.8

5.2
3.0

5.3
2.7

2.7
.1

–.4
–3.4

5.2
3.0

3.9
1.7

3.1
1.4

3.0
1.3

3.7
2.2

3.4
2.2
1.3
2.2
1.6
3.7
1.1
3.4
3.2
4.0
4.9

3.4
1.7
.8
1.8
.9
3.6
1.3
3.2
3.2
4.2
7.3

2.7
3.2
3.9
2.5
2.2
4.1
1.7
2.8
2.6
4.8
8.3

1.8
3.0
3.4
2.3
1.7
3.5
2.2
3.4
2.0
5.0
8.7

–.3
.4
.8
–.1
–1.2
.9
–1.0
–.8
1.2
1.7
5.8

–2.8
–4.4
–5.1
–3.1
–5.5
–3.8
–5.5
–5.2
–2.7
–1.1
3.1

2.5
2.0
3.9
1.7
1.7
–.2
4.7
1.7
3.4
5.9
7.5

1.8
1.5
3.4
2.0
.4
.1
–.6
1.1
2.5
3.2
6.2

2.8
–.7
.9
.0
–2.5
–1.6
1.4
.3
1.7
1.9
4.9

1.9
–.4
.5
.2
–1.8
–1.2
1.7
1.7
1.7
2.2
4.7

2.8
1.0
1.6
.9
.6
.6
1.7
2.4
2.2
3.0
5.1

4.0
2.9
2.8
3.2
7.1
9.2
6.2
4.0
2.5
2.5
2.4
4.6
4.5
3.1

5.9
6.7
6.4
7.7
9.5
11.3
9.3
5.4
4.7
3.2
3.2
6.0
6.3
5.3

6.4
8.8
8.2
10.6
10.3
12.7
9.3
5.5
5.6
4.0
5.0
6.7
6.4
5.6

5.4
8.9
8.5
9.9
11.5
14.2
9.8
6.2
5.7
6.1
3.1
5.9
7.1
5.5

3.2
5.3
5.2
5.6
7.3
9.6
3.9
4.7
4.2
5.2
1.2
5.0
5.7
3.6

–3.6
–6.4
–7.8
–3.1
7.7
9.2
8.5
1.8
–1.2
–.3
–4.5
2.8
2.6
–1.5

4.6
4.9
4.5
6.0
9.8
10.4
10.5
7.0
6.0
7.5
5.1
5.2
5.6
3.1

5.4
4.8
4.3
6.1
7.8
9.3
6.3
4.5
4.6
2.7
4.0
3.9
5.5
3.5

1.4
3.4
3.4
3.3
6.4
7.7
3.2
6.2
3.0
1.0
3.7
4.1
4.8
2.5

2.5
2.1
1.5
3.5
6.5
7.7
4.4
5.0
2.6
2.3
1.2
2.4
5.1
1.8

2.8
2.6
2.0
4.0
6.7
7.5
5.4
5.1
3.0
2.3
3.0
3.3
6.1
2.8

2012 2013 1 2014 1

1 All figures are forecasts as published by the International Monetary Fund. For the United States, the second estimate by the Department of Commerce
shows that real GDP rose 1.9 percent in 2013.
2 In 2014, consists of: Austria, Belgium, Cyprus, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Luxembourg, Malta, Netherlands, Portugal,
Slovak Republic, Slovenia, and Spain.
3 Includes Georgia, which is not a member of the Commonwealth of Independent States but is included for reasons of geography and similarity in economic
structure.
4 Data and forecasts are presented on a fiscal year basis and output growth is based on GDP at market prices.
5 Consists of Indonesia, Malaysia, Philippines, Thailand, and Vietnam.
Note: For details on data shown in this table, see World Economic Outlook, October 2013, and World Economic Outlook Update, January 2014, published by
the International Monetary Fund.
Sources: Department of Commerce (Bureau of Economic Analysis) and International Monetary Fund.

GDP, Income, Prices, and Selected Indicators  | 371

Table B–5. Real exports and imports of goods and services, 1999–2013
[Billions of chained (2009) dollars; quarterly data at seasonally adjusted annual rates]
Exports of goods and services
Year or quarter

1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2010: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������

      IV p �������������

Imports of goods and services

Goods 1
Total

1,174.1
1,272.4
1,200.5
1,178.1
1,197.2
1,309.3
1,388.4
1,512.4
1,647.3
1,741.8
1,583.8
1,765.6
1,890.5
1,957.4
2,010.0
1,700.4
1,739.3
1,784.9
1,837.7
1,854.7
1,876.9
1,908.9
1,921.7
1,941.4
1,959.8
1,961.6
1,967.0
1,960.5
1,998.4
2,017.6
2,063.5

Total
819.0
902.0
846.5
817.1
832.4
902.8
969.2
1,060.5
1,140.4
1,210.4
1,064.7
1,217.2
1,303.9
1,353.2
1,384.9
1,170.6
1,203.3
1,228.4
1,266.4
1,280.0
1,291.6
1,309.8
1,334.3
1,340.2
1,357.3
1,362.8
1,352.6
1,342.8
1,373.4
1,392.2
1,431.3

Durable
goods
533.8
599.3
549.5
518.7
528.0
586.0
641.1
710.1
771.1
810.4
671.9
784.6
855.5
896.4
913.7
743.5
781.5
794.9
818.5
832.0
851.3
864.9
873.9
897.5
898.3
897.8
892.0
890.5
921.2
916.6
926.7

Goods 1
Nondurable
goods
287.7
301.9
300.1
305.1
311.3
321.6
331.8
353.6
372.6
402.9
392.8
433.1
450.9
460.9
474.9
427.1
422.7
434.2
448.5
449.1
443.5
448.3
462.7
448.4
462.9
468.3
464.0
456.7
458.4
479.0
505.4

Services 1

354.4
368.2
352.3
360.5
364.1
406.3
418.4
450.8
506.2
530.5
519.1
548.1
586.3
603.7
624.8
529.6
535.6
556.3
571.0
574.3
585.0
599.2
586.6
600.7
601.9
598.0
614.2
617.5
624.9
625.1
631.8

Total

1,556.4
1,755.1
1,704.7
1,763.0
1,838.8
2,041.2
2,165.5
2,298.6
2,350.9
2,288.7
1,976.0
2,228.1
2,336.4
2,388.2
2,422.3
2,113.9
2,213.6
2,289.8
2,295.2
2,311.3
2,315.2
2,342.8
2,376.4
2,380.6
2,395.1
2,398.0
2,379.1
2,382.7
2,422.9
2,437.3
2,446.2

Total
1,286.1
1,454.4
1,407.3
1,459.9
1,531.3
1,701.4
1,814.7
1,922.2
1,957.5
1,885.1
1,587.3
1,828.0
1,923.4
1,964.3
1,988.4
1,722.9
1,818.4
1,881.4
1,889.2
1,909.8
1,906.5
1,923.1
1,954.4
1,958.6
1,970.7
1,972.7
1,955.1
1,954.0
1,989.6
2,001.4
2,008.9

Durable
goods
724.1
833.9
781.6
814.7
849.7
968.1
1,050.2
1,143.8
1,172.9
1,128.2
893.1
1,096.6
1,194.6
1,280.6
1,327.9
1,008.5
1,084.4
1,134.6
1,158.8
1,172.1
1,170.2
1,202.9
1,233.0
1,268.1
1,284.6
1,282.1
1,287.6
1,284.6
1,324.2
1,342.1
1,360.7

Nondurable
goods
572.2
623.9
640.5
658.7
697.9
744.0
773.0
785.9
792.3
764.2
694.2
735.7
740.7
710.3
694.7
715.6
737.4
751.8
737.9
745.9
744.8
734.2
738.0
715.2
712.8
716.3
696.8
698.1
698.5
695.0
687.3

Services 1

267.7
297.2
294.7
300.0
303.8
335.7
346.1
371.6
389.0
401.1
388.7
399.4
411.8
422.8
433.2
390.7
394.4
407.5
404.9
399.8
407.4
419.0
420.9
420.8
423.2
424.2
423.1
428.3
432.6
435.2
436.7

1 Certain goods, primarily military equipment purchased and sold by the Federal Government, are included in services. Repairs and alterations of equipment
are also included in services.
Source: Department of Commerce (Bureau of Economic Analysis).

372  |  Appendix B

Table B–6. Corporate profits by industry, 1965–2013
[Billions of dollars; quarterly data at seasonally adjusted annual rates]
Corporate profits with inventory valuation adjustment and without capital consumption adjustment
Domestic industries
Year or quarter

SIC: 2
1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
NAICS: 2
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2011: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2012: I ������������������

      II �����������������

      III ����������������

      IV ���������������
�
2013: I ������������������

      II �����������������

      III ����������������


Total

Financial
Total

Total

Federal
Reserve
banks

Nonfinancial
Other

Total

Manufacturing

TransWholeporta- Utilities sale
tion 1
trade

Retail
trade

Information

Other

Rest
of
the
world

81.9
88.3
86.1
94.3
90.8
79.7
94.7
109.3
126.6
123.3
144.2
182.1
212.8
246.7
261.0
240.6
252.0
224.8
256.4
294.3
289.7
273.3
314.6
366.2
373.1
391.2
434.2
459.7
501.9
589.3
667.0
741.8
811.0
743.8
762.2
730.3

77.2
83.7
81.3
88.6
84.2
72.6
86.8
99.7
111.7
105.8
129.6
165.6
193.7
223.8
226.4
205.2
222.3
192.2
221.4
257.7
251.6
233.8
266.5
309.2
305.9
315.1
357.8
386.6
425.0
511.3
574.0
639.8
703.4
641.1
640.2
584.1

9.3
10.7
11.2
12.9
13.6
15.5
17.9
19.5
21.1
20.8
20.4
25.6
32.6
40.8
41.8
35.2
30.3
27.2
36.2
34.7
46.5
56.4
60.3
66.9
78.3
89.6
120.4
132.4
119.9
125.9
140.3
147.9
162.2
138.9
154.6
149.7

1.3
1.7
2.0
2.5
3.1
3.5
3.3
3.3
4.5
5.7
5.6
5.9
6.1
7.6
9.4
11.8
14.4
15.2
14.6
16.4
16.3
15.5
16.2
18.1
20.6
21.8
20.7
18.3
16.7
18.5
22.9
22.5
24.3
25.6
26.7
31.2

8.0
9.1
9.2
10.4
10.6
12.0
14.6
16.1
16.6
15.1
14.8
19.7
26.5
33.1
32.3
23.5
15.9
12.0
21.6
18.3
30.2
40.8
44.1
48.8
57.6
67.8
99.7
114.1
103.2
107.4
117.3
125.3
137.9
113.3
127.9
118.5

67.9
73.0
70.1
75.7
70.6
57.1
69.0
80.3
90.6
85.1
109.2
140.0
161.1
183.1
184.6
169.9
192.0
165.0
185.2
223.0
205.1
177.4
206.2
242.3
227.6
225.5
237.3
254.2
305.1
385.4
433.7
492.0
541.2
502.1
485.6
434.4

42.1
45.3
42.4
45.8
41.6
32.0
40.0
47.6
55.0
51.0
63.0
82.5
91.5
105.8
107.1
97.6
112.5
89.6
97.3
114.2
107.1
75.6
101.8
132.8
122.3
120.9
109.3
109.8
122.9
162.6
199.8
220.4
248.5
220.4
219.4
205.9

11.4
12.6
11.4
11.4
11.1
8.8
9.6
10.4
10.2
9.1
11.7
17.5
21.2
25.5
21.6
22.2
25.1
28.1
34.3
44.7
39.1
39.3
42.0
46.8
41.9
43.5
54.5
57.7
70.1
83.9
89.0
91.2
81.0
72.6
49.3
33.8

������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������

3.8
4.0
4.1
4.7
4.9
4.6
5.4
7.2
8.8
12.2
14.3
13.7
16.4
16.7
20.0
18.5
23.7
20.7
21.9
30.4
24.6
24.4
18.9
20.4
22.0
19.4
22.3
25.3
26.5
31.4
28.0
39.9
48.1
50.6
46.8
50.4

4.9
4.9
5.7
6.4
6.4
6.1
7.3
7.5
7.0
2.8
8.4
10.9
12.8
13.1
10.7
7.0
10.7
14.3
19.3
21.5
22.8
23.4
23.3
19.8
20.9
20.3
26.9
28.1
39.7
46.3
43.9
52.0
63.4
72.3
72.5
68.9

������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������
������������

5.7
6.3
6.6
7.4
6.5
5.8
6.7
7.6
9.6
10.0
11.8
15.3
19.2
22.0
25.2
24.6
20.1
12.3
12.3
12.1
11.4
14.7
20.3
22.5
20.5
21.3
24.3
33.4
45.8
61.2
73.1
88.5
100.3
86.3
97.6
75.4

4.7
4.5
4.8
5.6
6.6
7.1
7.9
9.5
14.9
17.5
14.6
16.5
19.1
22.9
34.6
35.5
29.7
32.6
35.1
36.6
38.1
39.5
48.0
57.0
67.1
76.1
76.5
73.1
76.9
78.0
92.9
102.0
107.6
102.8
122.0
146.2

743.8
762.2
730.3
698.7
795.1
959.9
1,215.2
1,621.2
1,815.7
1,708.9
1,345.5
1,474.8
1,793.8
1,791.3
2,180.0
1,672.2
1,782.3
1,805.4
1,905.4
2,142.5
2,169.8
2,186.6
2,221.1
2,180.0
2,248.6
2,288.2

641.1
640.2
584.1
528.3
636.3
793.3
1,010.1
1,382.1
1,559.6
1,355.5
938.8
1,122.0
1,398.6
1,354.8
1,761.1
1,244.3
1,354.9
1,354.6
1,465.2
1,726.7
1,740.5
1,774.0
1,803.0
1,781.5
1,845.5
1,868.4

138.9
154.6
149.7
195.0
270.7
306.5
349.4
409.7
415.1
301.5
95.4
362.9
405.3
384.1
477.4
377.8
364.6
348.8
445.1
462.5
447.7
507.2
492.1
486.9
511.9
521.6

25.6
26.7
31.2
28.9
23.5
20.1
20.0
26.6
33.8
36.0
35.1
47.3
71.6
75.9
71.7
72.4
80.0
76.6
74.7
73.4
72.6
67.5
73.3
70.0
82.1
90.4

113.3
127.9
118.5
166.1
247.2
286.5
329.4
383.1
381.3
265.5
60.4
315.5
333.8
308.1
405.7
305.4
284.6
272.2
370.4
389.1
375.1
439.8
418.7
416.9
429.8
431.2

502.1
485.6
434.4
333.3
365.6
486.7
660.7
972.4
1,144.4
1,054.0
843.4
759.2
993.3
970.7
1,283.7
866.5
990.3
1,005.8
1,020.1
1,264.2
1,292.8
1,266.8
1,310.9
1,294.6
1,333.6
1,346.8

193.5
184.5
175.6
75.1
75.1
125.3
182.7
277.7
349.7
321.9
240.6
171.4
284.9
303.9
404.3
278.1
291.5
314.5
331.7
408.7
410.5
387.8
410.1
389.7
381.8
392.4

12.8
7.2
9.5
–.7
–6.0
4.8
12.0
27.7
41.2
23.9
28.8
22.4
44.6
32.1
51.5
29.8
33.3
30.3
35.1
53.4
53.5
52.2
47.1
54.5
57.6
61.3

33.3
34.4
24.3
22.5
11.1
13.5
20.5
30.8
55.1
49.5
30.1
23.8
29.8
11.1
37.1
3.9
29.7
3.2
7.9
34.5
39.4
40.8
33.6
38.3
47.2
50.2

57.3
55.6
59.5
51.1
55.8
59.3
74.7
96.2
105.9
103.2
90.6
89.3
102.2
96.3
137.8
74.4
94.7
110.3
105.9
128.8
146.5
131.6
144.4
150.2
151.1
154.7

62.5
59.5
51.3
71.3
83.7
90.5
93.2
121.7
132.5
119.0
80.3
108.7
118.3
116.1
149.2
112.2
109.1
114.9
128.2
149.9
145.3
142.5
159.0
148.9
169.9
166.0

33.1
20.8
–11.9
–26.4
–3.1
16.3
52.7
91.3
107.0
108.4
92.2
81.2
94.7
87.4
110.6
85.3
92.4
86.7
85.1
110.3
116.6
112.9
102.5
124.2
131.8
118.3

109.7
123.5
126.1
140.2
149.0
177.1
224.9
327.2
353.1
328.2
280.8
262.3
318.7
323.7
393.2
283.0
339.5
346.0
326.2
378.6
381.0
399.0
414.2
388.9
394.2
403.9

102.8
122.0
146.2
170.4
158.8
166.6
205.0
239.1
256.2
353.4
406.7
352.8
395.2
436.6
418.9
427.8
427.3
450.8
440.2
415.9
429.3
412.5
418.1
398.5
403.1
419.8

1 Data on Standard Industrial Classification (SIC) basis include transportation and public utilities. Those on North American Industry Classification System
(NAICS) basis include transporation and warehousing. Utilities classified separately in NAICS (as shown beginning 1998).
2 SIC-based industry data use the 1987 SIC for data beginning in 1987 and the 1972 SIC for prior data. NAICS-based data use 2002 NAICS.
Note: Industry data on SIC basis and NAICS basis are not necessarily the same and are not strictly comparable.
Source: Department of Commerce (Bureau of Economic Analysis).

GDP, Income, Prices, and Selected Indicators  | 373

Table B–7. Real farm income, 1950–2014
[Billions of chained (2009) dollars]
Income of farm operators from farming 1
Gross farm income
Year

Value of farm sector production
Total 2

1950 ����������������������
1951 ����������������������
1952 ����������������������
1953 ����������������������
1954 ����������������������
1955 ����������������������
1956 ����������������������
1957 ����������������������
1958 ����������������������
1959 ����������������������
1960 ����������������������
1961 ����������������������
1962 ����������������������
1963 ����������������������
1964 ����������������������
1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010  ���������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2014 p ��������������������

240.8
260.9
251.7
226.8
222.7
215.1
210.9
208.8
228.5
219.3
220.3
229.0
236.2
239.2
229.8
248.3
261.9
254.8
250.8
260.0
257.6
258.9
284.1
374.7
341.6
319.9
310.4
308.9
340.8
369.5
335.6
341.8
317.9
286.7
302.3
280.9
266.8
281.0
286.8
297.3
295.9
278.1
283.9
283.5
292.6
279.6
307.1
304.8
294.6
293.4
295.1
298.3
271.1
298.2
330.8
324.5
306.0
348.8
380.7
343.3
361.1
417.1
433.2
453.0
408.1

Total
238.8
258.9
249.9
225.4
221.1
213.6
207.5
202.7
222.1
215.4
216.3
220.5
226.5
229.9
218.0
235.2
244.9
239.2
234.0
242.6
241.3
245.8
268.3
364.8
339.7
317.3
308.1
303.7
332.8
366.1
332.7
337.8
311.2
269.4
287.1
267.4
246.6
253.0
263.4
280.4
282.0
266.2
270.9
265.0
281.9
270.0
297.6
295.2
278.9
266.5
266.7
271.5
256.5
279.2
316.3
298.0
289.4
336.6
368.3
331.2
348.9
407.0
423.1
442.5
402.4

Crops 3, 4
96.0
95.6
102.2
93.1
94.0
91.6
89.7
81.9
88.0
85.5
89.5
89.3
92.9
98.9
91.7
101.5
95.0
96.9
91.5
90.7
89.9
97.6
103.6
163.1
170.9
160.3
145.8
145.3
150.2
163.4
144.7
162.2
139.1
106.0
139.9
128.4
108.2
107.6
111.6
126.4
124.5
117.6
126.1
114.3
136.0
127.2
150.7
144.1
129.3
115.9
116.0
113.4
115.1
125.2
140.4
124.3
125.2
155.2
184.5
168.6
170.7
200.5
206.6
216.0
178.9

Livestock 4
132.0
152.1
135.7
120.1
115.3
110.0
106.2
109.0
121.9
116.8
113.5
117.4
119.5
116.4
111.2
118.4
134.2
126.0
126.3
135.2
134.7
131.1
147.4
183.2
148.9
136.8
140.6
134.4
156.2
174.5
158.1
144.7
136.5
130.5
129.6
120.3
120.9
126.4
126.7
129.5
134.7
126.3
123.4
127.2
121.5
116.4
119.9
123.3
119.3
118.9
121.0
127.0
109.9
121.0
139.4
137.5
125.8
142.2
141.5
119.8
139.1
159.5
162.3
169.9
169.6

Forestry
and
services
10.8
11.2
12.0
12.2
11.8
12.0
11.6
11.7
12.2
13.1
13.4
13.8
14.0
14.6
15.1
15.3
15.6
16.3
16.2
16.6
16.7
17.0
17.3
18.5
20.0
20.2
21.7
24.1
26.4
28.2
29.9
31.0
35.5
32.9
17.6
18.7
17.5
19.1
25.0
24.5
22.8
22.3
21.5
23.5
24.4
26.4
27.0
27.8
30.3
31.8
29.8
31.1
31.5
33.0
36.5
36.1
38.3
39.2
42.3
42.7
39.0
47.1
54.2
56.6
53.9

Direct
Government
payments
2.1
1.9
1.8
1.4
1.7
1.5
3.4
6.1
6.4
3.9
4.0
8.4
9.7
9.4
11.8
13.1
17.0
15.5
16.7
17.5
16.3
13.1
15.8
9.9
1.8
2.6
2.2
5.2
8.0
3.4
2.9
4.0
6.8
17.3
15.2
13.4
20.2
27.9
23.3
16.9
13.9
11.9
13.0
18.5
10.7
9.7
9.6
9.6
15.7
26.9
28.4
26.8
14.6
19.0
14.6
26.5
16.7
12.2
12.3
12.2
12.2
10.1
10.1
10.5
5.6

Production
expenses

Net
farm
income

141.5
152.3
152.0
141.3
142.1
142.4
141.0
142.2
151.3
157.3
156.3
161.4
168.9
174.3
172.8
179.5
189.5
192.5
191.2
194.1
194.7
196.3
206.4
244.5
246.8
238.7
249.5
252.4
274.0
302.3
299.3
286.5
271.8
260.1
255.5
231.2
213.7
217.6
222.9
225.1
226.7
219.8
212.9
218.9
221.4
226.9
230.4
239.1
234.9
233.8
233.2
232.8
225.1
227.9
232.8
238.9
245.5
276.9
296.3
283.0
284.0
302.8
324.8
330.5
320.0

1 The GDP chain-type price index is used to convert the current-dollar statistics to 2009=100 equivalents.
2 Value of production, Government payments, other farm-related cash income, and nonmoney income produced by farms including imputed rent of farm

dwellings.
3 Crop receipts include proceeds received from commodities placed under Commodity Credit Corporation loans.
4 The value of production equates to the sum of cash receipts, home consumption, and the value of the change in inventories.
Note: Data for 2013 and 2014 are forecasts.
Source: Department of Agriculture (Economic Research Service).

