An Urban Institute New Federalism Program to Assess Issues and Options for States Changing Social Policies THE URBAN INSTITUTE Series A, No. A-46, December 2000 Unemployment Insurance and Welfare Recipients: What Happens When the Recession Comes? Harry J. Holzer In many ways, our national experiment an era when jobs may not be so plentiful as with welfare reform has been more suc- they currently are, the “safety net” avail- cessful to date than many analysts had able to those who cannot find jobs may A recession will anticipated. Not only have welfare rolls have some significant gaps in it. declined by roughly half since the early Traditionally, the major “safety net” significantly limit 1990s, but also employment rates have program available to unemployed workers the employment risen for most former (and many current) during a recession has been the welfare recipients (Council of Economic Unemployment Insurance (UI) system. options of recent Advisers 1999). Nevertheless, several However, several authors (Kaye 1997; welfare recipients important questions about the success of Gustafson and Levine 1998; Vroman 1998) and will require welfare reform remain—and one of the have noted that, in the next recession, eligi- most pressing is what will happen when bility for UI among former welfare recipi- appropriate safety the current national economic boom ends ents will be limited for a variety of reasons, nets for unemployed and the next recession begins. particularly insufficient prior work experi- The importance of the strong national ence. At the same time, many of these indi- recipients. economy to the success of welfare reform viduals (and their families) will be ineligi- to date has been considerable. For one ble for Temporary Assistance for Needy thing, we have recently enjoyed the lowest Families (TANF) benefits if they have unemployment rates nationally in 30 years. exhausted their lifetime limits. Neither Virtually every recent analysis suggests program may be available to the (often that the strong economy of the 1990s has noncustodial) fathers in these families as contributed significantly to both the declin- well, whose contributions to their family’s ing caseload and the rise in employment financial well-being are increasingly cru- rates and earnings among single mothers cial (Sorensen 1999). (e.g., Council of Economic Advisers 1999; On the other hand, little is known cur- Meyer and Rosenbaum 2000). The continu- rently about how significant these prob- ation of extremely tight labor markets since lems are likely to be. Estimates in the federal reforms were implemented has cre- sources cited above are based almost ated an environment in which transitions exclusively on data from the 1980s and from welfare to work could proceed more early 1990s, during which time employ- easily than they otherwise would. A seri- ment among welfare recipients was much ous recession would eliminate these condi- lower than it is today. More recent data on tions and likely cause some reversal of the employment experiences of current these trends. Furthermore, we have moved and former welfare recipients are now from a social welfare system that was cen- available and might lead to new estimates tered around cash assistance to the nonem- of future UI eligibility. ployed to one that is based on assistance to This brief reviews evidence on these the “working poor” (e.g., Ellwood 1999). In issues and considers their implications for ASSESSING THE NEW FEDERALISM An Urban Institute Program to Assess Changing Social Policies policy. In particular, the following ques- considering current hiring patterns and tions are addressed: their relation to measures of labor market tightness. For instance, data from employer 1. By how much is employment likely to surveys recently administered in several decline among welfare recipients and large metropolitan areas show that the job other vulnerable groups of workers vacancy rate would likely decline by two- during a recession? thirds or more during a severe recession 2. How many former welfare recipients and by somewhat less in a milder recession and other vulnerable workers will be (Holzer and Stoll 2000a, b). Accordingly, eligible for UI? the new hire and employment rates of wel- 3. If UI is not likely to serve a large frac- fare recipients could decline by large tion of this population during a down- amounts as well.1 turn, what should state and federal pol- Of course, the impact of the next reces- icymakers do to address these issues? sion on the job status of welfare recipients remains uncertain, since many of its attrib- Only about 30 Employment Declines in utes—including its severity, duration, and distribution across states and/or economic percent of all the Next Recession sectors—are unknown. Despite this uncer- currently One way to estimate the extent to which tainty, however, nearly all of the above unemployed employment will decline among welfare