View original document

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

A Quarterly Review
of Business and
Economic Conditions
Vol. 23, No. 1

Unemployment Rate
Is It the Best Measure
of Labor Market Health?

President Bullard
Lessons from the Past
on Raising Rates

January 2015

The Federal Reserve Bank of St. Louis
Central to America’s Economy®

Growth around the World
Is Still Below the Trend
U.S. Faring Better than Most

c o n t e n t s

4

A Quarterly Review
of Business and
Economic Conditions
Vol. 23, No. 1

Growth around the World Still Below Trend

Unemployment Rate
Is It the Best Measure
of Labor Market Health?

President Bullard
Lessons from the Past
on Raising Rates

January 2015

THE FEDERAL RESERVE BANK OF ST. LOUIS
CENTRAL TO AMERICA’S ECONOMY®

By Juan M. Sánchez

The world’s output for 2014 is expected to end up being below trend,
and the forecast for this year doesn’t look much better. Once again,
the U.S. is performing better than most other developed countries.
Growth around the World
Is Still Below the Trend
U.S. Faring Better than Most

The Regional

Economist
january 2015 | VOL. 23, NO. 1

3

P resident ’ s M essa g e

8

European Debt Crisis
Recalls the Lost Decade

The Regional Economist is published
quarterly by the Research and Public Affairs
divisions of the Federal Reserve Bank
of St. Louis. It addresses the national, international and regional economic issues of
the day, particularly as they apply to states
in the Eighth Federal Reserve District. Views
expressed are not necessarily those of the
St. Louis Fed or of the Federal Reserve System.

By Maria A. Arias
and Paulina Restrepo-Echavarria
In many ways, the European
debt crisis is reminiscent of
Latin America’s experience in
the 1980s, characterized by a period of high growth interrupted
by an external shock. But there
are some notable differences.

Director of Research
Christopher J. Waller
Senior Policy Adviser
Cletus C. Coughlin
Deputy Director of Research
David C. Wheelock

10

Director of Public Affairs
Karen Branding

The Difficulty in Measuring
the Underground Economy
By Paulina Restrepo-Echavarria

Editor
Subhayu Bandyopadhyay

There are at least two main ways
to measure “the informal sector”
of an economy, both of which
entail difficulties. The effort is
needed, however, because the
underground economy accounts
for about 13 percent of GDP in
developed countries and almost
three times that in developing
countries.

Managing Editor
Al Stamborski
Art Director
Joni Williams

Please direct your comments
to Subhayu Bandyopadhyay
at 314-444-7425 or by email at
subhayu.bandyopadhyay@stls.frb.org.
You can also write to him at the
address below. Submission of a
letter to the editor gives us the right
to post it to our website and/or
publish it in The Regional Economist
unless the writer states otherwise.
We reserve the right to edit letters

12

for clarity and length.

Labor Force Participation
by Youth Drops; Why?
By Maria Canon, Marianna
Kudlyak and Yang Liu

Single-copy subscriptions are free
but available only to those with
U.S. addresses. To subscribe, go to

Workforce participation has
declined among those 16 to 24,
but there may be good reasons
for this. An analysis by age, gender and education looks at who
is in school and who is not.

www.stlouisfed.org/publications.
You can also write to The Regional
Economist, Public Affairs Office,
Federal Reserve Bank of St. Louis,
P.O. Box 442, St. Louis, MO 63166-0442.

The Eighth Federal Reserve District includes

all of Arkansas, eastern Missouri, southern
Illinois and Indiana, western Kentucky and
Tennessee, and northern Mississippi. The
Eighth District offices are in Little Rock,
Louisville, Memphis and St. Louis.

2 The Regional Economist | January 2015

of the labor market these days.
Are any of the other standard
indicators any better? What do
indexes that measure the labor
market conditions tell us?

14

Assessing the Health
of the Labor Market

16

and income over the past halfcentury. Although the area was
not immune to the Great Recession, Northwest Arkansas could
be on the verge of another spurt,
given that its economy often
follows that of the U.S. business
cycle, now in an upswing.

econom y at a g lanc E

17	national overview
Pace of Growth
Is Likely to Quicken

23

reader e x c h an g e

By Kevin L. Kliesen
There are a few negative developments that bode ill for the
U.S. economy this year, but they
are probably outweighed by
recent positive developments.
As a result, GDP growth is likely
to be stronger this year than
in 2014.

18	district overview
Income Inequality:
District vs. Nation
By Maximiliano Dvorkin
and Hannah Shell
Income inequality has increased
in the St. Louis Fed’s District
over the past 30 years, although
at a slower pace than in the
nation as a whole. In both areas,
the inequality is increasing primarily between the top-income
earners and the middle-income
earners.

20	metro profile
Widespread Growth
in Northwest Arkansas

By Maximiliano Dvorkin

By Charles S. Gascon
and Michael A. Varley

A question often on people’s
minds is whether the unemployment rate is capturing all the relevant information on the health

Home to Walmart and several
other large companies, this
region has experienced unusually strong growth in population

ONLINE EXTRA
Read more at www.stlouisfed.org/
publications/regional-economist.

Pattern in Job Gains
after Recessions Appears
To Be Changing
By Kevin L. Kliesen
and Lowell R. Ricketts

Following the two latest recessions, the growth in high-paying
jobs was stronger, on a percentage basis, than was the growth
in low-paying jobs. The opposite
happened after the previous two
recessions.
When Unemployment
Falls, Does the Average
Wage Go Up?
By James D. Eubanks
and David G. Wiczer

As the unemployment rate
declines, many people assume
that the average wage in the U.S.
will increase. However, the average doesn’t move that fast over
a single business cycle. And any
movement over the long term is
more in favor of high-wage earners than low-wage earners.

p r e s i d e n t ’ s

m e s s a g e

Liftoff: A Comparison
of Two Normalization Cycles

M

any Federal Open Market Committee
(FOMC) participants have said that
the policy rate (i.e., the target for the federal
funds rate) should come off the zero lower
bound in 2015, with the exact timing dependent on how key macroeconomic indicators
evolve. Given that this initial increase would
mark the start of a normalization cycle, now
is a good time to review the previous two
major normalization cycles to see what we
can learn from them.1
The first normalization cycle for comparison began in 1994. The policy rate since September 1992 had been at 3 percent, which
at the time was considered exceptionally
low relative to the federal funds rate during
the 1970s and 1980s. U.S. macroeconomic
data indicated a strong economy toward
the end of 1993. For instance, real gross
domestic product (GDP) growth accelerated in the fourth quarter, job growth was
slightly stronger on average and inflation
was threatening to move higher. In what
was largely a surprise to financial markets,
the FOMC began a normalization cycle
in February 1994 and continued raising
rates throughout that year.2 In contrast
to the second normalization cycle I will
highlight, the FOMC raised the policy rate
by 25 basis points sometimes, by 50 basis
points other times and by 75 basis points
on one occasion. Also, the policy rate was
left unchanged at a few meetings. The pace
was adjusted in reaction to the incoming
macroeconomic data and in this sense was
data-dependent, or state-contingent. The
normalization cycle ended in February
1995, with a policy rate of 6 percent.
Financial markets generally viewed this
adjustment to higher interest rates as disorderly. In fact, the bond market had one of its

worst years in 1994. The 10-year Treasury
yield, for instance, rose roughly 2 percentage
points that year. Despite being disorderly,
the 1994 normalization turned out to be a
success for the U.S. economy. The policy
rate was returned to a more normal level,
and the economy boomed in the second half
of the 1990s—one of the best periods for
economic growth in the postwar era.
The second normalization cycle for
comparison took place in 2004-06. The
policy rate had been 1 percent since June
2003. Leading up to the June 2004 FOMC
meeting, real GDP growth remained solid,
gains in nonfarm payroll employment had
increased in recent months and inflation
had risen. The FOMC raised the policy rate
to 1.25 percent in June 2004 and continued
with a mechanical pace of increase of 25
basis points at each of the next 16 meetings.
Thus, there was almost no state contingency
with this normalization cycle. In terms of
communication, the FOMC was more transparent regarding its expectations for future
increases in the policy rate than it had been
previously. This cycle ended in June 2006,
bringing the policy rate to 5.25 percent.
Financial markets viewed this form of
normalization as much more orderly than
the 1994 case and, therefore, a success.
However, this normalization cycle may have
been counterproductive. The housing bubble
inflated even more during this two-year
period as financial markets found ways to
create investments in housing based on cheap
financing—investments that ultimately
proved disastrous. Although policymakers
were cognizant that house prices were rising
and that mortgage finance was increasing,
the general view was that the air could be
let out of the bubble slowly and without

dramatic macroeconomic consequences. In
actuality, the opposite occurred. The housing
bubble burst, starting in 2006, right about the
time the normalization cycle ended. House
prices fell about 30 percent, and the U.S.
experienced a severe recession.
What are the lessons from these two
episodes? Although the 1994 normalization
cycle was considered disorderly (i.e., uneven
amounts that were somewhat unpredictable), it seemed to set up the U.S. economy
for success in the second half of the 1990s.
On the other hand, the 2004-06 normalization cycle was considered orderly (i.e.,
perfectly even amounts that were generally
anticipated) but, in retrospect, turned out
to be suboptimal because it allowed for
the continuation of speculation in housing
markets and in mortgage finance. For the
upcoming normalization cycle, some combination of the two—the data dependency
from the 1994 case and the transparency
from the 2004-06 case—would probably
provide the optimal method of returning
the policy rate to normal.

James Bullard, President and CEO
Federal Reserve Bank of St. Louis

ENDNOTES
1 For related commentary, see my essays “1994,”

Federal Reserve Bank of St. Louis National
Economic Trends, July 2004; and “The Taylor
Principle and Recent FOMC Policy,” Federal
Reserve Bank of St. Louis Monetary Trends,
September 2006. They can be found at http://
research.stlouisfed.org/econ/bullard.
2 See http://research.stlouisfed.org/fred2/
graph/?g=WYM.

The Regional Economist | www.stlouisfed.org 3

i n t e r n a t i o n a l

h e a d w i n d s

Growth around the World
Is Still Below the Trend
U.S. Faring Better than Most
By Juan M. Sánchez

T

he recovery of the U.S. economy from
the 2007-09 recession has been slow
but uninterrupted. Unfortunately, the same
cannot be said about other regions in the
world. Put in the context of international
headwinds, the U.S. recovery looks stronger
and the projected growth for this year more
valuable to overall world growth. As will be
seen below, more of the world’s new output
will be generated by the U.S. in 2014-2015
than in 2000-2007.
According to the Oct. 7 World Economic
Outlook (WEO) of the International Monetary Fund (IMF),1 during the first half of
2014 there were several pieces of disappointing news: weaker activity in Russia; slower
growth in Latin America, mainly Brazil;
weaker-than-forecast expansion of gross
domestic product (GDP) in Japan; weaker
activity in China; and stagnant growth in
the euro area. These international headwinds are not new, we argue here, and may
account, at least partly, for the slow recovery
of the U.S. economy.
The world’s output for 2014 is projected
to be only 3.3 percent. (See table.) This
is much lower than previous averages at
the world level. For instance, in 2006 and
2007, the two years prior to the recession,
yearly world output growth was about 5.6
percent. The poor performance of world
output in the past year is primarily due
to the low growth projected for advanced
economies,2 only 1.8 percent. Again, this is
4 The Regional Economist | January 2015

Projected Growth by Region
Percent Change
October
Projections
World Output

Difference from
July Projections

2014

2015

2014

3.3

3.8

–0.1

–0.2

2015

Advanced Economies

1.8

2.3

0.0

–0.1

United States

2.2

3.1

0.5

0.0

Euro Area

0.8

1.3

–0.3

–0.2

Germany

1.4

1.5

–0.5

–0.2

France

0.4

1.0

–0.4

–0.5

Italy

–0.2

0.8

–0.5

–0.3

Spain

1.3

1.7

0.1

0.1

SOURCE: The International Monetary Fund’s World Economic Outlook.

