Do Extended Unemployment Benefits Lengthen Unemployment Spells?

A. Normal and Extended Benefits

UI benefits are normally available for 26 weeks in the United States under the joint federal-state Unemployment Compensation program established under the Social Security Act of 1935. Unemployed individuals are eligible to receive benefits if they lost a job through no fault of their own (typically a permanent or temporary layoff) and they meet state-specific minimum requirements regarding work history and wages during the 12-to 15-month period preceding job loss. Availability for work and active job search typically are required for ongoing receipt of UI benefits, although the exact rules vary across states. In most states, UI benefits equal about half of the claimant’s preseparation weekly wage, up to a maximum that varies substantially across states.

Normal UI benefits periodically are supplemented and extended during episodes of economic distress through a combination of permanent and temporary legislation.² The federal Extended Benefits (EB) program, permanently authorized beginning in 1970, provides up to 20 weeks of additional unemployment compensation for unemployed individuals who lost jobs in states where the level and change in the state unemployment rate is above a specified threshold. The thresholds or triggers are state-specific but most commonly are based on an overall unemployment rate of 6.5 percent (for a 13-week extension) or 8.0 percent (for 20 weeks) combined with a 10 percent increase in the unemployment rate over the previous two years. The EB program has been supplemented by temporary programs that have been used eight times since 1958, with the most recent episode beginning in 2008. We focus on the two episodes of UI extensions since 2002.³

The severity of job loss and persistent labor market weakness during and after the recession of 2007–2009 resulted in an unprecedented expansion of UI benefit availability and takeup. Between mid-2008 and late 2009, a set of expansions raised the availability of UI benefits up to a maximum of 99 weeks in many states. A similar but much more limited extension of UI benefits occurred through the Temporary Extension of Unemployment Compensation (TEUC) legislation that was effective from March 2002 through early 2004. A maximum of 72 weeks of total benefits were phased in during this period. We describe the timing of both sets of expansions in detail in the appendix.

As suggested by this discussion and the detailed description in the appendix of the TEUC (2002–2004) and EUC (2008–forward) programs, the timing of the extended UI triggers and consequent maximum duration of UI eligibility varied substantially across states and over time. Different states surpassed the trigger levels for EB and TEUC / EUC availability at different times; some states never achieved the unemployment rates necessary for the complete extensions; and states saw available weeks rolled back as labor market conditions improved in 2011–12 and also as a result of the legislated rollback of maximum weeks available through the EUC program in late 2012.

Figure 1 illustrates the variation in eligibility for extended UI over time (years 2000–2012) based on the various programs in effect. Panel A displays the maximum and minimum number of total UI weeks available across states, and Panel B displays the average and standard deviation of the distribution of total weeks of UI available across unemployed individuals (measured using a sample of all individuals identified as unemployed and eligible to receive UI in the CPS microdata; see our definition of eligibility below). The spread between the maximum and minimum number of weeks was similar between the most recent episode and the preceding episode in the early 2000s at about 26 to 27 weeks. However, the number of states at or near the minimum in the recent episode, and their size, was much smaller than in the preceding episode. This is reflected in Panel B, which shows that the average weeks of total UI eligibility reached about 96 in late 2009, implying that the typical unemployed individual was located in a state in which maximum UI eligibility was 99 weeks. In the early 2000s, maximum weeks of eligibility reached 72, but few states triggered on to the maximum extensions and only about 13 additional weeks of UI beyond the normal 26 were available to the typical unemployed individual. The standard deviation displayed in Panel B (on a separate scale, on the right side of the graph) indicates that the dispersion in total weeks available was only slightly higher in the recent episode than in the preceding one, implying that there is a similar degree of cross-state variation used for our estimates in both episodes. Panel B shows a sharp drop in 2012 in the average number of weeks of UI for which unemployed individuals qualify, as implied by the discusion of the legislation in the appendix.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Variation in Total Weeks of UI Available

Notes: Panel B series calculated using monthly CPS observations (weighted) for UI-eligible unemployed individuals (see Section IVA).

