Evaluations of State Welfare-to-Work Programs


A central challenge in welfare policy arises from the dual imperatives to promote self-sufficiency among welfare recipients and to protect vulnerable families from economic deprivation. In the late 1980s and throughout the 1990s, the U.S. moved to expand the services it offered recipients to help them prepare for and find work, while also expanding the mandates on welfare recipients to participate in such services. This effort to move people from welfare to work was accompanied in the 1996 welfare reform law by new time limits on welfare receipt. Notably, the mandates and time limits applied to populations that had previously faced far fewer conditions on welfare receipt —particularly single mothers with young children.

MDRC’s evaluations of state welfare-to-work programs have provided rigorous evidence at every stage of this evolving story, looking at the effectiveness of various programs and services, the real results of time limits and participation requirements, and the outcomes for families and children.

Agenda, Scope, and Goals

MDRC began studying the effects of state welfare programs in the early 1980s, evaluating some of the early pilots that informed welfare reform in 1988 and, later, the federal 1996 welfare reform law. Over the following two decades, MDRC conducted a series of large-scale random assignment studies that provided reliable evidence both on the effects of welfare-to-work mandates and on the relative effectiveness and costs of different employment-service strategies — a knowledge base that has shaped policy and practice at the federal, state, and local levels.

MDRC’s post-1988 evaluations of state welfare-to-work programs began with the National Evaluation of Welfare-to-Work Strategies (NEWWS), which evaluated 11 mandatory welfare-to-work programs in seven sites. Subsequent state welfare-to-work evaluations also include studies of Los Angeles Jobs-First GAIN, the Minnesota Family Investment Program, Florida’s Family Transition Program, Vermont’s Welfare Restructuring Project, Connecticut’s Jobs First, and Wisconsin Works. All of these programs involved participation requirements coupled with work-related services; the services themselves varied from program to program. A number of these programs also sought to “make work pay” by offering earnings supplements. In addition, the Project on Devolution and Urban Change studied the implementation of the 1996 welfare reform law and its effects on families, neighborhoods, and institutions in four large urban counties.

Several of these evaluations included close examination of the effects of time limits on welfare receipt. The Florida Family Transition Program and Connecticut Jobs First included time limits that could result in the termination a family’s entire welfare grant, and MDRC’s evaluations of these programs are the only large-scale random assignment studies of the effects of such limits. In addition, MDRC has conducted studies examining time limits more broadly, including the Cross-State Study of Time-Limited Welfare and a federally funded synthesis of research on time limits.

Overall, many of these programs were successful in moving people from welfare to work, but generally did not improve family incomes or reduce material hardship, unless they provided earnings supplements. While the different programs varied and the results are far from uniform, several trends emerged. The work-first, employment-focused models that required initial job search  were successful at moving people from welfare to work, as were models that required initial participation in basic education and, to a lesser extent, vocational training. But, when the two models were tested side-by-side, the model with the requirement to participate in training and education, by and large, did not raise employment or employment retention rates or move people to higher-paying jobs more than the employment-focused models; on the contrary, the employment-focused models often moved people into jobs sooner and so increased overall earnings. At the same time, the programs with the most success at increasing earnings used a mixed approach, incorporating initial assignments to job search or to education or training, with case managers often creating different plans for different people.

Importantly, although the work requirements did often result in reductions in the use of welfare and greater employment, with some savings for government, overall income and material hardship remained more or less unchanged. With the exception of a few programs that also included components to “make work pay,” as people moved off of welfare into work, their new earnings generally brought dollar-for-dollar cuts in their welfare benefits. In addition, the long-term differences between those in such programs and those in the control groups, most of whom were subject to the previously existing welfare laws, were often notable at first but dissipated with time as the control group members often also found employment.

Several of the studies also looked closely at effects on child well-being. Overall, the programs did not produce major changes in child well-being or school performance. Some programs saw minor shifts, but these were not universally positive or negative.

