Hard problems in economics

This is a summary of a presentation I gave in Zurich earlier in June to FuturICT, one of the candidate flagship European Union research projects, each worth 1 billion Euros.

1. Financial markets is a very hard problem, issues of agent heterogeneity, networks, learning, financial innovation, regulation – all these and more are important. Mainstream economics has largely avoided the topic, content to believe in the efficient markets hypothesis. A great deal of distinguished work has already been done by econophysicists.

2. Inequality: where does it come from?

This is not just a matter of the distributions of income and wealth. A major concern of policy makers on say, outcomes in health care or education across hospitals and schools, is that such outcomes are ‘unequal’ in the key sense that they differ at any point in time. This seems inevitable in any complex system of interacting agents, but it is often a major concern to voters and hence to politicians.

3. Shortages: food/water/energy – how do we avoid/mitigate these key social, economic and security-related problems

Both these very hard problems require trans-disciplinary teams. For example, Elinor Olstrom showed how societies evolve ways of dealing with the ‘problem of the commons’ to give outcomes which are quite different to those predicted by economic theory.

4. What does it mean for an agent to be rational in the world of the 21^st century?

We face a stupendous number of choices, many of which are complex and difficult to evaluate and distinguish between them, and we are increasingly both aware of and influence directly by the behaviour and decisions of others across networks. A sub-theme of this is: how do agents cope with the massive explosion in information and turn it into useful knowledge.

The old concept of rationality is no longer in general valid.

This has very deep and widespread policy implications. Almost all existing policy is based on a view of the world in which autonomous agents maximise subject to constraints. In a world of interconnected agents where maximisation does not make much sense, what becomes the basis of policy? How will agents react? What are better ways to get them to react in ways which policy makers want?

This involves matters such as, individual learning, social learning, the evolution of self-image, networks (the relevant topology; how they evolve; who/where/when agents might copy)

So a very hard task is: what is the new paradigm for agent behaviour, one which in principle is general across the social sciences.