Figure 2 displays the pattern of UI recipiency over our sample frame based on publicly available administrative data on total UI claims. This figure illustrates the large expansion of UI receipt during the recent recession and subsequent reduction as the labor market recovery proceeded. The weekly flow of new UI claims peaked at about 660,000 in early 2009 (slightly below the peak of nearly 700,000 reached in late 1982; not shown). As of early 2013, new UI claims had declined to nearly their prerecession level. In addition to the weekly flow of new UI claims, two series for the level of ongoing UI claims are displayed in Figure 2: (1) regular UI claims (26 or fewer weeks), and (2) regular UI claims plus UI claims available through extensions. After peaking in late 2009 and early 2010, the level of both series had declined by about half as of early 2013. The sharp, temporary drop in mid-2010 corresponds to the suspension period of June–July 2010 (discussed in the appendix). A much smaller but still substantial number of extended benefit claimants also is evident during the earlier episode of UI extensions from 2002–2004.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Unemployment Insurance Claims (Regular and Extended)

Notes: Gray areas denote NBER recession dates.

B. Past Research on UI Disincentive Effects

Standard search models of unemployment imply that UI payments are likely to reduce job finding and prolong unemployment spells for eligible individuals, and a voluminous empirical literature exists that attempts to quantify such effects. Much of this research has focused on unemployment exit rates around the time of UI benefit exhaustion, often using administrative data on UI recipients (Moffitt 1985, Katz and Meyer 1990, Meyer 1990, Card and Levine 2000, Jurajda and Tannery 2003). The findings from this research generally reveal the expected disincentive effects of UI availability on unemployment exits. Some analysts have relied on such estimates to simulate the likely effects of extended UI on unemployment durations and the unemployment rate during the recent downturn (Fujita 2010a, Mazumder 2011). However, the estimated magnitude of such effects has varied widely across different studies. Recent research suggests that the disincentive effects of extended UI are likely to be muted relative to the effect implied by earlier research based on the exhaustion of regular UI benefits. This may be due to the decreased reliance on temporary layoffs over the past few decades and the unusually weak labor market conditions prevailing during the recent extension episode (Katz 2010; Landais, Michaillat, and Saez 2010; Kroft and Notowidigdo 2011).

Given these concerns about relying on past research findings to assess disincentive effects of extended UI in the recent episode, some researchers have relied on concurrent labor market data to directly assess the impact of extended UI. These analyses have focused on comparing unemployment durations for unemployed individuals distinguished by their likely eligibility for UI (Valletta and Kuang 2010) and examining changes in transition rates out of unemployment measured across duration spans that are distinguished by their exposure to extended UI (Fujita 2010b, Howell and Azizoglu 2011). Such studies provide broad empirical guidance about the likely impacts of extended UI. However, they are not based on formal statistical analyses that attempt to isolate the direct effects of UI extensions while controlling for the effects of labor market conditions and individual characteristics that are related to extended UI eligibility and receipt. As such, it is difficult to conclude that the relationships between extended UI and exit from unemployment found in these papers are causal.

We build on the existing literature by directly exploiting variation in the duration of extended UI available in different months in different states. In addition, our matched CPS data enable us to incorporate detailed controls for state economic conditions and individual characteristics into the analyses.

A. A Competing Risk Model

We apply the probit model developed here to estimate the probability that an individual exits unemployment regardless of the subsequent labor force state. It is also interesting to explore how extended benefits affect both the probability of exiting unemployment to a job (employment) and the probability of exiting unemployment to leave the labor force (cease searching). A competing risk framework is natural for this purpose and can be implemented as a straightforward generalization of the discrete choice hazard model outlined above. The key is to define cause-specific hazard functions and use them in the same likelihood context.

In the competing risk framework, there are two durations associated with each spell of unemployment:

1. Time until exit to employment. Call this time the UE duration.

2. Time until exit to NILF. Call this time the UN duration.

There are three types of spells in the data.

1. Spells that end in employment. For these spells, we observe the end of a UE spell. The UN spell is censored at the duration of the UE spell.

2. Spells that end in exit to NILF. For these spells, we observe the end of a UN spell. The UE spell is censored at the duration of the UN spell.

3. Spells that do not end during the observation period. For these spells, both the UE and UN durations are censored at the last observed unemployment duration.

We estimate three versions of the unemployment exit model: (1) exit to either employment or NILF, (2) exit to employment, and (3) exit to NILF in order to determine the effect of extended benefits on each of these exit rates.

B. Identifying the Effects of Extended Benefits

Our hazard model of the exit from unemployment includes fixed effects for each state and each month along with individual duration, demographic characteristics, and measures of local economic conditions. The effects of extended UI are identified by exploiting differences in UI availability at the individual level, which are determined by the comparison of individual unemployment duration with the number of weeks of UI available in that individual’s state at the same point in time (measured at a monthly frequency). These effects are estimated conditional on individual characteristics and differences in local labor market conditions. Given that UI extensions and rollbacks are triggered by deterioration or improvement in state labor market conditions, including the measures of local economic conditions in the model is crucial for this identification scheme.