374  |  Appendix B

99.3
108.6
99.8
85.4
80.6
72.6
69.9
66.5
77.2
62.0
64.0
67.5
67.3
64.9
57.0
68.8
72.4
62.2
59.6
65.9
62.9
62.6
77.7
130.1
94.8
81.2
60.8
56.5
66.9
67.2
36.3
55.2
46.2
26.6
46.7
49.7
53.2
63.4
63.9
72.1
69.2
58.3
71.0
64.6
71.2
52.7
76.8
65.7
59.7
59.6
61.9
65.5
46.0
70.3
98.1
85.6
60.6
71.9
84.3
60.4
77.1
114.3
108.4
122.5
88.0

Table B–8. New private housing units started, authorized, and completed and houses sold,
1970–2014
[Thousands; monthly data at seasonally adjusted annual rates]
New housing units started
Type of structure

Year or month
Total
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2012: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2013: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec p ����������
2014: Jan p �����������

1,433.6
2,052.2
2,356.6
2,045.3
1,337.7
1,160.4
1,537.5
1,987.1
2,020.3
1,745.1
1,292.2
1,084.2
1,062.2
1,703.0
1,749.5
1,741.8
1,805.4
1,620.5
1,488.1
1,376.1
1,192.7
1,013.9
1,199.7
1,287.6
1,457.0
1,354.1
1,476.8
1,474.0
1,616.9
1,640.9
1,568.7
1,602.7
1,704.9
1,847.7
1,955.8
2,068.3
1,800.9
1,355.0
905.5
554.0
586.9
608.8
780.6
926.7
723
713
707
754
711
757
741
749
854
864
842
983
898
969
1,005
852
919
835
891
883
873
899
1,101
1,048
880

New housing units authorized 1
Type of structure

1 unit
812.9
1,151.0
1,309.2
1,132.0
888.1
892.2
1,162.4
1,450.9
1,433.3
1,194.1
852.2
705.4
662.6
1,067.6
1,084.2
1,072.4
1,179.4
1,146.4
1,081.3
1,003.3
894.8
840.4
1,029.9
1,125.7
1,198.4
1,076.2
1,160.9
1,133.7
1,271.4
1,302.4
1,230.9
1,273.3
1,358.6
1,499.0
1,610.5
1,715.8
1,465.4
1,046.0
622.0
445.1
471.2
430.6
535.3
618.3
513
462
483
505
515
530
512
537
591
595
576
620
614
652
623
593
597
605
587
620
580
600
713
681
573

2 to 4
units 2
84.9
120.5
141.2
118.2
68.0
64.0
85.8
121.7
125.1
122.0
109.5
91.2
80.1
113.5
121.4
93.5
84.0
65.1
58.7
55.3
37.6
35.6
30.9
29.4
35.2
33.8
45.3
44.5
42.6
31.9
38.7
36.6
38.5
33.5
42.3
41.1
42.7
31.7
17.5
11.6
11.4
10.9
11.4
13.8
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������

5 units
or more

Total

535.9
780.9
906.2
795.0
381.6
204.3
289.2
414.4
462.0
429.0
330.5
287.7
319.6
522.0
543.9
576.0
542.0
408.7
348.0
317.6
260.4
137.9
139.0
132.6
223.5
244.1
270.8
295.8
302.9
306.6
299.1
292.8
307.9
315.2
303.0
311.4
292.8
277.3
266.0
97.3
104.3
167.3
233.9
294.6
194
243
214
240
181
219
217
205
254
252
256
345
273
307
356
244
311
219
285
251
283
289
379
344
300

1,351.5
1,924.6
2,218.9
1,819.5
1,074.4
939.2
1,296.2
1,690.0
1,800.5
1,551.8
1,190.6
985.5
1,000.5
1,605.2
1,681.8
1,733.3
1,769.4
1,534.8
1,455.6
1,338.4
1,110.8
948.8
1,094.9
1,199.1
1,371.6
1,332.5
1,425.6
1,441.1
1,612.3
1,663.5
1,592.3
1,636.7
1,747.7
1,889.2
2,070.1
2,155.3
1,838.9
1,398.4
905.4
583.0
604.6
624.1
829.7
976.4
714
739
785
749
806
785
839
827
921
908
933
943
915
952
890
1,005
985
918
954
926
974
1,039
1,017
991
945

1 unit
646.8
906.1
1,033.1
882.1
643.8
675.5
893.6
1,126.1
1,182.6
981.5
710.4
564.3
546.4
901.5
922.4
956.6
1,077.6
1,024.4
993.8
931.7
793.9
753.5
910.7
986.5
1,068.5
997.3
1,069.5
1,062.4
1,187.6
1,246.7
1,198.1
1,235.6
1,332.6
1,460.9
1,613.4
1,682.0
1,378.2
979.9
575.6
441.1
447.3
418.5
518.7
617.5
461
486
477
484
499
501
520
520
559
570
574
584
588
600
599
614
620
625
609
627
615
621
641
610
599

2 to 4
units
88.1
132.9
148.6
117.0
64.4
63.8
93.1
121.3
130.6
125.4
114.5
101.8
88.3
133.7
142.6
120.1
108.4
89.3
75.7
66.9
54.3
43.1
45.8
52.4
62.2
63.8
65.8
68.4
69.2
65.8
64.9
66.0
73.7
82.5
90.4
84.0
76.6
59.6
34.4
20.7
22.0
21.6
25.9
26.6
22
26
23
23
23
24
29
28
29
26
29
30
26
31
25
25
27
26
27
23
28
27
24
26
27

5 units
or more
616.7
885.7
1,037.2
820.5
366.2
199.8
309.5
442.7
487.3
444.8
365.7
319.4
365.8
570.1
616.8
656.6
583.5
421.1
386.1
339.8
262.6
152.1
138.4
160.2
241.0
271.5
290.3
310.3
355.5
351.1
329.3
335.2
341.4
345.8
366.2
389.3
384.1
359.0
295.4
121.1
135.3
184.0
285.1
332.3
231
227
285
242
284
260
290
279
333
312
330
329
301
321
266
366
338
267
318
276
331
391
352
355
319

New
housing
units
completed
1,418.4
1,706.1
2,003.9
2,100.5
1,728.5
1,317.2
1,377.2
1,657.1
1,867.5
1,870.8
1,501.6
1,265.7
1,005.5
1,390.3
1,652.2
1,703.3
1,756.4
1,668.8
1,529.8
1,422.8
1,308.0
1,090.8
1,157.5
1,192.7
1,346.9
1,312.6
1,412.9
1,400.5
1,474.2
1,604.9
1,573.7
1,570.8
1,648.4
1,678.7
1,841.9
1,931.4
1,979.4
1,502.8
1,119.7
794.4
651.7
584.9
649.2
765.1
540
566
588
667
613
628
673
686
651
741
677
672
720
727
810
698
711
759
783
765
762
814
826
778
814

New
houses
sold
485
656
718
634
519
549
646
819
817
709
545
436
412
623
639
688
750
671
676
650
534
509
610
666
670
667
757
804
886
880
877
908
973
1,086
1,203
1,283
1,051
776
485
375
323
306
368
428
338
366
349
352
369
360
369
374
384
365
398
396
458
445
443
446
429
450
373
388
403
452
444
427
468

1 Authorized by issuance of local building permits in permit-issuing places: 20,000 places beginning with 2004; 19,000 for 1994–2003; 17,000 for 1984–93;
16,000 for 1978–83; 14,000 for 1972–77; and 13,000 for 1970–71.
2 Monthly data do not meet publication standards because tests for identifiable and stable seasonality do not meet reliability standards.
Note: One-unit estimates prior to 1999, for new housing units started and completed and for new houses sold, include an upward adjustment of 3.3 percent
to account for structures in permit-issuing areas that did not have permit authorization.
Source: Department of Commerce (Bureau of the Census).

GDP, Income, Prices, and Selected Indicators  | 375

Table B–9. Median money income (in 2012 dollars) and poverty status of families and
people, by race, 2003-2012
Families 1

People below
poverty level

Below poverty level
Race,
Hispanic origin,
and
year

Number
(millions)

TOTAL (all races) 3
2003 ���������������������������������������
2004 4 �������������������������������������
2005 ���������������������������������������
2006 ���������������������������������������
2007 ���������������������������������������
2008 ���������������������������������������
2009 5 �������������������������������������
2010 6 �������������������������������������
2011 ���������������������������������������
2012 ���������������������������������������
WHITE, non-Hispanic 7
2003 ���������������������������������������
2004 4 �������������������������������������
2005 ���������������������������������������
2006 ���������������������������������������
2007 ���������������������������������������
2008 ���������������������������������������
2009 5 �������������������������������������
2010 6 �������������������������������������
2011 ���������������������������������������
2012 ���������������������������������������
BLACK 7
2003 ���������������������������������������
2004 4 �������������������������������������
2005 ���������������������������������������
2006 ���������������������������������������
2007 ���������������������������������������
2008 ���������������������������������������
2009 5 �������������������������������������
2010 6 �������������������������������������
2011 ���������������������������������������
2012 ���������������������������������������
ASIAN 7
2003 ���������������������������������������
2004 4 �������������������������������������
2005 ���������������������������������������
2006 ���������������������������������������
2007 ���������������������������������������
2008 ���������������������������������������
2009 5 �������������������������������������
2010 6 �������������������������������������
2011 ���������������������������������������
2012 ���������������������������������������
HISPANIC (any race) 7
2003 ���������������������������������������
2004 4 �������������������������������������
2005 ���������������������������������������
2006 ���������������������������������������
2007 ���������������������������������������
2008 ���������������������������������������
2009 5 �������������������������������������
2010 6 �������������������������������������
2011 ���������������������������������������
2012 ���������������������������������������

Median
Female
money
householder,
Total
income
no husband
(in
present
Number
2012
(milPercent
dollions)
Number
lars) 2 Number
(milPercent
(milPercent
lions)
lions)

Median money income (in 2012 dollars)
of people 15 years old and over
with income 2
Males

All
people

Yearround
full-time
workers

Females

All
people

12.5 $37,367 $51,813 $21,547
12.7 37,094 50,649 21,476
12.6 36,784 49,619 21,848
12.3 36,744 51,198 22,792
12.5 36,761 51,188 23,169
13.2 35,363 50,952 22,253
14.3 34,452 52,629 22,434
15.1 33,915 52,814 21,878
15.0 33,675 51,367 21,543
15.0 33,904 50,683 21,520

Yearround
full-time
workers

76.2 $65,767
76.9 65,715
77.4 66,092
78.5 66,514
77.9 67,944
78.9 65,607
78.9 64,323
79.6 63,434
80.5 62,248
80.9 62,241

