The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) (P.L. 104-193) marked its 15th anniversary on August 22, 2011. Although amendments to the Social Security Act of 1935 (P.L. 74-271) had altered Title IV-A, Aid to Dependent Children (ADC), with some frequency, the PRWORA was arguably the most substantial reconfiguration of American poor relief since the New Deal. Opponents of the law offered dire warnings as it moved toward enactment: One study commissioned by the U.S. Department of Health and Human Services (HHS) estimated that more than 1 million children would be made newly poor (Edelman, 1997), leading New York Senator Daniel Patrick Moynihan (1996) to call it "the most regressive and brutal act of social policy since Reconstruction" (p. 8076). Proponents, by contrast, predicted that the PRWORA would reduce out-of-wedlock births, increase marriage rates, improve poor women's labor force participation rates, and, as a result of these factors, improve the material and moral well-being of "dependent" families.
Curiously, it remains unclear which side was right. Comprehensive policy evaluation is always difficult: Policy goals can be conflicting or ambiguous, determining causality remains an imperfect science, policies may have diffuse impacts that are impossible to fully trace, and accurate, consistent, and comprehensive data may be unavailable or hard to acquire. But the evaluation challenge posed by the PRWORA seems to be greater than that which normally prevails.
First, although it has been common to speak of "welfare reform" in the singular, the PRWORA established a national framework within which states were expected to innovate to suit their own preferences, and their implementing legislation has varied along multiple dimensions, including thresholds for determining eligibility; benefit amounts; categories of eligible individuals; time limit policies; work requirements (including the definition of work itself); the availability and generosity of child care subsidies; sanction policies; the use and content of diversion programs or payments to discourage applications; whether families are required to meet school attendance, immunization, drug testing, or health screening requirements; which policies and programs are carried out by public agencies and which are contracted or subcontracted to for-profit or not-for-profit organizations (and how adequately they are funded and monitored and the manner in which their incentives are structured); and the extent to which street-level implementation reflects the law as written (Urban Institute, 1997-2008).
Further complicating evaluation, these policies, their funding, and the manner in which they were implemented vary within states from county to county and even city to city. As a consequence, there are many more than merely 50 configurations or "bundles" of policy reforms, and each was carried into effect in locations that differed in their unemployment rates, prevailing wages, and opportunities available to the lower-skilled and less-educated women who are the typical beneficiaries of welfare. Although there was variation across states under ADC and its successor, Aid to Families with Dependent Children (AFDC), regarding eligibility rules and benefit levels, states were otherwise constrained by national standards. One way we can productively alter our thinking about PRWORA is to take more seriously the fact that there has been no such thing as welfare reform: There has, instead, been a multiverse of welfare reforms, with variation far surpassing that which is regularly produced by American-style federalism.
Isolating and gauging the effects of PRWORA have been additionally bedeviled by economic circumstances. Most reforms went into effect during the boom of the late 1990s, a period of rapid growth and .job creation, rising wages, and low unemployment during which the Earned Income Tax Credit (EITC) underwent a large expansion and supplemented the income of the very population most affected by PRWORA, making it difficult to isolate the effects of PRWORA during these years (Blank, 2007). …