Consider this specific example of identifying information. Between April and May of 2012, maximum available weeks of UI benefits fell from 99 weeks to 73 weeks in Washington State and from 99 weeks to 89 weeks in Oregon. This was due to legislative changes and idiosyncratic features of prior and contemporaneous labor market conditions in those states. During this time, the actual unemployment rates were stable at 8.4 percent in Washington State and 8.8 percent in Oregon. Consider two individuals, one in each of these states, who had been unemployed for 78 weeks (18 months) as of April 2012 and who were observationally equivalent more generally. In our model, both individuals have UI available to them in April 2012 because their unemployment duration is less than their maximum eligiblity in that month. If they remain unemployed in the subsequent month, however, their unemployment duration will reach 82–83 weeks. At that time, the individual in Oregon will still have UI available, while the individual in Washington State will not. Our model will uncover an increase in unemployment duration due to extended UI if the individual in Washington State is more likely to exit unemployment between April and May of 2012 than is the individual in Oregon (conditional on their observational equivalence and unrestricted state and time effects). In this specific example, the difference in unemployment rates between the two states is small, and its impact on job prospects and search behavior for the two individuals is captured in our model by the local labor market controls (polynomials in the state unemployment rate and rate of payroll job growth).

It is important to note that, because our empirical models include indicator variables for each of the first six months of unemployment, the EB availability indicator captures the effects of extended benefits only for individuals unemployed for at least six months. In other words, the specification excludes any potential effects of EB availability on search behavior for the short-term unemployed. Our specification was chosen to capture the rapid decline in unemployment exit rates during the first six months of unemployment and to avoid the estimation biases that might result from a parametric specification of duration dependence. The resulting restriction is innocuous: Rothstein (2011) found that the recent episode of extended benefits had meaningful effects on exit rates only for individuals unemployed for at least six months. More generally, our specification enables separate estimation of the effects of UI remaining and unemployment duration at the individual level.

A. Defining the Relevant Sample: UI Eligibility

The most appropriate sample for analysis of the effect of UI benefits is unemployed individuals who are eligible to receive UI. However, there is no direct information in the CPS on UI eligibility, hence we rely on a proxy measure based on the reported reason for unemployment.

Unemployed individuals who report having lost a job as the reason for unemployment are, in principle, eligible to receive UI. In contrast, individuals are, in principle, not eligible to receive UI if they report voluntarily leaving a job, new entry, or reentry into the labor force as the reason for unemployment. Although losing a job is a necessary condition for being eligible to receive UI benefits, it is not sufficient. For example, a worker who reports a job loss may not have sufficient previous employment experience to qualify for unemployment insurance or may have been fired for cause.⁸

Because we do not have the detailed work-history information needed to impute eligibility, we proceed by classifying unemployed job losers as the UI-eligible group that we use in our analysis. Job leavers and new labor force entrants are classified as UI-ineligible and are not included in our analysis. Later, we present a “placebo” analysis of the effect of extended benefits on exit from unemployment for the UI-ineligible sample (job leavers and new entrants to the labor force). We expect the effects of extended UI for this group to be smaller than for UI-eligible individuals. However, after conditioning on the other variables in our model, there may still be variation in economic conditions across states and over time that is correlated with the availability of extended benefits. If so, this omitted variable problem could cause us to falsely attribute an effect of extended UI on unemployment exits for the ineligible group. Such a finding would imply that our estimates of the effect of extended UI on the UI-eligible group may be too large and would represent an upper bound on the true effect. Working in the opposite direction, there may be spillover effects on job search and job finding from eligible to ineligible individuals that would show up as a positive relationship between extended UI and exit from unemployment for the UI-ineligible group.⁹

