Repeated Evidence of Minority Underrepresentation in Special Education

• In Event: Roundtable Session 9
  In Roundtable Session: Are We Identifying the Right Students for Special Education? Emerging Directions in Disproportionality Research

Abstract

1. Objectives and Perspective

We summarize findings from five recent studies examining minority disproportionate representation. We situate these five studies within conflicting theoretical and empirical work regarding whether minority children are disproportionately over- or under-represented in special education. We briefly summarize this conflicting work, provide both methodological and substantive limitations that may account for these conflicting findings, and report on mechanisms (e.g., school-level economic and racial composition) that may explain disparities in disability identification and treatment.

2. Methods and Data Sources

We overview multivariate analyses of two nationally representative and longitudinal datasets maintained by the U.S. Department of Education. We report results using a range of analytical methods, including multilevel logistic regression modeling and event history modeling, as well as estimates for special education placement generally and across specific disability conditions. These analyses were designed to account for an unusually extensive set of potential confounding factors (e.g., individual-level academic achievement, family economic resources, school-level racial composition) when estimating minority children’s likelihood of being identified as disabled and so placed into special education, thereby allowing for contrasts with “otherwise similar” white, English-speaking children.

3. Results

Our results repeatedly indicate that minority children are less likely than otherwise similar white, English-speaking children to be identified as disabled and so receive special education services for which they may be legally entitled. The magnitude of these disparities can be quite large (e.g., an 60% difference in the estimated odds ratios). We find that under-identification is evident both prior to and following school entry, in separate analyses of the two datasets, and both for special education generally as well as across a range of specific disability conditions. This includes under-identification for both intellectual disabilities and emotional disturbances, the two conditions which minority children are often reported to be over-identified for. For example, we find that black children are 57% (covariate-adjusted odds ratio = .43) and 64% (covariate adjusted odds ratio = .36) less likely to be identified as having intellectual disabilities and emotional disturbances, respectively, than otherwise similar white children. We provide empirical evidence indicating that attending racially segregated schools may be a mechanism that helps explain minority children’s under-representation.

4. Scholarly Significance

Results from the five studies help address substantial ambiguity for researchers, practitioners, and policymakers as to whether minority children are over- or under-represented in special education. Our work addresses previously identified substantive and methodological limitations in the existing work (e.g., National Research Council, 2002). These results also have far-reaching implications. For example, our results suggest that current federal legislation and policy, which is largely directed towards reducing minority over-representation, may be inadvertently exacerbating racial/ethnic disparities in disability identification and treatment.

Authors

• Paul L. Morgan, The Pennsylvania State University

• George Farkas, University of California - Irvine