SPA Assistant Professor Claudia Persico recently published a paper in the Journal of Labor Economics identifying which environmental and social factors affect the diagnosis of disabilities in children.

The article, “School Segregation and Racial Gaps in Special Education Identification,” was co-authored by Persico, David Figlio of Northwestern University, and Todd Elder and Scott Imberman of Michigan State University.

Using data from Florida that linked birth and education records, the researchers discovered that Black and Hispanic students were overrepresented in special education programs in schools with relatively small proportions of minorities and under-identified in schools with large minority shares.

“Social comparisons among kids seem to be important in terms of diagnosing disability,” Persico said. “Students who stick out because of race are more likely to potentially get a disability diagnosis.”

However, the picture is complex. A 2020 companion paper by Elder, Figlio, Imberman, and Persico published in the American Journal of Health Economics found low birth weight to be the strongest predictor of disability. Because of health disparities affecting birth outcomes, Black and Latino babies tend to weigh less than white babies.
 
Persico hopes that these findings influence policymakers as they work to correct disproportionality in special education. Under the federal Individuals with Disabilities Education Act, this might mean that states are denying access to specialists to disabled kids who may benefit from special education, she continued.

“There are biological factors involved in disability that need to be taken into account when thinking about who gets a diagnosis,” she said.

Persico said that she and her colleagues were amazed by the consistency of the patterns in over- and under-identification. The Florida dataset was ideal: the large state maintained consistent school performance data over time, and is demographically representative of the United States, with Blacks and Latinos each comprising about 20% of its residents. These characteristics make the findings of the study generalizable to other parts of the country, Persico said.

“I hope that states and the federal government potentially take this information into consideration when thinking about disproportionality,” she continued. “There are situations in which there is significant disproportionality that happened because of structural racism . . . On average, we know that race and social class are important predictors of diagnosis with disability—and social context really does seem to matter.”