Which Welfare Reforms are Best for Children?

Greg J. Duncan and
Greg Duncan
Greg J. Duncan Distinguished Professor, School of Education - University of California, Irvine
Pamela A. Morris
Pamela A. Morris Professor of Applied Psychology - New York University

September 1, 2001

The Temporary Assistance for Needy Families (TANF) program, the centerpiece of the 1996 federal welfare reform law, imposed time limits and work requirements on welfare recipients in all states. TANF also freed states to formulate a variety of sanction and incentive packages, some of which have been relatively generous. In this policy brief, we summarize results from a synthesis of nearly a dozen welfare experiments to identify which ones are most favorable for children.

How Might Child Well-Being Be Affected by Reforms?

Over the past 30 years, policymakers have struggled to devise a welfare system that would simultaneously protect children and encourage parents’ self-sufficiency. Until recently, however, there has been precious little information to inform our understanding of the effects of welfare reform policies on children. Proponents of changes in welfare policy have argued that parental employment benefits children by providing them with family role models who work and are self-sufficient, by introducing a regular schedule into the family routine, and perhaps by increasing the income available to families. But critics have argued that employment may also create stress in the family by reducing parents’ opportunities to spend time with their children and interfering with parents’ monitoring of their children’s activities, particularly in single-parent families. And for families unable or unwilling to comply with work mandates, work requirements may actually increase poverty.

Policymakers have tried a number of approaches to increase employment among single parents on welfare. One of the earliest of these were coercive policies, like those that required single parents to participate in employment-related activities as a condition of receiving their welfare benefits. Even those who believed in the importance of such requirements were concerned about the harmful effects of “forcing” single mothers to participate in activities away from their children, particularly very young children. Accordingly, parents with children under the age of three were exempted from these policies when they were first considered as part of the Family Support Act of 1988.

While these programs were effective in moving parents into employment, the jobs welfare recipients found paid very little, leaving parents only slightly better off financially than when they were receiving welfare benefits. This motivated a second approach to increasing the self-sufficiency of welfare recipients: supplementing the earnings of those who moved from welfare into employment, or in effect, increasing the incentive to go to work. Unlike policies mandating participation, there was little reason to predict any negative effects of these policies on children. The thinking was that the positive effects of increased income might counteract any negative effects of maternal employment, protecting and perhaps even improving the well-being of children.

More recently, policymakers have instituted time limits on the receipt of welfare. Critics of this approach feared that parents would not be able to find work and support their families without the safety net of welfare. The families least able to make the transition into employment could experience pronounced income loss, which ultimately might harm children. As with programs requiring participation, however, proponents of time limits hoped for better role modeling, higher maternal self-esteem, more stable family routines, and higher family incomes as parents relied more on employment than on the welfare system.

The Experimental Evidence

Five large-scale studies collectively examined the effects on children of 11 different employment-based welfare and anti-poverty programs aimed primarily at single-parent families (see “Additional Reading”). Although most of the studies were under way by 1996, they were designed to test the effects of many program features that have been implemented by the states since the federal welfare law of 1996 was passed.

These programs tested three basic approaches that are currently being used in many state welfare policies to increase the self-sufficiency of welfare recipients: earnings supplements, mandatory employment services, and time limits. Four of the programs offered generous earnings supplements designed to make work more financially rewarding by providing families with monthly cash supplements or by increasing the amount welfare recipients could keep when they went to work. Six of the programs provided only mandatory employment services-such as education, training, or immediate job searches-in which parents were required to participate to be eligible for cash welfare benefits.

One of the programs put time limits on families’ eligibility for welfare benefits, restricting eligibility to a certain number of months in a specified period. A great virtue of these experiments is that participants were randomly assigned to a “program group” that received the welfare reform package or to a “control group” that continued to live under the old Aid to Families with Dependent Children (AFDC) rules. Random assignment provides a strong basis for assessing causal impacts of the reform packages relative to the old AFDC system.

Relying on evidence from experiments has its limitations, however. The treatments in these experiments represent neither the full range of TANF programs implemented by states nor of the macroeconomic conditions-both good and bad-that states currently face or are likely to face in the next decade. Furthermore, because the experiments followed families for only a short period of time, they probably detected few of the longer-term changes in norms and expectations regarding work and childbearing that might accompany these programs.

The Effects on Elementary School-Age Children

In order to understand how various welfare policies may effect children, we classified programs into the three categories described earlier (mandatory employment services, earnings supplements, and time limits) and then examined the impact on children for each category.

In figure 1, we focus on program impacts on school achievement outcomes for younger children, most of whom were preschool and elementary school-aged when these programs started, and in elementary school when the achievement outcomes were assessed. Mothers reported most of the measures, and others were drawn from standardized tests and surveys conducted with both teachers and the children themselves.

