NEW DIRECTIONS IN EARLY CHILDHOOD EDUCATION RESEARCH (APPAM Panel Proposal) It is now well-established that high-quality pre-K programs can enhance school readiness (Currie and Thomas 1999; Reynolds et al. 2002; Barnett et al. 2005; Gormley, Phillips, and Gayer 2008). Partly because of that evidence, 40 states have now established state-funded pre-K programs. The federal government has also decided to support such programs in various ways. Yet many questions remain to be answered: Do the positive short-term effects of pre-K persist over time? Does pre-K have any unintended negative effects on child care supply? Can we predict which pre-K programs will do the most to improve elementary school test scores, based on either teacher characteristics or classroom observations? In this panel, we propose to address all of these questions, by showcasing research by both younger and older scholars in three different settings: Boston, Mass., Florida, and Tulsa, Okla. Christina Weiland, Kchersti Ulvestad, Jason Sachs, and Hirokazu Yoshikawa (Harvard University and Boston Public Schools) will examine associations between children’s receptive vocabulary and executive function skills and several indicators of classroom quality, using data from a large urban pre-K program in the Boston Public Schools. They will use several classroom observation measures (ECERS, CLASS, ELLCO) and several tests (of cognitive development and executive functioning) to explore the relationship between classroom pedagogy and student performance. Daphna Bassok and Luke Miller (University of Virginia) will assess the effects of Florida’s universal voluntary pre-K (VPK) program on the overall number of child care slots, the distribution of slots between home and center based care, and the types of services provided. They will use a series of interrupted time-series models to examine impacts, leveraging the dramatic “shock” in state preschool spending that occurred in 2005 when VPK was created. Carolyn Hill, William Gormley, and Shirley Adelstein will estimate the persistence of pre-K program effects in Tulsa, for two cohorts (pre-K participants in 2000-01 and in 2005-06). They will use propensity score matching and boosted regression models to determine the effects of pre-K program participation on 3rd grade test scores, including both reading and math scores. Together these papers will illuminate both the consequences of pre-K program participation and the methods used to study these programs. Results will be of interest to public officials at both the state and federal levels, as well as the scholarly community.