Florida's test-based promotion policy requires third-graders scoring at the lowest reading level to repeat the grade and receive a remediation bundle (summer reading program, assignment to a higher-rated teacher, intensive daily reading instruction). Schwerdt, West and Winters (2017) used a regression discontinuity design to show that this policy produced substantial short-run achievement gains that fade out within five years, better high-school performance, no effect on graduation, and no measurable effect on one cohort's in-state college enrollment. The open problem is the policy's long-run net-benefit evaluation: whether retention with remediation improves full postsecondary attainment (including delayed entry, out-of-state enrollment, and degree completion) and labor-market outcomes; which components of the retention-year bundle generate the gains; what cutoff-based estimates actually identify given that students just above the cutoff may also receive services; and how early third-grade retention compares with the counterfactual of later retention across student subgroups. As of 2026 no study answers these questions for Florida; adjacent causal evidence from Indiana, Louisiana, New York City, and France is divergent in sign and setting.

Background

Grade retention — requiring a student to repeat a grade rather than move on with their peers ("social promotion") — is among the most debated instruments of education policy. Since the 2002–03 school year, Florida has operated a test-based promotion policy: third-grade students scoring at the lowest achievement level on the state reading assessment must be retained and must receive a bundle of remedial services — a mandatory summer reading program, assignment to a higher-rated teacher when repeating the grade, and intensive daily reading instruction — unless they qualify for a "good cause" exemption. Over twenty US states have since enacted comparable third-grade reading retention laws, so the policy's evaluation bears on active legislation.

Schwerdt, West and Winters (2017) evaluate this policy with a regression discontinuity design (RD), which compares students scoring just below the promotion cutoff — who are therefore very likely to be retained — with students scoring just above it. Because crossing the cutoff raises the probability of retention only imperfectly (exemptions exist), the actual retention indicator is instrumented with cutoff crossing; the resulting estimates are local average treatment effects, valid for "compliers" — the students retained because, and only because, they missed the cutoff. The authors find large short-run reading and math gains that become statistically insignificant within five years under same-age comparisons, higher grade-level performance in high school (higher GPA, less remedial coursework), no effect on high-school graduation, and no significant effect on college enrollment — where the postsecondary analysis could observe only the first exposed cohort and only enrollment at Florida in-state institutions, leaving delayed college entry, out-of-state enrollment, degree attainment, and earnings unobserved.

Three features of this evidence base frame the open questions. First, retention is administered jointly with the remediation bundle and take-up of the individual components is unobserved, so the design cannot separate the effect of repeating the grade from the effects of the summer program, teacher reassignment, or instructional time; the paper also lacks pre-retention test data to construct value-added measures (teachers' causal contributions to achievement) for third-grade teachers. Second, and now a live identification concern, Berne, Jacob, Weiland and Strunk (2026) document for Michigan that students who just clear such a cutoff also receive intensified reading support, violating the exclusion restriction (the requirement that cutoff crossing affects outcomes only through retention) — implying that published cutoff-based retention estimates may conflate retention with remediation, with implausibly large implied retention effects. A partial-identification response now exists: Winters and Francis (2026) show that when promoted students who failed the gateway test must also receive remediation, the reduced-form effect of failing is a weighted average of retention and remediation effects, and they derive bounds on the retention effect under stated assumptions — applied to Indiana's third-grade policy, not Florida's. Third, the long-run evidence base that does exist is contextually split: Jacob and Lefgren (2004, 2009), in Chicago, find retention at best neutral for younger students but harmful for older ones; later-grade retention in Louisiana raises adult crime (Eren, Lovenheim and Mocan, 2022) and lowers attainment through college (Larsen and Valant, 2024); New York City retention at the cutoff reduces high-school progress (Mariano, Martorell and Berglund, 2025); Indiana's third-grade retention raises achievement into middle school without nonacademic harms (Hwang and Koedel, 2025); and French data show negative associations of retention with attainment and wages under non-experimental matching (Moulin and Sari, 2026). None of these identifies the Florida third-grade long-run effects, the mechanism bundle, or the early-versus-later timing comparison within one policy system. Related Florida work shows the required K-12-to-college administrative linkage, through degree attainment, is feasible (Özek, 2021), and that school-level interventions interact with the retention policy's effectiveness (Özek, 2024).

Problem Statement

The open problem is the long-run net-benefit evaluation of Florida's test-based third-grade retention policy: the long-run follow-up anticipated by its evaluators (Schwerdt, West and Winters 2017), extended with the identification concern subsequently raised for retention RD designs in the later literature. Using the policy-exposed Florida cohorts whose K-12 records can now be linked to national postsecondary and labor-market data, determine:

  1. (Long-run outcomes) The causal effects of third-grade retention under the policy on postsecondary attainment — including delayed college entry, out-of-state enrollment, and degree completion, not only single-cohort in-state enrollment — and on labor-market outcomes such as employment and earnings.
  2. (Identification) What cutoff-based regression discontinuity estimates of this policy actually identify, given that students just above the promotion cutoff may also receive remedial services (the exclusion-restriction concern raised for retention RDs by Berne, Jacob, Weiland and Strunk 2026); in particular, estimates of — or credible bounds on — the separate effects of retention itself versus the mandated remediation bundle.
  3. (Mechanism) The relative contribution of the retention-year components — mandatory summer reading program, assignment to a higher-rated teacher, and intensive daily reading instruction — to the observed short- and medium-run achievement gains.
  4. (Timing and targeting) How the long-run effects and net benefits of early (third-grade) retention compare with the counterfactual of retention in later grades for comparable students, and how effects and net benefits vary across student subgroups (for example by prior achievement, English-learner status, or school context).

