View original document

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

The Economic Case for Strengthening Unemployment Insurance
Jason Furman
Chairman, Council of Economic Advisers
Center for American Progress
Washington, DC
July 11, 2016
This is an expanded version of these remarks as prepared for delivery.
Good morning. I want to thank the Center for American Progress (CAP), the National
Employment Law Project (NELP), and the Georgetown Center on Poverty and Inequality
(GCPI) for inviting me to offer opening remarks for today’s event. Unemployment insurance
(UI) is usually absent from the highest-profile debates over economic policy, except in brief
periods around the expiration of extended benefits. This is unfortunate because although UI only
benefits 2.1 million workers at this point in time, it provides a safety net for a much wider
group—as evidenced by the fact that more than 70 million American workers and their families,
including more than 17 million children, were supported by UI extensions in the Great Recession
(CEA and DOL 2014). Moreover, UI plays an important role in addressing three of the major
economic issues we face—the declining labor force participation rate, the increasing policy
challenges in addressing future recessions, and the changing nature of the employer-employee
relationship—and with thoughtful reforms could be even more effective in addressing all three.
The President put forward a detailed set of UI reforms in his Fiscal Year 2017 Budget, many of
them building on reforms enacted together with Congress in 2009 and 2012 and many of them in
common with the reforms so thoughtfully motivated and proposed in the report by CAP, NELP,
and GCPI in their report (West et al. 2016). The report also presents thought-provoking new
ideas, like a Jobseekers’ Allowance to help support those currently ineligible for UI, that are
worthy of discussion and consideration by policymakers.
Unfortunately, people often only pay attention to these issues at the wrong time: in the middle of
a recession, or a week before people are going to lose their extended benefits—or, even worse, a
week after they have lost their extended benefits, as happened more than once in recent years.
Instead, it is a discussion we should be having now while the labor market is strong.
In my discussion today I will start with the broader economic context that should inform UI
reform. Then I will turn to the benefits and tradeoffs in the current UI system, specifically
highlighting recent research which has tended to find considerably larger benefits and smaller
costs than previously thought. Finally, I will talk about six specific shortcomings of the current
system and some of the President’s proposals to address them by further building on the
strengths of the current UI system.

The Economic Context for Unemployment Insurance Reform: Falling Labor Force
Participation, Addressing Recessions, and the Changing Nature of Employment
Three broader economic issues inform and motivate the need for unemployment insurance
reform.
The first is the decline in recent years of labor force participation, particularly among those in
their prime working years (ages 25 to 54). The labor force participation rate—the share of
Americans working or actively seeking work—increased dramatically from the end of World
War II until around 2000 (as shown in Figure 1), as millions of women entered the workforce
and the Baby Boomers reached their prime working years.

Percent
68

Figure 1
Labor Force Participation Rate, 1948-2016

66
Jun-16

64
62
60
58
56
1948

1958

1968

1978

1988

1998

2008

Since 2000, however, the labor force participation rate has declined noticeably. In part, the
decline in labor force participation in recent years has been due to the aging of the U.S.
population, as more and more Baby Boomers have entered retirement. But the increase in overall
labor force participation in the second half of the twentieth century amid the large-scale entry of
women into the workforce masked a profoundly troubling trend: participation in the workforce
among men ages 25 to 54 has been declining for more than sixty years, from a peak of 98 percent
in 1954 to 88 percent today. This decline has been most pronounced among those with a high
school degree or less, as Figure 2 shows. Moreover, in the last decade and a half, prime-age
women have seen a similar pattern.

