Put To the Test, Unearned Passes Don't Help Kids

Of all of the controversial education policies in Chicago, perhaps none causes more heated debate than the requirement that students in third, sixth, and eighth grades must pass a standardized test to advance to the next grade. These test-based promotion requirements are intended to end “social promotion,” the widespread practice of promoting students regardless of their academic proficiency. A new Manhattan Institute study indicates that such policies lead to real academic improvement for low-performing students.

Proponents of the policies claim that simply promoting low-performing students because they are a year older does them no favors. They argue that students who cannot read at the third-grade level will fall further behind when confronted with fourth-grade material, and their problems will only compound in later grades as the gap between student skills and the material being presented grows ever larger. Thus, social promotion might be like moving a child to solid foods simply because he is “at the proper age,” regardless of whether he has teethed yet. Try as he might, the child will not be able to chew the solid food well and will not thrive.

Critics of the program consider linking promotion to test results to be at best heartless and at worst devastating to a child’s academic progress. They argue that retaining students destroys their self-esteem and stunts their academic growth. Each year, opponents of the program find sad children who will be retained because of the standardized test and parade them in front of the news cameras, insisting that the law be abandoned.

In a new empirical study, we evaluate the effect that ending social promotion has had in Florida, which has a policy very similar to Chicago’s. Florida requires third-grade students to pass a standardized reading test, the FCAT, in order to be promoted. We analyzed the test score gains of the first set of students affected by the program: third-graders in 2002-03 who failed to reach the minimum academic benchmark.

We compared the academic gains made by these students with those made by the previous year’s third-graders who had the same low test scores, but who were not subject to the retention policy because it was not yet in effect. To conduct our analysis we obtained the individual test scores of all students in the state of Florida who were in each of these groups, more than 99,000 students overall.

Our results support the use of the standardized tests to make promotion decisions. Students who were retained made test score gains greater than those of promoted students by about 4 percentile points in reading and 10 percentile points in math. Low-performing students who were given another chance to master third-grade skills improved faster than students who were socially promoted.

Critics might respond by pointing to recent studies by the Consortium on Chicago School Research, a respected research organization. They found that Chicago’s policy has had no effect on third-grade students and might have decreased the proficiency of sixth-graders.

However, their findings are by no means the final word. Instead of comparing all low-performing students in classes before and after the policy was implemented, the consortium researchers compared the test-score gains of Chicago students who scored just below and just above the passing threshold on the test. This method is less desirable; we know that the two groups they compare are substantially different from one another because they had different test scores.

The students in our analysis are more comparable because they have the same low test scores and are primarily separated only by the year in which they happened to have been born.

Furthermore, while our analysis includes all students subject to the retention policy, the consortium looks only at low-performing students who just barely missed the test score cutoff. They will miss any gains that have been made by Chicago’s lowest-performing students.

More evidence on the effects of retention policies in Chicago and elsewhere is necessary for policymakers to make informed decisions. We intend to perform future research following the same cohorts of Florida students over time to see if their academic gains continue. Until then, however, this study provides rigorous evidence that the level of a student’s achievement is a better foundation for promotion decisions than how old he is.