Association for Public Policy Analysis & Management

While teacher shortage for students with disabilities (SWD) is well documented, we know far less about how the shortages translate into teacher quality differences and how those teacher quality gaps vary with school-level characteristics such as poverty and achievement. Goldhaber et al (2015) and Goldhaber et al (2018) establish that there exist substantial gaps in teacher quality across the socioeconomic distribution where low-income and minority students have less access to highly qualified teachers. Given the aforementioned special education teacher shortage and the higher costs and support needs associated with students with disabilities, it seems likely that there may exist even more severe quality gaps between teachers who teach SWDs and those who teach other students. Moreover, these gaps may exacerbate already existing teacher quality gaps between low-income and minority students and their wealthier and non-minority peers. However, because SWDs are afforded more legal protections than their non-SWD peers, SWDs may potentially enjoy greater access to high quality teachers. Indeed, Gilmour and Henry (2018) find little evidence of teacher quality gaps for SWDs overall in North Carolina, though they find some gaps for specific disabilities.

In this paper we extend the work of Goldhaber et al. (2015, 2018), applying the same methods in the SWD context to examine the extent to which teacher quality gaps exist and how those gaps differ by school disadvantage. As such, we compare teacher quality gaps in two ways – within schools with similar income or achievement levels and across different income or achievement levels. To conduct this analysis, we use three-years (SY15-17) of student-teacher match data from the Los Angeles Unified School District (LAUSD). Preliminary results indicate that overall, SWDs in LA are significantly more likely to have low quality teachers as measured by teacher experience, evaluation scores, and value-added. SWDs are 4.8pp more likely to have a novice teacher, 2.6pp more likely to have a teacher in the bottom quartile of evaluation scores, and 2.3pp more likely to have a teacher in the bottom quartile of VAM scores. Of these measures, access to experienced teachers for SWDs drops precipitously as free/reduced-price lunch rates rise or average math achievement falls relative to both non-SWDs in similar schools and SWDs in more advantaged schools. Results are more mixed for other teacher quality measures.

We also examine heterogeneity by disability type (both compared to each other and compared to non-SWD students) and find some interesting differences in terms of students’ access to high quality teachers. For example, while students with autism and those who are diagnosed with specific learning disabilities are more likely to access novice teachers when they are in high poverty and low achieving schools, this is not the case for students who have Language and Speech Impairments. Finally, we explore how shifts in classroom composition (either more/less SWD) are correlated with teacher quality measures (Value-Added Measures and teacher evaluation scores) as well as other teacher outcomes such as attendance and retention.