significantly lower than the growth rate of
advanced economies in previous years—
close to 3 percent in 2006 and 2007. Among
the advanced economies, the U.S. has the
highest projected growth rate for 2014, at 2.2
percent. Growth in the euro area, however,
is projected to be much lower, 0.8 percent.
Projections for the year 2015, also shown
in the table, are slightly better. The world
output is projected to increase 3.8 percent in
2015 vs. 3.3 in 2014. Advanced economies’
output is projected to grow 2.3 in 2015 vs.
1.8 in 2014. The euro area is projected to
grow 1.3 in 2015 vs. 0.8 in 2014. Although
growth is projected to be better in 2015, the
projections are below previous averages. The

average growth for the years 2000-2007 was
4.5 percent for the world, 2.6 percent for
advanced economies and 2.2 for the euro area.
The last two columns of the table contain
useful information to understand how
things changed since July 2014, when the
previous projections of the WEO were
published. Those columns report the change
in the projections between July and October 2014. If we focus on projected output
growth, an increase in the forecast (a positive number in the last two columns of the
table) means that forecasters must have
received positive news during those periods.
If changes in projections were negative, bad
news must have been received. Notice that
most of the numbers in the last two columns
are negative, except for those for the U.S.
and Spain (and Spain’s numbers are very
close to zero). And some of those changes
are very large. For instance, for 2014 the
growth projected for Germany, France and
Italy was reduced by 0.5, 0.4 and 0.5 percentage points, respectively.
How We Got Here: A “News Index”

Based on the table discussed above, we
created a “news index” that accumulates the
revisions to the forecast. As mentioned above,
in the context of output growth, the sign of
the revision in the projections indicates the
sign of the news. Moreover, the size of the
revision is indicative of the dimension of the
news. In particular, the size of the revision

can be compared with the usual growth of
output (or any other variable for that matter) to have a sense of the size of the news.
Considering output growth, a 0.1 revision (as
observed for Spain for 2014 and 2015) can be
interpreted as mildly positive news, while a
–0.5 revision (as observed for Italy for 2014)
indicates very negative news.
Moreover, note that this metric can be
computed for the forecast of any macro-

The Deflation Scare

Figure 1 shows inflation in the advanced
economies since September 2011. The red
line fluctuates around zero, with the largest
accumulated value at 0.7 percentage point. As
mentioned above, fluctuations around zero
are expected if revisions do not contain much
news but just noise corrections.
However, note the decline of almost 1 percentage point from April 2013 to April 2014.

figure 1

0.8
0.6
0.4
0.2
0

Oct. ’14

July ’14

May ’14

Feb. ’14

Nov. ’13

Sept. ’13

June ’13

March ’13

Jan. ’13

Oct. ’12

July ’12

May ’12

Feb. ’12

Dec. ’11

–0.4

Sept. ’11

–0.2
June ’11

Cumulative Percentage Revised since Sept. 2011

“News Index,” Advanced Economies’ Inflation

These international
headwinds are not
new, we argue
here, and may
account, at least
partly, for the
slow recovery of
the U.S. economy.

SOURCE: World Economic Outlook.

economic variable. The difference is that
the sign of the revision may not necessarily coincide with the sign of the news.
For instance, in an inflationary context, a
reduction in the predicted inflation would
be good news.
Since we intend to use this metric to analyze how we got to the forecasts of October
2014, we accumulated this news (changes in
forecasts) since September 2011. In particular,
we accumulated the revisions for both years
(the year of the report and the next year) for
each report. The fact that we used both years
explains why there are two observations for
every date.3 At any particular period, the
index indicates the sum of all revisions in
all the WEO publications for that variable
forecast starting in September 2011.
Since forecasters use data with “noise,”
every time they produce a new forecast we
expect it to be different from the previous
one. But if these changes in the projections
are based on noise and not on news, we
expect no systematic pattern when we accumulate the series. If forecasts are accurate,
we also would expect little noise; so, changes
should be small unless there is news about
the state of the economy.

Although the decline may seem small, it
represents an adjustment of about 50 percent,
given that inflation in advanced economies is
usually about 2 percent. These accumulated
downward revisions of inflation for advanced
economies are in line with news during that
time that generated fear of deflation.
Bad News on Output Growth

Figure 2 shows the same news index but
constructed with the projections of output
growth in the world and advanced economies.
The results are a clear indication of international headwinds since September 2011. Focus
first on world output growth (blue dotted line).
It reached a depth of 4.5 percent in October
2014. To realize how large this number is,
recall that average growth for the years 20002007 was 4.5 percent. Notice that about half of
the decline occurred during the first quarter
considered (fall 2011-winter 2012), when
Europe entered into a recession. But after a few
reports with good news (until July 2012), there
was a very long sequence of negative news,
accounting for the other half of the decline.
The index for advanced economies (red
line) is even more surprising. It declines to
almost 4 percentage points. (Recall that the
The Regional Economist | www.stlouisfed.org 5

average growth for advanced economies
during 2000-2007 was only 2.6 percent.) A
large part of the bad news about world output
growth can be accounted for by the performance of advanced economies. In particular,
as it will be clearer below, international headwinds fly to North America over the North
Atlantic Ocean from Europe.

figure 2
0.0
–0.5
–1.0
–1.5
–2.0
–2.5
–3.0
–3.5
–4.0
–4.5
–5.0

Oct. ’14

July ’14

May ’14

Feb. ’14

Nov. ’13

Sept. ’13

June ’13

March ’13

Jan. ’13

Europe, Again

Oct. ’12

July ’12

May ’12

Feb. ’12

Dec. ’11

Sept. ’11

World Output
Advanced Economies’ Output

June ’11

Cumulative Percentage Revised since Sept. 2011

“News Index,” World and Advanced Economies’ Output Growth

SOURCE: World Economic Outlook.

figure 3
Euro Area GDP
1.5
1.45

14%
2.3% growth

1.35
1.30

2% growth

1.25

GDP Euro Area

1.20

Trend 1995:Q1 to 2008:Q3

1.15

June ’14

June ’13

June ’12

June ’11

June ’09

June ’08

June ’07

June ’06

June ’05

June ’04

June ’03

June ’02

June ’00

June ’01

Trend 2009:Q2 to 2011:Q2

1.10

June ’10

1995:Q1=1

1.40

SOURCE: Organization for Economic Cooperation and Development.

figure 4
Low Inflation in Europe: September 2014 vs. September 2013
2.0
1.5

Percent

1.0

Double Dip or Triple Dip?

0.5
0.0
–0.5

Greece
Hungary
Spain
Poland
Slovenia
Italy
Slovakia
Portugal
Lithuania
Cyprus
Switzerland
Sweden
Belgium
Croatia
Estonia
Netherlands
Denmark
Luxembourg
Iceland
Ireland
Malta
Germany
Czech Republic
Latvia
UK
Austria
Finland
Romania
Norway

–1.0
–1.5

Europe’s economic performance has been
quite poor, with some exceptions, since the
third quarter of 2008. Figure 3 shows euro
area GDP since 2000. As shown above, the
projected growth for 2014 and 2015 is well
below average. Figure 3 shows GDP in the
euro area at constant prices, normalized such
that 1995:Q1 is equal to 1.
Two other series are included in Figure 3.
The first one is the trend growth before the
crisis (from 1995:Q1 to 2008:Q3). Growth
averaged 2 percent during that period. The
other series is the trend for the two years
following the end of the recession (2009:Q2
to 2011:Q2). The growth during that period
was back to trend (actually a bit higher, 2.3
percent vs. 2.0 percent before). This figure
also shows that GDP is 14 percent below the
prefinancial crisis trend and about 7 percent
below the trend from 2009:Q2 to 2011:Q2.
Importantly, the recovery that followed
the 2011-12 recession in the euro area was
very weak, with GDP growing slower than
the two trends discussed above. Thus, the
gap between GDP and its trend (sometimes
referred to as “the output gap”) is not narrowing but expanding.

SOURCE: Haver Analytics.

6 The Regional Economist | January 2015

As shown in Figure 3, Europe had what is
called a double-dip recession. This term is
used to describe a recession that follows right
after another one. Recent data indicate that
Europe may be close to a new recession. This
would be a very rare triple-dip recession.
There exists an index that can be used
to evaluate the performance of Europe in
the month after the last report of GDP.
This index, called EuroCOIN, is a real-time
cyclical indicator for the euro area; it is
constructed using 1,000 macroeconomic
time series from the major countries in the
area. EuroCOIN is available quickly and
at monthly frequency. While the last data
available (at this writing) on GDP correspond

figure 5
Where Will Growth Come from in 2015?
30
25

26%
22%

20
Percent

24%

18%

15
10%

10
5
0

6%

Euro Area

United States
2000-2007

China
2014-2015

SOURCE: World Economic Outlook.

to the second quarter, EuroCOIN is already
available for September 2014. In this last
month, the indicator fell to 0.13 (from 0.19
in August), the lowest level in 12 months.
This indicates that Europe, again, will be the
source of headwinds for U.S. growth during
the coming years.
Monetary Policy in Europe

The European Central Bank (ECB) has
announced a range of actions to bolster
its economy: a reduction in policy rates in
September, targeted credit easing and other
measures to boost liquidity. In particular, the
ECB has declared an objective of expanding
its balance sheet back to early 2012 levels,
implying a 1 trillion euro expansion.
This policy is a response to the poor
projected growth and the low inflation in
Europe. Figure 4 displays the change in
prices (inflation) between September 2013
and September 2014. Only two countries,

Norway and Romania, had inflation close to
2 percent (2.1 and 1.8 percent, respectively).
Several countries had negative inflation. For
instance, Greece’s inflation was –1.1 percent.
The ECB’s actions to bolster its economy
and boost liquidity may include purchases
of member countries’ government bonds.
For most of the countries in the euro area,
the yields for 10-year government bonds
are already very low. They actually declined
abruptly during September, perhaps in
anticipation of this policy. This plan to buy
bonds may improve economic conditions, to
the extent that it helps in reducing financing
costs and in stimulating demand.

ENDNOTES
1

2

3

The World Economic Outlook is available on the
website of the International Monetary Fund at
www.imf.org/external/ns/cs.aspx?id=29.
The main criteria used by the WEO to classify the
world into advanced economies and emerging
market and developing economies are (1) per
capita income level, (2) export diversification—so
oil exporters that have high per capita GDP would
not make the advanced classification because
about 70 percent of their exports are oil, and (3)
degree of integration into the global financial system. The list of countries that are considered to be
advanced economies includes Australia, Austria,
Canada, France, Germany, Italy, Japan, the United
Kingdom and the United States.
For some periods, there is only one observation.
That happens when one of the revisions is zero.

China and the U.S.:
Engines of Growth in 2015

Finally, the conditions described above
for Europe imply that China and the U.S.
will probably account for most of the world’s
growth in 2015. The blue bars in Figure 5
display the share of the growth in output that
was contributed between 2000 and 2007 by
the euro area, the U.S. and China: 26, 18 and
10 percent, respectively. The contributions
for the expected growth between 2014 and
2015 are very different. The euro area will
contribute only 6 percent; the lion’s shares
will be contributed by the U.S. (22 percent)
and China (24 percent).
Juan Sánchez is an economist at the Federal
Reserve Bank of St. Louis. For more on his
work, see http://research.stlouisfed.org/econ/
sanchez.

The Regional Economist | www.stlouisfed.org 7

i n t e r n a t i o n a l

Sovereign Debt Crisis
in Europe Recalls
the Lost Decade
in Latin America
By Maria A. Arias and Paulina Restrepo-Echavarria

T

he recent European debt crisis may
seem like déjà vu. Many of its characteristics are reminiscent of the Latin
American debt crisis of the 1980s, which
led to what is known as the lost decade.
In this article, we explore the similarities
of and point out the differences between
both crises.
Similarities

During the 1970s, Latin America was
experiencing an era of high growth. Output, investment and per capita consumption were surging. The excess liquidity
generated by oil-exporting countries when
oil prices rose and the resulting high savings of those countries facilitated borrowing abroad. This borrowing was supposed
to finance infrastructure projects but
ended up financing consumption.
However, after 1979, an increase in oil
prices by the Organization of the Petroleum Exporting Countries (OPEC) led to
the start of what is known as the Volcker
era. Paul Volcker, then the chairman of
the Federal Reserve, increased interest
rates sharply in order to control inflation
in the U.S. economy, causing payments on
foreign debt to become more expensive
for Latin America, which had borrowed
heavily from U.S. banks. Most Latin
American countries were oil importers at
the time; so, higher prices for imported oil,
combined with the now more expensive
debt, should have generated an adjustment in borrowing and spending. Instead,
debt went from being 30 percent of gross
domestic product (GDP) on average in
1979 to nearly 50 percent in 1982 for
the larger Latin American countries.
(See Figure 1.) This situation became
8 The Regional Economist | January 2015

unsustainable and ended up with Mexico’s
default in 1982, followed soon by the
default of other countries in the region.
The picture is quite similar for peripheral Europe. Greece, Spain, Portugal and
Ireland were getting capital inflows since
the beginning of the 2000s. (See Figure 2.)
These newfound resources were meant to
finance investment. Instead, as in Latin
America, the excess liquidity went to
finance a consumption boom. Debt went
from being 90 percent of GDP on average
in 2000 to 200 percent in 2009 (see Figure
1), right before Greece first requested
financial aid from the International
Monetary Fund. Similarly, debt-to-GDP
ratios soared in Spain, Portugal and
Ireland, which also sought financial support to pay their sovereign debts in the
following years.
So, in both Latin America prior to the
1980s and peripheral Europe at the start
of the 21st century, output, investment
and consumption were growing rapidly.
Liquidity levels were extraordinarily high
and were accompanied by capital inflows
and fast-rising levels of debt to GDP.
Then, an external shock struck, making
the situation unsustainable. Capital flows
reversed (see Figure 2), and many countries defaulted.
What Was Different?

