Journal of Economic Literature Vol. XXXVI (March 19981, pp. 166-1 92 The Approach of Institutional Economics GEOFFHEYM . HODGSOM The Judge hstitute of Management Studies, University of Cambridge TIw outllov 17 pnrti~ulnrlyg rnfefifill to Peter Corning, Mo,taslti Moriokn, Rtchnrd Nelroa, Jol~n Nigl~tirsgcrle, Douglnss Nosth, Pnul Ttco?ney, ord three (I~LO~I~II MrefUerEee .^ for ~~SCUSY~OIIS, cr~t~caols ?~~i~oer~ 0~ts,h n~~~i ~tnnTche Inpn~z Socicty fop. tlte Pronmteor~o f Sclenze, nnd the Elr ropen11 Conls~~isrioPjl~~,n reA- CE pwgfo~n r~e ~tIu for their .fitutncial asrirto fzce with t11i s By means of the old, we corne to know the new. Confucius I. Introduction spread unfamiliarity with the "old" imerican institutionalism, despite its ODAY, THE TERM "new institutional favored geographic location and ac- Te conomics" is in widespread use and cessible language. The second reason is associated with a vast literature. is that since its decline in America Clearly, the temporal adjective in the after 1930 the "old" institutionalism adopted title of this broad set of postwar has been repeatedly written off, and is theories and approaches has been in- dismissed for failing to provide a sys- tended to demarcate the "new institu- tematic and viable approach to eco- tional economics" from the "old" institu- nomic theory. It is also widely-and tional economics of Thorstein Veblen, wrongly-believed that institutionalism John Commons, and Wesley Mitchell. was essentially anti-theoretical and de- This earlier institutionalism had actually scriptive. been dominant in economics depart- However, characterizations of the ments in American universities just after "old" institutionalism as purely descrip- the First World tive or anti-theoretical do not bear up Despite this, little detailed reference to close scrutiny. Particularly in the has been made by leading exponents of writings of Veblen and Commons, there the "new" institutional economics to is a strong emphasis on the importance this ~redecessor. Two factors may and priority of the tasks of theoretical help to explain this oversight. The explanation and theoretical develop- first is that the history of economic ment. Whatever their limitations, the thought is currently a much neglected early institutionalists addressed crucial subdiscipline, and there is now wide- theoretical issues. 1 Throughout this article, the terms "lnstitutlon- For example, Veblen (1899, 1919) allsrn" an8 "institutional economics,’’ when with- was the firs;social scientist to attempt out a temporal adjective, will be taken to refer to the institutionalism in the tradition of Veblen, a and in- Commons, and Mitchell. stitutional evolution along essentially
Hodgson: Institutional Economics 167 Darwinian lines (Hodgson 1993). Veb- not because of a naive and unsustain- len’s work shares common features with able belief tbat economics can proceed the much later attempts by economists with data alone, and without any theory. to use evolutionary metaphors from bi- Although several institutionalists put ology by Armen Alchian (1950), Fried- their faith in data, they all retained rich Hayek (19881, Kenneth Boulding some degree of belief in the importance (1981), and Nelson and Sidney Winter of a theoretical project. (1982). In addition, Commons (1924, The primary reasons for the failure of 1934) has been acknowledged as a ma- institutionalism lie elsewhere. In par- jor influence on, for example, the be- ticular, the old institutionalistn was havioral economics of Herbert Simon partially disabled by a combined re- (1979) and even the "new" institutional- sult of the profound shifts in social sci- ism of Oliver Williamson (1975). Insti- ence in the 1910-1940 period and of tutionalists also developed a number of the rise of a mathematical style of neo- theories of pricing behavior in imper- classicaI economics in the depression- fectly competitive markets (Marc Tool stricken 1930s. Behaviorist psychology 1991). Traces of the surviving influence and positivist philosophy displaced the of "old" institutionalist ideas are found instinct psychology and pragmatist phi- in many other areas of theoretical and losopby upon which the early institu- applied economics. Indeed, the influ- tionalism had been built. With their use ence of institutionalism persisted for of formal techniques, mathematical some time after the Second World economists caught the imagination of War." both theorists and policy makers. In Nevertheless, there is a shred of jus- comparison, institutionalism was re- tification for the dismissive statements. garded as technically less rigorous, and Ever since Veblen there has been a fail- thereby inferior. ure of the "old" institutionalists to An adequate history of American in- agree upon, let alone develop, a system- stitutionalism remains to be written. atic theoretical core. American institu- The purpose of the present essay is tionalism has bequeathed no integrated quite different. The main aims are: to theoreticaI system of the stature or outline the institutionalist approach in scope of that of Karl Marx, Alfred Mar- broad terms, and to address and evalu- shaI1, Lkon Walras, or Vilfredo Pareto. ate a few "hard core" propositions that The reasons for this failure cannot be were prominent in early institutional- discussed here, save to note that it was ism. A key argument in this paper is tbat the "old" institutionalism offers a radicaIly different perspective on the 2 Notably, several institutionalists ox their fellow nature of human agency, based on the travelers have been elected as Presidents of the American Econornic Association since 1945. John concept of habit. Habits and rules are Kenneth Galbraith, Edwin Witte, Morris seen as necessary for human action. A Copeland, George Stocking, and Bouldlng. In ad- habit-dominated conception of human dition, the "old" institutionalists Simon Kuznets and Gunnar Myrdal received Nobel Prizes in 1971 behavior not only has significant sup- and 1974 respectively. Other schools of thought port from psychology, it is also worthy resemble institut~onalism For example, much of of development and further elaboration pos t-Keynes "Cambridge" economics had a stron institutionalist flavor, particularly the work o! by economists. Nicholas Kaldor and Joan Robinson. The more xe- In an institutionalist view, the con- cent work of Hobert Boyer, Michel Aglietta, and cept of habit connects crucially with the other members of the French rigr~lntions chool also has strong institutionalist affinities. analysis of institutions. This issue has
168 Journal of Economic Literature, Val. XXXVI (March 1998) important implications for both micro- 11. Institutionalist Approaches to economic and macroeconomic analysis. Economic Analysis Some illustrative applications of this The core ideas of institutionalism general approach are considered in both these domains, with arguments concern institutions, habits, rules, and why their evolution. However, institutional- it is important to take account of ists do not attempt to build a single, habit in human behavior. These ap- proaches do not rely on the general mode1 on the basis of those standard as- sumptions ideas. Instead, these ideas facilitate a of individual rationality. However, while strong impetus toward specific and his- outlining the essentials torically located approaches to analysis. of a broadly institutionalist approach, it is conceded that institutionalism In this respect there is an affinity be- itself requires much more theoretical and tween institutionalism and biology. Evo- methodological development. lutionary biology has a few laws or gen- eral principles by which origin and The structure of this paper is as fol- lows. Section development can be explained. Analysis I1 sketches how institu- of the evoIution of a specific organism tionaI economists proceed in practice. It is noted that institutionalism does not requires detailed data concerning the organism and its environment, and also attempt to build an all-embracing, gen- specific explanations relevant to the eral theory. Instead, complex phenom- ena are approached with species under consideration. Evolution- a limited num- ber of common concepts and specific ary biology requires both specific and general theories. In contrast, in physics theoretical tools. Section I11 of this pa- per defines and elaborates the core con- there are repeated attempts to formu- late the general theory of all material cepts of habits and institutions, as phenomena-the so-called "theory of rooted in the early institutionalist the- ory of Veblen and Commons. Section everything" (Jack Cohen and Ian Stew- IV art 1994). In its relatively greater em- shows how much work in the "new" in- stitutionalist vein, including the prob- phasis upon specificities, institutional economics resembles biology rather lem of "institutional infinite regress," than physics. points to a required reformulation of the "new" institutionalist project The institutionalist approacb moves and a possibIe convergence with the thinking from general ideas concerning human of agency, institutions, and the evolution- the "old" institutionalists. Section V considers the circumstances in which ary nature of economic processes to it specific ideas and theories, related to is necessary or convenient for an agent specific economic institutions or types to rely on habits and rules. Not only are habits and rules ubiquitous, but we of economy. Accordingly, there are mul- are typicaIly required to rely on them tiple levels, and types, of analysis. Nev- whether or not (bounded) optimization ertheless, the different levels must be linked together. A crucial point here is is possible. Section VI concludes the es- that the concepts of habit and of an in- stitution (both defined in Section 111) help to provide the link between the 3 Restrictions of space revent extensive refer- ences to specific and the general. the literature. Tie reader is referred to Hodgson et al. (1994) for an encyclopedic treat- In contrast, neoclassical economics ment of some of the to ics discussed here, for moves from an universal theoretical wider references, and gr dircussions af many other issues pertinent to institutionalism. framework concerning rational choice
Hodgson: Institutional Economics 169 and behavior, and moves directly to Unlike the stationary outcomes theories of price, economic welfare, and of general equiIibrium theory, partiaI so However, institutional econom- equilibrium models could seemingly be ics does not presume that its habit- placed in an ongoing, evolutionary con- based conception of human agency it- text. Where institutionalists did agree, self provides enough to move toward however, was that it was necessary to operational theory or analysis. Addi- develop specific theories of pricing that tional elements are required. In par- reflected tlze institutional and market ticular, an institutionalist would stress structures of the modern economy. Fur- the need to show how specific groups of thermore, any attainable general theory common habits are embedded in, and of prices would necessarily be highly reinforced by, specific social institu- limited in its explanatory properties, be- tions. In this manner, institutionalism cause of the variety of institutional pro- moves from the abstract to the con- cesses of price formation in the real crete. Instead of standard theoretical world. models of given, rational individuals, Tbe basis of price tlzeory in institu- institutionalism builds upon psycho- tionalism is thus quite different from logical, anthropological, sociological, that in other schools of economics. and other research into how people be- Neoclassical economics relies on the have. Indeed, if institutionalism had a universal concepts of supply, demand, general theory, it would be a general and marginal utility. Adam Smith, theory indicating how to develop spe- David Ricardo, and Marx relied on the cific and varied analyses of specific phe- labor tlzeory of value. By contrast, in in- nomena. stitutionalism prices are social conven- tions, reinforced by habits and embed- A. An Illustration: Theorim of Pricing ded in specific institutions. Such conventions are varied and reflect the As an illustration, consider the theory different types of commodity, institu- of price formation. Subsequent to Ve- tion, mode of calculation, and pricing blen’s iconoclastic attacks on rational process. economic man, institutionalists were If prices are conventions then they themselves divided on whether Mar- depend in part on ideas and habits. A shallian or other neoclassical price theory of price must in part be a theory theories were acceptable and compat- of ideas, expectations, habits, and insti- ible with institutionalism. While in- tutions, involving routines and pro- stitutionalists generalIy rejected ra- cesses of valuation. Without such a the- tional economic man, this did not nec- ory, there is no adequate explanation of essarily mean the abandonment of all how individuals calculate or form expec- the apparatus of MarshalIian price tations of the future. 4 Neoclassical economics (a term originally It was in this vein that a great deal of coined by Veblen) ma be conveniently defined as empiricaI and theoretical work on the an approach whlch (lr assumes rational, marirniz- ing behavior by agents with given and stable pref- pricing process was carried out by insti- erence functions, (2) focuses on attained, or move- tutionalists and others in the first half ments toward, equilibrium states, and (3) excludes of the twentieth century. Instead of a chronic information roblems (such as uncertainty of the t e explore! by Frank Knight and John ~aynar&e~nes).N otably, some recent develop- 5 Note that "new" inst~tutionalistR onald Coase ments in modern economic theory-such as in has also rejected the standard rationality assump- ame theory-reach to or even lie outside the tions while working within a broadly Marshallian foundarles of this definition. framework.
