Long-Term Unemployment and Recessions
Gerald Mayer
Analyst in Labor Policy
Linda Levine
Specialist in Labor Economics
April 14, 2010
Congressional Research Service
7-5700
www.crs.gov
R41179
CRS Report for Congress
P
repared for Members and Committees of Congress
Long-Term Unemployment and Recessions
Summary
The recession that began in the United States in December 2007 has been one of the deepest and
longest since World War II. One feature that distinguishes the recent recession from its postwar
predecessors is the historically high percentage of unemployed persons without jobs for more
than six months (the long-term unemployed). This report analyzes the trend in long-term
unemployment over the postwar period and compares the individual, job, and household
characteristics of the long-term unemployed during the latest recession (2007-2009) with the
long-term unemployed at the end of the two previous recessions (1990-1991 and 2001).
In each of the last three recessions, older unemployed workers were more likely than younger
workers to have been unemployed for more than six months. On the other hand, during the last
two recessions, an equal share of unemployed men and women were without work for over half a
year. At the end of the 1990-1991 recession, unemployed women were less likely than men to
have been out of work for more than six months. Unlike the two previous recessions, in 2009,
unemployed workers with less than a high school education were more likely than unemployed
workers with more education to have been out of work for at least six months.
Long-term unemployment varies by industry and occupation. In 2009, workers laid off from the
financial activities and information industries were the most likely to have been jobless longer
than 26 weeks. Workers displaced from management, business, and financial occupations were
most at risk of long-term unemployment during recent recessions.
Unemployment affects both the individuals who are without work and their families. Households
of the long-term unemployed have lower earnings and income than other households (where
households include married couples, single parents, and single individuals). In 2008, the most
recent year for which data are available, the long-term unemployed were more likely than all
unemployed workers to live in households with incomes below the official poverty line. They
were more likely than other unemployed workers to receive benefits from the Supplemental
Nutrition Assistance Program (SNAP, formerly the Food Stamp program) or be covered by
Medicaid. In 2008, only 2.7% of the long-term unemployed received public assistance.
Over half (55%) of the long-term unemployed had some type of health insurance coverage at
some time during 2008, compared to 84% of employed workers. Although a majority of the long-
term unemployed (58%) were homeowners in 2008, they were less likely than employed workers
(72%) to own their own homes.
As the economy recovers and employers increase hiring to meet the growing demand for goods
and services, many currently unemployed workers will be able to find new jobs. However,
finding work may be more difficult for the long-term unemployed if, for example, employers
think their skills have deteriorated during their lengthy time away from the workplace. The long-
term unemployed displaced from industries in which restructuring has occurred may also have a
hard time finding new jobs in other industries, especially if the jobs require skills different from
those they possess. Policies to encourage employers to hire the long-term unemployed include
wage and training subsidies. Offering wage insurance and reemployment bonuses to unemployed
workers may encourage them to accept jobs sooner than they otherwise might have. This report
may be updated if issues warrant it.
Congressional Research Service
Long-Term Unemployment and Recessions
Contents
The Trend in Long-Term Unemployment .................................................................................... 1
Who is Likely to Experience Long-Term Unemployment?........................................................... 6
Individual Characteristics...................................................................................................... 6
Age................................................................................................................................. 6
Gender ............................................................................................................................ 7
Educational Attainment ................................................................................................... 8
Industry .......................................................................................................................... 9
Occupation.................................................................................................................... 10
Household Characteristics ................................................................................................... 11
Earnings and Income..................................................................................................... 12
Unemployment Benefits................................................................................................ 15
Poverty ......................................................................................................................... 15
Public Assistance and Supplemental Nutrition Assistance.............................................. 16
Health Insurance Coverage............................................................................................ 17
Number of Earners in the Household............................................................................. 18
Homeownership ............................................................................................................ 19
Implications for Public Policy ................................................................................................... 21
Figures
Figure 1. Long-Term Unemployed as a Percentage of the Labor Force,
1948 to 2010, Seasonally Adjusted ........................................................................................... 2
Figure 2. Long-Term Unemployed as a Percentage of Unemployed Workers,
1948 to 2010, Seasonally Adjusted ........................................................................................... 3
Figure 3. Percentage of Unemployed by Age Group Who are Long-Term
Unemployed, 2009, 2001, and 1991 ......................................................................................... 7
Figure 4. Percentage of Unemployed by Gender Who are Long-Term
Unemployed, 2009, 2001, and 1991 ......................................................................................... 8
Figure 5. Percentage of Unemployed by Education Who are Long-Term
Unemployed, 2009, 2001, and 1991 ......................................................................................... 9
Figure 6. Percentage of Unemployed by Industry Who are Long-Term Unemployed, 2009
and 2001 ................................................................................................................................ 10
Figure 7. Percentage of Unemployed by Occupation Who are Long-Term
Unemployed, 2009 and 2001.................................................................................................. 11
Figure 8. Average Annual Real Earnings, All Households, 2008, 2001, and 1991....................... 13
Figure 9. Average Annual Real Income, All Households, 2008, 2001, and 1991 ......................... 13
Figure 10. Average Annual Real Earnings, Married Couples, 2008, 2001, and 1991................... 14
Figure 11. Average Annual Real Income, Married Couples, 2008, 2001, and 1991 ..................... 15
Figure 12. Percentage of Individuals with Health Insurance Coverage,
2008, 2001, and 1991 ............................................................................................................. 18
Figure 13. Percentage of Individuals Who are Homeowners, 2008, 2001, and 1991 ................... 20
Congressional Research Service
Long-Term Unemployment and Recessions
Tables
Table 1. Percent of Employed, Unemployed, and Long-Term Unemployed by Individual
or Household Characteristic, 2008, 2001, and 1991 ................................................................ 16
Table 2. Number of Earners in the Households of the Employed, Unemployed, and Long-
Term Unemployed, 2008, 2001, and 1991 .............................................................................. 18
Appendixes
Appendix. Data and Methodology ............................................................................................. 24
Contacts
Author Contact Information ...................................................................................................... 25
Congressional Research Service
Long-Term Unemployment and Recessions
he recession that began in the United States in December 2007 has been one of the deepest
and longest since World War II. In response, Congress enacted legislation to stimulate
T aggregate demand and provide assistance to unemployed workers and their families. The
American Recovery and Reinvestment Act of 2009 (ARRA, P.L. 111-5) temporarily raised
spending by increasing funding for a number of programs, creating new programs, and providing
tax benefits for individuals, families and businesses. Congress also authorized a federally funded
temporary program (P.L. 110-252 as amended) to provide additional weeks of unemployment
benefits to those who exhausted their eligibility under the federal-state Unemployment
Compensation (UC) program. Although some in the public policy community are concerned that
extending benefits may discourage recipients from looking for work,1 Congress has extended the
program on multiple occasions.2
It is not unusual for improvement in the labor market, as measured by a declining unemployment
rate, to begin some months after the official ending date of a recession.3 It is also not unusual for
the proportion of the labor force unemployed for more than six months (the long-term
unemployment rate) to continue rising and peak after a recession’s end. What is unusual,
however, is that the incidence of long-term unemployment currently is higher than at any time in
the postwar period.