These sub-questions jointly support one scientific objective — whether the short-run gains documented for this policy are net beneficial in the long run and which of its ingredients are responsible — with sub-questions 3 and 4 as the mechanism and policy-design components of the same net-benefit judgment. All four sub-questions bear jointly on this objective; its effects may be identified up to bounds or decompositions rather than by new estimates, and its target is long-run postsecondary attainment and labor-market outcomes rather than short-run achievement outcomes or a single cohort's in-state enrollment.

References

  1. Schwerdt, Guido, Martin R. West, and Marcus A. Winters. 2017. 'The effects of test-based retention on student outcomes over time: Regression discontinuity evidence from Florida.' Journal of Public Economics 152: 154–169. DOI: 10.1016/j.jpubeco.2017.06.004; https://doi.org/10.1016/j.jpubeco.2017.06.004 (working-paper version NBER WP 21509, revised April 2017)
  2. Berne, Jordan S., Brian A. Jacob, Christina Weiland, and Katharine O. Strunk. 2026. 'Assessing the impact of grade retention: A cautionary tale of exclusion restriction violations.' Journal of Public Economics 253: 105534. DOI: 10.1016/j.jpubeco.2025.105534; https://doi.org/10.1016/j.jpubeco.2025.105534
  3. Eren, Ozkan, Briggs Depew, and Stephen Barnes. 2017. 'Test-based promotion policies, dropping out, and juvenile crime.' Journal of Public Economics 153: 9–31. DOI: 10.1016/j.jpubeco.2017.07.002; https://doi.org/10.1016/j.jpubeco.2017.07.002
  4. Eren, Ozkan, Michael F. Lovenheim, and H. Naci Mocan. 2022. 'The Effect of Grade Retention on Adult Crime: Evidence from a Test-Based Promotion Policy.' Journal of Labor Economics 40(2): 361–395. DOI: 10.1086/715836; https://doi.org/10.1086/715836
  5. Figlio, David, and Umut Özek. 2023. 'The Unintended Consequences of Test-Based Remediation.' NBER Working Paper 30831. DOI: 10.3386/w30831; https://doi.org/10.3386/w30831
  6. Gilraine, Michael. 2020. 'A Method for Disentangling Multiple Treatments from a Regression Discontinuity Design.' Journal of Labor Economics 38(4): 1267–1311. DOI: 10.1086/706740; https://doi.org/10.1086/706740
  7. Hwang, NaYoung, and Cory Koedel. 2025. 'Helping or Hurting: The Effects of Retention in the Third Grade on Student Outcomes.' Educational Evaluation and Policy Analysis 47(1): 65–88 (first published online October 2023). DOI: 10.3102/01623737231197639; https://doi.org/10.3102/01623737231197639
  8. Jacob, Brian A., and Lars Lefgren. 2004. 'Remedial Education and Student Achievement: A Regression-Discontinuity Analysis.' Review of Economics and Statistics 86(1): 226–244. DOI: 10.1162/003465304323023778; https://doi.org/10.1162/003465304323023778
  9. Jacob, Brian A., and Lars Lefgren. 2009. 'The Effect of Grade Retention on High School Completion.' American Economic Journal: Applied Economics 1(3): 33–58. DOI: 10.1257/app.1.3.33; https://doi.org/10.1257/app.1.3.33
  10. Larsen, Matthew F., and Jon Valant. 2024. 'The Long-Term Effects of Grade Retention: Evidence on Persistence through High School and College.' Journal of Research on Educational Effectiveness 17(4): 615–646 (first published online August 2023). DOI: 10.1080/19345747.2023.2240323; https://doi.org/10.1080/19345747.2023.2240323
  11. Mariano, Louis T., Paco Martorell, and Tiffany Berglund. 2025. 'The Effects of Grade Retention on High School Outcomes: Evidence from New York City Schools.' Journal of Research on Educational Effectiveness 18(1): 1–31 (first published online March 2024). DOI: 10.1080/19345747.2023.2287607; https://doi.org/10.1080/19345747.2023.2287607
  12. Moulin, Léonard, and Florent Sari. 2026. 'Grade retention and long-term outcomes: evidence from French data.' Empirical Economics 71(2): article 28. DOI: 10.1007/s00181-026-02957-8; https://doi.org/10.1007/s00181-026-02957-8
  13. Özek, Umut. 2021. 'The effects of middle school remediation on postsecondary success: Regression discontinuity evidence from Florida.' Journal of Public Economics 203: 104518. DOI: 10.1016/j.jpubeco.2021.104518; https://doi.org/10.1016/j.jpubeco.2021.104518
  14. Özek, Umut. 2024. 'Does One Plus One Always Equal Two? Examining Complementarities in Educational Interventions.' SSRN working paper. DOI: 10.2139/ssrn.4756067; https://doi.org/10.2139/ssrn.4756067
  15. Opper, Isaac, and Umut Özek. 2024. 'A Global Regression Discontinuity Design: Theory and Application to Grade Retention Policies.' SSRN working paper. DOI: 10.2139/ssrn.4748020; https://doi.org/10.2139/ssrn.4748020
  16. Winters, Marcus A. 2023. 'The Cost of Retention Under a Test-Based Promotion Policy for Taxpayers and Students.' Educational Evaluation and Policy Analysis 45(4): 688–695 (first published online December 2022). DOI: 10.3102/01623737221138041; https://doi.org/10.3102/01623737221138041
  17. Winters, Marcus A., and Abigail Francis. 2026. 'Identifying Retention Effects in the Presence of Remediation-Based Exemptions.' SSRN working paper. DOI: 10.2139/ssrn.7194887; https://doi.org/10.2139/ssrn.7194887