Elementary Models for a Sustainable Economy

On 21-24 january 2010, a workshop on Elementary Models for a Sustainable Economy was held in Utrecht.

This workshop was organized to explore the contributions of Complex Systems Science (CSS) to economic science and to, more specifically, the representation of (human) agents and the ecology-economy interface. A second objective was to identify (relatively simple) models which can be used to teach and communicate the insights from CSS-based models to policy makers and broader, interested audiences.

It is obvious that economic models based on the rational representative agent homo economicus as the micro-economic foundation cannot explain the rich behaviour of real-world economic systems. But – what is a better representation? During the first day, we explored various ways to represent (human) agents in the economic process. Various novel, and interrelated, ways are attempted: generalized functionals including heterogeneity and interaction, dynamic formation of networks connecting people, and evolutionary mechanisms of replication-selection-diffusion-mutation.

As yet, there is apparently no clear recipe which method should be used, and when, to make the economic agents in (macro-)economic models more in line with real-world diversity and interactivity. Evidently, the modeller’s choice should always be in relation to the model objective. It is this situation, however, which gives many of the CSS applications an air of arbitrariness in their methods and results. A direction for future work may be to list key economic phenomena (observations) and different explanations (mechanisms) and see whether there is complementarity and/or scope for integration.

From a sustainable development perspective, mainstream economic science fails to take into account the risks of irreversible overexploitation and quality loss of resources and, broader, ecosystem services. It also fails to acknowledge the role of human behaviour, notably the social dilemma aspect – which connects this issue to the previous one of agent representation. In this second part of the workshop we addressed approaches to remedy this situation, by for instance including ecosystem service quality explicitly in the management model, by addressing more explicitly the risks of ‘bad attractors’ which are to be avoided in resource use strategies, and by linking the economic agent behaviour more explicitly to underlying resource dynamics in an integrated model. An important issue here is the existence of different time-space scale levels in the natural and the economic systems and the variety in their connections.

Many CSS-researchers feel that important contributions have been made in order to understand and manage complex systems. Yet, many of the new insights – which indeed represent parts of a new emerging paradigm – are fragmented as to methods and outcomes. They are often at least partly misunderstood despite successful attempts at popularization and are only slowly penetrating the policies which aim to intervene in those complex systems. We spent the third day of the workshop to see what can be learnt from interactive modeling and simulation gaming, preferably in a ‘web2.0’ environment. In vivid discussions, it became clear that novel ways of engaging people (‘stakeholders’) into the decision processes is a promising and necessary way to go. Not only can it communicate CCS-insights more effectively, in the form of simple games for instance, but it can also provide the data and observations which permit a gradual improvement in the representation of agents in the models.

Workshop overview by the organiser Prof. Bert de Vries

More information:

Presentations from participants:

Report from day 1 of the workshop, by Peter Baudains, UCL Research Assistant