Carnegie Mellon University

John Chin

October 29, 2021

Seven years in the making: A look at John Chin's coup database

By Bill Brink

Judging the success of a coup attempt is easy. Either a new regime seized power or it did not. Analyzing failed power grabs – specifically, what constitutes one – is harder, and that’s what John Chin has been working on for the past eight years.

Chin was studying US-China relations at Princeton. He had no interest in coups until his PhD advisor, David Carter, invited him to help code different types of coups for a new database. At long last, the work is complete: Revealed in a paper in International Studies Quarterly during the summer of 2021 called "The Varieties of Coups D’état: Introducing the Colpus Dataset,” the database includes 1,923 pages of analysis covering 1,172 possible coup attempts from 1946 to 2019.

“We’re literally writing the book on post-World War II coups,” said Chin, an Assistant Teaching Professor at the Institute for Politics and Strategy. The book version, Historical Dictionary of Modern Coups D'état with co-authors Carter and Penn State’s Joseph Wright, will be published soon. 

The term coup d’état, which comes from French and literally means “stroke of state,” references a sudden and unlawful overthrow of a government by a politician or military leaders. Chin named his dataset “colpus” after the Greek word for coup. Chin and his fellow researchers relied upon preexisting databases, but this work required a deeper dive into the archives – encyclopedias, news articles, books – to accurately affix a political event with the coup moniker.

“Most coup datasets that exist, the most popular ones, are pretty basic,” Chin said. “They tend to give you a list of coup attempts in a certain set of countries on a particular date and whether they succeeded or failed. But what we wanted to do, and what I was brought on to do, was to try to distinguish between types of coups. Not all coups are the same, and theoretically scholars have known that for a long time, but empirically, with the data, all the statistical work that’s happened in the field, they hadn’t been able to measure those distinctions.”

The new dataset empowers scholars to conduct further research and allows for a more accurate estimate of the possibility of attempted challenges to power.

“Which countries are more likely to experience coups? Most people have thought about coup forecasting in a binary sort of way, ‘Is it going to have a coup attempt or not?’” Chin said. “Well, our data allows questions that are more detailed. Not only is a country more prone to a coup in general, but is it more prone to a particular type of coup than another?”