Association for Public Policy Analysis & Management

Recent proposals at the federal and state levels aim to expand public early childhood education (ECE) programs, and specifically push for universal access to age-4 preschool programs. Yet the empirical work on the long-run impacts of ECE are generated from high-quality, small-scale interventions targeted to low-income children, and there exists limited evidence on how such a universal state-level intervention may influence children’s long-run achievement and attainment.

Our study looks to the origins of, and attendance mandates for, now universal American kindergarten programs, to identify the impact of ECE on children’s long-run outcomes. While in most states kindergarten (KG) began as a voluntary program, between 1970-2015 some states evolved to mandating KG attendance. Our quasi-experimental approach exploits different sources of policy variation to causally identify the influence of an additional year of ECE on individual outcomes. Because states adopted policies in different years, children from the same birth cohort but born in different states are subject to different KG requirements. As policies change, variation within states across time leaves some cohorts affected by mandatory KG and others not.

We use multiple data sources to construct a dataset with policy and outcome information. Key state variables include mandatory KG laws, KG enrollment, state-by-year covariates (e.g., child poverty and welfare use rates, educational expenditures), and school exit and entry laws. Data come from the Digest of Education Statistics, Bureau of Labor Statistics, and National Conference of State Legislatures. These data show that 24 states have implemented mandatory KG since 1970, and there exist very few differences between states that ever adopted and never adopted mandates across a large set of characteristics. We also use restricted access American Community Survey (2001-2015) and 2000 decennial Census data, which contain information on individual outcomes and month and year of birth, allowing us to assign respondents to KG policies. Our sample comprises pooled repeated cross-sections for individuals born between 1965 and 1995, with outcomes observed between ages 20 and 50. Outcomes include: high school completion, college attendance, years of schooling, and earnings. Individual’s month, year, and state of birth allow matching of respondents to their state policies.

We will estimate the impact of mandatory KG laws on individual outcomes using a difference-in-difference design with state fixed effects, which also include birth- and survey-year fixed effects that account for unobserved factors common to individuals born or surveyed in a specific year. To ensure that we are truly capturing the effect of mandatory KG policies on outcomes, we estimated our first-stage equation, regressing cohort-level KG enrollment on mandatory KG requirements for each state following state’s adoptions. These results show that adoption of KG mandates increased KG enrollment by between 5 and 7 percentage points. This finding serves as key evidence to further study the effects of state mandatory KG requirements on later life outcomes.

Our study results will provide a first look at how an additional year of schooling during early childhood influences long-run outcomes, in a policy context where state governments are actively considering universal preschool programs.