The kind of external shock that triggered each of the crises, the composition of
the debt, the interest rates that the regions
were facing and the relationships among
the countries involved were different.
For Latin America, the external shock
was the hike in U.S. interest rates, which
was a consequence of the rise in oil prices.

For Europe, the Great Recession of 200809 triggered the crisis.
Figure 1 shows the composition of the
debt-to-GDP ratio in both regions. The
solid lines depict total debt, while the
dashed lines show only public debt. Note
that for Latin America, public debt was
driving the increase in total debt, while in
Europe, private debt was actually driving
the increase in total debt.
Figure 3 shows the respective real
interest rate that was being faced by Latin
America and peripheral Europe in the
years preceding and during the crises.1
As discussed above, it is clear that Latin
America was enjoying low interest rates
prior to the crisis and then experienced a
sharp rise in rates at the beginning of the
1980s. Europe, on the other hand, faced
much higher interest rates from the start;
those interest rates continuously decreased
over time.
Figures 1 and 3 portray the main differences between both crises. The jump in
interest rates for Latin America made the
debt very expensive to repay, leading
countries to default by failing to service
their debt. The relatively high level of debt
held by countries in Europe’s periphery
simply became unsustainable when
output, investment and per capita consumption started to decline after the
Great Recession.
The Aftermath

The Latin American debt crisis resulted
in the well-known lost decade for the
region, during which initial fiscal readjustments and austerity did little but reinforce
anemic growth. Currency devaluation, an
emphasis on trade expansion (see Figure 2)

FIGURE 1

ENDNOTE

Total Debt and Public Debt by Region
Total Debt (Latin America, left)

60

190

Public Debt (Latin America, left)

50
Percent of GDP

210

170

Total Debt (Europe, right)

40

150

Public Debt (Europe, right)

30

130

20

110

10

90

0
1972
1998

1974
2000

1976
2002

1978
2004

1980
2006

1982
2008

1984
2010

1986
2012

Percent of GDP

70

1

70
Latin Am.
Europe

SOURCES: International Monetary Fund, Penn World Tables, Reinhart and Rogoff, and author’s calculations.

FIGURE 2

FIGURE 3

Net Exports by Region

Real (Inflation Adjusted) International
Lending Rate

8
6

Europe

4

r e f e r en c es
Eichengreen, Barry. “Latin Lessons for the Euro Zone.”
Manuscript, University of California, Berkley,
January 2010.
European Central Bank. “ECB Assumes Responsibility
for Euro Area Banking Supervision.” Press release.
Nov. 4, 2014. See www.ecb.europa.eu/press/pr/
date/2014/html/pr141104.en.html.
Reinhart, Carmen M.; and Rogoff, Kenneth S. This
Time Is Different: Eight Centuries of Financial Folly.
New Jersey: Princeton University Press, 2009.

Latin America

8

Europe

6
Percent

Percent of GDP

10

Latin America

The interest rate for Latin America corresponds to
the rate posted by a majority of the top 25 insured
U.S.-chartered commercial banks (by assets in
domestic offices), while that for Europe is the
average of the respective rates posted by banks in
Germany, France and Great Britain—the primary
lenders for the peripheral European countries. The
real lending interest rate is calculated using the
GDP deflator.

2

4

0

2

–2

0

–4
’72
’98

’74
’00

’76
’02

L.A.=Latin America

’78
’04

’80
’06

’82
’08

’84
’10

’86 L.A.
’12 E.

E.=Europe

SOURCE: World Bank

–2
’72
’98

’74
’00

’76
’02

L.A.=Latin America

’78
’04

’80
’06

’82
’08

’84
’10

’86 L.A.
’12 E.

E.=Europe

SOURCE: International Monetary Fund

NOTE: Net exports can be used as a proxy to reflect net capital flows, whereby
negative net exports represent a positive capital inflow to the region.

NOTE FOR ALL FIGURES: Data in each figure represent aggregate averages for select Latin American countries between 1972
and 1987 and for select European countries between 1998 and 2013. The Latin American countries in the sample are Argentina, Brazil, Chile and Mexico, while the European countries in the sample are Greece, Ireland, Italy, Portugal and Spain. The
gray vertical bar represents the start of the debt crisis in both regions: 1982 in Latin America and 2008 in Europe.

and eventually debt restructuring through
what was known as the Brady Plan helped
the countries in the region regain strength
and return to economic growth.
As for the situation in Europe, being
part of the Economic and Monetary Union
(EMU) without having broader fiscal
integration limited what policies could be
implemented by the individual countries
to jump-start the economy after the crisis.
Austerity measures and some debt restructuring have been part of monetary authorities’ response. But the European Union has
moved to integrate even further by creating

joint supervisory authorities and a closer fiscal union, including the most recent banking union, which took effect Nov. 4, 2014.
Ultimately, more unified coordination and
governance could strengthen the fiscal union
and lead to greater economic stability.
Paulina Restrepo-Echavarria is an economist
and Maria A. Arias is a research associate,
both at the Federal Reserve Bank of St. Louis.
For more on Restrepo-Echavarria’s work, see
http://research.stlouisfed.org/econ/restrepoechavarria.
The Regional Economist | www.stlouisfed.org 9

m e t h o d o l o g y

Measuring Underground
Economy Can Be Done,
but It Is Difficult
By Paulina Restrepo-Echavarria
© thinkstock

T

he informal economy, also known as
the underground economy or black
market, is very hard to measure. A good
example is the produce vendor on the street
who sells the same vegetables you find in the
supermarket but handles only cash and pays
little or no taxes. Nevertheless, this sector
adds considerable value to the economy. In
developing countries, the informal sector
has been estimated to account for about 36
percent of gross domestic product (GDP). In
developed countries, it has been estimated
to be about 13 percent of GDP.1 (See table.)
So how do economists measure the informal
sector? This article explains the two main
approaches—direct and indirect—and the
difficulties that each entails.
Average Size of the Informal Economy (% of GDP, 2002-2003)
Developed Countries

13%

Developing Countries

36%

SOURCE: Schneider.

Direct Approaches

These methods rely on surveys, samples
based on voluntary replies, tax audits and
other compliance methods. The problem is
that the results depend directly on the questions asked by the survey, and few surveys
are alike. As a result, it is very difficult to
use the same parameters to measure and
compare the informal economy in different countries.
Usually, what ends up happening is that
the definition that is used has to be very
simple and contain only one parameter. For
example, the informal sector may be defined
as those people who do not have the right to a
pension when they retire. Clearly, this definition excludes several important elements
10 The Regional Economist | January 2015

that would describe the informal economy
differently. Another very common definition
is that people are considered to work in the
informal economy if they work for a firm that
has N or fewer workers. But a firm can be
very small and still comply with the law, and
its production can be reported to the authorities, meaning that its value added will appear
in the GDP despite being a small firm.
If what is used is a direct questionnaire,
people are not usually willing to admit that
they are not reporting taxes or that they
are engaging in fraudulent behavior, either
because they feel afraid of getting caught or
because they feel ashamed since they know
this is a moral issue. This makes it difficult
to estimate the extent of undeclared work.
Finally, a direct estimate of the informal
economy can also be obtained by calculating the discrepancy between income
declared for tax purposes and that measured
by selective checks. For example, one can
compare the number of jobs declared by
firms with the number of employed people
found through household surveys. The
number of employed people exceeding the
number of jobs represents the informal
workforce. Once the informal number of
workers is identified, informal workers can
be attributed the same net compensation as
similar workers in the formal economy.2
Indirect Approaches

These are macroeconomic approaches
that try to use an indicator of the informal
economy as a proxy for its size or growth.
Discrepancy between the National
Expenditure and Income Statistics

In theory, the income measure of GDP
and the expenditure measure should be

equal to each other. However, informal
activities can show up in the expenditure measurement but not in the income
measurement. This is because the income
side is measured through the value added
of registered firms (the formal economy),
while on the expenditure side there is some
self-reporting. Thus, the difference between
these two measures is an indicator of the
size of the informal economy. The problem
with this estimate is that statisticians would
like to make the difference between the two
as small as possible; so, using the initial
measure rather than the published measure would be ideal.3 Moreover, there are
differences due to sampling and statistical
errors, which cannot be disentangled from
the amount that can be explained by the
informal economy.
Discrepancy between Official
and Actual Labor Force

Assuming that the total labor force
participation is constant, all else being the
same, then any decrease in the labor force
participation in the official economy can
be seen as an indicator of an increase in
the activity in the informal economy.4 The
problem with this method is that changes
in labor force participation can be due to
other causes. For example, following the
recent recession, many people have exited
the labor force. It could also be the case
that people work in both the informal and
formal economy; so, this is not a very good
estimator.
The Transactions Approach

In 1979, economist Edgar Feige developed
this approach based on the quantitative theory of money MV = pT, where M is money,

V is velocity, p is prices and T is total transactions. The main assumption is that the
relationship of the volume of transactions
and official gross national product (GNP) is
constant over time.5 Using the value of total
transactions (pT) as an estimate of nominal
GNP, he calculated the informal economy
as the difference between nominal GNP and
the official GNP. Several issues arise with
this approach. He had to assume there is a
base year when there was no informal economy. Then, the assumption that the ratio of
transactions to official GNP is constant over
time was quite strong. Additionally, obtaining accurate estimates of the total number of
transactions was difficult.
The Currency Demand Approach

This approach uses the correlation between
currency demand and tax pressure, assuming
that informal activities operate with cash.6
Thus, if the tax burden increases and so does
the demand for money, then that increase in
the demand for money reflects an increase in
the informal economy.
In order to calculate the excess in money
demand, the economists behind this
approach estimated an equation for money
demand using econometric methods. They
controlled for development of income,
payment habits, interest rates and other
related variables. In the equation, they also
included government regulation, direct and
indirect tax burden, and the complexity of
the tax system. The most common critiques
to this approach are the following:
• Not all the transactions in the shadow
economy are paid in cash.
• Most studies using this approach include
only the tax burden factor and ignore
others, such as “tax morality,” regulation
and attitudes toward the state. (There are
usually no reliable data on these factors.)
• A rise in currency demand deposits is usually due in large degree to a slowdown in
demand deposits and not to a rise in currency due to informal economic activity.
• Also, most studies assume that both the
formal and informal economy have the
same velocity of money.7

of both formal and informal economic
activity. It has been observed that the
electricity/GDP elasticity is usually close
to 1. 8 So, by using electricity as a proxy
for the overall economic activity and then
subtracting from it the official estimates
of GDP, we get an indicator of informal
economic activity. The difference between
the growth of electricity consumption and
official GDP is then attributed to the growth
of the informal economy.
The critiques to this approach rely on the
fact that not all informal activities require a
considerable amount of electricity, or, if they
do, other energy sources such as gas, oil and
coal could be used. Also, the use of electricity has become more and more efficient in
both types of economies. Finally, there may
be differences in the elasticity of electricity/
GDP across countries or changes over time.
Ultimately, the approach used to measure the informal economy depends on
the specific question being asked by the
researcher. For macroeconomic studies,
indirect approaches usually suffice, but
direct approaches are more generally used
for microeconomic studies. Newer methods
being developed to better gauge the size of
the informal economy involve more-technical, model-based estimations.
Paulina Restrepo-Echavarria is an economist at
the Federal Reserve Bank of St. Louis. For more
on her work, see http://research.stlouisfed.org/
econ/restrepo-echavarria.

