Testing for Structural Breaks in the Evaluation of Programs

Abstract

A standard methodology in program evaluation is to use time series variation to compare pre-and post-program outcomes. However, when the timing of a break in a statistical relationship can be determined only by looking at the data, then the usual distribution of the test statistic which assumes exogenous timing of the break is no longer valid. Tests for parameter instability provide a flexible framework for testing a range of hypotheses commonly posed in program evaluation. These tests help pinpoint the timing of maximal break and provide a valid test of statistical significance. These tests are particularly useful when the start date of the intervention and any effect is unclear and possibly endogenous due to implementation lags.