Association for Public Policy Analysis & Management

As parents and political leaders instinctively assume that missing school is detrimental to student achievement, a growing body of literature is beginning to provide causal evidence to substantiate this belief (Gershenson et al, 2014; Goodman, 2015; Gottfried 2014). In another unrelated literature, researchers have found large inequalities in academic performance for students receiving special education services (SWDs) compared to their general education peers (Hanushek et al. 2002). The National Assessment of Educational Progress (NAEP) tells a similar story: in 2015, 84% of SWDs score less than proficient, while 56% of general education students fall into the same category (US DOE 2015). Given that SWDs tend to miss more school than their nondisabled counterparts, we propose that absence patterns of SWDs might begin to explain some of the achievement gap. We conduct the first known empirical study examining the relationship between having a disability and absenteeism in an effort to explain some of the inequality in achievement levels between SWDs and the nondisabled student population.  

We utilize student-level, longitudinal data from the New York City Department of Education for two unique cohorts of first graders in the public school system. Students in the two cohorts enter the first grade in September 2006 or 2007, and remain in the system for five consecutive school years. This rich data set includes semester and annual measures of each student’s absence rate, as well as daily absence indicators for three scholastic years in the time period of interest.  The data also include student demographic characteristics, details on disabilities and assigned special education services, as well as school and classroom membership indicators. To examine the relationship between absences and SWDs, we analyze the association between SWDs (compared to their general education peers) and school absences.  We estimate models with grade, year, school, and individual fixed effects, as well as an array of increasingly detailed descriptions of classrooms experienced by both SWDs and general education students.  We then consider heterogeneity by gender and race. 

In preliminary results, we find statistically-significant evidence that SWDs do in fact miss more school that general education students. As we disaggregate students by types of disabilities (such as learning disability and emotional disability), we find differential relationships, such as those with emotional or learning disabilities missing the most school out of any disability group. Going further, we examined students based on homogeneous and heterogeneous classroom settings (all general education, all special education, mixed). We find no significant difference in classroom setting for SWDs, excluding those diagnosed with a speech impairment. More specifically, students with a speech impairment are more likely to miss school if placed in a homogeneous classroom. 

This study provides new evidence on one mechanism through which achievement inequalities might be generated.  If absences are higher for students with special needs compared to their general education peers, policy makers may want to turn attention to developing policies to target those students at greatest risk of engaging in this academically-damaging behavior.