Economic Conditions and Welfare Reform.

Edited by Sheldon H. Danziger.

Kalamazoo, MI: W. E. Upjohn Institute for Employment Research, 1999. Pp. vii, 321. $22.00 (paperback).

Perhaps nothing makes a good empirical researcher more uncomfortable than being asked to speculate on the effects of policy changes that represent a big departure from previous policy. Economic Conditions and Welfare Reform is an intriguing collection of essays that presents this exact challenge to several of the top welfare policy researchers across the country. The nine papers in this volume (presented as part of a conference organized by the Joint Center for Poverty Research) ask the authors to "use their analyses to predict what is likely to happen to welfare caseloads, to recipient well-being, and to state budgets and policies when the next recession arrives" (p. 2). Given that the "next recession" is now upon us, it is hard to imagine a more timely collection of essays.

The papers in this volume are among the first to explore the implications of the new era for the welfare system in the United States, ushered in with the passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) in August 1996. PRWORA ended the entitlement to cash support for low-income single-parent families by abolishing Aid to Families with Dependent Children (AFDC) and replacing it with block grants to the states, that is, the Temporary Assistance for Needy Families (TANF) program. TANF allows states more discretion in setting rules for their welfare programs, and has intensified the emphasis on work for most single parents receiving public assistance by placing time limits on cash support (five years for most families) and by mandating some form of work as a requirement for cash support.

Yet policy changes aimed at increasing work among welfare recipients started well before 1996. For example, beginning in the early 1990s, many states were granted waivers to their AFDC programs, which states used to encourage work among AFDC recipients, often instituting time limits or work requirements similar to those implemented on a broader scale as part of PRWORA in 1996. It is important to note that most of the papers in this book use these early experiences with state waivers to the AFDC program to predict how later welfare reform efforts will affect caseloads and state budgets. [1]

The first four papers of this volume tackle an important but difficult question: How much of the recent decline in AFDC/TANF caseloads (they fell by 35% between 1994 and 1998) is due to welfare reform and how much is due to the improving economy? [2] Disentangling the effects of welfare reform and the economy are crucial for predicting how caseloads will respond during the next recession. Taken together, the first four chapters provide a remarkable lesson on the sensitivity of empirical findings to alternative sets of reasonable modeling choices.

Using state AFDC caseload data from 1976-1996, David Figlio and James Ziliak (Chapter 1) find that the improving economy was responsible for 75% of the decrease in welfare caseloads between 1993 and 1996. On the other hand, state AFDC waivers, which instituted work requirements and time limits before "welfare reform," were responsible for at most 1%. They argue that the much smaller role attributable to AFDC waivers (compared with a 1997 Council of Economic Advisers report) is due to using a dynamic rather than static model to account for "the sluggish adjustment of current caseloads to past caseloads, from lags in the response of caseloads to changes in unemployment rates, and from nonstationarities in caseloads" (p. 19). Geoffrey Wallace and Rebecca Blank (Chapter 2) examine both AFDC and Food Stamp caseload data from 1980-1996 and come to a very different conclusion, arguing that state AFDC waivers appear to have played a substantial role in the recent declines in both AFDC and Food Stamp caseloads. They also find that a wide range of demographic, political, and policy-related variables do not explain much of recent declines in welfare caseloads. It appears that the different conclusions in these two papers hinge on whether state-specific time trends in welfare caseloads are modeled as linear or quadratic, a difficult modeling choice given that the data span only a couple of business cycles.

Robert Moffitt (Chapter 3) used microdata (from the Current Population Survey) to examine the relative contributions of the economy and welfare reform on AFDC participation, labor supply, and wages for various demographic groups. Moffitt finds that less-educated single mothers in states with AFDC waivers experienced falls in AFDC participation and gains in employment and hours worked, whereas only better-educated single mothers experienced wage gains. Moffitt also finds that the effects of the economy on AFDC participation and labor supply are quite sensitive to the time period examined. Timothy Bartik and Randall Eberts (Chapter 4) argue that the effects of the economy should be modeled with a richer set of variables than the often-used state unemployment rate. They find that "areas with higher concentrations of industries that hire welfare recipients and demand workers with higher education levels have higher caseloads" (p. 153). They also argue that increasing educational requirements in many industries c an explain some of the increase in welfare caseloads in the late 1980s and early 1990s.

Maria Cancian et al. (Chapter 5) indicate that those who leave welfare typically earn between $6.50 and $7.50 per hour, often in part-time or part-year jobs. This results in the majority of those leaving welfare having incomes below the poverty line. Furthermore, Harry Holzer (Chapter 6), reporting on a survey of employers of welfare recipients, suggests that the willingness of employers to hire former welfare recipients may evaporate during a recession, which implies that former welfare recipients may be particularly vulnerable to economic fluctuations.

The last three chapters examine questions related to how states will respond to fiscal downturns. LaDonna Pavetti (Chapter 7) is highly valued in the welfare reform literature for her extensive knowledge about how welfare reform has been implemented at the state and local levels. She speculates that work requirements are likely to remain a strong focus, even during a fiscal downturn, but state and local areas may become more liberal in defining what constitutes work activity. Phillip Levine extrapolates from experiences with the Unemployment Insurance system to argue that the block grant nature of TANF (with its relatively small and hard-to-access rainy day fund) is likely to make it difficult for many states to weather the next recession. Finally, Howard Chernick and Therese J. McGuire (Chapter 9) contend that moving from matching grants (AFDC) to block grants (TANF) is likely to reduce state-financed cash assistance to low-income families, but it is unlikely to result in the "race to the bottom" that many welfare analysts have predicted.

I highly recommend Economic Conditions and Welfare Reform for anyone interested in welfare reform. More importantly, this book provides a refreshing glimpse into how good empirical researchers use data and creativity to answer difficult but important questions.

(1.) Note that because of data limitations, all of these papers examine the effect of AFDC waiver approval rather than implementation. Implementation often lagged approval by a year or more, which might explain why some of these papers find small effects of AFDC waivers.

(2.) Surprisingly, very little attention is paid to other policy changes in the 1990s that might have affected welfare caseloads, such as the large expansions of the Earned Income Tax Credit (Meyer and Rosenbaum, forthcoming; Yelowitz 1995).

References

Meyer, Bruce D., and Dan T. Rosenbaum. In press. Welfare, the earned income tax credit, and the labor supply of single mothers. Quarterly Journal of Economics.

Yelowitz, Aaron S. 1995. The Medicaid notch, labor supply and welfare participation. Quarterly Journal of Economics 110:909-40.