7.6
7.8
7.7
7.7
7.6
8.1
8.8
9.4
9.5
9.5

10.0
10.2
9.9
9.8
9.8
10.3
11.1
11.8
11.8
11.8

3.9
4.0
4.0
4.1
4.1
4.2
4.4
4.8
4.9
4.8

28.0
28.3
28.7
28.3
28.3
28.7
29.9
31.7
31.2
30.9

35.9
37.0
37.0
36.5
37.3
39.8
43.6
46.3
46.2
46.5

$39,516
39,039
39,114
39,846
40,051
39,125
39,858
40,480
39,493
40,019

54.0
54.3
54.3
54.7
53.9
54.5
54.5
53.8
54.2
54.0

74,827
74,143
74,280
75,008
77,447
74,724
72,087
72,561
71,288
71,478

3.3
3.5
3.3
3.4
3.2
3.4
3.8
3.9
4.0
3.8

6.1
6.5
6.1
6.2
5.9
6.2
7.0
7.2
7.3
7.1

1.5
1.5
1.5
1.6
1.5
1.5
1.7
1.7
1.8
1.7

20.4
20.8
21.5
22.0
20.7
20.7
23.3
24.1
23.4
23.4

15.9
16.9
16.2
16.0
16.0
17.0
18.5
19.3
19.2
18.9

8.2
8.7
8.3
8.2
8.2
8.6
9.4
9.9
9.8
9.7

40,363
40,938
41,571
41,639
41,386
39,893
39,377
39,127
38,945
38,751

57,795
57,110
56,595
57,446
56,992
55,822
56,167
57,555
56,928
56,247

22,847
22,409
22,877
23,604
24,016
23,193
23,485
22,868
22,690
22,902

42,493
42,451
42,102
42,006
42,832
42,091
43,103
43,526
42,237
42,171

8.9
8.9
9.1
9.3
9.3
9.4
9.4
9.6
9.7
9.8

42,907
42,725
41,711
43,581
44,454
42,528
41,116
40,643
41,341
40,517

2.0
2.0
2.0
2.0
2.0
2.1
2.1
2.3
2.3
2.3

22.3
22.8
22.1
21.6
22.1
22.0
22.7
24.1
24.2
23.7

1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.7
1.7
1.6

36.9
37.6
36.1
36.6
37.3
37.2
36.7
38.7
39.0
37.8

8.8
9.0
9.2
9.0
9.2
9.4
9.9
10.7
10.9
10.9

24.4
24.7
24.9
24.3
24.5
24.7
25.8
27.4
27.6
27.2

27,448
27,581
26,643
28,543
28,595
26,931
25,411
24,533
23,965
24,923

41,734
38,558
40,263
40,401
40,681
41,176
42,136
39,727
41,114
39,816

20,700
21,101
20,737
21,755
21,873
21,538
20,842
20,689
20,168
20,021

34,484
35,428
35,711
35,230
34,984
34,324
34,758
35,850
35,880
35,090

3.1
3.1
3.2
3.3
3.3
3.5
3.6
3.9
4.2
4.1

78,964
79,523
81,103
84,968
85,416
78,465
80,315
79,210
74,521
77,864

0.3
0.2
0.3
0.3
0.3
0.3
0.3
0.4
0.4
0.4

10.2
7.4
9.0
7.8
7.9
9.8
9.4
9.3
9.7
9.4

0.1
.0
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1

23.8
13.6
19.7
15.4
16.1
16.7
16.9
21.1
19.1
19.2

1.4
1.2
1.4
1.4
1.3
1.6
1.7
1.9
2.0
1.9

11.8
9.8
11.1
10.3
10.2
11.8
12.5
12.2
12.3
11.7

40,313
40,137
40,242
42,611
41,187
39,038
39,961
37,725
37,093
40,227

57,702
56,906
58,487
59,333
56,713
55,224
57,193
55,293
57,459
60,253

22,071
24,946
25,453
25,283
26,970
24,644
26,059
24,814
22,499
23,335

43,176
44,508
43,297
45,835
45,752
47,144
47,772
44,146
42,276
46,371

9.3
9.5
9.9
10.2
10.4
10.5
10.4
11.3
11.6
12.0

42,786
43,080
44,537
45,552
44,922
43,153
42,530
41,387
40,898
40,764

1.9
2.0
1.9
1.9
2.0
2.2
2.4
2.7
2.7
2.8

20.8
20.5
19.7
18.9
19.7
21.3
22.7
24.3
22.9
23.5

0.8
0.9
0.9
0.9
1.0
1.0
1.1
1.3
1.3
1.3

37.0
38.9
38.9
36.0
38.4
39.2
38.8
42.6
41.2
40.7

9.1
9.1
9.4
9.2
9.9
11.0
12.4
13.5
13.2
13.6

22.5
21.9
21.8
20.6
21.5
23.2
25.3
26.5
25.3
25.6

26,283
26,203
25,980
26,707
27,077
25,597
23,825
23,610
24,227
24,592

32,976
32,696
31,716
33,676
33,724
33,292
33,868
33,534
32,758
32,516

17,031
17,567
17,684
17,945
18,547
17,507
17,352
17,157
17,181
16,725

28,791
29,532
29,429
29,260
30,070
29,263
29,848
30,641
30,731
29,508

1 The term “family” refers to a group of two or more persons related by birth, marriage, or adoption and residing together. Every family must include a
reference person.
2 Adjusted by consumer price index research series (CPI-U-RS).
3 Data for American Indians and Alaska natives, native Hawaiians and other Pacific Islanders, and those reporting two or more races are included in the total
but not shown separately.
4 For 2004, figures are revised to reflect a correction to the weights in the 2005 Annual Social and Economic Supplement.
5 Beginning with data for 2009, the upper income interval used to calculate median incomes was expanded to $250,000 or more.
6 Reflects implementation of Census 2010-based population controls comparable to succeeding years.
7 The Current Population Survey allows respondents to choose more than one race. Data shown are for “white alone, non-Hispanic,” “black alone,” and
“Asian alone” race categories. (“Black” is also “black or African American.”) Family race and Hispanic origin are based on the reference person.
Note: Poverty thresholds are updated each year to reflect changes in the consumer price index (CPI-U).
For details see publication Series P–60 on the Current Population Survey and Annual Social and Economic Supplements.
Source: Department of Commerce (Bureau of the Census).

376  |  Appendix B

Table B–10. Changes in consumer price indexes, 1945–2013
[For all urban consumers; percent change]
December
to
December
1945 ����������������������
1946 ����������������������
1947 ����������������������
1948 ����������������������
1949 ����������������������
1950 ����������������������
1951 ����������������������
1952 ����������������������
1953 ����������������������
1954 ����������������������
1955 ����������������������
1956 ����������������������
1957 ����������������������
1958 ����������������������
1959 ����������������������
1960 ����������������������
1961 ����������������������
1962 ����������������������
1963 ����������������������
1964 ����������������������
1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 ����������������������

All items less food and energy
All items

2.2
18.1
8.8
3.0
–2.1
5.9
6.0
.8
.7
–.7
.4
3.0
2.9
1.8
1.7
1.4
.7
1.3
1.6
1.0
1.9
3.5
3.0
4.7
6.2
5.6
3.3
3.4
8.7
12.3
6.9
4.9
6.7
9.0
13.3
12.5
8.9
3.8
3.8
3.9
3.8
1.1
4.4
4.4
4.6
6.1
3.1
2.9
2.7
2.7
2.5
3.3
1.7
1.6
2.7
3.4
1.6
2.4
1.9
3.3
3.4
2.5
4.1
.1
2.7
1.5
3.0
1.7
1.5

Total 1

Shelter 2

Medical
care 3

���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
1.7
2.0
1.0
1.3
1.3
1.6
1.2
1.5
3.3
3.8
5.1
6.2
6.6
3.1
3.0
4.7
11.1
6.7
6.1
6.5
8.5
11.3
12.2
9.5
4.5
4.8
4.7
4.3
3.8
4.2
4.7
4.4
5.2
4.4
3.3
3.2
2.6
3.0
2.6
2.2
2.4
1.9
2.6
2.7
1.9
1.1
2.2
2.2
2.6
2.4
1.8
1.8
.8
2.2
1.9
1.7

���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
3.2
1.8
.9
2.6
3.4
.8
2.0
1.6
.8
.8
1.9
1.5
2.2
4.0
2.8
6.5
8.7
8.9
2.7
4.0
7.1
11.4
7.2
4.2
8.8
11.4
17.5
15.0
9.9
2.4
4.7
5.2
6.0
4.6
4.8
4.5
4.9
5.2
3.9
2.9
3.0
3.0
3.5
2.9
3.4
3.3
2.5
3.4
4.2
3.1
2.2
2.7
2.6
4.2
3.1
1.9
.3
.4
1.9
2.2
2.5

2.6
8.3
6.9
5.8
1.4
3.4
5.8
4.3
3.5
2.3
3.3
3.2
4.7
4.5
3.8
3.2
3.1
2.2
2.5
2.1
2.8
6.7
6.3
6.2
6.2
7.4
4.6
3.3
5.3
12.6
9.8
10.0
8.9
8.8
10.1
9.9
12.5
11.0
6.4
6.1
6.8
7.7
5.8
6.9
8.5
9.6
7.9
6.6
5.4
4.9
3.9
3.0
2.8
3.4
3.7
4.2
4.7
5.0
3.7
4.2
4.3
3.6
5.2
2.6
3.4
3.3
3.5
3.2
2.0

Apparel

Energy 4

Food
New
vehicles

4.9 ���������������
18.1 ���������������
8.2 ���������������
5.1
11.5
–7.4
4.0
5.3
.2
5.7
9.7
–2.9
4.4
.7
–1.7
–.7
1.3
.5
–2.3
2.5
7.8
.9
2.0
.2
6.1
1.3
–.2
1.5
–3.0
.4
.2
.6
–1.0
1.7
–.4
.4
–.6
1.3
–2.9
3.9
.0
4.2
2.8
6.3
1.4
5.2
2.1
3.9
6.6
2.1
–3.2
2.6
.2
4.4
1.3
8.7
11.4
2.4
7.3
4.6
4.8
4.3
7.2
3.1
6.2
5.5
7.4
6.8
7.4
3.5
6.8
1.6
1.4
2.9
3.3
2.0
2.5
2.8
3.6
.9
5.6
4.8
1.8
4.7
2.2
1.0
2.4
5.1
2.0
3.4
3.2
1.4
2.3
.9
3.3
–1.6
3.3
.1
1.9
–.2
1.8
1.0
–.9
–.7
.0
–.5
–.3
–1.8
.0
–3.2
–.1
–1.8
–2.0
–2.1
–1.8
–.2
.6
–1.1
–.4
.9
–.9
–.3
–.3
–1.0
–3.2
1.9
4.9
–1.1
–.2
4.6
3.2
1.8
1.6
.6
.4

Total 1

At home

Away
from
home

3.5 ��������������� ���������������
31.3 ��������������� ���������������
11.3 ��������������� ���������������
–.8
–1.1 ���������������
–3.9
–3.7 ���������������
9.8
9.5 ���������������
7.1
7.6 ���������������
–1.0
–1.3 ���������������
–1.1
–1.6 ���������������
–1.8
–2.3
0.9
–.7
–1.0
1.4
2.9
2.7
2.7
2.8
3.0
3.9
2.4
1.9
2.1
–1.0
–1.3
3.3
3.1
3.2
2.4
–.7
–1.6
2.3
1.3
1.3
3.0
2.0
1.6
1.8
1.3
1.5
1.4
3.5
3.6
3.2
4.0
3.2
5.5
1.2
.3
4.6
4.4
4.0
5.6
7.0
7.1
7.4
2.3
1.3
6.1
4.3
4.3
4.4
4.6
5.1
4.2
20.3
22.0
12.7
12.0
12.4
11.3
6.6
6.2
7.4
.5
–.8
6.0
8.1
7.9
7.9
11.8
12.5
10.4
10.2
9.7
11.4
10.2
10.5
9.6
4.3
2.9
7.1
3.1
2.3
5.1
2.7
1.8
4.1
3.8
3.6
4.2
2.6
2.0
3.8
3.8
3.7
4.3
3.5
3.5
3.7
5.2
5.6
4.4
5.6
6.2
4.6
5.3
5.8
4.5
1.9
1.3
2.9
1.5
1.5
1.4
2.9
3.5
1.9
2.9
3.5
1.9
2.1
2.0
2.2
4.3
4.9
3.1
1.5
1.0
2.6
2.3
2.1
2.5
1.9
1.7
2.3
2.8
2.9
2.4
2.8
2.6
3.0
1.5
.8
2.3
3.6
4.5
2.3
2.7
2.4
3.0
2.3
1.7
3.2
2.1
1.4
3.2
4.9
5.6
4.0
5.9
6.6
5.0
–.5
–2.4
1.9
1.5
1.7
1.3
4.7
6.0
2.9
1.8
1.3
2.5
1.1
.4
2.1

Total 1
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
���������������
–0.9
4.7
1.3
–1.3
2.2
–.9
.0
1.8
1.7
1.7
1.7
2.9
4.8
3.1
2.6
17.0
21.6
11.4
7.1
7.2
7.9
37.5
18.0
11.9
1.3
–.5
.2
1.8
–19.7
8.2
.5
5.1
18.1
–7.4
2.0
–1.4
2.2
–1.3
8.6
–3.4
–8.8
13.4
14.2
–13.0
10.7
6.9
16.6
17.1
2.9
17.4
–21.3
18.2
7.7
6.6
.5
.5

Gasoline
–1.4
7.8
16.4
6.2
1.6
1.6
2.1
.5
10.1
–1.4
4.2
3.1
2.2
–3.8
7.0
1.2
–3.2
3.8
–2.4
.0
4.1
3.2
1.5
1.5
3.4
2.5
–.4
2.8
19.6
20.7
11.0
2.8
4.8
8.6
52.1
18.9
9.4
–6.7
–1.6
–2.5
3.0
–30.7
18.6
–1.8
6.5
36.8
–16.2
2.0
–5.9
6.4
–4.2
12.4
–6.1
–15.4
30.1
13.9
–24.9
24.8
6.8
26.1
16.1
6.4
29.6
–43.1
53.5
13.8
9.9
1.7
–1.0

C-CPI-U 5

����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
����������������
2.6
1.3
2.0
1.7
3.2
2.9
2.3
3.7
.2
2.5
1.3
2.9
1.5
1.3

1 Includes other items not shown separately.
2 Data beginning with 1983 incorporate a rental equivalence measure for homeowners’ costs.
3 Commodities and services.
4 Household energy--electricity, utility (piped) gas service, fuel oil, etc.--and motor fuel.
5 Chained consumer price index (C-CPI-U) introduced in 2002. Reflects the effect of substitution that consumers make across item categories in response to

changes in relative prices. Data for 2013 are subject to revision.
Note: Changes from December to December are based on unadjusted indexes.
Series reflect changes in composition and renaming beginning in 1998, and formula and methodology changes in 1999.
Source: Department of Labor (Bureau of Labor Statistics).