The purpose of this report is to assist policymakers who may consider legislation to stimulate job
growth, provide additional benefits to unemployed workers, or assist families whose incomes
have fallen because of a job loss or reduction in hours worked. The report first compares the
prevalence of long-term unemployment across business cycles during the postwar period. It next
analyzes the likelihood of unemployed workers with different demographic, job, and household
characteristics experiencing a very long period without paychecks. The report closes with a
discussion of long-term unemployment’s implications for public policy.
The Trend in Long-Term Unemployment
The percentage of the labor force without jobs for longer than six months (the long-term
unemployment rate)—like the unemployment rate—generally rises as economic activity falls.
This countercyclical pattern was evident during the deep recessions of the 1970s and 1980s, when
the long-term unemployment rate rose 0.4 percentage points to 0.8% at the 1973-1975 recession’s
end and 1.1 percentage points to 2.1% at the 1981-1982 recession’s end.4 The same pattern was
evident during milder recessions, when the long-term unemployment rate rose 0.3 percentage
points to 0.8% at the trough of the 1990-1991 and 2001 recessions. As the latest recession
unfolded, the long-term unemployment rate increased from 0.9% in December 2007 to 3.3% in
August 2009 (the recession’s ending date used in this report).5 Not only was this 2.4 percentage
1 Michael A. Fletcher and Dana Hedgpeth, “Are Unemployment Benefits No Longer Temporary?,” Washington Post,
March 9, 2010, pp. A1, A11.
2 For additional information, see CRS Report RS22915, Temporary Extension of Unemployment Benefits: Emergency
Unemployment Compensation (EUC08), by Julie M. Whittaker and Alison M. Shelton.
3 CRS Report R40798, Unemployment and Employment Trends Before and After the End of Recessions, by Linda
Levine.
4 Calculated by the Congressional Research Service (CRS) from the Current Population Survey (CPS). Individuals are
counted as unemployed if they are not working, have actively looked for work during the four weeks before the CPS
survey, and are currently available for work.
5 The National Bureau of Economic Research is the official arbiter of turning points in the business cycle. It typically
(continued...)
Congressional Research Service
1
Long-Term Unemployment and Recessions
point increase much greater than recorded during prior recessions, but it also put the long-term
unemployment rate at a postwar high. (See Figure 1.)
Figure 1. Long-Term Unemployed as a Percentage of the Labor Force,
1948 to 2010, Seasonally Adjusted
5%
4%
3%
2%
1%
0%
8
5
75
5
0
5
10
194
195
1960
1965
1970
19
1980
198
199
199
2000
2005
20
Recessions
Long-term unemployed as a percent of the labor force
Source: CRS analysis of data from the Current Population Survey (CPS).
As also shown in Figure 1, the long-term unemployment rate continues to rise during the early
stage of recoveries. For example, the share of the labor force without jobs for more than six
months increased from 3.3% in August 2009 to 4.1% in January 2010. The rate rose further to
4.3% in March, the latest month for which data from the Current Population Survey (CPS) are
available from the U.S. Bureau of Labor Statistics (BLS).
Another way to look at the trend in long-term unemployment is as a proportion of total
unemployment. The long-term unemployed’s share of all unemployed individuals was 33.6% in
August 2009, up from 17.3% in December 2007.6 This measure also continues to rise during the
early stage of recoveries. It rose to 44.1% in March 2010, as fewer people flowed into (short-
(...continued)
has not announced the starting and ending dates of recessions until well after the fact. For purposes of comparison with
prior recessions, August 2009 was used in this report as the month in which the latest recession may have ended
because it is the consensus among economists that the business cycle reached its nadir in the third quarter of 2009.
6 Data derived by BLS from the CPS.
Congressional Research Service
2
Long-Term Unemployment and Recessions
term) unemployment. The proportion of unemployed persons out-of-work over 26 weeks has
been considerably above the rate recorded during prior downturns and nascent recoveries. (See
Figure 2.)
Figure 2. Long-Term Unemployed as a Percentage of Unemployed Workers,
1948 to 2010, Seasonally Adjusted
50%
40%
30%
20%
10%
0%
8
55
60
65
75
90
5
00
05
194
19
19
19
1970
19
1980
1985
19
199
20
20
2010
Recessions
Long-term unemployed as a percent of unemployed
Source: Bureau of Labor Statistics (BLS) data from the Current Population Survey (CPS).
The historically high incidence of long-term unemployment observed starting in mid-20097
coincides with Congress’ lengthier than usual temporary extension of UC benefits beyond the
duration afforded by the permanently authorized federal-state UC system and Extended Benefit
programs.8 The availability of a maximum of 53 weeks of benefits through the Emergency
Unemployment Compensation (EUC) program probably has prompted individuals to continue
searching for jobs instead of leaving the labor force due to discouragement over their job
7 The long-term unemployment rate exceeded 2.6% starting in May 2009; 2.6% was the previous record-high for the
postwar period. It was reached in June 1983, seven months after the end of the 1981-1982 recession.
8 The Extended Benefit (EB) program is automatically triggered when a state surpasses a statutorily set unemployment
rate. Workers in high-unemployment states are eligible for up to 20 weeks of benefits beyond the duration provided
under their state’s UC program (typically 26 weeks). With ARRA, as amended, temporarily paying 100% of EB costs,
many workers unemployed during this recession were eligible for the EB benefit as well as the EUC benefit.
Additionally, during prior recessions when Congress enacted a temporary unemployment benefit program, states were
allowed to suspend the EB program.
Congressional Research Service
3
Long-Term Unemployment and Recessions
prospects upon exhausting their regular UC benefits.9 The currently low hiring rate of employers
makes workers refraining from accepting jobs in order to continue receiving EUC benefits a less
plausible explanation for the historically high incidence of long-term unemployment.10
The shorter-lived temporary UC benefit programs enacted to mitigate the effects of seven prior
recessions coincided with much less long-term unemployment. These programs provided a
maximum of 13 to 33 weeks of benefits. After the seven recessions ended, the long-term
unemployment rate and share of long-term unemployment ranged from 0.8% to 2.6% and from
13.6% to 25.7%, respectively.11 The 111th Congress provided a maximum of 33 weeks of benefits
from about November 2008 through October 2009. Both the long-term unemployment rate
(3.7%) and the share of long-term unemployment (36.0%) in October 2009 were well outside
these ranges. The subsequent extension of benefits to a maximum of 53 weeks starting in
November 2009 has been accompanied by a still higher long-term unemployment rate and share
of long-term unemployment in early 2010. (See Figure 1 and Figure 2.)
The greater industry restructuring that arguably characterized the 2001 recession may have
coincided with the latest recession as well, and made it more difficult for workers to find new
jobs as quickly as they did after earlier recessions. Workers displaced by industries permanently
reducing their demand for labor will take longer to become reemployed than workers waiting to
be recalled by their employers from temporary layoffs. The period of job search could be
especially prolonged when the skill sets needed for employment in growing industries differ from
the skills sets of workers laid off by restructuring industries and when displaced workers must
substantially lower their wage expectations to obtain new jobs.12
9 The official count of unemployment is developed by the BLS from the CPS, not from UC administrative records that
cover a subset of unemployed persons who are benefit-eligible. (In 2009, according to Labor Department data, insured
unemployment accounted for 40% of total unemployment.) Nonetheless, it is likely that survey respondents are saying
that more unemployed members of their households are seeking work than would have been case in the absence of
EUC program.