ENDNOTES
1 See Restrepo-Echavarria.
2 This is the approach used in Italy. See Bovi.
3 There is usually some degree of statistical dis-

4
5

6

7

8

crepancy between the income and expenditure
measures because of how the data are constructed.
The initial estimations, before the data are revised
to sort the majority of this discrepancy and balance both sides of the equation, are not usually
published. Only the final measures are published,
once the discrepancy is accounted for.
The labor force participation rate is calculated as the
labor force divided by the working-age population.
GNP is often used to estimate total transactions
as it also includes national currency transactions
that originate in other countries, whereas GDP is
a measure of transactions only within the particular country.
This approach was first proposed by Cagan, and
then Tanzi took the method a step further. See
Cagan, as well as Tanzi.
The velocity of money is the rate at which money
circulates in the economy or the rate at which
people spend money.
The electricity/GDP elasticity is a measure of how
sensitive GDP growth is to changes in electricity
consumption. If the absolute value of the elasticity
is greater than 1, a larger change in electricity consumption is needed to achieve a 1 percent change
in GDP; if the elasticity is less than 1, a smaller
change in electricity consumption is needed to
achieve a 1 percent change in GDP; if the elasticity
is equal to 1, a 1 percent change in electricity
consumption is associated with a 1 percent change
in GDP.

Re f e r en c es
Bovi, Maurizio. “Shadow Employment and Labor
Productivity Dynamics.” Labour, December 2007,
Vol. 21, No. 4-5, pp. 735-61.
Cagan, Phillip. “The Demand for Currency Relative
to the Total Money Supply.” Journal of Political
Economy, August 1958, Vol. 66, No. 4, pp. 302-28.
Feige, Edgar L. “How Big Is the Irregular Economy?”
Challenge, November/December 1979, Vol. 22,
No. 5, pp. 5-13.
Restrepo-Echavarria, Paulina. “Macroeconomic
Volatility: The Role of the Informal Economy.”
European Economic Review, August 2014, Vol. 70,
pp. 454-69.
Schneider, Friedrich. “Shadow Economies and
Corruption All Over the World: New Estimates
for 145 Countries.” Economics: The Open-Access,
Open-Assessment E-Journal, July 2007, Vol. 1,
No. 2007-9, pp. 1-66.
Tanzi, Vito. “The Underground Economy in the
United States: Estimates and Implications.”
Banca Nazionnale del Lavoro Quarterly Review,
December 1980, Vol. 33, No. 135, pp. 427-53.
Tanzi, Vito. “The Underground Economy in the
United States: Estimates and Implications.” Staff
Papers—International Monetary Fund, 1983,
Vol. 30, No. 2, pp. 283-305.

The Physical Input (Electricity
Consumption) Method

This method assumes that electricity
consumption is the best physical indicator
The Regional Economist | www.stlouisfed.org 11

work

Youth Labor Force
Participation Continues
To Fall, but It Might Be
for a Good Reason
By Maria Canon, Marianna Kudlyak and Yang Liu
© thinkstock

T

larger fraction of young people are attending
school today than in the 1980s or 1990s, then
the currently low labor force participation
rate of youth might signal good news, implying a more-skilled prime-working-age labor
force and possibly higher aggregate LFP rates
in the future. On the other hand, if young
potential workers are neither in the labor
force nor in school, incorporating them into
the labor force in the future might not be an
easy task.
In this article, we review the trends in youth
labor force participation by age, gender and
education, focusing on the distinction between
those in school and those not in school.

FIGURE 1

FIGURE 2

Labor Force Participation Rate: 1955-2014

Labor Force Participation Rate

Less Education=Bigger Decline

There are two distinct age groups among
these youth: those between 16 and 19 years
old and those between 20 and 24 years old. A
large share of the first group is transitioning

80

66

70

62

60

Aged 16-24

Aged 16 and Older

Male 16-24 Years Old
16-19 Years Old

2014

2012

2010

2008

2006

2004

Jan. ’10

Jan. ’05

Jan. ’00

Jan. ’95

Jan. ’90

Jan. ’85

Jan. ’80

Jan. ’75

30
Jan. ’70

50
Jan. ’65

40

Jan. ’60

54

2002

50

2000

58

1998

Percent

70

Jan. ’55

Percent

he aggregate labor force participation
(LFP) rate measures the share of the
civilian, noninstitutionalized population
(16 years and older) that is either employed or
nonemployed but looking for work. The LFP
rate reached its peak of 67.1 percent in 2000
and has been declining since, accelerating
during the Great Recession.
Workers between 16 and 24 years of age
constitute the demographic group that has
experienced one of the most substantial
declines in labor force participation. Figure 1
shows participation rates for these youth
since 1955. The LFP rate for this group
increased more or less steadily until 1979,
reaching 68.8 percent in September 1979,
then remained above 65 until 2000 before
starting its sharp decline.1 The rate was down
to 54.9 percent in September 2014. Was the
decline homogeneous across different subsets
of youth? The question is important: If a

Female 16-24 Years Old
20-24 Years Old

SOURCES: Organization for Economic Cooperation and Development (OECD)
and Federal Reserve Economic Data (FRED).

SOURCES: U.S. Bureau of Labor Statistics (BLS) and National Bureau of
Economic Research (NBER).

NOTE: The gray bars represent recessions. The final data point is from
September 2014.

NOTE: The data start in 1998 for this and the remaining figures because that’s
the earliest year for which the needed microdata from the Current Population
Survey are available. The 2014 numbers are calculated using the average of
the numbers from January through September.

12 The Regional Economist | January 2015

from high school to college; thus, one should
expect low labor force participation rates for
this group. Workers between 20 and 24 years
old are, instead, transitioning from college to
either graduate school or to the labor market;
thus, one should expect the LFP rate of this
group to be closer to the LFP rate of the
prime working-age population.
As can be seen in Figure 2, the 16-19 group
experienced a large decline from 1998 until
2014 in the LFP rate, from 52.8 percent to 34.2
percent, a decrease of 35.2 percent. For the
20-24 group, the LFP rate declined from 77.5
percent to 71 percent, a decrease of 8.4 percent.
The decline in the LFP rate was similar
for men and women, 17.1 percent for men
and 15.2 percent for women for the entire
16-24 group.
Figure 3 shows that the decline of youth
labor force participation was not homogeneous across education groups. Those between
16 and 24 with less than a high school diploma
experienced the largest decline in the LFP rate:
from 50.3 percent in 1998 to 29.8 percent in
2014, a 40.8 percent decrease. This decline was
primarily driven by people 16-19; their LFP
rate declined by 45.4 percent. Young people
with at most a high school diploma also experienced a significant decline of their LFP rate,
from 78.2 percent to 68.4 percent. The 16-19
group drove this decline, as well. Those with
some college experienced a decline similar to
that of high school graduates. Finally, young
workers with at least a college degree did not
experience significant changes in their LFP
rate; it decreased from 84.5 percent in 1998 to
82.4 percent in 2014.
In School or Not in School?

Because of their ages, many of those not in
the labor force (nonparticipants) are expected

education groups. (See Figure 5.) Although
the NEET fraction for those with less than a
high school diploma decreased by about
1 percentage point between 1998 and 2014,
the NEET fraction increased significantly for
the population with a high school diploma
(in particular for the 16-19 group, whose
percentage rose from 8.2 percent in 1998 to
12.4 percent in 2014) and for the population
with some college education.
Interestingly, the 16-24 population with
at most a high school diploma has some
noticeable incidence of disability. This is not
observed for the young population with at
least some college.
In conclusion, the data from the Current
Population Survey show that since 1998 most
of the decline in youth labor force participation corresponds to an increase in school
attendance (in particular of the 16-19 population). The fraction of the NEET population
did not change significantly over this period,
but within education groups the trends have
been different. A more-detailed study of
these labor trends among youth is needed to
understand the future incorporation of these
people into the labor market.

FIGURE 3

FIGURE 4

FIGURE 5

Labor Force Participation Rate

In School to Population Ratio: 16-24

NEET to Population Ratio: 16-24

8

No High School Diploma
High School Diploma
Some College
Bachelor’s and Above

2014

2012

2010

0

2008

2014

2012

2010

2008

2006

2004

2002

4

2000

1998

0

2014

20

2012

10
2010

30
2008

20

2006

12

30

40

NEET: Not in Education, Employment or Training

2006

40

16

2004

60

2004

Canon, Maria; Debbaut, Peter; and Kudlyak, Marianna. “A Closer Look at the Decline in the Labor
Force Participation Rate.” Federal Reserve Bank
of St. Louis’ The Regional Economist, October
2013, pp. 10-11. See https://www.stlouisfed.org/
publications/regional-economist/october-2013/acloser-look-at-the-decline-in-the-labor-forceparticipation-rate.

2002

50

2002

Re f e r en c e

2000

70

2000

on the decline of the aggregate labor force participation rate.
2 Being retired is one of the options given to everyone who is asked in the Current Population Survey
why he or she is are not working. Fewer than 0.03
percent of young people pick this option.

1998

60

16-19 No High School Diploma
16-24 No High School Diploma
16-24 High School Diploma
16-24 Some College
16-24 Bachelor’s and Above

1 See Canon, Debbaut and Kudlyak for an analysis

Percent

80

Percent

70

50

ENDNOTES

Maria Canon is an economist at the Federal
Reserve Bank of St. Louis. Marianna Kudlyak
is an economist at the Federal Reserve Bank
of Richmond. Yang Liu is a senior research
associate at the Federal Reserve Bank of
St. Louis. For more on Canon’s work, see
http://research.stlouisfed.org/econ/canon.

90

1998

Percent

to be still in school. But are they? In the
Current Population Survey, conducted by
the Census Bureau for the Bureau of Labor
Statistics, a young individual who is out of
the labor force can, in principle, be classified
as either in school, not in school, disabled or
retired.2 A young individual who is not in
school of any kind and not working is often
referred to by the acronym NEET: Not in
Education, Employment or Training.
The decline in youth labor force participation corresponds to a higher fraction of them
attending school. Figure 4 shows that school
attendance for the 16-24 population without
a high school diploma increased from 38
percent in 1998 to 60 percent in 2014. This
increase was driven by the younger population. While 39.8 percent of those 16-19
were attending school in 1998, 58.5 percent
of them were attending school in 2014.
(School attendance for the 20-24 population
increased significantly less, going from 11.6
percent in 1998 to 17.4 percent in 2014.)
The second most-prevalent reason for not
being in the labor force was NEET. Among
those in the 16-19 group, 6.1 percent were
NEET in 2014; for those in the 20-24 group,
9.2 percent were NEET in 2014. Lowereducated individuals (those with a high
school diploma at most) were more likely to
be NEET.
The fraction of NEET did not change
significantly from 1998 until 2014 for the
entire 16-24 population. But its relative
stability masks heterogeneous trends across

No High School Diploma
High School Diploma
Some College
Bachelor’s and Above

SOURCES: BLS and NBER.

SOURCES: BLS and NBER.

SOURCES: BLS and NBER.

NOTE: The 2014 numbers are calculated using the average of the numbers
from January through September.

NOTE: The 2014 numbers are calculated using the average of the numbers
from January through September.

