Do educated leaders matter?

Tim Besley and co-authors have extended the brilliant (if morbid) Jones-Olken paper which estimates the impact of national leaders on economic growth, to consider the impact of education upon leadership performance.

"On average," they write, "the departure of an educated leader" -- a leader with a postgraduate education, like a Ph.D or law degree -- "leads to a 0.713 percentage point reduction in growth. This contrasts with the reduction of just 0.05 percentage points after the death of a leader who does not have a post-graduate qualification." It appears that the more educated leaders were doing something right before they were suddenly removed from office -- and that replacing them with less-educated leaders resulted, on some level, in less effective policy. There's a reverse effect as well: when comparatively less-educated leaders die, their replacements are statistically likely to be more educated, and so growth tends to increase after those transitions.

If you believe these numbers, then that is a large effect, and possibly a boon to brain gain/brain circulation arguments regarding migration given the number of African Presidents with advanced education from Europe and America?

Then again, given the robust correlation between penis length and economic growth, we probably shouldn't be putting too much faith in this kind of cross-country empirical work.