Academic journal article Contemporary Economic Policy

Interactions between Welfare Caseloads and Local Labor Markets

Academic journal article Contemporary Economic Policy

Interactions between Welfare Caseloads and Local Labor Markets

Article excerpt

I. INTRODUCTION

Welfare reform and the introduction of work requirements have generated considerable interest in the consequences of the post-Aid to Families with Dependent Children (AFDC) policy regimes that spread across the states in the 1990s. There have been several lines of inquiry. (1) One has sought to identify the influence of state-specific reforms like waivers and time limits on caseloads and participation rates. A second and closely related line of inquiry has focused on the relative roles of welfare reform and state economic conditions in explaining the sharp decline in caseloads in the late 1990s. The state unemployment rate has served as the primary measure of economic conditions in this empirical research.

A third and smaller strand of research has explored the influence of welfare policy on economic well-being. Most of this emerging research examines the consequences of policy for the welfare population itself, while a small subset of the literature examines spillover effects for the broader population. The spillover hypothesis suggests that modern welfare programs and their attendant work requirements push the welfare population into the workforce, in turn depressing earnings, reducing hours of work, and raising unemployment rates.

Together, the research has led to some tentative conclusions regarding the consequences of welfare reform. Generally, policy parameters (including waivers and Temporary Assistance for Needy Families [TANF]) and economic conditions (in particular the unemployment rate) have had a significant influence on caseloads. There is also some evidence that welfare reform has made program participants better off, but several questions and puzzles remain.

This paper provides several contributions to the literature. First, we disentangle the potential two-way dynamic between caseload activity and labor market outcomes as described above using appropriate econometric techniques. Two approaches could conceivably be used. The first is an instrumental variables approach.

Because we are unable to find an appropriate instrument for the unemployment rate, we turn to a second approach and apply Granger-causality tests to allow inferences to be made on the direction of causality. As described below, we use the multiple-rank F test developed by Holmes and Hutton (1990, 1992) to test whether unemployment rates Granger-cause caseload activity and whether caseload activity Granger-causes unemployment rates.

We also make a number of more modest contributions. Unlike most research in the area, we focus on local labor markets, including both metropolitan and nonmetropolitan areas, for a single state. The use of local data allows for a closer correspondence between welfare cases and the market conditions individuals confront than is the case with state-level labor market data. The single-state focus allows us to ignore state welfare program parameters and isolate the interactions between caseload activity and local market conditions. Monthly data on caseload activity and local unemployment rates are used in the empirical analysis. In contrast to the existing literature that relies on data for the strong period of growth prior to the 2001 recession, we use data that begin in 1997, a period of strong economic growth, and close in September 2002, a period of stagnant job growth and rising unemployment rates.

We use unique data from Tennessee, a state that has imposed time limits on the receipt of welfare benefits. Data on caseload stocks and flows are examined to determine their influence on the local unemployment rate and the effect of the unemployment rate on caseload activity. We cannot directly address the spillover hypothesis in terms of how time limits affect individual labor market outcomes. However, we can explore the way in which caseload activity influences aggregate labor market outcomes, thus offering further evidence on labor market-caseload dynamics. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.