WHAT ROLE DOES WELFARE RECEIPT PLAY IN CHILDREN'S DEVELOPMENT?
Many Americans are concerned about how children are affected by their parents' welfare receipt. It is commonly assumed that welfare, in and of itself, undermines the well-being and development of children. On economic grounds, some argue that receiving AFDC creates a disincentive to employment that exposes children to both lower household income and negative role models. Some also believe that welfare receipt affects noneconomic facets of family functioning, such as marriage, timing of childbearing, and the consistency and warmth of parenting in ways that expose children to potentially harmful circumstances. Research has addressed these issues by attempting to disentangle the influence of welfare receipt from that of poverty and other factors that characterize families that receive AFDC. This research is particularly critical because one of the rationales for moving families off welfare has been that such efforts will enhance children's development in both the short and long term.
Children on welfare have been found to have lower cognitive attainment and more behavior problems than other children (Duncan and Yeung, 1995; Moore and Driscoll, 1996; Santiago, 1995). However, the relationship between child outcomes and welfare receipt is clouded by the fact that, although not all poor children receive welfare, most children who live in welfare households are poor. Welfare and poverty are so closely intertwined, and it is extremely important to distinguish their independent effects on children. Doing so will answer the question of whether the relationships between welfare and negative child outcomes are due to welfare receipt, the poverty that accompanies welfare receipt, or some other factor.
Research presented at the briefing by Kristin Moore and Anne Driscoll of Child Trends, Inc. (Moore and Driscoll, 1996) assessed whether negative child outcomes related to welfare receipt are an effect of welfare dependency, a reflection of poverty, or due to some other factors that are associated with participation in welfare programs. The study used children's scores on behavior, reading, and math tests as measures of child outcomes.3
Children whose families received welfare were found to have more negative outcomes than other children. Moore and Driscoll found that welfare receipt is associated with worse reading, math, and behavior problem scores for white children and with worse reading scores for black children, but that
|
Demonstration Project The Canadian Self-Sufficiency Project The Self-Sufficiency Project is a research demonstration project begun in 1992 designed to determine the effectiveness of an earnings supplement for single-parent Income Assistance (welfare) recipients in Canada who take jobs and agree to leave Income Assistance (Social Research and Demonstration Corporation, 1992). Project participants are randomly assigned to a program group or a control group. Eligible participants in the program group are offered an earnings supplement for a 3-year period. The offer is employment driven in that only those who work full time are eligible, and it is generous enough to make work financially preferable to continued receipt of income assistance. The project seeks to determine to what extent eligible Income Assistance recipients respond to the supplement offer, what proportion of eligible recipients find jobs and receive a supplement, and whether there is variation across major population subgroups or under different labor market conditions. The study is designed to assess the effectiveness of the earnings supplement in enabling Income Assistance recipients to take jobs, earn more money, and become self-reliant. The project will also explore why some recipients refuse the supplement offer, identify the average wage and supplement payment in the jobs taken by project participants, and seek to determine what happens to participants at the end of the eligibility period. The Self-Sufficiency Project is also designed to determine if participation in the program group has any effect on the well-being of children in participating families. Assessing the impact on children's well-being will be accomplished with cognitive tests and questionnaires completed by children in selected age groups and questionnaires completed by parents. These questionnaires and tests will provide information about a set of outcomes relating to the children's academic achievement and emotional and behavioral characteristics. The project is funded by Human Resources Development Canada, a federal government department. A consortium of organizations is carrying out the project, including the Social Research and Demonstration Corporation, Statistics Canada, the provincial Income Assistance agencies, and the Manpower Demonstration Research Corporation. The project is being carried out in New Brunswick and British Columbia. |
Figure 1 Differences in Scores Between Children Never on Welfare and Children Ever on Welfare
NOTE: Controls include child's age, sex, birth order; mother's marital, fertility, work histories, and educational attainment, and score on AFQT; length and depth of poverty; contextual factors.
SOURCE: Data from Moore and Driscoll (1996)
this relationship is weakened when demographic and background factors are held constant; see Figure 1. The factors included were the child's age, sex, and birth order and the mother's marital, fertility, and work histories, educational attainment, and score on the Armed Forces Qualifying Test. When length of time spent in poverty and depth of poverty and a variety of contextual factors (such as local rates of unemploy-
ment, families in poverty, and births to teenagers) are also taken into account, the magnitude of the negative associations between welfare and child outcomes is further reduced. For example, white children whose families have ever received welfare have math scores 6.7 points lower, on average, than white children whose families have never received welfare; after controlling for demographic and background factors, the difference is reduced to 2.1 points.
These findings can be interpreted as providing evidence that welfare receipt adversely affects children's development. It is equally plausible, however, that the reported relationships between welfare and child development are due to unmeasured ways in which the home environments of the AFDC children in Moore's and Driscoll's sample were less optimal than those of the children not on welfare, and that this—rather than welfare receipt—was the ''true'' cause of the differences in children's outcomes. In fact, a long-standing research literature (Corcoran, 1995; Duncan and Brooks-Gunn, in press; Haveman and Wolfe, 1995; Hill and Duncan, 1987; Mayer, in press; Sameroff et al., 1987) has documented that children's development is compromised by a combination of factors—both economic and noneconomic—that operate in concert and tend to co-occur in low-income families. Moore's and Driscoll's research suggests that welfare receipt may well be one of these factors, particularly given that the study took into consideration an unusually large number of selection factors.
Efforts to distinguish welfare from selection effects hinge, in part, on explaining the alternatives that would face these families if they were simply removed from welfare. In some cases, the best a given mother could do off of welfare would involve stable, economically viable employment and a stable marriage or partnership; in others, it would involve very sporadic and poorly paid employment and no marriage or an unstable marriage. These disparate alternatives would likely have very different ramifications for the children involved. The challenge, then, for research and policy is to try to determine those factors that predispose families toward improved or deteriorating circumstances if they do not receive welfare.
The research presented at the briefing raises additional questions. Would the findings be the same in the context of mandated time limits for welfare receipt, which are likely to affect the characteristics of families on welfare and not on welfare? What other factors—such as the timing of welfare receipt—are important to consider in understanding the role of welfare in the lives of children? What aspects of the social and economic context within which families are embedded need to be considered in future studies of the effects of welfare on children's development? What additional indicators of children's well-being need to be assessed to more fully understand the effects of welfare receipt and welfare reform on children and families?