170 Journal of Economic Literature, Vol. XXXVI (March 1998) general theory of price, attempts were cial markets (Wayne Baker 1984). None made to develop specific theories of of these studies assumes perfect infor- pricing, each related to real-world mar- mation or perfect competition. The ket starting point is an investigation of how structures and types of corporate or- ganization. It was in this context that prices are actuaIly formed in specific in- much of the early work on oligopoly stitutional contexts, followed by the for- pricing was pioneered, including mulation of a pricing theory that is spe- several theories of "mark-up," "administered," cific to the type of institution being or "full cost" pricing (Tool 1991). No investigated. less than other economists, institution- Institutionalism has no general theory aIists want to develop theoretical expla- of price but a set of guideline ap- proaches to specific problems. These nations of such crucial real world phe- nomena as price. Where they differ, lead to historicalIy and institutionally however, is in stressing the practical specific studies which are arguabIy of and explanatory limitations of any possi- more operational value than any all- ble general theory of all prices. embracing general theory. Regrettably, An institutionalist approach to the specific studies of market institutions theory of pricing proceeds by first ex- and pricing processes have received far amining the institutions in which the less research resources and prestige prices are being formed. All aspects of than general equiIibrium and other the institutions that are closely bound highly abstract approaches. up with the process of price formation B. Starting from Habit Some are relevant. What are the costs and Macroeconomic Illustrations how are they evaluated? What routines govern the calculation of prices? What Much empirical data in economics information is avaiIable and what is un- are consistent with the prevaIence of known? By what routines is information habitual activity, even at the macro- obtained and used? What routines are economic level. Consider, for example, used to revise prices in line with the ex- the now neglected theory of the con- perience on the market? Wliat is the sumption function developed by James strategy concerning competitive pric- Duesenberry (1949). This theory was ing? How does this relate to market heavily influenced by Veblen and structure? stressed the role of habit in consumer Obviously, a process of abstraction behavior. Duesenberry’s theory did not and simplification is required to deaI fall out of favor because it did badly on with all this complexity. As a result of empirical tests. In fact it predicted detailed investigations, it may be possi- rather well. Instead, the theory was dis- ble to abstract some key processes gov- carded primarily because it was not erning the formation of prices. One of seen to conform with the presumptions the best examples of this approach is of rational choice theory (Francis Green the work on "administered pricing" by 1979). Duesenberry’s theory proceeded Gardiner Means and his collaborators on the assumption that an estabIished (Caroline Ware and Means 1936). level of income, plus prevailing cultural There is also a close institutionalist af- norms, would lead to an habitual pat- finity with the behavioraI theory of the tern of consumer behavior. The con- firm (Richard Cyert and James March sumer acts imitatively and adaptively, 1964) and with the theory of informa- and also on the basis of ingrained hab- tion cIustering and networking in finan- its. SimiIarly, the major subsequent
Hodgson: Institutional Econonzics 171 study of aggregate consumer demand in regularities from micro-stabilities-is the USA, by Hendricks Houthakker and quite different. In contrast, institutional Lester Taylor (1966), also found that a economics sees regularities at the sys- major part of consumer spending is sub- temic level as being reinforced through ject to inertia, that is, primarily depen- positive feedbacks that act, in part, dent on preceding consumption. upon the microeconomic eIements. It is beside the point to suggest that Hence the latter are not taken as given. all such phenomena can be recast into a The institutionalizing function of insti- more compIex model where habit is for- tutions means that macroeconomic or- mulated as a convoluted result of utility der and relative stability is reinforced maximizing behavior. In principle, the alongside variety and diversity at the possibility of such a recasting cannot be microeconomic Ievel. IronicalIy, by denied. The point is tliat tIze evidence assuming given individuals, the micro- alone does not bestow theoretical pri- foundations project in orthodox eco- macy to rational choice modeling. nomics lzad typically to assume further- (Theoretical arguments against the pri- more that each and every individual was macy of the rationality assumptions are identical in order to attempt to make raised later in tlzis essay.) Furthermore, the analysis tractable. In contrast, insti- the standard principle of parsimony can tutionalism points not to a spurious su- be used to support a fundamental as- pra-individual objectivity, nor to the sumption of human inertia or habit, no uniformity of individual agents, but to less that the standard axioms of ration- tlze concept of socio-economic order, ality. arising upon variety at the micro-level. In general, institutional economists Individual habits both reinforce, and approach the anaIysis of macroeconomic are reinforced by, institutions. Through systems by examining patterns and this circle of mutual engagement, insti- regularities of human behavior, expect- tutions are endowed with a stabIe and ing to find a great deal of imitation, inert quality. Further, institutions play inertia, lock-in and "cumulative causa- an essential role in providing a cogni- ti~n."~Im portantly, regularities or tive framework for interpreting sense- stability at the systemic level may arise data and in providing intellectual habits not despite, but because of variations at or routines for transforming informa- the micro-level. In complex systems, tion into useful knowledge. The strong macro-stability may depend on micro- influence of institutions upon individual chaos (Francesca Chiaromonte and cognition provides some significant sta- Giovanni Dosi 1993; Cohen and Stewart bility in socioeconomic systems, partly 1994). Or systemic constraints may pre- by buffering and constraining the di- vail over micro-variations (Gary Becker verse and variable actions of many 1962; Dhananjay Gode and Shyam Sun- agents. der 1993). With tlzis line of argument the rela- The prominent view that it is neces- tive autonomy of ~nacroeconomicsa nd sary to build microeconomics on "sound the idea of the workability of aggregates microfoundations"-to derive macro- is reinstated. This contrasts with reduc- tionist modes of theoretical analysis GVeblen was one of the first to develop the con- that see macroeconomic phenomena as cept of cumulative causation that directly or indi- necessariIy explained by the micro- rectly i~~fluenceAdl lyn Young, Myxdal, Kaldox, economic. Here the institutionalists and others. See Veblen (1919, pp. 70-77, 173-77, 240-43, 370, 436) and Hodgson (1993, ch. 9). made a significant contribution. MitchelI
172 Journal of Economic Literature, Vol. XXXVI (March 1998) and his colleagues in the National Bu- 1994). Furthermore, institutionalists re- reau for Economic Research in the ject the idea of the primary ontological 1920s and 1930s played a vital role in unit of the given, institution-free indi- the development of national income ac- vidual, upon which the microfounda- counting and suggested that aggregate, tions project rested. (Th1*s issue is ex- macroeconomic phenomena have an on- pIored in more detail below.) Arguably, toIogicaI and empirical legitimacy. This the failure of the mainstream micro- important incursion against reduction- foundations project points to the need ism created space for the Keynesian to develop a quite different overall ap- revolution. Through the deveIopment of proach. In this there are both micro- national income accounting, the work of economic and macroeconomic levels of MitcheIl and his colleagues influenced analysis, each with a degree of relative and inspired the macroeconomics of theoretical autonomy, but at the same John Maynard Keynes. Partly for this time both levels are connected by con- reason, there is a close and explicit af- ceptual links and spanning explanations. finity between institutionalism and what The abandonment of a standard mi- is often described as "Post Keynesian" crofoundations approach does not mean macroeconomics. that institutionalists are necessarily de- The fact that institutions typically prived of the capacity to build models portray a degree of invariance over long or make predictions. On the contrary, periods of time, and may outlast indi- repeated studies-including those cited viduals, provides a reason for choosing above-have shown that models with institutions rather than individuals as a strong elements of inertia, explained in basic unit. Most institutions are tempo- terms of habit persistence, are good rally prior to the individuals that relate predictors in the macroeconoinic to them. We are all born into and so- sphere. It is also well-known econo- cialized within a world of institutions. metric folklore that naive prediction Recognizing this, institutionalists focus models, based on simple extrapolations on the specific features of specific insti- from the recent past into the future, are tutions, rather than building a general often much better predictors of macro- and ahistorical model of the individuaI economic performance than much more agent. sophisticated economic models. Institu- However, the proposed alternative is tionalists regard such results as confir- not a methodoIogica1 colIectivism where mation of the phenomena of habit per- individual behavior is entirely explained sistence, and of institutional lock-in and by the institutional or cultural environ- self-reinforcement. ment. Complete explanations of parts in Institutionalism works from the "styl- terms of wholes are beset with prob- ized facts" of the macroeconomic sys- lems of equivalent stature to those of tem and attempts to uncover the under- the inverse procedure. Just as struc- lying structural features of the system tures cannot be adequateIy explained in that help explain these outcomes. This terms of individuals, individuals cannot requires analyses that are both quantita- adequately be explained in terms of tive and qualitative. Consider an exam- structures. ple, starting from the stylized fact that The failure of the mainstream micro- productivity growth in the United foundations project confirms the diffi- States economy has been lower in the culty of modeling the whole in terms of last 40 years or so than in many East the individual parts (S. Abu Turab Rizvi Asian and other competing countries.
