Wednesday, December 17, 2008

The impact of leader assassinations on institutions

Zimbabwe is a mess and voices are growing louder that only a violent overthrow of Robert Mugabe can solve the humanitarian, political and economic crisis of the country. While the leader is clearly a problem, would his disappearance really change institutions?

Benjamin Jones and Benjamin Olken ask this question in a broader framework: does assassinating a leader matter? In the case of autocratic leaders, it increases the probability to democracy if the assassination attempt is successful, but decreases it if not. In ex-ante expectation, attempts only increase the likelihood of democracy by 2-3%. Democracries, however, seem to be robust to this kind of violence.

The empirics exploit the randomness of the success of an attempt. But is this randomness truly random? For example, is the leader really ripe for an attempt and/or is not protected. Thus the optimal control group should be leaders who die in office, at least among autocrats. This may make the slim positive anticipated effect on autocracies vanish.

So, all in all, it does not appear to be appropriate to call for the assassination of Mugabe. But once his reign is over, here is how to fix hyperinflation.

No comments: