CHILDREN: Whether or not children benefit depends on the program strategy and the age of the child.

The AFDC welfare system was originally created in the 1930s to protect the children of indigent widows. While normative changes in divorce and out-of-wedlock childbearing dramatically altered the composition of the welfare rolls, one of welfare's goals has remained the same: to protect children in poor, usually single-parent families headed by women. Following dramatic increases in the welfare rolls in the 1970s, however, a new goal was added: to increase parents' employment and reduce family welfare dependency. The effects on children of welfare dependency, poverty, and requirements that their mothers work have been much debated, but there has been little reliable evidence to provide solid answers. A new generation of program studies that sheds light on how welfare reform programs affected children's behavior, development, and progress in school has begun to fill in the blanks.

Logically, one might expect changes in parental employment and income to affect children differently depending on their age. For example, time spent with one's child is thought to be especially important for infants' and toddlers' development. Parental employment might be of less concern for elementary school-aged children, as long as adequate after-school care is available. Teenagers are presumably self-sufficient when it comes to after-school care, although this also means they will receive less supervision at an age when youngsters are more likely to experiment with risky behaviors. For these reasons, child outcomes are examined separately by age of child.

Welfare reform programs that led to increases in mother's employment and income - specifically, those that included earnings supplements - consistently improved the school performance of elementary school-age children. By contrast, work mandates or time limits alone had few effects on young children; there was no consistent pattern of benefit or harm.  Three-year follow-up studies of children's well-being in the three programs that offered working mothers an earnings supplement that boosted their family's income when they took jobs - Minnesota's program, Canada's Self-Sufficiency Project, and Milwaukee's New Hope program - found consistent, statistically significant positive impacts on the academic achievement of elementary school-aged children (Figure 5). Though small, the positive impacts were potentially important — equivalent to lifting average achievement levels from the 25th to the 30th percentile on a standardized test. In all three evaluations of programs that offered earnings supplement, the child well-being findings were based on reports by the children's parents and, in two of the three, were confirmed by more objective measures: In the New Hope Project, teachers were asked to rank children's school performance, and in the Self-Sufficiency Project, a standardized test was administered to assess school readiness. In both cases, program group children either ranked or scored higher than their control group peers whose parents were not eligible for the earnings supplement program. The Canadian program's positive effects, moreover, were sustained into the fifth year of follow-up (not shown), when fewer program group than control group children were reported by their parents to be performing below average in school or to be enrolled in special education classes. These differences were generally smaller in magnitude at the 54-month follow-up than at the 36-month point, but their persistence was encouraging because parents' earnings were supplemented only for the first three years and the program's impacts on adult employment and income had fallen to zero by the four-and-one-half year follow-up period. Several of the earnings supplement programs also produced small positive impacts on elementary school-aged children's social behavior. Programs that increased employment without increasing income (such as the mandate programs in Figure 5) did not have strong effects, either positive or negative, on elementary school-aged children. Mandatory employment-services programs produced no statistically significant differences in cognitive performance (matching shapes and letters, for example) on a test of school readiness at either the two- or five-year follow-up point, with one exception from the NEWWS study: In the Atlanta job-search-first program, results were positive at the two-year point but not at the five-year point (not shown in Figure 5). Program impacts on the children's social behavior were more common, but they were neither consistently positive nor nega-tive at the five-year follow-up point. Some programs seemed to benefit children, increasing positive behavior (such as being sensitive to others) and decreasing negative behavior (such as fighting with others), while other programs had the opposite effects. The story may differ in states that use full-family sanctions rather than the partial sanctions in place when these studies were conducted, particularly if full-family sanctions are more likely to result in loss of income for families. A similar lack of consistent positive or negative findings was found in the two programs that combined mandates and welfare-based work incentives with time limits (not shown in Figure 5) - Florida's Family Transition Program and Connecticut Jobs First - but neither was there evidence of pervasive harm. Recall that these programs typically increased parents' employment and income until recipients hit the benefit time limit, after which the income gain disappeared while the employment gain was sustained. In both studies, effects on children were assessed only after parents reached the time limit. No effects on academic