GDP, Income, Prices, and Selected Indicators  | 377

Labor Market Indicators
Table B–11. Civilian population and labor force, 1929–2014
[Monthly data seasonally adjusted, except as noted]

Year or month

Civilian
noninstitutional
population 1

Civilian labor force
Employment
Total

Total

NonAgricultural agricultural

Not in
labor
force

Civilian
labor force
participation rate 2

Civilian
employment/
population
ratio 3

1,550
4,340
8,020
12,060
12,830
11,340
10,610
9,030
7,700
10,390
9,480
8,120
5,560
2,660
1,070
670
1,040
2,270
2,356

������������������
������������������
������������������
������������������
������������������
������������������
������������������
������������������
������������������
������������������
������������������
44,200
43,990
42,230
39,100
38,590
40,230
45,550
45,850

�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
55.7
56.0
57.2
58.7
58.6
57.2
55.8
56.8

�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
47.6
50.4
54.5
57.6
57.9
56.1
53.6
54.5

3.2
8.7
15.9
23.6
24.9
21.7
20.1
16.9
14.3
19.0
17.2
14.6
9.9
4.7
1.9
1.2
1.9
3.9
3.9

2,311
2,276
3,637
3,288
2,055
1,883
1,834
3,532
2,852
2,750
2,859
4,602
3,740
3,852
4,714
3,911
4,070
3,786
3,366
2,875
2,975
2,817
2,832
4,093
5,016
4,882
4,365
5,156
7,929
7,406
6,991
6,202
6,137
7,637
8,273
10,678
10,717
8,539
8,312
8,237
7,425
6,701
6,528

42,477
42,447
42,708
42,787
42,604
43,093
44,041
44,678
44,660
44,402
45,336
46,088
46,960
47,617
48,312
49,539
50,583
51,394
52,058
52,288
52,527
53,291
53,602
54,315
55,834
57,091
57,667
58,171
59,377
59,991
60,025
59,659
59,900
60,806
61,460
62,067
62,665
62,839
62,744
62,752
62,888
62,944
62,523

58.3
58.8
58.9
59.2
59.2
59.0
58.9
58.8
59.3
60.0
59.6
59.5
59.3
59.4
59.3
58.8
58.7
58.7
58.9
59.2
59.6
59.6
60.1
60.4
60.2
60.4
60.8
61.3
61.2
61.6
62.3
63.2
63.7
63.8
63.9
64.0
64.0
64.4
64.8
65.3
65.6
65.9
66.5

56.0
56.6
55.4
56.1
57.3
57.3
57.1
55.5
56.7
57.5
57.1
55.4
56.0
56.1
55.4
55.5
55.4
55.7
56.2
56.9
57.3
57.5
58.0
57.4
56.6
57.0
57.8
57.8
56.1
56.8
57.9
59.3
59.9
59.2
59.0
57.8
57.9
59.5
60.1
60.7
61.5
62.3
63.0

3.9
3.8
5.9
5.3
3.3
3.0
2.9
5.5
4.4
4.1
4.3
6.8
5.5
5.5
6.7
5.5
5.7
5.2
4.5
3.8
3.8
3.6
3.5
4.9
5.9
5.6
4.9
5.6
8.5
7.7
7.1
6.1
5.8
7.1
7.6
9.7
9.6
7.5
7.2
7.0
6.2
5.5
5.3

Unemployment

Thousands of persons 14 years of age and over
1929 ����������������������
1930 ����������������������
1931 ����������������������
1932 ����������������������
1933 ����������������������
1934 ����������������������
1935 ����������������������
1936 ����������������������
1937 ����������������������
1938 ����������������������
1939 ����������������������
1940 ����������������������
1941 ����������������������
1942 ����������������������
1943 ����������������������
1944 ����������������������
1945 ����������������������
1946 ����������������������
1947 ����������������������

�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
�������������������
99,840
99,900
98,640
94,640
93,220
94,090
103,070
106,018

49,180
49,820
50,420
51,000
51,590
52,230
52,870
53,440
54,000
54,610
55,230
55,640
55,910
56,410
55,540
54,630
53,860
57,520
60,168

1947 ����������������������
1948 ����������������������
1949 ����������������������
1950 ����������������������
1951 ����������������������
1952 ����������������������
1953 ����������������������
1954 ����������������������
1955 ����������������������
1956 ����������������������
1957 ����������������������
1958 ����������������������
1959 ����������������������
1960 ����������������������
1961 ����������������������
1962 ����������������������
1963 ����������������������
1964 ����������������������
1965 ����������������������
1966 ����������������������
1967 ����������������������
1968 ����������������������
1969 ����������������������
1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������

101,827
103,068
103,994
104,995
104,621
105,231
107,056
108,321
109,683
110,954
112,265
113,727
115,329
117,245
118,771
120,153
122,416
124,485
126,513
128,058
129,874
132,028
134,335
137,085
140,216
144,126
147,096
150,120
153,153
156,150
159,033
161,910
164,863
167,745
170,130
172,271
174,215
176,383
178,206
180,587
182,753
184,613
186,393

59,350
60,621
61,286
62,208
62,017
62,138
63,015
63,643
65,023
66,552
66,929
67,639
68,369
69,628
70,459
70,614
71,833
73,091
74,455
75,770
77,347
78,737
80,734
82,771
84,382
87,034
89,429
91,949
93,775
96,158
99,009
102,251
104,962
106,940
108,670
110,204
111,550
113,544
115,461
117,834
119,865
121,669
123,869

47,630
45,480
42,400
38,940
38,760
40,890
42,260
44,410
46,300
44,220
45,750
47,520
50,350
53,750
54,470
53,960
52,820
55,250
57,812

10,450
10,340
10,290
10,170
10,090
9,900
10,110
10,000
9,820
9,690
9,610
9,540
9,100
9,250
9,080
8,950
8,580
8,320
8,256

37,180
35,140
32,110
28,770
28,670
30,990
32,150
34,410
36,480
34,530
36,140
37,980
41,250
44,500
45,390
45,010
44,240
46,930
49,557

Unemployment
rate,
civilian
workers 4

Percent

Thousands of persons 16 years of age and over
57,038
58,343
57,651
58,918
59,961
60,250
61,179
60,109
62,170
63,799
64,071
63,036
64,630
65,778
65,746
66,702
67,762
69,305
71,088
72,895
74,372
75,920
77,902
78,678
79,367
82,153
85,064
86,794
85,846
88,752
92,017
96,048
98,824
99,303
100,397
99,526
100,834
105,005
107,150
109,597
112,440
114,968
117,342

7,890
7,629
7,658
7,160
6,726
6,500
6,260
6,205
6,450
6,283
5,947
5,586
5,565
5,458
5,200
4,944
4,687
4,523
4,361
3,979
3,844
3,817
3,606
3,463
3,394
3,484
3,470
3,515
3,408
3,331
3,283
3,387
3,347
3,364
3,368
3,401
3,383
3,321
3,179
3,163
3,208
3,169
3,199

1 Not seasonally adjusted.
2 Civilian labor force as percent of civilian noninstitutional population.
3 Civilian employment as percent of civilian noninstitutional population.
4 Unemployed as percent of civilian labor force.

See next page for continuation of table.

378  |  Appendix B

49,148
50,714
49,993
51,758
53,235
53,749
54,919
53,904
55,722
57,514
58,123
57,450
59,065
60,318
60,546
61,759
63,076
64,782
66,726
68,915
70,527
72,103
74,296
75,215
75,972
78,669
81,594
83,279
82,438
85,421
88,734
92,661
95,477
95,938
97,030
96,125
97,450
101,685
103,971
106,434
109,232
111,800
114,142

Table B–11. Civilian population and labor force, 1929–2014—Continued
[Monthly data seasonally adjusted, except as noted]

Year or month

Civilian
noninstitutional
population 1

Civilian labor force
Employment
Total

Total

NonAgricultural agricultural

Unemployment

Not in
labor
force

Civilian
labor force
participation rate 2

Thousands of persons 16 years of age and over
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 5 ��������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 ����������������������
2011: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2012: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2013: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2014: Jan �������������


189,164
190,925
192,805
194,838
196,814
198,584
200,591
203,133
205,220
207,753
212,577
215,092
217,570
221,168
223,357
226,082
228,815
231,867
233,788
235,801
237,829
239,618
243,284
245,679
238,704
238,851
239,000
239,146
239,313
239,489
239,671
239,871
240,071
240,269
240,441
240,584
242,269
242,435
242,604
242,784
242,966
243,155
243,354
243,566
243,772
243,983
244,174
244,350
244,663
244,828
244,995
245,175
245,363
245,552
245,756
245,959
246,168
246,381
246,567
246,745
246,915

125,840
126,346
128,105
129,200
131,056
132,304
133,943
136,297
137,673
139,368
142,583
143,734
144,863
146,510
147,401
149,320
151,428
153,124
154,287
154,142
153,889
153,617
154,975
155,389
153,198
153,280
153,403
153,566
153,526
153,379
153,309
153,724
154,059
153,940
154,072
153,927
154,328
154,826
154,811
154,565
154,946
155,134
154,970
154,669
155,018
155,507
155,279
155,485
155,699
155,511
155,099
155,359
155,609
155,822
155,693
155,435
155,473
154,625
155,284
154,937
155,460

118,793
117,718
118,492
120,259
123,060
124,900
126,708
129,558
131,463
133,488
136,891
136,933
136,485
137,736
139,252
141,730
144,427
146,047
145,362
139,878
139,064
139,869
142,469
143,929
139,287
139,422
139,655
139,622
139,653
139,409
139,524
139,904
140,154
140,335
140,747
140,836
141,677
141,943
142,079
141,963
142,257
142,432
142,272
142,204
142,947
143,369
143,233
143,212
143,384
143,464
143,393
143,676
143,919
144,075
144,285
144,179
144,270
143,485
144,443
144,586
145,224

3,223
3,269
3,247
3,115
3,409
3,440
3,443
3,399
3,378
3,281
2,464
2,299
2,311
2,275
2,232
2,197
2,206
2,095
2,168
2,103
2,206
2,254
2,186
2,130
2,270
2,266
2,260
2,143
2,230
2,253
2,225
2,344
2,232
2,211
2,251
2,362
2,211
2,193
2,246
2,203
2,278
2,221
2,212
2,106
2,178
2,176
2,126
2,066
2,057
2,070
2,020
2,048
2,081
2,091
2,171
2,205
2,208
2,208
2,139
2,229
2,183

115,570
114,449
115,245
117,144
119,651
121,460
123,264
126,159
128,085
130,207
134,427
134,635
134,174
135,461
137,020
139,532
142,221
143,952
143,194
137,774
136,858
137,615
140,283
141,799
137,036
137,182
137,471
137,438
137,395
137,136
137,215
137,470
137,904
138,283
138,500
138,454
139,437
139,782
139,888
139,712
139,980
140,246
140,020
140,017
140,773
141,379
141,110
141,121
141,234
141,393
141,350
141,604
141,860
142,021
142,081
141,918
142,058
141,449
142,317
142,337
142,970

Civilian
employment/
population
ratio 3

Unemployment
rate,
civilian
workers 4

Percent
7,047
8,628
9,613
8,940
7,996
7,404
7,236
6,739
6,210
5,880
5,692
6,801
8,378
8,774
8,149
7,591
7,001
7,078
8,924
14,265
14,825
13,747
12,506
11,460
13,910
13,858
13,748
13,944
13,873
13,971
13,785
13,820
13,905
13,604
13,326
13,090
12,650
12,883
12,732
12,603
12,689
12,702
12,698
12,464
12,070
12,138
12,045
12,273
12,315
12,047
11,706
11,683
11,690
11,747
11,408
11,256
11,203
11,140
10,841
10,351
10,236

63,324
64,578
64,700
65,638
65,758
66,280
66,647
66,837
67,547
68,385
69,994
71,359
72,707
74,658
75,956
76,762
77,387
78,743
79,501
81,659
83,941
86,001
88,310
90,290
85,506
85,571
85,597
85,580
85,787
86,110
86,362
86,147
86,012
86,330
86,368
86,658
87,942
87,610
87,793
88,218
88,019
88,022
88,384
88,897
88,754
88,476
88,895
88,865
88,963
89,317
89,896
89,815
89,754
89,730
90,062
90,524
90,695
91,756
91,283
91,808
91,455

66.5
66.2
66.4
66.3
66.6
66.6
66.8
67.1
67.1
67.1
67.1
66.8
66.6
66.2
66.0
66.0
66.2
66.0
66.0
65.4
64.7
64.1
63.7
63.2
64.2
64.2
64.2
64.2
64.2
64.0
64.0
64.1
64.2
64.1
64.1
64.0
63.7
63.9
63.8
63.7
63.8
63.8
63.7
63.5
63.6
63.7
63.6
63.6
63.6
63.5
63.3
63.4
63.4
63.5
63.4
63.2
63.2
62.8
63.0
62.8
63.0

