Book Description

Poverty rates in the United States fell from a 25-year high of 15.1 percent in 1993 to near record lows of 11.3 percent in 2000 and have since increased steadily to 12.7 percent in 2004 (U.S. Census Bureau 2004a). The poverty rates for children and for people in single female-headed families followed a similar pattern, although at considerably higher rates -- 17.8 percent and 30.5 percent in 2004, respectively (U.S. Census Bureau 2004a, 2004b). Many political leaders pointed to poverty rate declines along with increases in employment and falling welfare caseloads that occurred in the late 1990s as evidence that the 1996 federal welfare reform had been a success (Kaus 2001). During the late 1990s, however, there was concern that welfare reform was leading to increases in deep poverty (living below 50 percent of the poverty threshold) (e.g., Sherman et al. 1998, as cited in Haskins 2001), as deep poverty rates increased in 1996 and were unchanged in 1997 (U.S. Census Bureau 2004c). Deep poverty rates subsequently fell, however. While trends in poverty and deep poverty generated discussion and speculation about the effect of welfare reform on poverty and deep poverty, there is limited and mixed information on welfare reform's effect on these outcomes. Moreover, the literature provides no guidance on how specific welfare reform policies affect poverty and deep poverty. Some of the welfare reform policies implemented by states are hypothesized to increase poverty (e.g., family cap), while others are hypothesized to decrease poverty (e.g., increases in the earned income disregard). Thus, on net, one could find an overall reform effect of zero, when in fact specific policies have affected families' economic well-being, but in off-setting ways. This paper contributes to the literature by examining the effects of a rich and comprehensive set of specific welfare policies on poverty and deep poverty among women and children. We capture objective and detailed measures of states' policies by measuring policies individually, and in continuous values such as dollars, wherever possible. Nineteen specific policies are included in our analysis. These polices are grounded in a conceptual framework of how policies can influence poverty and are measured in great detail on a monthly basis from 1986 through 2000. Our approach leads to results that are robust to alternate specifications. Variation in welfare policies over time and across states enables us to measure the relationship between policy and poverty. States implemented changes to their welfare programs via welfare waivers in the early to mid-1990s and then used the flexibility provided by federal welfare reform's 1996 Temporary Assistance for Needy Families (TANF) program to further change policies. We also examine how the effects of these policies change over time, as there can be short-run mechanical effects that are simply due to changes in grant size and eligibility rules, for example, as well as medium-run behavioral responses as families alter their work effort in response to changes in program rules. In addition, we contribute to the literature by examining the impact of welfare reform on the economic well-being of children as well as adults. We use longitudinal data from the 1988, 1990, 1993, 1996, and 2001 panels of the Survey of Income and Program Participation (SIPP). We use these panels to provide data from January 1988 through December 2002, allowing us to capture the period prior to the implementation of state waivers, during the implementation of state waivers, and after the 1996 federal welfare reform. Further, these data capture periods of strong and weak economic conditions. We use the Urban Institute's Welfare Rules and related databases to measure our welfare policies. Overall, we find evidence that more lenient eligibility requirements for welfare receipt and more generous financial incentives to work generally reduce deep poverty, as hypothesized. We also find evidence that eligibility requirements for welfare receipt and financial incentives to work affect poverty. Time limits are hypothesized to have ambiguous effects on poverty and our results suggest that some stricter time limit policies may lead to lower rates of deep poverty and poverty. Our findings are generally consistent with our hypotheses and are also consistent across our population of ever-single mothers and children of ever-single mothers.