Welfare’s New Rules: A Pox on Children

Six decades of guaranteed government aid for economically deprived children ended on August 22, 1996, when President Clinton signed the Personal Responsibility and Work Opportunity Act. The law eliminated the open-ended federal entitlement program Aid to Families with Dependent Children (AFDC). In its place, it provided block grants to states under the new Temporary Assistance for Needy Families (TANF) program.

Two vocal camps have already boldly predicted the act’s consequences. Proponents say that new time limits on the receipt of cash assistance and sanctions for failure to comply with work and other requirements will propel welfare mothers into the labor force and produce abundant benefits. These include higher family income; more regular family routines; greater maternal self-esteem; more positive role models for children; and, in the longer run, declining out-of-wedlock teen births as children learn that welfare no longer provides a viable alternative to marriage.

Opponents are convinced that time limits and sanctions will alter the source but not increase the average level of family income. They predict increased stress on single mothers; lower quality of care of younger children; a reduction in the ability of mothers to monitor the activities of adolescents; and, for the many women likely to be unable to find and hold jobs, deepened family poverty to the point that even basic needs cannot be met, with attendant increases in homelessness, hunger, foster care, and health problems.

There is little scientific basis for either set of predictions. Rigorous random-assignment welfare-to-work experiments were conducted during the 1980s. Some featured sanctions for noncompliant behavior, but none imposed time limits, and few looked beyond work and welfare receipt to evaluate the impact of maternal employment on family process and child development. Also, it will be a few years before we have meaningful analysis from research projects funded by the legislation itself.

Recent declines in the number of welfare recipients, coupled with media accounts of hopeful mothers beginning to work after years on welfare, have created a kind of euphoria over welfare reform. Some think that few if any recipients will ever hit the new five-year time limit. More than a decade of research, however, suggests otherwise.

Families use welfare in disparate ways, and the majority of women who have received AFDC assistance are neither extremely long-term or shorter-term recipients. The current declines in welfare recipients undoubtedly reflect transitions off welfare by the more work-ready, short-term segment of welfare recipients. However, substantial numbers of longer-term recipient families remain, and it is unrealistic to believe that most will not have benefits cut off by the time limits. Still others will face benefit reductions or will lose benefits altogether because of sanctions.

Reform will indeed spur a substantial number of welfare-to-work transitions. Just as certainly, it will increase the depth of the poverty among families in which mothers cannot make successful transitions to full-time work. In other words, the 1996 legislation will have a much bigger impact on the distribution than on the average level of the economic well-being of children. Even if the total number of children living in low-income families does not increase dramatically, the gap between the family incomes of the poorest and better-off low-income children almost certainly will.

Recent studies suggest that such deepening poverty, especially if it occurs early in childhood, can have detrimental effects on the cognitive development of children. This is our greatest concern about the changes in the welfare system. However, we believe that various policy changes, such as gearing assistance programs to the needs of the youngest children, could help minimize the negative effects of welfare reform on children.

Sweeping change

In addition to eliminating AFDC, the new welfare law made changes affecting child care, the Food Stamp program, Supplemental Security Income for children, benefits for legal immigrants, and the Child Support Enforcement program. Also, it offers states numerous options, such as capping benefits so payments do not increase to recipients who have more children and denying assistance to unmarried teen parents and their children.

Another major provision has implications for the well-being of children and youth. Parents must engage in work or work-related activities to receive TANF. Those not participating in their state’s work requirements face reduction or termination of benefits. The legislation introduced two provisions linked to the length of welfare receipt. First, after 24 months, it requires recipients to participate in “allowable work activities” or face sanctions or penalties. Second, it sets a 60-month lifetime limit, regardless of work effort. This limit applies to the entire household and to all forms of federal assistance. States can impose shorter time limits on total receipt, and nearly half have already done so. For families currently receiving assistance, the five-year clock started when their state of residence implemented the federal block grant.

Early returns on the new state-designed welfare reforms appear to be stunningly positive. Caseloads have fallen dramatically in the past couple of years, by as much as 80 percent in some Wisconsin counties. Some pundits and politicians assure us that the five-year time limit will not, as some fear, deprive the least-able welfare recipients of their basic needs, because safeguards allow states to exempt up to one-fifth of their familes from these limits for reasons of hardship. By and large, they say, the cutoff threat provides just the jolt many welfare recipients need to get their acts together.

Much of the euphoria over welfare reform is unfounded. Specifically, the number of welfare recipients started to fall well before the legislation took effect. In fact, a 25-year look at U.S. welfare recipients shows little change, except for a rise in the 1990s. The drop in the past three years followed a period of soaring welfare rolls and actually represents a return to more typical levels. In Wisconsin, whose economy began booming earlier than most other states, the number of welfare recipients has indeed fallen dramatically. However, the typical state reductions are more modest.

Research by one of the authors of this article (Duncan) and colleagues Kathleen Harris and Johanne Boisjoly indicates that states should prepare for almost half of their welfare families to hit the five-year benefit cutoff within eight years. These estimates suggest that nearly two million families and four million children will accumulate five years of welfare receipt-more than twice as many as states can exempt for hardship. And although some affected families will make successful transitions into the labor force, many will not.