Figure 1 shows program impacts-the standardized differences in the school achievement of the children in both program and control groups-for each category of program. Each bar represents the effect of a single program, and bars above the horizontal axis indicate that the program had a positive effect on that outcome. Bars below the axis indicate that the program had a negative effect on that outcome. Stars on top of the bars indicate program effects that are large enough to be statistically significant.

As figure 1 shows, programs that offer the most generous earnings supplements appear to have more consistently positive impacts on children than programs without these supplements. In all four of the programs that provided earnings supplements, children in the program group had significantly higher academic achievement than children in the control group. (These impacts are about the same size as moving the children from the 25th percentile to the 30th percentile on a standardized IQ test.) In comparison, just one of the six programs that provided only mandatory employment services had such positive effects and the single time-limited program had no effect on children’s school achievement.

Although less consistently than for the achievement outcomes, programs with earnings supplements appear to benefit children’s behavior and health outcomes as well, with beneficial impacts found for some of the programs with earnings supplements (not shown in figure 1). By comparison, programs with mandatory employment services or time limits had few effects across children’s behavioral and health outcomes, and the effects that were found were sometimes positive and sometimes negative.

While there are several negative effects on children’s behavior and health outcomes across all three types of programs, the positive and neutral effects are far more common than the negative effects, suggesting little evidence of the harm that critics feared. Concerns that the development of young children, especially elementary school-aged children, might be compromised by the stresses and disruptions wrought by welfare-to-work transitions receive virtually no support from these studies. This is the case even in the single study of Florida’s time-limited welfare program. Even so, it is important to recognize that in Florida, welfare officials bundled time limits together with intensive case management, a practice that many states do not follow. Moreover, the findings occurred only shortly after recipients began reaching the time limit, so we do not know what the long-term effects of such a program may be. Finally, there was little evidence of income loss from this program; a time-limited program that results in a loss of income to families may have very different effects on children.

At the same time, the positive effects of the earnings supplement programs show that certain welfare policies, when designed in ways that increase both parents’ employment and income, can benefit children. These programs benefited children’s school achievement, and sometimes benefited their behavior and health as well.

Children in Generous Programs Remain at Considerable Risk

The positive evidence of favorable program impacts on elementary school-aged children in several studies do not indicate that all the children lucky enough to be in the most generous programs are now doing well. Among children in the four earnings supplement programs, a little more than 10 percent of children had repeated a grade, and 15-20 percent had received special education in the two- to four-year period of these studies. A much smaller proportion of children in this age group had a reportedly “high” level of behavioral and emotional problems, however. One-third of the children scored in the bottom 25th percentile on a nationally standardized test of language skills, and almost 40 percent of children had long-term health problems, although most of the parents reported that their children were in very good or excellent health. Thus, even in families offered generous work supports, there were many children with school or health-related problems.

What Changes in Family Functioning May Account for Changes in Child Well-Being?

We look now at the evidence of which components of family functioning appeared to have caused the beneficial changes in child well-being in the earnings supplement programs. While we know that these programs caused the effects on children, the design of the experiments does not allow us to know exactly how these effects occurred, and therefore the conclusions here are somewhat tenuous.

First, we know that all of the generous programs increased both employment and income. What is not clear, however, is whether it was the increased income alone or the combination of increased income and employment that brought about the benefits to children.

In three of the four earnings supplement programs, mothers in the program groups were more likely to enroll their children in formal child care programs or after-school programs and extracurricular activities than were mothers in the control group. Thus, evidence from three of the programs suggests that structured programs outside of the home may be one of the pathways by which the beneficial effects to children occurred.

Surprisingly, effects on parenting behavior (including parental warmth, control, and cognitive stimulation) were remarkably limited. Also, contrary to the hopes of many welfare reformers, work preparation or employment itself did little to improve mothers’ mental health, as there were only scattered impacts on depression and stress across these programs. A likely reason for the general lack of improvement in parenting and mental health is the difficulties inherent in combining child rearing with employment in the context of economic hardship.

These findings point to parents putting their children in formal child care or after-school activities as one important way earnings supplement programs may have affected the well-being of children. However, one study puts this conclusion into question because it found that the benefits to children and increases in formal child care were caused by different policy dimensions (the former by earnings supplements, the latter by mandatory employment services). If the benefits to children were a result of increases in formal child care, we would have expected both to be caused by the same policy approach. Further research is currently being conducted to help us better understand the pathways to improved child well-being in these programs.

Impacts on Very Young Children and Adolescents

How do these programs affect infants and toddlers? Only one of the earnings supplement programs assessed impacts on very young children (in this case children aged 0-2) at the time of program enrollment. In contrast to the favorable picture for elementary school children, no impacts were found on young children’s achievement and behavior. It is unwise to generalize from a single source of data, although it seems plausible that the positive effects of increased income may have canceled out any negative effects of full-time maternal employment in this study. Some non-experimental studies have shown that very young children may be more vulnerable to the ill effects of an employment-induced separation from their mothers than older children, although the evidence here is limited. Increases in income as a result of welfare reforms are probably more important for very young children given their greater sensitivity to spells of economic deprivation and the significance of the early years for healthy development later on.

