Briana Ballis and Katelyn Heath examined the effects of the Texas (US) policy that essentially capped identification of students with disabilities as eligible for special education (see the excellent series of articles reported by Brian Rosenthal in 2016 in the Houson Chronicle that exposed the policy and it consequences: Denied). For the Brookings Institution, they reported that students who were denied special education services as a consequence of that policy experienced negative outcomes, principally reduced chances of graduating from high school and of enrolling in post-secondary education.
However, Ms. Ballis and Ms. Heath provide some confusing statements. Here, for example, is a direct quote:
When students of any race are removed from SE [special education] in response to the cap on overall SE enrollment, their long-run outcomes suffer. However, when Black students are removed from SE in response to capping disproportionality, their outcomes improve.
It appears to me that the basis for the conflicting statements in the foregoing quotation [(a) students of any race have worse outcomes and (b) Black students have improved outcomes] is either ill-chosen use of the word “any” in the first part of it or an observation based on their research methods—they were able to separate main effects for average students who were denied services from specific effects for students denied services because of disproportionality. They predicate their analysis on two papers they have recently published.
One is in journals of the American Economic Association and the other in Brown University’s Annenberg Center education working papers.
Resolving the conflict will require detailed examination of the methods in those two reports, I suspect. I encourage readers who have a chance to review the papers to contribute to the comments on this SpedTalk entry.