10 According to BLS data from the Job Openings and Labor Turnover Survey (JOLTS), the hiring rate fell from 3.6% in
December 2007 to 3.1% in August 2009. It remained largely unchanged through February 2010, the latest month for
which data are available. Data from JOLTS first became available in December 2000. Thus, JOLTS data are available
for only one earlier recession. The hiring rate during the March-November 2001 recession and ensuing “jobless
recovery” exceeded the hiring rate reported during and after the 2007-2009 recession. The hiring rate fell from 4.4% in
March 2001 to 3.9% in November 2001, after which it reached a low of 3.5% in March 2003 before gradually trending
upward.
11 The peak rate rather than the rate at the end of recessions was used in the analysis because in five instances, the
temporary federal UC program was enacted after the trough and in two instances, it was enacted very near the trough.
Only in the latest case did Congress enact the program well before the recession’s end. According to CRS Report
RL34340, Extending Unemployment Compensation Benefits During Recessions, Congress provided a maximum of 13
weeks of benefits in response to four recessions: 1957-1958, 1960-1961, 1969-1970, and 2001. The peak long-term
unemployment rates and peak shares of long-term unemployment after the end of these recessions ranged from 0.8% to
1.5% and from 13.6% to 23.4%, respectively. Congress provided 26 weeks of benefits in response to the severe 1973-
1975 recession; the long-term unemployment peaked at 1.7% and the share of long-term unemployment peaked at
21.1% after the recession’s end. Congress next provided 16 weeks of benefits in response to the severe 1981-1982
recession; both the long-term unemployment rate (2.6%) and the share of long-term unemployment (25.7%) exceeded
the rates attained after Congress provided UC benefits for shorter (13 weeks) and longer (26 weeks) periods. Until the
2007-2009 recession, the longest period for which temporary federal UC benefits were authorized came in response to
the 1990-1991 recession. Despite the 33-week benefit period, the peak long-term unemployment rate (1.7%) and peak
share of long-term unemployment (23.3%) after the end of the 1990-1991 recession were comparable to figures
associated with shorter UC extensions.
12 When unemployment occurs because of a mismatch between the skills supplied by workers and the skills demanded
by employers, it is called structural unemployment. In contrast, cyclical unemployment results from the temporarily
(continued...)
Congressional Research Service
4
Long-Term Unemployment and Recessions
The so-called jobless recovery from the 2001 recession prompted several economists to analyze
whether increased industry restructuring had led comparatively large numbers of unemployed
workers to move from downsizing industries to industries increasing their demand for labor. No
consensus was reached, however. Groshen and Potter estimated that, according to their measure,
many more industries were undergoing restructuring during the 2001 recession and subsequent
recovery than during prior business cycles.13 In contrast, Aaronson, Rissman, and Sullivan
estimated that, based on a different measure of labor reallocation across industries, the degree of
restructuring during and after the 2001 recession was less than that during earlier periods.
Although this second group of researchers did not find evidence that the rate of permanent job
loss was atypically elevated following the 2001 recession, they noted that “the unusually high
duration of unemployment [in 2002 and 2003] is consistent with [an] increased mismatch”
between the skills of displaced workers and the skills being demanded by employers in growing
industries.14 Similarly, Figura and Wascher estimated that workers displaced from downsizing
industries experienced almost a doubling in their time spent not working (from 19 to 35 weeks)
between the 1997-1999 and 2001-2003 periods, while workers displaced from other industries
exhibited a much smaller increase (from 19 to 28 weeks). The two researchers concluded that
while it appears downsizing industries (e.g., manufacturing, information and data processing
services, broadcasting and telecommunications, publishing) were quick to layoff workers in the
early 2000s, “it took a considerable amount of time for these resources to be re-employed
elsewhere.”15
Rissman’s more recent work, covering the period from the first quarter of 1984 to the second
quarter of 2009, suggests that the impact of structural realignment on unemployed workers varies
by industry. While restructuring in the finance, insurance, and real estate industries resulted in
worker displacement, employees from these industries may be “better able to find alternative
employment in other sectors of the economy because the skills they possess are more readily
transferable to employment in other industries.” Alternatively, workers who lose jobs in the
construction, manufacturing, and transportation and utilities industries “may be less readily
absorbed into other sectors,” thereby lengthening the time spent unemployed.16
Distinguishing cyclical from structural unemployment is more than an academic exercise. If job
losses are the result of a temporary (cyclical) reduction in the demand for goods and services, the
use of fiscal and monetary policies to stimulate economic activity is appropriate. However, if job
losses also result from changes in technology, trade, or consumer taste that prompt permanent
shifts in the demand for labor across sectors, then employment policies are suitable measures to
enact as well. This matter will be revisited in “Implications for Public Policy” at the conclusion of
this report.
(...continued)
reduced demand for labor that occurs during recessions when the demand for goods and services is depressed.
13 Erica Groshen and Simon Potter, “Has Structural Change Contributed to a Jobless Recovery,” Federal Reserve Bank
of New York Current Issues in Economics and Finance, vol. 9, no. 8 (August 2003), pp. 1-7.
14 Daniel Aaronson, Ellen R. Rissman, and Daniel G. Sullivan, “Can Sectoral Reallocation Explain the Jobless
Recovery,” Federal Reserve Bank of Chicago Economic Perspectives, vol. 28, no. 2 (second quarter 2004), pp. 36-49.
15 Andrew Figura and William Wascher, “The Causes and Consequences of Sectoral Reallocation: Evidence from the
Early 21st Century,” Business Economics, vol. 45, no. 1 (2010), pp.49-68.
16 Ellen R. Rissman, “Employment Growth: Cyclical Movements or Structural Change?,” Federal Reserve Bank of
Chicago Economic Perspectives, fourth quarter 2009, pp. 40-57.
Congressional Research Service
5
Long-Term Unemployment and Recessions
Who is Likely to Experience Long-Term
Unemployment?
Knowledge of which jobless workers are more likely to suffer very long spells of unemployment
may help to identify policies that address their plight. This report compares the long-term
unemployed at the end of the three most recent recessions (2007-2009, 2001, and 1990-1991).
The report uses data from the monthly Current Population Survey (CPS), which is a household
survey conducted by the U.S. Census Bureau for BLS. The report also uses data from the Annual
Social and Economic (ASEC) Supplement to the CPS. The supplement collects information on
the sources and amounts of household and family income and benefits. (The ASEC Supplement
was formerly called the March Supplement.) 17
Because of changes in the CPS over time, not all variables analyzed in this report can be
compared across the last three recessions. For example, beginning in January 2003, the CPS
adopted new industry and occupation classification systems that substantially changed the way in
which some industries and occupations are categorized. In January 2003, the CPS also changed
the way survey respondents can respond to the question on race. While in the past survey
respondents had to select one race (e.g., white, black), respondents can now pick more than one
race, which makes meaningful comparisons over time difficult. Finally, because the CPS is a
household survey, the size of some cells (e.g., number of workers with an advanced or
professional degree) limits the detail of the analysis presented below.
Individual Characteristics
Age
Age appears to be positively related to long-term unemployment. As shown in Figure 3, older
unemployed workers are more likely to have been without jobs for longer than half a year in
2009, 2001, and 1991. In 2009, for example, 30.4% of unemployed 25-34 year olds and 33.1% of
unemployed 35-44 year olds had been out of work for at least 27 weeks, compared to 37.5% of
unemployed 45-54 year olds, 39.2% of unemployed 55-64 year olds, and 40.0% of unemployed
workers age 65 and older.
