Riots And Revolutions In The Digital Age – Analysis


By John Fender and Chris Ellis

Major world events, such as the Arab Spring of 2011, the fall of communism just over 20 years ago in Eastern Europe and the Soviet Union, and the financial crisis of 2007-09, sometimes come as a complete surprise to almost everyone. It is of course important to try to understand such events, but it is sometimes not easy to know how even to begin to analyse them. However, some recent work by economists offers considerable promise in explaining such events.

Explaining regime change

In recent years, economists have put considerable effort into explaining regime changes and democratisation. A common explanation put forward by Daron Acemoglu and James Robinson in a number of articles and their book (Acemoglu and Robinson 2006) is that dictatorial regimes democratise because of a threat of revolution by workers.

According to their analysis, the ruling elite may wish to redistribute income to workers to avert a revolution. But a mere promise to redistribute may not be credible since the regime could renege on its promise when the revolutionary threat has subsided. Democratisation is a way for rulers to make the promise to redistribute credible.

Social media
Social media

This theory arguably explains a great many democratisations, but there are some loose ends. It is not clear why revolutions actually occur in reality, since the theory predicts that if workers have a credible threat of revolution, democratisation will take place without the need for a revolution. 1

Explaining revolutions: information cascades

Another question is how do workers coordinate their actions? In a recent paper (Ellis and Fender 2011) we use the idea of information cascades to develop a theory of political regime change brought about by the occurrence or threat of revolution.

An information cascade, speaking loosely, is where people make decisions on the basis of their observations of other peoples’ actions. 2 According to the analysis in our study, workers decide whether or not to rebel by observing other workers’ behaviour, as well as by observing any ‘signals’ that they may receive about the state of the regime. So if some people rebel, others may follow, thinking that their rebellion may be a sign of the regime’s weakness. If enough of them rebel, there is a successful revolution and the rulers are overthrown.

Rulers anticipate the possibility that workers may rebel, and may react in various ways. They may seek to redistribute income or perhaps democratise. Of particular relevance is the ‘quality of information’ – information flows are crucial to the occurrence of revolutions. This is consistent with the behaviour of many autocratic regimes, which seek to prevent the gathering and dissemination of information.

The analysis shows why dictatorships may not want to eliminate the risk of revolution entirely since that may involve considerable redistribution. Instead, the elite may prefer to enjoy their perquisites of living in a highly unequal society, calculating that it is worth running what they consider to be a very small risk of revolution. However, such societies may be particularly susceptible to revolution.

The model can explain why revolutions are often a considerable surprise to virtually everyone, participants and spectators alike – something that is clearly applicable to the currently unfolding events of the ‘Arab Spring’, where there are information cascades both within and between countries. We would argue that developments in information technologies, such as Twitter and Facebook, are likely to make such information cascades even more powerful. Another result is that rebellions need not always succeed, and sometimes revolutions can be a mistake in the sense that everyone is worse off after they occur.

The theory is illustrated with discussion of a number of historical episodes, including the French and Russian revolutions, the East European revolutions of 1989, the Tiananmen Square massacre in China in 1989 and the events in Burma (Myanmar) in 2007. Our paper was completed before the events of the Arab Spring, but is clearly relevant to explaining them.

Explaining rioting

An information cascade approach may also help to explain the recent riots in the United Kingdom – information transmission enabled rioters to coordinate their activities and facilitated the spread of the rioting to other cities. But we need to supplement the information cascades approach with some theory of why rioting may take place. Here, another paper by one of us develops a multiple equilibrium theory of crime that may well be relevant (see Fender 1999).

The idea behind the multiple equilibrium story is as follows. Suppose there are some potential criminals who have no conscience about committing crime, but decide whether to commit crime rationally by balancing the expected benefits of committing crime against the costs, which depend on the probability of being punished as well as the actual punishment itself.

The probability of being punished depends on a number of factors, most notably the resources devoted to spending on preventing and punishing crime (police force, prisons, courts, etc) but also the amount of crime taking place. If crime is high, the resources the authorities may be able to devote to preventing, investigating, and punishing individual acts of crime may be quite low, so the probability of punishment may also be low.

We can now see how there can be multiple equilibria – there can be a low-crime equilibrium, where agents believe that if they do commit crime, they will probably be punished, which means crime will be low. On the other hand, if potential criminals believe the chances of being caught are low, they will have an incentive to commit crime, and the fact that a high level of crime is taking place means that the probability of being punished is low, hence justifying the decision to commit crime.

The hypothesis is that there was a jump from a low-crime to a high-crime equilibrium for a brief period in August and then a reversion back to the low-crime equilibrium. How and why did this jump take place? One account would be that rioting started in London over what was perceived to be a legitimate grievance; the police response was fairly muted and rioters found they were getting away with it. This was communicated to other potential rioters who joined the riots; as the number of participants grew, the police’s resources became more stretched and the expectation that rioting and looting would be unlikely to be punished grew, and the whole process escalated.

This was reversed when the police became sufficiently organised as to raise the probability of punishment enough to induce a shift to the low-crime regime. So, for a short period of time potential criminals believed that they could riot and loot and get away with it.

Our explanation of the recent riots is that there was a temporary jump from a low-crime equilibrium to a high-crime equilibrium. There may have been no sudden increase in the underlying propensity to commit crime at all – ie no ‘moral deficit’ had suddenly emerged. The propensity to commit crime for any given probability of detection and punishment may not have increased – rather, it was a temporary fall in this probability that caused the increase.

Another implication from our explanation is that measures to prevent a repeat of the rioting may be very different from those that might be envisaged to reduce the ‘underlying’ level of crime in the low-crime equilibrium, which may still be quite high. Such measures might include the police reacting strongly to any temporary increase in crime and monitoring and possibly disrupting communication between potential rioters.


Information cascades may help to explain a huge range of economic, social, and political behaviour. We have not mentioned the role of such cascades in the financial crisis, but undoubtedly such cascades were important for many of the events of the crisis, and more generally may well be important for the stock market. Economists (and others) need to be aware of their importance.

John Fender
Professor of Macroeconomics, University of Birmingham

Chris Ellis
Professor of Economics, University of Oregon


Acemoglu, Daron and James Robinson (2006), Economic Origins of Dictatorship and Democracy, Cambridge University Press.

Bikhchandari, Sushil, David Hirshleifer, and Ivo Welch (1992), “A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades”, Journal of Political Economy, 100(5):992-1026.

Ellis, Christopher J and John Fender (2011), “Information Cascades and Revolutionary Regime Transitions”, The Economic Journal, 121(553):763-792.

Fender, John (1999), “A General Equilibrium Model of Crime and Punishment”, Journal of Economic Behavior & Organization, 39(4):437-453.

Lizzeri, Allesandro and Nicola Persico (2004), ‘Why did the Elites Extend the Suffrage? Democracy and the Scope of Government with an Application to Britain’s “Age of Reform”’, Quarterly Journal of Economics, 119(2):705-763.

1 There are other explanations for democratisation, based on division within the elite, which seeks to explain regime change in other ways (eg Lizzeri and Persico 2004). These explanations should be seen as complementary to those based on a threat of revolution – regime change is undoubtedly a highly complex phenomena and many different theories may have some relevance.

2 Bikhchandari et al (1992) is one paper that develops the theory of information cascades and applies it to a number of phenomena.

To view the original article, click here. was a policy portal set up by the Centre for Economic Policy Research ( in conjunction with a consortium of national sites. Vox aims to promote research-based policy analysis and commentary by leading scholars. New content can be found at

Leave a Reply

Your email address will not be published. Required fields are marked *