Hodgson: Institutional Economics 173 Data may also reveal the stylized fact than mere correlation between vari- that the proportion of GNP devoted to ables. Hence it is important to explain investment in the United States has the causal mechanisms linking share been relatively low. But, for the institu- ownership structures with low invest- tionalist, the analysis does not stop with ment, and, in turn, with lower growth mere statistical correlation. The task is rates of productivity. Such causal expla- to explain the institutional constraints nations could involve many factors, in- and causal processes giving rise to both cluding national cultures, political sys- low investment and low productivity tems, and so on. InstitutionaIists are not growth. wedded to any one hypothesis or any Consider a tentative hypothesis. The one theory on this issue, but in general functional and cultural separation of fi- the institutional approach stresses the nancial from industrial institutions may importance of comparative institutional have encouraged a short-term orienta- analysis, and the examination of a broad tion toward investment returns. The set of factors, in searching for an ade- sparse institutional connections be- quate causal explanation. tween finance and industry, and the C. The institutionalist Approach. relative lack of shared personnel and Some General Remarks shared vested interests, may aIlow the financial sector to concentrate on maxi- For some, the general approach out- mizing its returns to short-, rather than lined above may seem fairly obvious, Iong-term, investment. Furthermore, adding nothing new. Several points can the relatively low degree of cross own- be made in response to such a remark. ership between industrial corporations First, there is a degree of emphasis on may further encourage a primary orien- institutional and cultural factors that is tation of the corporation toward short- not found in mainstream economic the- term financial markets and short-term ory. Second, the analysis is openly inter- investment decisions. A first step in ap- disciplinary, in recognizing insights praising this hypothesis wouId be to from poIitics, sociology, psychology, and vaIidate its key assumptions, for exam- other sciences. Third, there is no re- pIe by looking at the . data on the course to the model of the rational, util- distribution of share ownership, includ- ity-maximizing agent. Inasmuch as a ing cross share ownership between fi- conception of the individual agent is in- nancial and industrial corporations. An volved, it is one which emphasizes both important next step would be to carry the prevalence of habit and the possibil- out simiIar and comparative investiga- ity of capricious novelty. Fourth, tions in relevant economies with higher mathematica1 and statistical techniques rates of productivity growth, such as Ja- are recognized as the servants of, rather pan. The existence elsewhere of higher than the essence of, economic theory. degrees of corporate cross ownership of Fifth, the analysis does not start by stocks and shares would provide a basis buiIding mathematical models: it starts for attempting to evaluate the hypothe- from stylized facts and theoretical con- sis still further. Econometric evidence jectures concerning causal mechanisms. of a significant statistical correlation be- Sixth, extensive use is made of historical tween the relevant variabIes would be and comparative empirical material con- important, but far from enough. Institu- cerning socio-economic institutions. In tionalists stress tlze need to outline the several of these respects, institutional real causal linkages involved, rather economics is ’at variance with much
174 Jouraal of Economic Literatuw, Vol. XXXVI (March 1998) of modern mainstream economic theory. the methodoIogica1 and meta-theore- This should not and does not mean, tical foundations of the institutionalist however, that institutionalists become approach. Constraints of space prevent mere data-gatherers. No understanding further elaboration. What is clear, how- nor explanation is possible without the- ever, is that such metlzodoIogica1 issues ory. Veblen and Commons, as founders have become lively topics of debate in of the "old" institutionalism, knew that the 1980s and 1990s, with a veritable theory does not arise by induction from explosion of Iiterature in the area of data. All empiricaI anaIyses presuppose economic methodoIogy. Institutionalists a set of concepts and an implicit or ex- have made significant contributions to plicit theory. For this reason, to start this literature and are thus playing a from stylized facts must itself require a part in the development of appropriate prior conceptual framework. Institu- approaches to economic analysis. tionalism attempts to provide this broad D. Soma Contemporary Issues framework in terms of a set of meta- in litstitutionalist Theory tlzeoretic and methodological guide- Iines. There is no single, agreed set of It has already been admitted that in- definitive guidelines among institution- stitutionalism lacks a systematic core alists, but a number of common themes theory. Institutionalism does not seek a general theory of everything but it does emerge. For example, a central problem is the require a coherent framework of anaIy- identification of what may be termed sis and a workable methodology. "ideal types." These are abstract de- In particular, there remains impor- scriptions of situations, phenomena, or tant scope for the development of an in- persons that indicate the general fea- stitutionalist microeconomics. Although tures on which a theorist will focus as past institutionalists made significant crucial for purposes of explanation. It is progress in the development of theories impossibIe to include a11 details and all of pricing in imperfectly competitive features in such a venture because so- markets, there is still a great of work cio-economic systems are too complex that can be done. In economics since and are open in the sense in which they the Second World War such alternative interact with their outside environment. approaches have received negligible re- A process of abstraction must occur search resources. However, alternative where the essential structures and fea- theories of consumer behavior have tures of the system are identified. The flowered in other disciplines, such as crucial question, of course, is which marketing (Roger Mason 1995). Some ideal type is to be selected in the analy- of this research has a strong institution- sis of a given phenomenon. To answer aIist flavor, in part because it brings to- this question requires a methodology to gether insights from psychology and distinguish between the general and the other social sciences. In developing specific aspects of any given phenome- theories of individual economic behav- non. By making this distinction, and ior, as well as eIsewhere, institutional- perhaps by using comparative material ists look forward to the possibility of a from other socio-economic systems, it is much more extensive and fruitful dia- possible to construct and develop hy- logue across disciplinary barriers. potheses concerning the key causal link- The institutionalist emphasis on habit ages behind the observed phenomena. and routine also fits in well with the This is an all-too-sketchy account of evolutionary models developed by Nel-
Hodgson: Institutional Economics 175 son and Winter (1982) and their follow- they operate, in place of former concep- ers. As Veblen (1899) himself sug- tions and habits of thought. In particu- gested, the evolutionary paradigm pro- lar, if the metl~odsa nd criteria of "opti- vides a basis for encapsulating both mization" are themselves being learned continuity and change, both inertia and how can learning itseIf be optimal? novelty. Habits or routines may adapt InstitutionaIists bring a different per- slowly or "mutate" as agents attempt spective to the analysis of learning by purposeful improvements. In addition, seeing it, in part, as a transformative there is some selection process by and reconstitutive process, involving which some habits and routines are re- the creation of new habits, propensities, tained and imitated, and others fall out and conceptual frameworks (Veblen of use. Institutionalism is congenitally 1919; James Murphy 1994; Henry Plot- an "evolutionary economics ." Like all kin 1994). InstitutionaIists need to capi- work in this vein, it is biased toward dy- talize on their prima facie conceptual namic rather than equilibrium-oriented and methodological advantage in this modes of theorizing. area and develop theories of learning Problems of cognition and learning that are appropriate for a knowledge- have been thematic to institutionalism intensive and rapidly changing world. since its inception. Instead of the pre- The remainder of this essay explores sumed bedrock of given individuaIs, further some of the methodologica1 and there is the idea of interactive and par- theoretical issues that have been raised. tially malleable agents, mutually en- The following section outlines some of twined in a web of partially durable and the core theoretical characteristics of self-reinforcing institutions. Institution- institutionalism. Subsequent sections alist theory is admittedly underdevel- address the problem of reductionism in oped in this area but institutionalists economic theory, and show that main- may potentiaIIy be in a relativeIy strong stream explanations of economic and in- theoretical position. Although main- stitutional phenomena in terms of stream economics has addressed the given, rational individuals are not as ro- concept of learning in recent years, at bust as is often supposed. These argu- root there are severe problems in the ments give further credence to the in- approach based on the assumption of a stitutionalist approach. rationaI actor. The key question is what is meant by "rational learning." How 111. Some Core Characteristics of can agents be said to be rational at any institutionalist Theory given moment when they are in the pro- A. The "New" and the "Old" cess of learning? The very act of learn- institutionalism Cornparad ing means that not all information is possessed and global rationality is com- What is the essential difference be- promised or ruled out. Furthermore, tween the "old" and the "new" institu- much more than inputting data or esti- tionalism? Answering this question is mating probabilities is involved. Learn- made more difficult because there is no ing is more than the acquisition of in- unanimity, even among its adherents, as formation; it is the development of new to what is precisely to be included in means and modes of cognition, calcula- the "new" variety. Nevertheless, an an- tion, and assessment. This means that swer is possibIe if we focus on the com- agents are building up new repre- mon theoretical core of some of the sentations of the environment in which most prominent and influential "new"
176 Journal of Economic Literature, Vol. XXXVI (March 1998) institutionalist writings, such as by North (1981), Richard Posner (1973), Andrew Schotter (1981), and Williamson (1975). Despite their analytical and pol- icy differences, there are some common presumptions behind a11 these works. The characteristic "new" institution- alist project is the attempt to explain the emergence of institutions, such as the firm or the state, by reference to a mode1 of rationaI individua1 behavior, Figure 1. The Inst~tutior~liAstc tion-Iufor~nationC oop tracing out the unintended conse- quences in terms of human interactions. An initial institution-free "state of na- As we11 as in Menger’s writings, this important core theme of an action-in- ture" is assumed. The explanatory move- ment formation loop is cIearIy evident, for ex- is from individuals to institutions, taking individuals as given. This ap- ample, in North’s (1981) theory of the proach is often described as "methodo- development of capitaIism, WiIIiam- logical individualism. "7 son’s (1975) transaction cost analysis of Along these lines, Carl Menger the firm, and Schotter’s (1981) game- theoretic analysis of institutions. The (1892) long ago saw the institution of money as emanating in an undesigned value of this core idea shouId not be de- manner from the communications and nied. However, despite the temporal adjec- interactions of individual agents. Once convenient regularities become promi- tive, the "new" institutionaIism is built nent, a circuIar process of institutional upon some antiquated assumptions con- cerning the human agent, derived from self-reinforcement takes place. E merg- ing to overcome tlze difficulties of bar- the individualism of the Enlightenment. ter, money is chosen because it is con- In this 300-year tradition, a key idea is venient, and it is convenient because it the notion that the individual can, in a is chosen. Other similar exampIes in the sense, be "taken for granted." Accord- "new" institutionalist literature include ingly, the individual is taken as the ele- traffic conventions (Schotter 1981; mentaI budding block in economic the- Robert Sugden 1986). For example, ory. Strictly, it is not a question of once a majority of car drivers stick to whether or not a theorist is found to ad- the right-hand side of the road, it is mit that individuals-or their wants and preferences-are changed by circum- clearly rational for all drivers to follow the same rule. Accordingly, the emer- stances. Indeed, many economists admit gent convention is reinforced and insti- that individuals might so be changed. tutionalized by imitation, and by effi- What is crucial is the individualistic economist assumes, for the purposes of cient use of "all relevant information" (Schotter 1981, p. 160). We may stylize economic enquiry, that individuals and a core idea involved here in terms of an their preference functions should be action-information loop, as shown in taken as given. Thus the demarcating Figure 1. criterion is not the matter of individual malleabiIity per se, but the willingness, 7The enormous and controversial literature on or otlzerwise, to consider this issue as this topic is too large to address here. See Hodg- son (1988) for a survey and d~scussion an important or legitimate matter for
