Explaining the Variation in Non-Classroom Staff in Texas Public Schools

This paper begins to address a gap in the education literature regarding nonclassroom staff. I study non-classroom staff across Texas and find substantial
variation in the staff-to-student ratio at both the school and district level. This
variation has significant ramifications for per-student spending. For instance, the
difference between the 10th and 90th percentile districts in terms of spending is
approximately $700 for support staff and $300 for educational aides. However,
much of this variation remains a puzzle. It cannot be explained by exogenous
school and district demographics, student achievement, or revenue increases.
Although I do find that non-classroom staff are correlated with policy choices like
smaller school size (controlling for population density) and more Special
Education students, these correlations capture only a small fraction of variation.
The wide variation in spending suggests that some staffing models are likely more
cost-efficient than others. If Texas can incentivize or teach local decision makers
to make cost-efficient hiring choices, it may thus be able to reduce unnecessary
costs and improve quality of education by better allocating the cost savings.