DRI Working Paper No. 94
By William Easterly and Steven Pennings
High average rates of economic growth are often seen as evidence of good national leadership. This commonly- used metric substantially upward biases the growth contribution of leaders because transitory shocks also affect average growth rates. The bias is much larger in autocratic countries, because the transitory component of their growth rates tends to be much more volatile. Even identification of the best and worst leaders is prone to error due to differences in growth rates across data sources. Assuming a contemporaneous relationship between leader quality and growth, we decompose growth into a leader-specific component, a country-specific component and a (possibly auto-correlated) random component using four growth datasets and two leader datasets over 50 years and more than 100 countries. We find a very small variance in leader effects, even in autocracies. We find that only a small fraction of the variation in growth in autocracies can be explained by variation in leader quality.