Association for Public Policy Analysis & Management

Public pre-kindergarten (pre-k) has captured the nation’s attention, reflected in historic state pre-k expansions (Phillips et al., 2017). Most pre-k research relies on samples of 4-year-olds who constitute the major share of public pre-k participants (Barnett et al., 2014). Yet 26 states now serve 3-year-olds in their pre-k systems (Barnett et al., 2014). Head Start, the nation’s oldest public preschool program, is required to serve 3-year-olds; more than 1/3 of Head Start’s nearly 1 million enrollees are 3-year-olds (US DHHS Office of Head Start, n.d.). Other early care and education (ECE) programs targeted to low-income children, like Educare and the federal child care subsidy program (the Child Care and Development Fund), also serve 3-year-olds. All together across these programs, approximately 700,000 3-year-olds are in some center-based publicly-funded preschool. But because few studies capture data on children’s 3- and 4-year-old preschool experiences, little is known about (a) low-income children’s patterns of enrollment in publicly-funded preschool across both years, (b) predictors of these patterns, or (c) how these patterns affect children’s early educational experiences prior to kindergarten entry. This descriptive study uses newly collected data from an ongoing project in Tulsa, OK (Project SEED) to address these unanswered questions.

Project SEED is a longitudinal study in Tulsa OK, that sampled low-income 3-year-olds in 2016-2017 (N=700) and will track them through 4^th grade. In study year 1 (2016-2017), sampled children were followed into the full range of publicly-funded center-based preschool settings available to low-income 3-year-olds: 8 Tulsa Community Action Program (CAP-Tulsa) Head Start centers and 3 Educare schools in Tulsa, as well as a random sample of classrooms in each of 10 community-based child care centers serving 3-year old children on child care subsidies. In 2017-2018, we continued to follow the sample as they either (1) remained in their 3-year-old preschool setting or (2) migrated into Tulsa Public Schools (TPS) universal school-based pre-k program, available when children are 4 years of age. In both 2016-2017 and 2017-2018, comprehensive child assessment and classroom quality observation data were collected for the full sample of children and classrooms; rich covariates were collected via parent and teacher surveys.

Descriptive statistics will reveal enrollments by program type at ages 3 and 4 years, as well as patterns of migration within and across programs. Of particular interest will be factors that differentiate low-income families who do and do not migrate into TPS pre-k. OLS regression models will be estimated to predict migration into TPS pre-k (and other patterns of change vs. continuity in enrollment) from child and family variables. Consequences of differing enrollment patterns for children’s classroom experiences across the 3- and 4- year old years will also be examined.

Illuminating patterns of 3- and 4-year-old preschool enrollment, predictors of these patterns, and consequences for children’s preschool experiences will address policy-relevant questions regarding equity of access to preschool that best supports school readiness. Results will also inform program outreach and parent education efforts, as states increase their capacity to serve both 3- and 4-year-old children in publicly-funded preschool programs.