Abstract
Parents consider whether to enroll their child in kindergarten on time, or to delay entry, known as “redshirting.” States have also moved their birthdate cutoffs to earlier in the year to create older, higher achieving cohorts. However, the impacts on the intermediate- and long-run outcomes of delayed entry are unclear, as are the mechanisms through which advantages or disadvantages may accrue. We provide plausibly causal estimates of delayed kindergarten entry impacts on achievement by exploiting a policy change in the birthdate enrollment cutoff for North Carolina public schools requiring children born in a six-week window to redshirt. We compare the outcomes of “forced” redshirts with several different counterfactual peer groups using statewide administrative data to triangulate the mechanisms through which delayed entry alters outcomes and inform educational theory. Delayed entry provides benefits to math and reading achievement, and reduced identification of having a disability; these impacts operate through cohort position and maturation advantages, and not from hold-out year experiences. Forced redshirting was differentially beneficial for students who are low-income, but further disadvantaged students of color.
Acknowledgements
The authors gratefully acknowledge the support of Gary T. Henry, Charles L. Thompson, and graduate research assistants and staff at the Carolina Institute for Public Policy. We also thank Rachel Baker, Shanyce Campbell, and Keren Horn for helpful comments on prior drafts. This research depends, in part, on data infrastructure partially funded by the North Carolina Department of Public Instruction and the North Carolina General Administration. Any errors are the sole responsibility of the authors.
Open Research Statements
Study and Analysis Plan Registration
There is no registration associated with this study.
Data, Code, and Materials Transparency
The materials, data, and code underlying the results reported in this manuscript are not openly available.
Design and Analysis Reporting Guidelines
Not applicable.
Transparency Declaration
The lead author (the manuscript’s guarantor) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Replication Statement
This manuscript reports an original study.
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Notes
1 In addition, delaying public school enrollment eligibility can have a negative effect on the maternal labor supply, especially for single mothers (Gelbach, Citation2002). In response to these concerns, states such as California have created transitional kindergarten programs, available for children born in the window between the old and new birthdate cutoffs. However, no published studies have evaluated the effect of these programs. Estimates of the impact of kindergarten birthdate policy changes on student academic outcomes would be a timely contribution to the research and policy literature.
2 Lenard and Peña (Citation2018) examine the impacts of the NC birthdate policy change on white-non-white achievement gaps for a single county (Wake).
3 Indeed, it is now widely understood across educational and policy researchers that the group to whom you compare your treated participants matters substantially for estimating and interpreting impacts (Feller et al., Citation2016). This is particularly true in early childhood education, where the counterfactual environments are more variable and selected (i.e., center-based care vs. parental care, Head Start vs. family child care home) (Friedman-Krauss et al., Citation2017; Kline & Walters, Citation2016; Morris et al., Citation2018). There is not a generally accepted counterfactual of delayed entry, despite the sole focus in relative age studies on children who just missed the chance to enter kindergarten on time. The counterfactual may be those in the same cohort not subjected to the policy change, or those at the top of the class in biological age and position in the prior cohort. We construct these analyses and explain in detail the information gleaned from them in the following section.
4 We use enrollment comparisons of adjacent kindergarten year data to identify students who were voluntarily redshirted, excluding students who were retained in kindergarten, given the extensive research on the negative impacts of retention on later outcomes (Andrew, Citation2014; Frey, Citation2005), and that retention is a school’s decision rather than the parents’.
5 The non-compliance rate based on date of birth only (regardless of complete data) was 7.2%.
6 Cook and Kang (Citation2020) demonstrate that the covariates of students in this comparison are smooth across the cutoff which we confirm, but do not reproduce the table.
7 We also estimated IV probit models for the dichotomous outcomes to assess the robustness of the OLS estimates, finding nearly identical coefficient magnitudes. The results in column 2 that include only the treated cohort and omit all non-compliers are nearly identical to our main results presented in Table 1, column 2.