62.8
61.7
61.5
61.7
62.5
62.9
63.2
63.8
64.1
64.3
64.4
63.7
62.7
62.3
62.3
62.7
63.1
63.0
62.2
59.3
58.5
58.4
58.6
58.6
58.4
58.4
58.4
58.4
58.4
58.2
58.2
58.3
58.4
58.4
58.5
58.5
58.5
58.5
58.6
58.5
58.6
58.6
58.5
58.4
58.6
58.8
58.7
58.6
58.6
58.6
58.5
58.6
58.7
58.7
58.7
58.6
58.6
58.2
58.6
58.6
58.8

5.6
6.8
7.5
6.9
6.1
5.6
5.4
4.9
4.5
4.2
4.0
4.7
5.8
6.0
5.5
5.1
4.6
4.6
5.8
9.3
9.6
9.0
8.1
7.4
9.1
9.0
9.0
9.1
9.0
9.1
9.0
9.0
9.0
8.8
8.6
8.5
8.2
8.3
8.2
8.2
8.2
8.2
8.2
8.1
7.8
7.8
7.8
7.9
7.9
7.7
7.5
7.5
7.5
7.5
7.3
7.2
7.2
7.2
7.0
6.7
6.6

5 Beginning in 2000, data for agricultural employment are for agricultural and related industries; data for this series and for nonagricultural employment are
not strictly comparable with data for earlier years. Because of independent seasonal adjustment for these two series, monthly data will not add to total civilian
employment.
Note: Labor force data in Tables B–11 through B–13 are based on household interviews and usually relate to the calendar week that includes the 12th of
the month. Historical comparability is affected by revisions to population controls, changes in occupational and industry classification, and other changes to the
survey. In recent years, updated population controls have been introduced annually with the release of January data, so data are not strictly comparable with
earlier periods. Particularly notable changes were introduced for data in the years 1953, 1960, 1962, 1972, 1973, 1978, 1980, 1990, 1994, 1997, 1998, 2000,
2003, 2008 and 2012. For definitions of terms, area samples used, historical comparability of the data, comparability with other series, etc., see Employment
and Earnings or concepts and methodology of the CPS at http://www.bls.gov/cps/documentation.htm#concepts.
Source: Department of Labor (Bureau of Labor Statistics).

Labor Market Indicators  | 379

Table B–12. Civilian unemployment rate, 1970–2014
[Percent 1; monthly data seasonally adjusted, except as noted]

Year or month

1970 �������������������
1971 �������������������
1972 �������������������
1973 �������������������
1974 �������������������
1975 �������������������
1976 �������������������
1977 �������������������
1978 �������������������
1979 �������������������
1980 �������������������
1981 �������������������
1982 �������������������
1983 �������������������
1984 �������������������
1985 �������������������
1986 �������������������
1987 �������������������
1988 �������������������
1989 �������������������
1990 �������������������
1991 �������������������
1992 �������������������
1993 �������������������
1994 �������������������
1995 �������������������
1996 �������������������
1997 �������������������
1998 �������������������
1999 �������������������
2000 �������������������
2001 �������������������
2002 �������������������
2003 �������������������
2004 �������������������
2005 �������������������
2006 �������������������
2007 �������������������
2008 �������������������
2009 �������������������
2010 �������������������
2011 �������������������
2012 �������������������
2013 �������������������
2012: Jan ����������

      Feb 
����������
      Mar 
���������
      Apr 
����������
      May ��������
�
      June ��������

      July 
���������
      Aug ���������

      Sept ��������

      Oct ����������

      Nov ���������

      Dec ���������
�
2013: Jan ����������

      Feb 
����������
      Mar 
���������
      Apr 
����������
      May ��������
�
      June ��������

      July 
���������
      Aug ���������

      Sept ��������

      Oct ����������

      Nov ���������

      Dec ���������
�
2014: Jan ����������


All
civilian
workers
4.9
5.9
5.6
4.9
5.6
8.5
7.7
7.1
6.1
5.8
7.1
7.6
9.7
9.6
7.5
7.2
7.0
6.2
5.5
5.3
5.6
6.8
7.5
6.9
6.1
5.6
5.4
4.9
4.5
4.2
4.0
4.7
5.8
6.0
5.5
5.1
4.6
4.6
5.8
9.3
9.6
8.9
8.1
7.4
8.2
8.3
8.2
8.2
8.2
8.2
8.2
8.1
7.8
7.8
7.8
7.9
7.9
7.7
7.5
7.5
7.5
7.5
7.3
7.2
7.2
7.2
7.0
6.7
6.6

Males
Total
4.4
5.3
5.0
4.2
4.9
7.9
7.1
6.3
5.3
5.1
6.9
7.4
9.9
9.9
7.4
7.0
6.9
6.2
5.5
5.2
5.7
7.2
7.9
7.2
6.2
5.6
5.4
4.9
4.4
4.1
3.9
4.8
5.9
6.3
5.6
5.1
4.6
4.7
6.1
10.3
10.5
9.4
8.2
7.6
8.2
8.5
8.4
8.3
8.4
8.4
8.3
8.3
8.0
7.9
7.9
7.9
8.0
7.9
7.6
7.8
7.9
7.8
7.7
7.7
7.7
7.5
7.3
6.8
6.8

Females

20
16–19 years
years and
over
15.0
16.6
15.9
13.9
15.6
20.1
19.2
17.3
15.8
15.9
18.3
20.1
24.4
23.3
19.6
19.5
19.0
17.8
16.0
15.9
16.3
19.8
21.5
20.4
19.0
18.4
18.1
16.9
16.2
14.7
14.0
16.0
18.1
19.3
18.4
18.6
16.9
17.6
21.2
27.8
28.8
27.2
26.8
25.5
25.7
27.0
26.9
26.9
26.7
26.4
26.1
28.5
27.0
26.8
26.8
26.7
26.5
27.2
25.8
25.9
26.8
27.7
26.9
25.0
24.1
24.4
23.3
21.1
22.6

3.5
4.4
4.0
3.3
3.8
6.8
5.9
5.2
4.3
4.2
5.9
6.3
8.8
8.9
6.6
6.2
6.1
5.4
4.8
4.5
5.0
6.4
7.1
6.4
5.4
4.8
4.6
4.2
3.7
3.5
3.3
4.2
5.3
5.6
5.0
4.4
4.0
4.1
5.4
9.6
9.8
8.7
7.5
7.0
7.6
7.8
7.7
7.6
7.7
7.7
7.7
7.5
7.3
7.2
7.2
7.2
7.4
7.2
6.9
7.1
7.2
7.0
7.0
7.0
7.0
6.9
6.7
6.3
6.2

Total
5.9
6.9
6.6
6.0
6.7
9.3
8.6
8.2
7.2
6.8
7.4
7.9
9.4
9.2
7.6
7.4
7.1
6.2
5.6
5.4
5.5
6.4
7.0
6.6
6.0
5.6
5.4
5.0
4.6
4.3
4.1
4.7
5.6
5.7
5.4
5.1
4.6
4.5
5.4
8.1
8.6
8.5
7.9
7.1
8.2
8.1
8.0
8.0
7.9
7.9
8.0
7.8
7.6
7.7
7.6
7.9
7.7
7.6
7.5
7.2
7.1
7.3
6.9
6.8
6.7
6.9
6.7
6.5
6.4

20
16–19 years
years and
over
15.6
17.2
16.7
15.3
16.6
19.7
18.7
18.3
17.1
16.4
17.2
19.0
21.9
21.3
18.0
17.6
17.6
15.9
14.4
14.0
14.7
17.5
18.6
17.5
16.2
16.1
15.2
15.0
12.9
13.2
12.1
13.4
14.9
15.6
15.5
14.5
13.8
13.8
16.2
20.7
22.8
21.7
21.1
20.3
21.3
20.6
22.7
22.1
21.6
21.0
21.2
20.2
20.5
20.7
21.0
21.3
20.6
23.2
22.1
21.6
21.3
19.7
19.8
20.1
18.1
19.6
18.3
19.3
18.7

4.8
5.7
5.4
4.9
5.5
8.0
7.4
7.0
6.0
5.7
6.4
6.8
8.3
8.1
6.8
6.6
6.2
5.4
4.9
4.7
4.9
5.7
6.3
5.9
5.4
4.9
4.8
4.4
4.1
3.8
3.6
4.1
5.1
5.1
4.9
4.6
4.1
4.0
4.9
7.5
8.0
7.9
7.3
6.5
7.6
7.6
7.4
7.4
7.4
7.4
7.5
7.3
7.0
7.1
7.0
7.3
7.2
7.0
6.9
6.6
6.5
6.8
6.4
6.2
6.2
6.4
6.2
6.0
5.9

By race
Hispanic Married Women
Both
or
who
sexes
men,
Latino spouse maintain
Black Black or Asian
16–19
White 2 and African (NSA) 2, 3 ethnic- present families
years
ity 4
(NSA) 3
other 2 American 2
15.3
16.9
16.2
14.5
16.0
19.9
19.0
17.8
16.4
16.1
17.8
19.6
23.2
22.4
18.9
18.6
18.3
16.9
15.3
15.0
15.5
18.7
20.1
19.0
17.6
17.3
16.7
16.0
14.6
13.9
13.1
14.7
16.5
17.5
17.0
16.6
15.4
15.7
18.7
24.3
25.9
24.4
24.0
22.9
23.5
23.8
24.8
24.6
24.2
23.7
23.7
24.4
23.8
23.8
23.9
24.0
23.5
25.2
23.9
23.7
24.1
23.8
23.4
22.6
21.3
22.0
20.8
20.2
20.7

4.5
5.4
5.1
4.3
5.0
7.8
7.0
6.2
5.2
5.1
6.3
6.7
8.6
8.4
6.5
6.2
6.0
5.3
4.7
4.5
4.8
6.1
6.6
6.1
5.3
4.9
4.7
4.2
3.9
3.7
3.5
4.2
5.1
5.2
4.8
4.4
4.0
4.1
5.2
8.5
8.7
7.9
7.2
6.5
7.4
7.4
7.3
7.4
7.4
7.3
7.4
7.2
7.0
6.9
6.8
6.9
7.1
6.8
6.7
6.6
6.6
6.6
6.6
6.4
6.3
6.3
6.1
5.9
5.7

8.2 ����������� ��������������� ������������
9.9 ����������� ��������������� ������������
10.0
10.4 ��������������� ������������
9.0
9.4 ���������������
7.5
9.9
10.5 ���������������
8.1
13.8
14.8 ���������������
12.2
13.1
14.0 ���������������
11.5
13.1
14.0 ���������������
10.1
11.9
12.8 ���������������
9.1
11.3
12.3 ���������������
8.3
13.1
14.3 ���������������
10.1
14.2
15.6 ���������������
10.4
17.3
18.9 ���������������
13.8
17.8
19.5 ���������������
13.7
14.4
15.9 ���������������
10.7
13.7
15.1 ���������������
10.5
13.1
14.5 ���������������
10.6
11.6
13.0 ���������������
8.8
10.4
11.7 ���������������
8.2
10.0
11.4 ���������������
8.0
10.1
11.4 ���������������
8.2
11.1
12.5 ���������������
10.0
12.7
14.2 ���������������
11.6
11.7
13.0 ���������������
10.8
10.5
11.5 ���������������
9.9
9.6
10.4 ���������������
9.3
9.3
10.5 ���������������
8.9
8.8
10.0 ���������������
7.7
7.8
8.9 ���������������
7.2
7.0
8.0 ���������������
6.4
�����������
7.6
3.6
5.7
�����������
8.6
4.5
6.6
�����������
10.2
5.9
7.5
�����������
10.8
6.0
7.7
�����������
10.4
4.4
7.0
�����������
10.0
4.0
6.0
�����������
8.9
3.0
5.2
�����������
8.3
3.2
5.6
�����������
10.1
4.0
7.6
�����������
14.8
7.3
12.1
�����������
16.0
7.5
12.5
�����������
15.8
7.0
11.5
�����������
13.8
5.9
10.3
�����������
13.1
5.2
9.1
�����������
13.6
6.7
10.5
�����������
14.0
6.3
10.6
�����������
14.1
6.2
10.4
�����������
13.2
5.2
10.4
�����������
13.6
5.2
11.1
�����������
14.1
6.3
11.0
�����������
14.2
6.2
10.2
�����������
13.9
5.9
10.1
�����������
13.5
4.8
9.9
�����������
14.2
4.9
10.0
�����������
13.3
6.4
9.9
�����������
14.0
6.6
9.5
�����������
13.8
6.5
9.7
�����������
13.8
6.1
9.5
�����������
13.2
5.0
9.2
�����������
13.1
5.1
9.0
�����������
13.5
4.3
9.1
�����������
13.5
5.0
9.1
�����������
12.6
5.7
9.5
�����������
12.9
5.1
9.3
�����������
13.0
5.3
8.9
�����������
13.0
5.2
9.0
�����������
12.4
5.3
8.7
�����������
11.9
4.1
8.3
�����������
12.1
4.8
8.4

