Navigation – Plan du site

AccueilAppels à contributionAppels passésFrom thought experiments to Agent...

From thought experiments to Agent Based Models and calibration. Reflecting (on) the many facets of simulations in economics.

Editors of special issue: Muriel Dal-Pont Legrand (Université Côte d'Azur), John Davis (Marquette University and University of Amsterdam), Magda Fontana (Università di Torino), Cyril Hédoin (Université de Reims – Champagne-Ardenne), Alan Kirman (EHESS).

Expression of interest: July 30th 2018.

Notification by the editors: August 30th 2018.

Dedicated Workshop at the University Paris 1 Panthéon-Sorbonne, February-March 2019 (to be confirmed)

Deadline for submission of revised papers: May 30th 2019.

Planned publication of the issue: March 2020.

Simulation as a method of investigation or of validation in economics has been used steadily although with parsimony since the second half of the 20th century. Perhaps it would be more correct to say that simulation methods have been confined to the work done by economists that has never come to the foreground and rarely been debated, in contrast to the extensive literature on the use of experimental methods in economics. However, there is now a growing awareness that a body of practices in economics can be grouped together under the heading of simulation methods, and shares a number of specific goals within economics as a whole. Historians of economic thought still largely ignore this territory (Colander (ed.), 2000; Fontaine and Leonard (eds), 2005) and it is only over the last decade that there has been a growing interest in methodological and epistemological issues linked with simulation proper (e.g. Reiss, 2011; Fontana, 2006; Grüne-Yanoff & Weirich, 2010). Recent advances in the use of Agent-Based Models in economics, as well as continuous developments of technologies and transformations in the representations of economic practices, make simulation an ever more attracting or even a necessary step in theoretical and applied work in economics. The time is ripe for reflecting on the recent and older trends pushing forward simulation in economics, discussing and confronting their contexts of development and purposes, and elaborating on the various kinds of epistemological or methodological stances that account for their proper use in economics and other social sciences.

Œconomia intends to publish a special issue covering the various historical and epistemological aspects of simulation. The remainder of this call provides an overview of the different directions of research that may be usefully investigated regarding simulation in economics. It then provides a short (non-exclusive) list of possible themes of investigation and gives some practical information about the schedule of the special issue.

The dominant view of simulation is that the economics profession is still reluctant to place simulation on the same footing as other methodological tools of research for reasons that have to do with the epistemic qualities of the outcomes obtained from simulations. This deserves to be discussed on different grounds. First, there is a question of definition. Second, there is the historical aspect of the issue. Third, there are the methodological issues, which leads to discussing the goals of simulation within the various steps of any economic inquiry.

1. Defining simulation

Simulation as a specific and well-identified method in the social sciences must be questioned, precisely because the use of this label designates more a set of practices and innovative know-how than an established method of investigation.

In economics, simulation refer to a gamut of practices used when one wants to extract or reveal some hidden information about the properties of a model or of some part of a model. Those various practices can serve different goals. They can serve as provisional steps in the building of an integrated model, they can serve to extract information about the scope of behavioral properties that the model describes, and ultimately they can enhance specific properties of the model which could not have been obtained by other more conventional (analytic) methods.

This provisional definition can “materialize” in a variety of practices and technical devices. After all, even a thought experiment could be regarded as a simulation, even though the ‘experimenter’ cannot be taken as providing all the required characteristics of a neutral (de-biased) system to accomplish the task objectively. That said, it is striking that many attempts at defining simulations often employ a two-part definition. The first part involves the material technical device involved in simulations (most often a computer program)—with the view of enhancing technical boundaries to distinguish simulations from other kinds of experiments—while the second part refers to the specific goals of using that device, such as drawing inferences about properties of an economic system.

It is not the purpose of this call to impose a definition, but at least it seems necessary to identify boundaries and related issues. The common idea is that at some moment in the development of a theoretical construct, some partially uncontrolled and indirect human process has been involved to solve it (typically a computer program). Hence, the concept of simulation seems to encompass to some degree both the idea of mimicking or producing a simplified description of the behavior and standard properties of a model on the one hand and the idea that this process can be handled in a variety of situations, or under various initial conditions, ad libitum on the other.

Actually, even such a definition of what simulation is in economics may be too restrictive and unsatisfying. Facing the variety of practices, one cannot but feel that there is still a need for thinking about the boundaries of simulation and how they have evolved in various material and technical contexts, from thought experiments to complex situations of man-machine interactions.

