Working Paper: CEPR ID: DP13050
Authors: Christian Dustmann; Thomas Cornelissen; Anna Raute; Uta Schoenberg
Abstract: In this paper, we examine the heterogeneous treatment effects of a universal child care (preschool) program in Germany by exploiting the exogenous variation in attendance caused by a reform that led to a large staggered expansion across municipalities. Drawing on novel administrative data from the full population of compulsory school entry examinations, we find that children with lower (observed and unobserved) gains are more likely to select into child care than children with higher gains. This pattern of reverse selection on gains is driven by unobserved family background characteristics: children from disadvantaged backgrounds are less likely to attend child care than children from advantaged backgrounds but have largertreatment effects because of their worse outcome when not enrolled in child care.
Keywords: universal child care; child development; marginal treatment effects
JEL Codes: J13; J15; I28
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| unobserved family background characteristics (I24) | selection into child care (J13) |
| lower observed and unobserved gains (D80) | selection into child care (J13) |
| child care attendance (J13) | treatment effects for disadvantaged children (I24) |
| treatment on untreated (TUT) (C22) | average treatment effect (ATE) (C22) |
| family care (J13) | worse outcomes (I14) |
| treatment on treated (TT) (C22) | lower probability of immediate school entry (I21) |
| unobserved characteristics (D80) | benefits from early child care (J13) |
| disadvantaged backgrounds (I24) | less likelihood of attending child care (J13) |
| children of immigrant ancestry (J11) | school readiness benefits (I21) |