2

Percent
100

Figure 2
Prime-Age Male Labor Force Participation
by Educational Attainment
2015

98

Bachelor's or More

96
94
92
90

Some College

High School or Less

88
86
84
82
80

1964

1974

1984

1994

2004

2014

As documented in a recent CEA (2016) report, very little of the fall in prime-age male labor
force participation can be explained by reductions in labor supply (in other words, individuals
choosing not to work), including as a result of public assistance such as Social Security
Disability Insurance or UI, both of which are less generous in the United States than in many
other countries that have higher labor force participation rates. Instead, reductions in the demand
for labor, especially for lower-skilled men, appear to be an important component of this decline.
However, the design of U.S. labor market institutions—including the fact that U.S. labor markets
are relatively “flexible” (with few barriers to hiring and firing workers) but provide low levels of
support for jobseekers—also appears to be important, and may help explain some of the
differences in the U.S. experience both over time and compared to other countries (CEA 2016).
This is one issue that unemployment insurance reform can help to address by keeping workers
attached to the labor market and helping them find a job, increasing the likelihood of
reemployment.
The second issue is how to address future recessions. The United States has sufficient fiscal
space to take whatever steps are needed to counteract a future recession, as recent experience in
advanced economies and financial markets indicates (Furman and Shambaugh 2016). But given
the political climate, it may be difficult for Congress to move rapidly on countercyclical fiscal
policy measures in future downturns. Additionally, while the Federal Reserve counteracts
recessions by lowering interest rates, with both low inflation and a long-term trend toward lower
real interest rates, the Fed may have less room to respond to the next recession—whenever it
occurs. As Figure 3 shows, real interest rates have been falling since the 1980s in major
advanced economies and were already relatively low even before the extraordinary steps taken to
combat the crisis. There is substantial evidence that, going forward, the equilibrium interest rate
will continue to be low, as posited by either the secular stagnation or global savings glut
hypotheses (Summers 2014; CEA 2015). In a world where both the political space for fiscal
policy and the economic space for conventional monetary policy may be constrained, other
measures, including automatic stabilizers like unemployment insurance, are even more important
in counteracting cyclical downturns.
3

Figure 3
Real 10-Year Benchmark Rate in Selected Countries

Percent
8
7
6

France

5
4

Germany

3
2

2015

Japan

1

United States

0

United Kingdom

-1
-2
-3
1985

1990

1995

2000

2005

2010

2015

Finally, the third issue is the changing nature of the employment relationship, and in particular
the movement towards models of the employer-employee relationship that have shifted greater
risk onto workers. This includes greater “fissuring” of the workforce: firms’ shedding of
functions that were previously controlled internally to contractors and franchisees, leaving many
workers outside of labor protections that are predicated on the traditional employer-employee
relationship (Weil 2014). Katz and Krueger (2016), for example, have found that “nonstandard
work arrangements” (including independent contractors, on-call workers, and temporary help
agency employees) grew from 9 percent of employment in 1995 to 16 percent in 2015, as shown
in Figure 4.
Figure 4
Alternative Work Arrangements, 1995-2015

Percent of Employed Who Worked During Survey Week
20
18

15.8

16
14
12
10

9.3

10.1

8
6
4
2
0

CPS, Feb-1995

CPS, Feb-2005

4

RAND, Oct/Nov-2015

The changing nature of the employer-employee relationship also includes the potential for the
growth of the so-called “sharing economy,” which appears to be only a small portion of the U.S.
economy today but also appears to be growing rapidly. While bulk of employment relationships
are likely to remain relatively traditional, and thus a wholesale redo of public policy is not
necessary, policymakers do need to rethink how some of the existing models of labor policy can
be better adapted to keep up with changes in the nature of employment.

Why UI is Economically Important
Unemployment insurance has been a cornerstone of the American social insurance system since
the 1930s. I want to talk about four reasons why UI is economically important. Many of these
have been long-known or suspected, but several are the result of recent research using credible
research designs analyzing natural experiments that have led us to revise our traditional
understanding of UI.
First and foremost, unemployment insurance provides households with income when it is needed
most—that is, it provides “insurance” in the traditional sense. Involuntary job loss can have
striking effects on income and consumption: research examining UI suggests that, in the absence
of the UI system, a typical family whose head of household becomes unemployed would spend
22 percent less on food, compared with the 7 percent drop that is actually observed because of
income replacement by UI (Gruber 1997). Unemployment compensation played a substantial
role in maintaining household income levels during the Great Recession: CEA calculations show
that, from 2007 to 2010, the share of households receiving income from unemployment
compensation rose from 4.1 percent to 9.6 percent, and the average amount received by these
households rose from $4,400 to over $8,000 (CEA and DOL 2014). The Census Bureau
estimates that from 2008 to 2012, unemployment insurance lifted 11 million Americans above
the poverty line, although this is not a causal estimate because it does not account for what their
income would have been in the absence of UI (CEA 2014).
Second, unemployment insurance acts an “automatic stabilizer,” reducing the depth of economic
downturns. The loss of income associated with involuntary joblessness does not just affect
unemployed workers and their families; their reduced consumption contributes to deepening
recessions and slowing recoveries. Unemployment compensation is a particularly effective
automatic stabilizer for two reasons. First, it is timely, with claims automatically going up when
more people lose their jobs without any action from policymakers or any lag to determine
whether the economy is in recession. And second, beneficiaries tend to spend a high fraction of
their UI benefits because they want to smooth their consumption in the face of temporarily lower
income, precisely the insurance role that UI is designed to play (CEA and DOL 2014). The
Congressional Budget Office (2011) found that increasing aid to the unemployed is among the
policies that would have “the largest effects on output and employment per dollar of budgetary
cost,” and Auerbach and Feenberg (2000) found that UI played a stabilization role well in excess
of its relatively small share of the Federal Budget in recent downturns
Third, recent research has found that UI can help improve the functioning of labor markets by
keeping workers attached to the labor force who would otherwise drop out. The long-run decline

