An iridescent sunset: An empirical analysis of sunset legislation

Abstract Sunset provisions gained popularity in the 1970s as a means of ensuring state flexibility and accountability. Inefficiencies due to out-of-date agencies or regulations could be more easily curbed when there is a fixed termination date for a policy, program, or agency, pending review. This view of sunset legislation has come to be known as the Good Government hypothesis. Despite this intent, empirical studies have yet to find that sunset legislation improves efficiency or decreases waste. Rather, agencies seem to be renewed from the political inertia that they were meant to overcome. Similarly, no difference in expenditures has been observed between states with more sunset provisions compared to those with fewer. Still, past research remains correlative. The present study attempts to isolate the effect of agency-based, rule-based, and executive-led sunset provisions on state GDP by Difference-in-Difference (DiD) with staggered timing and synthetic control designs (SCM). Under a conditional parallel trends assumption, we estimate that sunset provisions lead to a $14,000 GDP per capita increase in 2022 real dollars (or a 59.8 % increase). With SCM, the only state for which we observe robust findings is Tennessee, which evinced a $3,000 per capita increase.