Hodgson: Institutional Economics 177 economic analysis. The commonplace difficuIties with the "new" institutional- statement by mainstream economists ist approach. It is argued below that to that tastes and preferences are not the proceed from the assumption of given explananda of economics thus derives individuaIs in an institution-free "state directly from the individualist tradition. of nature" is theoretically misconceived. Likewise, the conception of economics Accordingly, developments in the "new" as "the science of choice" takes the institutionalism show some sign of choosing individual and her preference yielding some ground to the "old," or at functions as given. Unlike the "old" in- least creating the possibility of a fruitful stitutionaIism, the "new" institutional- dialog between the two approaches. ism has also taken such individuaIistic NotabIy, some Ieading mainstream presuppositions on board. economists seem to be moving toward A common thread in the literature of the view that the individual should not "old" institutionalism, from Veblen be taken as given. Joseph Stiglitz (1994, through Commons and Mitchell, to pp. 272-73) has accepted that "certain Myrdal and Galbraith, is the idea that in aspects of human nature are endoge- economic analysis the individuaI should nous to the system . . . traditionaI eco- not always be taken as given. Tlze gen- nomic theory was cIearly wrong in treat- eral use of given preference functions ing individuals as immutable." A to model individuals is rejected by insti- foremost agenda item for institutional tutionalists. IndividuaIs interact to form economics is to incorporate a richer, institutions, while individual purposes context-dependent conception of hu- or preferences also are molded by so- man agency within a systematic and rig- cio-economic conditions. The individual orous theory. is both a producer and a product of her B. Agency and Habit circums tances .8 It is possibIe to distinguish the new Rejecting the mainstream approach institutionalism from the "oId" by to economic theory, wit11 its conception means of the above criterion. This dis- of the utility-maximizing individual, the tinction holds despite important theo- founders of the "old" institutionalism retical and policy differences within promoted an alternative conception of both the new and the old institutionalist human agency. Such an alternative was However, there are conceptual well developed by the early twentieth century, in the influentia1 writings of in- SThe recent work of Becker (1996) might at stinct psychoIogists such as William first sight appear to "take account of tastes" by building a model wlth "endogenous prefer- James and William McDougall, and ences"-in which individual utillty functions con- pragmatist philosophers such as CharIes tain arguments such as "culture" This is not Sanders Peirce. For all these writers the the case because an Immanently conceived (meta)pxeference function for each individual influence of Darwinian biology was cru- is still assumed at the outset, it is an unexpIored ciaI. Although instinct psychology was "black box" that still remains to be expla~ned. subsequently eclipsed by behaviorism 9 Notably, and in contrast to the accounts in Richard Langlols (1986) and Malcolm Rutherford (Carl Degler 1991) it enjoys a rehabili- (1994), this criterion places the work of Nelson tation today (Leda Cosmides and John and W~nter( 1982) squarely in the "old" ~nstitu- Tooby 1994a, 199413; Plotkin 1994; Ax- tionalist category. This affinit with the "o1d" insti- tutlo~lallsmi s now reco nize 1by Nelson and Win- tlzur Reber 1993). ter themselves In ad$ition, some of the later Following these leading psychologists writings of Hayek (1988) come cIoser to the 01d and phiIosophers of their time, the early institutionalism, as do recent works by North (1990) institutionalists saw habit as the basis of
Journal of Economic Literature, Vol. XXXVI (March 1998) human action and belief. Habit can be It is as if institutionalists and rational defined as a largely non-deliberative choice theorists have been engaged in a and self-actuating propensity to engage century-long game of Go. Each has been trying to place the pieces of an ar- in a previously adopted pattern of be- havior. A habit is a form of self-sustain- gument in an intricate attempt to encir- ing, nonreflective behavior that arises cle the postulates of the other. Notably, however, leading advocates of the ra- in repetitive situations. Many modern economists have ad- tionaI choice paradigm, such as Becker dressed habit. However, mainstream (1962) have demonstrated that an "irra- economists typically regard habit as an tional" mode of behavior, in which evocation or appendage of rational agents are ruled by habit and inertia, is choice, and thereby explicable in its just as capable of predicting the stan- terms. Habits are seen as the result of dard downward-sloping demand curve and the profit-seeking activity of firms. an earlier choice, or as a deliberate means of avoiding endIess deliberation. Becker showed how the negatively in- Rationality thus retains explanatory pri- clined market demand curve can result macy (Becker 1992; Robert Pollak from habitual behavior. Actors "can be 1970). said to behave not only "as if’ they were The treatment of habit by the phi- rational but also "as if’ they were irra- tional: the major piece of empirical evi- losophers and psychologists who influ- enced the early institutionaIists was dence justifying the first statement can quite different. The explanatory arrow equally well justify the second" (Becker ran in the opposite direction: instead of 1962, p. 4). habits being explained in terms of ra- Kenneth Arrow has also accepted the tional choice, rational choice was ex- possibility of an alternative approach plained in terms of habits. Further, based on habit. After outlining a possi- habit was linked with knowledge and ble "non rational" and habit-based belief, seeing the essence of belief as mode1 of human behavior, Arrow (1986, the establishment of habit. All ideas, in- p. 5386) remarked: "Without belaboring cluding beliefs, preferences, and ra- the point, I simply observe that this the- tional modes of calculation, were re- ory is not only a logically complete ex- garded as evolutionary adaptations to planation of behavior but one that is circumstances, established through the more powerful than standard theory acquisition of habitual propensities. and at least as capable of being tested." At first sight, both approaches seem Accordingly, the "accuracy of the pre- feasible: habit can be regarded as the dictions" or other familiar criteria for basis of rational choice, or rational theory selection do not give outright choice can be seen as the procreator of victory to rational choice. Common ar- habits. The fact that economists dispIay guments for regarding rational choice as an habitual inclination to assert the lat- pre-eminent do not carry as much ter should not obscure the possibility of weight as has been supposed. This ques- a reverse ordering. (If the alIegation of tion of the explanatory primacy of habit the priority of rationality over habit is versus that of rationaI choice wiIl be ex- itself simply a matter of habit then it is plored again later in this essay. For by that fact undermined.) The problem now, there is at least a prima facie case of adjudication between these two para- for examining the distinctive treatment digms is not as straightforward as is of habit in the work of the "oId" institu- often assumed. tionalists.
Hodgson: Institutional Economics 179 To many mainstream economists, (1932, p. 84). He saw an institution as such a habit-driven characterization of "a way of thought or action of some the agent is excessively deterministic, prevalence and permanence, which is seemingIy denying free will and embedded in the habits of a group or However, it can be argued tlze customs of a people." This elabo- that the conception of the agent as an rates Veblen’s (1919, p. 239) earlier utility-maximizer based on fixed prefer- definition of an institution as "settIed ence functions itself denies free wiI1 habits of thought common to the gener- and choice. An individual ruled by her ality of men." Notably, in the "old" in- preferences is made a prisoner, not sim- stitutionalism, the concept of habit ply of her social environment, but also plays a central role both in its definition of her utility function. It is as if she is a of an institution, as in its picture of hu- robot, programmed by her preferences. man agency. Within such a deterministic machinery, Although, by contrast, definitions of critics find it difficult to find any place an institution in the "new" institutional- for real choice. As James Buchanan ism do not typically include the notion (1969, p. 47) once remarked: "Choice, of habit, they often share with the oIder by its nature, cannot be predetermined institutionaIism a broad, rather tlian a and remain choice." narrow, conception of an institution. In- In trying to gain a deeper under- stitutions, are regarded as general regu- standing of the nature of human agency larities in social behavior (Schotter we are addressing an age-oId, highIy 1981, p. 11) or "the ruIes of the game in complex, and inadequately resolved society or . . . the humanly devised con- philosophical and psychological prob- straints that shape human interaction" lem. In some of the best writings in the (North 1990, p. 3). area we find attempts to reconcile ha- All these definitions, "new" and "old" bitual behavior, on the one hand, and institutionalist alike, involve a relatively purposefuIness, choice, novelty, and broad concept. They encompass not creativity, on the other (Michael Po- simply organizations-such as corpora- lanyi 1967; Murphy 1994; Plotkin 1994). tions, banks, and universities-but also Common to these approaches is the integrated and systematic social entities idea of habits being the foundation of such as money, language, and law. The learned behavior. Accordiugly, examin- case for such a broad definition of insti- ing the ways in which new habits are tutions is that all such entities invoIve acquired is as important as a recogni- common characteristics: tion of their durability. This duaI stress is thematic to the institutionaIism of Allinstitutions involvetbeinterac~ono fagents,with Veblen and others, and retains its cmcial information Fedbacks A11 institutions have a number of charactenstic and theoretical relevance today. common conceptions and routines. Institutions sustain, and axe sustained by, shared C. From Habits to institutions conceptions and e~ectations. Although they axe nelthex immutable nor immortal, One of the most useful definitions of institutions have relatively durable, self-reinforcing, an institution was provided by the insti- and persistent qudtles. tutional economist WaIton Hamilton Institutions incoxpol-ate values, and pxocesses of normative evaluation. In particular, ~nstitut~ons ’ONote that, contrary to a w~despreadm iscon- reinforce their own moral legitlmatlon that which ception, VebIen (1899, p. 15; 1914, p 31 32 endures is often-rightly or wrongly-seen as 1919, p 75; 1934, pp 80, 175) re eate$ as’serted morally just. that human behavror war purpore!ul,