If simulations cannot be defined exclusively through their material devices, would it not be necessary to also consider the kind of uses that the modeler makes of the outcomes as part of the definition? Here, it would then be necessary to look at simulations with a broader description of the research strategy at work and with a broader process of economic inquiry, whereby simulations make sense in an interactive way with other methodological types of economic inquiry. Then, simulations are defined relative to the kind of knowledge that is expected from them and from the kind of use made of this piece of knowledge in later steps of economic inquiry.

2. Historicizing simulations

The use of simulation and algorithms as a step in the work of researchers must be examined in historical context. The boundaries of simulation and what exactly characterizes simulation as a well-identified method of investigation answering a specific kind of question also deserves to be put in historical perspective.

In this respect, this call for papers aims at exploring the meanings of simulation throughout history, beyond the computer implemented representation usually put forth, and also distinguishing between various types of computer simulations (Phan and Varenne, 2010). Thus, the history of economics contains episodes when simulation practices have been dealt with or discussed in pro and con terms. Consider for instance the “thought experiments” that have been argument devices in economics since Hume’s famous thought experiment on money (Schabas, 2008). They deserve an economic investigation in their own right, considering the arguments raised by scholars for or against such kind of arguments throughout history. Likewise, Quesnay’s Tableau économique certainly could be viewed in many respects as a simulation device. Another famous and original instance of simulation would be Irving Fisher’s hydraulic machine (Fisher, 1892).

Throughout 20th century, the issue of simulation appears here and there on different occasions. It is implicit and sometimes explicit in the famous debates about planning (Pegrum, 1941) or in the Phillips study on the stability of a closed economy (Phillips, 1954). The 1960s seems to be a turning point in the development of simulation practices in economics. As Kelman (1960) writes, “Within the past decade there has been a steadily increasing use by operations researchers of computer simulation techniques in studying complex industrial and military systems. Economists are now beginning to recognize that similar techniques may be helpful in analyzing complex economic systems.” A number of contributions in the 1960s and 1970s testify to this increasing interest for simulations in economics (Duesenberry et al., 1960; Houssiaux, 1960; Naylor, 1971 and 1972). The most programmatic work, in this respect, is probably that of Orcutt who had aimed at devising a disaggregated socioeconomic system by the late 1950s (Orcutt, 1957). And the most well-known work is probably Schelling’s spatial segregation model (Schelling, 1969) in which a set of rules regarding the moves of coins on a chessboard was the starting point for reflection upon segregation issues. In addition, the use of Monte Carlo methods from Yule to Hendry is clearly among the early uses of simulation techniques that deserves a specific place in this history (see Orcutt, 1960; Fontana, 2006).

In the 1980s, the interest in simulation seems to be more connected to evolutionary models in economics, hence to a subfield of approaches focused on biological models in the social sciences (Axelrod, 1986). Of course, the use of simulations in macroeconomics, under the heading of calibration, is also a theme for inquiry that has triggered passionate debates (Gregory and Smith, 1991; Hoover, 1995). The increasingly widespread use of agent-based modeling practices, seems to further promote simulation as part of the usual toolbox of economics, though this needs to be documented. On this topic, the multidisciplinary research program at Santa Fe institute since the middle of the 1980s is dedicated to simulating artificial societies and understanding how different patterns of social structures emerge from agent-based interactions (Epstein and Axtell, 1996)

If simulation is an auxiliary complement to modeling practices, then simulations are likely to be used in all fields of economics, theoretical and applied. So, there is need for a better understanding of how various areas of economics, theoretical and applied, have dealt with simulation and how those practices were adopted by their communities of researchers. This seems to be notably the case in evolutionary economics, ecological economics, agricultural economics, resource and energy economics, transportation economics, but also in industrial organization. Broadly, in all those fields, there has been increasing investments in understanding externalities and behavioral assumptions, with parallel efforts to model economic and ecosystem uncertainty, irreversible thresholds (e.g., bankruptcy, destructions), as well human-environment feedbacks. This situation makes the use of computation and simulation technics inescapable—e.g. stochastic dynamic programming (SDP) and agent-based modeling (ABM).

3. Identifying epistemic virtues of simulations

From a methodological perspective, the status of different types of simulation produces a variety of conceptions of their uses and contributions to economic knowledge.