5

in prime-age male labor force participation that I noted earlier has not been slow and steady;
instead, it has tended to be concentrated in recessions, as more individuals are more likely to exit
the labor force altogether in cyclical downturns than in economic expansions. By including
active work-search requirements, the unemployment insurance system provides incentives for
individuals to continue to search for jobs rather than leave the workforce. Jesse Rothstein (2011)
has found that the majority of the modest increase in measured unemployment caused by the
availability of extended UI benefits in the Great Recession was attributable to lowering the
number of people who give up looking for work and leave the labor force rather than to lowering
the number who become employed. Likewise, Farber and Valletta (2015) examine benefit
extensions in both the early 2000s recession and the most recent recession and find a small but
statistically significant reduction in exits from unemployment to nonparticipation in the labor
force. Meanwhile, Farber, Rothstein, and Valletta (2015) find that the phase-out of extended
benefits since 2012 has lowered the unemployment rate by increasing exits from the labor force.
This “attachment effect” of UI is particularly important given the many negative effects—on
everything from future earnings to wellbeing and even mortality—associated with nonemployment, which is likely easier to exit from unemployment than from being outside of the
workforce altogether (CEA 2016).
Fourth, UI may also help improve the functioning of labor markets by providing liquidity and, in
turn, helping workers form better job matches. In the absence of unemployment insurance,
workers would be more likely to take one of the first jobs offered to them because they could not
afford to continue searching for a better job. In many cases, this could result in less efficient
matches between employees and employers, leaving workers and the economy worse off in the
long run. While empirical research on the existence and magnitude of the effect of UI on match
quality is limited, in part because of the difficulty of isolating how much of unemployment
duration is due to workers holding out for a better job match, recent work by Nekoei and Weber
(2015) finds that unemployment insurance does, in fact, increase the quality of recipients’ job
matches.

How Serious Are the Tradeoffs Associated with Unemployment Insurance?
Of course, as with many policies, the benefits of UI come with tradeoffs. But here, too, recent
research has cast substantial light, in many cases finding these tradeoffs to be considerably
smaller than previously appreciated. The principal tradeoff with unemployment insurance, like
much of social insurance, is “moral hazard,” the idea that UI may dampen the incentives for
recipients to exert effort to search for a job, leading to increased unemployment. The traditional
evidence for this was the “spike” in exits from unemployment insurance when unemployment
insurance expired, which some argued was indicative that, but for unemployment insurance,
more people would get jobs (Moffit 1985; Katz 1986; Katz and Meyer 1990; Feldstein 2005).
In the last decade, however, research has greatly qualified this interpretation. As I just discussed,
some of the benefits of unemployment insurance are that it can keep people attached to the labor
force instead of dropping out (the “attachment effect”) and that it can keep people looking for the
best job match rather than taking the first job that comes along regardless of its suitability (the
“liquidity effect”). But these, too, manifest themselves in the form of a higher measured

6

unemployment rate—albeit for good reasons. The relevant question requires us to tease out the
difference between the good and bad reasons for increased unemployment as a result of UI.
The “spike” itself turns out to be more an artefact of the measurement of labor force attachment
than an indicator of changing job seeking behavior. Card, Chetty, and Weber (2007) have found
that there is no spike in the percentage of people finding a job after unemployment insurance
benefits expire (as shown in Figure 5), at least in the Austrian data they study—though this is
consistent with other evidence from the United States as well. This suggests that the main effect
of benefit expiration is for people to stop reporting that they are looking for jobs—thus
disappearing from the labor force and the official measure of the unemployed.
Figure 5
Job Finding vs. Unemployment Exit Probabilities:
20-Week UI, Austria

