The Downside of Imperial Collapse
When Empires or Great Powers Fall, Chaos and War Rise
According to many commentators, especially those with left-wing inclinations, we are living through an era of political protest unprecedented in its radical militancy. The influential British journalist Paul Mason, for instance, argued in his 2012 book, Why It’s Kicking Off Everywhere, that the Arab Spring of 2011 was only one of many acts in a “global revolution” that is “here to stay.” Riots in particular are allegedly becoming more common. Alain Badiou, arguably the preeminent French philosopher working today, wrote in 2012 that we are living through “a time of riots,” a “global popular uprising” that marks nothing less than a “rebirth of History,” following the “end of history” first theorized by political scientist Francis Fukuyama in 1989.
This notion, far from being a product of the heady events of 2011–12, continues to hold sway today. In 2016, the journalist and English professor Joshua Clover published Riot. Strike. Riot., heralding a “new era of uprisings.” Clover claims, with surprising confidence, that “the riot has returned as the leading tactic in the repertoire of collective action” in the “overdeveloped countries.” High-profile examples from Berkeley, California to Hamburg, Germany appear to support the view that protestors in the Western democracies have never been more riotous than in recent years.
Although it may appear to some journalists and philosophers that global protest culture is radicalizing, this is mostly because they ignore empirical data suggesting otherwise. In what follows, I use one of the most comprehensive and up-to-date datasets on protest activity around the world, the 2017 version of the Cross-National Time-Series Data Archive by Arthur S. Banks and Kenneth A. Wilson, to paint a more systematic picture of global rioting. (For their measure of rioting, Banks and Wilson count “any violent demonstration or clash of more than 100 citizens involving the use of physical force.”) After accounting for population growth, the dynamics of peaceful demonstrations, and some illustrative country histories, I find a relative worldwide pacification of rioting over the past several decades. I also checked that my findings are robust with regard to to different ways of defining the system of states.
Figure 1 plots the total number of riots observed each year, around the world, from 1919 to 2016. Between 2011 and 2016, the average number of annual riots around the world surged to about 253. This is the real empirical basis for claims about the resurgence of global rioting, but Figure 1 gives a glimpse of how much our era has also been characterized by a peculiar dearth of riots. For most of the historical record, from 1919 to 1992, the world saw an average of roughly 53 riots per year. But for the 18 years before 2011, from 1993 to 2010, the world averaged only 35 riots per year with little year-to-year variance. After the collapse of the Soviet Union, it appeared as though the triumph of liberal democracy might have finally removed any reason for militant, anti-systemic forms of protest such as rioting.
At first glance, Figure 1 seems like overwhelming evidence of a global surge in rioting since 2011. But it partially reflects the far larger number of potential rioters in the world today. In 1964, on the cusp of the spike in rioting between 1965 and 1970, the global population was about 3.2 billion. By 2016, the world’s population had more than doubled to 7.3 billion. This poses a problem when analyzing the world’s riotousness over time because every new person slightly increases the probability of a riot occurring. The absolute number of riots can very well increase over time, even if attitudes and behaviors stay the same or become less riotous, or if a decreasing percentage of the population participates in riots. (This is not to mention the role of “youth bulges” in rioting, which would likely add further qualifications to contemporary assertions about the prevalence of rioting.)
To get a better sense of global rioting dynamics, then, we can divide the number of riots each year by the total population to see how “riotous” the average person is for each year. Figure 2 plots the total number of riots around the world divided by the global population in each year. Note that the data are scaled and centered at zero to reflect relative departures from the long-term historical average. This method suggests a very different story than the one presented in Figure 1. The spike in rioting since 2011 now appears more like a return to historically normal levels, whereas the low level of rioting from 1993 to 2010 looks even more abnormal and persistent, with strikingly little year-to-year variance. This perplexing, global depression of per-capita militant political behavior since the 1960s is the object of my own academic research on the global pacification of protest culture, which suggests that transformations in information technology have had profound effects on political attitudes and behaviors, in ways we are only just beginning to understand.
One further consideration is that if all types of protests increase, riots might rise in absolute terms even if they make up a decreasing share of overall protests. Indeed, that is what the data suggest: despite the apparent eruption of rioting since 2011, what has really happened is an explosion of peaceful protests at an unprecedented scale. Figure 3 compares total riots around the world to total number of peaceful demonstrations. In 2016, for instance, there were 917 peaceful demonstrations and 313 riots—both large quantities relative to historical averages. But 917 peaceful demonstrations is about 11 times greater than the historical average of 85 demonstrations per year, whereas 313 number of riots is only about 4.6 times the historical average of 67 riots per year. There has been a particular relative dearth of riots since 1968. From 1919 to 1968, there were 21 riots for every ten demonstrations; from 1969 to 2016, however, there were only six riots for every ten demonstrations. Within that latter range, the year of greatest rioting, 2014, still only saw about eight riots for every ten demonstrations (393 to 502, respectively). The real puzzle today is why so many people take to the streets without rioting.
WORLD ON FIRE?
Global averages could be misleading if there are substantial differences at the region- or country-level. Figure 4 plots riots as a percentage of total protest events (riots plus peaceful demonstrations), this time broken out by region. (When zero riots and demonstrations occur in a country, that country in that year is removed from the analysis because one cannot divide by zero.)
In every region, rioting is less prevalent today, as a share of protest events, than in the period before 1968. Latin America and the Caribbean has seen the greatest pacification of rioting, while sub-Saharan Africa’s decline has been the most modest. Prior to 1968, riots accounted for about 80 percent of the protest events in Latin America and the Caribbean, but have been only 33 percent of protests since then. In sub-Saharan Africa, those numbers are 74 percent and 52 percent, respectively. What is unique about the present moment is the historically extreme peacefulness of protests. The overall increase in protests appears to have been so great that, although riots have increased notably, they have never been a less common way for people to express their political frustrations.
Finally, Figure 5 considers the five countries that made the greatest contribution to global rioting: China, Egypt, India, Turkey, and the United States, which together accounted for one-quarter of the world’s riots between 2011 and 2016. I also include France because it is mentioned by Mason, Badiou, and Clover—rioting in the banlieues being a favorite piece of evidence for those who claim we are living through a period of unprecedented militancy. Of the six cases, Turkey is the only example of an extraordinary level of per-capita rioting, relative both to peaceful protests and its own history. In China, India, and the United States, rioting in 2011–16 surged relative to the preceding few decades, but remained far below each country’s historical highwater mark. In each case, peaceful demonstrations substantially exceeded riots. France falls somewhat within this group, although its recent riots are comparable to its highest peaks in the 1940s and 1968.
The quantity of riots around the world has certainly spiked since 2011, but the raw number of riots has been widely misunderstood. Why, then, have so few people chosen to riot over the past several decades?
As with most complex political phenomena, there are likely multiple causes. Many states have become much stronger, especially in terms of surveillance, which would increase the risks of rioting while making peaceful, legal demonstrations more attractive. Many countries have democratized over the past several decades, and many developing countries now have rapidly rising middle classes. Both of these dynamics likely push individuals toward legal, cooperative forms of political activity, and away from militant, anti-systemic types of activity. A final, more speculative possibility—and the subject of my own research—is that modern information technology, especially in communications and mass media, tends to pacify certain psychological and behavioral circuits that historically led human beings to principled rebellion against systemic injustices.
Scholars, of course, will continue to debate the causes and consequences of recent trends in global rioting. If there is one conclusion we can draw at present, it is that the appearance of increased rioting almost certainly hides more than it reveals.