Journal of Economic Literature, Vol. XXXVI (March 1998) A broad definition of an institution is and made meaningful by agents. In- consistent with long-standing practice in deed, as discussed below, rationality it- the social sciences. More narrowly, or- self is regarded as reliant upon institu- ganizations may be defined as a special tionaI props. subset of institutions, involving deliber- The availability of common cognitive ate coordination (Viktor Vanberg 19941, tools, as well as congenital or learned dispositions for individuals to conform and recognized principles of sovereignty and command. Language is an example wit11 other members of the same group, of an institution that is not an organiza- work together to mold individuals goals tion. A business corporation is an institu- and preferences. Accordingly, individu- tion and also an organization. All institu- als are not taken as given. In main- tions and organizations exhibit the five stream economics, widespread lip-ser- characteristics above. vice to notions of individuality and However, a key difference between choice may have helped to obscure the the "old" and the "new" institutionalism degree in reality to which conformism here is that in the former the concept of or emulation actuaIly occur, even in habit is central. For the "old" institu- modern competitive economies. For an tionalists, habit is regarded as crucial to "old" institutionalist, such outcoines are the formation and sustenance of institu- an important part of the institutional tions. Habits form part of our cognitive self-reinforcing process. abilities. Cognitive frameworks are From an "old" institutionaIist per- learned and emulated within institu- spective, the institutional action-infor- tional structures. The individual relies mation loop in Figure 1 stands out in on the acquisition of such cognitive higher reIief. The imitation and emuIa- habits, before reason, communication, tion of behavior leads to the spread of cboice, or action are possible. habits, and to the emergence or rein- Learned skills become partiaIly em- forcement of institutions. In turn, insti- bedded in habits. When habits become tutions foster and underline particular a common part of a group or a social behaviors and habits, and help transmit culture they grow into routines or cus- them to new members of the group. toms (Commons 1934, p. 45). Institu- The additional emphasis here concerns tions are formed as durable and inte- the roIe of habit both in sustaining indi- grated complexes of customs and vidual behavior, and in providing the in- routines. Habits and routines thus pre- dividual with cognitive means by which serve knowledge, particularly tacit incoming information can be inter- knowledge in relation to skiIls, and in: preted and understood. Our under- stitutions act through time as their standing of the durable and self-rein- transmission belt. forcing qualities of institutions is Institutions are regarded as imposing enhanced. form and social coherence upon human The thrust of the "old" institutionalist activity partly through the continuing approach is to see behavioraI habit and production and reproduction of habits institutional structure as mutually en- of thought and action. This involves the twined and mutually reinforcing: both creation and promulgation of concep- aspects are relevant to the full picture tual schemata and learned signs and (Commons 1934, p. 69). Choosing insti- meanings. Institutions are seen as a cru- tutions as units of analysis does not nec- cial part of the cognitive processes essarily imply that the role of the indi- through which sense-data are perceived vidual is surrendered to the dominance
Hodgson: Institutional Economics 181 of institutions. A dual stress on both more open-ended and ap- agency and structure is required, redo- proach, redolent of the earlier institu- Ient of similar arguments in sociology tionalists. and philosophy (Roy Bhaskar 1979; An- Two opposite types of error are thony Giddens 1984; Harrison White possible. The "cultural determinists" 1992). Both individuals and institutions place too much stress on the molding of are mutually constitutive of each other. individuals by Such Institutions mold, and are molded by, "oversociaIized" views of human be- human action. havior have been widely criticized Institutions are both "subjective" (Mark Granovetter 1985). At the oppo- ideas in the heads of agents and -objec- site end of the spectrum, the "new tive" structures faced by them. The twin institutional economics" gives no more concepts of habit and institution may than weak stress to the processes of in- thus help to overcome the philosophical stitutional conditioning, and focuses dilemma between realism and subjectiv- primarily on the emergence of institu- ism in social science. Actor and struc- tions out of the interactions of given in- ture, although distinct, are thus con- dividuals. This section explores the nected in a circle of mutual interaction problems that may arise if exclusive and interdependence. stress is put on the latter direction of Summarizing the argument so far, it causaIity. has been shown in this section that the The distinctive "new" institutionalist wider recognition of the importance of project has been identified as the at- institutions and rules in human society tempt to explain the existence of insti- is endorsed by institutional economics tutions by reference to a given model of in the Veblen-Commons tradition, but individual behavior, and on the basis of with an additionaI and cruciaI stress an initial institution-free "state of na- on the role of habit. A circle of interac- ture." The procedure is to start with tion between actor and structure is given individuals and to move on to ex- based on the linked concepts of habit plain institutions. and institution. Section IV below adds Admittedly, substantial heuristic in- further credence to this argument by sights about the development of insti- considering some difficulties that are tutions and conventions have been raised when the explanatory circle is gained on the basis of the assumption of broken and the individuaI is given un- given, rationaI individuaIs. The main warranted ontological and explanatory problem addressed here is the incom- priority. pleteness of the research program in its attempt to provide a general theory of the emergence and evolution of IV. Explaining lmtitutions: The institutions. Some tentative and "evolu- Problen~o f Institutional lnjinite Regress tionary" approaches to solving this In this section it is argued that at- probIem are addressed at the end of tempts to explain the origin and suste- this section. nance of institutions on the basis of the assumption of given individuaIs have in- llflutherford (1994, pp. 40-41) notes how ternal flaws and inconsistencies. Ac- American institutionalism itself moved toward cul- tural determlnism and methodological collectivism cordingly, attempts to explain institu- in the post-1940 eriod. Such a one-sided empha- tions in this way may have to be sls was not foun 2l n the earller institutional~smo f abandoned. The door is opened to a Veblen and Commons
Journal of Economic Literature, Vol. XXXVI (March 1998) A. infernal Problems in Explanafiofls "in the beginning there were markets" Based ofl Giuen Indiuiduals (Williamson 1975, p. 20); this starting point is characteristic of his approach. Alexander Field (1979, 1984) bas ad- From this original context, some indi- vanced a key problem in the new insti- viduals go on to create firms and hierar- tutionalist project to attempt to explain chies. These endure if they involve institutions solely in terms of given, ra- lower transaction costs. tional individuals. In trying to explain However, the market itseIf is an insti- the origin of social institutions from tution. The market involves social given individuals, the new institutional norms, customs, instituted exchange re- economics has to presume a socia1 lations, and-sometimes consciously or- framework governing their interaction. ganized-information networks that In any original, hypothetical, "state of themselves have to be expIained (Dosi nature" from which institutions are seen 1988; Hodgson 1988). Market and ex- to have emerged, a number of rules, change relations themselves involve and cultural and social norms are al- complex rules. In particular, the institu- ready presumed. No "thought experi- tion of private property itself requires ment" involving an institution-free explanation. Markets are not an institu- "state of nature" has yet been postu- tion-free beginning. As if in search of lated without them. the original, institution-free, state of For example, game theorists such nature prior to property and markets, as Schotter (1981) take the individual Williamson (1983) argues that private "for granted," as an agent unambigu- property can emerge through "private ously maximizing her expected pay- ordering," that is, individual-to-individ- off. Yet, in attempting to expIain the uaI transactions, without state legisla- origin of institutions through game tion or interference. theory, Field (1984) points out that The question of the possibility of certain norms and ruIes must inevitabIy property and contract without any role be presumed at the start. There can for the state is debated in legal theory. be no games without prior rules, and However, there is another fundamental thus game theory can never explain the objection to the idea of attempting to elemental rules themselves. Even in a ground explanations of property or in- sequence of repeated games, or of stitutions on individuals alone. Even games about other (nested) games, at if the state is absent, individuaIs rely least one game or meta-game, with a on customs, norms, and, most em- structure and payoffs, must be assumed phatically, the institution of language, at the outset. Any such attempt to deal in order to interact. Interpersonal with history in terms of sequentiaI or communication, which is essential to all nested games is thus involved in a stories of institutionaI emergence, itself problem of infinite regress: even depends on linguistic and other rules with games about games about games and norms. The shared concept of indi- to the nth degree there is still at least vidual property requires some means of one preceding game left to be ex- communication using common concepts plained. and norms, both before and after ex- As another example, Williamson’s pIicit or tacit recognition of property transaction cost theory of the firm takes rights can be established. Even if the its originaI state of nature as the mar- state can be absent from these pro- ket. In a famous passage he writes that cesses, some prior institutions are still
Hodgson: Institutional Economics 183 Before an individua1 may tion. Information cannot be received choose, he or she requires a conceptual unless the individua1 has to some de- framework to make sense of the gree been enculturated through prior More generaIly, considering engagement with Com- the action-information loop in Figure 1 munication requires an institutionalized above, the reception of "information" in individual. the new institutionalist story requires a In the "old" institutional economics, paradigm or cognitive frame to process cognition and habit have a central and make sense of that information. place. Knowledge and learning are Further, our interaction with others re- stressed. There is also an insistence that quires the use of the institution of the perception of information is not language. We cannot understand the possible without prior habits of thought world without concepts and we cannot to endow it with meaning. Without such communicate without some form of habits, agents cannot perceive or make Ianguage. Without the prior use of the data received by their senses. institutionalization of individuals, the Habits thus have a cruciaI cognitive action-information loop cannot be com- role: "All facts of observation are neces- pleted. As the "old" institutionalists ar- sarily seen in the light of the observer’s gue, the transmission of information habits of thougbt" (Veblen 1914, p. 53). from institution to individual is impossi- Such habits are acquired through in- ble without a coextensive process of en- volvement in institutions. culturation, in which the individual The central "new" institutionalist pro- learns the meaning and vaIue of the ject of explaining institutions from indi- sense-data that is communicated. The viduals alone is thus misconceived. The "information" arrow on the right hand problem of infinite regress identified side the figure always and necessarily here undermines any "new institutional- involves such a process of encultura- ist" claim that the explanation of the emergence of institutions can start from 19 Whether common agreement to and practical some kind of original, institution-free enforcement of property can arise wlthout the ex- ensembIe of (rational) individuals, in lstence of a state is a matter of continuing debate, The classic "old" institutionalist argument that the which there is supposedly no rule or in- state was required h~storicallyi n order to establish stitution to be the verylabor, land, and mane markets fi In Karl Polanyi minimum, "new" institutionalist stories (1944). Desnite his cerebrated arpument concern- ung the urtdLsugrted evolution of ;oney from irtdi- of the development of institutions de- vidual interactions, even Menger (1936) recog- pend upon interpersonal communica- nized that the state may be necessary to maintain tion of information. And the communi- the integrity of the monetary unit. ’,]This raises a problem in Young-Back Choi’s cation of information itself requires (1993) lrtrtovat~vew ork. While develo ing a theory shared concepts, conventions, rules, of conventions and institutions he taKes the indi- vidual as "the basic unit of analysis" (p. 5). Choi routines, and norms. These, in turn, (pp. 32-39) is thus unpelled to have this ~ndivudual have to be explained. somehow choosing the couzceptual "paradigm" by wh~chs he makes sense of the uncertaun world in which she is situated The unanswered question is l*This raises the question of how a newborn on what basls has such an individual infant may acquire lrtformation. Our ~ncapacity "elected to 39) one paradigm rather than another? to learn without prior conceptual frameworks Surely (t% 1 s choice Itself requires a cortceptual means-as James ar ued lon ago and Plotkin (1994) and others elaborate-tfat much of our in- framework or paradigm, and nor-ms and crlteria of judgement, to make some sense of the s~tuatloi~nl itial capacity to learn must be genetically Inherited which the choice is made The choice of paradigm and instinctive. In the eyes of modern sychology> ~tselfr equlres a paradigm. Again we have a prob- Veblen’s adherence to the concept oflnstinct IS not as old-fashioned as it used to seem. lem of infinite regress.
