In the next few years, the major sources of new information will be the continuing evaluations of the 1988 welfare reform law, the Family Support Act (FSA), and of the waiver-based state welfare experiments that followed. Each is described next.

Evaluations of the Family Support Act

The last major attempt to reform welfare was the Family Support Act (FSA) of 1988, which created the Job Opportunities and Basic Skills Training (JOBS) Program within the Aid to Families with Dependent Children (AFDC) Program. The JOBS Program provided AFDC recipients with employment, training, and education-related activities, as well as supportive services.

The FSA also required states to impose work or "participation" mandates on adult recipients. States were to enroll increasing percentages of their "mandatory" AFDC caseload¹ in JOBS activities and to target expenditures on individuals most likely to become long-term welfare recipients. These participation rates were gradually increased from 7 percent in fiscal year 1990 to 20 percent in 1995. For the first time, mothers with children under the age of six were expected to participate and states could extend this requirement to those with children as young as one year of age.

The FSA imposed considerably higher participation requirements on two-parent families, who were eligible under the Unemployed Parent (AFDC-UP) Program. Between fiscal years 1994 and 1997, participation was to rise from 40 percent to 75 percent. Moreover, the required activities were generally limited to employment and work activities, with basic education encouraged only for younger recipients who had not finished high school.

The FSA also required states to ensure that teen mothers on welfare finished high school or other education or training programs. (This latter mandate, however, seems initially to have been ignored by most states.)

To ease the transition from welfare to work, the FSA provided up to 12 months of extended child care and Medicaid benefits. The act also required states to adopt even stronger child support enforcement measures, including immediate income withholding, mandatory guidelines for establishing support awards, and periodic review and adjustment of support orders. Finally, it required all states to operate an AFDC-UP program. (At that time, only about half the states provided such aid.)

The major evaluation of the JOBS Program is being conducted by the Manpower Demonstration Research Corporation (MDRC). It includes process, impact, and cost-benefit analyses, as well as special studies related to child well-being and the effectiveness of adult education. A 1995 report describes the program's preliminary, two-year impacts on employment, earnings, and welfare receipt in three sites (Atlanta, Georgia; Grand Rapids, Michigan; and Riverside, California).² Its two most notable findings were:

• "Labor force attachment" strategies: Mandatory job search activities followed by work experience or short-term education or training for those who did not find employment reduced welfare receipt 11 percentage points (68 percent for the control group vs. 57 percent for the treatment group). Monthly welfare payments declined by 22 percent ($276 vs. $216). Employment increased 8 percentage points (34 percent vs. 42 percent) and average monthly earnings increased 26 percent ($226 vs. $285).
• "Human capital development" strategies: Longer-term education and training activities also produced positive results, although not as great as those for labor force attachment approaches. Welfare receipt decreased by 4 percentage points (69 percent vs. 65 percent) and monthly welfare payments declined 14 percent ($285 vs. $247). However, there were no significant impacts on overall employment or earnings.³

In addition, the Rockefeller Institute of Government conducted a ten-state, three-year study of the implementation of the JOBS program by state and local governments. The study described how different states used the flexibility under JOBS to implement their programs. The findings covered a wide range of topics, including the design of state JOBS programs, child care and supportive services, case management, and federal participation and targeting requirements. Although it had no impact data, the study concluded that JOBS was a promising approach, but would benefit from additional funding and strong leadership from federal and state lawmakers. In fact, it argued that a major overhaul of the welfare system was not desirable, given the promise of JOBS.⁴

Several demonstration projects are testing JOBS-like services for special populations, such as teen parents, families no longer receiving AFDC, and noncustodial parents of AFDC recipients.

The Teenage Parent Demonstration is a randomized experiment being evaluated by Mathematica Policy Research (MPR), Inc. It required teen parents in three cities (Chicago, Illinois; and Newark and Camden, New Jersey) to participate in an education or training activity and then seek employment. Case management and a rich array of services, such as child care, transportation assistance, and counseling, were also provided. During the two years following intake, participation in school, job training, or employment was 19 percent higher for those in the treatment group than the control group (66 percent vs. 79 percent). However, impacts on employment and earnings were more modest. Employment for the treatment group was 12 percent higher (43 percent vs. 48 percent) and monthly earnings were up 20 percent ($114 per month vs. $137 per month). Monthly AFDC benefits were 7 percent lower ($242 per month vs. $261 per month). However, over 60 percent of the mothers experienced another birth and the birth rate was actually somewhat higher for the treatment group.⁵ A study soon to be released will provide the results for a longer follow-up period.

