Possible Effects of Welfare
There are numerous ways that receiving welfare may affect children's development (Zaslow et al., 1994). The cash payments provided by AFDC may allow mothers to provide better environments for their children (through better food, housing, clothing, etc.) than they could without such assistance. Alternatively, if a parent could both earn more income and provide a constructive role model to her children through employment than by receiving AFDC, welfare may deprive children of beneficial resources and experiences.
AFDC receipt may also have noneconomic effects. If, for example, the alternative to receiving AFDC is child care of substandard quality, then the additional time in maternal care that AFDC receipt permits may offer a better alternative for children. However, if welfare receipt discourages marriage or otherwise compromises family stability, then its developmental effects may be negative.
Welfare receipt, as these examples illustrate, can plausibly have both positive and negative effects. The net effect on children depends on whether the positive effects outweigh the negative ones. Identifying and measuring those effects is the key task of research on welfare and children's development. However, the ability to test theories about the causal role of AFDC receipt in children's development is seriously constrained by the limitations of available data. The central challenge associated with this area of research is that of selection effects.
Selection effects occur because low-income families that receive AFDC are not the same as low-income families that do not receive AFDC. For example, in the National Longitudinal Survey of Youth (NLSY)—a data set commonly used to study poverty, welfare, and child development—mothers in AFDC families had their first children at a younger age and had completed less education than mothers in non-AFDC families; in addition, AFDC families had lower marital and employment rates than non-AFDC women; see Table 1 (Moore and Driscoll, 1996). Other unmeasured differences between these groups may include worse health and mental health, fewer job skills, and higher levels of conflict in the home for the families receiving AFDC, as well as a host of other factors that are often not explicitly measured.
These differences make it extremely difficult to ascertain whether welfare itself creates problems for children, or whether children would fare poorly in these families even if they didn't receive welfare because of other factors that distinguish AFDC families. It seems likely that the selection effects themselves are most likely to tip the scales toward negative findings for welfare receipt: that is, for two women objectively similar in age, education, and family background, it is likely that the one who has received more education and training and has more initiative and better life-coping skills is more
Table 1 Differences Between AFDC Families and Non-AFDC Families
|
Receipt of AFDC |
|
Family Characteristics |
Never on Welfare |
Ever on Welfare |
Mother's age at first birth |
20.6 years |
19.0 years |
Mother's education |
12.1 years of schooling completed |
11.0 years of schooling completed |
Mother's marital status in 1992 |
81.2 percent married |
48.0 percent married |
Mother's employment status in 1992 |
68.3 percent employed |
45.5 percent employed |
Years in poverty, 1986-1991 |
0.8 years |
3.5 years |
SOURCE: Data from Moore and Driscoll (1996). |
likely to have better labor market opportunities, and thus, to work. If these qualities also make women better mothers, then selection effects will make welfare look bad for children. Because none of the results in this report is based on research that used random assignment,2 selection effects must always be considered as an alternative interpretation of the results.
Without careful efforts to separate the effects of welfare from selection effects that co-occur with welfare receipt, public policies run the risk of failure. If, for example, it is mothers' lower education levels and life skills—rather than the fact that they receive welfare—that are the more potent influences on children's development, then policies focused on removing families from welfare that do not include efforts to improve mothers' education and competence will be mistargeted, and they may well be ineffective in the long run. If the underlying problems are those that are reflected in selection effects, solving them is much harder than just changing welfare programs.
Another challenge associated with efforts to understand the developmental consequences for children of receiving AFDC is the time frame of welfare receipt. Most families that receive welfare do so for brief, often sporadic, periods of time. Repeated spells on welfare are not unusual: for example, one-half of recipients stop receiving AFDC during the first year and three-quarters stop during the first 2 years (Greenberg, 1993; Pavetti, 1993). But many families