Association for Public Policy Analysis & Management

In late 2014, the Massachusetts Department of Early Education and Care (EEC) was awarded a federal Preschool Expansion Grant (PEG) to support the expansion of high-quality early childhood education to high-needs communities, with particular focus on serving children from very low-income families who in most cases had not received prior formal care. The Massachusetts PEG model involves collaboration between local school districts and community-based programs, with preschool services provided by the community-based programs. The 48 PEG classrooms serve four-year-olds (as of September 1 of the incoming school year) who are eligible for kindergarten in the upcoming fall and who, with some exceptions, had not yet attended a formal child care program (licensed center-based or family child care). To study the impacts of PEG on children’s school readiness, an age cut-off regression discontinuity design (RDD) study was conducted to examine whether children who had attended a PEG program had greater skills at kindergarten entry compared with similar children who did not attend PEG. The study compared the outcomes for students who attended PEG classrooms in the 2016-17 school year to outcomes for students who were just entering PEG classrooms in the 2017-18 school year. Child outcome measures included the Woodcock Johnson® III Letter-Word Identification subtest (early literacy skills) and Applied Problems subtest (early math skills); the Peabody Picture Vocabulary Test (vocabulary skills), and the Hearts & Flowers task (executive function skills). Covariates included gender, home language, and prior child care exposure.

Robust, statistically significant impacts were found on children’s early literacy and early math achievement (effect sizes of .92 and .45 standard deviations, respectively). Impacts were also seen for vocabulary skills (effect size of .21 standard deviations), but that effect was found to be sensitive to specific RDD-related analytic decisions and, therefore, less robust. No effects were seen for children’s executive function skills. Exploratory analyses also suggested differential program effects; the program had greater impacts on children who were not native English speakers and had no experience with formal child care prior to PEG entry. There were also no clear patterns of differential effects by classroom characteristics including observed classroom quality and teacher supports.