Understanding how to change behaviour is a major problem for policy makers at the moment. Just think of the problems which could be solved if we knew how to do this effectively. For example, climate change. To reduce emissions, behaviour needs to change.

In financial markets, bankers have always wanted to make money ever since banks were first invented. But there has been a profound change in culture over the past two decades or so, which led directly to the financial crisis. It became acceptable for bankers and traders to behave exactly as pure Economic Man, without regard to the consequences of their actions on other people.

More generally in the corporate world, senior executives now feel able to pay themselves hundreds of times the wages received by ordinary workers. They deserve to be rewarded, but behavioural change has led to gross excesses. In the UK in the mid-1980s when the domestic gas industry was privatised, the Chief Executive was vilified for paying himself what was at the time considered to be a stupendous amount. How much was it? Half a million Euros a year. Most CEOs of major companies would now simply laugh if that was all they were offered.

Think of social problems such as obesity, drug taking and alcoholism. In each of these areas, behaviour has changed. What was previously regarded as unacceptable behaviour has now become permitted in the relevant communities.

Standard economics has an answer. It is not wrong. But it is at best a very partial solution to these major policy challenges. To change behaviour, you change incentives. Most public policy is based on this view, that people operate purely as individuals, and systematically compare the costs and benefits of different alternatives. On this theory, they then make the best choice for themselves. So policy becomes based on changing the incentives that people face, altering the costs and benefits of different choices.

Global systems science has the potential to revolutionise the range of tools available to policy makers. Network science has expanded dramatically over the last decade or so. We now know much more about how ideas and behaviours either spread or are contained across networks.

Peer effects, the copying or imitation of the behaviour of others, are now a dominant feature of behaviour in the interconnected world of the 21st century. Incentives still matter, but in many contexts they are of second order of importance compared to the influence of peers on the relevant social network.

Global systems science needs to make operational the insights of network theory. We can provide the practical tools for policy makers which will transform their effectiveness by understanding much better how to change behaviour. We need to:
• Develop heuristics to identify in any given situation the relative strength of network effects versus individual selection on behaviour
• Develop ICT tools to obtain good approximations to the underlying topology whenever network effects are a strong influence on behaviour
• Develop strategies on how to alter behaviour in these contexts

Paul Ormerod www.paulormerod.com