The Debate on Unemployment Benefit in the US

Are Unemployment Benefit Extensions Responsible for High Long-Term Unemployment.? 

The 2007-09 recession produced the greatest labor market meltdown in the U.S. since
the Great Recession. Three indicators underscore the collapse in labor demand responsible
for the record rates of long-term unemployment that have persisted into 2011.

Job destruction, measured as the share of the work force that experienced permanent
job loss (not temporarily laid off) rose to 4.5 percent in late 2009, far above its worst level
of 2 percent reached in the aftermath of the 2001 recession. The job openings rate fell from
a normal range of well above 3 percent (2006-7) to 1.6 percent in mid-2009 and has
remained at only around 2 percent between March and January 2011. And finally, the hires
rate, which in good times is around 4 percent (2005-6), has been stuck at an abysmal 3
percent since mid-2008.

Large-scale job destruction and feeble hiring have had predictable consequences for the
employment numbers. The employment-to-population rate (EPR) fell from 63.4 percent in
March 2007 to just 58.2 percent at the end of 2009, an unprecedented collapse in recent
memory. Tthe number of unemployed workers increased from 7.7 to
14.8 million between December 2007 and June 2009, peaked at nearly 16 million in late
2009, and was still at 14.9 million in October 2010. Long-term unemployment increased
from 1.3 million to 4.4 million over the course of the official recession. By mid-2010 the
long-term unemployed had increased by another 50%, to 6.6 million workers.

Policy makers responded to this employment calamity by extending the maximum
duration of unemployment insurance (UI) benefits. Beginning in May 2008, Federal
Extended Benefits (EB) and Emergency Unemployment Compensation (EUC) extensions
raised the regular 26-week limit to as many as 99 weeks for some workers.2 As Figure 2
shows, total UI claimants rose from a “normal” level of 2-3 million to about 4 million in the
fall of 2008, and then exploded to nearly 12 million early 2010. At least as striking, Federal
extended UI claimants increased from zero in May 2008 to 6 million in January-March

In response to this dramatic increase in the duration of eligibility for UI benefits,
leading economists and business press editorials have warned that these policy
interventions must reduce incentives to take jobs and predicted that the extensions had
become a leading source of the persistence of high unemployment. NBER President Martin
Feldstein testified before Congress at the very bottom of the trough of the Great Recession
(September 2009) that extending benefit eligibility would “create undesirable incentives
for individuals to delay returning to work. University of Chicago’s Casey B. Mulligan
(2010a) predicted that passage of UI benefit extensions “would raise unemployment rates
and reduce both employment and economic output” and that to believe otherwise would be
to suspend “the laws of economics.”  Writing in the Wall Street Journal, Harvard’s Robert
Barro (2010) estimated that two-thirds of 6.7 million long-term unemployed would have
been employed in the summer of 2010 but for the eligibility extensions. Barro’s colleague
at Harvard, Gregory Mankiw (2010), explained that the reason for his “ambivalence” about
whether UI benefit extensions should be reauthorized at the height of the unemployment
crisis was because “UI reduces the job search efforts of the unemployed.”

With such professional luminaries sounding the alarm over work disincentives, The
(November 2009) needed no reference to the evidence to inform its readers that
“It may seem heartless to counsel against too much support for the unemployed but
incentives matter even when unemployment is high…. More generous benefits will mean
vacancies are filled less quickly, pushing up unemployment.” The Wall Street Journal’s
editorial page responded with numerous attacks on the extensions, with such titles as
“Incentives Not to Work” (April 13, 2010) and “Stimulating Unemployment” (July 20,

The prediction of such large work disincentive effects is consistent with the
conventional wisdom in academic and policy circles that labor market rigidities caused by
social protection policies like UI were at the root of high European unemployment in the
1980-90s (for a critical assessment see Howell et al., 2007; Howell and Rehm, 2009). As it
turned out, the U.S. unemployment advantage was actually limited to the 1990s and turned
out to be short lived.

By the mid-2000s, the U.S. was no longer among the top employment
performers. The OECD’s harmonized figures for November 2010 put U.S. unemployment at
9.8 percent, far above the German and Italian rates (6.6% and 8.6%) and about the same as
the French rate (9.7%). But labor market rigidities have remained at the heart of the
orthodox explanation for persistent high unemployment. As Mulligan (2010b) put it, “It is
no surprise that adopting a European safety net is giving us a European unemployment
problem” (Mulligan, 2010b).

We argue that the case against the 2008-10 UI extensions on work disincentive
grounds has been vastly exaggerated and suggest that this misdiagnosis of the direction of
causation between unemployment and UI benefits reflects the orthodox theoretical lens
through which the labor market and the work decision are viewed. In short, it is theory-
driven. The policy consequences of getting the direction of causation right are obvious.

