The Washington PostDemocracy Dies in Darkness

A coup for democracy?

[Joshua Tucker: One of our regular features here at The Monkey Cage is summaries from political scientists of recently published research. Thanks to a generous collaboration from journal publishers, we have arranged for articles that are featured in this series to be “ungated" and made freely available to the public for a period of time following the post on The Monkey Cage. During this time, you can download (and keep) the article described in the post. So if you are interested in keeping up on recent political science research but do not currently have access to political science journals, these posts can be one way to get access to interesting recent articles.  The current post (begins below the five stars) is from political scientist Hein Goemans of University of Rochester. The article is “Coups and Democracy," which is co-authored by Goemans and University of Mannheim political scientist Nikolay Marinov, appears in the current issue of the British Journal of Political Science and will be available for free download through the end of October.]


One of the less heralded consequences of the dissolution of the Soviet Union in the early ’90s — and the subsequent financial difficulties of newly independent Russia — was that a major source of foreign aid to poor and developing countries more or less vanished overnight. In a recent article in the British Journal of Political Science, “Coups and Democracy” (available here) we show this may have promoted democracy.

In the ’90s, a new international norm was promoted by the EU and later by the US:  in the aftermath of a coup, any new regime must quickly schedule and hold elections. Failure to do so will result in the denial or withholding of foreign aid. With no alternative aid sources, coup leaders in countries dependent on foreign aid were forced to comply. (Note the example of General Salou Djibo of Niger – pictured above – who scheduled and held elections after his coup in 2010, but not before he had secured immunity for himself and his fellow coup plotters.) Thus, as we show in the article, after the fall of the Soviet Union, elections were scheduled and held significantly faster after a coup than before. If they were not, foreign aid was indeed withheld.

This sets up a selection mechanism whereby we systematically no longer see some kinds of coups. Now comes the kind of clever twist that political scientists particularly enjoy. Think of the incentives to potential coup plotters: if they successfully execute a plot, they will have to schedule elections in which the winner will be the median voter. The unpleasant alternative is a lack of funds to pay the troops and keep the economy going. Any coup plotter with policies closer to the median voter than to the current ruler won’t mind: elections will leave him or her better off than under the current leader. But if potential plotters prefer policies closer to those of the current ruler than those of the median voter, elections would leave them worse off after a coup. In anticipation of this sequence of events, such potential plotters may prefer not to launch a coup in the first place. As a result, we end up with fewer coups overall. And indeed, as expected, we did indeed find a structural break in the data in the number of coups after 1992. There are fewer coups overall, because there are fewer coups by extremists but there still are coups by plotters who launch coups that move a country closer to the median voter — and closer to democracy.