IDEA-Aligned Estimates of Racial Disproportionality in Special Education versus Conventional Approaches: A cautionary note on included-variable bias when achievement and socioeconomic status proxy for special education need

Racial disproportionality in special education is a contested policy space. Federal oversight has traditionally focused on minority over-representation through IDEA’s significant disproportionality framework. However, observational studies report that Black students appear under-identified based on a canonical model that regresses special education receipt on race and controls, notably prior achievement and socioeconomic status (SES). Drawing on this evidence, recent federal proposals would scale back oversight tied to significant disproportionality determinations. We formalize an IDEA-aligned estimand and show analytically that the canonical model recovers it only under strong assumptions. Most notably, it requires that the residual Black--White achievement gap net of SES reflects a gap in disability-related need rather than residual differences in opportunity to learn and other exclusionary factors. In simulations calibrated to empirically plausible environments, the canonical model overstates Black under-representation to an extent sufficient to fully account for previously reported levels. This bias stems from over-controlling for components of social inequality IDEA regards as distinct from disability. We further show popular sensitivity analyses perform poorly in addressing this issue. The implication is that negative coefficients from standard adjustments should not be interpreted as evidence of IDEA-aligned under-service without designs that better separate disability from opportunity.