Just two days after Turkey’s failed coup attempt, dissidents in neighboring Armenia seized police headquarters, took several hostages, demanded the release of a popular opposition leader and called for President Serzh Sargsyan to resign immediately. The hostage-takers never presented a credible threat to political elites and there was no evidence of a broader conspiracy to oust the government, but observers were quick to call it an unfolding coup nonetheless. The government reaction was both swift and disproportionate, as the regime blocked social media, suspended news broadcasts and arrested political opponents throughout the country.

These events highlight an overlooked consequence of coup attempts. They can drive vulnerable leaders in nearby countries to repress their rivals as they work to quickly “coup-proof” their own governments. The failed coup in Turkey altered the lens through which Armenians and the international media viewed the hostage crisis. A recently published dataset on coup-related repression shows this is not rare. Unsubstantiated accusations of coup plotting are especially likely soon after coup attempts in other countries.

Repression spreads with coup fears

Evidence from sub-Saharan Africa, the world’s most coup-prone region, shows how coup attempts motivate coup-proofing repression in nearby countries. We examined whether recent coup attempts in the region affected the most explicit form of coup-proofing: instances in which leaders – usually without presenting any evidence whatsoever – justified political purges and arrests by claiming they had uncovered a secret coup conspiracy. These events are common in sub-Saharan Africa, having happened nearly 250 times between 1960 and 2012. A recent example is Uganda’s June purge of some 30 military elites.

We found that the number of coup attempts around sub-Saharan Africa in the previous three months was a very strong predictor of this type of coup-related repression. When there had not been any recent coup attempts, sub-Saharan Africa averaged 1.43 coup-related arrests over the subsequent three months. One recent coup attempt did not meaningfully change this average (1.38), but multiple coup attempts sharply increased the number of coup-related arrests in nearby countries. At times when sub-Saharan Africa had suffered three or more coup attempts during the previous three months, the region averaged 1.91 coup-related arrests over the next three months. This equates to a 35 percent increase in this type of repression.

An example of this coup-proofing contagion occurred in the spring of 2009. Shortly after a successful coup in Madagascar, leaders around the region began accusing their rivals of plotting their own coups, usually without presenting any evidence of a conspiracy. Over the next five weeks, public announcements of foiled coup plots were used as a pretense for preemptive arrests in Cameroon, Togo, Lesotho, Guinea, Ethiopia and Kenya. Before this spree, there had been no similar accusations in sub-Saharan Africa for six months.

Coup-related repression spreads, but coup attempts do not

Coup-related repression spreads around regions because leaders fear coup attempts can also cross borders. This fear is evident in the statements of world leaders and early scholarship on coup activity. Turkish President Recep Erdogan recently likened his post-coup crackdown to stamping out a virus, an analogy that reflects the common belief that coups can be contagious. Scholars like Samuel Huntington and Ruth First observed waves of coup attempts sweeping Latin America and sub-Saharan Africa in the 1960s and similarly warned of international coup contagion.

Yet more recent research offers little reason for leaders to aggressively coup-proof when coup attempts happen in other countries. The most comprehensive and rigorous study on this topic is a newly published paper by Michael K. Miller, Michael Joseph and Dorothy Ohl. They find no evidence that coups act as a virus, either globally or regionally. This does not mean coups are never inspired by events in other countries, only that it does not hold as a general trend.

It may be the case that coup attempts are not spreading like a virus because leaders learn from events in other countries and successfully undermine coup plots before they can occur. Sean Yom offers this kind of transnational learning as a reason that Arab monarchies like Morocco, Saudi Arabia and Kuwait were more stable during the 2010-11 Arab Spring than non-monarchies like Yemen, Libya, Egypt and Syria. Unfortunately, what leaders tend to learn from foreign coup attempts is that they need to identify and remove potential coup plotters before they face a similar challenge.

All of this suggests Turkey’s coup attempt may have worse consequences for vocal dissidents in other countries than it will for other world leaders. Though coup attempts rarely occur in regional waves, coup fears spread with pernicious consequences. Erdogan’s aggressive post-coup purge is more likely to be emulated than the poorly conceived plot that sparked Turkey’s latest crisis.

Jonathan Powell is an Assistant Professor of Political Science at the University of Central Florida. Curtis Bell is a Research Associate for the One Earth Future Foundation.