NOTE: The 2014 numbers are calculated using the average of the numbers
from January through September.
The Regional Economist | www.stlouisfed.org 13

w o r k

Assessing the Health
of the Labor Market:
The Unemployment Rate
vs. Other Indicators
By Maximiliano Dvorkin
© Thinkstock

O

ne of the goals of the Federal Reserve
System, particularly of the Federal
Open Market Committee (FOMC), is to
achieve maximum employment. Therefore,
staff and officials across the System put great
effort into analyzing the current conditions
of the labor market. Unfortunately, there is
no widespread consensus on the definition of
maximum employment or how far the economy is from it. Until recently, the unemployment rate has been the hallmark indicator of
labor market health;1 even Fed Chair Janet
Yellen argued in 2013 that “the unemploy-

It has been argued that the
current level of the unemployment rate may not be capturing all the relevant information
about the health of the labor
market.
ment rate is probably the best single indicator
of current labor market conditions.” 2
Currently, the unemployment rate stands
very close to its natural level, indicating that
the labor market has returned to some semblance of normal.3 However, unemployment
may fall for reasons other than improved
economic conditions. For example, it may fall
when unemployed workers become discouraged and stop looking for work; then, they
are no longer being counted as a part of the
labor force.
In much of the recent policy debate, it
has been argued that the current level of the
unemployment rate may not be capturing
all the relevant information about the health
of the labor market and that it is best to look
14 The Regional Economist | January 2015

at a broad range of labor market indicators.
Most notably, attention has shifted to variables like labor force participation, involuntary part-time employment and long-term
unemployment.
Summarizing all this data is not a simple
task since these different variables do not
always move in tandem. For this reason,
economists have developed several tools to
distill key information that might be common to many of these observed variables
(unemployment rate, labor force participation rate and dozens of other indicators) and
might be driving those variables. In particular, the Federal Reserve Board of Governors
and the Federal Reserve Bank of Kansas City
introduced in 2014 indexes that they developed on labor market conditions.4 The goal
of these indexes is to get a handle on “labor
market conditions” or “labor market health”;
these are terms that are not precisely defined
and are even harder to measure.
Although these tools are indexes, much
like the Consumer Price Index or the Industrial Production Index, the labor market conditions indexes use relatively sophisticated
statistical procedures to weight some labor
market variables more heavily than others.
These statistical procedures let the data determine which indicators are more informative
of the movements in the underlying labor
market conditions.
In this article, I compare and contrast the
labor market conditions indexes with one
another and with the unemployment rate to
see what labor market insights can be gained.
I found a couple of things. First, despite
some differences in their construction and
the variables used, the three indexes that
I reviewed seem to provide essentially the
same information. This similarity is not

surprising since they are, after all, trying to
capture the same object, namely the general health of the labor market. Second, the
indexes have a strong negative correlation
with the unemployment rate, that is, a rising
index is associated with a falling unemployment rate. This strong link confirms that the
unemployment rate is a reliable proxy for
unobserved labor market health.
Measuring Labor Market Conditions

The goal of these indexes is to distill the
information from a large set of observed
labor market variables using a statistical
model. Once the final index is calculated,
the levels are interpreted as relative labor
market conditions. A level higher than zero
indicates that labor market conditions are
above the historical average, while a level
below zero indicates that labor market conditions are relatively poor compared with
historical averages.
The index developed by the Federal
Reserve Board of Governors (FRB) uses 19
labor variables of the U.S. economy. These
variables are measured monthly; the sample
starts in July 1976. The index is reported
in average monthly changes instead of the
index levels.
The Federal Reserve Bank of Kansas City
has developed two indexes on labor market conditions; it uses 24 variables with a
monthly frequency, and the sample starts in
January 1992. The first index is interpreted as
the level of conditions in the labor market; the second reflects the momentum, or
changes, in these conditions.
Figure 1 shows the evolution of the changes
in labor market conditions as captured by
the different indexes, and Figure 2 shows the
same evolution for the level of labor market

ENDNOTES

FIGURE 1
Changes in Labor Market Conditions Indexes (LMCI) and in the Unemployment Rate

0

0.0

–10

0.1

–20

0.2

–30

0.3

–40

0.4

–50

0.5

2

–0.2

1

–0.1

0

0.0

–1

0.1

–2

0.2

–3

0.3

–4

0.4

–5

0.5

Change in Unemployment Rate (inverted)

–0.1

1992:05
1994:05
1996:05
1998:05
2000:05
2002:05
2004:05
2006:05
2008:05
2010:05
2012:05
2014:05

–0.2

10

Change in Index Values

20

Change in Unemployment Rate (inverted)

k ansas cit y fed ’ s inde x

1992:05
1994:05
1996:05
1998:05
2000:05
2002:05
2004:05
2006:05
2008:05
2010:05
2012:05
2014:05

Change in Index Values

federal reser v e board ’ s inde x

1 The unemployment rate is defined as the ratio of

Date
Recession
KC Fed level (standardized and smoothed changes, left)
KC Fed momentum (left)
Unemployment Rate (smoothed changes, right)

Date
Recession
FRB’s LMCI (changes, left)
Unemployment Rate (smoothed changes, right)

2

3

4
5
6

SOURCES: Author’s calculations using data from the Federal Reserve Board of Governors, the Federal Reserve Bank of Kansas City and the Bureau of Labor Statistics.
7

FIGURE 2
Levels of Labor Market Conditions Indexes and the Unemployment Rate

Re f e r en c es

k ansas cit y fed ’ s inde x

2

6
7
8
9
10
11

Date
Recession
FRB’s LMCI (implied level, standardized, left)
Unemployment Rate (right)

2
3
4
5
6
7
8
9
10

1992:05
1994:05
1996:05
1998:05
2000:05
2002:05
2004:05
2006:05
2008:05
2010:05
2012:05
2014:05

5

Index Values

4

Unemployment Rate (inverted)

3

2.5
2.0
1.5
1.0
0.5
0.0
–0.5
–1.5
–1.0
–2.0
–2.5
–3.0

Unemployment Rate (inverted)

2.0
1.5
1.0
0.5
0.0
–0.5
–1.0
–1.5
–2.0
–2.5
–3.0
1992:05
1994:05
1996:05
1998:05
2000:05
2002:05
2004:05
2006:05
2008:05
2010:05
2012:05
2014:05

Index Values

federal reser v e board ’ s inde x

people actively looking for work to the sum of
people actively looking for work and those currently employed.
In 2013, Janet Yellen was vice chair of the Federal
Reserve. For more information on her remarks on
the unemployment rate and its role in monetary
policy, see Yellen.
Under this view, the concept of natural rate of
unemployment, the rate that will prevail in the
long run in the absence of short-term cyclical
factors, can provide adequate information on the
level of maximum employment. The Congressional
Budget Office most recently estimated the natural
rate of unemployment to be between 5.5 percent
and 5.8 percent in 2014.
See Hakkio and Willis, as well as Chung et al.
I call these implied levels.
To eliminate some of the very high frequency volatility on the monthly changes in the unemployment rate, I take a seven-month centered moving
average on these changes with equal weights,
which is what I present in the graph. At the end of
the period, this average contains only current and
past values.
In statistical lexicon, the R 2 for these regressions is
close to 1.

11

Chung, Hess; Fallick, Bruce; Nekarda, Christopher;
and Ratner, David. “Assessing the Change in Labor
Market Conditions.” FEDS Notes, May 22, 2014.
See www.federalreserve.gov/econresdata/notes/
feds-notes/2014/assessing-the-change-in-labormarket-conditions-20140522.html.
Hakkio, Craig S.; and Willis, Jonathan L. “Kansas
City Fed’s Labor Market Conditions Indicators
(LMCI).” The Macro Bulletin, Aug. 28, 2014. See
http://kansascityfed.org/publicat/research/macrobulletins/mb14Willis-Hakkio0828.pdf.
Yellen, Janet L. Speech at the 2013 National Association for Business Economics Policy Conference.
See www.federalreserve.gov/newsevents/speech/
yellen20130302a.htm.

Date
Recession
KC Fed’s LMCI level (left)
KC Fed’s LMCI momentum (implied level, standardized, left)
Unemployment Rate (right)

SOURCES: Author’s calculations using data from the Federal Reserve Board of Governors, the Federal Reserve Bank of Kansas City and the Bureau of Labor Statistics.

conditions. For the FRB’s index and the KC
Fed’s momentum index, I recovered the levels
from the reported changes.5 As illustrated
by the figures, the information that these
three different measures provide is remarkably similar, which is not surprising since the
statistical method to construct the different
indexes is the same and they employ similar
labor market variables.
In each of these figures, I also plotted the
unemployment rate for the same period.
Since in Figure 1 I analyzed the changes in
the indexes, I also plotted the changes in
the unemployment rate, while in Figure 2
I plotted the levels.6 In periods of expansion,

labor market conditions, as captured by the
different indexes, improve and the unemployment rate falls. The opposite happens
in downturns. To ease the comparison
between the indexes and the unemployment
rate, I inverted the axis for the unemployment rate in the figures.
It is evident from the figures that the
unemployment rate and the indexes are
highly synchronized. While nothing in the
statistical procedure behind the indexes
imposes this strong link between them and
the unemployment rate, the data suggest that
continued on Page 16
The Regional Economist | www.stlouisfed.org 15

e c o n o my

RE A L G D P GR O W T H

16 The Regional Economist | January 2015

4
2
0
–2
–4
–6
–8
–10

Q3
’09

’10

’11

’12

’13

PERCENT CHANGE FROM A YEAR EARLIER

6

6

CPI–All Items
All Items, Less Food and Energy

3

0

–3

’14

December

’09

’10

’11

’12

’13

’14

NOTE: Each bar is a one-quarter growth rate (annualized);
the red line is the 10-year growth rate.

I N F L A T I O N - I N D E X E D T RE A S UR Y Y IE L D S P RE A D S

RATES ON FEDERAL FUNDS FUTURES ON SELECTED DATES

3.00
2.75
2.50
PERCENT

PERCENT

2.25
2.00
1.75

5-Year

1.50

10-Year

1.25

20-Year

1.00

’11

’12

Jan. 9

’13

’14

’15

0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

09/17/14

C I V I L I A N U N E M P L O Y M E N T R AT E

12/17/14

10/29/14

1st-Expiring
Contract

NOTE: Weekly data.

3-Month

6-Month

01/14/15

12-Month

CONTRACT SETTLEMENT MONTH

I N T ERE S T R A T E S
4

11

10-Year Treasury

10
3

8

PERCENT

PERCENT

9

7
6

2

1

Fed Funds Target
1-Year Treasury

5
4
’09

December

’10

’11

’12

’13

0

’14

’09

’10

’11

’12

’13

December

’14

NOTE: On Dec. 16, 2008, the FOMC set a target range for
the federal funds rate of 0 to 0.25 percent. The observations
plotted since then are the midpoint of the range (0.125 percent).

U . S . A GRI C U L T UR A L T R A D E
90

AVERAGE LAND VALUES ACROSS THE EIGHTH DISTRICT

7,000

Exports

6,000

75
BILLIONS OF DOLLARS

Maximiliano Dvorkin is an economist at the
Federal Reserve Bank of St. Louis. Hannah
Shell, a research analyst at the Bank, provided
research assistance. For more on Dvorkin’s
work, see http://research.stlouisfed.org/econ/
dvorkin.

C O N S U M ER P RI C E I N D E X ( C P I )

8

Conclusion

The U.S. economy has recently experienced the largest economic downturn in
postwar history. Five years have passed since
the official end of the recession, yet the difficult question on how far we are from full
employment remains.
With unemployment returning to normal
levels, it has been argued that the unemployment rate may not properly capture the
current amount of slack in the economy; as
a result, labor market conditions indexes
have been proposed as a new measure of
labor market health. These indexes have
the advantage of summarizing information
from many different variables. At the same
time, they are the result of a statistical procedure requiring several steps to compute
and a nontrivial amount of judgment.
In this article, I showed that the unemployment rate is reflective of underlying labor
market health, as represented by the indexes.
In addition, a closer inspection of the figures
suggests that this strong link between the
indexes and the unemployment rate does
not appear to have changed recently, which
suggests that the unemployment rate is still as
good at measuring labor market conditions as
it has been in the past.

g l a n c e

Imports

60

DOLLARS PER ACRE

the unemployment rate is very informative
of the underlying conditions in the labor
market. In fact, a simple linear regression
between the unemployment rate and the different indexes suggests that the bulk of the
variability of the unemployment rate is due
to movements in the indexes.7
While labor market conditions are not
directly observed, the previous results lead
to an important conclusion. If we were
to use only the unemployment rate, or its
changes, to predict the conditions in the
labor market, the prediction error would be
small. In other words, the unemployment
rate has a very high signal-to-noise ratio for
measuring labor market conditions.

a

Eleven more charts are available on the web version of this issue. Among the areas they cover are agriculture, commercial
banking, housing permits, income and jobs. Much of the data are specific to the Eighth District. To see these charts, go to
www.stlouisfed.org/economyataglance.

PERCENT

continued from Page 15

a t

45
30
15

Trade Balance

0
’09

’10

’11

’12

’13

NOTE: Data are aggregated over the past 12 months.