Table 1 provides supporting evidence for our working definition of UI eligibility. Each March, the regular monthly CPS is accompanied by an extensive set of supplemental questions regarding income receipt in the prior calendar year; UI is separately identified as an individual income source. The rotating sample structure of the CPS enables matching of observations on unemployed individuals for selected months in year t with the information on their income receipt in year t recorded in the year t + 1 March supplement. Based on this match, Table 1 provides a breakdown of UI income receipt (percent reporting positive UI income) by measured eligibility status for individuals who are unemployed in March or December of the calendar year corresponding to reported income receipt for income years 2005–2011. The percentage of the UI-eligible who report receiving UI income reached about 50 percent in 2009 and declined somewhat in 2010–11. The percentage of the UI-ineligible who report receiving UI income is much lower, usually at about 5–10 percent, with a few higher values recorded. Because individuals may be subject to multiple unemployment spells over a calendar year, an unknown proportion of individuals identified as ineligible in a particular month may have received UI income based on a separate spell of unemployment that year for which they were eligible for UI receipt based on their stated reason for unemployment. On balance, we interpret these figures as suggesting that our eligibility indicator is strongly correlated with actual eligibility status, although takeup of unemployment insurance by those eligible is clearly not universal.

B. Sample Construction: Matching the Monthly CPS

Given its rotation group structure, the CPS interviews a given address (housing unit) for four consecutive months, does not interview for eight months, and reinterviews the address for four more months. In other words, over a 16-month period, the household is surveyed for four months, left alone for eight months, and surveyed for four more months. This sample structure allows us to match households in consecutive months up to three times. Failures to match primarily occur when a household moves to a new housing unit between interviews. This generally occurs less than 5 percent of the time.

One key concern with regard to use of the matched data is the likelihood of spurious transitions in labor force status, particularly spurious exits from unemployment, that can lead to an overestimate of the probability of exit from unemployment (Abowd and Zellner 1985, Poterba and Summers 1986, 1995). As demonstrated by Rothstein (2011), the unemployment durations implied by the frequency and duration structure of exits from unemployment are much lower than the reported durations of spells in progress in the cross-section.¹⁰ This difference in distributions is due in part to the length-biased sampling inherent in an analysis of durations of spells in progress in a cross-section. But the difference is also the result of the likely presence of spurious transitions. These spurious transitions are potentially problematic for the estimation of our models because errors in the identification of exit from unemployment could bias seriously the estimates of the effects of extended benefits.

Following Rothstein (2011), we address this problem through direct adjustments to reported transitions following specific patterns that are indicative of misclassification. In particular, for individuals who report a transition out of unemployment in month one followed immediately by an entry back to unemployment in month three, we recode the intervening month as a continuation of the initial unemployment spell. We do this whether the unemployment exit is due to job finding or labor force withdrawal. In other words, letting U represent unemployment, E employment, and N out of the labor force, we recode three-month transition patterns of UEU and UNU to UUU, and we retain both of the resulting UU “transitions” in the matched data. We recoded 3,016 UEU transitions and 7,119 UNU transitions (a total of 10,135 transitions) in our estimation samples in this way.¹¹

Some support for the idea that UEU and UNU transitions are likely to be spurious is that reported unemployment durations in the third month of observed UEU and UNU transitions is, on average, far greater than one month (average of 5.7 months, median of three months). As noted by Elsby et al. (2011), it is likely that individuals’ reported unemployment duration reflects the time elapsed since the loss of a salient job, which is likely the one that enabled them to qualify for UI benefits. The reported durations are therefore much more likely to capture the duration of a spell of ongoing UI receipt than are the durations implied by reported unemployment exits, which may reflect stopgap jobs and temporary labor force withdrawals in addition to direct reporting errors (see also Poterba and Summers 1995).

Figure 3 contains a plot of average (over the 2000–2012 period) exit rates from unemployment by duration of unemployment. Two exit rates are presented: (1) the observed exit rate and (2) the exit rate adjusted by recoding UNU and UEU sequences to UUU (no exit). The adjustment substantially reduces the exit rates from unemployment, implying an increase in the survivor function and associated unemployment durations.¹²

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Exit Rate from Unemployment, Observed and Adjusted (UEU,UNU —> UUU)

Source: Authors’ calculations from matched CPS data, January 2000–October 2012.

Imposing this adjustment to observed transitions requires restricting the set of observations we use to the first two of each set of four consecutive CPS rotation groups so that we have at least two subsequent matched observations. We impose this restriction and the transition adjustment for all of our analysis samples. As such, although we use CPS data through December 2012, our final observation is for October 2012. In addition, to ensure valid matches of individuals across months, we dropped a small number of observations for which reported age, gender, race, and educational attainment are not consistent across months (for example, age changes by more than one year). We limit our sample to individuals ages 20–64.

Table 2 contains, for each period and by eligibility status, counts of the number of spells that end with employment, end with NILF, and are right censored. When considering the competing risk model, (1) the number of censored UE spells is the sum of the number of spells that end in NILF and the number that do not end within the observation period and (2) the number of censored UN spells is the sum of the number of spells that end in employment and the number that do not end within the observation period.

