Association for Public Policy Analysis & Management

It is well established that there is a strong link between K-12 performance and later adult outcomes, such as post-secondary education attainment, teenage pregnancy, criminal activity, and adult employment and earnings. Given that differences in educational performance appear early in life and the fact that it is increasingly difficult to remediate disparities in education as children age, many have suggested prioritizing early educational interventions as a means of improving performance both in childhood and later in life. This view has its theoretical foundations in child psychology literature and is supported by early studies of high quality but small-scale pre-K programs such as the Perry Pre-School Program and the Carolina Abecedarian Project, which find substantial benefits in both the short and long-run. Fueled in part by evidence from these small-scale experiments like Perry Preschool and Abecedarian, some states initiated or significantly expanded pre-K education programs in the 1980s and 1990s. While most of these state-funded pre-K programs are means tested, as of 2017, 11 state programs (including Georgia’s) are “universal” programs that have no income restriction for participation.

This study evaluates the impact of Georgia’s state-run, universal, school-based pre-k program (Georgia Pre-K Program) on kindergarten readiness, elementary school test scores, attendance, and discipline by exploiting the random assignment of pre-K slots in over-subscribed schools. The study also evaluates heterogeneous effect of school based pre-K program based on student’s socio-economic status. We use a unique longitudinal dataset created by Metro Atlanta Policy Lab for Education (MAPLE) that includes information for four large school districts in metro Atlanta, each of which operates dozens of pre-K classes, many of which are oversubscribed. Correlational evidence from the study suggests that school-based Georgia Pre-K program is associated with 3% of a standard deviation increase in Math and English test scores. Among FRL eligible sample the same Pre-K program is associated with almost 4.5% of a standard deviation increase in math and English test scores. A trend study shows that among FRL sample, school-based Georgia Pre-K students do better than others in Math from grades 3 to 7 but others catch up by grade 8. For English the school-based Georgia Pre-K students do better than others all through grades 3 to 8. School-based Georgia Pre-K students do not do statistically significantly better than others in Social Science and Science in most grades between 3 and 8.