Quality Farmland

Ranchland or Pastureland

5,000
4,000
3,000
2,000
1,000

November

’14

0

2013:Q3 2013:Q4 2014:Q1 2014:Q2 2014:Q3
SOURCE: Agricultural Finance Monitor.

o v e r v i e w

Pace of Growth
Is Expected
To Quicken

The FOMC’s December 2014 Economic Projections for 2014-2016

By Kevin L. Kliesen

Percent

n a t i o n a l

D

espite an early stumble in 2014, the
U.S. economy has performed well over
the past year. Job growth has been strong,
inflation expectations remain low and financial market stresses are lower than average.
Prevailing economic conditions at the end
of 2014 suggest that the probability of faster
real GDP growth and continued low inflation in 2015 outweighs the probability of
slower growth and higher inflation.
A Look Back

Forecasters and policymakers were optimistic in late 2013 about the U.S. economy’s
prospects for 2014. Over the second half of
2013, growth of real gross domestic product
(GDP) had increased at a 4 percent annual
rate. In addition, the unemployment rate
had fallen to 6.7 percent in December 2013,
and inflation remained unusually low
(1.2 percent). The consensus of private forecasters and the majority of the Federal Open
Market Committee (FOMC) was that real
GDP would increase by about 3 percent in
2014 and that inflation would remain below
the FOMC’s target of 2 percent.
However, the economy stumbled coming out of the gate, as real GDP fell at a 2.1
percent annual rate in the first quarter of
2014. Although some were alarmed by this
development, most viewed the unexpected
decline in economic activity as a temporary setback, influenced in part by adverse
weather. Indeed, over the remainder of the
year, the stock market would reach record
highs, measures of business and consumer
confidence would reach multiyear highs,
and the unemployment rate would fall
below 6 percent—much faster than most
forecasters had anticipated. Importantly,
inflation and interest rates would remain
quite low and stable.
In short, following the first-quarter hiccup,
the economy began developing some significant
forward momentum in the spring: Growth of
real GDP measured 4.6 percent in the second
quarter and 5 percent in the third quarter.
The pace of economic activity is expected to

8

2014
2015

5.8

6

5.3

2016

5.1

4
2.4

2.8

2.8
1.9

2

1.3

0

Real GDP Growth

Unemployment Rate

1.3

PCE Inflation

NOTE: Projections are the mid-points of the central tendencies. The projections for real GDP growth and inflation are the percentage change
from the fourth quarter of the previous year to the fourth quarter of the indicated year. Inflation is measured by the personal consumption
expenditures (PCE) chain-price index. The projection for the unemployment rate is the average for the fourth quarter of the year indicated.

slow modestly in the fourth quarter to about
2.75 percent. Overall, real GDP growth is
expected to be about 2.5 percent in 2014 and
accelerate to about 3 percent this year.
A Look Ahead

When forecasting the macroeconomy, it is
important to consider how developments in
2014 could affect the U.S. economy’s performance in 2015. Some of these developments
have been positive, and some have been negative. Two potential negative developments
stand out: concerns about the economic
health of the global economy and an increase
in the trade-weighted value of the dollar.
At the conclusion of the European Central
Bank’s meeting Dec. 4, bank President Mario
Draghi reported that the staff’s forecast for
European real GDP growth in 2015 had
been revised “substantially downward” to
1 percent. But Europe’s slow growth is not an
outlier. Following a tax hike in the first quarter of 2014, Japan slipped into a recession;
the country is expected to grow by less than
1 percent in 2015. Growth in Asia’s other large
economy, China, is also expected to be lower
in 2015 than it has been in years.
The corrosive effect of slower global
growth on the U.S. economy is likely to be
magnified by a substantial appreciation of
the U.S. dollar since late July 2011. During December 2014, the real value of the
dollar rose to its highest level since June
2009. Although Canada and Mexico are the
United States’ two largest trading partners,
and their economies show few obvious
strains, slow global growth and a stronger
dollar could lead to weaker growth of U.S.
exports in 2015. If so, manufacturers may
slow their planned capital expenditures.

Three positive developments stand out.
First, and perhaps most important, is the
aforementioned forward momentum, which
is being manifested by strong employment
growth and a larger-than-expected decline
in the unemployment rate. Second, crude oil
prices have fallen substantially since midJune. Historically, falling energy prices have
helped to boost the real purchasing power
of consumers, spurring faster growth of
consumer expenditures. As an example, car
and light-truck sales in 2014 were at their
highest since 2006. Falling oil prices also tend
to lower headline inflation and—for a time
at least—inflation expectations and nominal
interest rates. However, some policymakers
are worried that inflation will drift too far
below the FOMC’s target of 2 percent.
Macroeconomic policy is the third positive development. On the fiscal side, government expenditures are no longer a negative
contribution to real GDP growth, as they
were from 2011 to 2013. Of note, state and
local government finances have improved.
Regarding monetary policy, financial markets expect the stance of monetary policy to
remain extraordinarily accommodative in
2015 even if, as some Federal Reserve policymakers have suggested, the Fed’s target
for short-term interest rates rises slightly.
Overall, then, key economic and financial
market indicators at the end of 2014 suggest
that the U.S. economy is likely to strengthen
further in 2015.
Kevin L. Kliesen is an economist at the Federal
Reserve Bank of St. Louis. Lowell R. Ricketts, a
senior research associate at the Bank, provided
research assistance. See http://research.stlouisfed.
org/econ/kliesen for more on Kliesen’s work.
The Regional Economist | www.stlouisfed.org 17

d i s t r i c t

o v e r v i e w

Income Inequality Is Growing
in the District,
but Not as Fast as in the Nation

The Eighth Federal Reserve District
is composed of four zones, each of
which is centered around one of
the four main cities: Little Rock,
Louisville, Memphis and St. Louis.

By Maximiliano Dvorkin and Hannah Shell

T

he evolution of national income
inequality is a major topic in current
economic discussions. In a recent speech,
Federal Reserve Chair Janet Yellen
described the issue as one of the most
important of our time: “By some estimates,
income and wealth inequality are near their
highest levels in the past hundred years,
much higher than the average during that
time span and probably higher than for
much of American history before then.” 1
Less discussed is the evolution of income
inequality on a subnational level. Not all
regions in the U.S. exhibit the same inequality patterns as the nation does. Using data
from the annual March supplement of the
Current Population Survey, we analyzed the
long-term trends of income inequality in
the Eighth District. We found that although
inequality in the District has increased, it has
done so at a slower pace than has occurred in
the nation as a whole.2
There are several ways to measure inequality. The most common is in terms of income,
but inequality can also be measured by
consumption and by wealth.3 Each measure
has different implications for the level of
inequality found. In general, measuring
inequality through wealth yields the most
unequal distribution, while income inequality is slightly less unequal, and consumption
inequality is even less unequal.4
This article focuses on income inequality,
using total earnings and disposable income
as the variables for analysis. We chose total
earnings because this variable represents
gross labor income—it excludes income
earned from financial wealth, income from
government transfers (such as welfare) and
deductions due to income taxes. Disposable
income is the amount individuals have left
18 The Regional Economist | January 2015

from all income sources after paying taxes
and receiving government benefits. Comparing disposable income to total earnings
can show how effective the government is at
mitigating inequality.
Income Inequality in the District

Overall, income inequality has increased
in the Eighth District. From 1979 to 2009,
the income ratio of a person in the top 10
percent of the income distribution to that of
a person in the bottom 10 percent has grown
from 5.7 to 6.2.5 This means an individual at
the top of the distribution now earns slightly
more than six times as much as someone at
the bottom. Moreover, the entirety of this
increase is due to larger earnings in the top
of the distribution.
Over the 30 years studied, income in the
90th percentile has grown by more than
8 percent in real terms, while income in the
10th percentile has remained essentially
flat in those same inflation-adjusted terms.
These numbers indicate that the top-echelon
income earners are taking home more,
but the rest of the population’s purchasing
power is about the same or even less than
30 years earlier.
Eighth District vs. U.S.

Clearly, income inequality has increased
in the Eighth District, but how does this
compare with the evolution of income
inequality in the U.S.? Economists commonly use Gini coefficients to answer
these types of questions. A Gini coefficient
measures inequality across a distribution
of individuals, giving a value between zero
(expressing perfect equality) and 1 (expressing perfect inequality). The table shows the
average Gini coefficients over two five-year

periods for the U.S. and the Eighth District.
In terms of inequality measured by annual
earnings, inequality in the Eighth District
rose 0.03 points, from 0.36 in the first period
to 0.39 in the second period. In the U.S.,
inequality increased 0.05 points, from 0.36
to 0.41. By this measure, the U.S. and the
Eighth District started at the same place in
the beginning of our analysis but ended with
income inequality for the nation higher than
income inequality in the Eighth District.
In terms of inequality measured by
disposable income, the Eighth District has
kept pace with the U.S. The Gini coefficients
on disposable income have increased 0.07
points in both the U.S. and the District
between the two periods. However, the
absolute level of inequality in the Eighth
District remains below the U.S. in both
periods reported.6
Average Gini Coefficients for Annual
Earnings and Disposable Income
Eighth District

U.S.

Earnings
1979-1984

0.36

0.36

2004-2009

0.39

0.41

1979-1984

0.33

0.34

2004-2009

0.40

0.41

Disposable Income

SOURCES: Authors’ calculations using data from the U.S. Census Bureau and
the U.S. Bureau of Labor Statistics provided by Unicon Corp.

Another way to compare inequality over
time and regions is to calculate ratios of various percentiles of the income distribution.
For example, the 90th to 50th percentile
ratio compares the income of a person who
stands at the 90th percentile of the income
distribution, that is, a top earner, to that of
a person who stands at the 50th percentile

Income Inequality Ratios for Total Earnings and Disposable Income

ENDNOTES

A) Percentile Ratios, Total Earnings

1 See Yellen.
2 There are some limitations when using data from

3.05

Percentile Ratios

2.85
2.65
2.45
2.25
2.05
1.85
1979

1984

1989

Eighth District, 90th/50th Percentile Ratio
Eighth District, 50th/10th Percentile Ratio

1994

1999

2004

2009

U.S., 90th/50th Percentile Ratio
U.S., 50th/10th Percentile Ratio

3

B) Percentile Ratios, Disposable Income
3.05

Percentile Ratios

2.85

4

2.65

5

2.45

6

2.25
2.05
1.85
1979

1984

1989

Eighth District, 90th/50th Percentile Ratio
Eighth District, 50th/10th Percentile Ratio

1994

1999

2004

2009

U.S., 90th/50th Percentile Ratio
U.S., 50th/10th Percentile Ratio

SOURCES: Authors’ calculations using data from the U.S. Census Bureau and the U.S. Bureau of Labor Statistics provided by Unicon Corp.

of the distribution, that is, a middle-income
earner. Figures A and B show the 90th to
50th and the 50th to 10th percentile ratios
for earnings and disposable income in the
Eighth District and the U.S. In 1979, the
90th to 50th percentile ratios for earnings
and disposable income started out at about
2 in both the U.S. and the Eighth District.
A ratio of 2 means that the top-income
earners in both the measures studied made
about twice as much as the middle class.
In the 30 years since, the ratios for the U.S.
(dashed lines) have increased more rapidly
than the ratios for the District. Both ratios
appear to have followed a similar trend until
the early 1990s, when the income inequality
in the U.S. began to increase more rapidly
than in the Eighth District. In 2009, the U.S.
top-income earners were earning more than
2.4 times the middle class. In the District,
these top individuals were earning about
2.3 times more.
In sum, income inequality is increasing
primarily in the upper end of the distribution, between the top-income earners and

the middle-income earners. Although the
level of income inequality is higher in the
lower end of the distribution, this level has
not dramatically increased over the period
studied. In the upper end of the distribution,
an increasing trend is clearly visible.
Conclusion

Economists continue to debate the source
of the increase in inequality in the past few
decades. The above analysis shows that
although income inequality in the Eighth
District has increased, it has done so at a
slower pace than in the nation as a whole.
Moreover, despite the different paces of
increase, both the U.S. and the Eighth
District have experienced increased income
inequality primarily in the upper end of the
distribution.

the annual March supplement of the Current
Population Survey. These data are somewhat
limited at the subnational level. In particular,
geographic identifiers do not follow the Federal
Reserve district boundaries. For this reason, we
compute measures of income inequality for the
Eighth District by including individuals living in
the following states at the time of the survey: Missouri, Arkansas, Mississippi, Tennessee, Kentucky,
and Indiana. This is, therefore, an approximation
to the population living in the Eighth District’s
territory. While part of the state of Illinois lies in
the Eighth District, we decided to exclude it in the
analysis since most of Illinois’ population, including that of the city of Chicago, does not.
For more details on various measures of inequality
in the U.S., see speech by St. Louis Fed President
James Bullard to the Council on Foreign Relations
on June 26, 2014, at https://www.stlouisfed.org/
from-the-president/speeches-and-presentations/
2014/income-inequality-and-monetary-policy.
See the study by Heathcote, Perri and Violante and
the study by Ricketts and Waller.
We study the evolution of income inequality from
1979 to 2009 because of data availability.
The higher Gini coefficient for disposable income
as compared to earnings doesn’t necessarily mean
the government hasn’t been effective at mitigating
inequality. Disposable income starts with earnings,
then adds income from financial wealth and subtracts
government transfers. Therefore, before government transfers take place, we would expect income
to be more unequal than earnings.

