This is the headline from a recent study I conducted in Indiana with my coauthor NaYoung Hwang. We study a statewide policy in Indiana that required students who failed a literacy test in the third grade to be held back. Students had two chances to pass the test—once in the spring of the third grade, and once during the summer after a mandatory remediation program. Students who failed both assessments were held back.
Our study shows that retained students improved tremendously in terms of their on-grade academic performance. They did not suddenly become top performers in their new third-grade classrooms, but they moved meaningfully toward the middle of the performance distribution. The positive effect of third-grade retention on test scores persisted through at least seventh grade, which was as far as we could track with our data. We also found no evidence of harmful side effects. In particular, retained students were no more likely to be subject to school discipline and had no changes in attendance after their retention.
How do we know the retention itself is responsible for the improvement in test scores, and not something else? We used what researchers call a “regression discontinuity design” to estimate the retention effect. This approach compares students whose test scores fall just above the cutoff for promotion with those just below. When we compare students very close to the cutoff, the only difference between the ones on different sides is whether they were held back—in all other respects, they are the same, at least on average. This design mimics random assignment, and gives us strong confidence that the improvements we observe were caused by retention.
Our findings in Indiana corroborate similar findings in several other cities and states showing that early-grade retention greatly improves academic outcomes without negative behavioral consequences. Retention in later grades is more problematic. This has led researchers to theorize that the negative stigma and weakened sense of belonging associated with retention are problems that affect older students, but not younger ones.
It is increasingly well understood that grade retention for struggling readers is an important part of a robust early literacy policy (e.g., see here). If students can’t read by the end of the third grade, it is in their best interest to be held back while they catch up. Missouri currently has a weak and ineffectual retention policy that as far as I can tell, is not in active or meaningful use. We should update our state retention policy to make it objective and rigorous, and stop socially promoting children who can’t read.
