A new study from Sakib Mahmud of Florida State University finds that compulsory retention of underperforming third-grade students does not, on its own, lead to meaningful gains in reading achievement. For state policymakers considering this strategy, preliminary analysis reveals that retention only contributes to improved outcomes when paired with broader literacy initiatives.
Using district-level data from the Stanford Education Data Archive (SEDA), the study isolates the causal effect of state-mandated third-grade retention policies on fourth-grade Reasoning Through Language Arts (RLA) scores between 2010 and 2019. During this period, Arizona, Mississippi, North Carolina, and Ohio implemented statewide retention mandates.
The author compares outcomes in these four treatment states to six control states that allow districts discretion over retention policies—Colorado, Maryland, Minnesota, New Mexico, Texas, and West Virginia, New Mexico—and finds that the effect of a statewide mandate on student achievement is generally not statistically significant. Students living in states with compulsory retention policies do not achieve greater score improvements on the RLA test than those in states that don’t require struggling students to repeat the third grade.
Only in Mississippi does state-level compulsory retention improve RLA scores, and Mahmud argues the positive result should be attributed to the state’s comprehensive Literacy-Based Promotion Act, which combines retention with intensive teacher training, literacy coaches, and diagnostic supports.
The author notes that the study is constrained by the use of aggregated analysis and potential overlap with confounding policies and could be improved with longitudinal data on student outcomes and more detailed state policy audits. But the study provides initial evidence that state-mandated retention in third grade alone is unlikely to improve literacy achievement. Only when integrated into broader literacy strategies are such policies likely to make a meaningful difference in student outcomes.