Our matched CPS sample covering January 2000 through February 2005 contains 75,089 monthly observations on 54,678 spells of unemployment for workers ages 20–64. The analogous sample for January 2007 through October 2012 contains 126,062 monthly observations on 86,869 spells of unemployment. In order to examine the representativeness of our matched sample, we compared the reported unemployment duration in our sample, based on the survey responses regarding in-progress spells, with published statistics on unemployment duration from the U.S. Bureau of Labor Statistics (BLS). Figure 4 contains plots of three statistics from the two distributions of unemployment durations. These are mean duration, median duration, and the fraction of long-term unemployed (at least six months). While the BLS series is seasonally adjusted and our series is not, it is clear that there is little difference between our sample of unemployment durations from the matched CPS and the BLS reported statistics from all sampled unemployment spells in the CPS. The notable exception is average duration, for which the reported BLS series has systematically higher values than our series late in the sample frame. This pattern arises due to the increase in the survey top-code value of unemployment duration beginning in early 2011, which applies to the data files used to compute the BLS series but not the public use microdatafiles.¹³

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Unemployment Duration, BLS and Matched CPS Data

Notes: From U.S. BLS (seasonally adjusted) and authors’ tabulations of matched CPS data (weighted). Based on reported duration of in-progress spells. Gray areas denote NBER recession dates.

Figure 4 also provides a sense of changes in unemployment duration over the 2000–2012 period. The sharp increase in duration during and after the most recent recession is evident in all panels, with each of the duration measures approximately doubling or more relative to their prerecession values.

The unemployment durations in Figure 4 come from the cross-sectional distribution of spells in progress. As noted earlier, due to the length-biased sampling problem, this distribution is likely to overstate the duration distribution for all spells of unemployment. This is implied by the tabulations of survivor rates of unemployment spells in Table 3, for which we used our matched data to calculate survivor rates across months during the first year of unemployment. This table displays the tabulations separately for our two estimation samples (January 2000–February 2005 in Column 1 and January 2007–October 2012 in Column 3). In order to highlight the higher unemployment survivor rates in the weakest labor market periods, the table also displays survivor rates for the subperiods with relatively high availability of extended benefits and unemployment rates (March 2002–February 2004 in Column 2 for the earlier period and January 2009–December 2011 in Column 4 for the later period). The exit rates from unemployment are sufficiently high that only a small fraction of individuals remain unemployed after the first six months. In the weak labor market from 2002–2004, only about 10 percent of unemployment spells for job losers (our UI-eligible sample) lasted past six months. In the very weak labor market from 2009–2011, 13.7 percent of job losers remained unemployed for at least six months.

On inspection, it appears that the survivor rates presented in Table 2 are not consistent with the cross-section distribution of durations of incomplete spells calculated directly from the CPS. For example, in the 2009–2011 period, the survival rate at six months is 13.7 percent while 39.8 percent of spells in progress in the cross-section in this period were at least six months. This is largely a result of the length-biased sampling built into the cross-section that results in oversampling of long spells. In Section VIC, we present estimates of the steady-state cross-section distribution of spells in progress implied by our model of exits that reconciles much of this difference. Our analysis highlights the inappropriateness of making inferences about the distribution of completed spells from the cross-section distribution of spells in progress.

Figure 5 complements Table 3 by showing monthly exit rates from unemployment over our complete sample tabulated for all exits in Panel A and exits by type (to employment or out of the labor force) in Panels B and C. In each case, we display the exit rates for all unemployed individuals and also for UI-eligible and ineligible individuals separately. A sharp decline in exit rates, particularly exits to employment, is evident during the recent recession, with a minor rebound evident beginning in 2010. Overall exit rates are higher for UI-ineligible individuals than for eligible individuals, which reflects the large gap between the two groups for exits out of the labor force (Panel C). The rates of labor force exit from unemployment exhibit very little cyclicality, with only a slight net decline evident during the recent recession and essentially no cyclicality evident for UI ineligibles.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Unemployment Exit Rates, Matched CPS

Notes: Authors’ tabulations from matched CPS data (weighted), expressed as three-month moving averages. Gray areas denote NBER recession dates.

