Crowd behaviour: united they stand

Abstract: Is there a link between the Arab Spring uprisings and the behaviour of market traders?…

From Cairo to Tunis, demonstrators united by emotion, purpose and social media gather in their thousands to topple long-standing despotic regimes. Across the UK, crowds of a different temperament surge through summer streets, leaving behind them a trail of destruction. Meanwhile, in the world’s financial centres, bond traders linked by the internet behave like a “virtual” crowd as they sell the securities of increasingly embattled eurozone members, leading yields to soar.

The crowd was at the heart of some of the most memorable events of 2011, demonstrating the power of the group driven by common identity and capacity for decision-making. They are classic examples of the herd mentality – the shared and self-regulated thinking of individuals in a group – an area of study popular with sociologists and psychologists.

And while a social, economic and cultural chasm separates wealthy financial traders fixated by screens in corporate offices from unemployed youths hurling bricks, recent research into herd mentality suggests the two groups are, in fact, closer than is commonly believed. Moreover, their behaviour can be examined, plotted and, to a certain degree, predicted.

Research now shows that similar behaviour can occur in both real and virtual crowds where they share a sense of collectivity, driven by common goals and interests. Whereas individuals in a physical crowd may take their cue from the visible behaviour of others, fund managers and online gamers take theirs from changes on a screen.

“Some studies suggest that it’s enough just to imagine a situation – you don’t actually have to be there,” says Michelle Baddeley, a behavioural economist at Cambridge university, “and people’s emotions can easily be engaged in virtual emotions, for example in computer games and on-line gambling.”

Emotions – innate and instinctive responses to stimuli – are, according to Ms Baddeley, an integral component in financial decision-making, and help explain the seemingly irrational herd mentality displayed during good times and bad.

While psychologists have long talked about concepts such as the “wisdom of the crowd”, many of today’s economists remain reluctant to embrace the concept of group behaviour. The discipline has steered away from the sociological and psychological factors in decision-making and focused instead “on narrow behavioural assumptions in which expectations are formed on the basis of mathematical algorithms”, Ms Baddeley says. To them, crowd behaviour is just too messy, irrational and unpredictable.

John Maynard Keynes, by contrast, believed professional investors would “follow the herd”, at least partly, for the sake of their reputation. “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally,” he wrote in 1935, in the midst of the Great Depression.

The idea of a neurological component in financial decision-making is of central interest to those digging around in the relatively new field of neuroeconomics, where neurology, psychology and economics overlap.

Vasily Klucharev, a neuroeconomics researcher at the university of Basel, is particularly interested in the part of the brain that triggers a “learning signal” when our behaviour differs from that of the crowd we are in, and which makes us adjust to match the group. He says this is an automatic process in which people form their own opinion, face the group view then quickly shift their attitude to make it more compliant.

But the group view is not always healthy or in the best interest of the individuals in the crowd. Automatic, uncritical adjustment can backfire, whether in the case of overexuberance in a bull market that precedes a crash or in mass looting of shops that destroys a rioter’s own neighbourhood.

Mr Klucharev points out that the processes that occur rapidly in crowds are akin to changes in attitude that occur slowly in more diffuse groups. “Think of geographic areas in the United States where obese people are a very common sight,” he says. “In that case, the conflict area in a person’s brain no longer sees being overweight as negative, and the eating behaviour is sustained rather than corrected. Ironically, in these circumstances, it is the more healthily eating people who get the basic brain message: I am doing something wrong.”

Research in social psychology is shedding more light on the processes that drive humans to separate into groups. According to the “minimal group paradigm”, people will do this based on almost any characteristic, and will even show loyalty to groups formed at random.

Seeing other members of the group in a real crowd enhances the effects of herding, but it is not indispensable, Mr Klucharev says. Take the role of social media in the Arab uprisings and the UK riots. The point here is not whether Facebook, Twitter or BlackBerry Messenger were organising tools for mass action but that they helped to create a common identity.

Recent events in Russia are another case in point. Despite state media’s depiction of widespread support for Vladimir Putin, the prime minister, social networks unified those objecting to the results of recent elections. As a result, tens of thousands found a common voice, pouring on to the streets of Moscow to challenge Mr Putin.

Neuroscientists such as Mr Klucharev are studying the neural origins “informational cascade” – the process by which individuals change their minds when they learn about the decisions of others in the group. “Such changes occur in individuals independently of their own private information signals,” he says. “You can see it in the ‘herding behaviour of traders, especially in times of uncertainty.”

The modelling of crowd behaviour – an important component in urban planning and the design of large buildings – depends on the notion that groups of individuals operate as a co-ordinated unit. To some extent, they can act like a living, thinking form and thus their movements can be predicted and plotted.

The resultant mapping of the interaction between crowds and geographical spaces is known as geosimulation. Computer models are constructed based on the movement of groups under different geographical and behavioural scenarios, including emergency evacuations and riots.

But geosimulation is undergoing something of a change, as research reveals more about the intricacies of herd behaviour. Previously, modelling employed a physics-based approach, assuming individuals respond individually but predictably to forces of attraction and repulsion in their physical environment and others in the crowd – a bit like molecules in a stream of gas.

