An empirical comparison is made of the impact on disaggregated education costs of three types of education spending control regimes — state-level control of education spending; local school board control of spending with non-overlapping jurisdictions; and local school board control with overlapping jurisdictions. A model is specified which describes the determination of teacher wages, the teacher–student ratio, administrative costs, and other operating costs under each regime. The model is estimated using Canadian provincial-level data and the entire estimated structure of behaviour is allowed to vary across the three regimes. The estimates imply that education costs are regime dependent and that a single proxy variable is unlikely to reflect the differences between regimes. Simulations indicate that centralized state or provincial control of spending leads to the lowest teacher costs. The local control regime with non-overlapping jurisdictions is the most successful at controlling administrative and other operating costs.
Read full abstract