Clarence Ayres Memorial Lecture (2007): evolutionary institutional economics.

Author:Potts, Jason

The economic system is not ultimately made of resources, but of institutions. And because economic systems differ in their institutions, their comparative performance can be (partially) explained by these differences. Moreover, institutional differences are the outcome of an evolutionary process, implying that as institutions evolve, so do economic systems. Institutional significance, institutional difference, and the causal connection to economic growth and development are all therefore central to the revival of Veblenian institutional economics (Brette 2006; Hodgson 2007).

Yet, this will require more than just a refurbishment of "original" institutional economics with new micro insights from evolutionary psychology, experimental and behavioral economics, complexity theory, replicator dynamics, and game theory. Why? Because as Brette and Mehier (2005) also argue, it is better to begin with a general theory of economic evolution and locate institutions within this, than to start with institutions and derive a theory of the economy.

Toward this end, I propose an evolutionary theory of economic institutions derived from the rule-based analytic framework of micro meso macro recently developed by Dopfer, Foster and Potts (2004); Dopfer (2005); and Dopfer and Potts (2004; 2007). In this framework, an institution is defined as a stable (meso) rule population as the outcome of a (meso) trajectory. The central analytic implication of the evolutionary theory of institutions is that an institution is essentially a meso--not a micro or macro--concept. A meso-centered theory of institutions may provide, in this view, a useful map of the various schools of institutional economics in terms of the different aspects of the meso unit they represent.

What is an Institution?

"Institution" is a well-defined concept in modern economic analysis. Unfortunately, however, it runs to multiple definitions. Worse, they range not only over how institutions function, or how they emerge and change, but extend to seemingly incommensurable definitions of what institutions actually are. The theory of economic institutions is certainly not in conceptual disarray, but it is still well short of a unified analytic framework.

Instead, what currently exists are a number of distinct schools of institutional thought. These draw on different analytic foundations derived from distinct ontological commitments, theoretical models and empirical methods (e.g., Veblen 1899; Commons 1931; Ayres 1944; Williamson 1985; Schotter 1981; Hodgson 1988; 1997; Heiner 1990; North 1990; Vanberg 1994; Young 1998; Loasby 1999; Gintis 2000; Aoki 2001; Greif 2006). In all cases, institutions are of fundamental importance to economic analysis, yet their specific analytic nature remains a point of serious contention.

I shall not rehearse these methodological (or even ideological) differences in detail here (see Rutherford 1994; Dequech 2002; Hodgson 2006), but rather shall seek to unpack only the broad analytic distinctions between original and new institutional economics, along with Austrian, Schumpeterian and game theoretic approaches to the theory of institutions. The purpose of this is to show how they might all then be re-constituted with respect to a general evolutionary analysis based about the micro-meso-macro framework (Dopfer and Potts 2007). I shall argue that "an institution is a 'meso 3' phenomenon," and that many different aspects of this have already been captured across the otherwise disparate schools of institutional economic analysis.

Our starting point is the four points of agreement that connect all institutional analysis:

1) Institutions are artificial in being human artifacts, but also natural in being widely self-organizing and emergent.

2) Institutions are individual in that they relate to human action, but also social in that they are transactions between a system of agents that produce coordination.

3) Institutions are structures that coordinate a system of agents, but also processes that exist in historical time subject to evolution and entropy.

4) Institutions compose the economy as markets, organizations and correlated behaviors, but also constitute the economy as the legal/social/political/ cultural "rules of the game" that define the space of economic opportunity.

These may of course be argued to be the same point: that institutions are coordinating mechanisms between the individual and social process of the creation of economic value, and that these process-structures of coordination are just as important in explaining economic activity as relative endowments of factors of production or the particular array of prices. Yet, within these points of agreement lie many divergent points of analytic emphasis and implied causality.

In Original Institutional Economics (OIE), an institution is a population of systemic behaviors (e.g. Veblen's (1899) "settled habits and routines common to the generality of men"), and also the emergent system that population generates, which Commons (1931, 651) defined as "collective action in control, liberation or expansion of individual action." This sets up analysis in terms of the interaction between different types of institutions, such as ceremonial and instrumental institutions (Ayres 1944), or between innovation systems and economic systems (Nelson 1993). In OIE, an institution is therefore both a rule as a population--a process-structure of correlated behaviors--and a population as a rule--a social coordinating mechanism.

In New Institutional Economics (NIE), the transaction, as the operation of a rule, is the unit of analysis (Commons 1931; Williamson 1985). An institution consists of the population of operations of a rule and institutional analysis is then concerned with analysis of the most efficient level of application of such a "social technology," and also with the particular economic form this takes (e.g., firms or markets). Like OIE, however, NIE also defines institutions in a "macroscopic" sense, through its role in furnishing the necessary conditions for economic coordination to occur at all. This proceeds under the heading of "political economy" in relation to the efficacy of general rules of political organization, law, property-rights, bargaining structures, regulations, etc. Political economy thus also has old (i.e., Marx/Gailbraith) and new (i.e., Hayek/Buchanan, et al.) variants.

The crucial difference between the OIE and NIE therefore resides in their analytic foundations in terms of an open system analysis of an historical process and a closed-system analysis of equilibrium outcomes, respectively. These incompatible analytical views have resulted in an ongoing methodological "cold-war" that has mostly just simmered throughout the twentieth century (Hodgson 2001).

Yet, in both cases, we are dealing with a conception of institutions as the outcome of an evolutionary process in which novel rules become populations, populations become...

To continue reading