2.6
3.2
2.8
2.3
2.7
5.1
4.2
3.6
2.8
2.8
4.2
4.3
6.5
6.5
4.6
4.3
4.4
3.9
3.3
3.0
3.4
4.4
5.1
4.4
3.7
3.3
3.0
2.7
2.4
2.2
2.0
2.7
3.6
3.8
3.1
2.8
2.4
2.5
3.4
6.6
6.8
5.8
4.9
4.3
5.1
5.1
5.2
5.1
5.2
5.0
5.0
4.9
4.6
4.6
4.6
4.6
4.6
4.5
4.3
4.4
4.4
4.4
4.3
4.3
4.3
4.5
4.2
3.8
3.8

5.4
7.3
7.2
7.1
7.0
10.0
10.1
9.4
8.5
8.3
9.2
10.4
11.7
12.2
10.3
10.4
9.8
9.2
8.1
8.1
8.3
9.3
10.0
9.7
8.9
8.0
8.2
8.1
7.2
6.4
5.9
6.6
8.0
8.5
8.0
7.8
7.1
6.5
8.0
11.5
12.3
12.4
11.4
10.2
12.0
11.7
10.8
10.2
10.9
11.8
11.7
12.3
11.3
11.5
10.7
11.3
11.3
11.0
10.7
10.3
9.9
10.7
10.5
11.0
8.8
9.5
9.7
8.7
9.1

1 Unemployed as percent of civilian labor force in group specified.
2 Beginning in 2003, persons who selected this race group only. Prior to 2003, persons who selected more than one race were included in the group they
identified as the main race. Data for “black or African American” were for “black” prior to 2003. Data discontinued for “black and other” series. See Employment
and Earnings or concepts and methodology of the CPS at http://www.bls.gov/cps/documentation.htm#concepts for details.
3 Not seasonally adjusted (NSA).
4 Persons whose ethnicity is identified as Hispanic or Latino may be of any race.
Note: Data relate to persons 16 years of age and over.
See Note, Table B–11.
Source: Department of Labor (Bureau of Labor Statistics).

380  |  Appendix B

Table B–13. Unemployment by duration and reason, 1970–2014
[Thousands of persons, except as noted; monthly data seasonally adjusted 1]
Duration of unemployment
Year or month

1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 ����������������������
2012: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2013: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2014: Jan �������������


Unemployment
4,093
5,016
4,882
4,365
5,156
7,929
7,406
6,991
6,202
6,137
7,637
8,273
10,678
10,717
8,539
8,312
8,237
7,425
6,701
6,528
7,047
8,628
9,613
8,940
7,996
7,404
7,236
6,739
6,210
5,880
5,692
6,801
8,378
8,774
8,149
7,591
7,001
7,078
8,924
14,265
14,825
13,747
12,506
11,460
12,650
12,883
12,732
12,603
12,689
12,702
12,698
12,464
12,070
12,138
12,045
12,273
12,315
12,047
11,706
11,683
11,690
11,747
11,408
11,256
11,203
11,140
10,841
10,351
10,236

Less
than 5
weeks
2,139
2,245
2,242
2,224
2,604
2,940
2,844
2,919
2,865
2,950
3,295
3,449
3,883
3,570
3,350
3,498
3,448
3,246
3,084
3,174
3,265
3,480
3,376
3,262
2,728
2,700
2,633
2,538
2,622
2,568
2,558
2,853
2,893
2,785
2,696
2,667
2,614
2,542
2,932
3,165
2,771
2,677
2,644
2,584
2,461
2,584
2,724
2,621
2,575
2,741
2,708
2,832
2,517
2,619
2,636
2,688
2,753
2,677
2,497
2,491
2,704
2,665
2,548
2,527
2,571
2,794
2,439
2,255
2,434

5–14
weeks
1,290
1,585
1,472
1,314
1,597
2,484
2,196
2,132
1,923
1,946
2,470
2,539
3,311
2,937
2,451
2,509
2,557
2,196
2,007
1,978
2,257
2,791
2,830
2,584
2,408
2,342
2,287
2,138
1,950
1,832
1,815
2,196
2,580
2,612
2,382
2,304
2,121
2,232
2,804
3,828
3,267
2,993
2,866
2,759
2,880
2,842
2,792
2,839
3,018
2,804
3,037
2,834
2,825
2,850
2,777
2,876
3,077
2,788
2,843
2,844
2,642
2,848
2,826
2,738
2,685
2,636
2,585
2,506
2,429

15–26
weeks
428
668
601
483
574
1,303
1,018
913
766
706
1,052
1,122
1,708
1,652
1,104
1,025
1,045
943
801
730
822
1,246
1,453
1,297
1,237
1,085
1,053
995
763
755
669
951
1,369
1,442
1,293
1,130
1,031
1,061
1,427
2,775
2,371
2,061
1,859
1,807
1,942
2,021
1,924
1,951
1,677
1,839
1,780
1,845
1,853
1,774
1,796
1,862
1,867
1,735
1,779
1,969
1,934
1,892
1,786
1,704
1,802
1,777
1,742
1,651
1,689

27
weeks
and
over
235
519
566
343
381
1,203
1,348
1,028
648
535
820
1,162
1,776
2,559
1,634
1,280
1,187
1,040
809
646
703
1,111
1,954
1,798
1,623
1,278
1,262
1,067
875
725
649
801
1,535
1,936
1,779
1,490
1,235
1,243
1,761
4,496
6,415
6,016
5,136
4,310
5,524
5,352
5,292
5,106
5,392
5,331
5,166
5,003
4,875
5,021
4,767
4,772
4,707
4,750
4,576
4,360
4,353
4,325
4,246
4,269
4,125
4,047
4,044
3,878
3,646

Reason for unemployment

Average Median
(mean)
duration duration
(weeks) 2 (weeks)
8.6
11.3
12.0
10.0
9.8
14.2
15.8
14.3
11.9
10.8
11.9
13.7
15.6
20.0
18.2
15.6
15.0
14.5
13.5
11.9
12.0
13.7
17.7
18.0
18.8
16.6
16.7
15.8
14.5
13.4
12.6
13.1
16.6
19.2
19.6
18.4
16.8
16.8
17.9
24.4
33.0
39.3
39.4
36.5
40.1
40.0
39.4
39.3
39.6
40.0
38.8
39.1
39.4
40.3
39.2
38.0
35.4
36.9
37.0
36.6
36.9
35.7
36.7
37.0
36.8
36.0
37.1
37.1
35.4

4.9
6.3
6.2
5.2
5.2
8.4
8.2
7.0
5.9
5.4
6.5
6.9
8.7
10.1
7.9
6.8
6.9
6.5
5.9
4.8
5.3
6.8
8.7
8.3
9.2
8.3
8.3
8.0
6.7
6.4
5.9
6.8
9.1
10.1
9.8
8.9
8.3
8.5
9.4
15.1
21.4
21.4
19.3
17.0
20.9
20.0
19.6
19.2
19.8
19.8
17.2
18.2
18.7
20.0
18.6
17.8
16.0
17.7
18.1
17.3
16.9
16.2
15.8
16.5
16.4
16.5
17.0
17.1
16.0

Job losers 3
Total
1,811
2,323
2,108
1,694
2,242
4,386
3,679
3,166
2,585
2,635
3,947
4,267
6,268
6,258
4,421
4,139
4,033
3,566
3,092
2,983
3,387
4,694
5,389
4,848
3,815
3,476
3,370
3,037
2,822
2,622
2,517
3,476
4,607
4,838
4,197
3,667
3,321
3,515
4,789
9,160
9,250
8,106
6,877
6,073
7,270
7,167
7,051
6,859
6,980
7,106
7,121
6,885
6,508
6,511
6,434
6,475
6,675
6,495
6,321
6,367
6,094
6,089
5,894
5,887
5,803
6,162
5,731
5,366
5,407

On
layoff
675
735
582
472
746
1,671
1,050
865
712
851
1,488
1,430
2,127
1,780
1,171
1,157
1,090
943
851
850
1,028
1,292
1,260
1,115
977
1,030
1,021
931
866
848
852
1,067
1,124
1,121
998
933
921
976
1,176
1,630
1,431
1,230
1,183
1,136
1,253
1,160
1,148
1,099
1,143
1,264
1,383
1,231
1,170
1,058
1,082
1,110
1,164
1,091
1,118
1,179
980
1,195
1,197
1,059
1,091
1,507
1,128
997
986

Other
1,137
1,588
1,526
1,221
1,495
2,714
2,628
2,300
1,873
1,784
2,459
2,837
4,141
4,478
3,250
2,982
2,943
2,623
2,241
2,133
2,359
3,402
4,129
3,733
2,838
2,446
2,349
2,106
1,957
1,774
1,664
2,409
3,483
3,717
3,199
2,734
2,400
2,539
3,614
7,530
7,819
6,876
5,694
4,937
6,017
6,007
5,903
5,760
5,838
5,842
5,738
5,654
5,338
5,452
5,351
5,365
5,511
5,404
5,204
5,188
5,114
4,894
4,697
4,828
4,712
4,655
4,603
4,369
4,421

Job
ReNew
leavers entrants entrants
550
590
641
683
768
827
903
909
874
880
891
923
840
830
823
877
1,015
965
983
1,024
1,041
1,004
1,002
976
791
824
774
795
734
783
780
835
866
818
858
872
827
793
896
882
889
956
967
932
928
1,035
1,101
987
906
929
873
953
956
1,018
929
1,000
984
952
978
857
944
1,034
970
890
984
842
890
862
818

1,228
1,472
1,456
1,340
1,463
1,892
1,928
1,963
1,857
1,806
1,927
2,102
2,384
2,412
2,184
2,256
2,160
1,974
1,809
1,843
1,930
2,139
2,285
2,198
2,786
2,525
2,512
2,338
2,132
2,005
1,961
2,031
2,368
2,477
2,408
2,386
2,237
2,142
2,472
3,187
3,466
3,401
3,345
3,207
3,303
3,360
3,300
3,360
3,395
3,193
3,365
3,336
3,303
3,321
3,336
3,615
3,520
3,330
3,182
3,131
3,326
3,240
3,234
3,116
3,165
3,104
3,065
3,036
2,937

504
630
677
649
681
823
895
953
885
817
872
981
1,185
1,216
1,110
1,039
1,029
920
816
677
688
792
937
919
604
579
580
569
520
469
434
459
536
641
686
666
616
627
766
1,035
1,220
1,284
1,316
1,247
1,252
1,383
1,392
1,357
1,347
1,318
1,298
1,264
1,268
1,306
1,349
1,296
1,274
1,276
1,304
1,268
1,257
1,250
1,246
1,295
1,211
1,217
1,169
1,201
1,184

1 Because of independent seasonal adjustment of the various series, detail will not sum to totals.
2 Beginning with January 2011, includes unemployment durations of up to 5 years; prior data are for up to 2 years.
3 Beginning with January 1994, job losers and persons who completed temporary jobs.

Note: Data relate to persons 16 years of age and over.
See Note, Table B–11.
Source: Department of Labor (Bureau of Labor Statistics).

