For thousands of years, close observers of politics have claimed that economic inequality causes political turmoil. In 350 B.C., Aristotle identified inequality as a principal driver of revolution and state collapse. In 19th-century Europe, Karl Marx concurred (and, in a fashion, applauded). In 2011, as the Arab Spring unfolded, Kenyan journalist John Githongo claimed in The New York Times that “radical and growing economic inequality animated much of what was at stake in the various Arab uprisings, and it will play a major role in shaping African politics for years to come.” And in a recent column for Le Monde, economist Thomas Piketty argued that “terrorism feeds on the Middle Eastern powder keg of inequality.”
Just because a belief is widely held, however, does not make it true. In fact, it’s still hard to establish with confidence whether and how economic inequality shapes political turmoil around the world. That’s largely because of the difficulty in measuring inequality; on this subject, the historical record is full of holes. Social scientists are busy building better data sets, but the ones we have now aren’t sufficient to make strong causal claims at the global level.
Inequality seems like a simple and obvious thing, but it’s hard to quantify in a routine and reliable way. Statistics on income inequality — what economists usually use to measure economic inequality — depend on careful and accurate record-keeping or high-quality household surveys. These are expensive and politically sensitive propositions. Until the 1980s, many countries weren’t even keeping such records or conducting such surveys. Some still aren’t.
In countries that do conduct the necessary surveys, however, comparing the data we get is often hard because the content and quality of the surveys vary widely. Some surveys treat the individual as the unit of analysis; others focus on households. Some ask about income; others consumption — that is, the value of goods and services people buy.
Even among countries that have tracked inequality for decades, it is often hard to compare measures for the same country over time because survey methods have occasionally changed. Eurostat, the statistical office of the European Union, publishes measures of income inequality for 28 EU members, and its tables — some of the best data on income inequality — are dotted with the “b” superscript that the agency uses to indicate a break in the time series. In the mid-2000s, a massive survey in India asked respondents for the first time about income instead of consumption. That methodological change produced a 20-point increase in the country’s Gini index, a commonly used measure of economic inequality that ranges from 0 to 100.
The result is a global patchwork quilt of data with many tattered or missing sections. The Journal of Economic Inequality devoted an entire special issue last year to appraising data sets on the subject. In the introduction, the editors noted the many gaps and discrepancies in the databases. “In some cases, different databases would lead users to radically different conclusions about inequality dynamics in certain countries and periods,” the editors wrote.
Given the intense academic and popular interest in the topic, it’s not surprising that many scholars and statistical agencies are working hard to improve measures of economic inequality. The Cross-National Data Center in Luxembourg, for example, statistically adjusts relevant data from countries that collect it to develop measures that are directly comparable across cases and over time. The Commitment to Equity project is hoping to render those kinds of adjustments unnecessary by repeating the same survey across many countries. So far, though, it has completed surveys in fewer than 20 countries and has work in progress in just 15 more.1
Even if we had reliable inequality measures, however, assessing how they relate to political crises like coups, popular uprisings and armed rebellions would still be hard. The problem is that those kinds of crises are rare — most only happen a few times each year worldwide — so the foreshortened view of history we get with available data on inequality doesn’t leave us with much statistical leverage.
Worse, the inequality data isn’t missing or flawed at random. From other work on political instability, we know that the kinds of countries that are least likely to produce reliable measures of inequality — the poorest or most repressive states — are often the ones most likely to suffer turmoil. Those kinds of countries also happen to be the ones least covered by the international press and the most opaque to foreign scholars. So our ability to observe both inequality and instability is often obscured in the most critical cases.
With such incomplete and blurry information about the crucial quantities, why are so many of us so sure that economic inequality is a principal cause of political turmoil? Careful observation is one answer. Aristotle and Marx drew inferences about the destabilizing effects of inequality from their deep knowledge of the societies around them. In the past several years, claims about inequality’s role in spurring popular unrest often point to protesters’ demands, which frequently reference inequity (e.g., “We are the 99 percent!”).
Those are important pieces of evidence, but they hardly close the case. Humans are notoriously unreliable at understanding and explaining their own behavior, so protesters’ statements about why they have taken to the streets don’t necessarily capture the true or full causes of their actions. Sometimes people participate in protests as a way to connect or improve their image with others or even just for the fun of it.2
Some statistical studies have confirmed the expected association between inequality and instability, but others have not — and all of them have depended on data that suffers from the flaws enumerated above. The seminal economic study that found a link between inequality and instability, Alberto Alesina and Roberto Perotti’s 1996 article, looked at a non-random sample of 71 countries for a single time slice (1970-86) using data on income distributions that did not deal carefully with the methodological issues described above.
Here is an alternative explanation for our collective confidence in this hypothesis: Humans are hard-wired to spot and feel uncomfortable about unfairness. Economic inequality is unfair; why should some have so much while others suffer with so little? Economic inequality is also ubiquitous. Its extent varies widely, but it is always present to some degree. So every time we see an uprising or rebellion or coup, we also see inequality, and we often hear the participants in that turmoil talking about inequality as a motivation for their actions. Because this is emotionally salient, we are more likely to see and remember it. Then, when we recall and compare instances of political unrest, a clear pattern emerges: Inequality and unrest seem inextricably linked. Maybe inequality isn’t a sufficient condition for political crisis, but it sure looks like a necessary and important one.
This alternative explanation may not be true, of course. But it does show that it’s possible to explain our collective confidence in the effects of inequality on political stability without a strong relationship actually existing. That so many careful observers of politics in so many different contexts have reached the same conclusion certainly seems to support the belief that inequality causes political crises. Still, we’ve been collectively wrong about plenty of other things, so it would be useful to subject this belief to more rigorous testing before accepting it as axiomatic. Unfortunately, we’re not there yet.