Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Childcare subsidies are a critical work support for low-income families. Studies consistently demonstrate a positive relationship between childcare subsidies and employment outcomes (e.g., Schaefer, Kreader, & Collins, 2006). Few studies examine whether subsidies alter the kinds or amount of care parents use. Yet, by reducing expenses, subsidies may allow parents to purchase more care or enter the more expensive formal care market. Formal arrangements, such as centers and preschools, have been found to foster cognitive development and school readiness, especially for low-income pre-school children (e.g., Fuller, Kagan, Caspary, & Gauthier, 2002) and may be indicative of higher quality settings overall (Coley, Chase-Landsdale, & Li-Grining, 2001). As such, using subsidy policy to encourage formal care use may be a worthy policy goal.

In this paper, Women's Employment Study (WES) data are used to examine the relationship between subsidies and the kinds and amount of care low-income mothers select. Whereas most childcare research focuses on a single arrangement, this study examines how subsidy use is related to the entire package of care used for all children in a household. Moving research beyond a single arrangement is important, given that 40% of parents use multiple arrangements (Presser, 2003).

Methods. The analyses use three waves of WES data (1999, 2001, 2003). WES is a longitudinal panel study of over 700 low-income families in one Michigan county. The survey collected information on each care arrangement in the previous year for every child under 14. This allowed us to determine whether any center or preschool care was used, the proportion of all care hours in these formal arrangements compared to relative and nonrelative homes, the total number of all care hours, and the total number of care arrangements.

We first examine the cross-sectional relationship between subsidy use and the type and amount of care, controlling for important covariates. We then estimate fixed effects models using three waves of data to account for unmeasured characteristics that may bias cross-sectional statistical estimates. Person fixed-effects control for everything about a respondent that does not change over time, and year fixed effects control for year specific changes, such as the possibility that demand for subsidies changed between waves.

Results. The children of subsidy users are significantly more likely to use any formal care and report significantly more formal care hours as a proportion of all care hours. Moreover, subsidy users report significantly more care hours overall as compared to nonsubsidy users. These effects are robust to alternative model specifications. Subsidy use was not associated with more arrangements across the year; thus, subsidy users are not facing more care disruptions than nonsubsidy users.

Implications. Despite significant increases in childcare funding from 1996-2000, funding has remained flat since 2000 and only a minority of eligible families use subsidies. The results of this study demonstrating positive associations between subsidies and amount and type of care, together with previous research demonstrating positive employment effects, suggest the promise of using subsidy policy as a means of encouraging formal care and facilitating employment.