For instance, the use of Monte Carlo methods in econometrics should be appraised within strict boundaries and through careful discussion (Fontana, 2006). Simulation can also produce clues about the likelihood of certain formal properties of a behavioral assumption (Kehoe, 1992). More generally, discussion about the merits and defects of simulation deal with the relative weight of their conclusions.

On the one hand, simulations and numerical experiments are flexible in that they allow for changing values of parameters, and make it possible to simulate situations that would have been infeasible in real experiments, either for cost reasons or for ethical reasons (Axelrod, 1997; Humphreys, 2004). Because they are not constrained by the necessity of formal proofs, they can deal with more complex systems and behavioral assumptions, and allow for more specific theoretical or empirical implications.

On the other hand, simulations offer no certainty and no specific kind of understanding of why the properties of an economic system are what they are. Therefore, simulation results are not as certain as those reached by deduction. One way of considering this situation is, as Reiss (2011) noted, that the high degree of certainty obtained in deducing results is paid for by a high degree of uncertainty regarding the use of the outcome outside the model. Another way could be that the value of simulation remains dependent on how its outcome is part of a broader set of practices in economics (Phan and Varenne, 2010). This stance should certainly be qualified, given that many simulations (e.g. with ABM) are conducted when analytical solutions are known to exist. The same could be said also of various kinds of calibration practices in macroeconomic modeling. Again, searching for a synthetic view of the epistemic values of simulations may seem to be hopeless, and there may be a need for a more case-based approach.

There are a number of other pending issues regarding simulation in the social sciences. What kind of probabilistic knowledge is obtained once various parameters have been combined and outcomes have been analyzed? How are standards of internal and external validity dealt with (Gilbert, 2008)? What are current views about calibration of agent-based models (Bianchi et al, 2007)? What kind of practices might usefully be devised to complement the use of simulations in economics? Etc.

From this overview of simulations in economics, there is a need for collective inquiry as to the contributions that simulation can provide to economics and according to its various uses. This call is both for a systematic effort at thinking about simulation and for case-by-case examinations of their specificity. Œconomia – History, Methodology, Philosophy invites papers for a theme issue which looks at how simulation can be understood to be part of the development of economic thought and how simulation has been integrated into economic inquiry throughout history.

4. Items

A non-exclusive list of subjects that come out from the previous presentation are:

  • Historical accounts of early uses of simulations (Quesnay’s Tableau économique, , Fisher’s hydraulic machine, Phillips machine, early econometrics)

  • Historical accounts of simulation programs for market design and implementation.

  • Historical accounts of the use of simulation in various fields, both in theoretical and applied models (transportation economics, computable general equilibrium, resource and energy economics, industrial organization) and in various periods.

  • Robustness and simulation

  • Simulation and evolutionary systems in economics

  • The use of algorithms in economics

  • Historical aspects of ABM modeling

  • Probability arguments and simulation

  • The use of Monte Carlo methods in economics

  • Do simulations exclude performativity in the model’s outcome?

  • Do simulation results serve as arguments in favor of policy implementation?

  • Thought experiments in economics.

  • Simulation and the figure of the engineering economist

  • Simulation and the socialist calculation debate

  • Simulation as theorizing or as experimenting

  • Simulation in industrial organization (mergers, antitrust policies)

  • Simulation and the boundaries of economics and other disciplines.

  • Calibration in Macroeconomics

  • Various aspects of Agent Based Modeling in economics

  • The epistemology of simulation

Procedure and timeline: Researchers who would like to be considered for participation in this special issue of Œconomia should submit, via email attachment, the paper title, an extended (1000-1500 words) abstract, and the affiliations of all authors. This information should be sent to and is due by July 30th, 2018. Authors whose contributions are selected will be notified by August 30th, 2018. On February-March 2019, the university of Paris 1 Panthéon-Sorbonne will organize a workshop at which authors will be invited to present their completed papers. This event will depend on financial constraints and has still to be confirmed. Revised versions of the papers will be due by May 30th, 2019 and will go through the normal refereeing process of Œconomia. Publication of the special issue is planned for Spring 2020. For further information, please contact the editors or send a message to


Anderson, Jock R. 1974. Simulation: methodology and application in agricultural economics. Review of Marketing and Agricultural Economics, 42(1): 3-55.