A. A Placebo Test: UI-Ineligible Spells

We defined the UI-eligible group to be those who report a job loss as their reason for unemployment. The remaining unemployed report being a job leaver (quit) or a labor force entrant (new entry or reentry) as their reason for unemployment, and we classify these individuals as UI ineligible. While this classification scheme is not perfect, we presented evidence in Table 1 based on the March CPS that only a small fraction of job leavers and new entrants report having received UI. This group should be largely unaffected by the availability of extended benefits. On this basis, we reestimate our probit model of exit from unemployment on samples of UI-ineligible unemployed individuals in order to provide a placebo test of our estimation strategy. If we find effects of extended benefits on exit from unemployment for the ineligible that are similar to those we present in Table 3, it could be that we have not adequately controlled for state / month specific economic conditions, and hence our estimates of the effect of extended benefits are too large.¹⁴ A potential factor working in the opposite direction is that there may be a positive effect of extended benefits on exit from unemployment among the UI ineligible resulting from spillovers from eligible to ineligible individuals. As we noted earier, the reasoning is that, if extended benefits reduce the job finding rate among the UI eligible, there may be improved job opportunities for the UI ineligible that increase their exit rate from unemployment (Levine 1993; Lalive, Landais, and Zweimuller 2013).

Table 5 contains the estimates of the key UI parameters ( and ) estimated for the sample of UI-ineligible spells. In the earlier period, we estimate there to be no effect of either the availability of extended benefits or the exhaustion of benefits on the exit rate in either the single risk or the competing risks model. In the later period, we do find a statistically significant negative effect of the availability of extended benefits on the unemployment exit rate in the single risk model and in the exit-to-NILF model. These estimates suggest that our estimates for the UI-eligible sample in Table 3 may be biased upward in absolute magnitude.

Interestingly, our estimates show a marginally significant (though small) increase in the exit rate to employment due to extended benefits for the UI-ineligible individuals in the 2007–2012 period. As noted above, this may reflect a spillover from those eligible for UI to those not eligible for UI.

Our conclusion from the placebo exercise is that in the earlier period we appear to have adequately controlled for labor market conditions while in the later period it may be the case that the availability of extended benefits is correlated with unmeasured (unfavorable) economic conditions. As such, our estimates of the effect of extended benefits on exit from unemployment in the later period could be considered an upper bound on the true effect.

A. The Effect of Extended Benefits on the Survivor Function of Unemployment Durations

Figure 6 contains plots of the survivor functions for the sets of hypothetical spells of unemployment corresponding to the weakest labor market in the early 2000s (spells starting in 2002) and the weakest labor market later in the decade (spells starting in 2009).¹⁷ Each panel of Figure 6 presents a comparison of the survivor functions (beginning in month 6) for two scenarios. The first row of the figure shows comparisons for the hypothetical spells beginning in 2002, whereas the second row shows comparisons for hypothetical spells beginning in 2009. The left panel in each row shows a comparison of the Observed-EB and No-EB scenarios while the right panel in each row shows a comparison of the Full-EB and No-EB scenarios.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Unemployment Survivor Functions, by Time Period and EB Scenario

Source: Authors’ calculations using hypothetical unemployment spells.

At six months, a small difference in survival probabilites emerges for each comparison. The plots in the first row of Figure 6 for spells beginning in 2002 indicate that the Observed-EB – No-EB difference in survival probabilities is fairly small while the Full-EB – No-EB difference is larger. This reflects the fact, illustrated in Figure 1, that extended benefits were relatively less generous in the 2002–2004 period, so that the No-EB scenario is much closer to the Observed-EB scenario than to the Full-EB scenario. The plots in the second row of Figure 6 for spells beginning in 2009 show a different pattern. The Observed-EB – No-EB difference in survival probabilities is similar in magnitude to the Full-EB – No-EB difference. This reflects the fact that 99 weeks of extended benefits were almost universally available in the 2009–2011 period so that the Observed-EB and Full-EB scenarios are very similar to each other.

In order to see clearly the magnitude of the effect of extended benefits on the survivor probabilities, Figure 7 contains plots for the 2002 and 2009 hypothetical spells of the difference in survival functions between the Observed-EB and No-EB scenarios (left panel) and the difference between the Full-EB and No-EB scenarios (right panel). The comparison of the Observed-EB and No-EB scenarios suggests that extended benefits had a more persistent effect on the duration of unemployment at higher durations (after about ten months) in the later period than in the earlier period. This is not surprising given the relatively limited duration and availability of extended benefits in the earlier period. Indeed, the comparison of the Full-EB and No-EB scenarios makes it clear that much of the difference across periods is due to the higher level of availability and generosity of UI in the 2009–2011 period relative to the 2002–2004 period. Interestingly, the Observed-EB – No-EB difference in survival probabilities falls sharply after about one year in the later period despite the fact that almost two years of extended benefits were available.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7

Difference in Unemployment Survivor Functions, by EB Scenario and Time Period

Source: Authors’ calculations using hypothetical unemployment spells.

