Engineering of markets and artifacts

https://doi.org/10.1016/j.elerap.2004.06.002Get rights and content

Abstract

In this paper, we continue the dialog started by Roth [The Economist as Engineer: Game theory, Experimentation, and Computation as Tools for Design Economics, Fisher Schultz lecture, Econometrica 70 (2002) (4) 1341–1378] between economics and engineering in the context of design of markets. We take the position that markets and engineered artifacts are the products of a social process of design formulation. Our perspective is that designing of markets and artifacts follow the same kind of problem formulation and solution testing process. Further, we show using two case studies that the design of engineered artifacts and markets are often interdependent. In light of their similarities, methodologies for modeling and testing using theoretical, analytical, computation and physical models in engineering can inform the development of methods for testing in the design of markets. We also illustrate an example of devising a test and results of its execution thorough the use of a simple computational model of an electric power market. We conclude by summarizing the similarities between the designing of markets and artifacts and call for a continued dialogue between engineers and economists.

Introduction

A recent New York Times column by Hal Varian headlined, “Avoiding the pitfalls when shifting from science to engineering,” reviewed a paper by Alvin Roth entitled, “The Economist as an Engineer” [19], [27]. Why do we suddenly see economists, purists in their claim to science, starting to see that engineering is an integral part of economics? Is it a startling return to often-forgotten Thorstein Veblen’s view of economists as engineers of the price system? Is it a renunciation of the claim that economics is a pure science? Science without experimentation is no science at all for all the mathematical devices are not enough to guarantee the validity of the theory if it is not tested in practice; it is here engineering becomes very important. While these are fascinating questions, this paper will focus on a much narrower question. This paper takes as its starting point that the development and use of a market is as much a socially designed mechanism as any other artifact [5]. We take up this perspective – from the engineering side – in what we hope will be a dialogue between economists and engineers about how the experience with engineering design of artifacts can help inform the discussion in economics’ new focus on design of markets. Here the lessons from testing from engineering become important. Perhaps more importantly, this debate between engineer and economist will change the status of designing of artifacts and markets themselves. While we are not economists, but more students of design of engineered artifacts, we want to view economics with the lens of engineering design.

Our goal is to show that understanding design and engineering can create insightful approaches to practical economics that have been brought to fore by the emergence of computing in economics, experimental economics and e-commerce [14], [17], [26], [29]. In these areas, the focus has shifted from being solely on mathematical elegance to also including as equally important, practical design, simulation and creation of markets. This shift in economics from purely analytical to explore various approaches to simulation through computing has been profound. In this shift, Simon and Von-Neumann have been most forceful in constructing a vision of economics as simulated Cyborg Science [14]. Of course, in this new context the issues of methodology of research in design of markets have raised questions that economists, until recently, have not asked themselves. We address some of these issues by staking the claim that design of markets has important similarities to design of engineered artifacts. We further argue that markets and the nature of engineered artifacts are more closely inter-related than most economists tend to acknowledge and engineers dare to state for fear of trespassing.

The paper has five sections. The next section is on the similarities between design of artifacts in engineering and markets in economics. Section 3 addresses the interdependence between market structure and structure of designed artifacts. Section 4 then turns to methodologies in modeling and testing for failures in engineering design and its relationship to design economics. Section 5 is an example of devising a test for excessive profits and testing it using a simple model of the electric power market. Section 6 concludes the paper with some discussions and conclusions on the relationship between design of markets and artifacts.

Section snippets

Markets and artifacts

“Economics is one of the Artificial Sciences” [21, p. 83].

Economics is part of the sciences of the artificial as Simon claims in his well-known paper on bounded rationality and his book on the sciences of the artificial [20], [21]. The mechanisms of economics, “markets (artifacts),” are designed to sustain a particular set of desired behaviors (e.g., to measure and keep time) through constraints on behavior of the participants (functional/structural parts) and on transactions (interactions)

Market design affects artifact design and artifact design affects market design

We will show two cases to illustrate these relationships. The first case is the history and emergence of the electric power industry in the US [10]. The electric power industry in the US was mainly a regulated monopoly until recently when efforts to introduce markets in this industry began. The original electric power industry in the US was decentralized with even small buildings having their own generators and with small companies supplying power to small but often disconnected sets of