This is now seen as simplistic because individuals can spontaneously behave in a more socially coherent manner. Carol O’Sullivan, professor of visual computing at Trinity College, Dublin, explains that for computer modelling to be accurate, it has to “give a sense of intelligence to the crowd”. Individuals in geosimulation models are therefore programmed to react to different alternative scenarios as well as to other individuals, especially in situations of panic or high emotion.

Further research is needed to examine what percentage of a crowd can change the emotional character of the whole under varying circumstances. What does it take for a powerful individual or small contingent to sway opinion? “It would also be interesting”, she adds, “to try and model the influence of agent provocateurs in crowd behaviour.”

Nonetheless, the concept of a crowd psychology is gaining ground. According to Stephen Reicher, professor of social psychology at the University of St Andrews: “Individuals don’t lose identity in the crowd and they don’t lose control over their behaviour or rationality. Rather they shift to a shared social identity and seek to act in terms of that shared identity.”

But the process is not mechanical or thoughtless, he insists. “Shared identity helps the creation of consensus through debate. For that reason, I deeply dislike notions such as ‘herd mentality’, which derives from the idea of an inferior animal mentality in crowds. Group or crowd members are as thoughtful as isolated individuals. The difference is that the processes of thought and discussion are scaffolded by social as opposed to individual identity.”

Although psychologists recognise that links between, say, political demonstrations and financial trading are a fruitful field for research, they warn against pushing the similarities too far. “Financial and political environments will be affected in similar ways,” says Ms Baddeley. “However it is important to recognise that for political protests the goals are quite different – in a financial environment, the goals are at least a lot cleaner and starker.”

“The riots and the Arab spring are about inter-group behaviour,” says Prof Reicher. “People cohere and work together because they are acting as one group against another. Traders, on the other hand, don’t have an obvious ‘out group’ but are rather engaged in intra-group behaviour. They compete against their fellow traders and are set against each other as individuals.

“Problems may arise precisely because of this lack of ‘groupness’, and the competition, mistrust and negative inferences about others that this entails,” he continues. “Perhaps, then, the answer is to increase a sense of common group membership.”

It seems shared emotions can give a seething crowd a common purpose – destructive or constructive, according to the observer’s perspective – by activating neural areas usually associated with fear and greed. If the same brain regions are activated in financial herds, is the image of the cool, calculating, rational trader outdated?

Ms Baddeley put the question to people in the financial services sector in a recent study. They agreed that, for high-frequency trading in very volatile markets, it would be hard to imagine traders remaining cool and calculating – after all, they are only human – and felt it was not necessarily bad to be guided by emotions.

Neuroscientists are still far from understanding – let alone controlling – all the factors that lead crowds to behave as they do. But with more volatility forecast for 2012, in both the streets and in the financial markets, they are likely to gain plenty of fresh material for their studies.

FINANCE AND FEELINGS: The neural ‘pleasure centre’ at the heart of herd behaviour’

Which part of the brain is powerful enough to overwhelm individual thought and generate the desire to be part of a herd? Neuroscientists believe a neural network that includes the nucleus accumbens – more popularly known as the “pleasure centre” – is responsible. It is part of the limbic system associated with generating feelings and emotions such as fear, reward, punishment and pleasure.

Exposing subjects to situations that forced them to change their individual decisions to fall into line with those of a group, Basel university’s Vasily Klucharev and colleagues saw neural activity in the posterior medial frontal cortex – which monitors behaviour – and the nucleus accumbens using functional magnetic resonance imaging.

But when they temporarily shut down the cortices of the volunteers with a non-invasive procedure known as transcranial magnetic stimulation, the subjects ceased to adjust their behaviour to comply with that of the group. In other words, deactivating a specific part of the brain made subjects temporarily immune to social influence, and thus incapable of joining the herd mentality.

The nucleus accumbens is also important for understanding the primary drivers of financial traders’ decision-making: risk and reward. “Whereas a neurological basis for risk is difficult because the concept of ‘risk’ is hard to define – it is different for economists and psychologists – the experience of ‘reward’ can be traced to the nucleus accumbens,” says Gregory Berns, professor of neuroeconomics at Emory University in Atlanta.

His studies have underscored the pleasure centre’s role in mediating the release of dopamine, a powerful neurotransmitter (chemical messengers in the brain). Dopamine is also linked to addiction; cocaine and amphetamines stimulate the release of dopamine, which produces the characteristic “high”.

According to Prof Berns, actually receiving the reward is not necessary. “Just the anticipation of a reward is sufficient to generate activity in this part of the brain,” he says. “It’s the expectation of what’s going to happen in the future that helps guide financial decision-making.”

The system sheds light on the herd mentality, he says: “It helps explain our need for a sense of belonging. Just as teenagers may rate music according to a star rating others have accorded it, our judgments are influenced by the desire to belong to a crowd, and the expectations of the reward of the acknowledgement of that crowd.”

You do not even have be in the physical presence of others to receive the feedback that encourages a change of opinion. “Our own processing stimuli can be turned off and our decisions can be swayed, simply by the suggestion that others thought differently,” says Prof Berns, whose studies show this.

(With Clive Cookson, Science Editor Financial Times) Originally published in The Financial Times, 27 December 2011