achievement were found in either study, and effects on children's social behavior were mixed. Some small positive effects occurred among elementary school children in the Connecticut study; their mothers reported fewer problem behaviors and more positive behaviors (such as how much they helped, shared, or cooperated with other children). Overall, few effects were found in the Florida study, including somewhat more negative effects for a more advantaged group of children whose parents were both better educated and had more work experience and fewer years on welfare than other members of the study sample. Back to children summary Regardless of program approach, policies that led to increases in mothers' employment plausibly led, in turn, to small negative effects on adolescents' progress in school. On average, however, these policies did not lead to increases in more serious problems like school suspensions, dropout rates, or teenage childbearing.  When the state evaluations discussed in this guide began, the focus was on young children, not adolescents. As a result, most of the studies asked parents only a few questions about their adolescent children (although in the Canadian evaluation, both parents and teenagers were asked about both school performance and behavior). Somewhat surprisingly, when effects for adolescents were found, they were almost always negative rather than positive. In particular, when impacts were averaged across seven welfare-to-work programs, an increase in mother's employment had deleterious effects on adolescents' school-related achievement in three areas: their academic performance as reported by their mothers, the percentage repeating a grade, and the percentage who received special educational services and assistance (Figure 6). In addition, at the time of the three-year follow-up survey in the Canadian study, both mothers and teenagers in the program were more likely than their counterparts in the control group to report that the teenager engaged in smoking, drinking, drug use, and staying out late. But while these effects raised cause for concern, they were not large and, more important, the four-and-a-half-year follow-up found no evidence that these problems led to more deleterious outcomes like increased school dropout. Interestingly, negative effects were found in mandatory employment services programs, time-limit programs, and earnings supplement programs, suggesting that adolescents can be adversely affected when their mothers take jobs, whether or not income also rises. Small negative effects were also found in the voluntary New Hope program, which served a wide range of low-income single- and two-parent families, not just welfare recipients. These findings imply that any development that increases job-taking (such as a strong economy or an expanded Earned Income Credit and not just welfare reform) could have some negative impacts on the school performance of low-income adolescents. These persistent negative impacts on teens were concentrated notably among adolescents with younger siblings. As well as showing larger unfavorable effects on school performance and receipt of special educational services than did the full sample, program group adolescents with younger siblings were more likely than their control group peers to be suspended or expelled from and to drop out of school. One explanation for this impact is that as mothers' labor force participation rates have increased, their adolescent children may have received less supervision, even as they had to provide care for their siblings - an activity that takes away from schoolwork. Although the average impacts on school outcomes are not large, they appear often enough to warrant the attention of policymakers. Back to children summary The data on infants and toddlers are too limited to permit definitive conclusions, though available evidence reveals little systematic harm or benefit to very young children's later achievement or schooling when their mothers go to work.   From the few studies that have examined effects for very young children, there is little evidence that their development is either harmed or helped when their parents take jobs as a result of programs designed to encourage their employment. Probably the most extensive evaluation of program effects on very young children was undertaken in Canada's Self-Sufficiency Project, the voluntary earnings supplement program that had large initial effects on full-time employment and income. Despite concerns that increases in full-time work would reduce parents' time with their young children, no effects on children were found either at the 36-month follow-up point, when a test of language comprehension was administered to the children who would then have been age 3 or 5, or at the 54-month follow-up point, when parents were asked about their children's school performance. Nor were behavioral differences between the program and control groups found, based on maternal reports. However, maternal reports on children's development are not the ideal source of information for assessing very young children's development. A more limited set of questions about the cognitive and behavioral performance of 1- and 2-year-old children was asked of parents who were enrolled in four mandatory employment services programs in Detroit, Grand Rapids, Oklahoma City, and Portland. Here, too, no systematic effects, either negative or positive, were found. These results are encouraging but not definitive, because the few studies that examined program impacts on very young children, again, relied on their mothers to answer only a handful of questions and the programs themselves did not reflect the full range of welfare reform policies being implemented today. Back to children summary