Taken together, Figures 6 and 7 imply that there is a measurable effect of a generous extended benefits program on unemployment durations for the small fraction (about 15 percent) of unemployment spells that last at least six months. Focusing on the Full-EB – No-EB comparison in the righthand columns of the two figures, at 12 months-duration in 2009 there is about a two percentage point higher survival probability attributable to the availability of full extended benefits. Given a predicted survival probability at 12 months in the 2009 Observed and Full-EB cases of about 0.06, this implies that about 1 / 3 (0.02 / 0.06) of the surviving unemployment spells at one year are due to the availability of extended benefits. Although this seems substantial, this effect only applies to the very small fraction of spells (about 6 percent) that survive at least one year.

B. The Effect of Extended Benefits on Expected Duration

A useful summary measure of the overall effect of extended benefits on the distribution of duration of unemployment spells is the expected (average) duration of an unemployment spell. Table 6 contains our estimates of the expected duration of unemployment spells for the 96,575 hypothetical spells for each of the four scenarios (Observed-EB, No-EB, Full-EB, One-Month-EB). Each panel of the table contains estimates for one of the two time periods we defined earlier. The first four rows of each panel contain predicted expected durations for the four extended benefit scenarios described above. The remaining three sections of each panel show both the absolute and proportional differences in expected durations between: (1) the Observed-EB and No-EB scenarios, (2) the Full-EB and No-EB scenarios, and (3) a One-Month-EB scenario and the No-EB scenario. We provide the final comparison using the One-Month-EB scenario as a common metric for assessing the impact of a specific extension of UI availability.

Consider first the estimates for Observed-EB in the single risk model in the first column. The expected duration of unemployment spells beginning in March 2002 (Panel A) averaged 3.56 months. This expected duration is intermediate between those for the No-EB and Full-EB scenarios. Our estimates suggest that the expected duration of unemployment was 4 percent higher due to the existence of extended benefits during the 2002–2004 period. Compared with this earlier period, the average expected duration of unemployment spells beginning in January 2009 was substantially higher (4.89 months in Panel B). Our estimates suggest that the expected duration of unemployment was 7 percent higher due to the existence of extended benefits during the 2009–2011 period. As we noted earlier, the larger effect of extended benefits in the 2009–2011 period is due to the wider availability of extended benefits during this period. This is demonstrated in the penultimate sections of each panel, which show the proportional difference in average expected duration between the Full-EB and No-EB scenario (thereby holding constant the extent of availability). This comparison shows a very similar effect of Full-EB on expected durations across the two periods (7 percent in both periods).

At the bottom of each panel, an additional month of extended benefits (from the base of no extended benefits) increases expected duration of unemployment spells by about 0.06 months (0.25 weeks) in both periods. This translates to a 2 percent increase in expected duration in the earlier period and, given the relatively longer expected durations in the later period, a 1 percent increase in expected duration in that period.

The nonlinear relationship between the duration of extended benefits and the expected duration of unemployment is clear from the pattern of results. Consider the earlier period. The 7 percent increase in expected duration implied by the Full-EB scenario is 3.5 times the 2 percent increase in expected duration implied by the One-Month-EB scenario despite the fact that the Full-EB scenario incorporates a 17-fold increase in extended benefits. Although not as dramatic as in the earlier period, the nonlinearity in response is also substantial in the later period (a ratio of 17 to 7 rather than 17 to 3.5).

Finally, we note that the magnitude of our estimates is slightly below the bottom end of the range of estimated effects of UI extensions from past research using U.S. administrative data. The results for both periods in Panels A and B of Table 6 show that an additional month of extended benefits raises unemployment duration by 0.06 months. By comparison, the estimates from Katz and Meyer (1990) imply that a one-month extension raises unemployment duration by 0.16 to 0.20 months. Estimates based on more recent data from Card and Levine (2000) and Jurajda and Tannery (2003) imply that a one-month extension raises unemployment duration by 0.07 to 0.08 months. Our estimate is slightly below the bottom end of this range, consistent with the view that the impact of extended UI benefits on job search have declined over time (Katz 2010).¹⁸

C. The Cross-Section Distribution of Unemployment Duration in a Steady State

The weak labor market in the Great Recession and its aftermath is notable not only for the high level of unemployment but also for the historically high average duration of unemployment and incidence of long-term unemployment as measured based on the reported duration of in-progress spells (Figure 4).²¹ An interesting question regards the extent to which these long durations observed in the cross-section are attributable to the availability of extended benefits.

