Sunday, 7 August 2011

Are crowds to blame for the plunging financial markets?

Economics seems to be the last remaining social science discipline to retain a place for ‘mass panic’ in its explanatory models. Thus according to economics commentators, the sharp drop in share values at the end of last week is due to ‘panic’ among investors in the financial markets which in turn is attributed to the supposedly collective nature of markets.

For example, on Friday’s ‘The World Tonight’ on Radio 4,‘City Boy’ author Geraint Anderson explained that the trading floor is an emotionally volatile environment which leads to irrational and damaging behaviour because of the presence of other people. While individuals may be rational, ‘crowds are foolish’.

He is not alone in making such claims. Standard sources describe all the major financial crashes of the last 100 years in terms of ‘panic’. The classic case study of a stock market bubble and crash, ‘Tulipolamania’, was also explained in similar terms by Charles Mackay in his book Extraordinary popular delusions and the madness of crowds (1841):

‘The more prudent began to see that this folly [investing in tulips] could not last for ever. Rich people no longer bought the flowers to keep them in their gardens, but to sell them again at cent per cent profit. It was seen that somebody must lose fearfully in the end. As this conviction spread, prices fell, and never rose again. Confidence was destroyed, and a universal panic seized upon the dealers. … The cry of distress resounded every where, and each man accused his neighbour… Many who, for a brief season, had emerged from the humbler walks of life, were cast back into their original obscurity. Substantial merchants were reduced almost to beggary…’
(p. 95)

To those of us working in other social sciences, the reliance by economists on the concept of panic is both interesting and perplexing. In the study of emergencies and disasters, panic has been largely discredited as an explanation for mass behaviour. Most commentators regard panic in emergencies and disasters as either rare or a myth. Where people flee from harm, this is often the most reasonable course of action rather than an over-reaction. Coordination and helping is common in emergencies, and people even take personal risks to assist others. Where survivors fail to help others, or even harm them during their own escape, it is usually because they are physically unable to do otherwise rather than because they are overwhelmed by selfish motives. In short, outside of economics, there is little evidence for a supposed tendency to rash behaviour in the face of danger brought on by the fact of crowd membership.

The rational individual and the intrusion of mass psychology

The economic subject is in essence a cognitive individual, not an emotional crowd member. S/he is homo economicus, the rational maximizer of self-interest. Mass psychology impinges upon this rational individual as an external and unwelcome intrusion, foisting upon her not just panic but a range of other emotions, including ‘confidence’, ‘optimism’, ‘nervousness’ and ‘fear’:

‘The key to such widespread phenomena [i.e., panic selling] lies in the nature of the crowd: the way in which a collection of usually calm, rational individuals can be overwhelmed by such emotion [mania, then fear] when it appears their peers are behaving in a certain universal manner.’

While ‘confidence’ may be a good thing, ‘over-confidence’ is not, and can lead to irrational behaviour (such as buying too many tulips). Indeed, if too many individuals display either ‘over-confidence’ (or ‘excessive caution’), this can lead to collective disaster.

Economic commentaries offer various accounts of how social influence occurs and hence emotional and irrational behaviour spreads between people. These include ‘herding’ and ‘mimicry’. The most common concept for explaining social influence within and between financial markets, however, is ‘contagion’. The first person to apply this medical concept, which describes the spread of a disease, to social behaviour was the historian and crowd psychologist Hippolyte Taine. Like ‘herding’ and ‘mimicry’, the concept of contagion conveys mindlessness, and hence uncritical and simple effects of social co-presence. And like ‘mass panic’, ‘contagion’ and these other concepts are today more usually found in the biological sciences than the social sciences.

 A frantic scene on a trading floor

The subjects and social relations of economic production and circulation can be characterized, and indeed reified, in various different ways to achieve different effects. For example, the expression ‘money makes the world go round’ places a human product in the position of producer. After endowing the products of labour with their own agency and power, a second step is to naturalize and eternalize these qualities and the relationships they imply. When ‘the economy’ seems to exist as a separate entity in its own right, then those attempting to ‘steer’ it cannot be blamed so easily when there is a recession; and when ‘market forces’ are inevitable, then social change becomes unthinkable. 

In the present case, over-emotional and rash behaviour brought on by supposed ‘crowd’ membership is reified as a separate ‘social realm’, external to individual ‘human nature’. This psychology of the crowd, rather than other features of the system, can then be blamed for the volatility of the financial markets. Explaining financial market crashes in terms of this mass psychology can thereby draw attention away from other possible causes of crisis. Thus, it could be the lack of proper sociality, rather than its excess, that is at the heart of what goes on in a financial crash.