Re f e r en c es
Heathcote, Jonathan; Perri, Fabrizio; and Violante,
Giovanni L. “Unequal We Stand: An Empirical
Analysis of Economic Inequality in the United
States, 1967-2006.” Review of Economic Dynamics,
January 2010, Vol. 13, No. 1, pp. 15-51.
Ricketts, Lowell R.; and Waller, Christopher J.
“U.S. Income Inequality May Be High, but It
Is Lower Than World Income Inequality.” The
Regional Economist, July 2014, Vol. 22, No. 3.
See https://www.stlouisfed.org/publications/
regional-economist/july-2014/us-income-inequality-may-be-high-but-it-is-lower-than-worldincome-inequality.
Yellen, Janet. Speech at the Conference on Economic
Opportunity and Inequality at the Federal Reserve
Bank of Boston on Oct. 17, 2014. See www.federalreserve.gov/newsevents/speech/yellen20141017a.
htm.

Maximiliano Dvorkin is an economist and
Hannah Shell is a research analyst, both at the
Federal Reserve Bank of St. Louis. For more on
Dvorkin’s work, see http://research.stlouisfed.org/
econ/dvorkin.
The Regional Economist | www.stlouisfed.org 19

m e t r o

Sam Walton, the founder of the Walmart chain,
started with this “dime store” in Bentonville, Ark.
The store now serves as the Walmart museum.

p r o f i l e

A Tale of Four Cities:
Widespread Growth
in Northwest Arkansas
By Charles S. Gascon and Michael A. Varley
© 2012 Wal-Mart Stores, Inc. All Rights Reserved.

When Sam Walton opened in 1950 the dime store in Bentonville, Ark., that would evolve into
the giant retailer Walmart, to say the area was sparsely populated would be an understatement.
Even 20 years later, when the first Walmart distribution center was opened in Bentonville, the
city had just over 5,000 people. Today, these environs have grown so much that they comprise
their own metropolitan statistical area (MSA).

20 The Regional Economist | January 2015

these rates, Northwest Arkansas doubles its
income every 35 years, while the U.S. needs
50 years to do the same.
As a result, per capita income in Northwest Arkansas has converged with the
national average. As of 2012, income per
capita in Northwest Arkansas was about
$36,000, about $6,000 below the national
average. After adjusting for the lower cost

of living in the region, per capita income
in Northwest Arkansas was at that time,
in effect, $40,000, just $2,000 shy of the
national average.
Annual, inflation-adjusted income
growth has been relatively uniform across
all counties since 1971, although Benton
(2 percent per year) and Washington
(1.8 percent) exhibited greater growth than

FIGURE 1
Per Capita Personal Income Growth
15
10
5
0
–5

2011

2009

2007

2005

2003

2001

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

–10

1999

U.S.
Northwest Arkansas
1997

Year-over-Year Percent Change

T

he Fayetteville-Springdale-Rogers MSA
(which includes Bentonville) is home to
almost a half-million people. A region more
often referred to as Northwest Arkansas,
this MSA covers four counties, one of which
is across the border in Missouri.
This strong growth is relatively dispersed
among the MSA’s counties. Although the
core counties of Benton and Washington
have been leading the way with an average
population growth of at least 2.4 percent,
every county in the region has experienced
average growth faster than that of the nation
since 1971. Thus, not only has Northwest
Arkansas experienced the fastest population
growth of any MSA located in Arkansas, but
it is also the only MSA in Arkansas (with at
least two counties in the state) that has experienced an average population growth above
the national average in all counties.
Strong growth in real income per capita
has gone hand-in-hand with the growth in
population. Per capita income in Northwest
Arkansas has grown about 2 percent annually since 1970, compared with 1.4 percent
in the U.S. as a whole. (See Figure 1.) At

SOURCES: Census Bureau/Federal Reserve Economic Data (FRED) and Haver.
NOTE: Data are presented at an annual frequency, and the range is 1971-2012. In 2012, per capita income growth was 2.33 percent in the U.S. and 1.96 percent
in Northwest Arkansas.

MSA Snapshot
Northwest Arkansas
Population.............................................................................................491,966
Personal Income (Per Capita)...............................................$35,980
Cost of Living........................................................................................–13.8%
Employment........................................................................................219,300
Unemployment Rate............................................................................4.5%
Pop. (Age 25+) w/Bachelor’s Degree or Higher........ 27.3%
NOTES: Population is from the Bureau of Economic
Analysis (BEA), as of 2013. Per capita income is the
2012 annual figure from the BEA and was created averaging county income data using population weights.
Cost of living figure is from Sperling’s Best Places and
is an annual figure for 2013; it is shown above as relative to the national average. The unemployment rate
and employment figure are from the November 2014
release from the Bureau of Labor Statistics. The percentage of the population with a bachelor’s degree or
© T yson foods inc.

Northwest Arkansas is home to not just one but multiple “city centers,” each with its own specialty. Fayetteville, for example, is a source of college-educated,
skilled labor, thanks to the university there. Tyson Foods (above) is a key industry in the Rogers area. It’s also the second-largest employer in the MSA.

Madison (1.3 percent) and the Missouri
county of McDonald (1.5 percent).
Multiple City Centers

Although the growth rates alone are noteworthy, arguably more impressive is how this
growth has occurred across multiple cities in
the region. Economic activity in most metro
areas typically revolves around one city
center. The economic growth in Northwest
Arkansas is supported by four cities: Fayetteville, Springdale, Rogers and Bentonville.
More than half the metro area’s residents
reside in these four cities, each of which
brings prosperity to the region in its own
way. Fayetteville, where the University of
Arkansas is located, is a source of skilled
labor. Of its residents age 25 and older, 44.8
percent have a bachelor’s degree or higher,
placing it between the cities of Boston (43.9
percent) and Austin, Texas (45.6 percent).
A few miles north of Fayetteville are the
cities of Springdale and Rogers. This area is
home to two of the region’s major employers: Tyson Foods, a multinational food
corporation, and J.B. Hunt, a trucking and
transportation company. Both cities also
have strong manufacturing and construction sectors. In Springdale, 34 percent of
the workforce is employed in one or the
other sector; in Rogers, 22 percent. As a
result of widespread growth in the region
and opportunities for employment, many
families have moved to this part of the state

and now call Springdale and Rogers home.
About 20 percent of the population in this
area is foreign-born, and nearly 30 percent
of residents in Springdale and in Rogers are
Hispanic or Latino, about twice the national
rate. More than 10 percent of the firms in
each of these cities are Hispanic-owned.
The smallest of the four cities in the MSA
is Bentonville, but as home to Walmart, it
has experienced some of the fastest growth
in the region. The population of the city
increased about 14 percent from April 2010
to July 2013, nearly twice the growth rate
of the other three cities and almost six
times the national rate. Bentonville is also
the wealthiest of the four cities; its median
household income tops $60,000—nearly 20
percent higher than each of the other three
aforementioned cities.

higher is from the U.S. Census Bureau; it is a five-year
estimate from 2008-2012.

largest sectors by Employment
(Percent of Nonfarm Employment as of October 2014)
professional and business services
GOVERNMENT
manufacturing
education and health services
leisure and hospitality
0

2

4

6

12

10

14

16

18

20

largest local employers

1. Walmart.................................................................................................28,000
2. Tyson Foods Inc..............................................................................12,000
3. University of Arkansas.................................................................4,000
4. J.B. Hunt....................................................................................................2,600
NOTE: Totals are from the U.S. Department of Housing
and Urban Development, as of January 2012.

Northwest arkansas

McDonald
Benton

MISSOURI

Bentonville
Rogers
Springdale

Economic Outlook

The region has been growing at a remarkable pace for nearly 50 years. However, such
robust growth was not immune to the financial crisis of 2008 and subsequent recession.
Since 2007, Northwest Arkansas has been
experiencing population growth that is
below average for the area, with Madison
County losing population in 2011 and 2012,
and McDonald County doing the same in
2010 and 2011. It is too early to tell if this is a
cyclical phenomenon or if the MSA is experiencing a permanent slowdown in population growth. Long-term trends indicate that

8

Fayetteville

Madison

Washington
OKLAHOMA

ARKANSAS

NOTES: The colors do not correspond to any specific
values; they are used solely to identify the outline of
the metro area. McDonald County is in Missouri but is
included as part of the metropolitan statistical area.

The Regional Economist | www.stlouisfed.org 21

FIGURE 2
Population Growth: U.S. vs. Northwest Arkansas
6

Northwest Arkansas
4
3
2

2012

2010

2008

2006

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

0

1972

1
1970

Percent Change from a Year Ago

U.S.
5

SOURCES: Census Bureau and Haver.
NOTE: Data are presented at an annual frequency, and the range is 1971-2013. In 2013, population growth was 0.7 percent in the U.S. and 2.08 percent in
Northwest Arkansas.

FIGURE 3
Real GDP Growth 2013

Percent Change from a Year Ago

6
© thinkstock

5

Thanks to the University of Arkansas, Fayetteville is a major source in the
state for skilled labor. Of the city’s residents 25 and older, 44.8 percent have a
bachelor’s degree or higher—on par with Boston and Austin, Texas.

4
3
2
1
0

U.S.*

Arkansas

Northwest
Arkansas

Little Rock, Ark.

Springfield, Mo.

Tulsa, Okla.

*U.S. is average of all metro areas.

the region’s population growth is positively
correlated with the U.S. business cycle. As
the national economy gains momentum,
population growth in the region may pick
up again.
While population growth has slowed,
income growth has returned to the longrun average rate after being hit by the Great
Recession. Real income per capita in the
region declined during the first two years
of the recession, but incomes rebounded in
2011 and 2012 (the latest year data are available). All the Arkansas counties are doing
particularly well with all three experiencing above-average income growth (relative
to the long-run trend) at some point after
the end of the Great Recession; however,
McDonald County (Missouri) is showing
the effects of the downturn, with a decline
in income in both 2011 and 2012.
This recovery in the MSA is particularly notable considering the significant
22 The Regional Economist | January 2015

impact of the recession on the local housing
market. From 2006 to 2013, new permits
for private housing fell 60 percent, worse
than both national and state declines of 45
percent. The ability of Northwest Arkansas
to weather such a downturn in housing is a
good sign of stable growth in the MSA.
Despite the uncertainty surrounding
recent population growth, which is most
likely a reflection of economic growth in the
MSA, comparing the region to the nation
as a whole paints an optimistic outlook. In
2013, the region’s economy (measured by
real gross metropolitan product) grew by
5.6 percent, three times the national rate
and much faster than the nearby MSAs of
Little Rock, Ark.; Tulsa, Okla.; and Springfield, Mo. In fact, only 25 of the nation’s 381
MSAs experienced faster growth in 2013,
placing Northwest Arkansas firmly in the
top 10 percent of the fastest-growing MSAs
in the nation. (See Figure 3.)
Charles S. Gascon is a regional economist and
Michael A. Varley is a research analyst, both at
the Federal Reserve Bank of St. Louis. For more
on Gascon’s work, see http://research.stlouisfed.
org/econ/gascon.

READER

E X CHANGE

ASK AN ECONOMIST

Guillaume Vandenbroucke is a senior economist at
the Federal Reserve Bank of St. Louis, where he has
worked since mid-2014. His research focuses on
human capital and schooling, as well as on demographic questions related to such topics as fertility
and marriage. He enjoys reading, swimming and
spending time with his family.
For more on his research, see http://research.
stlouisfed.org/econ/vandenbroucke.

Q: How much education do Americans get?
How has this figure changed over time?
A: These are important questions. Education is a primary determinant of an individual’s lifetime earnings. At a macroeconomic level, understanding the evolution of
educational attainment is relevant, given the importance of human capital to the
national income of countries.
On average, Americans spend about 14 years in school. Educational attainment
has increased remarkably since early in the 20th century, as can be seen in the
chart. In 1940, 76 percent of those 25 and older had not completed high school;
by 2013, only 12 percent hadn’t.
Diego Restuccia at the University of Toronto and I have a paper in which we asked
what caused this substantial trend.1 We developed a model in which individuals can
accumulate human capital (i.e., become educated) and assessed how much technological progress and changes in life expectancy contributed to the increase of educational attainment. We found that skill-biased technical change represented the most
important factor in accounting for the increase in educational attainment. In other
words, the main reason why more people sought education was because technology
keeps rewarding educated people with better and better paychecks. This may sound
obvious, but there are many other reasons for people to continue their education.
Knowledge could be enjoyable, for instance. After all, even retired people sometimes
go back to school to learn about something they are interested in. Yet, we are finding
that the strongest of all reasons is that education simply is a good investment.