Weekly Hazard Rate
0.20
0.18
0.16
0.14

Unemployment
Exit Probabilities

0.12
0.10
0.08
0.06
0.04
0.02
0.00

Job Finding Probabilities
0

10

20

30

40

50

But although not nearly as large as earlier interpretations of the data had suggested, there is still
clear evidence that unemployment insurance has some effect on how long it takes people to get
jobs. The question is whether this delay is for good reasons or for bad reasons. In the economic
framing, “moral hazard” would be the bad reason. In this context, by providing extra income
contingent on not having a job, unemployment insurance creates a disincentive for unemployed
workers to take jobs and thus leads to a distortion and too little job-taking. In the economic
framing, “liquidity” would be the good (or at least neutral) reason because UI benefits would not
affect the return to going to work and thus would not be distort recipients’ choices. 1 This effect
means that the extra income provided by UI decreases the sharp financial constraints that an

1

In the terms of introductory economics, the moral hazard effect is a “substitution effect” that comes about because
the relative price of work is changed—and thus people will inefficiently choose not to work even though the benefit
they would get (excluding taxes and transfers) from working would exceed the cost to them. The liquidity effect is
an “income effect” that may change individuals’ desire to engage in various forms of consumption but would still
leave them making an undistorted choice between the benefits of working and the costs of foregone time.
Technically, the income effect is purely a transfer and does not have any welfare cost or benefit. It is reasonable to
think that the consumption-smoothing benefit of the insurance it provides in the face of capital market imperfections
exceeds the cost of the distortionary taxes required to pay for it.

7

unemployed worker would face, helping him or her to smooth consumption and make the most
efficient choice about when or whether to take a future job opportunity.
Economists have studied this tradeoff extensively for several decades, providing a large body of
evidence on the relative importance of liquidity and moral hazard effects (e.g. Katz and Meyer
1990, Meyer 1990, Card and Levine 2000). As Raj Chetty pointed out in a 2013 op-ed, these
studies have come to a near-uniform conclusion: the liquidity/income effects of unemployment
insurance far outweigh any moral hazard/substitution effects, and that, as a result, policymakers
“can extend unemployment benefits to provide assistance to those out of work without
substantially increasing unemployment rates” (Chetty 2013). 2 In his own research studying both
differential responses to unemployment insurance by people who are more or less liquidityconstrained and also lump-sum severance payments, Chetty (2008) finds hardly any evidence of
a moral hazard effect once the liquidity effect is accounted for.
Of course, moral hazard is not entirely absent, and there is still some tradeoff between moral
hazard and liquidity effects. This is why unemployment insurance benefits replace only a
fraction of previous earnings and last for a finite period of time. But the magnitude of the
tradeoff varies over time and is itself a function of economic conditions. When jobs are scarce,
reduced search effort has a smaller effect on the unemployment rate than when jobs are plentiful.
This is due to the fact that when jobs are scarce, if one person doesn’t take an available job as
quickly it just means someone else will be able to take it. This intuition has been confirmed by
recent papers, including Kroft and Notowidigdo (2016) and Schmieder, von Wachter, and
Bender (2012). An important implication of this finding is that increasing or extending
unemployment insurance benefits during recessions can actually improve overall social welfare,
since their liquidity (i.e. socially beneficial) effects are more likely to predominate during
downturns. So in addition to its macroeconomic effect as an automatic stabilizer, expanding UI
during recessions also has important positive microeconomic effects.

Six Shortcomings of the Current UI System—and the President’s Proposals to Address
Them
While the current unemployment insurance system provides a number of important benefits to
workers, their families, and to the U.S. economy as a whole, there is still room to do better. UI
can be improved to better deal with the three challenges I began with: the declining labor force
participation rate, the need for better options for fighting future recessions, and the changing
nature of the workforce. The Administration has already made a number of important reforms to
UI in the 2009 Recovery Act and the Middle-Class Tax Relief and Job Creation Act of 2012,
including measures that have helped expand UI to more part-time workers, prevent layoffs via
greater use of work-sharing, and modernize the administration of the UI system. The President’s
FY 2017 Budget contains an ambitious reform proposal building on these ideas that has much in
common with the recommendations of the report by CAP, NELP and GCPI. Let me highlight six
more specific shortcomings of the current UI system and the President’s proposals to address
them.
2

Research has found similar results for Supplemental Security Income for children (Deshpande, forthcoming) and
Social Security Disability Insurance (Gelber, Moore. and Strand 2016).