We contrast the conventional “labor-leisure” model of the work
decision, which provides the theoretical framework for the work disincentive view, with an
alternative “employment-idleness” view. In the conventional model, workers allocate their
time between two activities, labor, which is market work done only to get the income
necessary for consumption, and leisure, which is the residual. Workers want both
consumption and leisure and they choose the combination they prefer. “Unemployment”
falls under leisure since it is chosen: workers are assumed to be able to reduce their
reservation wage to get a job if they really want one, so there is no involuntary
unemployment. The function of UI benefits is to reduce the need for market work to satisfy
a given level of consumption. Such strategic use of UI benefits requires that the UI system’s
work requirements either do not exist or are not effectively enforced (requirements of
adequate pre-benefit employment, job loss without cause, active job search, and job-
taking). If the sum of the value of leisure and the UI benefit are greater than the value of
income from labor, workers choose leisure (“unemployment”), and the longer the duration
of benefits the longer the period of full-time leisure. In this way, the “laws of economics”
require direct UI-related work disincentive effects.

In contrast, in an employment-idleness labor market, labor markets are characterized
by nominal wage rigidity (workers cannot get jobs by lowering their wage demands), there
is job rationing (especially in downturns), and there is assumed to be a more expansive set
of relevant benefits and costs. Workers get substantial nonpecuniary benefits to
employment, experience large nonpecuniary costs from unemployment, can expect
significant “scarring” effects of on future wages and employment that increase with the
duration of unemployment, and confront effective enforcement of work requirements by
the UI system. In a labor market of this sort, even quite generous benefits available for long
durations may have little or no work disincentive effects for a rational worker, especially
when job rationing is severe. But UI benefits can be expected to promote labor market
attachment and job search (required for the benefits), and may therefore increase the
formal unemployment rate.

Considering  the literature that makes use of pre-Great Recession studies
to estimate effects of the 2008-9 UI extensions on unemployment. We address
three main approaches. The first extrapolates from earlier labor markets under
the assumption that the U.S. labor market would have produced outcomes, such as
long-term unemployment and the job finding rates of the unemployed, similar to
those in the past in the absence of the policy shock of large-scale UI extensions
A second approach sees shifts in the Beveridge curve (the relationship between job vacancies
and unemployment) in the Great Recession as evidence of UI work disincentive effects. The
third and most credible extrapolation approach has relied on evidence on the exit rate from
unemployment at benefit exhaustion (e.g., at 26 weeks) – the so-called “spike” evidence. It
is notable that these supply elasticities are typically taken from a small selective set of
studies that used data from the U.S. in the early 1980s, and not from the many studies, most
notably those published since 2000, that have found little evidence of such “spikes”. No
justification is offered in any of the studies we reviewed for the unconditional application
of elasticity coefficients generated from earlier labor markets to the Great Recession

Finally we consider some suggestive evidence from 2008-10, both from recently
published studies and from our own examination of the data. We find no support for UI
extended benefit effects in either benefit exhaustion rates or labor flow rates (from
unemployment to employment). On the other hand, the labor flow data appear consistent
with a “nonparticipation” effect on long-term unemployment, since the timing of flows
from unemployment to nonparticipation for the long-term unemployed closely correspond
to changes in extended benefit claimant numbers and rates.

Focusing on the Great Recession in the U.S.,examination of 2008-10 data offers no
support for the work disincentive account. If UI benefit extensions are responsible for
increasing unemployment, it appears to have been more by increasing the incentive to
remain connected to the labor market in a time of severe job rationing than by increasing
the incentive to avoid working.

If future empirical work supports this conclusion, it would be consistent with the
heterodox vision that most workers value a job apart from the income it provides and
abhor unemployment, because it is idleness, shame, inadequate income, and fear of a
jobless future that they wake up to every morning, not leisure. In times of deficient
aggregate demand, UI spending increases job openings and encourages job search. The
policy implications are straightforward. On both equity and efficiency grounds, the correct
policy response to large-scale job destruction and job rationing is to dramatically increase
the duration of benefits, just as U.S. policymakers did in The Great Recession. 


Barro, R. (2010), ‘The Folly of Subsidizing Unemployment’, Wall Street Journal, August 30.

Howell, D. R., D. Baker, A. Glyn and J. Schmitt. (2007), ‘Are Protective Labor Market
Institutions at the Root of Unemployment? A Critical Review of the Evidence’, Capitalism
and Society, 2(1), 1-71.

Howell, D. R. and M. Rehm. (2009), ‘Unemployment compensation and high European
unemployment: a reassessment with new benefit indicators’, Oxford Review of Economic
Policy, vol. 25, no. 1, 60-93.

Mankiw, G. (2010), ‘My Agnosticism about UI’, Greg Mankiw’s Blog, December 4.

Mulligan, C. B. (2010a). ‘Do Jobless Benefits Discourage People From Finding Jobs?’,
Economix Blog, The New York Times, March 17.

Mulligan, C. B. (2010b). ‘Eurosclerosis Comes to America’, Economix Blog, The New York
Times, August 4. 

This policy brief is a short version of the paper co-authored with Bert Azizoglu,
“Unemployment Benefits and Work Incentives: The U.S. Labor Market in the Great
Recession,” forthcoming in the Oxford Review of Economic Policy. It can be downloaded
from the Schwartz Center for Economic Policy Analysis (The New School) or from the
Political Economy Research Institute (UMass-Amherst). 

David R. Howell