Population Age 25 and Over by Educational Attainment: 1940-2013

200,000

65,506
Bachelor’s Degree or Higher
116,876

150,000

Less than High School

SOURCES: www.census.gov/hhes/socdemo/education/data/cps/historical/index.html and
www.oecdbetterlifeindex.org/topics/education.

endnote
1

See www.economics.utoronto.ca/public/workingPapers/tecipa-446.pdf.

2013

2010

2007

2004

2001

1998

1995

1992

1989

1986

1983

24,517
1980

1977

1960

1952

1940

0

High School or Some College

1974

50,000

1971

3,407
14,627
56,742

1968

100,000

1965

Population in Thousands

250,000

What Does the St. Louis Fed Do?
Find Out in “100 Years of Service”
In a report published recently to mark the 100th anniversary of the St. Louis Fed and the Fed System, you will
learn not only about the founding and history of these
institutions, but you will get a first-person account of
the work of each department at the St. Louis Bank
today. For example, Chris Waller, the head of our
Research department, explains the academic-style
research that our economists undertake. Julie Stackhouse talks about the sophisticated approach taken to
bank supervision these days. Karen Branding highlights
the importance of earning the public’s trust. Other
essays discuss the payments system, our work for the
Treasury, internal and external audits of the St. Louis Fed,
and many other aspects of day-to-day operations.
The core of “100 Years of Service,” however, is the
history. You will read about the financial instability in the
country that led to the birth of the Fed, our nation’s
third attempt at a central bank. You will also find out why,
a half-century later, the St. Louis Fed came to be known
as the maverick in the Fed System.
You can scroll through the book—or download it in
the iTunes store—online at www.stlouisfed.org/annualreport/2013. There, you will also find a seven-minute
video that captures some of the highlights.
Deck the Halls—with infographics
from the st. louis fed
As you take down holiday decorations in your workspace, consider putting up something almost as colorful
and even more thought-provoking: some of our new
infographics on subjects related to the economy. One
illustrates the history of dissenting votes on the Federal
Open Market Committee; it highlights key data from a
recent article in our research journal, the Review. Another
one breaks down the pros and cons of traditional and
alternative providers of financial services (from banks
to pawnshops). The changing
landscape of housing market
conditions around the country is
the subject of a third
graphic; it, too, highlights
key data from indepth
reports that are also
available from the
St. Louis Fed. These
and more can be seen
at www.stlouisfed.org/
infographics. When you
follow the links, you can
print the infographics
yourself.
We welcome letters to the editor, as well as questions for
“Ask an Economist.” You can submit them online at www.
stlouisfed.org/re/letter or mail them to Subhayu Bandyopadhyay, editor, The Regional Economist, Federal Reserve
Bank of St. Louis, P.O. Box 442, St. Louis, MO 63166-0442.

The Regional Economist | www.stlouisfed.org 23

Change Service Requested

N E X T ISS U E

The Synchronization of Business Cycles across Countries

There’s Still Time
To Tell Us What You Think
Our survey of Regional Economist readers
is still open. If you didn’t return the postcard
that was attached to the last issue, you still
have time to go online to take this quick
survey. Go to www.stlouisfed.org/
publications/regional-economist and
look for the orange survey button in the
right-hand column.

Tell us how to improve
The Regional Economist
—take our survey!

s the world shrinks, countries’ business cycles—those shifts from expansion to
recession and back again—are becoming more synchronized. This means that a
shock—good or bad—that is felt in one country may reverberate in other countries.
Countries whose cycles are not connected to anyone else’s do not suffer—or enjoy—
the ripple effects from other countries. This interconnectedness is often thought of
as a global phenomenon, but increasingly of late, its importance is being felt more
on the regional level. Find out more in the April issue of The Regional Economist.

Your responses will help us determine
what changes, if any, we should make in the
coming year. In addition, we want to know a
bit about you—to ensure that we are writing
for the right audience.
Sincerely,
The RE team at the St. Louis Fed

printed on recycled paper using 10% postconsumer waste

economy

at

a

The Regional

glance

Economist

january 2015

REAL GDP GROWTH

4
2
0
–2
–4
–6
–8

Q3
’09

’10

’11

’12

’13

PERCENT CHANGE FROM A YEAR EARLIER

6

6

PERCENT

VOL. 23, NO. 1

CONSUMER PRICE INDEX

8

–10

|

’14

CPI–All Items
All Items, Less Food and Energy

3

0

–3

December

’09

’10

’11

’12

’13

’14

NOTE: Each bar is a one-quarter growth rate (annualized);
the red line is the 10-year growth rate.

I N F L AT I O N - I N D E X E D T R E A S U RY Y I E L D S P R E A D S

RATES ON FEDERAL FUNDS FUTURES ON SELECTED DATES

3.00
2.75
2.50
PERCENT

PERCENT

2.25
2.00
1.75

5-Year

1.50

10-Year

1.25

20-Year

1.00

’11

’12

Jan. 9

’13

’14

0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

09/17/14

1st-Expiring
Contract

’15

NOTE: Weekly data.

C I V I L I A N U N E M P L O Y M E N T R AT E

12/17/14

10/29/14

3-Month

6-Month

01/14/15

12-Month

CONTRACT SETTLEMENT MONTH

I N T E R E S T R AT E S

11

4

10-Year Treasury

10
3

8

PERCENT

PERCENT

9

7
6

2

1

Fed Funds Target
1-Year Treasury

5
4
’09

December

’10

’11

’12

’13

0

’14

’09

’10

’11

’12

’13

December

’14

NOTE: On Dec. 16, 2008, the FOMC set a target range for
the federal funds rate of 0 to 0.25 percent. The observations
plotted since then are the midpoint of the range (0.125 percent).

U . S . A G R I C U LT U R A L T R A D E
90

AVERAGE LAND VALUES ACROSS THE EIGHTH DISTRICT
7,000

Exports

6,000
Imports

60

DOLLARS PER ACRE

BILLIONS OF DOLLARS

75

45
30
15

Trade Balance

0
’09

’10

’11

’12

’13

NOTE: Data are aggregated over the past 12 months.

Quality Farmland

Ranchland or Pastureland

5,000
4,000
3,000
2,000
1,000

November

’14

0

2013:Q3 2013:Q4 2014:Q1 2014:Q2 2014:Q3
SOURCE: Agricultural Finance Monitor.

U.S. CROP AND LIVESTOCK PRICES / INDEX 1990-92=100
140
120

Crops
Livestock

100
80
60

December

40
’99

’00

’01

’02

’03

’04

’05

’06

’07

’08

’09

’10

’11

’12

’13

’14

YEAR

commercial bank performance ratios
U . S . B an k s by A sset S i z e / T hird Q U A R T E R 2 0 1 4

All

$100 million­$300 million

Less than
$300 million

$300 million$1 billion

Less than
$1 billion

$1 billion$15 billion

Less than
$15 billion

More than
$15 billion

Return on Average Assets*

1.02

0.99

0.96

1.02

0.99

1.02

1.01

1.02

Net Interest Margin*

3.09

3.80

3.80

3.82

3.81

3.88

3.85

2.92

Nonperforming Loan Ratio

2.13

1.48

1.50

1.52

1.51

1.47

1.48

2.32

Loan Loss Reserve Ratio

1.55

1.58

1.58

1.55

1.56

1.40

1.47

1.58

R E T U R N O N AV E R A G E A S S E T S *

NET INTEREST MARGIN*
1.10

0.94

1.27
1.25
1.02
0.98

0.91

1.03
1.09

Indiana

1.02

Kentucky

.00

.20

.40

.80

Third Quarter 2014

1.00

1.20

1.40

PERCENT

0.0 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50

2.31

1.49

1.35

1.41

1.60
2.25

2.00

2.50

Third Quarter 2013

NOTE: Data include only that portion of the state within Eighth District boundaries.
SOURCE: FFIEC Reports of Condition and Income for all Insured U.S. Commercial Banks
* Annualized data

1.58
1.89

1.49
1.62

Tennessee
PERCENT

1.65

1.68

Missouri

1.57

1.50

1.52

1.44

Mississippi

1.55

1.87

1.15

Kentucky

2.11

1.26

0.97

Indiana

1.94

1.71

1.55

Arkansas
Illinois

1.30
1.16

Third Quarter 2013

Eighth District

1.91

1.70

Third Quarter 2014

3.45
3.42

L O A N L O S S R E S E RV E R AT I O

1.25
1.37

1.00

3.63
3.64

Third Quarter 2014

1.61

.50

3.80
3.65

Third Quarter 2013

1.40

4.17

3.77
3.83

Tennessee

N O N P E R F O R M I N G L O A N R AT I O

.00

3.90

Missouri

1.13

.60

3.60
3.47

Mississippi

1.05
0.98
0.44

4.27
4.11

Arkansas
Illinois

0.97

0.82

3.81
3.73

Eighth District

.00

.30

.60

Third Quarter 2014

.90

1.20

1.50

1.80

Third Quarter 2013

For additional banking and regional data, visit our website at:
www.research.stlouis.org/fred/data/regional.html.

2.10

regional economic indicators
nonfarm employment growth / third Q U A R T E R 2 0 1 4
year-over-year percent changE
United
States

Eighth
District †

Arkansas

Illinois

Indiana

Kentucky

Mississippi

Missouri

Tennessee

Total Nonagricultural

1.9%

1.5%

1.4%

0.7%

2.3%

1.3%

1.0%

1.9%

2.2%

Natural Resources/Mining

5.7

2.2

2.7

1.7

3.8

0.9

4.3

0.8

NA

Construction

3.8

2.5

7.9

5.8

5.1

–6.3

–4.6

0.4

3.8

Manufacturing

1.4

1.9

2.1

–0.7

4.6

1.2

3.6

1.8

1.9

Trade/Transportation/Utilities

2.1

0.6

0.3

0.2

0.5

1.1

0.4

–0.1

2.1

Information

0.3

–1.5

–4.2

–1.5

–1.7

2.9

–2.4

–1.6

–2.4

Financial Activities

1.0

0.6

0.1

–0.1

1.6

–2.8

0.1

2.8

1.6

Professional & Business Services

3.5

3.2

1.8

2.4

2.5

4.5

2.4

3.7

5.8

Educational & Health Services

1.9

1.2

2.1

0.5

1.4

1.2

1.7

2.1

1.0

Leisure & Hospitality

2.5

2.7

3.1

0.1

2.9

4.7

1.5

4.9

4.3

Other Services

0.9

1.2

1.7

0.8

3.9

1.7

–2.4

–0.4

1.2

Government

0.3

0.7

–0.2

0.7

1.8

0.4

0.8

1.6

–1.0

† Eighth District growth rates are calculated from the sums of the seven states. For Natural Resources/Mining and Construction categories, the data exclude
Tennessee (for which data on these individual sectors are no longer available).

District real gross state product by industry-2013

U nemployment R ates

Information 3.7%

Financial Activities

III/2014

II/2014

III/2013

United States

6.1%

6.2%

7.2%

Arkansas

6.2

6.4

7.7

Illinois

6.7

7.5

9.2

Indiana

5.8

5.8

7.5

Kentucky

7.1

7.6

8.4

Construction 3.6%

Mississippi

7.9

7.7

8.6

Natural Resources
and Mining 2.0%

Missouri

6.4

6.6

6.6

Tennessee

7.3

6.4

8.3

Trade,
Transportation
and Utilities

Professional and
Business Services

17.7%

18.1%

11.3%
Manufacturing

17.2%

Education and
Health Services

9.0%

Leisure and
Hospitality 3.6%

11.4%
Government

Other Services 2.2%

United States $15,527 Billion
District Total		 $ 1,877 Billion
Chained 2009 Dollars

H ousing permits / third quarter

REAL PERSONAL INCOME* / third QUARTER

year-over-year percent change in year-to-date levels

year-over-year percent change

6.7

–10.0

United States

21.1

–0.8

Arkansas
27.0

18.6
6.9

–3.3

Mississippi
23.7
24.7

8.4

2014

–5

0

5

10

15

20

–0.8
2.2
1.4

25

All data are seasonally adjusted unless otherwise noted.

30

35

PERCENT

1.9
2.2

Tennessee

2013

2.6

0.9

Missouri

17.7

1.9
1.6

0.7

Kentucky

5.5
7.7

–15 –10

1.2

Indiana

27.8

1.5

1.0

Illinois

5.0

2.4

1.8

1.3

–1.50 –1.00 –0.50 0.00 0.50 1.00 1.50 2.00 2.50 3.00
2014

2013

*NOTE: Real personal income is personal income divided by the PCE
chained price index.