8

1. Declining Coverage
UI benefit receipt has fallen to historically low levels in recent years, declining 8 percentage
points in the past decade to just 27 percent of unemployed workers nationwide, as shown in
Figure 6. In part, this drop is due to the expiration in December 2013 of the Federal Emergency
Unemployment Compensation (EUC) program first authorized by Congress in 2008 and
expanded under the Recovery Act. However, and more importantly, declining recipiency rates
are also the result of reductions in coverage within States’ regular UI programs. For the first time
in over fifty years, nine States have cut their unemployment insurance maximum benefit duration
to less than 26 weeks, and some, such as North Carolina, now provide only up to 13 weeks of
coverage. A number of States have also tightened eligibility requirements for UI. These recent
reductions follow a longer-run pattern of reduced coverage: following the 1980s recessions,
eleven States reduced the duration of benefits down to 26 weeks and many others took steps to
raise eligibility requirements.

Percent
80

Figure 6
Share of Unemployed Workers Receiving
UI Benefits, 1972-2015
Regular programs

70

Federal programs

60
50
40
30
20
10
0
1972

1977

1982

1987

1992

1997

2002

2007

2012

Low levels of coverage limit the extent to which UI can act effectively, both on a micro level by
protecting families from poverty as a result of involuntary job loss, and on a macro level by
automatically countering downturns in the business cycle. To address this issue, the President’s
proposal would require that State UI programs cover many of the workers who currently fall
through the cracks, including part-time workers, newer labor market entrants, certain low-income
and intermittent earners, and workers who leave work for compelling family reasons such as to
move with a spouse, escape domestic violence, or care for an ill family member. It would also
reverse the damaging erosion of State UI benefit duration by requiring that all State programs
provide at least 26 weeks of coverage, consistent with the historic norm.

9

2. Insolvency of State Programs
In part, reductions by States of UI coverage have come as the result of increased fiscal pressures
during the previous recession. Thirty-six States exited the Great Recession having relied on
emergency Federal loans for UI, with more than $3 billion in loans still outstanding, and several
States have relied on private borrowing to meet shortfalls in their respective UI trust funds. Low
State UI trust fund reserves remain a serious threat to unemployment insurance for working
Americans. Three out of five State UI programs currently do not have enough in their trust funds
to pay for a single year of benefits in a recession—a common measure of program solvency.
It is important to note, though, that declining solvency was an issue even before the Great
Recession. According to a 2010 report from the Government Accountability Office, from 1938
to 1973, State UI trust funds held average year-end reserves, net of loans, equal to 5.1 percent of
wages, and never dropped below 2 percent. From 1974 to 2008, that average fell to 1.0 percent
of wages and has never been as high as 2 percent. UI tax contributions as a percentage of UIcovered wages have trended downward in recent decades, from an annual national average of
1.15 percent from 1979 to 1988 to 0.65 percent from 1999 to 2008 (GAO 2010).
The President’s Budget proposal would put State UI programs on a path to permanent solvency
while ensuring they have sufficient reserves to weather the next economic crisis. Starting in FY
2018 the proposal would broaden the taxable wage base from $7,000 to $40,000 under the
Federal Unemployment Tax Act (FUTA), while cutting the tax rate from 0.8 percent to 0.167
percent to keep the proposal revenue-neutral at the Federal level. States would be effectively be
required to set their taxable wage bases at least to the level of the FUTA tax base, and although
States could also cut their rates to offset the additional revenues, in practice many would not and
thus would see improvements in the solvency of their UI systems. This would help ensure that
States are better able to deal with increased demands on their UI systems in a future recession.
3. Ineffective Countercyclical Triggers
One of the key effects of the UI system is its ability to act as an automatic stabilizer during
recessions by providing much-needed economic stimulus. Moreover, whatever one thinks of the
optimal balance of “moral hazard” and “liquidity” effects, it is the case that the optimal policy in
recessions should be to provide longer and potentially more generous benefits.
Under current law, the Extended Benefits (EB) program provides 50 percent of the funding for
States that choose to extend benefits in a downturn, defined as a period of sustained high
unemployment. This provision, however, has frequently been ineffective because States do not
want to pay their half of the cost often due to balanced budget requirements that result in
pressures to cut spending during downturns. Furthermore, the triggers themselves can turn off for
States that experience persistently high unemployment rates because eligibility is based on
changes in unemployment rates, not levels.
Congress has, on an ad hoc basis, established 100-percent Federally funded Extended
Unemployment Compensation (EUC) in recessions since 1972. But the length and timing of
these extensions have varied widely, most recently with Congress allowing them to lapse at the