184 Journal of Literature, Vo L. XXXVI (March 1998) B. Which Came First: The Chicken nisms of reproduction and procreatioil or are very different. Individuals are pur- the Egg? poseful, whereas institutions are not, at What is being contested here is the least in the same sense. Institutions possibility of using given individuals as have different life spans from individu- the institution-free starting point in the als, sometimes enduring the passing of explanation. It is not possible to under- the individuals they contain. Crucially, stand how institutions are constructed each individual is born into, and molded without seeing individuals as embedded by, a world of pre-existing institutions: in a culture made up of many interact- even if these institutions were made by ing institutions. Institutions not only others and can be changed. constrain but also influence individuals. We have seen that the new institu- Accordingly, if there are institutional tionalist project to explain the emer- influences on individuals and their gence of institutions on the basis of goals, then these are worthy of explana- given individuals runs into difficulties, tion. In turn, the explanation of those particularly with regard to the concep- may be in terms of other purposeful in- tualization of the initiaI state from dividuals. But where should the analysis which institutions are supposed to stop? The purposes of an individual emerge. This does not mean that all could be partly explained by relevant in- "new" institutionalist research is with- stitutions, culture, and so on. These, in out value, but it indicates that the start- their turn, would be partly explained in ing point of explanations cannot be in- terms of other individuals. But these in- stitution-free: the main project has to dividual purposes and actions could be reformulated as just a part of a wider then be partly explained by cultural and theoretical analysis of institutions. The institutional factors, and so on, indefi- reformulated project would stress the nitely. evolution of institutions, in part from We are involved in an apparently infi- other institutions, rather than from a nite regress, similar to the puzzle hypothetical, institution-free "state of "which came first, the chicken or the nature." What is required is a theory of egg?" Such an analysis never reaches an process, evolution, and learning, rather end point. It is simply arbitrary to stop than a theory that proceeds from an at one particular stage in the explana- original, institution-free, "state of na- tion and say "it is all reducible to indi- ture" that is both artificial and unten- viduals" just as much as to say it is "all able. social and institutional." The key point C. Toward Euolufionary Explanations is that in this infinite regress, neither of Institutional Change individual nor institutional factors have complete explanatory primacy. The idea In some cases the "comparative stat- that all explanations have ultimately to ics" character of such "new" institution- be in terms of individuals (or institu- alist explanations is obvious. However, tions) alone is thus unfounded. one of the reasons for the rise of "evolu- There is thus an unbreakable circle of tionary" thinking in economics since the determination. This does not mean, earIy 1980s has been an attempt to however, that institutions and individu- break the constraints of this mode of ex- als have equivalent ontological and planation with its two fixed end-points. explanatory status. CIearly, they have Because there is no answer to the different characteristics, Their mecha- chicken-or-egg question, the question
Hodgson: Institutional Economics 185 itself has to be changed. The question encompassing assumption of the "ra- should no longer be "which came first?" tional" agent is deemed to be sufficient. but "what processes explain the devel- The contention bere, however, con- opment of both of them?" This implies cerns the explanatory primacy of lzabit a movement away from comparative over such all-encompassing conception statics and toward a more evolutionary of rational behavior. We start by posing and open-ended framework of analysis. a question: under what circumstances is Some moves in this direction by two it necessary or convenient for an agent prominent "new" institutionalists have to rely on habits or rules?’" The ques- already led to a degree of convergence tions of how habits and rules are repli- with the evolutionary and open-ended cated and transmitted in society is side- ideas of the "old" institutionalists. This stepped here to focus on a decision is apparent in the Iater works of Hayek situation giving rise to their use. It is (1988) and the recent writings of North argued that even optimization requires (1990). the deployment of rules, and for this Such evolutionary explanations in- reason mainstream economics cannot volve the search for "a theory of the legitimately ignore these questions. process of consecutive change, realized This suggests that a detailed analysis of to be self-continuing or self-propagat- the evolution of specific habits and ing and to have no final term" (Veblen rules-including the pecuniary rational- 1919, p. 37). Emphatically, abandoning ity of agents in a market economy- the attempt to explain all institutions in should be instaIled at the core of eco- terms of given individuals does not nomics and social theory. mean the abandonment of tlzeoretical Rules are conditional or uncondi- explanation. Instead, the origins and de- tional patterns of thought or behavior velopment of organizations and institu- which can be adopted either con- tions are seen as an evolutionary pro- sciously or unconsciously by agents. cess. Today, there is a substantid GeneraIly ruIes have the form: in cir- amount of work going on in this area, cumstances X, do Y. Habits may have a with extensive use being made of evolu- different quality: rule-following may be tionary metaphors taken from biology. conscious and deliberative whereas ha- bitual action is characteristically unex- amined. Rules do not necessarily have a V. The Necessity of Habits and Rules self-actuating or autonomic quality but This section extends the argument clearly, by repeated application, a rule further by showing how rational indi- can become a habit. Often it is easier to viduals depend on habits and rules. The break a rule than to change a habit, be- prominent idea of the utility-maximiz- cause our awareness of our own habits ing individual has permitted economists is often incomplete; they have a self-ac- to ignore the procedures and rules that tuating character, being established in are knowingly or unwittingly employed subliminal areas of our nervous system. by agents. Most explanations of behav- However, habits still have the same ior, including the role-driven and the general form: in circumstances X, action habitual, can seemingly be encom- Y follows. passed within the framework of utility- A familiar question of enduring con- maximization. Accordingly, the underly- ing psychological and other explanatory ’"ox a treatment of this issue in greater depth issues have been largely ignored. The see Hodgso11 (1997).
186 Journal of Economic Literature, Vol. XXXVJ (March 1998) troversy is the extent to which optimiz- Conventional accounts sometimes ne- ing techniques are applicable to deci- glect the universal need for rules of cal- sion situations in the real world. Much culation to reach optima. One reason of modern economics is founded on the for this is that optimization processes assumption that they are. If assump- are often confused with optimal out- tions, of perfect information are comes. However, statements of equilib- dropped, it is typicaIly assumed that un- rium conditions are not the same thing certain or complex decision problems as the specification of algorithmic or can still be accommodated using prob- other procedures required to attain abilities. Against this, a number of crit- equilibria: outcome is not the same as ics have argued that a significant pro- process. Another reason for the neglect portion of decision problems are not is the widespread belief that optimiza- amenable to probabilistic or other opti- tion involves choice and rule-folIowing mization techniques (., Veblen 1919; denies it. On the contrary, as argued Keynes 1937; Simon 1957, 1979). above, mechanical optimization ex- cludes genuine choice. A. Optimizatio~ta nd Rules ~lelxp licit optimization procedures must involve rules. This raises the sec- However, we shall side-step this well- ondary but important question of their known controversy here, to focus on a origin. Notably, optimization itself can- (large or smaI1) cIass of decision situ- not provide a complete explanation of ations in which (bounded) optimization either the origin of rules or the adop- may be possible. Consider mathematical tion of rule-driven behavior. As a11 opti- optimization problems and their solu- mization involves intrinsic rules, the tions. The procedures of Iinear pro- idea of explaining all rules on the basis gramming and differential calculus, for of the optimizing behaviors of agents example, sustain methods of optimiza- involves circular reasoning and is thus tion with strict rules. Optimizing proce- misconceived. Hence the question of dures always involve rules: namely the "where do the original rules come rules of computation and from?" remains, and it cannot ha an- ’W otably, Vanber (1994) has suggested that swered completely in terms of opttmiza- rational choice and r3e-following behavior are io- tion itself. It is necessary to consider compatible. He argues that ~t is illhererttly incon- sistent to speak of a "rational chorce to follow additional explanations of their genesis, rules" or a "rational choice among rules." To Van- at least to supplement the optimization berg, the essence of following a rule 1s to be par- story. In search of this "first cause" we tially unresponsive to the changing particularities of each choice situation. This is contrasted with are forced to consider explanations the concept of choice, where an individual is other than optimization for the reIiance deemed free of such "pre-pro rammed behavior." of the individual upon rules. However, first, the quality of$eing unresponsrve to changing particularities is not an universal There is also the case of the intuitive fea- ture of rule-following behavior There are condi- optimizer, with tacit skills. Although the tional rules that discriminate between different skills may not be codifiable, they are environmental conditions and point to different embedded in habits of the same general outcomes in different circumstances. Second, the very idea of rational calculntzon, as shown below, form: in circumstances X, do Y. Like- must itself de end on computational rules. Van- wise, the formation of these habits can- berg has also eiled to recognize that strict optimi- zation must necessarily exclude choice As noted not be explained by optimization alone, above, an ut~llty-maximizeri s essentially a taste- without addressing the other rules, hab- satisfying machine. If choice means the possibility its, or instincts that Ied to their origin. of acting otherwise then it cannot be predeter- mined by either preference functions or rules. This primary reliance on habits or
Hodgson: lastit?~ tionaEl conomics 187 rules limits the scope of rational optimi- involve acquired habits. Habits are used zation. Rationality always depends on even by firms or individuals attempting prior habits or rules as props (Hodgson to optimize in some way. As the "old" 1988). Hence rational optimization institutional economist John Maurice alone can never supply the complete ex- Clark (1918, p. 26) noted, "it is only by pIanation of human behavior and insti- the aid of habit that the marginal utility tutions that some theorists seem to be principle is approximated in real life." striving for. Given that explanation in It is an institutionalist tenet that habit social science requires more than this has ontological and explanatory primacy powerful idea at its core, arguably we over rational choice. Again this implies must rely on more complex, contingent, a greater level of generality for the core and multifaceted behavioral specifica- institutionalist tions. In practice, the human agent cannot As a result, neocIassicaI economics be a "lightning calculator" (Veblen could be regarded as a special and 1919, p. 73), quickly, effortlessly, and (highly) restricted case of the "old" in- inexplicably finding the optimum just as stitutional economics, which accepted we can readily locate the lowest point of the ubiquity of habits and rules. In con- a U-curve in a simple textbook diagram. trast to their image as myopic and anti- Even with given and unambiguous in- theoretical data-gatherers, institutional- formation, complex optimization prob- ists bave the potential to achieve a lems typically involve difficulties not higher level of theoretical generality. only of specification but of comput- Winter (1971) has argued that neoclas- ability. Artificially intelligent systems sical economics is a special case of be- even in moderately complex environ- havioral economics. We may conclude, ments require "inherited" framing pro- further, that both behavioral economics cedures to structure the incoming infor- and neoclassical economics are special mation (Zenon Pylyshyn 1986). cases of institutional economics. At its uolution and the Limits foundations, institutional economics has to Rationality greater generality and encompasses neoclassical economics as a special case. Strikingly, recent developments in B. The Ubiquity of Habits and Rules evolutionary psychology (Cosmides and Tooby 1994a, 1994b; Plotkin 1994; Re- The importance of habits and rules is ber 1993) give strong support to the underIined by consideration of types of "old" institutionalist idea of the primacy decision situation or procedure other of habits. The key argument in this lit- than optimization, such as decision erature is that postulates concerning making in the context of complexity or the rational capacities of the human uncertainty. In particular, Ronald brain must give an explanation of their Heiner (1983) has shown that individu- evolution according to established prin- als are obliged to rely on relatively sim- ciples of evolutionary biology. ple procedures and decision-rules in What may be termed the Principle of such situations. Tbere are long and es- tabIished arguments that individuals 17 Clark (1918, p. 25) also wrote: "a good kedon- must rely on "conventions" or "ruIes of 1st would stop calculatirtg when it seemed likely to i~lvolvem ore trouble than it was worth." Hence thumb" in situations of radical uncer- Simon’s concept of satisficing behavior finds a tainty (Keynes 1937; Simon 1957). clear precedent 111 the work of an "old instltu- Skilled depIoyment of rules must also tional economist.
