Association for Public Policy Analysis & Management

There is plentiful evidence that high-quality early childhood education (ECE) can increase school readiness and academic skills, and help close the achievement gap, with some studies even showing effects that last into adulthood ^1,2,3. However, other studies have found that initial benefits of attending early childhood programs often “fade out” over time, with some studies even finding that the comparison group begins to out-perform the intervention group after many years⁴. Policy-makers want to see sustained effects of their investments in ECE, and thus it is important to understand predictors of, and processes involved in, such pre-K fadeout effects.

The present study uses data from a prospective, longitudinal study (Miami School Readiness Project; n = 9,994) to examine effect sizes contrasting the academic performance of fifth and eighth graders who attended Title 1 pre-K, center-based care (CBC), or family childcare (FCC) at age 4. Individual child data included their GPA and standardized math and reading test scores in fifth and eighth grades. Multivariate GLM analyses (controlling for the selection effects of poverty status in fifth grade and child cognitive skills prior to kindergarten entry had the three-level preschool type (Title 1 pre-K, CBC, or FCC) as the IV. In fifth grade, there was a still a significant main effect for preschool type (F [1, 9,993] = 184.56, p < 0.00) for standardized math and reading tests. Children who attended public school pre-K had significantly higher math and reading scores than those who attended CBC, and children who attended CBC had significantly higher scores than children who attended FCC. On the other hand, all three ECE groups had nearly identical GPAs in fifth grade.

By eighth grade, however, not only was complete fadeout present, the pattern for the main effect for preschool type was in a different direction for math (F [1, 3,781] = 13.36, p < 0.00), reading (F [1, 3,781] = 13.36, p < 0.00), and GPA (F [1, 4,744] = 3.16, p < 0.05). Children who had previously attended CBC at age 4 now significantly out-performed children who attended public school pre-K on all outcomes, although mean differences were small. Children who attended FCC were no better or worse than pre-K or CBC children.

In summary, sustained positive effects of public school pre-K were sustained through 5^th grade, but by 8^th grade, the pattern reversed, with children who attended center-based care in the community out-performing those who attended public school pre-K on all academic outcomes. Clearly, more longitudinal research on preschool fadeout is needed to investigate how and why these patterns change over time. These are initial preliminary results. For the conference, we will have the results from additional mediational and subgroup results as well. We will investigate the role of elementary and middle school quality and school mobility as moderating factors, and determine the role that systematic attrition and student retention in grade may account for the results.