Age (a proxy for seniority) tends to protect older workers from being laid off and thereby
contributes to the comparatively lower unemployment rate of older workers. But, seniority
affords older workers little protection if companies go out of business or close plants. Once they
are let go, the higher earnings that often accompany more job tenure would tend to lead older
workers to remain unemployed longer than younger workers. Joblessness could be especially
prolonged among high-wage, high-tenure persons displaced from industries undergoing structural
17 In a 2007 report, the Congressional Budget Office (CBO) examined the extent to which unemployment is
concentrated among workers who are unemployed for more than six months. The report examined the characteristics of
workers who experienced unemployment at any time from 2001 to 2003. Although the report uses data from the
monthly CPS, it mainly uses data from the 2001 panel of the Survey of Income and Program Participation (SIPP). SIPP
is a longitudinal survey of the population of the United States and is conducted by the U.S. Census Bureau. Unlike the
CBO study, this report examines the characteristics of the long-term unemployed during recent recessions.
(Congressional Budget Office, Long-Term Unemployment, October 2007, pp. vii, 21.)
Congressional Research Service
6
Long-Term Unemployment and Recessions
change in addition to temporarily reduced product demand (e.g., workers in the motor vehicle
manufacturing industry).18
Figure 3. Percentage of Unemployed by Age Group Who are Long-Term
Unemployed, 2009, 2001, and 1991
50%
40%
30%
20%
10%
0%
2009
2001
1991
16-24
25-34
35-44
45-54
55-64
65 and over
Source: CRS analysis of data from the Current Population Survey (CPS).
Gender
Unemployment among men has been greater than among women during and shortly after
recessions since the 1980 recession. The latest recession is no exception, with men’s higher
unemployment rate related in part to their dominance of employment in more cyclically sensitive
industries (e.g., manufacturing and construction).19
A notable change across recent recessions involves the likelihood of long-term unemployment by
gender. Figure 4 shows that in 2009 and 2001, unemployed women were as likely as unemployed
men to be out-of-work for over half a year. In 1991, in contrast, unemployed women were
significantly less likely than men to experience long-term joblessness.20 The gap between their
probability of long-term unemployment narrowed from 4.8 percentage points in favor of men in
1991 to a statistically insignificant 0.5 percentage points in 2001 and 2009. Women’s increased
18 For information on the reemployment problems faced by older workers who lose long-held jobs, see CRS Report
RL33054, Older Displaced Workers in the Context of an Aging and Slowly Growing Population, by Linda Levine.
19 Aysegul Sahin, Joseph Song, and Bart Hobijn, “The Unemployment Gender Gap during the 2007 Recession,”
Federal Reserve Bank of New York Current Issues in Economics and Finance, vol. 16, no. 2 (February 2010).
(Hereafter cited as “The Unemployment Gender Gap.”)
20 Unless otherwise noted, the percentage comparisons in this report are significant at the 90% confidence level or
better. See the discussion of confidence intervals in the Appendix.
Congressional Research Service
7
Long-Term Unemployment and Recessions
attachment to the labor force may partly explain their labor market behavior becoming more like
that of men: rather than immediately withdrawing from the workforce upon being laid off,
women have become more likely to seek new jobs and to persist in their job search for a very
long time.21
Figure 4. Percentage of Unemployed by Gender Who are Long-Term
Unemployed, 2009, 2001, and 1991
50%
40%
30%
20%
10%
0%
2009
2001
1991
Men
Women
Source: CRS analysis of data from the Current Population Survey (CPS).
Educational Attainment
Another notable change concerns the probability of long-term unemployment among the
unemployed who ended their formal schooling upon attainment of a high school diploma. They
were more likely to be without jobs for longer than 26 weeks in 2009, while their risk of long-
term unemployment did not differ substantially from that of other educational groups in 2001 and
1991. (See Figure 5.) In part, this change may have occurred because major employers of high
school graduates age 25 and older were harder hit during this recession than during the two prior
recessions. In 2009, the construction (10.8%) and manufacturing (13.6%) industries employed
almost one-fourth of high school graduates in the labor force.22 Construction employment
plummeted by 21.4% between December 2007 and August 2009, compared to only 1.1% during
the 2001 recession and 7.5% during the 1990-1991 recession.23 Similarly, manufacturing
employment dropped by 14.9% during the latest recession compared to only 6.6% during the
2001 recession and 3.2% during the 1990-1991 recession.
21 Another contributory factor may be the 1994 revision of the CPS, which led to more women reporting they were
unemployed rather than not being members of the labor force.
22 Unpublished BLS data from the Current Population Survey (CPS).
23 Calculated by CRS from BLS establishment survey data.
Congressional Research Service
8
Long-Term Unemployment and Recessions
Figure 5. Percentage of Unemployed by Education Who are Long-Term
Unemployed, 2009, 2001, and 1991
50%
40%
30%
20%
10%
0%
2009
2001
1991
Less than high school
High school
Some college
Bachelor's, advanced, or professional degree
Source: CRS analysis of data from the Current Population Survey (CPS).
Industry
Seemingly unique to this recession is the significantly greater risk of long-term unemployment
among workers laid off from the information (e.g., telecommunications, publishing) and financial
activities (i.e., finance, insurance, and real estate) industries. Almost two out of five workers laid
off from the two industries did not receive a paycheck for over six months in 2009. In fact, long-
term unemployment accounted for a larger percentage of total unemployment in the two
industries than it did in the cyclically sensitive manufacturing industry (at a little more than one-
third). Nonetheless, displaced manufacturing workers consistently were among the industries
most likely to experience long-term unemployment. (See Figure 6.)
Congressional Research Service
9
Long-Term Unemployment and Recessions
Figure 6. Percentage of Unemployed by Industry Who are Long-Term
Unemployed, 2009 and 2001
50%
40%
30%
20%
10%
0%
2009
2001
Construction
Manufacturing
Wholesale & retail trade
Transportation and utilities
Information
Financial activities
Professional, business, & other services
Educational & health services
Leisure & hospitality
Public administration
Source: CRS analysis of data from the Current Population Survey (CPS).
Note: This chart does not include mining or agriculture. In 2009, each of these industries accounted for less
than 2% of total employment.
Occupation
Both in 2009 and 2001, workers displaced from management, business, and financial occupations
had a greater probability of being unemployed for more than six months compared to workers
laid off from other occupations. The relatively high wages of individuals in management,
business, and financial occupations probably contributed to their lengthy unemployment while
searching for jobs offering comparable compensation. So, too, would their industry-specific (not-
easily transferable) skills if they needed to change industries to obtain new jobs. In contrast, only
in 2009 were workers laid off from office and administrative support jobs significantly more
likely to experience long-term unemployment. The overrepresentation of office and
administrative support workers in two industries (information and financial activities) hit
uncharacteristically hard during the 2007-2009 recession might have contributed to the
occupational group’s elevated risk of long-term unemployment.24 (See Figure 7.)
24 In 2009, office and administrative support workers accounted for 13.0% of total employment. They accounted for a
larger share of employment in the information industry (18.6% ) and in the financial activities industry (24.6%),
according to CRS calculations utilizing BLS data from the CPS.