188 Journal of Economic Literature, Vo L. XXXVI (March 1998) EvoIutionary Explanation demands that as adaptive or (near) optima1 behavior any behavioral assumption in the social differs markedIy from situation to situ- sciences must itself be capable of expla- ation. Second: "Combinatorial explosion nation along (Darwinian) evolutionary paralyzes even moderately domain-gen- lines. 1s However, the empirical and eral systems when encountering real- theoreticaI work of modern evolution- world complexity. As generaIity is in- ary psychologists suggests that, even in creased by adding new dimensions to a highly intelligent organisms, globaI ra- problem space or new branch points to tionality is very unlikely to emerge a decision tree, the computationa1 load increases with catastrophic rapidityM through evolution. In other words, the standard assumption of the rationaI ac- (Cosmides and Tooby 1994a, p. 56). tor fails to satisfy the Principle of Evo- Third, the generality of all-purpose lutionary Explanation. mechanisms undermines their perfor- In an evolutionary view of intelli- mance: "when the environment is gence it is recognized that tacit knowl- clueless, the mechanism will be too. edge and implicit learning of an habit- Domain-specific mechanisms are not uaI character are ubiquitous even in limited in this way. They can be con- higher animals, including humans. This structed to fill in the blanks when per- is because higher levels of deliberation ceptual evidence is lacking or difficult and consciousness are recent arrivals on to obtain" (p. 57). As a result: "The the evolutionary scene, and certainly mind is probably more like a Swiss army came after the development of more ba- knife than an alI-purpose b1ade" (p. 60). sic mechanisms of cognition and learn- In evolutionary terms, time does not ing in organisms. That being the case, "hammer logic into men." Cosmides many of our evolved cognitive processes and Tooby produced evidence that hu- must be able to proceed below the Ievel mans are generally poor at solving gen- of full deliberation and awareness. eral, logical problems. However, when Cosmides and Tooby (1994a) postu- these problems are reformulated in Iate that the mind is riddled with func- terms of social interactions, our abi1ity tionally specific circuits. This contrasts to solve them increases markedly, de- wit11 what they describe as the "Stan- spite the fact that the logical structure dard Social Science Model," where the of the problem has not changed. This is mind harbors general cognitive pro- clear evidence of special design in cesses-such as "reasoning," "induc- brain, rather than ability to solve gen- tion," and "learning"-that are "con- eral logical problems. 19 text-independent," "domain-general," Theories of human behavior must be or "context-free." They show that this consistent with our understanding of abstract and generalist view of the mind evolutionary origins: "The human brain is difficult to reconcile with modern did not fall out of the sky, an inscruta- evoIutionary biology, giving experimen- bIe artefact of unknown origin, and tal evidence to support their argument. there is no longer any sensible reason A key argument is that all-purpose for studying it in ignorance of the optimizing techniques are difficult to causal processes that constructed it" construct and utilize. First, what counts 19 Significantly, Plotk~nh as d~stancede volution- 18 Wlthout glvin ~t a name, Veblen (1934, p ary psychology from enetic reductionism. He ar- 79-80) clearly mase use of this principle, in Ris gues that mtelligent%ehauior ’cannot be reduc- discussions of the origlrts of habits, purposeful be- tively explained by genetrcs or genetics and havior, and so on development’’( Plotkln 1994, p. 176).
Hodgson: Institutional Economics 189 (Cosmides and Tooby 1994a, p. 68). Ap- more widely accepted-and equally plied to economics itself, this principle valid-notion that institutions, know- establishes the untenability of the ingly or unknowingly, are formed and prominent assumption "that rational be- changed by individuals. havior is the state of nature, requiring It has been suggested above that the no explanation" (Cosmides and Tooby breakdown of the microfoundations 1994b, p. 327). If rational behavior is to project provides institutionalism with a be assumed, then its evolution has to be significant entrCe. Its focus on institu- explained. tions as durable and typically self-rein- The reintroduction of the concepts of forcing entities provides a convenient habit and instinct into a theory of hu- micro-macro link. To see the role of the man behavior helps to provide a founda- individual in relation to institutions is to tion upon which a theory of institutions focus on the micro aspect. To take the can be buiIt. We have shown that the institution as a socially constructed in- grounding of such a theory on the idea variant-or emergent property-is a ba- of the given, rationaI individuaI is both sis for consideration of macroeconomic unsuccessful, and untenable in ongoing, dynamics and behavior. Accordingly, evolutionary terms. The introduction of "old" institutionalists and mainstream habit and instinct provides a consis- economists have a great deal to learn tency between the socio-economic and from each other. biotic levels of analysis, and estabIishes However, some would dismiss those an important link between the socio- scholars that reject the rational actor economic and the natural world. The in- paradigm as being out side "economics," stitutionaI action-information loop no and consign them to "sociology." In re- longer hangs in empty space: it has a sponse, these vigilantes of "economic biotic grounding. correctness" have to face two severe This does not mean that explanations problems. First, leading economists of socio-economic phenomena have to such as Smith, Ricardo, Marx, Keynes, be in biological terms. Socio-economic Hayek, Simon, and Coase, all failed to reality has emergent properties that incorporate the standard picture of "ra- defy such a reduction. They are differ- tional economic man" in their writings ent levels of analysis, but, ultimately, or expressed ~rofoundm isgivings about propositions at one level do have to be his behavior. Second, the problem also consistent with those at another. This is has to be faced that much of "sociology" a key reason why economics has to take has now embraced rational choice (for account of evolutionary bioIogy (Hodg- example, James Coleman 1990). son 1993). Among twentieth century economists the prevailing practice has been to re- gard the subject as being defined not as VI. Conclusiofi a study of a reaI object-the economy- By institutions, individuals are not but in terms of a single approach and merely constrained and influenced. set of core assumptions. If the subject is Jointly with our natural environment defined in this way then not much and our biotic inheritance, as social be- theoretical pluralism within economics ings we are constituted by institutions. is possible; we are stuck with a single They are given by history and constitute type of theory or approach. Elsewhere, our socio-economic flesh and blood. however, a science is normally defined This proposition must cohabit with the as the study of a particular aspect of ob-
190 Journal of Economic Literature, Vol. XXXVI (March 1998) jective reality: physics is about the na- and fundamentaI precepts will be the ture and properties of matter and en- principles of evolution themselves, stilI ergy, biology about living things, psy- taking strong inspiration from the meth- chology about the psyche, and so on. It odology and approach of Darwin. We shaIl is on the basis of their wish to study thereby take a huge step back in and understand real world economies that time and visit the evolutionary contro- the "oId" institutionalists can retain versies of the 1890s and early 1900s- claim to the title of being economists. and the intellectual world of Peirce, Tames, VebIen, and Commons-and dis- In fact, the boundary between eco- nomics and sociology that has endured cover that much of what we want to say by the prevaiIing consensus for the last has already been said before. Only the; 60 years or so is now being violated on will we be able to read the works of the both sides. The line of demarcation de- "old" institutionalists and fully appreci- ate their achievement. fined by "the science of rational choice" is thus losing its Iegitimacy, and the REFERENCES most reasonable alternative is to at- ALCHLAN, ARMEN A "Uncertainty, Evolution and tempt once again to redefine economics Economic Theory," J. Polit. Econ., June 1950, 58(2),p p. 21 1-21. as the intellectual discipline concerned ARROW, KENNETH J "Rationality of Self and 0th- with the study of economic systems. In ers in an Econornlc System," 1. Bus., 59(4),O ct. other words, it should be defined, as in 1986, pp S385-99; reprinted rn ROBIN M. other HOCARTHA ND MELVIN W REDER, eds. Ba- sciences, in terms of its object of tional choice: ths contrast Between ecoszomics analysis, rather than by any set of prior and psychology. Chicago and London, U. of tenets. Chicago Press, . 