TY  - JOUR
AB  - I estimate the technical efficiency of U.S. school districts in producing middle-grades achievement with a stochastic frontier model built on the National Education Cost Model (NECM), which measures each district’s spending relative to the predicted cost of national-average outcomes. Because outcomes are measured in standard-deviation units, efficiency is expressed in input terms: the share of a district’s resources, relative to cost, that a frontier district would need to reach the same outcome. Using 128,713 district-years for 8,088 districts in 38 states (2009–2024), I find a concave frontier: a 10 percent increase in resources relative to cost raises frontier outcomes by 0.066 SD in the pooled model and 0.086 SD within states. States account for 20 percent of the variance in pooled efficiency but 1 percent once state frontier intercepts are included, and those intercepts track pooled state rankings almost exactly (r = 0.96). This pattern is consistent with cross-state measurement error in linked assessments and reported spending, which attenuates the pooled frontier and inflates state differences in efficiency. District rankings are otherwise stable across specifications (rank correlation 0.88). The share of district-years in the two lowest efficiency bands doubled between 2011 and 2024, from about 11 to 22 percent, and the widening occurs within states. Relative to their own state’s averages, the least efficient districts are smaller, spend more and are more heavily staffed in every category, about 1 SD above the state mean in total staff per pupil. Differences in administrative spending, by contrast, are small in budget terms.
AU  - Bruce, Baker D.
PY  - 2026
ST  - Cost Efficiency of U.S. School Districts: Stochastic Frontier Evidence from the National Education Cost Model
TI  - Cost Efficiency of U.S. School Districts: Stochastic Frontier Evidence from the National Education Cost Model
UR  - http://www.edworkingpapers.com/ai26-1626
ER  -