10

end of 2013 despite that fact that, in the aftermath of the worst downturn since the Great
Depression, the mean and median durations of unemployment were still higher than in any
recession since World War II.
Changes to the UI system to ensure that it acts as a better countercyclical measure are some of
the most urgently-needed reforms in this area, particularly given concerns around the efficacy of
available policy tools to mitigate the effects of recessions. The President’s Budget would
modernize the Extended Benefits program into a permanent, 100-percent Federally funded
program that would automatically provide up to 52 additional weeks of benefits for States
experiencing either rapid increases in unemployment or a high level of unemployment. By
making triggers for extended benefits automatic, the revised Extended Benefit program would
help to ensure that stimulus is provided where it is needed most, when it is needed most—
strengthening the UI system’s ability to protect against recessions.
4. Incentives to Reduce Employment, Not Hours
Currently, workers may qualify for regular UI benefits if they are laid off and otherwise eligible,
but they are often ineligible for benefits if their hours are reduced by their employer. This may
discourage employers from avoiding layoffs by temporarily reducing hours across the board
when demand for their products or services falls and also discourages workers from accepting
lighter schedules. As a result, these incentives limit the ability of the UI system to reduce the
number of families who face the trauma (and lasting economic consequences) of a layoff and
keep individuals attached to the labor force, and may encourage employers to lay off workers
rather than reduce their hours and keep them on payrolls.
One solution to this problem would be to allow groups of workers whose hours were temporarily
reduced to receive unemployment benefits in order to offset some of their lost earnings. By
removing the incentives for firms to cut jobs rather than merely reduce hours, these programs
would help prevent job losses during an economic downturn, as a similar program during the
most recent recession did in Germany—which saw larger output losses than the United States but
a much smaller increase in its unemployment rate (Baker 2011). Abraham and Houseman (2014)
estimate that if the United States had had work-sharing at European levels during the Great
Recession, as many as one in eight of the roughly 8 million jobs lost during the recession could
have been saved. Economists on both the left and the right—including Dean Baker of the Center
for Economic and Policy Research and Kevin Hassett from the American Enterprise Institute—
have supported expanding work-sharing programs (Baker and Hassett 2012).
Progress has been made in this area via the Short-Time Compensation (STC) program, a
voluntary program in which States can enter into agreements with employers to provide
employees whose hours have been reduced with a portion of weekly UI payments to compensate
for lost income as a result of hours reductions. Twenty-seven States have Short-Time
Compensation programs established in law, with twenty-five of them currently operational. The
President’s Budget would provide implementation grants and other incentives to States without
work-sharing programs, including a cost-splitting mechanism for Short-Time Compensation
benefits when a State is triggered onto Extended Benefits.

11

5. Insufficient Support for Job Search
While the United States has the types of flexible labor markets that traditional economic
prescriptions recommend for a well-functioning labor market, U.S. labor markets are much less
“supportive” than in other OECD countries. The United States spends just 0.1 percent of GDP on
“active labor market policies” such as job-search assistance and job training, much less than the
OECD average of 0.6 percent of GDP, and less than every other OECD country except Chile and
Mexico (OECD 2015). These policies provide more help for individuals in other countries to
find jobs by providing more support during unemployment (when individuals must also search
for work) and making reemployment more likely through skill-building and facilitating jobmatching.
Reforming the UI system so that it provides better support for jobseekers, as well as those who
want to retool and retrain to take advantage of new opportunities in the job market, would help
ensure that UI recipients are more likely to remain attached to the workforce. The President’s
proposals would provide States with incentives to directly link jobseekers to work as a
reemployment strategy by providing States with incentives to adopt work and training-related
reforms. For example, States would receive incentives to create temporary work-based training
programs—including apprenticeship programs—to help workers get back on the job, with
considerable flexibility in the design of such programs to better match local labor needs. States
would also receive incentives to allow workers to continue receiving UI benefits while
participating in an apprenticeship or on-the-job training. States could also receive funds for
relocation assistance programs to help workers pursue new opportunities in a different
community.
6. Insufficient Insurance Against More Adverse Job Prospects
Unemployment insurance provides liquidity benefits that help workers smooth consumption but
at some modest cost in terms of the moral hazard associated with taking a job. One way to
handle this tradeoff would be to make unemployment insurance benefits less generous but, as
discussed earlier, this would be costly to job losers. A better way is to raise the return to finding
a new job in a manner that both preserves the current insurance against job loss and also better
encourages people to get back to work. Workers seeking reemployment also face the risk of not
being able to find a new job at the same earnings level, since experienced workers starting over
find themselves earning wages 10 percent or more below what they earned in the jobs they lost
(Farber 2004) and a reluctance to adjust earnings expectations can discourage them from
accepting new jobs (Krueger and Mueller 2016). While this can be a long-term shock to income,
many workers would still benefit from some ability to smooth this shock, not just to help them
maintain their standard of living but also to encourage them to take a lower paying job rather
than continuing to wait for a position closer to their old salary and potentially risking a longer
spell of unemployment.
One way to create this incentive is to offer wage insurance to help protect workers against lost
income. Evidence suggests that workers transition out of work—and thus back into paying
employment—more quickly when they have wage insurance: an experiment in Canada found
that wage insurance increased entry into full-time work by 4.4 percentage points at the 26-week

