Manhattan Institute Bogus Research on Florida Retention Policy

In September the sludge tank, Manhattan Institute, pumped out a smelly report requested by Jeb Bush to justify his immoral and abusive retention policy in Florida, which has used a single test to label hundreds of thousands of children failures at an early age and to leave them behind permanently. Walmart scholar, Jay Greene, concluded that the retention policy is actually good for children, since retention raises subsequent test scores. Leaving out for a second Greene's self-imposed blindness to the psychological damage that results from school failure, we now find that, surprise, surprise, the research is rigged, faulty, and, otherwise, sludge.

Thanks to Marty Solomon for the tip to the new evaluation of the study from the Think Tank Review Project at Arizona State. Summary findings of the review:

1. Florida’s retention policy has three major elements; it includes more than just repeating the same grade twice. Retained students are also required to attend a summer school intervention and to receive ongoing intensive reading instruction. The effects estimated by Greene and Winters include all of these experiences. This makes it impossible to isolate the effect of repeating the same grade from the effect of attending the summer intervention and of receiving intensive reading instruction.

2. While the study’s methodological approach is in general appropriate for the analysis Greene and Winters have conducted, the authors omit important information necessary to understand and evaluate the particular model they specified. Particularly problematic is the omission of key descriptive statistics about the characteristics of samples analyzed over the study’s two-year time period.

Even under the assumption that their instrumental variable regression analysis has been appropriately specified, the authors appear to misinterpret the retention effect they have estimated. The upshot of this misinterpretation is that the magnitude of both the one- and two-year retention effects are overstated.