1986, pp. 201-15. BAKERW, AYNEE "The Social Structure of a Na- It may be conjectured that the loss of t~onalS ecurities Market," Amer. J. Sociology,, the reigning twentieth-century bounda- Jan. 1984, 89(4),p p 775-81 1. ries between the social sciences fore- BECKER,G ARYS . "Irrational Behavior and Eco- nomic Theory," J. Polit. Econ. 70(1), Feb. shadows a great centennial climacteric 1962, pp. 1-13. in these intellectual disciplines. For -. -Habits, Addictions, and Tradltions," Kyk- long the butt of the critic from the het- los, 1992, 45(3),p p. 327-45 erodox for tastes. Cambridge, MA: fringes, rational economic man Harvard U. Press, 1996. has increasingly come under challenge BHASKARR,O Y The possibility of naturalism A from the mainstream in recent years, philosophical critique of the contenxpoa y hu- partly because of developments in game man scieszces. Brighton: Harvester, 1979. BOULDING, KENNETHE . Eoolutionat-y economics. theory. It has been suggested by leading Beverly Hills, CA: Sage Publications, 1981. economists such as Frank Hahn (1991) BUCHANANJA, MESM . "Is Economics the Science that one of the central responses to the of Cho~ce?"i n Roads to freedom, Essays in honor of Friedrich A, uon Hayek. Eds.: ERICH developing crisis will be the deconstruc- STREISSLERE T AL. London: Routledge and tion of the rational actor that has long Kegan Paul, 1969, pp. 47-64. CHIAROMONTE, FRANCESCA AND DOSI, ruled economics. Optimizing activity GIOVAMNI. "Heterogeneity, Competitiolz, and will be recognized as no more than a Macroeconomic Dy~zamics,"S tructural Change special case of a larger set of possible and Ecoaomic Dynun~ics,J une 1993, 4(1), pp. modes of behavior, with all of them be- 3963. CHOI,Y OUNG BACK Paradignu adc coauentioas: ing required to render viable explana- Uncertainty, decision making, and en- tions of their origin and evolution. trepreneurship. An11 Arbor. U, of Michlgan Press, 1993. Biology is widely seen as the science CLARKJ, OHNM AURICE".E conomics and Modern of the twenty-first century. Following Psychology" Parts I and TI, J. Polit. Econ , Jan.- this lead-and to conjecture further-in Feb. 1918,26(1-2), pp. 1-30, 136-66, repri~lted in JOHNM AURICEC LARKP reface to social eco- the renewed social sciences the uniting
Hodgson: Institutional Economics 191 nomics. Mew York Augustus Kelley, 1967, pp. Friedrich August Hayek. Vol. I. Ed.: W. W. 92-169. BARTLEYII I Londoli Routledge, 1988. COHEN, JACKA ND STEWART1, .4~ e collapse of HEINERR, ONALDA . "The Orig~no f Predictable chaos; Discouertn simplicity in a compler Behavior," Amer. Econ. Reu , Dec. 1983, 73(4), world London andgNew York: Vikin pp. 560-95. COLEMAN,A MES S Foundations afit%’the- HODGSONG, EOFFREMY . Economics and institu- oty. Camb ridge, MA. Harvard U. Press, 1990. tions: A nmnifesto for n nwdern institutional COMMONS, JOHN R. The le a1 foundutions of economzcs. Cambridge- Polity Press; Philadel- capitnhsm New York Macmi~an1, 924. phia: U, of Pen~zsylvaniaP ress, 1988. -. Institutional economics Its place in politi- - Economics and euolutioa; Bringing cal economy. New York: Macm~llan,1 934. life back into economics. Cambridge: Polity CosMrnEs, LED4 AND Toofly, JOHN. "Beyond Press; Ann Arbor, MI. U of Michrgan Press, Intuitron and Instinct Blindness: Toward an 1993. Evolutionarily Rigorous Cognitive Science," .-"The Ubiquity of Habits and Rules," Cognition, Apr.-June 1994a, 50(1-3), pp. 41- Cnmbridge J Econ., Nov. 1997, 21 (6), pp. 663- 77. 84. -. "Better than Rational: Evolutionary Psy- HODGSON, GEOFFREY M.; SAMUELSW, ARRENJ chology and the Invisible Hand," An~ mn. AND TOOL, MARCR , eds. The Elgar compan- Reu., May 1994b, 84(2),p p. 377-32 on to instztutional and euolutioaary economics. CYERT,R ICHARDM . AND MARCH, JAMESG A Aldershot: Edward Elgar, 1994 hehauwrnl theor of the firn~.E ngelwood Cliffs, HOUTHAKKER, HENDRICKS S. AND TAYLOR, NJ: Prentice-Ha{, 1963 LESTEHD Consumer demand in the Uaited DEGLER,C ARLN . In search of human nature: States, 1929-1 970: Analyses and projections. The decltne and reoiual of Darwiaisn~in Ameri- Cambridge, MA: Harvard U. Press, 1966. can social thou ht Oxford and New York Ox- KEYNESJ, OHNM AYNARD"T. he General Theory ford u Press, 1891. of Employment," Quart ] Econ., Feh. 1937, Dost, GIOVANNI".I nstitutions and Markets in a Sl(l),p 209-23; reprinted in Keynes, John Dynamrc World," Manchester School of Eco- Maynard!’The collected writings nomics and Social Studies, June 1988, 56(2), nard Keynes. Vol. XIV The pp. 119-46. er: Part 11: Defence and deuelopment. Lon- DUESEN~ERRJAYM, ESS . Imm, saving aad the on. Macmrllan, 1973, pp. 109-23. theory of consumer behauior. Cambridge MA, LANGLOIS, RICHARD N , ed. Economics as a pro- Harvard U. Press, 1949. cess: Essays in the new instituttonal aconomics. FIELD. ALEXANDER J. "On the Ex lanation of Cambridge and New York Carnbridge U Press, Rules Using Rational Choice ~ode!s," / Econ, 1986. Issues, Mar. 1979, 13(1),p p. 49-72 MASONR, OGER "Interpersonal Effects on Con- .-"Microeconomics, Norms, and Rational- sumer Demand In Econom~cT heory and Mar- ity," Econ. Deoel Cult. Change, July 1984, keting Thought, 1890-1950," ] Econ Issues, 32(4),p p 683-711. Sept. 1995, 29(3),p p. 871-81 GIDDEN S, he constetution of society. MENGER,C ARL. "On the Origins of Money," Outline of the theory of structurutton Cam- Econ ] , June 1892,2(2), p 239-55 bridge: Polity Press, 1984. . "Celd," in The cal&c;ed works of Carl CODE,D HANANJAKY AND SUNDERS, HYAM"A. [- Menger. Vol. IV. Schrijlen iiber Geldtheorie locative Efficiency of Markets with Zero-lntelli- und Wcihrungspolitik. London. London School ence Traders: Market as a Partial Substitute of Economics, 1936, p 1 116 for Individual Rationality," I, Polit Econ., Feb, MURPHYJ,A MES B "T& ids of Order in Soci- 1993, 101( l), pp 119-37. ety," in Natut-al inza es ia econontic thought: GRANOVETTERM, ARK". Economic Act~ona nd So- Markets read in tooti and claw. Ed.: PHILIP cial Structure: The Problem of Embeddedness," MIROWSKIC. ambridge and Mew York. Cam- Amer J. Sociology, Nov 1985, 91(3], pp. 481- bridge U Press, 1994, pp. 536-82. 510. NELSON,R ICHARDR . AND WINTER,S IDNEYG . GREEN, FRANCIS" The Consumption Function: A An euolut~onary theory of economic change,. Study of a Failure in Positive Economics," in Cambridge, MA: Harvard U Press, 1982 Issues in political ecoraomy, A critical approach NORTH, DOUCLASSC Structure and change zn Eds: FRANCIS GREEN AND PETTER NORE econonttc histoy New York: Norton, 1981. London: Macm~llan,1 979, pp. 33-60. . Institutions, institutional change, and eco- HAHN,F RANKH "The Next Hundred Years," nomic pelfot-maace. Cambridge. Cambridge U. Econ. J. , Jan. 199 1, 101( 4041, pp 47-50 Press, 1990. HAMILTONW, ALTONH "Inst~tution,"in Encyclo- PLOTKINH, ENRYC . Dartotn machines ad the paedia ofthe social sciences Eds.: EDWIN R A natut-e of knowledge: Concerning ahptations, SELIGMANA ND ALVIN JOHNSONN. ew York: instinct, and the evolution of intelligence. Har- Mac~rlillan,1 932, Vol 8, pp 84-89, rnondswortll: Pen uin, 1994 HAYEKF,R IEDRICH A. VON The POLANPIK, ARL. Ta e great transforn~ation. New ert-ot-s of socialesn~. The co York. Rinehart, 1944.
Journal of Economic Literature, Vo l. XXXVI (March 1998) POLANYIM, ICHAEL. The twit clin~ on- TOOL,M ARCR . "Contributions to an Tnstitutlonal don: Routledge and Kegan Paul, 1966 Theory of Price Determination," in Rethinktng POLLAK,R ORERTA . "Habit Formation and Dy- econonzzcs: Markets, technology and economic namic Demand Functions,"] Polit. Econ , July- eoolution. Eds : GEOFFREYM HODGSONA ND Aug. 1970, 78(4), pp. 745-63. ERMESTOS CREPANTIA. ldershot: Edward El- PYLYSWNZ, ENONW ., ed. The Robot’s dilemma; gar, 1991, pp. 19-39. The frame prohlenz in artificial tntelligence. VANBERGV, IKTORJ . Rules and choice tn econom- Norwood, NJ:A blex, 1986. ics. London: Routledge, 1994. POSNERR, ICHARDA . Economic analysis of law. VEBLEN,T HORSTEINB. The theory o the leisure Boston: Little, Brown, 1973. class; An economic study in the euof u tion of in- REBER,A RTHURS . Implicit leaf-ning and tacit stitutiom. New York: Macmillan, 1899. . The instinct ofworkmmhip, and the state knowledge: An essay on the cognitive uncon- scious. Oxford: Oxford U. Press, 1993. of the industrial arts. New York. Augustus Rrzvr, su TURAB".T he Microfoundations Pro- Kelley, 1914 ject in General Equilibrium Theory," Cam- . The place of science in modern ciuilisation bridge] Econ., Aug. 1994, 18(4),p p- 357-77. and other essays. New York: Huebsch, 1919. RUTHERFORDM, ALCOLMC , Institutions in eco- . Essays on our changing order Ed.. LEON nomics: The old and the new kmtitutiomlism. ew York: The Viking Press, Cambridge and New York: Cambridge U. Press, 1934. 1994 WAREC, AROLINE F. AND MEANSG, ARDINERC . SCHOTTERA, he econonzic theory of so- The modern economy in axtion. Mew York: Har- cial institutiom. Cambridge and New York: court Brace, 1936. Cambridge U. Press, 1981. WHITE, HARRISONC . Identity and control: A SIMON, HERBERTA . Models of man: social and structural theory of social action. Princeton. rational; Mathematical essays on ratzonal hu- Princeton U Press, 1992. man behauior in a social setting. New York: WILLIAMSONO,L IVER E. Markets and hierar- Wiley, 1957. chies, analysis and anti-trust in~plications:A - "Rational Decision Making in Business study in the economics of internal organization Organ~zat~onsA,"n wr. Econ. Rev., Sept. 1979, New York. Free Press, 1975. pp. 69(4), 493513, "Credible Commitments: Using Hostages STIGLITZ,J OSEPH E. Whither socialism? Cam- to Support Exchange," Amer. Econ. Reu., Sept. bridge, MA- MRP ress, 1994 1983, 74(3), pp 5 19-40 SUGDENR, ORERT. The economics of ri hts, co- WINTER,S IDNEY G., JR. "Satisficing, Selection, operation, and wel/nre Oxford: Hfackwel~, and the Innovatrng Remnant," Quart. J. Econ., 1986. May 1971, 85(2), pp. 237-61.