12

mark (Bloom et al. 1999). The President’s Budget contains a proposal to establish a wage
insurance system that would replace up to 50 percent of lost wages (up to a limit of $10,000, paid
over two years) for unemployed workers with at least three years of tenure who take new, lowerpaying jobs earning less than $50,000. This system, which would be administered through State
UI programs, would offer protection against reduced earnings and create an incentive for the
unemployed to get back into employment quickly and to remain in the workforce, preventing
them from facing the challenges associated with long-term unemployment.

Conclusion
Unemployment insurance is a pillar of the U.S. social safety net: economic research has shown it
to be effective in keeping individuals attached to the labor force, helping households weather
downturns, and lifting the economy out of recessions. The proposals put forward by the
Administration and by CAP, NELP, and GCPI represent positive steps forward in modernizing
and strengthening the UI system so that it meets the challenges of the 21st century economy—
promoting participation in the labor force, acting as an effective automatic stabilizer, and
responding to the changing needs of the U.S. workforce. Reforming the UI system may not be
the most talked-about policy issue today, but it is urgently needed in order to make sure that both
workers and the economy as a whole are better protected against the next recession.

13

Notes to Figures
Figure 1
Note: Shading denotes recession.
Source: Bureau of Labor Statistics, Current Population Survey.
Figure 2
Source: Bureau of Labor Statistics, Current Population Survey (Annual Social and Economic
Supplement); CEA calculations.
Figure 3
Source: National sources via Haver Analytics.
Figure 4
Note: Alternative work arrangements include independent contractors, on-call workers,
temporary help agency workers, and workers provided by contract firms.
Source: Katz and Krueger (2016).
Figure 5
Source: Card, Chetty, and Weber (2007).
Figure 6
Note: Shading denotes recession.
Source: West et al. (2016).

14

References
Abraham, Katharine G. and Susan G. Houseman. 2014. “Encouraging Work Sharing to Reduce
Unemployment.” In Melissa S. Kearney and Benjamin H. Harris, eds., Policies to Address
Poverty in America. Washington: The Hamilton Project.
Auerbach, Alan J. and Daniel Feenberg. 2000. “The Significance of Federal Taxes as Automatic
Stabilizers.” Journal of Economic Perspectives 14 (3): 37-56.
Baker, Dean. 2011. “Work Sharing: The Quick Route Back to Full Employment.” Center for
Economic and Policy Research. Report.
Baker, Dean and Kevin Hassett. 2012. “The Human Disaster of Unemployment.” New York
Times. May 12.
Bloom, Howard, Saul Schwartz, Susanna Lui-Gurr, and Suk-Won Lee. 1999. “Earnings
Supplement Project (ESP) – Testing a Re-employment Incentive for Displaced Workers.” Social
Research and Demonstration Corporation. Report.
Card, David, Raj Chetty, and Andrea Weber. 2007. “The Spike at Benefit Exhaustion: Leaving
the Unemployment System or Starting a New Job?” American Economic Review 97 (2): 113118.
Card, David and Phillip B. Levine. 2000. “Extending Benefits and the Duration of UI Spells:
Evidence from the New Jersey Extended Benefit Program.” Journal of Public Economics 78 (1):
107-138.
Chetty, Raj. 2008. “Moral Hazard versus Liquidity and Optimal Unemployment Insurance.”
Journal of Political Economy 116 (2): 173-234.
______. 2013. “Yes, Economics is a Science.” New York Times. October 20.

Congressional Budget Office (CBO). 2011. “Policies for Increasing Economic Growth and
Employment in 2012 and 2013.” Testimony of Douglas W. Elmendorf, Director, before the
Committee on the Budget, United States Senate.
Council of Economic Advisers (CEA). 2014. “Chapter 3: The Economic Impact of the American
Recovery and Reinvestment Act Five Years Later.” Economic Report of the President.
______. 2015. “Long-Term Interest Rates: A Survey.” Report.
______. 2016. “The Long-Term Decline in Prime-Age Male Labor Force Participation.” Report.