The Post-Employment Services Demonstration is another randomized experiment being evaluated by MPR. Although many welfare recipients leave welfare for work, many return. This evaluation tests the impact of providing services, such as assistance with family and social problems, on the likelihood that such families will keep their jobs and stay off welfare.⁶

The Parents' Fair Share (PFS) Demonstration is being evaluated by MDRC. It is intended to help reduce the welfare dependency of children by requiring unemployed noncustodial fathers to participate in employment-related activities. The goal of the project is to increase the earnings of these absent parents so that they can provide more financial support for their children.⁷

Evaluations of State Waiver Experiments

In his 1992 State of the Union Address, President Bush encouraged states to seek waivers of federal welfare rules in order to test innovative new programs and policies. President Clinton heightened federal support for the waiver process and also streamlined it.

The actions of both presidents sparked a flurry of state activity. By August 1996, when the new welfare law was signed, 43 states already had waivers authorizing significant changes in their programs. Waivers authorized states to impose time limits on assistance, strengthen work and training requirements, allow recipients to keep more of their welfare income when they go to work, expand child care and other services for families in work or training, impose requirements and incentives for teen parents to live at home and stay in school, and test many other policies aimed at improving the well-being of needy families with children.⁸

Since the official purpose of these waivers was to allow states to experiment with changes in federal programs, such "experiments" were required to be systematically evaluated. As a result, dozens of large-scale evaluations are in progress, with many initial reports published or soon to be published.

Many of the program changes made through waivers closely resemble the changes that are likely under the new welfare law. Hence, those evaluations that were soundly designed and implemented should contain valuable information to guide state planning and implementation decisions. In fact, for some years to come, the waiver evaluations will likely constitute the best source of information about the probable effects of various programmatic changes.

For example, MDRC recently issued a report on the early experiences in Florida, Vermont, and Wisconsin in implementing time-limited welfare under waivers.⁹ Some of its findings were: (1) Implementing time limits without adequate planning poses significant risks to both recipients and the program's credibility. States must carefully assess the need for additional staff and staff training, job training, child care, and management information systems. (2) Communicating the new program rules to recipients is extremely important but can be difficult, especially when several far-reaching changes are implemented simultaneously. It is essential to explain the new policies to recipients clearly and repeatedly. (3) Each state has expanded its JOBS program and is trying to focus on employment, but their approaches vary. For example, some place greater focus on quick job entry and others on longer-term approaches. (Impact data from these state demonstrations are only now becoming available.)

In a 1995 study, Pavetti and Duke of the Urban Institute examined waiver programs in five states (Utah, Colorado, Iowa, Michigan, and Vermont).¹⁰ The authors focused on implementation issues raised by attempts to increase JOBS participation substantially and to change the culture of welfare. Their report concluded that states have taken different approaches to reach similar goals.

Some key findings include: (1) Participation rates for work or work-related activities can be substantially raised in a relatively short period of time, but program costs rise. (2) Child care plays an important role in transforming the welfare system into a more work-oriented system. (3) If large numbers of recipients are placed in unsubsidized employment and caseloads decline substantially, those recipients left behind are likely to have multiple barriers to employment. (4) The sanctioning of clients is an important strategy for reforming the welfare system.

In November 1996, Besharov and his colleagues at the American Enterprise Institute collected information from 21 states on the application of health-related rules authorized by the waiver process.¹¹ Their major findings include: (1) Considerable state interest exists for using financial sanctions or support services to change the behavior of welfare mothers. (2) Most states adopted a program that required recipients to establish their compliance with immunization mandates, sanctioned recipients for noncompliance (either initially or after a warning), and provided "good cause" exemptions. (3) Sanctions for failing to have children immunized ranged from $25 to the entire portion of the mother's grant, usually for as many months as the family was not complying with the requirements. The limited data available suggest that the sanctions were not severely burdensome on the families involved. (4) Neither the monitoring and sanctioning process nor the provision of support services seem to have created an undue or prohibitive administrative burden. (5) Subject to various methodological and implementation problems, the results of two early evaluations suggest significant increases in immunization rates.

