Association for Public Policy Analysis & Management

A growing number of school districts and states have instituted or are considering implementing policies that require students to pass standardized tests in order to be promoted to the next grade. The idea behind these policies is that struggling students will benefit from an additional year of instruction in the current grade. However, critics of these policies argue that grade retention has high costs, both for the school district that needs to provide another year of instruction and also for the student who may lose a year of labor market earnings due to having to spend more time in school. While a number of studies have found that grade retention has sizable positive effects on standardized test scores that persist for at least several years, only a small number of studies have examined the impact of grade retention on high school completion.

This study will examine the impact of grade retention in New York City Department of Education (NYCDOE) elementary and middle schools on high school performance and completion. We will use a quasi-experimental regression discontinuity research design. During our study period, students who did poorly on a standardized test given in spring are required to attend summer school. Among those assigned to summer school, grade retention is largely determined by whether or not the student scores below a test score cutoff. As long as students scoring just above or just below this cutoff are likely to be similar in all other dimensions, comparisons between students who score just above and below the cut score can be used to identify the effect of grade retention.

This paper will build on earlier work in the following ways. First, this will be the first study to estimate the causal impact of grade retention on high school completion separately by grades for grades 3-8. Second, using the score from the test given at the end of summer school as the “running variable” in the regression discontinuity analysis allows us to isolate the impact of grade retention rather than the combined impact of grade retention and remedial summer services. It also means that there is a very strong “first stage” effect of test failure on grade retention, which helps generate precise estimates of the grade retention effects. Finally, we will examine a variety of “medium-run” outcomes including dropping out, credit accumulation, graduation, and performance on statewide Regents exams.