The Impact of State Intervention on School District Fiscal Performance: Evidence from a Regression Discontinuity Design

EasyChair Preprint 1104

Abstract

State monitoring and intervention has been implemented to reveal and address fiscal problems in local governments, yet we know very little about its role in promoting financial performance in a causal sense. This paper estimates the causal effect of state intervention on fiscal performance of monitored school districts by using district-level administrative data from Illinois State Board of Education. Utilizing the system design that only low-performing districts receive intervention from the state, I employ a regression discontinuity design based on financial indicators that are introduced in the system to evaluate fiscal positions and determine intervention. Results indicate that there are precisely zero effects on future financial outcomes of school districts. However, in heterogeneity analysis, I document statistically significant positive impacts on financial indicators reflecting long-run fiscal health in a relatively long term for districts with certain characteristics. For elementary school districts, I show that state intervention improves the long-term debt capacity remaining by 15-20 percentage points more on average for districts just receiving the intervention in three to four years since the intervention, compared to those barely not. This indicates that they are less reliant on issuing long-term debt in order to meet obligations. Similarly, among accrual basis school districts, I find that the intervention decreases the value of Expenditure to Revenue Ratio by 0.035-0.050 unit more for districts barely receiving the intervention, suggesting that their budget is becoming more structurally balanced.

Keyphrases: Education Finance, Fiscal performance, School Districts, State Monitoring and Intervention

@booklet{EasyChair:1104,
  author    = {Chuanyi Guo},
  title     = {The Impact of State Intervention on School District Fiscal Performance: Evidence from a Regression Discontinuity Design},
  howpublished = {EasyChair Preprint 1104},
  year      = {EasyChair, 2019}}