Methodological experiences in designing engineered artifacts

The design of aerodynamic shape for aircraft provides an example of methodological approaches to testing a designed artifact with respect to a specified formulation. We draw this example from a case study of the design of Boeing 777 and the use of computing tools [3]. There are four types of methods used in the designing of tests. They are (a) analytical, (b) simulation (CFD), (c) scaled mock up (wind tunnel), and (d) flight tests. In characterizing these methods and their use the author

An example in devising a test for a market design using a simple model for power markets

In a recent paper, we proposed a test to identify market designs that are prone to excess prices and profits, whether from learning or capacity concentrations [25]. In devising a model for this test, we use computational models of agents that are evolutionary in terms of learning from experience, similar to those in the literature on computational economics, evolutionary economics and distributed agents. Even without an elaborate model of the electric power grid, a simple model of the market

Discussions and conclusions

The important message of this paper is that formulation of engineered artifact design problems and market design problems are contextual social processes involving multiple goals set by the actors. In the words of Roth, the importance viewing economics from a design perspective would be the following insight [19]:

“The largest lesson in all this is that design is important because markets don’t always grow like weeds – some of them are hothouse orchids. Time and place have to be established,

Acknowledgements

We would like to thank Raghu Arunachalam for inviting and urging us to write this paper, and Ira Monarch, Robin King and Arthur Westerberg for discussions on this topic and comments.

References (31)

  • P.R. Wurman et al.

    A parameterization of auction design space

    Games and Economic Behavior

    (2001)
  • B. Arthur, Inductive reasoning and bounded rationality (The El Farol Problem), American Economic Association Papers,...
  • C.Y. Baldwin et al.
    (2000)
  • Boeing 777, 100% digitally designed using 3D solids technology. Available from: <http://www.cds.caltech....
  • S. Borenstein, Power politics hurts consumers, The Financial Times, Aug 19,...
  • M. Callon

    Introduction

  • M. Callon

    An essay on Framing and Overflowing: economic Externalities revisited by Sociology

  • E. Chattoe, Just how (Un)realistic are evolutionary algorithms as representations of Social process? Journal of...
  • A. Debs, C. Hansen, Y. Wu, Effective electricity Market Simulators, IEEE computer applications in Power, January...
  • Guesnerie, 1996, 1’economie de Marche’, Dominos, Paris: Flammarion (as referenced in...
  • M. Granovetter et al.

    The making of an Industry: electricity in the United states

  • S.L. Konda, I. Monarch, P. Sargent, E. Subrahmanian, Shared Memory: A Unifying Theme for Design Research and Practice,...
  • R.P. McAfee et al.

    Analyzing Airwave auctions

    Journal of Economic Perspectives

    (1996)
  • P. Mirowski

    Machine Dreams: Economics becomes a Cyborg, Science

    (2002)
  • Cited by (8)

    • Markets come to bits: Evolution, computation and markomata in economic science

      2007, Journal of Economic Behavior and Organization
      Citation Excerpt :

      Improvements in these methods of testing come from advances in several areas including theoretical or analytical modeling, physical testing and computational modeling… This cycle of formulation and reformulation takes place in an evolutionary manner as experience in identification and testing of unanticipated externalities and failures accumulate. ( Subrahmanian and Talukdar, 2003) Here is where matters become contentious, since this literature is clearly divided between those who insist, generally without direct documentation, that economic engineers have successfully designed modern auctions and other markets by means of the application of neoclassical theory (usually, but not exclusively, Nash game theory), and those who come closer to admitting that their experience with engineering rarely validates the existing toolbox.13

    • Aided design of market mechanisms for electricity clusters

      2020, Central European Journal of Operations Research
    • Allocative Efficiency and Traders’ Protection Under Zero Intelligence Behavior

      2011, Dynamic Modeling and Econometrics in Economics and Finance
    • A trading system for flexible VWAP executions as a design artifact

      2009, PACIS 2009 - 13th Pacific Asia Conference on Information Systems: IT Services in a Global Environment
    View all citing articles on Scopus
    1

    Tel.: +1-412-268-8778

    View full text