The Head Start Impact Study’s experimental design facilitates analysis of child care’s effect on mothers’ participation in work and school

In an article out this month in the Review of Economics of the Household, author Cuiping Schiman notes that “there is not a consensus as to how important child care costs are to maternal labor supply and debate continues today about whether there needs to be greater government support.” The existing literature also is inconsistent about whether low-income mothers are more likely to participate in the workforce when their preschool children have access to the federally supported Head Start program. To investigate this, Schiman was able to take advantage of the experimental design of two studies distributed by the Child and Family Data Archive: the Head Start Impact Study (HSIS), 2002-2006 [United States] (ICPSR 29462) and the Third Grade Follow-up to the Head Start Impact Study (HSIS), United States, 2007-2008 (ICPSR 35003). With these data, Schiman estimated “the effect of Head Start attendance and, more broadly, non-parental care on maternal labor supply including whether a mother is employed, working full-time, working part-time, maternal schooling, household income, and receipt of public assistance.” Since the HSIS was a large randomized control trial study, its design allowed Schiman “to avoid the empirical difficulties caused by nonrandom family selection into the program.”

The HSIS was mandated by Congress after it reauthorized the Head Start program in 1998, to determine any differences between the outcomes observed for Head Start participants and what would have been observed for these same individuals had they not participated in the program. Begun in 2002, the HSIS was conducted with a nationally representative sample of nearly 5,000 three- and four-year old preschool children, who were randomly assigned to either a treatment group (which had access to Head Start services) or a comparison group (which did not have access to Head Start services, but could receive other community resources). Additionally, according to Schiman, after the second year, the children in the comparison group could apply to be enrolled in Head Start, thus creating a “reverse experiment” in which the comparison group “experienced a substantial increase in non-parental care when they became eligible for Head Start, while the treatment group experienced little change.” Study participants were followed through 2008 when they were in third grade.

Schiman’s focus was on the data collected from mothers in the study’s parental interviews, including information on demographic characteristics, employment, household income, course enrollment, and welfare assistance for both the mother and her family. Schiman found that “the availability of Head Start significantly increased the proportion of mothers using non-parental care but had no effect on the overall probability of maternal employment.” On the other hand, Schiman found that “Head Start availability facilitated full-time employment and decisions over course enrollment for some mothers.” And for mothers who first applied to Head Start when her child was three years old, “the availability of Head Start increased full-time employment by 4.7 percentage points (14.7 percent), decreased part-time employment (21 percent), and increased course enrollment (35 percent).” Schiman added that “for these mothers there is suggestive evidence of persistence in full-time employment after their child exited Head Start.” Conversely, “for mothers who first applied to Head Start when their child was four, the availability had little effect on maternal employment and schooling.” Schiman also found that “for both groups of mothers, Head Start availability did not significantly impact household income or welfare participation in SSI, TANF, food stamps, or WIC.” More publications that analyze HSIS data can be found on the recently created Head Start Impact Study (HSIS) Series page.