The Economic Record, March 2003

The New Evolutionary Microeconomics: Complexity, Competence and Adaptive Behavior, by Jason Potts, Cheltenham, UK: Edward Elgar. Xii, 239 pp.

The various branches of heterodox economics have lacked a unifying theme. Jason Potts argues that each branch has attacked the orthodox neowalrasian model from a different perspective. Schumpeterian's say it lacks imagination, Post-Keynesians say it lacks emotion, institutionalists say it lacks structure (attention to habits and routines), and the behavioralists say it lacks passion and an accurate picture of the brain. Heterodox schools have attacked the axioms of continuity, completeness, transitivity, reflexivity, montonicity, regularity, and convexity. Instead they insisted on uncertainty, bounded rationality, disequilibrium, increasing returns, and hysteresis. But Potts believes that all have missed the unifying point, namely the geometry of economic space, which he observes is not that of a integral field, but one of complex systems. And, the understanding of complex systems requires nothing less than a revolution in ontology and methodology .

Potts, Lecturer in Economics at the University of Queensland, Australia, argues that the following characteristics of economic systems destroy the assumption of integral space. All elements (such as products, consumers, firms) can't be connected because of uncertainty and bounded rationality and cognitive, heuristic, organizational, spatial, temporal, market, and social structures. At the level of the human brain, it is composed of modules with varying degree of connections. At the level of the supermarket, the consumer can't have a prior preference map of its many products. Out of the few products, organizational forms, and institutions that the brain can account for, the selection process can't be regarded as optimal because fitness is local and time dependent. All elements are not and cannot be connected. Uncertainty cannot be brushed aside by assuming a complete set of forward contingent markets; rather cognitive imagination matters. Increasing returns and path dependency destroy general equilibrium possibilities. Open systems destroy positivistic and deductive methodology.  Economic systems have multiple agents with different mental maps of the world. This lack of homogeneity in cognition (necessarily selective perception) means that connections of elements will be differently made by different people.

"Heterodox economics comes together about a single point, that systems exist and that the crucial fact of their existence is structural incompleteness: that not every element is connected to every other (182)," says Potts. What is the essence of a systems approach? A system is made up of elements and connections chosen by the observer to suit a purpose. Any system can be further disaggregated into subsystems and in turn is a part of a higher order system. The connections within a particular system and between it and other systems are not given, but generated endogenously. "These dimensions do not exist a priori, but are created in the process of economic coordination (79)." The whole can be greater (or less) than the sum of its parts. Potts is frequently describing circular and cumulative causation, but with only tangential reference to Gunnar Myrdal and concepts of beneficial and vicious circles. This is the property of emergence. The formation of new connections and switching from one regime to another is typical of complex systems, not an aberration. A system evolves by changing connections.

Learning and knowledge are at the heart of emergence and the formation of new connections. This conception unites technology, institutions, organizations, and cognitive processes. New technology makes new connections among materials. New institutions and organizations make new connections among people and between people and materials. Thus Potts brings together such otherwise disparate topics as strategic planning and other entrepreneurial processes, imagination, and creativity, in the face of uncertainty. All are evolving, emerging connections. "Change and existence are the same thing (39)."

While Potts argues for an evolutionary perspective, he warns of narrow Darwinism. We must be careful of attributing any special standing to existing connections. Fitness and optimality have questionable meaning when the connection and the environment are both changing. Conception of both fitness and the environment are both cognitive acts, not simply givens. Prior preferences influence choices, but preferences are worked out in the act of choice. Institutions affect individuals and individuals affect institutions.

Systems thinking with real time leads to a different conception of efficiency. One system could make an efficient combination of resources at each point in time, but be inferior to another system over time. The latter could be more flexible in adapting to unpredictable events. There is a continual tension between order and change. Stable systems are more predictable and perhaps less stressful. But, Potts suggests that systems can be so stable and appropriate to a given environment that they are destroyed by small changes from outside. On the other hand, they can be so interconnected that small changes are so magnified that chaos ensues.

Potts suggests that graph theory is appropriate to model the connections of complex systems. The density of connections can thus be seen and related to evolutionary paths. He goes so far as to say, "Systems are characterized by their diversity not so much by what they are made of but by how they are connected together (198)". Potts leaves the reader thinking that the only thing that matters is what is connected to what (on or off) and not the character of the connection. But surely in the human brain, some people have connections across the synapses that are so strong that it leads to addiction while it does not in others. The parties exchanging goods behave differently when they have affinity or malevolence toward each other. Any buyer or seller is connected, but it makes a difference who is the owner of a particular opportunity.

Applications are brief in this theoretical work, but the transaction cost minimization ( economizing) explanation of the firm is critiqued by Potts using his emergent systems perspective (98). If the firm already knew how to maximize profits, then it could design incentive structures to achieve it and prevent costly opportunistic behavior by its contractees made possible by unforeseen events. But, if the organizational structure also produces benefits in the form of knowledge about how to achieve profits, then the means and ends are both products of organization. The problem is not merely to prevent shirking and opportunism, but to create a competent organization that can learn and evolve. Different organizational forms with different transaction costs create different competencies and ideas on future products and technologies. One does not just design institutions (connections) to implement a previously known smooth running machine, but rather to build ( create) competence. This story does not formalize and graph easily and Potts does not try .

Institutions are clearly connections in Pott's view. Still, they are described at a very abstract level. The reader is left to think about what kind of connections in what context are important to evolution and performance (who gets what). There is no typology of useful categories for particular systems within systems. As noted above, there is some attention given to an environment marked by uncertainty. Increasing returns is also mentioned. But, there are other situations creating human interdependencies where the connecting structure produces different performance outcomes. There is little blood in Potts-no sense of how past systems created winners and losers or how these might be changed in the future. Different interests struggle to make their preferred connections count. Property rights are mentioned, but no taxonomy of alternative structures is offered that are relevant to each source of interdependence.

This is a deep and profound book that deserves a deep reading. I have added it to the reading list for my graduate seminar in institutional and behavioral economics.

A. Allan Schmid

Michigan State University

1 "a single (mathematical) operation links any point in economic space with any other point (17)." If economic space "is non-integral, in the sense that the set of interactions cannot be collapsed to a field of actions, then the geometry of economic space must be mapped by a set of specific connections (on a lattice.)" (19)