Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Agent-based modelling in economics
Hamill L., Gilbert N., Wiley Publishing, Chichester, UK, 2016. 256 pp. Type: Book (978-1-118456-07-1)
Date Reviewed: Aug 25 2016

The social sciences have long been plagued by physics envy. The seductive simplicity of closed-form equations with strong predictive power in dealing with billiard balls, planets, and electrons suggests that we ought to be able to analyze the behavior of people with similar precision. Unfortunately, much of the elegance of physics depends on the essential identity of the interacting entities, whether at sub-atomic or planetary scale. Each electron is just like every other electron. But the actors in social systems, people, are very unlike one another, leading to a breakdown in the application of classical physical formalisms to social behavior.

Around 1970, researchers began to explore an alternative approach, representing each actor with a separate process that can sense and respond autonomously to its environment. Perhaps the earliest such model was reported in 1971 by Thomas Schelling, recipient of the 2005 Nobel Prize for Economics, and ran, not in software, but with coins on graph paper. In the early 1980s, Robert Axelrod studied the evolution of cooperation with a computer tournament in which different software agents competed with one another. By 1996, Epstein and Axtell could issue a book-length treatment on Growing artificial societies [1], and Leigh Tesfatsion at Iowa State University began the Agent-Based Economics website, later renamed Agent-Based Computational Economics (http://www2.econ.iastate.edu/tesfatsi/ace.htm), a major resource that this volume mentions only in passing. The same decade saw the first issue of the Journal of Artificial Societies and Social Science (JASSS) (1998), and the field has been exploding ever since.

Nigel Gilbert, the founder of JASSS and second author of this volume, has been a pioneer in the field. The first author is one of his PhD students. Together, they have assembled a collection of economic models that provide a two-track introduction: to agent-based modeling in general, and to economic theory from an agent-based perspective.

The first chapter motivates agent-based modeling around three themes: the heterogeneity of actors (in contrast with the homogeneity assumed by classical differential equation models), the dynamics of the system as it evolves (in contrast with the emphasis on static equilibria in classical analyses), and the importance of tracking interactions among entities. Together, these characteristics often result in a radical distinction between the micro-level behavior of individual agents and the emergent macro-level behavior of the system.

Chapter 2 introduces the reader to agent-based modeling. All the models in the book are constructed in the NetLogo environment and are available from the book’s website, and the authors walk the reader through a simple introductory model.

Chapters 3 through 10 present agent-based models of eight important economic systems or phenomena, explaining both the economic theory behind the phenomenon and how that theory is reflected in an agent-based model, and then analyzing the behavior of the model. The individual subjects discussed include heterogeneity of demand, the influence of social networks in modulating demand, the benefits of barter, various models of market interaction, the function of the labor market, international trade, banking systems, and the tragedy of the commons.

The volume will find extensive use in educational settings, where it would be an excellent supplement to introductory economics courses. It would be even more valuable if a revised edition could address some of its shortcomings. In particular:

While the literature in agent-based economics is vast, some major contributions are not mentioned. One thinks immediately of the immensely influential minority game [2], a formalization of Brian Arthur’s El Farol bar problem [3] that has stimulated a wide range of agent-based and other models for dealing with resource-constrained situations.

Agent-based models have numerous advantages over equation-based models, but they have their own limitations. Since 1996 [4], Robert Axtell has argued eloquently that a proper analysis of a social system should combine, or “dock,” an agent-based model with an equation-based analysis. Much can be learned by observing where these two approaches converge, and the points on which they differ. An equation-based model offers a level of synthesis that is difficult to understand from an agent-based model alone.

In particular, the growing field of econophysics (http://www3.unifr.ch/econophysics/) is finding good success applying the mathematics of statistical mechanics to economic problems. Closer interaction between agent-based modelers and econophysicists promises great gains in our understanding of complex economic and other social systems.

An example of the need for such a breadth of perspective, even in this volume, is the model of international trade presented in chapter 8. This model does not involve either heterogeneity or interaction among the actors, and its dynamics could be presented more parsimoniously and with greater clarity as a conventional system of differential equations. Encouraging computational economists to adopt a richer toolbox, and understanding the appropriate context for applying each tool, should be a central agenda item for courses that might adopt this volume as a supplementary text.

Reviewer:  H. Van Dyke Parunak Review #: CR144711 (1611-0795)
1) Epstein, J. M.; Axtell, R. Growing artificial societies: social science from the ground up. Brookings Institution Press, Washington, DC, 1996.
2) Challet, D.; Zhang, Y.-C. Emergence of cooperation and organization in an evolutionary game. Physica A 246, 3-4(1997), 407–418.
3) Arthur, W. B. Inductive reasoning and bounded rationality. The American Economic Review 84, 2(1994), 406–411.
4) Axtell, R.; Axelrod, R.; Epstein, J. M.; Cohen, M. D. Aligning simulation models: a case study and results. Computational and Mathematical Organization Theory 1, 2(1996), 123–141.
Bookmark and Share
  Featured Reviewer  
 
Multiagent Systems (I.2.11 ... )
 
 
Economics (J.4 ... )
 
 
Model Development (I.6.5 )
 
Would you recommend this review?
yes
no
Other reviews under "Multiagent Systems": Date
Engineering intelligent hybrid multi-agent systems
Khosla R., Dillon T. (ed), Kluwer Academic Publishers, Norwell, MA, 1998. Type: Book (9780792399827)
Aug 1 1998
Linguistic geometry: from search to construction
Stilman B., Kluwer Academic Publishers, Norwell, MA, 2000.  395, Type: Book (9780792377382)
Jan 1 2001
 Transactional agents: towards a robust multi-agent system
Nagi K., Springer-Verlag New York, Inc., New York, NY, 2002.  205, Type: Book (9783540430469)
Apr 13 2004
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy