Leveraging quasi-experimental methods to estimate model structure: Understanding school funding changes in response to court orders

Abstract

We discuss an approach to using the event study, a common experimental design in the social sciences, to parameterize delays and develop other insights into system structure. We show a step-by-step process for undertaking a delay event study, discuss some of the conceptual reasons that this provides information about the delay, and illustrate the process for a typical example. We find evidence that school funding changes following court orders do not adjust quickly, and likely follow a higher-order, as opposed to a first-order, delay process. Our tests also suggest that school district budget makers appear to forecast revenue pessimistically, contributing an additional source of delay to the system.