Axelrod, Robert. 1997. Advancing the art of simulation in the social sciences. In Simulating social phenomena (pp. 21-40). Berlin and Heidelberg: Springer.

Bianchi, Carlo, Pasquale Cirillo, Mauro Gallegati, and Pietro Vagliasindi. 2007. Validating and calibrating agent-based models: a case study. Computational Economics, 30(3), 245-264.

Cohen, Kalman J. 1960. Simulation of the Firm. American economic review, 50(2): 534-540.

Duesenberry, James S., Otto Ekstein, and Gary Fromm. 1960. A Simulation of the United States Economy in Recession. Econometrica, 28(4): 749-809.

Epstein, Joshua M. and Robert Axtell. 1996. Growing Artificial Societies. Social Science from the Bottom Up. Cambridge, MA: MIT Press.

Fontana, Magda. 2006. Simulation in economics: evidence on diffusion and communication. Journal of Artificial Societies and Social Simulation, 9(2).

Gilbert, Nigel. 2008. Agent-based models (No. 153). Sage.

Gregory, Allan W., and Gregor W. Smith. 1991. Calibration as testing: inference in simulated macroeconomic models. Journal of Business & Economic Statistics, 9(3), 297-303.

Grüne-Yanoff, Till, and Paul Weirich. 2010. The philosophy and epistemology of simulation: a review. Simulation & Gaming, 41(1): 20-50.

Hoover, Kevin D. 1995. Facts and artifacts: calibration and the empirical assessment of real-business-cycle models. Oxford Economic Papers, 24-44.

Houssiaux, Jacques. 1960. Renouveau de l’expérimentation en sciences économique: la simulation par les jeux d’entreprises. Revue d’économie politique, 70(2): 271-300.

Huberman, Bernardo A. and Nathalie S. Glance. 1993. Evolutionary games and computer simulations. Proceedings of the National Academy of Sciences, 90(16), 7716-7718.

Kehoe, Timothy. 1992. Gross substitutability and the weak axiom of revealed preference. Journal of mathematical economics, 21: 37-50

Lehtinen, Aki, and Jaakko Kuorikoski. 2007. Computing the perfect model: Why do economists shun simulation? Philosophy of Science, 74(3), 304-329.

Meisnerr, Werner. Zur Methodologie der Simulation. Zeitschrift für die gesamte Staatwissenschaft / Journal of institutional and theoretical economics, 126(3): 385-397.

Morgan, Mary S. 2004. Simulation: The birth of a technology to create «evidence» in economics. Revue d'histoire des sciences, 339-375.

Naylor, Thomas H. 1972. Experimental economics revisited. Journal of Political Economy, 80(2): 347-352.

Nayor, Thomas H. 1971. Computer simulation experiments with models of economic systems. New York: John Wiley & Sons.

Odum, Howard T. 1989. Simulation models of ecological economics developed with energy language methods. Simulation, 53(2): 69-75.

Orcutt, Guy H. 1957. A new type of Socio-economic System. The Review of Economics and Statistics, 39(2): 116-123

Orcutt, Guy H. Simulation of Economic Systems. American Economic Review, 50(5): 893-907.

Pegrum, Dudley F. 1941. Economic planning and the science of economics. American Economic Review, 31(2): 298-307.

Phan, Denis, and Franck Varenne. 2010. Agent-Based Models and Simulations in Economics and Social Sciences: from conceptual exploration to distinct ways of experimenting. Journal of Artificial Societies and Social Simulation, 13(1): 5. [online at:]

Phillips, A. William. 1954. Stabilisation policy in a closed economy. The Economic Journal, 64(254): 290-323.

Reiss, Julian. 2011. A plea for (good) simulations: nudging economics toward an experimental science. Simulation & gaming, 42(2), 243-264.

Reiss, Julian. A Plea for (Good) Simulation. Nudging Economics Toward Experimental Science. Simulation and Gaming, 42(2): 243-264.

Schabas, Margaret. 2008. Hume’s monetary thought experiments. Studies in History and Philosophy of Science Part A, 39(2): 161-169.

Schelling, Thomas. 1969. Models of Segregation. American Economic Review, 59(2): 488-493.

Taylor, John B. 1999. The robustness and efficiency of monetary policy rules as guidelines for interest rate setting by the European Central Bank. Journal of Monetary Economics, 43(3): 655-679.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search