This paper investigates the timing of the impact of changes in checks and balances on macroeconomic efficiency in a panel of countries. Using Battese and Coelli's one-stage approach, we find that increasing checks and balances results in greater efficiency. Furthermore, the impact of institutional changes on efficiency is delayed, with a full effect showing after five to six years. This finding remains robust after distinguishing developed and developing countries, including country-fixed effects and using alternative functional forms of the production frontier and alternative measures of checks and balances. The relation is also robust to controlling for endogeneity.