Congressional Research Service
10
Long-Term Unemployment and Recessions
Figure 7. Percentage of Unemployed by Occupation Who are Long-Term
Unemployed, 2009 and 2001
50%
40%
30%
20%
10%
0%
2009
2001
Management, business, & financial
Professional & related
Service
Sales & related
Office & administrative support
Construction & extraction
Installation, maintenance & repair
Production
Transportation & material moving
Source: CRS analysis of data from the Current Population Survey (CPS).
Note: This chart does not include farming occupations.
Household Characteristics
Unemployment affects both the individuals who are without work and their families. Workers
who become unemployed face a loss of earnings and, in some cases, fringe benefits, such as
employer contributions for health insurance or to a retirement plan. As a result, their families may
experience a loss of household income and health insurance coverage. Unemployed homeowners
may find it difficult to keep up with their mortgage payments and could face foreclosure.
At the same time, workers who become unemployed may be eligible for UC benefits, food
assistance, or cash public assistance. Under the Consolidated Omnibus Budget Reconciliation Act
of 1985 (COBRA, P.L. 99-272), they may be able to purchase health insurance through their
former employer. Workers could be eligible for a temporary subsidy to buy health insurance
under COBRA. Some workers may be eligible for a health coverage income tax credit (HCTC).
Unemployed workers may also be covered by the insurance policy of another family member or
they may be eligible for Medicaid.
This section compares workers who have been unemployed for 27 weeks or more with all
unemployed workers and all employed workers. The analysis compares “all households” with
Congressional Research Service
11
Long-Term Unemployment and Recessions
married couple households. All types of households includes married couples, single parent
households, and single individuals.
The analysis in this section is based on data from the Annual Social and Economic (ASEC)
Supplement to the monthly CPS. The supplement collects information on household earnings and
income for the previous year. The supplement also asks individuals whether they had health
insurance at any time during the previous year. The supplement is conducted at the beginning of
each year, mainly in March. The most recent data available are for 2008.
The data from the CPS supplement have some limitations in analyzing the household
characteristics of workers unemployed for more than six months. Since the supplement collects
information for the previous year (when a long-term unemployed worker may have been
employed), it may not fully capture the household characteristics of workers at the time the
survey is conducted.
In household surveys like the CPS, households tend to under report their income. Households
report their wages and salaries more accurately than their income from interest, dividends, public
assistance, or other sources. 25 Thus, the estimates in this section may understate the percentage of
individuals or households receiving different types of income or benefits.
Earnings and Income
Figure 8 shows the average real earnings of households in 2008, 2001, and 1991. Average
earnings are shown for employed workers, unemployed workers, and long-term unemployed
workers. Unemployed workers include both the long-term unemployed and other unemployed
persons. Average earnings are in constant (i.e., inflation-adjusted) 2009 dollars.
In 2008, the average household earnings of workers unemployed for more than six months were
lower than the average household earnings of employed workers. The long-term unemployed had
average household earnings that were about 55% of the average household earnings of employed
workers ($39,780 compared to $72,450). The difference in average household earnings between
the long-term unemployed and employed workers was similar in both 2001 and 1991.
25 U.S. Census Bureau, Current Population Survey, 2009 Annual Social and Economic (ASEC) Supplement, available at
http://www.census.gov/apsd/techdoc/cps/cpsmar09.pdf, p. 9-4. (Hereafter cited as U.S. Census Bureau, Current
Population Survey, 2009 Annual Social and Economic (ASEC) Supplement.)
Congressional Research Service
12
Long-Term Unemployment and Recessions
Figure 8. Average Annual Real Earnings, All Households, 2008, 2001, and 1991
(in 2009 dollars)
$100,000
$80,000
$60,000
$40,000
$20,000
$0
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Figure 9. Average Annual Real Income, All Households, 2008, 2001, and 1991
(in 2009 dollars)
$100,000
$80,000
$60,000
$40,000
$20,000
$0
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Congressional Research Service
13
Long-Term Unemployment and Recessions
Households may have income from sources other than work. For example, they may receive rent,
interest, or dividends, as well as cash income in the form of UC benefits or public assistance. In
2008, the average household income of long-term unemployed workers was approximately
$43,390. Again, this was about 55% of the average household income of employed workers
($78,350), and similar to the patterns in 2001 and 1991. (See Figure 9.)
Married couples have higher average earnings than all household types considered together. In
2008, as shown in Figure 10, married couples with at least one spouse unemployed for over six
months had average earnings of $60,210 (compared to $39,780 for all household types). The
average earnings of married couples with at least one long-term unemployed spouse were
approximately three-fifths (61%) of the average earnings of married couples in which neither
spouse was unemployed ($60,210 compared to $98,440). This percentage was slightly lower than
in 2001 (64%), but higher than in 1991 (54%).
The average income of married couples is higher than their average earnings. Figure 11 shows
that in 2008, the average income of married couples with at least one long-term unemployed
spouse was $70,370, which was about two-thirds of the average income of married couples with
no unemployed spouse ($106,130). This was the same percentage as in 2001, but higher than in
1991 (61%).
Figure 10. Average Annual Real Earnings, Married Couples, 2008, 2001, and 1991
(in 2009 dollars)
$100,000
$80,000
$60,000
$40,000
$20,000
$0
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Congressional Research Service
14
Long-Term Unemployment and Recessions
Figure 11. Average Annual Real Income, Married Couples, 2008, 2001, and 1991
(in 2009 dollars)
$120,000
$100,000
$80,000
$60,000
$40,000
$20,000
$0
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Unemployment Benefits
In 2008, almost a third (31.7%) of long-term unemployed workers received unemployment
benefits (e.g., UC, EB and EUC benefits, Trade Adjustment Assistance).26 As shown in Table 1,
this was the same as in 2001 (30.6%), but lower than in 1991 (48.5%).27
Poverty
Poverty is a common measure of economic well-being. In 2008, almost a third (31.0%) of the
long-term unemployed lived in households that had incomes below the official poverty level. This
percentage was not statistically different from either 2001 (27.1%) or 1991 (27.4%). (See Table
1.)
26 The 2009 ASEC Supplement asks survey participants the following question: “At any time during 2008 did
(name/you) receive any State or Federal unemployment compensation?” U.S. Census Bureau, Current Population
Survey, 2009 Annual Social and Economic (ASEC) Supplement, p. D-20.
27 Estimates for 2008 may not be comparable to estimates for 2001 or 1991, the official end dates of the two previous
recessions. NBER has not announced the official end date of the recent recession, but many economists believe that it
will be in the third quarter of 2009.
Congressional Research Service
15
Long-Term Unemployment and Recessions
Table 1. Percent of Employed, Unemployed, and Long-Term Unemployed by
Individual or Household Characteristic, 2008, 2001, and 1991
Labor Force Status
Characteristic
of Individual
2008 2001 1991
Individual received
Employed 2.8
3.5
5.0
unemployment compensation
Al unemployed
23.5
22.0
29.9
Long-term unemployeda 31.7 30.6 48.5
Individual lived in a household
Employed 6.3
5.8
6.3
below poverty
Al unemployed
21.0
20.5
23.5
Long-term unemployed
31.0
27.1
27.4
Someone in the household
Employed 4.1
2.6
3.8
received benefits under the
Al unemployed
15.7
13.1
18.0
Supplemental Nutrition
Assistance Program (SNAP,
Long-term unemployed
formerly the Food Stamp
program)
20.3 15.7 19.3
Individual was covered by
Employed 4.3
3.1
2.3
Medicaid
Al unemployed
12.1
11.1
11.9
Long-term unemployed
15.5
12.8
11.1
Individual received public
Employed 0.3
0.4
0.9
assistanceb
Al unemployed
2.0
3.0
7.7
Long-term unemployed
2.7
4.3
7.6
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Note: The unit of analysis is either the individual or the household. Thus, in 2008, 2.8% of persons employed at
the time of the 2009 ASEC Supplement received unemployment compensation in 2008. But 6.3% of the
employed lived in a household with income in 2008 that was below the official poverty level.
a. A person is unemployed if they are not working, have actively looked for work during the four weeks
before the CPS is conducted, and are currently available for work. The long-term unemployed are persons
who have been unemployed for 27 weeks or more.
b. Public assistance consists of cash payments under the Temporary Assistance for Needy Families (TANF)
program, state and local general assistance programs, and other forms of cash assistance.
