How is the development of children being studied in the Child Outcomes Study of the National Evaluation of Welfare-to-Work Strategies?
The present report focuses on a sample of about 3,000 families with young children whose development is being followed over a period of five years in the Child Outcomes Study of the National Evaluation of Welfare-to-Work Strategies. The Child Outcomes Study focuses in depth on children's development and family life in a subset of the approximately 55,000 families participating in the full seven-site evaluation of economic impacts of JOBS programs as part of the National Evaluation of Welfare-to-Work Strategies.
The children whose development is being tracked over time were all preschool-age when their mothers enrolled in the evaluation (enrollment occurred between September 1991 and January 1994). The families in the study come from three of the seven sites of the National Evaluation of Welfare-to-Work Strategies: Atlanta, Georgia; Grand Rapids, Michigan; and Riverside, California. These are study sites in which two different JOBS program approaches were tested: a labor force attachment approach, which encouraged welfare recipients to make a rapid transition to employment, and a human capital development approach, which encouraged recipients to engage first in education and training as a longer-term strategy to improve employment prospects.
What are the characteristics of the families at enrollment in the Child Outcomes Study?
Looking at the characteristics of the families in the sample when they entered the evaluation, there are numerous indications of disadvantage. For example, for the sample as a whole, a majority of mothers had never been married, though there was substantial variation across the sites (71 percent of mothers in Atlanta, 59 percent in Grand Rapids, and 43 percent in Riverside had never been married). In addition, in each of the sites, a majority of the mothers had received welfare for two years or more (about 75 percent in Atlanta; 72 percent in Grand Rapids; and 65 percent in Riverside). Between a third and half of the mothers in the three sites had limited literacy according to an assessment completed at baseline (about 48 percent in Atlanta, 33 percent in Grand Rapids, and 35 percent in Riverside). More than a third of the mothers in the sample reported a moderate to high number of depressive symptoms (about 36 percent in Atlanta; 43 percent in Grand Rapids; and 35 percent in Riverside), a prevalence of symptoms substantially greater than in community samples.
Yet in some respects the characteristics of the families contradict commonly held assumptions about families receiving welfare. For example, while it is common to think of mothers receiving public assistance as having low levels of education, many of the mothers in this sample had completed high school or received a General Educational Development (GED) degree. Indeed, in two of the study sites (Atlanta and Grand Rapids), a majority of mothers had a high school diploma or GED. Many think of families receiving public assistance as having large numbers of children, yet most of the families in the sample had only one or two children. Some view welfare as an intergenerational pattern. Yet in the present sample, most of the mothers did not report receiving welfare when they were themselves children (between one-fifth in Riverside, and one-third in Grand Rapids). Finally, most of the mothers in the sample agreed with the statement that "It's wrong to stay on welfare if you can get a job, even a job you don't like," suggesting positive attitudes about employment.
Even though all of the mothers in the sample had applied for or were receiving welfare at the start of the evaluation, there was substantial variation in terms of key background characteristics. This variation underscores the importance of asking whether children were affected differently in families of different backgrounds. For example, there were sizeable subgroups within this sample who had and had not completed high school or a GED; with higher and lower reading and math literacy scores; with few and many symptoms of depression.
How were the data for this report collected?
This is the second report from the Child Outcomes Study. The first report described the family context and the children's development in one study site at the start of the evaluation.(5) In the current report, for the first time, the possibility of program impacts on young children is being examined. The results come from a follow-up interview carried out with the families in all three sites approximately two years after enrollment in the evaluation, when the children were about five to seven years old. The interviews were conducted in the families' homes. During the course of the interview, a direct assessment of the child's cognitive development was carried out, and mothers completed further measures regarding the children's development. The interview also collected information about aspects of family life that may have been affected by the JOBS programs, and thus, may help to explain program impacts on the children (for example, measures of family economic circumstances, maternal psychological well-being, the home environment, and the children's child care participation).
The measures we report on here from the two-year follow-up, particularly those pertaining to the children's behavioral and emotional adjustment and health and safety, while often well validated in other research, must be seen as having the limitation of relying exclusively on mothers' reports. For example, in the area of health and safety, we do not have doctors' assessments of overall health, nor do we have hospital records data concerning emergency room visits. Mothers' reports carry with them all of the influences that can affect maternal perceptions. For example, a mother who is concerned about being able to make it to work (particularly in a job without benefits such as sick leave) may find child health issues more salient and, thus, any indications of the child's compromised health (e.g., the common cold) may lead these mothers to rate the focal child's overall health less favorably than mothers without similar employment concerns.
In the future, results will be presented from a final follow-up with the Child Outcomes Study families and children. This follow-up was conducted about five years after the families enrolled in the evaluation, when sufficient time had passed for longer-term changes in the families and children in response to JOBS to have manifested themselves. This report will include, in addition to information on assessments of the children's cognitive development and maternal report measures of the children's health and safety, and behavioral and emotional adjustment, data from teacher reports of the children's progress and adjustment in school, and children's reports of their own behavior and engagement in school.(6) The direct assessments of academic achievement are also more extensive at the five-year follow-up point.
The larger, seven site evaluation of primarily economic impacts in the National Evaluation of Welfare-to-Work Strategies includes a more limited set of questions addressed to mothers about the well-being of all of the children in the family. Impact results for these questions in the larger evaluation sample are included in the two-year follow-up report focusing on economic impacts in the full sample (Freedman, Friedlander, Hamilton, Rock, Mitchell, Nudelman, Schweder, Storto, 2000).
In addition, a brief Synthesis Report (Hamilton, with Freedman, and McGroder, 2000) is also being made available, drawing together and integrating the two-year follow-up results from the Child Outcomes Study, and the set of brief measures focusing on all children in the family in the larger evaluation.