In looking at the implementation of federal time limits on welfare receipt, MDRC found wide variance among the states in what time limits were set, how they were enforced, what exceptions were granted, and what the results were for families. By the mid-2000s, nationally, a relatively small proportion of families had had their benefits revoked because of a time limit. This was in part because of differing state laws, as well as exceptions and extensions that were granted to many families. In addition, many families left welfare on their own, at least for some period of time, before running up against a time limit. Families that did have their benefits revoked, however, often became more reliant on family and friends as well as on other government benefits, and saw a marked increase in hardship.

It is worth noting that most of these studies took place in a climate of general economic prosperity. Work requirements and time limits would likely pose different challenges at a time when jobs are scarce.


Design, Sites, and Data Sources

MDRC’s evaluations of state welfare-to-work policies include large-scale, random assignment tests, as well as implementation studies, covering eleven states and involving tens of thousands of participants. The state policies that MDRC evaluated either anticipated or responded to new federal requirements, particularly the 1996 welfare reform law; all contained a core quid pro quo arrangement in which the government would offer education, training, job search assistance, and support services to people receiving cash welfare, while most recipients — the majority of them single parents — would be required to participate in such services in order to qualify for benefits. MDRC’s evaluations offer key evidence about the impacts of different services and requirements on families and provide important lessons about what works to increase employment and improve family well-being among welfare recipients.

The National Evaluation of Welfare-to-Work Strategies (NEWWS): With an overarching goal to find out what welfare-to-work strategies work best for what groups of recipients, NEWWS examined 11 programs in seven locations, tracking more than 40,000 single-parent families over a five-year period. The 11 programs were each characterized by one of three approaches: an employment-focused approach, which emphasized short-term job search assistance and encouraged recipients to find jobs quickly; an education-focused approach, which emphasized the importance of recipients’ participating in longer-term skill-building activities (primarily basic education) before entering the labor market; or a mixed approach, which determined on the basis of recipient characteristics and caseworker judgments whether recipients would benefit more from job search first or education and training first.

Los Angeles Jobs-First GAIN: Between 1993 and 1995, responding to earlier MDRC evaluation results for California’s welfare-to-work program, Greater Avenues for Independence (GAIN), the Los Angeles County Department of Public Social Services shifted the focus of its GAIN program from providing basic education to moving people into employment as quickly as possible. With the largest welfare caseload of any county in the nation, Los Angeles targeted its new mandatory program, Jobs-First GAIN, at a broad range of welfare recipients. The Jobs-First GAIN program had four important attributes: It imparted a strong work-first message through an unusually time-intensive program orientation and alerts about impending welfare reforms; it emphasized the financial benefits of combining work and welfare in the short term under California’s rules for disregarding earnings in its welfare benefit calculations; it assigned most people to a job club as a first activity, providing high-quality job search assistance that included the services of job developers, who cultivated relationships with local employers and tried to match recipients to available jobs; and it strictly enforced work participation requirements and frequently imposed financial sanctions for nonparticipation.

Minnesota Family Investment Program: Minnesota Family Investment Program (MFIP), piloted from 1994 through 1998, combined employment mandates with financial work incentives. MFIP was characterized by four key features: a requirement that long-term recipients work or participate in employment-focused services; financial work incentives for recipients who worked; payment of working recipients’ child care costs directly to providers (rather than reimbursement of recipients later); and simpler public assistance rules and procedures that combined different programs into one and provided food stamps as part of the cash welfare grant. The hope was that combining welfare-to-work strategies with financial incentives designed to “make work pay” would make both aspects more effective and reduce hardship (for more information on MFIP and similar programs, see MDRC’s studies of “make work pay” experiments).