Public Assistance and Supplemental Nutrition Assistance
Households that receive public assistance or benefits under the Supplemental Nutrition Assistance
Program (SNAP, formerly the Food Stamp program) have low incomes. In 2008, about one in
five (20.3%) long-term unemployed workers lived in a household where someone received SNAP
benefits. This was higher than in 2001 (15.7%) but statistically the same as in 1991 (19.3%).
In 2008, only 2.7% of the long-term unemployed received income from public assistance.28 This
percentage was down from an estimated 4.3% in 2001 and 7.6% in 1991. Public assistance
28 Public assistance consists of cash payments under the Temporary Assistance for Needy Families (TANF) program,
state and local general assistance programs, and other forms of cash assistance.
Congressional Research Service
16
Long-Term Unemployment and Recessions
caseloads for 1991 and 2001 are not comparable, however. The number of families receiving cash
welfare assistance declined after Congress enacted the Temporary Aid to Needy Families (TANF)
program in 1996.29 In addition, the CPS may underestimate the percentage of unemployed who
receive public assistance. A comparison of estimates from the CPS with administrative data from
the TANF program suggests that the CPS does not capture a significant portion of persons who
receive public assistance and that this underreporting has worsened over time.30 (See Table 1.)
Health Insurance Coverage
Unemployed workers may lose employer contributions for health insurance.31 They may,
however, be eligible for coverage under Medicaid, be covered by the insurance policy of a family
member, or buy insurance through their former employer or directly from an insurer.
The American Recovery and Reinvestment Act of 2009 (ARRA, P.L. 111-5) included a subsidy
for up to 15 months for unemployed workers who buy health insurance under COBRA. The
subsidy is available to workers who lost their jobs between September 1, 2008, and March 31,
2010. The subsidy consists of an employer tax credit equal to 65% of the unemployed worker’s
health insurance premium.32
Some unemployed workers may be eligible for a health coverage income tax credit (HCTC),
which may bolster the share with benefits. The credit against health insurance premiums is
available to workers who are receiving income or wage subsidies under the TAA program or are
between the ages of 55 and 64 and are receiving payments from the Pension Benefit Guaranty
Corporation (PBGC).33
In 2008, as shown in Table 1, approximately 15.5% of long-term unemployed workers were
covered by Medicaid. Statistically, this percentage was not different from either 2001 (12.8%) or
1991 (11.1%).
Figure 12 shows that an estimated 55% of long-term unemployed had health insurance coverage
at some time during 2008. This is significantly less than the 84% of employed workers who had
insurance coverage. It is also less than the percentage of long-term unemployed who had health
insurance coverage in 2001 (60.3%) but statistically the same as in 1991 (57.3%).
29 See CRS Report RL32760, The Temporary Assistance for Needy Families (TANF) Block Grant: Responses to
Frequently Asked Questions, by Gene Falk.
30 See CRS Report RL30797, Trends in Welfare, Work, and the Economic Well-Being of Female-Headed Families with
Children: 1987-2008, by Thomas Gabe.
31 For information on employer-provided fringe benefits, see CRS Report RL33835, Real Earnings, Health Insurance
and Pension Coverage, and the Distribution of Earnings, 1979-2007, by Gerald Mayer.
32 Both the House and Senate have passed legislation that would extend the eligibility period for the COBRA subsidy.
On December 16, 2009, the House passed H.R. 2847, the Jobs for Mainstreet Act, 2010, which would extend the
COBRA subsidy to workers who lose their jobs through June 30, 2010. On March 10, 2010, the Senate passed H.R.
4213, the American Workers, State and Business Relief Act, which would extend eligibility for COBRA subsidies to
workers who lose their jobs through December 31, 2010. See CRS Report R40165, Unemployment and Health
Insurance: Current Legislation and Issues, by Janemarie Mulvey.
33 For more information on the HCTC, see CRS Report RL32620, Health Coverage Tax Credit, by Bernadette
Fernandez. Beginning in 2014, the Patient Protection and Affordable Care Act (H.R. 3590, P.L. 111-148) would
provide subsidies to individuals and families with incomes between 133% and 400% of the federal poverty level who
buy health insurance through new insurance exchanges. CRS Report R40942, Private Health Insurance Provisions in
Senate-Passed H.R. 3590, the Patient Protection and Affordable Care Act , by Hinda Chaikind et al.
Congressional Research Service
17
Long-Term Unemployment and Recessions
Figure 12. Percentage of Individuals with Health Insurance Coverage,
2008, 2001, and 1991
100%
80%
60%
40%
20%
0%
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Number of Earners in the Household
Many households have more than one member who is employed. If a worker becomes
unemployed, other household members may continue to work, work more hours, or enter the
workforce. In 2008, however, one-fourth of the households of the long-term unemployed had no
earned income. As shown in Table 2, the percentage with no earners in 2008 was higher than in
2001 (17.2%) but not statistically different from 1991 (22.5%).
Table 2. Number of Earners in the Households of the Employed, Unemployed, and
Long-Term Unemployed, 2008, 2001, and 1991
(percent)
Labor Force Status of
Individual
Number
of
Earners
2008 2001 1991
All Households
Employed
No earnersa
1.0 1.0 0.9
One
earner
47.6 45.9 44.1
Two earners
41.5
42.3
42.7
Three or more earners
9.9
10.9
12.3
Unemployed
No earners
9.9
8.2
10.8
One
earner
51.8 51.6 47.5
Two earners
30.8
32.8
34.5
Three or more earners
7.4
7.4
7.2
Congressional Research Service
18
Long-Term Unemployment and Recessions
Long-Term Unemployed
No
earners
24.9 17.2 22.5
One
earner
48.4 51.9 41.2
Two earners
21.6
25.9
29.8
Three or more earners
5.1
4.9
6.5
Married Couples Only
Employed
No
earners
0.4 0.3 0.5
One
earner
18.0 17.6 20.1
Two earners
65.7
64.8
60.6
Three or more earners
15.9
17.3
18.9
Unemployed
No
earners
2.1 2.2 4.1
One
earner
26.9 23.5 27.6
Two earners
57.2
61.0
55.1
Three or more earners
13.8
13.3
13.2
Long-Term Unemployed
No earners
4.2
7.4
12.2
One
earner
40.3 33.6 31.5
Two earners
43.3
49.9
44.4
Three or more earners
12.3
9.1
12.0
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
Note: Percentages may not add to 100% because of rounding.
a. The ASEC Supplement collects earnings information for both wage and salary workers and self-employed
persons. Some self-employed workers report negative earnings if the losses from their business are greater
than their wages. In this report, workers with negative earnings are not counted as earners. Only workers
with positive earnings are counted as earners.
