Quick Report on the first NESS meeting on the shores of Geneva Lake

Attendees:

Marina Bianchi, Cassino university, Italy
Mark Buchanan
Ping Chen, Peking university, China
Mauro Gallegati, University of Ancona
Jeff Johnson, Open University, UK
Alan Kirman, Universite de Marseille
Imre Kondor, Budapest university
Andrzej Nowak, University of Warsaw and University of Florida
Paul Ormerod, Volterra Consulting and Durham University
Bridget Rosewell, Volterra Consulting and Greater London Authority
J Barkley Rosser, George Madison University
Gabriele Tedeschi, University of Ancona
You-Gui Wang, Beijing normal university
Yi-Cheng Zhang, University of Fribourg

The meeting was convened to discuss how to most effectively take forward social science which was able to go beyond the standard models, particularly in economics, whose assumptions are restrictive and fail to capture important aspects of reality.

Those attending all shared an interest in complex systems approaches of various kind and most had either physics or economics backgrounds, though one is a psychologist and a number had roles in policy in various forms.  Most participants were European, but there were attendees from both Asia and the US.  A number of additional invitees supported the idea of NESS but were unable to attend the meeting.

The first session was spent discussing the issues that participants felt were particularly relevant to taking forward the development of social science in general and economics in particular, with short informal presentations from Paul Ormerod, Alan Kirman and Jeff Johnson.  The following morning developed these themes and discussed ways of taking them forward, with a further presentation from Yi-Cheng Zhang.

The Issues Raised

The first two presentations concentrated on the nature and role of the agent in the social sciences.  Paul Ormerod concentrated on alternative definitions of rationality to that of the standard paradigm, in particular in situations where tastes and preferences could not reasonably be supposed to be fixed. Alan Kirman focused more on a critique of economic rationality, and described how it was insufficient to conclude that agents could be treated as if they were maximisers and or that the statistical description of market prices was sufficient to establish the efficient market hypothesis.

The discussion focussed on how agents might develop and operate rules of behaviour.  Two further aspects emerged from this of how a model of the self might be integrated into social and economic decision making and the development of agent based modelling techniques to capture interactions between agents.  Social interactions were considered to be key by many for whom the assumption of fixed tastes and preferences is the special and not the general case.

A number of attendees raised the relationship between the standard models and new approaches and the importance of the current time for developing new approaches in the light of the visible failure of the existing models to predict the financial crisis.  There was a general acceptance that too much concentration on the failures of the existing models was a diversion from the intellectual effort to capture the actually existing system.

The presentation by Jeff Johnson focussed more on the construction of a system of analysis rather than the agents within it, and particularly on the question of how to address the levels at which any system should be considered – whether macro, meso or micro.  Discussion developed to consider the extent to which there were discrete or continuous differences, or a hierarchy of levels of social organisation and analysis.  The need to look at the empirical data and its patterns was stressed and the empirical pattern of the dynamics in particular.

At the end of the first session, the following themes had emerged.

• The nature of the agent in the social sciences
• The rules of interaction between agents, especially their networks and the existence and nature of feedback loops
• The use of information and its capability
• The evolutionary characteristics and vocabulary of hte social and economic system
• The hierarchy of systems and issues of aggregation
• Characterising institutions and assessing which simplifications are correct
• Descriptions of dynamics and patterns of data
• Identifying the relevant questions on concrete matters

The second session started with a presentation by Yi-Cheng Zhang on the role of information capability as a new element in economic decision.  This stressed the ability of information to create further information distinction and inherent instability in markets.  Discussion centred both on the relationship of this concept to standard concepts of imperfect information and the connection to the problem of value mediated through other’s choices (the prize based on guessing what others will think).

All of these problems also had implications for the timescales on which social and market forces might operate and the ability of analysis to capture this.

Taking the Issues Forward

When turning to how all these themes should be taken forward, two approaches emerged.  One was based on creating practical and concrete subjects of analysis, such as the theme of consumption, innovation, or financial markets.  Another was the theme of policy advice and development.  It was pointed out that central banks and government are increasingly interested in new approaches to policy making.  However, it became clear that attitudes to policy differed amongst participants and that there was no agreement about what a ‘better’ situation might look like.  All agreed that it was important that the science came first.

It was also identified that a wider range of disciplines needed to be represented including anthropology, sociology and political science to create a transdisciplinary approach.

All participants agrees that the title of the group – non-equilibrium social science – was important and captured the main purpose of the group, namely to analyse open rather than closed systems.  Non-equilibrium is a key characteristic of many situations in the social sciences.

The scientific themes are:
1    The nature of the agent and rules governing interaction in establishing social and market behaviour
2    The identification of the social system and organisations and the levels of analysis appropriate to individual problems
3    The potential for prediction and the ability to understand the consequences of any policy interventions

Finally, the meeting discussed how to take these themes forward.

It was agreed to investigate the next call from the EU for a Co-ordination Framework which could be used to take forward further workshops and more specialised studies of a particular topic, such as consumption.  It was agreed to develop an orientation paper, starting with the note of the meeting and setting up a wiki site to which participants could contribute, including those who had been unable to attend the meeting.

Yi-Cheng Zhang and his group at Fribourg will co-ordinate the development of the wiki with Andrzej Nowak’s expert doing wiki, University of Fribourg preparing the website] and of a website with assistance from Andrzej Nowak and Jeff Johnson.

Bridget Rosewell, Chair of the Meeting