Photographs of former welfare mothers now hard at work make wonderful copy, but the truth is that there have always been hundreds of thousands of women making successful exits from the welfare rolls every year. Numerous studies have shown that typical periods of AFDC receipt are often quite short, lasting less than two years. For many, welfare provides short-term insurance against the economic consequences of a divorce, nonmarital birth, or job loss. The booming economy of recent years has made the transition back to work that much easier and faster.

Eclipsed by the optimistic predictions are the long-term welfare recipients, the families least likely to be able to support themselves and most likely to reach the welfare time limits. Profiles of recipients most likely to hit the five-year time limits are very similar to those drawn by earlier studies of long-term recipients: Two-thirds lack high school diplomas; a majority lack work experience; two-thirds were age 21 or younger when they started receiving benefits; and most have low levels of cognitive skills.

Iowa’s experience may illustrate the likely events. Beginning in 1993, welfare recipients there had to help formulate and then follow a “Family Investment Agreement.” Failure to comply led to a series of sanctions, including a six-month cutoff from all cash benefits. A follow-up study of sanctioned families found an almost equal split between those working immediately after the cash benefits ended and those not working. Nearly half of those sanctioned enjoyed monthly income increases averaging $500, but fully half suffered drops averaging nearly $400. As with welfare recipients in general, the heterogeneity of these Iowa families is key to understanding the consequences of sanctions and time limits. Roughly half of recipients may indeed respond quite successfully to sanctions but the other half will not.

It is far too early to tell exactly how children’s family incomes will be affected by welfare reforms. Much will depend on how states use their new-found freedom to design replacement programs as well as on the economic fortunes of the nation and individual regions. From the point of view of children, the most important aspects of the new legislation are the various provisions that render families ineligible for receiving benefits, including time limits and sanctions for families that do not abide by the work requirements and other rules.

Consequences of child poverty

Although the literature on the effects of poverty on children is large, it has major shortcomings. Information on some topics is old or from studies narrowly focused on local communities, or it is of limited usefulness because children’s circumstances or outcomes are measured imprecisely. For example, although income and social class are far from synonymous, some studies used variables such as occupation, single-parenthood, or low maternal education to infer family income. Because family incomes are surprisingly volatile, there are only modest correlations between economic deprivation and typical measures of socioeconomic background.

Several longitudinal data sources do collect the requisite information, making it possible to distinguish between the effects on child development of income poverty and of its correlated events and conditions. The distinction is crucial, both conceptually and because welfare reform has a much bigger impact on family income than on correlates of poverty such as low levels of schooling or lone-parent family structure.

Research focused on isolating the impact of poverty as such suggests that family income has at times large but rather selective effects on children’s development. Most noteworthy is the importance of the type of outcome being considered. To the extent that family income influences children’s development, most affected are children’s ability and achievement rather than behavior, mental health, or physical health. Also important is the childhood stage. Family economic conditions in early childhood appear to be far more important in shaping ability and achievement than they are later.

A forthcoming study by Duncan and colleagues illustrates this. Controlling for income later in childhood as well as for demographic characteristics of households, it estimates that a $10,000 increment to income averaged over the first five years of life for children in low-income families is associated with a 0.81-year increment in completed schooling and a 2.9-fold increase in the odds of finishing high school. These were much larger than the effects of income measured later in childhood. Moreover, a similar picture of income effects emerges in a comparison of sibling differences in family income and completed schooling. This suggests that income differences, rather than unmeasured persistent family characteristics, cause the achievement differences.

Other studies have found associations between income and school-related outcomes prior to high school, including children’s achievement and cognitive and verbal ability test scores. They controlled for conditions such as maternal education, maternal age at child’s birth, single parenthood, and employment. Although transient poverty has effects, the most pronounced effects were for children who were persistently poor over multiple years, even though effects of transient poverty were also seen. Also, the lowest scores were for extremely poor children (50 percent of the poverty threshold). Early cognitive, verbal, and achievement test scores strongly predict completed levels of schooling. The few studies of this show that poverty exerts much of its influence upon high school completion through its effect upon early test scores.

How poverty acts

Four key aspects or pathways illustrate how income affects children and where policy intervention may be effective: quality of the home environment, school readiness, parental health, and neighborhood. In the home, warmth of mother-child interactions, the physical condition of the home, and especially opportunities for learning account for much of the effects of family income on cognitive outcomes in young children. Differences in the home learning environments of higher- and lower-income children account for up to half of the effect of income on the cognitive development of preschool children and between one-quarter and one-third of the effect of income on the achievement scores of elementary school children. These findings are based on several large and in one case nationally representative samples of children and their families. The learning environment includes such factors as access to a library card, the practice of reading to the children, availability of learning-oriented toys and experiences, and use of developmentally appropriate activities.