If a single parent or single individual becomes unemployed, there may not be anyone else in the
household with earnings. Among married couples, if one spouse becomes unemployed, the other
spouse (or someone else in the household) may be working or may go to work.
Among married couples in 2008, only 4.2% of the families of the long-term unemployed had no
earned income in 2008. This percentage was not statistically different from 2001 (7.4%), but was
lower than in 1991 (12.2%). The decline since 1991 may be due to a number of factors, including
the lack of data for 2009, an increase in the labor force participation rate of married women, a
change in the average age of married persons, and changes in the size and composition of married
couple households.34
Homeownership
Employed workers are more likely than unemployed workers to own their own home.
Nevertheless, a majority of unemployed workers regardless of unemployment duration are
34 From 1991 to 2008, the labor force participation rate of women increased from 57.4% to 59.5%. For men, it declined
from 75.8% to 73.0%. Bureau of Labor Statistics, Labor Force Statistics from the Current Population Survey, available
at http://stats.bls.gov/cps/.
Congressional Research Service
19
Long-Term Unemployment and Recessions
homeowners. Figure 13 shows that, in 2008, 72% of employed workers owned their own home,
compared to 57% of unemployed workers and 58% of the long-term unemployed.35
From 1991 to 2008, homeownership fell among workers who were unemployed for more than six
months (from 61% to 58%). However, homeownership rose among employed workers (from 69%
to 72%) and for all unemployed workers (from 54% to 57%). From 2001 to 2008, there were no
significant changes in the percentages of employed, unemployed, or long-term unemployed
workers who owned their own homes.
Job loss and a reduction in household income may cause some borrowers to become delinquent
on their mortgage loans. A borrower who has missed three or more mortgage payments is
generally considered to be in default. After a default, lenders may take a number of steps—
including rescheduling loan payments or restructuring a loan—before proceeding to foreclosure
(i.e., a lender may recover losses on a loan by repossessing the home).36 At the end of 2009, the
percentage of mortgages that were 90 or more days past due accounted for over half of all
delinquent loans, double the percentage from a year earlier.37 For this report, the most recent data
on homeownership are for 2008. Data for 2009 may or may not show a change in homeownership
from 2008 for either unemployed workers or the long-term unemployed.
Figure 13. Percentage of Individuals Who are Homeowners, 2008, 2001, and 1991
100%
80%
60%
40%
20%
0%
2008
2001
1991
Em ployed
Unem ployed
Long-term unem ployed
Source: CRS analysis of data from the Annual Social and Economic (ASEC) Supplement to the Current
Population Survey (CPS).
35 Most households that do not own their homes are renters. A small percentage of households neither own nor rent.
They receive nonmoney income from their employer in the form of rent free housing.
36 See CRS Report RL34232, The Process, Data, and Costs of Mortgage Foreclosure, coordinated by Darryl E. Getter,
and CRS Report R40210, Preserving Homeownership: Foreclosure Prevention Initiatives, by Katie Jones.
37 Renae Merle, “Mortgage Delinquency Rate Slows in Fourth Quarter,” Washington Post, February 20, 2010.
Congressional Research Service
20
Long-Term Unemployment and Recessions
Implications for Public Policy
Recessions are generally characterized by a drop in consumer demand and a decline in
investment. Consumers buy fewer cars and homes, and cut back on purchases from retail stores,
for example. For their part, businesses invest less in new buildings and equipment. A result of the
decline in aggregate demand is an increase in unemployment. Fiscal and monetary policy are the
macroeconomic responses to a recession. Fiscal policy consists of changes in government
spending and taxes. Monetary policy consists of actions by the Federal Reserve to affect the
supply of money and interest rates.38
The strength of an economic rebound largely determines the pace of net job growth (i.e., the
extent to which additions exceed separations from employer payrolls) and reduction in
unemployment, including long-term unemployment. In the early stages of an economic recovery,
increased demand for goods and services first slows the flow of workers into unemployment.
While short-term unemployment consequently falls, the average duration of unemployment
increases for some time following a recession’s end. Once employers become confident that an
economic upturn is firmly in place, they go beyond restoring the hours worked by employees
whom they retained during the recession to hire additional workers.
But, a return to economic growth may not be sufficient to fully overcome the reemployment
problems of the long-term unemployed. Employers may rank job applicants by their duration of
unemployment and hire from the front of the queue (i.e., the short-term unemployed) because
they consider lengthy unemployment to be a signal of poor worker quality (i.e., low productivity).
In effect, long-term unemployment stigmatizes workers. Firms may also be reluctant to hire the
long-term unemployed because they believe the group’s skills have atrophied during their lengthy
time away from the workplace. In addition, businesses may discriminate against older workers
who, as previously described, are more likely than younger workers to be among the long-term
unemployed.
A variety of employment policies might be considered to ameliorate these labor market
difficulties. Offering wage subsidies to firms that hire the long-term unemployed compensates
employers for the group’s actual or perceived skill shortcomings.39 So, too, does providing a
training subsidy to companies that hire workers with arguably deteriorated skills due to their
lengthy bout of unemployment.40 Passage of a temporary public jobs program early in a recession
may prevent the skill erosion potentially induced by long-term unemployment. The latter was one
of the rationales for enacting job creation programs during the Great Depression.41 And,
38 Fiscal policy (i.e., budget surpluses or deficits) may also affect interest rates.
39 For example, P.L. 111-147 includes a provision temporarily reducing the payroll tax of employers who hire persons
unemployed for 60 days or more. (For additional information, see CRS Report R41006, Unemployment: Issues and
Policies, by Jane G. Gravelle, Thomas L. Hungerford, and Marc Labonte.) Other legislation was introduced to add the
long-term unemployed to the groups for whom an employer can claim the Work Opportunity Tax Credit. (For
additional information, see CRS Report RL30089, The Work Opportunity Tax Credit (WOTC), by Linda Levine.)
40 For information about on-the-job training under the Workforce Investment Act, see CRS Report R41135, The
Workforce Investment Act and the One-Stop Delivery System, by David H. Bradley.
41 For additional information on two of these programs, see CRS Report R41017, Job Creation Programs of the Great
Depression: the WPA and the CCC, by Linda Levine.
Congressional Research Service
21
Long-Term Unemployment and Recessions
employers might be discouraged from failing to hire older displaced workers through stepped up
enforcement of workplace discrimination laws.42
A return to economic growth also may be insufficient to fully overcome the reemployment
problems of the long-term unemployed given the history of structural change coinciding with
cyclical downturns. Before the back-to-back recessions of the early 1980s, for example, workers
in the steel industry were about half as likely as the average worker to experience long-term
unemployment. With the advent of these recessions, long-term unemployment among
steelworkers became “among the worst of any worker group. …Considerable evidence indicates
that the Nation’s steel industry is suffering from some basic problems quite unrelated to cyclical
declines in demand.”43
One approach that has been suggested to reemploy workers displaced from jobs in restructuring
industries is wage insurance. Such a program may encourage high-tenure, high-wage workers to
accept lower paying jobs in growing industries by compensating them for the difference in wages
from their new and old jobs. Under the Alternative Trade Adjustment Assistance (ATAA)
program, older workers who become reemployed within 26 weeks of being laid off can receive
one-half of the difference in earnings from their former and new jobs up to a maximum of
$10,000 over two years.44
Another potential policy response is providing reemployment bonuses to unemployed workers
who accept new jobs within a given time frame, thereby shortening their spell of unemployment.
