Violence and Votes: Are Noncompetitive Elections in Dictatorships Worth the Risk?

a man waves an Egyptian flag after the new election
Photo credit: Sebastian Horndasch

This is the second post in a five-part, weekly series on elections in dictatorships. Visit our CoupCast site for information about our REIGN database and previous commentary on coups d’état.

Another election has come and gone in Egypt. Unsurprisingly, incumbent President Abdel Fattah al-Sisi has emerged victorious with some 92 percent of the vote. This would be more impressive had he not won 96.9 percent in the previous presidential election.

Our previous post in this series discussed some of the reasons nondemocratic leaders bother with uncompetitive elections. Here, we look to the future and ask whether these leaders suffer different coup risks in the months preceding and following these contests.

Coups and Nondemocratic Elections Since 1950

We can gain insight on how elections affect coup risk by turning to OEF Research’s REIGN dataset. Unlike other election resources, REIGN provides dates of election announcements. This allows us to compare how often coups have happened in periods with no upcoming election, the interim between an election announcement and an election, and in the months following an announced election result. Here’s what a global analysis of elections dating back to 1950 suggests.

Figure and Data Analysis by Clayton Besaw

This graph shows how often coup attempts occur in four distinct periods: the PRE-ELECTION period that begins on the election announcement date and ends on the day of the election, the six months following a GOVERNMENT VICTORY, the six months following a GOVERNMENT DEFEAT, and the NO ELECTION period when there has neither been an election in the previous six months nor an election announced for any time in the next six months.

We show these monthly coup frequencies for two sets of countries: NON-DEMOCRACIES in blue and MILITARY-PERSONALIST governments (like Egypt) in orange.

Coup Risk Rises Before Nondemocratic Elections…

Regardless of election timing, military-personalist governments face coup attempts more often than other nondemocratic political systems. Coup scholars posit that this could be due to the military being more willing to intervene in politics when another division of the military already holds power. Since 1950, military-personalist states with no election on the calendar have faced more than double the monthly coup risk relative to other types of dictatorship.

We can also see by comparing the orange bars for NO ELECTION and PRE-ELECTION that elections are an especially risky gamble for these kinds of regimes. In the months between an election announcement and election day, the frequency of coup attempts rises to 2.25 percent, or roughly 1 in 45 scheduled elections. Pre-election coup attempts like this have occurred in Guinea-Bissau (2012), Burkina Faso (2014), Guinea (2008), and elsewhere.

…But Leaders Face Fewer Coup Attempts After Electoral Wins

In the end, this risk may be worth it. In fact, there has not been a single post-victory coup attempt against a military-personalist dictatorship over the span of our dataset (1950 to present). 

Even in dictatorships, elections serve to insulate the government. Often, elections offer dictators pretexts for scrambling their cabinets to isolate potential rivals and further consolidate their core bases of power. This is why coup risk drops substantially in the six months following an election victory.

For this reason, nondemocratic leaders may find that the heightened short-term risk of a challenge before an election can be offset by the substantial decline in coup risk that follows. This, of course, assumes the leaders will win the elections they set up.

Just Don’t Lose

Nondemocratic leaders are rarely surprised by election results, but unanticipated outcomes can have disastrous consequences for political stability. In such situations, coup risk rises sharply as supporters of the outgoing government scramble to annul the results and keep their leader in power. This exact scenario unfolded in the run-up to the 1952 Cuban elections. Fulgencio Batista staged a coup and canceled elections. While a successful move, this triggered the Cuban Revolution and brought Fidel Castro to power in 1959.

The recent history of coups and elections strongly suggests President al-Sisi's hold on power is now stronger than ever. By winning the election and surviving all the strongest protests the political opposition could organize, he is now in a good position to hold power for a very long time. Our CoupCast forecasting model shows coup risk will rise again in the run-up to al-Sisi's next election, but the risk after the election will be lower than the risk immediately before it. The risks to the regime were very real before the election, but our data suggest this gamble was well worth it.