Do leaders matter? Evidence from 56 years of US Governor successions.
In this research we provide strong evidence that leaders do matter for economic growth by estimating the "Leadership effect" in a novel, highly controlled and comparable context; a sample of US governors from 50 US states and Washington DC from 1963 to 2019. We show that governors are responsible for about 5% of the variation in state GDP growth, despite multiple constraints on their power. To estimate this effect, we integrate the latest methodological learnings by using a custom maximum likelihood function in R with an autoregressive component, to decompose the variance explained by changes in leadership.