Florida’s Family Transition Program: Florida’s Family Transition Program (FTP) was one of the first welfare initiatives to impose a time limit on receipt of cash assistance. For this reason — and because Florida was the first place in the country where people reached a time limit and actually had their benefits cancelled — FTP was an important test case for states and localities across the country. Operated from 1994 through 1999 as a pilot program in Escambia County (Pensacola), FTP imposed a time limit of 24 out of 60 months on most families’ receipt of cash welfare and a time limit of 36 out of 72 months for those who were least job-ready. To help recipients — most of them single parents — prepare for, find, and keep jobs, it also offered an array of services and incentives, including: mandatory employment-related, social, and health services; intensive case management; a modest financial work incentive that increased the portion of recipients’ earnings that could be disregarded in calculating cash benefits; extended transitional child care assistance for recipients who left welfare for work; and a parental responsibility mandate that, among other things, required recipients to make sure their children attended school.

Vermont’s Welfare Restructuring Project: Vermont’s Welfare Restructuring Project (WRP) was one of the first comprehensive statewide reform initiatives piloted under waivers of pre-1996 federal welfare rules. Operated from 1994 through 2001, WRP had three key features: It required that recipients get a wage-paying job after a prescribed maximum number of months of benefit receipt (30 months for single-parent families and 15 months for two-parent families); it provided job search assistance and, for recipients who could not find regular jobs, subsidized, minimum-wage community service positions; and it encouraged work and self-sufficiency by changing some welfare rules (for instance, those governing how long a portion of a recipient’s earnings could be disregarded in benefit computations) and providing more generous transitional benefits to recipients who left welfare.

Connecticut’s Jobs First: Connecticut's Jobs First program attracted national attention because it included all the key elements of the 1990s welfare reforms: time limits, financial work incentives, and work requirements. Specifically, Jobs First limited families to 21 cumulative months of cash assistance unless they received an exemption or extension. It included an unusually generous financial work incentive that allows employed recipients to retain their full welfare grant as long as they earned less than the federal poverty level. And it required recipients to work or to participate in employment services designed to help them find jobs quickly.

Wisconsin Works: The Wisconsin Works (W-2) program was not only one of the nation’s earliest comprehensive welfare reform initiatives but, in the wake of the 1996 welfare reform legislation, also revealed the challenges of implementing a complex new set of administrative procedures and legal protocols. MDRC’s study set out to document how W-2 was implemented in Milwaukee County, the state’s largest metropolitan area. In the study, MDRC researchers examined such issues as how W-2 administrators and program contractors managed the client intake and assessment process, how determinations were made regarding whether clients would be eligible to continue to receive program benefits when they reached time-limit milestones, and how program procedures evolved under the new system to resolve clients' challenges of agencies' decisions. The W-2 program involved assigning welfare recipients to different “tiers” of programming; the most commonly assigned tier was the community service job (CSJ) tier. Intended for individuals who were deemed not ready for immediate regular employment, the CSJ tier was designed to provide an opportunity to practice work habits and skills. Most program participants entering the tier were assigned to work experience in community service jobs, but they could also be assigned to other activities such as further assessment, employment search, education, and training. MDRC conducted an in depth study of how assignments were made and how services were monitored and managed within this tier.

Project on Devolution and Urban Change: The 1996 federal welfare reform law devolved to states unprecedented control over welfare policy while requiring a dramatic overhaul of their welfare rules and services. Launched in 1997, the Project on Devolution and Urban Change chronicled the changes that reform wrought in the lives of low-income families and in the institutions that served them by examining welfare reform as it played out in big-city neighborhoods, where the nation’s caseload is increasingly concentrated. The research project examined a wide range of family, program, and neighborhood outcomes in four large urban counties — Cuyahoga (Cleveland), Los Angeles, Miami-Dade, and Philadelphia.

Cross-State Study of Time Limited Welfare: Time limits first emerged at the state level and subsequently became a central feature of federal welfare policy in the 1996 welfare reform law, which imposed a 60-month time limit on federally funded assistance for most families. The Cross-State Study of Time Limited Welfare looked at the results of seven of the earliest state welfare reform