Labor Market Indicators  | 381

Table B–14. Employees on nonagricultural payrolls, by major industry, 1970–2014
[Thousands of jobs; monthly data seasonally adjusted]
Private industries

Year or month

1970 ����������������������
1971 ����������������������
1972 ����������������������
1973 ����������������������
1974 ����������������������
1975 ����������������������
1976 ����������������������
1977 ����������������������
1978 ����������������������
1979 ����������������������
1980 ����������������������
1981 ����������������������
1982 ����������������������
1983 ����������������������
1984 ����������������������
1985 ����������������������
1986 ����������������������
1987 ����������������������
1988 ����������������������
1989 ����������������������
1990 ����������������������
1991 ����������������������
1992 ����������������������
1993 ����������������������
1994 ����������������������
1995 ����������������������
1996 ����������������������
1997 ����������������������
1998 ����������������������
1999 ����������������������
2000 ����������������������
2001 ����������������������
2002 ����������������������
2003 ����������������������
2004 ����������������������
2005 ����������������������
2006 ����������������������
2007 ����������������������
2008 ����������������������
2009 ����������������������
2010 ����������������������
2011 ����������������������
2012 ����������������������
2013 p ��������������������
2012: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec ������������
�
2013: Jan �������������

      Feb 
�������������
      Mar 
������������
      Apr 
�������������
      May �����������
�
      June �����������

      July 
������������
      Aug ������������

      Sept �����������

      Oct �������������

      Nov ������������

      Dec p ����������
2014: Jan p �����������

Total
nonagricultural
employment

71,006
71,335
73,798
76,912
78,389
77,069
79,502
82,593
86,826
89,933
90,533
91,297
89,689
90,295
94,548
97,532
99,500
102,116
105,378
108,051
109,527
108,427
108,802
110,935
114,398
117,407
119,836
122,951
126,157
129,240
132,019
132,074
130,628
130,318
131,749
134,005
136,398
137,936
137,170
131,233
130,275
131,842
134,104
136,368
133,188
133,414
133,657
133,753
133,863
133,951
134,111
134,261
134,422
134,647
134,850
135,064
135,261
135,541
135,682
135,885
136,084
136,285
136,434
136,636
136,800
137,037
137,311
137,386
137,499

Goods-producing industries
Total
private

58,318
58,323
60,333
63,050
64,086
62,250
64,501
67,334
71,014
73,865
74,158
75,117
73,706
74,284
78,389
81,000
82,661
84,960
87,838
90,124
91,112
89,881
90,015
91,946
95,124
97,975
100,297
103,287
106,248
108,933
111,230
110,956
109,115
108,735
110,128
112,201
114,424
115,718
114,661
108,678
107,785
109,756
112,184
114,503
111,246
111,474
111,720
111,822
111,953
112,028
112,200
112,336
112,495
112,750
112,961
113,176
113,395
113,658
113,822
114,010
114,232
114,433
114,603
114,783
114,936
115,183
115,455
115,544
115,686

Total

22,179
21,602
22,299
23,450
23,364
21,318
22,025
22,972
24,156
24,997
24,263
24,118
22,550
22,110
23,435
23,585
23,318
23,470
23,909
24,045
23,723
22,588
22,095
22,219
22,774
23,156
23,409
23,886
24,354
24,465
24,649
23,873
22,557
21,816
21,882
22,190
22,530
22,233
21,335
18,558
17,751
18,047
18,420
18,700
18,304
18,327
18,377
18,396
18,394
18,411
18,465
18,452
18,436
18,452
18,484
18,536
18,579
18,651
18,680
18,669
18,671
18,684
18,679
18,696
18,718
18,756
18,824
18,811
18,887

Mining
and
logging
677
658
672
693
755
802
832
865
902
1,008
1,077
1,180
1,163
997
1,014
974
829
771
770
750
765
739
689
666
659
641
637
654
645
598
599
606
583
572
591
628
684
724
767
694
705
788
848
868
840
846
849
850
853
852
851
849
846
839
846
851
854
858
860
857
861
864
867
870
876
881
882
883
890

Private service-providing industries
Trade, transportation,
and utilities 1

Manufacturing
Construction

3,654
3,770
3,957
4,167
4,095
3,608
3,662
3,940
4,322
4,562
4,454
4,304
4,024
4,065
4,501
4,793
4,937
5,090
5,233
5,309
5,263
4,780
4,608
4,779
5,095
5,274
5,536
5,813
6,149
6,545
6,787
6,826
6,716
6,735
6,976
7,336
7,691
7,630
7,162
6,016
5,518
5,533
5,646
5,827
5,627
5,622
5,627
5,630
5,613
5,620
5,635
5,647
5,648
5,666
5,687
5,720
5,743
5,789
5,813
5,811
5,816
5,829
5,830
5,836
5,849
5,864
5,896
5,874
5,922

Total
17,848
17,174
17,669
18,589
18,514
16,909
17,531
18,167
18,932
19,426
18,733
18,634
17,363
17,048
17,920
17,819
17,552
17,609
17,906
17,985
17,695
17,068
16,799
16,774
17,020
17,241
17,237
17,419
17,560
17,322
17,263
16,441
15,259
14,509
14,315
14,227
14,155
13,879
13,406
11,847
11,528
11,726
11,927
12,005
11,837
11,859
11,901
11,916
11,928
11,939
11,979
11,956
11,942
11,947
11,951
11,965
11,982
12,004
12,007
12,001
11,994
11,991
11,982
11,990
11,993
12,011
12,046
12,054
12,075

Durable
goods
10,762
10,229
10,630
11,414
11,432
10,266
10,640
11,132
11,770
12,220
11,679
11,611
10,610
10,326
11,050
11,034
10,795
10,767
10,969
11,004
10,737
10,220
9,946
9,901
10,132
10,373
10,486
10,705
10,911
10,831
10,877
10,336
9,485
8,964
8,925
8,956
8,981
8,808
8,463
7,284
7,064
7,273
7,470
7,543
7,395
7,419
7,447
7,460
7,470
7,480
7,518
7,492
7,477
7,481
7,493
7,505
7,514
7,527
7,533
7,533
7,531
7,532
7,526
7,540
7,549
7,562
7,581
7,583
7,598

Nondurable
goods
7,086
6,944
7,039
7,176
7,082
6,643
6,891
7,035
7,162
7,206
7,054
7,023
6,753
6,722
6,870
6,784
6,757
6,842
6,938
6,981
6,958
6,848
6,853
6,872
6,889
6,868
6,751
6,714
6,649
6,491
6,386
6,105
5,774
5,546
5,390
5,271
5,174
5,071
4,943
4,564
4,464
4,453
4,457
4,463
4,442
4,440
4,454
4,456
4,458
4,459
4,461
4,464
4,465
4,466
4,458
4,460
4,468
4,477
4,474
4,468
4,463
4,459
4,456
4,450
4,444
4,449
4,465
4,471
4,477

Total
Total
36,139
36,721
38,034
39,600
40,721
40,932
42,476
44,362
46,858
48,869
49,895
50,999
51,156
52,174
54,954
57,415
59,343
61,490
63,929
66,079
67,389
67,293
67,921
69,727
72,350
74,819
76,888
79,401
81,894
84,468
86,581
87,083
86,558
86,918
88,246
90,010
91,894
93,485
93,326
90,121
90,034
91,708
93,763
95,804
92,942
93,147
93,343
93,426
93,559
93,617
93,735
93,884
94,059
94,298
94,477
94,640
94,816
95,007
95,142
95,341
95,561
95,749
95,924
96,087
96,218
96,427
96,631
96,733
96,799

14,144
14,318
14,788
15,349
15,693
15,606
16,128
16,765
17,658
18,303
18,413
18,604
18,457
18,668
19,653
20,379
20,795
21,302
21,974
22,510
22,666
22,281
22,125
22,378
23,128
23,834
24,239
24,700
25,186
25,771
26,225
25,983
25,497
25,287
25,533
25,959
26,276
26,630
26,293
24,906
24,636
25,065
25,476
25,871
25,355
25,368
25,396
25,417
25,457
25,447
25,451
25,470
25,495
25,545
25,618
25,638
25,691
25,691
25,683
25,718
25,760
25,811
25,862
25,911
25,973
26,017
26,090
26,172
26,182

Retail
trade
7,463
7,657
8,038
8,371
8,536
8,600
8,966
9,359
9,879
10,180
10,244
10,364
10,372
10,635
11,223
11,733
12,078
12,419
12,808
13,108
13,182
12,896
12,828
13,021
13,491
13,897
14,143
14,389
14,609
14,970
15,280
15,239
15,025
14,917
15,058
15,280
15,353
15,520
15,283
14,522
14,440
14,668
14,841
15,077
14,818
14,803
14,808
14,833
14,827
14,814
14,802
14,802
14,834
14,861
14,915
14,917
14,944
14,953
14,944
14,967
15,002
15,040
15,089
15,118
15,146
15,187
15,210
15,272
15,260

1 Includes wholesale trade, transportation and warehousing, and utilities, not shown separately.

Note: Data in Tables B–14 and B–15 are based on reports from employing establishments and relate to full- and part-time wage and salary workers in
nonagricultural establishments who received pay for any part of the pay period that includes the 12th of the month. Not comparable with labor force data
(Tables B–11 through B–13), which include proprietors, self-employed persons, unpaid family workers, and private household workers; which count persons as
See next page for continuation of table.

382  |  Appendix B

Table B–14. Employees on nonagricultural payrolls, by major industry,
1970–2014—Continued
[Thousands of jobs; monthly data seasonally adjusted]
Private industries—Continued

Government

Private service-providing industries—Continued
Year or month
Information
1970 ���������������������������������
1971 ���������������������������������
1972 ���������������������������������
1973 ���������������������������������
1974 ���������������������������������
1975 ���������������������������������
1976 ���������������������������������
1977 ���������������������������������
1978 ���������������������������������
1979 ���������������������������������
1980 ���������������������������������
1981 ���������������������������������
1982 ���������������������������������
1983 ���������������������������������
1984 ���������������������������������
1985 ���������������������������������
1986 ���������������������������������
1987 ���������������������������������
1988 ���������������������������������
1989 ���������������������������������
1990 ���������������������������������
1991 ���������������������������������
1992 ���������������������������������
1993 ���������������������������������
1994 ���������������������������������
1995 ���������������������������������
1996 ���������������������������������
1997 ���������������������������������
1998 ���������������������������������
1999 ���������������������������������
2000 ���������������������������������
2001 ���������������������������������
2002 ���������������������������������
2003 ���������������������������������
2004 ���������������������������������
2005 ���������������������������������
2006 ���������������������������������
2007 ���������������������������������
2008 ���������������������������������
2009 ���������������������������������
2010 ���������������������������������
2011 ���������������������������������
2012 ���������������������������������
2013 p �������������������������������
2012: Jan ������������������������

      Feb 
������������������������
      Mar 
�����������������������
      Apr 
������������������������
      May ����������������������
�
      June ����������������������

      July 
�����������������������
      Aug �����������������������

      Sept ����������������������

      Oct ������������������������

      Nov �����������������������

      Dec �����������������������
�
2013: Jan ������������������������

      Feb 
������������������������
      Mar 
�����������������������
      Apr 
������������������������
      May ����������������������
�
      June ����������������������

      July 
�����������������������
      Aug �����������������������

      Sept ����������������������

      Oct ������������������������

      Nov �����������������������

      Dec p ���������������������
2014: Jan p ����������������������

2,041
2,009
2,056
2,135
2,160
2,061
2,111
2,185
2,287
2,375
2,361
2,382
2,317
2,253
2,398
2,437
2,445
2,507
2,585
2,622
2,688
2,677
2,641
2,668
2,738
2,843
2,940
3,084
3,218
3,419
3,630
3,629
3,395
3,188
3,118
3,061
3,038
3,032
2,984
2,804
2,707
2,674
2,676
2,685
2,673
2,673
2,675
2,677
2,680
2,675
2,679
2,680
2,673
2,672
2,681
2,674
2,673
2,692
2,694
2,688
2,686
2,685
2,697
2,669
2,682
2,688
2,689
2,679
2,679

Financial
activities
3,532
3,651
3,784
3,920
4,023
4,047
4,155
4,348
4,599
4,843
5,025
5,163
5,209
5,334
5,553
5,815
6,128
6,385
6,500
6,562
6,614
6,561
6,559
6,742
6,910
6,866
7,018
7,255
7,565
7,753
7,783
7,900
7,956
8,078
8,105
8,197
8,367
8,348
8,206
7,838
7,695
7,697
7,784
7,880
7,733
7,741
7,765
7,766
7,778
7,781
7,781
7,789
7,803
7,812
7,816
7,827
7,835
7,847
7,853
7,863
7,872
7,885
7,901
7,897
7,896
7,903
7