estimates suggest that a recession will sig- recipients and other vulnerable groups nificantly limit the employment options of workers receive UI, during the next recession is to gauge recent welfare recipients and will require and only about 40 movements in employment and unem- appropriate safety nets for unemployed ployment over previous business cycles recipients. percent did so (figures 1 and 2). In general, the employ- during the most ment rates of women have been less sensi- Determining Eligibility for tive to the business cycle than those of recent economic men, even among the less educated. Unemployment Insurance downturn. Nonetheless, the data show that adult There are a number of reasons why wel- female high school dropouts, black fare recipients who lose their jobs might women, and teenage black women in par- not qualify for UI. Vroman (1998) identifies ticular experience large employment four: (1) insufficient prior work experience declines during recessions. Indeed, and earnings, (2) use of base periods for employment rates for black female teens calculating prior earnings that, in turn, dis- during a recession can decline by as much qualify up to six months of an employee’s as one-third. most recent work (since the current and So which group is most comparable to previous quarter’s earnings are generally former welfare recipients who are now omitted), (3) reasons for job departure or working? In terms of education and basic loss (since spells of unemployment result- cognitive (i.e., reading, writing, and arith- ing from employee quits or discharges for metic) skills, adult female high school just cause are generally not covered by UI), dropouts may be the most relevant com- and (4) lack of availability for full-time parison group and, therefore, the employ- work due to family responsibilities or ment losses of welfare recipients in a other personal problems. Applying these downturn may be relatively modest. On reasons to work and turnover patterns the other hand, welfare recipients’ vulnera- observed among welfare recipients in the bility to a downturn might be more like National Longitudinal Survey of Youth in that of teens—both have substantially less the 1980s and 1990s has led several ana- labor market experience than most work- lysts to conclude that no more than 20 per- ing adults, and experience is a very strong cent of unemployed welfare recipients predictor of job loss during a recession. would be eligible for UI in a recession. Another way to determine the vulnera- There are a number of reasons, howev- bility of welfare recipients’ employment er, to question whether these inferences during an economic downturn involves from past data are accurate predictors of 2 An Urban Institute Program to Assess Changing Social Policies ASSESSING THE NEW FEDERALISM FIGURE 1. Unemployment Changes over Previous Business Cycles (1979, 1982–1983, 1989, 1991–1992) All Men (Age 16 and All Women (Age 16 Black Women (Age 20 Adult Female High Teenage Black Women Older) and Older) and Older) School Dropouts (Ages 16–19) Source: Bureau of Labor Statistics. Note:Data on high school dropouts were not available for the earlier periods. FIGURE 2. Employment Changes over Previous Business Cycles (1979, 1982–1983, 1989, 1991–1992) All Men (Age 16 and All Women (Age 16 Black Women (Age 20 Adult Female High Teenage Black Women Older) and Older) and Older) School Dropouts (Ages 16–19) Source: Bureau of Labor Statistics. Note: Data on high school dropouts were not available for the earlier periods. 3 ASSESSING THE NEW FEDERALISM An Urban Institute Program to Assess Changing Social Policies UI eligibility during future downturns. average is roughly $2,000 of earnings over First, there are many more current and the previous four quarters, with most recent welfare recipients working now states falling in the range of $1,000 to than in the 1980s. Next, the longer the cur- $3,000.2 At 30 hours per week and roughly rent boom lasts, the more work experience $6.00 per hour (both of which are achieved recipients will have gained once the reces- by large majorities of these workers), recip- sion begins—and the less important the ients would need to work 5 to 17 weeks, or base period designation might be in deter- an average of 11 weeks, to meet base peri- mining their eligibility. Furthermore, many od earnings requirements. Even allowing more job losers in a recession will have for a base period that may exclude the been laid off from their jobs than is the most recent three to six months of earn- case currently, and thus fewer job ings, most of these workers would qualify leavers/losers will be disqualified from UI for UI if laid off so long as they would be eligibility due to reason for job departure. willing to work as many hours on their Recent estimates Finally, the growth of labor market activity subsequent jobs as they had worked on among single mothers in the 1980s and their most recent ones.