Council of Economic Advisers and the Department of Labor (CEA and DOL). 2014. “The
Economic Benefits of Extending Unemployment Insurance.” Report.

15

Deshpande, Manasi. Forthcoming. “The Effect of Disability Payments on Household Earnings
and Income: Evidence from the SSI Children’s Program.” Review of Economics and Statistics.
Farber, Henry S. 2004. “Job Loss in the United States, 1981-2001.” In Polachek, Solomon W.,
ed., Accounting for Worker Well-Being, Research in Labor Economics, vol. 23. Amsterdam, San
Diego, and Oxford: Elsevier.
Farber, Henry S. and Robert G. Valletta. 2015. Do Extended Unemployment Benefits Lengthen
Unemployment Spells? Evidence from Recent Cycles in the U.S. Labor Market.” Journal of
Human Resources 50 (4): 873-909.
Farber, Henry S., Jesse Rothstein, and Robert G. Valletta. 2015. “The Effect of Extended
Unemployment Insurance Benefits: Evidence from the 2012-2013 Phase-Out.” American
Economic Review 105 (5): 171-176.
Feldstein, Martin. 2005. “Rethinking Social Insurance.” American Economic Review 95 (1): 124.
Furman, Jason and Jay C. Shambaugh. 2016. “Fiscal Policy Remains Critical for Much of the
World Economy.” VoxEU.org. April 29.
Gelber, Alexander, Timothy Moore, and Alexander Strand. 2016 “The Effect of Disability
Insurance Payments on Beneficiaries’ Earnings.” NBER Working Paper No. 21851.
Government Accountability Office (GAO). “Unemployment Insurance Trust Funds: Longstanding State Financing Policies Have Increased Risk of Insolvency.” GAO Report No. GAO10-440.
Katz, Lawrence F. 1986. “Layoffs, Recall and the Duration of Unemployment.” NBER Working
Paper No. 1825.
Katz, Lawrence F. and Alan B. Krueger. 2016. “The Rise and Nature of Alternative Work
Arrangements in the United States, 1995-2015.” Working paper.
Katz, Lawrence F. and Bruce D. Meyer. 1990. “The Impact of the Potential Duration of
Unemployment Benefits on the Duration of Unemployment.” Journal of Public Economics 41
(1): 45-72.
Krueger, Alan B. and Andreas I. Mueller. 2016. “A Contribution to the Empirics of Reservation
Wages.” American Economic Journal: Economic Policy 8 (1): 142-179.
Kroft, Kory and Matthew J. Notowidigdo. 2016. “Should Unemployment Insurance Vary with
the Unemployment Rate? Theory and Evidence.” Review of Economic Studies 83 (3): 10921124.

16

Meyer, Bruce D. 1990. “Unemployment Insurance and Unemployment Spells.” Econometrica 58
(4): 757-782.
Moffit, Robert. 1985. “Unemployment Insurance and the Distribution of Unemployment Spells.”
Journal of Econometrics 28 (1): 85-101.
Nekoei, Arash and Andrea Weber. 2015. “Does Extending Unemployment Benefits Improve Job
Quality?” IZA Discussion Paper No. 9034.
Organisation for Economic Co-operation and Development (OECD). 2015. Economic Policy
Reforms: Going for Growth 2016. Report.
Rothstein, Jesse. 2011. “Unemployment Insurance and Job Search in the Great Recession.”
Brookings Papers on Economic Activity 43 (2): 143-213.
Schmieder, Johannes F., Till von Wachter, and Stefan Bender. 2012. “The Effects of Extended
Unemployment Insurance Over the Business Cycle: Evidence from Regression Discontinuity
Estimates Over 20 Years.” Quarterly Journal of Economics 127 (2): 701-752.
Summers, Lawrence H. 2014. “U.S. Economic Prospects: Secular Stagnation, Hysteresis, and the
Zero Lower Bound.” Business Economics 49 (2): 65-73.
Weil, David. 2014. The Fissured Workplace: Why Work Became So Bad for So Many and What
Can Be Done to Improve It. Cambridge, MA: Harvard UP.
West, Rachel, Indivar Dutta-Gupta, Kali Grant, Melissa Boteach, Claire McKenna, and Judy
Conti. 2016. “Strengthening Unemployment Protections in America: Modernizing
Unemployment Insurance and Establishing a Jobseeker’s Allowance.” Center for American
Progress, National Employment Law Project, and Georgetown Center on Poverty and Inequality.
Report.

17