In a recent piece for The National, Joshua Kurlantzick claims to spot a resurgence of military coups. He writes:
In Latin America, Africa and Asia, coups, which had been a frequent means of changing governments during the Cold War, had become nearly extinct by the dawn of the new century. But over the past decade, the military has made power grabs in at least 12 states, from Guinea to Honduras, from Thailand to Madagascar.
What does that 12 tell us, though? Have military coups really become more frequent in the past several years?
The answer is a flat “no.” The chart below plots annual counts of successful coups from 1946 through the first half of 2011, using data compiled by the Center for Systemic Peace. As the chart clearly shows, the incidence of coups has fallen substantially in the post-Cold War period and remains historically low. (NB: Those counts don’t adjust for the large increase in the number of countries worldwide in the post-Cold War era. Against that growing denominator, the rate of successful coups has declined even further than the raw counts suggest.)
Maybe coup attempts have become more common, but fewer of them are succeeding? Again, no. The following chart looks at the incidence of failed coup attempts over the same period. Again, there has been a clear decline in the past couple of decades, and that pattern has not changed noticeably in the past several years.
Kurlantzick’s story of a trend that isn’t reminds me of this anecdote from a post by Sarah Slobin at Mix Online called “The 7 1/2 Steps to Successful Infographics,” which I found through Andrew Gelman’s blog.
When I was at the NYT, there was this reporter who drove a thousand miles across country chasing this thesis that population growth was sparked near off-ramps on the interstate. It was a lovely road-trip story; he gathered amazing anecdotes and the editors loved it. Except that when we mapped the census data it didn’t support the thesis. Imagine how much gas he could have saved had he started by looking at the data.
Forget infographics for a moment; the moral of Slobin’s story applies to anyone looking for patterns or trends in the real world. Sure, some questions can’t be answered with numbers; the relevant data may not exist (yet), or the research question might involve aspects of process that are difficult to quantify. The rest of the time, however–and that’s going to be a lot of the time–it’s a good idea to look at available data before getting deeply invested in a particular answer.