Dozens of additional reports will be issued in the coming months and years. The following is a sampling of what the states are doing.

California's Work Pays Demonstration Project combines benefit reductions with expanded work incentives. An interim impact report suggests that the demonstration has had little effect on employment, earnings, or welfare receipt.¹² However, the implementation analysis concluded that eligibility workers did not effectively communicate to recipients the new rules, especially the financial incentives for working, which may have influenced the absence of effects.

Colorado's Personal Responsibility and Employment Program seeks to reduce welfare dependency by encouraging participation in job training programs and increasing work incentives. A preliminary analysis shows no reduction in welfare receipt and a small increase in employment.¹³

Florida's Family Transition Program combines a two-year time limit with increased incentives for work. An interim implementation study describes some of the early challenges related to implementing a time-limited welfare program. After 15 months, it produced modest increases in employment and earnings, but it had no impact on AFDC payments. In the sixth quarter, the demonstration increased employment 15.4 percent (39.6 percent vs. 45.7 percent) and average earnings by 23.9 percent ($708 per quarter vs. $877). However, there was no statistically significant reduction in AFDC payments. The authors of the Florida report conclude that "FTP's financial work incentives have helped generate an increase in family income without raising welfare spending; however, in part because of the incentives, FTP is not reducing the rate at which people are accumulating months toward the time limit." Because these findings reflect the period before any recipients had exhausted their time-limited benefits, longer-term follow-up is needed to gauge the program's impact.¹⁴

Georgia's Preschool Immunization Project requires parents to immunize their preschool children. Failure to comply can result in the imposition of financial sanctions. An interim report indicates that the program significantly improved immunization rates.¹⁵ While encouraging, the report's findings should be examined cautiously because only about half of the AFDC families in the treatment and control groups granted permission for evaluators to examine their children's immunization records. Consequently, the data indicating significant increases in all categories of vaccinations may have been distorted by self-selection biases. Families in the treatment group that were in full compliance with the immunization requirements were presumably somewhat more likely to open their children's records to evaluators.

Iowa's Family Investment Plan tests the impact of a social contract that provides time-limited services and tough penalties for those who fail to comply with the agreement. A special report summarizes the findings of a survey of 137 cases whose cash benefits had been terminated for noncompliance: 40 percent experienced an increase in total income, primarily earnings, but nearly half experienced a drop in income, averaging $384.¹⁶ The report is the first in-depth study of what happens to families who are terminated from welfare. Forthcoming reports will assess the demonstration's impact on welfare dependency.

Maryland's Primary Prevention Initiative requires parents to ensure that children meet certain education and preventive health requirements, with sanctions imposed for each child not in compliance with the demonstration's requirements. An interim report found that the demonstration has not had a significant impact on school attendance rates.¹⁷

Michigan's To Strengthen Michigan Families program was initially designed to test the impact of various work incentives and requirements, but has since been expanded to include other objectives as well, such as increasing immunization rates and requiring minor mothers to live at home. Annual evaluation reports indicate that the state's welfare reform program has led to modest improvements in employment and earnings, and small reductions in welfare receipt.¹⁸ For families that were receiving assistance when the demonstration started, the program increased employment by 1.2 percentage points and annual earnings by $223, or 7 percent, over the four-year period they were exposed to the new policies.

Minnesota's Family Investment Program combines AFDC and Food Stamps into a single cash grant and expands work incentives. An interim report found that after six months, the program significantly increased welfare receipt and the number of families combining welfare with work, but cautions that it is too early to draw any firm conclusions about the program's impacts.¹⁹

New York State's Child Assistance Program is a voluntary alternative to AFDC. It provides enhanced work incentives for single-parent families who work and include at least one child covered by a court order for child support from the noncustodial parent. A five-year impact report found a 4 percent reduction in welfare payments and a ²⁰ percent increase in earnings.20

Ohio's Learning, Earning, and Parenting demonstration tests the impact of using welfare bonuses and sanctions to enforce school attendance requirements on teen parents. Several reports suggest that the intervention has produced modest impacts on school attendance and completion rates, as well as subsequent employment.²¹ At the four-year point, 93 percent of those assigned to the program had qualified for a financial bonus or penalty. Although LEAP increased school enrollment, attendance, and progress through the 11th grade, it did not increase high school graduation for the full sample of teen parents. It also increased employment for this group by nearly 5 percentage points (60 percent vs. 65 percent) in the fourth year of follow-up.