What is a crowd?

Another way of looking at the problem of market crashes is to say that the financial market is not really a crowd at all. Market crashes are caused by people acting as individuals not as crowd members. In good times as much as bad, calm times as well as stressful times, individuals attempt to maximize their gains or limit their losses. But, because the individuals within each financial market do not communicate, coordinate or plan together, of course they are always trying to second-guess the intentions of other individuals. The imperfect attempt to second-guess, again, takes place in good times as well as bad times. Psychologically, therefore, there is no qualitative difference in behaviour between normal trading conditions and crisis: both are individual rather than collective.

The idea that the financial crash is explicable in terms of exactly the same individual imperfect second-guessing as in normal conditions, rather than being due to collective panic, is not new. Keynes explained market behaviour in roughly these terms many years ago. What we as social psychologists can add conceptually, however, is an elaboration of the idea of the crowd that economists use so loosely.

Hence we distinguish between a physical crowd and a psychological crowd. A physical crowd is simply an aggregate of people who are all present in the same physical space at the same time. A psychological crowds is a co-present set of people who see each other as belonging to the same social category: for example a crowd of protestors against racism or a crowd of Manchester United football supporters. While both may be referred to loosely as ‘crowds’, the behaviour of their ‘members’ towards each other will be very different. Moreover, in situations of possible danger or scarcity, the collective outcomes of this behaviour will also be very different.

In a physical crowd, people act for themselves as individuals. When there is a threat (of danger or missing out), those in a physical crowd will experience that threat in relartion to themselves personally only, and will act in competition with others in the crowd to gain the scarce resource or to escape. The net effect may be that there is less collectively beneficial behaviour across the ‘crowd’ as a whole: the exit is blocked, the market collapses, and so on. In the psychological crowd, by contrast, 'my interest' and ‘your interest’ is superseded by ‘ours’. Other people in the crowd are therefore assisted, and members of the crowd communicate in order to coordinate the most effective response in the interests of the group as a whole.

Evacuation behaviour in two virtual reality ‘crowds’

As well as offering conceptual clarification, we can also add an empirical illustration of some of these arguments. We carried out an experiment to look at effects of collective threat under different psychological conditions. The experiment employed a virtual reality visualization of a scene in an underground railway station. In the visualization, people in the scene run towards the exit. There are bottlenecks on the escalator, and there are a number of injured fellow passengers who need assistance. The participant has to decide how to respond. 

We ran the simulation at the immersion laboratory at St Andrews University, and introduced a number of conditions. In one condition, participants were told that their aim was to get to the sales as quickly as possible to get a bargain. This was the ‘physical crowd’ condition. In a second condition, participants were told that their aim was to escape from the station as quickly as possible because there was a fire that threatened the whole crowd. This was the ‘psychological crowd’ condition. In each condition, the visuals (number of other people running, number of bottlenecks and people needing assistance) were identical.

To check the manipulation, we took measures of participants’ social identification with others in the crowd. As planned and expected, there was greater social identification for those in the psychological crowd than the physical crowd. Importantly, there were also differences between these conditions on all three behavioural measures taken. Compared to participants evacuating the station in the ‘physical crowd’ condition, those in the psychological crowd were more likely to help those who needed assistance, quicker to offer such help, and less likely to push others out of the way. In other words, they acted in ways which were of benefit to other crowd members as crowd members and hence in the interest of the crowd as a whole.

What does this mean for real-world evacuating crowds? With a limited exit and a large crowd, the optimum response for the crowd as whole is collective coordination, such as queuing. By contrast, if people act simply as individuals rather than as fellow crowd members in such contexts, their individual ‘rational self interest’ translates into competition among individuals, such as pushing and shoving, which could block those same exits and hence endanger the crowd as a whole.

What does this mean for the psychology of the financial markets? It means that ‘crowds’ per se are not to blame for plunging financial markets. This in turn means that ‘panic’, conceptualized as the intrusion of irrational mass psychology into the smooth workings of the otherwise rational individual, as an explanation for financial crises obscures more than it reveals. If there is dysfunction in the markets – whole economies plunging into recession and even depression – then it is not because people are acting as crowd members. Quite the reverse: it is precisely the fact that people are acting simply as individuals attempting to maximize their own individual interests, and in a setting where others are involved and their actions have collective outcomes, that is the essence of the dysfunction.



Mackay, C. (1841). Extraordinary popular delusions and the madness of crowds. Ware: Wordsworth.