Methods: Data on child care subsidy receipt, earnings, and individual and family characteristics were drawn from Wisconsin administrative data. The study sample includes mothers who were eligible for child care subsidies between March 2000 and February 2001. This sample includes both mothers who received subsidies and mothers who were eligible but did not receive subsidies. Mother's subsidy receipt and earnings were followed over a duration of five years. Two methodological issues are addressed: treatment heterogeneity and self-selection by using a latent index model. A latent index model considers structural factors that link unobserved factors in child care subsidy use to unobserved factors determining labor earnings. This model estimates the overall impact of subsidies on earnings of those who received subsidies (Average Treatment Effect) and identifies mothers who benefit the most from subsidy use (Marginal Treatment Effect).
Findings: The study finds positive effects on earrings from subsidy receipt. On average, mothers who received subsidies have $1,474 more in quarterly earnings (Average Treatment Effect), but the Marginal Treatment Effect is much higher than the average treatment effect, around $3,500. The study finds that mothers who are already better off, for example, those with high school diploma, higher initial earnings, and work experiences are likely to have higher earnings than those with no high school diploma, lower initial earnings, no or less work experience, and welfare receipt.
Implications: These findings suggest that mothers who are better off upon entering the subsidy program have benefited the most from child care subsidies and suggests potential policy changes to improve the child care subsidy program for those who are disadvantaged.