Utah's Single Parent Employment Demonstration is a multi-faceted welfare reform demonstration that allows families that appear eligible for AFDC to be diverted from AFDC through payments that can be up to three times the regular monthly grant. An interim report indicates that about 10 percent of all applicants have been diverted, and only a small number of diverted families have returned to AFDC. Since diversion is but one of many components of the demonstration, it is not possible to know for certain what impact it has had, but it appears to be a promising approach.²²

Wisconsin's Learnfare program tested the impact of using welfare sanctions to enforce school attendance requirements on all teens receiving welfare, not just teen parents. The evaluation found that the demonstration has had little effect on improving school attendance. It also suggests that, at least in Milwaukee, the program has suffered from substantial implementation failures.²³

Some evaluations of the waiver experiments will be halted now that a welfare reform law has passed, but many others will continue. The Department of Health and Human Services (HHS) concluded that these experimental programs could provide important information about the implementation and impact of various welfare reform strategies. Hence, in November 1996, it announced support for the continuation of these evaluations (called "State Welfare Reform Evaluation" projects). At least $7.5 million in annual funding will be available. Thirty states, representing 43 demonstration projects, responded by the deadline. As of June 1997, nine states had been funded to continue their evaluations; the other states are eligible to receive planning grants to develop evaluation plans, with $4 million in funds to be awarded through a competitive process.

In addition, HHS is conducting a "Project on State-Level Child Outcomes: Enhancing Measurement of Child Outcomes in State Welfare Evaluations." Twelve states have been selected to participate in a one-year planning project to assess how their existing welfare reform evaluations could be supplemented to provide more in-depth and uniform measures of child outcomes. This will lead to the selection of about five states, with funding of $3.7 million over a three-year period, to assess the impact of welfare reform on child well-being. The effort is intended to help states expand their data capabilities and to measure and track child outcomes on an ongoing basis. Technical assistance is being provided by Child Trends to assist the states and the federal government in the planning and implementation of this project.

Within the next two or three years, therefore, there will be dozens of reports on a myriad of state experiments, including different approaches to earnings disregards, asset limits, work requirements, sanctions, time limits, transitional benefits, family caps, immunization and school attendance requirements, requirements on teen parents, and child support enforcement.

Related Evaluations

The foregoing discussion has concentrated on research and evaluation studies that focus directly on reforms to the AFDC Program. Many other studies of equal or greater importance are also under way.²⁴ Some of the most well known are described below.

The New Chance Demonstration provided comprehensive, multi-year education, training, parenting, child care, and other services to young mothers who had children as teenagers and were also high school dropouts. After 42 months, the program raised school attendance and education attainment rates; 52 percent of the treatment group had received a high school diploma or GED compared to just 44 percent of the control group. However, it had no significant impact on many other outcomes, such as employment, earnings, welfare receipt, reading skills, and health status. In addition, the New Chance mothers were more likely to have had another pregnancy than those in the control group, and equally likely to still be on welfare.²⁵

Canada's Self-Sufficiency Project offers a temporary earnings supplement to public assistance recipients. Employment increased by 13.1 percentage points (nearly 50 percent) and average monthly earnings rose $137 (almost 60 percent) in the fifth quarter after enrollment. During this period, the incidence of welfare receipt declined 14 percentage points and average monthly welfare payments fell by $117. However, when the supplemental payment is counted as a government payment, the treatment group is more likely to receive assistance and its payments are higher.²⁶

The New Hope Project is a test of a neighborhood-based antipoverty program and welfare alternative operating in Milwaukee.²⁷ The project is still in its early stages.

� 1997 by the University of Maryland, College Park, Maryland.  All rights reserved.  No part of this publication may be used or reproduced in any manner whatsoever without permission in writing from the University of Maryland except in cases of brief quotations embodied in news articles, critical articles, or reviews.  The views expressed in the publications of the University of Maryland are those of the authors and do not necessarily reflect the views of the staff, advisory panels, officers, or trusties of the University of Maryland

Back to top