Experiments were conducted in a few states in the 1980s to evaluate the efficacy of variously
designed reemployment bonus programs. The evaluations and subsequent reanalyses generally
concluded that offering bonuses to UC recipients shortened their length of benefit receipt,
especially when UC recipients also were offered job search assistance and when the bonus
programs were targeted at those UC recipients most likely to exhaust their benefits.45
If restructuring industries are concentrated geographically, then economic development and
relocation assistance may be appropriate measures as well. But, these strategies do not necessarily
focus on assisting the long-term unemployed. States typically take the lead in efforts to attract
growing industries to reemploy their laid off residents. To assist the workers and communities
adversely affected by the auto industry’s restructuring, however, President Obama created the
position of Director of Recovery for Auto Communities and Workers within the Department of
Labor to leverage available federal resources (e.g., ARRA funds). In contrast, providing
assistance to unemployed workers to encourage them to take jobs in other areas is not often
mentioned today. Relocation assistance currently is one of the benefits offered to TAA-eligible
workers, those receiving Department of Labor national emergency grants, and adult and
dislocated workers served by the Workforce Investment Act.
42 For additional information, see CRS Report RL34652, The Age Discrimination in Employment Act (ADEA): A Legal
Overview, by Jody Feder.
43 Philip L. Rones, “Recent Recessions Swell Ranks of the Long-Term Unemployed,” Monthly Labor Review, February
1984, pp. 25-29.
44 For additional information on the ATAA program, see CRS Report RL34383, Trade Adjustment Assistance (TAA)
for Workers: Current Issues and Legislation, by John J. Topoleski.
45 For additional information, see CRS Report RL31825, Personal Reemployment Accounts: Results from Bonus
Experiments, by Linda Levine and Ann Lordeman.
Congressional Research Service
22
Long-Term Unemployment and Recessions
Lastly, workers laid off due to structural changes in the economy may not possess the skills
needed to get jobs in expanding industries. While these individuals can undertake education and
training to acquire the requisite skills, these activities may not result in reemploying the long-term
unemployed as quickly as some of the approaches discussed above.
Congressional Research Service
23
Long-Term Unemployment and Recessions
Appendix. Data and Methodology
The analysis in this report is based on data from the Current Population Survey (CPS), which is a
household survey conducted by the U.S. Census Bureau for the Bureau of Labor Statistics (BLS).
The monthly CPS is the source of the national unemployment rate and other labor market
information. The survey is representative of the civilian, noninstitutional population. The sample
does not include persons living in institutions (such as mental hospitals, nursing homes, or
correctional facilities). The monthly survey does not include individuals who are on active duty in
the military. The monthly sample includes approximately 60,000 households.46
Each year, the CPS conducts the Annual Social and Economic (ASEC) supplement to the monthly
CPS. The supplement collects information on the types and sources of family income, health
insurance coverage, and other household information. The supplement collects information for
the previous year. The sample for the ASEC supplement is representative of the civilian,
noninstitutional population of the United States. The sample for the supplement includes
members of the Armed Forces living in civilian housing units on a military base or in a household
not on a military base. The 2009 supplement collected information from about 76,200
households.47
In the supplement, earnings include wage and salary income and income from self-employment.
Wages and salaries include earnings from self-employment, tips, commissions, and cash bonuses.
Total income includes earnings, interest and dividends, social security and pension benefits,
public assistance, unemployment and worker’s compensation, alimony and child support, and
other types of money income. Money income does not include in-kind transfers for food, housing,
healthcare, or energy assistance. Money income does not include capital gains and is income
before taxes or other deductions. 48 In this report, total family earnings and income consist of the
sum of earnings and income of all persons in a household or married couple family.
In the CPS, a family is defined as a group of two or more persons who are living together and
who are related by birth, marriage, or adoption.49 In the CPS, a family may include a related
subfamily, such as a married son or daughter or an unmarried son or daughter with a child. The
CPS defines single individuals as either “nonfamily householders” or “secondary individuals.”
Following the CPS definition of a family, in this report married couples and single parent families
may include one or more related subfamilies. A primary family and an unrelated subfamily living
in the same household are treated as two separate families. Unmarried parents are treated as a
single parent and a single individual. Unmarried couples without children are treated as two
single individuals.
The analysis in this report includes persons in the labor force, either employed or unemployed,
who are ages 16 or over. The analysis of household and family earnings and income includes
households and families with positive earnings or income.
46 Bureau of Labor Statistics, “Labor Force Data Derived from the Current Population Survey,” Handbook of Methods,
available at http://www.bls.gov/opub/hom/pdf/homch1.pdf, pp. 1-2.
47 U.S. Census Bureau, Current Population Survey, 2009 Annual Social and Economic (ASEC) Supplement, pp. 1-1, 9-
3, G-3.
48 Ibid., p. 9-4.
49 Ibid., p. 9-2.
Congressional Research Service
24
Long-Term Unemployment and Recessions
Topcoding
To protect the confidentiality of survey participants, the CPS assigns an income amount to higher
income persons. Changes in these amounts, or topcodes, over time can affect the observed trend
in real income and earnings. For example, in 1991, individual earnings were topcoded at $99,999;
in 2001, earnings were topcoded at $150,000; and in 2008, earnings were topcoded at $200,000.
Since 1996, for persons with earnings above the topcoded amounts, the amount of earnings
reported in the public CPS files is the average earnings of workers with similar characteristics.
Average earnings are calculated for persons based on gender, race, ethnicity, and whether or not a
person works full-time, year-round. Since 1999, for persons with incomes above the topcoded
amounts, the public CPS files report the average income of persons with topcoded income. For
consistency, for the years 2001 and 2008, this report uses average topcoded earnings and income
as reported in the CPS and, for 1991, average earnings and income from a report published by the
National Bureau of Economic Research (NBER).50
Confidence Intervals
Estimates based on survey responses from a sample of households have two kinds of error:
nonsampling and sampling. Examples of nonsampling error include information that is
misreported and errors made in processing collected information. Sampling error occurs because
a sample, and not the entire population, of households is surveyed. The difference between an
estimate based on a sample of households and the actual population value is known as sampling
error. When using sample data, researchers typically construct confidence intervals around
population estimates. Confidence intervals provide information about the accuracy of estimated
values. With a 90% confidence interval and repeated samples from a population, 90% of intervals
will include the average estimate of the population characteristic.
Author Contact Information
Gerald Mayer
Linda Levine
Analyst in Labor Policy
Specialist in Labor Economics
gmayer@crs.loc.gov, 7-7815
llevine@crs.loc.gov, 7-7756
50 Jeff Larrimore, Richard V. Burkhauser, Shuaizhang Feng, and Laura Zayatz, Consistent Cell Means for Topcoded
Incomes in the Public Use March CPS (1976-2007), NBER, Working Paper 13941, April 2008, available at
http://www.nber.org/papers/w13941, p. 46. U.S. Census Bureau, Current Population Survey, 2009 Annual Social and
Economic (ASEC) Supplement, p. 5-1.
Congressional Research Service
25