Instead of cross-sectional studies asking whether countries with better institutions are wealthier, economists should use time-series studies asking what happens to a particular country's outcomes when it changes its own policies or institutions, which is more relevant to governments.