E. Regression Discontinuity Design: An Effectiveness-Based Evaluation of Five State Pre-Kindergarten Programs

Since 1980, the number of state pre-kindergarten (pre-K) programs has more than doubled, with 38 states enrolling more than one million children in 2006 alone. This study evaluates how five state pre-K programs affected children's receptive vocabulary, math, and print awareness skills. Taking advantage of states' strict enrollment policies determined by a child's date of birth, a regression-discontinuity design was used to estimate effects in Michigan, New Jersey, Oklahoma, South Carolina, and West Virginia. For receptive vocabulary, only New Jersey and Oklahoma yielded significant standardized impacts, though two of the three other coefficients were in a direction indicating positive effects. For math, all the coefficients were positive but only Michigan and New Jersey yielded reliable results. The largest impacts were for print awareness, where all five coefficients were positive and four were reliable in Michigan, New Jersey, South Carolina, and West Virginia. The five states were not randomly selected and, on average, have higher quality program standards than non-studied states, precluding formal extrapolation to the nation at large. However, our sample of states differed in many other ways, permitting the conclusion that state pre-K programs can have positive effects on children's cognitive skills, though the magnitude of these effects varies by state and outcome. © 2008 by the Association for Public Policy Analysis and Management.