As we turn from young children to adolescents, the pattern of impacts changes for the worse. Two of the programs (one with earnings supplements and one with time limits) included assessments of adolescent well-being (see figure 2). The earnings supplement program showed that adolescent self-reports of drinking and smoking, as well as parental reports of school achievement and problem behavior, were significantly worse in the program group relative to the control group. Notably, not as many teens participated in the surveys as one would like to be truly confident in the findings. In the time-limited welfare program, parents in the program group reported lower levels of achievement and more suspensions than did parents in the control group, but no differences were found in adolescents’ fertility or involvement with police. Thus, there is some experimental evidence indicating that adolescents may be more at risk with welfare reforms that increase the self-sufficiency of welfare recipients. However, more than two studies are needed before more definitive conclusions can be made about the effects of welfare policies on adolescent children.

What Do These Findings Mean for Policy?

A key finding from the experiments is that impacts on child achievement and behavior were consistently more positive in programs that provided financial and in-kind supports (earnings supplements) for work than in those that did not. The packages of work supports were quite diverse, ranging from generous earnings supplements provided alone to more comprehensive packages of earnings supplements, child care assistance, health insurance, and even temporary community service jobs.

Although more costly than the “work first” approach taken by the programs with mandatory employment services only, two of the programs with earnings supplements had costs within the range of some of the actual welfare reform packages implemented by states in response to the 1996 legislation. Relative to the AFDC program, the average yearly cost for a participant in a program with mandatory employment services ranged from savings of $255 to a cost of $1,595. The annual costs per participant of the earnings supplement programs ranged from $2,000 to $4,000 above the costs of the AFDC program.

These findings suggest that policymakers face a choice when deciding which welfare reforms are best for children. They can increase parental self-sufficiency, provide few benefits to children, and save government money with mandatory employment service programs. Or they can increase parental employment, raise family income, provide benefits to children, and increase government spending with earnings supplement programs. Clearly, welfare policies can affect and improve the well-being of children if states or the federal government choose to spend additional money on work supports in the context of their welfare programs.

Potential ways to expand federal policies to include work supports are providing additional funds for the child care block grant, expanding health insurance coverage and participation for children, expanding participation in the food stamp program, and expanding the Earned Income Tax Credit (EITC). Notably, in three of the programs with earnings supplements, benefits to children were found even though the EITC was available for both the program and control groups. This implies that even if the EITC does have some similar positive effects on children, there is room for further improvement for families through additional work supports or through expansion of existing programs. For states that wish to increase their work supports, the challenge is how to use TANF and other dollars to fashion support packages that best meet the needs and concerns of local populations, labor markets, and policymakers.

Another policy implication of these findings is that states should be aware of the potentially differential consequences of their policies among children of different ages. Many critics have suspected that time limits, sanctions, and categorical restrictions may be more detrimental to families with very young children, as may participation requirements for mothers in the first months of their children’s lives. But the findings summarized here are the first to suggest that adolescents may experience negative effects when parents increase their self-sufficiency. If the findings hold when additional welfare reform efforts are examined, states may want to consider after-school and community-based programs for adolescents to help support working parents’ efforts to keep their children focused on school achievement and positive behavior.

With the reauthorization of TANF in 2002, policymakers will face some other difficult questions. One key question will be whether to maintain the same amount of money in the block grants that states received in 1996, given the sharp decline in welfare caseloads. For states with an interest in supporting working families, the research here suggests that those funds could be effectively used to supplement the earnings of low income workers, increase employment among welfare recipients, and give children a better start in school.

Additional Reading

Bloom, Dan, and others. 2000. The Family Transition Program: Final Report on Florida’s Initial Time-Limited Welfare Program. New York: Manpower Demonstration Research Corporation.

Bos, Johannes, and others. 1999. New Hope for People with Low Incomes: Two-Year Results of a Program to Reduce Poverty and Reform Welfare. New York: Manpower Demonstration Research Corporation.

Gennetian, Lisa, and Cynthia Miller. 2000. Reforming Welfare and Rewarding Work: Final Report on the Minnesota Family Investment Program: Volume 2: Effects on Children. New York: Manpower Demonstration Research Corporation.

McGroder, Sharon, and others. 2000. National Evaluation of Welfare-to-Work Strategies: Impacts on Young Children and Their Families Two Years After Enrollment: Findings from the Child Outcomes Study. Washington, D.C.: U.S. Department of Health and Human Services and U.S. Department of Education.

Morris, Pamela, and Charles Michalopoulos. 2000. The Self-Sufficiency Project at 36 Months: Effects on Children of a Program That Increased Parental Employment and Income. Ottawa: Social Research and Demonstration Corporation.

Morris, Pamela, and others. 2001. How Welfare and Work Policies Affect Children: A Synthesis of Research. New York: Manpower Demonstration Research Corporation.