Salah Ben Hammou and I published a commentary at The Conversation. In it we describe some simple, but quite revealing data on post-coup political trajectories in Africa. I re-post our figure from The Conversation below.

Previous decades had seen militaries quickly retreat from formal rule. In the above graph, we illustrate the number of days until the armed forces turned power over to a civilian.

This is not to say that the armed forces or an individual soldier did not remain influential, or even return to power later. What it means is that in the short term, the military at least went through the motions of turning power over to a civilian, and typically did this quickly.

Soldiers would occasionally test–and win–post-coup elections, of course. But they seemed to think it necessary to at least exit power first, even if they had little interest in remaining sidelined and were already planning to run for office.

Since 2021, this has completely reversed. None of the continent’s coup cases have seen militaries exit, not even going through the motions of appointing a civilian ally as a figurehead. In short, the armed forces no longer sees the need to even pretend to leave.

Back in late 2018, Mwita Chacha and I wrote something at the Washington Post‘s Monkey Cage in which we presented bivariate data on how often countries could be described as democracies after coups. Surprisingly, since 1990 (up to 2017 at that point), as many as half of coup cases could be described as being a democracy within three years.

This trend is important because this goes further than the military simply turning power over to someone. It suggests these political systems had hit enough criteria for comparativists to believe the country had transitioned to a bonafide democracy.

This makes the post-2020 shift all the more striking. It wasn’t just that the armed forces were pretending to exit, they previously seemed to be legitimately leaving the political scene in large numbers. This is simply no longer the case.

We have now entered a new era in which it would be difficult to assume the existence of an anti-coup norm.