The long-term consequences of early childhood care

A large literature exists on the effects on child development of maternal employment outside the home during a child’s early years (Blau & Currie, 2006). In a Danish context, the relevant choice for most parents is not whether or not to stay home with the child, but, instead, what type of daycare to select. The existing international literature has shown that high-quality programs targeted at children from lower socioeconomic strata can be highly cost-effective in the long-run (Heckman, 2008). These studies find that exposing vulnerable children to a stimulating environnment already from early childhood (for instance, the Perry Preschool program in the US) promotes the formation of later skills. There is little knowledge to date, however, on the effects of universal schemes covering all children. This is despite growing political interest in and support for such programs, not just in Scandinavia, but also in the US, as was highlighted recently in President Obama’s 2013 State of the Union speech. In the EU as well, policymakers are calling for wider provision of childcare for increasing women’s attachment to the labour force by reconciling work and family life and for responding to declining fertility rates. Our earlier work (Datta Gupta & Simonsen 2010, 2012) utilized variation across municipalities in the supply of family daycare and pre-school slots to estimate the effects of different modes of care on children’s behavior at school starting age and on a number of cognitive tests and noncognitive skills measured at age 11. These studies utilized a large random sample of around 6,000 children from the 1995 birth cohort followed over time in the DALSC child longitudinal survey. The results showed that there was a positive effect on behavior at school starting age of having attended center-based preschool compared to family daycare, however, at age 11, effects were imprecisely estimated. This paper uses comprehensive Danish register data to follow up on the earlier analyses and to investigate the effects on children’s final year school GPA, exit examination scores, and on the probability of enrolling in high-school. We contribute on two frontiers: first, the use of register data on the population of Danish children is expected to give more precise and robust evidence compared to the earlier survey-based analyses. Second, looking over a longer-time span allows us to map the long-term consequences of early childhood care. All analyses are carried out for the full sample and separately by subject (language and math), child gender and according to socioeconomic background. ∗Department of Economics, Aarhus University, and IZA, Bonn †Department of Economics, Aarhus University, and IZA, Bonn