School readiness depends also on the quality of the care children receive outside the home. High-quality developmentally appropriate early childhood education in the toddler and preschool years is associated with enhanced school readiness for poor and middle-income children alike. In addition, early childhood education programs for poor children increase verbal ability and reasoning skills through the early elementary years. Such programs may also decrease behavior problems and increase persistence and enthusiasm for learning.

For adolescents, family economic pressure can lead to conflict with parents, resulting in lower school grades, reduced emotional health, and impaired social relationships. Indeed, studies suggest this conflict may result from income loss or economic uncertainty due to unemployment, underemployment, and unstable work conditions, rather than from poverty or low income per se. In addition, parents who are poor are likely to be less emotionally and physically healthy. Parental mental health accounts for some of the effect of economic circumstances on child health and behavior and is associated with impaired parent-child interactions and fewer learning experiences in the home.

Finally, neighborhoods affect child and adolescent test scores and high school completion independent of family income. Poor parents are constrained in their choice of neighborhoods and schools. Low income may lead to residence in extremely poor neighborhoods characterized by social disorganization (crime, many unemployed adults, and neighbors who do not monitor the behavior of adolescents) and few resources for child development (playgrounds, child care, health care facilities, parks, and after-school programs). Families with preschoolers in poor neighborhoods tend to have fewer learning experiences, over and above the links seen between family income and learning experiences.

Judicious exemptions

Most worrisome among welfare reform provisions are time limits, sanctions for noncompliant behavior, and categorical restrictions on eligibility that drop cash assistance to zero. Some families hitting the limits or losing benefits when sanctioned for not following program rules will replace the lost welfare payments with income from work and other sources. Others, perhaps as many as half, will see their incomes fall well below the poverty line. State-specific provisions that deny cash assistance to children born to underage, unmarried women also lower dramatically the incomes of a subset of affected families.

When viewed with the welfare of young children in mind, time limits pose less of a threat than sanctions and categorical restrictions, especially in states that opt for the full five-year time limits. In those states, mothers are not likely to have young children in their households after the 60 months, unless they have additional children during that period. In contrast, sanctions and many of the categorical provisions are much more likely to deny benefits to families with very young children. Not only do young children appear to be most vulnerable to the consequences of deep poverty, but mothers with very young children are least able to support themselves.

An obvious recommendation is that states exempt families with young children from time limits, sanctions, and categorical restrictions. Some states now exempt families for up to a year after a child’s birth. Granting exemptions until a child’s second or third birthday would be far preferable. States could also establish more universal programs, such as a child allowance or refundable tax credits geared to children’s ages. Those who fear that such policies create incentives for mothers to bear additional children should be aware that evidence suggests at most weak links between fertility and the generosity of welfare benefits.

Several European countries gear time-limited benefits to the age of children in their assistance programs. In Germany, a modest parental allowance is available to a mother working fewer than 20 hours per week until her child is 18 months old. France guarantees a modest minimum income to most of its citizens, including families with children of all ages. Supplementing this basic support is the Allocation de Parent Isolé (API) program targeted at lone parents. Eligibility for generous income-tested API payments to families with children is limited to the period between the child’s birth and third birthday, even if low-income status persists beyond that point. In effect, API acknowledges a special need for income support during this period, especially if a parent wishes to care for very young children and forgo income from work. The elaborate state-funded system in France for providing child care beginning at age 3 lessens the problems associated with the parent’s transition into the labor force.

Yet another strategy is to liberate long-term recipients from welfare through a combination of cost-effective job-training and other skill-building programs. Also important are efforts to make work pay by increasing the after-tax family incomes of women who take low-wage jobs and by funding work-for-welfare jobs of last resort for those who are unable, despite effort, to find an employer.

If the goal is to promote the healthy development of children, we must go beyond cash transfer to service-delivery programs such as nutrition education and nutritional supplements, medical care, early childhood education, and housing. Existing programs in these areas could be expanded. The case for giving preference to such programs over income transfers is strongest for those addressing health and behavior, because there is little evidence that outcomes in these domains are responsive to improvements in family living standards.

Because of the relationship of home environment to cognitive ability, interventions might profitably focus on working with parents. An example is the Learningames curriculum, developed by Joseph Sparling and Isabelle Lewis at the Frank Porter Graham Center at the University of North Carolina at Chapel Hill. In the curriculum, parents receive instruction, materials, and role-playing practice in providing learning experiences. Programs that focus on teaching parenting skills and on encouraging and modeling reading skills alter parenting behavior as well as child language and school readiness. To be effective, home visits focusing on parenting skills have to be frequent (at least several times per month) and extensive (several years in duration) and have specific curricula focused on behaviors and interactions. More generally, economic logic requires a cost/benefit comparison of the various development, income transfer, and service delivery programs.

It will be years before we have a definitive accounting of the long-run effects of the 1996 welfare reforms. These reforms will almost certainly increase both the number of successful transitions from welfare to work and the number of severely economically disadvantaged children. Recent research shows that economic deprivation early in children’s lives is most harmful to their chances for achievement. Policies aimed at preventing either economic deprivation itself or its effects are likely to constitute profitable social investments.