The effects of extended benefits on expected durations and on the distribution of spells more generally pertains to the distribution of completed spells. The cross-sectional distribution of durations of spells in progress results from the dynamic process of births and deaths of spells over time, and it is known to overrepresent longer spells (length-biased sampling). The result is that the survivor function for a set of spells starting at a point in time appears on its face to be inconsistent with the observed cross-sectional distribution. For example, the raw survivor rates calculated from our matched CPS data (Table 2) show that for the 2009–2011 period 13.7 percent of unemployment spells last at least six months, while the cross-sectional distribution of unemployment spells for the same period shows 39 percent of unemployment spells in progress have lasted at least six months. The same comparison for March 2002 through June 2004 shows that 9.9 percent of unemployment spells from the matched CPS last at least six months versus 22 percent in the cross-section.

We use our estimates of the model of exit from unemployment together with the steady-state assumption that there is a constant monthly inflow into unemployment to estimate the effect of extended benefits on the cross-sectional distribution of duration of spells in progress. We calculate these cross-sectional distributions for each of the two time periods (March 2002–June 2004 and January 2009–April 2011) and for the Observed-EB and No-EB scenarios. The idea is that, in a steady state, the estimate of the survivor probability (G(•)) at a given duration S is also an estimate of the probability in a cross-section that a spell that started s periods earlier is still in progress. On this basis, the fraction of spells in progress in the cross-section with duration greater than s is

(11)

where is the average predicted probability that a spell of unemployment that started j periods earlier is still in progress (the survivor probability in the steady state).²²

Figure 8 contains plots for the 2002 and 2009 samples of simulated spells of the fraction of spells in a cross-section that have durations at least as long as the indicated number of months. These are calculated using the estimates for the relevant time period contained in Table 4 and the two sets of hypothetical spells of unemployment described earlier. In order to implement the steady state assumption, exit probabilities were predicted using the estimated parameters and assuming that the time-varying measures (the month indicators and the measures of economic activity in the state) are held constant at their state-specific mean values for each time period. As such, we are assuming not only that entry into unemployment is time invariant but also that economic conditions are stable as well. The figures include a vertical line at six months indicating the point beyond which unemployment duration is considered to be “long term.”

• Download figure
• Open in new tab
• Download powerpoint

Figure 8

Cross-Sectional Distribution of Duration of Unemployment Spells in Progress in the Steady State

Source: Authors’ calculations based on estimated exit model and steady-state assumption.

The lefthand panel of Figure 8 contains the plot for the March 2002–April 2004 simulated sample of unemployment spells. In the Observed-EB scenario, the average duration of spells in progress is 4.22 months and 20.3 percent of spells in the cross-section are long term (≥ 6 months). In the No-EB scenario, the average duration of spells in progress is 3.95 months and 17.3 percent of spells in the cross-section are long term.²³ This implies that the availability of extended benefits accounts for 0.27 months (about 6.4 percent) of average unemployment duration in the cross-section and about 14.8 percent of long-term unemployment for UI eligibles in the 2002–2004 period.

The righthand panel contains the analogous plot for the January 2009–April 2011 simulated samples of unemployment spells. In the Observed-EB scenario, the average duration of spells in progress is 5.74 months and 31.4 percent of spells in the cross-section are long term (≥ 6 months).²⁴ In the No-EB scenario, the average duration of spells in progress is 4.85 months and 24.3 percent of spells in the cross-section are long term. This implies that the availability of extended benefits account for 0.89 months (about 15.5 percent) of average unemployment duration in the cross-section and about 22.6 percent of long-term unemployment for UI eligibles in the 2009–2011 period.

These calculations imply that the availability of extended benefits had a larger impact on average unemployment duration and long-term unemployment in the 2009–2011 period (increases of 15.5 percent and 22.6 percent respectively) than in the 2002–2004 period (increases of 6.4 percent and 14.8 percent respectively). This is consistent with the findings we reported earlier and as expected given the substantially higher prevalence and duration of extended benefits in the Great Recession and its aftermath.