Between Hierarchies and Markets: An Explorations of Some Alternatives to imperfect Institutional Arrangements
Simon Fraser University
Paper presented to the fist annual meeting of the British Columbia Political Science Association, Simon Fraser University 5-6 May 1995.
Contemporary political debates often take the form of a confrontation between the logic of the market and the goals of the administrative state. Concepts like “government failure” now serve as rhetorical counter-arguments to the certainties of yesterday about the need (and the state’s capacity) to remedy “market failures.” At least in North America, this debate has captured the imagination of elites and citizens alike.
Simple dichotomies serve a useful didactic purpose, but they also rule out a number of interesting possibilities. As much as the opposition between market and non-market failures reveals after decades of almost exclusive fixation on the former at the exclusion of the latter,1 it also hides. State institutions have been a favourite instrument for carrying out massive projects since Antiquity, sometimes successfully but sometimes also with disastrous results.2 It would be a mistake, however, to equate all more or less hierarchical or coercive non-market institutions with the institutions of the modern administrative state. Similarly, property rights have taken, or can potentially take, forms that diverge from current practices in large market-based economic systems. And contractual arrangements can take many forms other than those most commonly observed in ordinary market transactions (e.g., voluntary associations, co-operatives, etc.). Finely tuned analytical models can help to reveal these interesting variations and, thereby, diffuse the ideological tensions that a simplistic dichotomy can only accentuate.
Some critics of Public/Rational Choice may be tempted to blame it for the displacement of the image of a benign state actively remedying “market failures” by the image of a bloated and incompetent welfare state. It is indeed true that this approach lends some measure of support to the rhetorical argument that since state institutions are “part of the problem” rather than an element of the solution to contemporary policy dilemmas, the only alternative rests with market-based policies. The popularity of rational choice theory must be understood within the context of the political gains achieved by free-market libertarianism–described by its critics as neo-conservatism or “the new right” — and the policy priorities pursued by governments sympathetic to these ideas. Theodore Lowi has underlined that connection very perceptively in the case of the United States.1 In the same vein, it could be ventured that the relatively more modest advances made in Canada by libertarian ideas, and the markedly more statist character of our political culture, explain, in part, the more limited impact of rational choice on Canadian political science.2 Thus it seems a priori somewhat paradoxical to turn to an approach that justifies the “downsizing” of the state for clues about the role of institutions in political economy issues, and for insight about the complexity of institutional arrangements–but perhaps no more so than exclusive reliance on Rousseauian or social democratic principles for understanding the problems of contemporary liberal democracies! My contention, however, is that something like a conceptual revolution has taken place in the last few years within the rational choice paradigm. This has come as a result of recent attempts to take institutions into account when constructing models of choice under constraints. Of course, some essential aspects the approach, such as a commitment to methodological individualism, remain largely unchanged, even though they may be qualified in some significant respects. In this respect, rational choice continues to be antithetical to radical collectivist political projects or holistic ideological discourse. But the general direction in which researchers are moving has shifted toward more complex situations than the simplistic a-social and a-political world posited by the rational actor model. This shift draws its inspiration from–but also re-fashions or even critically responds to–the intellectual victories achieved in the last ten to fifteen years by more or less “neo-” institutional approaches in political science, sociology and economics.
To recap, this paper aims at two goals: first, to provide an assessment of some of the most salient methodological tendencies in policy analysis and in political science in general. My purpose here is to underline the extent to which rational choice has become something rather different from the image that its critics prefer to attack, i.e., an approach concerned with the actions of individuals pursuing their myopic self-interest and inhabiting and a-political world;1 on the contrary, rational choice now constitutes one of the most interesting current within “the new institutionalism.” The second goal of this paper is to suggest that this current is not only interesting in terms of its methodological sophistication, but also can make a contribution to a concrete political problem. Namely, it points in the direction of a possible way out of the sterile confrontation between the advocates of market-based policies and the defenders of the interventionist administrative state.
In the next section, I present a brief account of this multifaceted rediscovery of institutions. Then I examine the role played by game theoretic models in opening up new perspectives within the rational choice tradition, but with significant potentials for conceptual innovation in other research frameworks. I argue, however, that further progress along that road depends on a more systematic exploration of recent advances in institutional economics. This is not to say that economics has all the answers! In the last section I offer a few reflections on the dilemmas facing political scientists who are, as I am, convinced that rational choice theory has jolted the discipline out of the complacent state in which it had fallen in earlier decades, but who also are sceptical about some of the claims made by the zealots of the new paradigm. Paradoxically, perhaps, I would argue that economics, once more, but this time Austrian economics rather than neo-classical economics, provide some elements of an answer–only some elements, however. The complexity of the political world ultimately must be addressed in its own terms.
I. WE’RE ALL INSTITUTIONALISTS NOW…
While it is probably futile to try to capture all the dynamic elements of a discipline in one single, sweeping generalization, there appears to be a tendency toward the recasting of societal problems in institutional terms. This trend was heralded about ten years ago in a seminal article by James March and Johan Olsen;1 since then the movement in that direction has, if anything, only accelerated–although not always in the same direction as advocated by March and Olsen. Evidence to that effect can be gathered by surveying research priorities and conceptual frameworks in political science, as well as in several other disciplines.
A/ From Post-Behaviouralism to the New Institutionalism
In both Canada and the United States, the study of public policy became more central to the discipline from the 1970s onward.2 While the behavioural revolution had not affected the research agenda of Canadian political scientists as deeply as it had in the United States, the study of electoral behaviour or cultural orientations had nevertheless attracted considerably more interest in the 1960s–indeed well into the 1970s–than the “nuts and bolts” issues typical of the “old” a-theoretical institutionalism. Therefore, theoretical reflections on the origin, nature and effect of institutions had not benefitted from many new insights until relatively recently. (In any case, whatever significant differences there exists between American and Canadian political science, the influence of the former on the latter is undeniable; for better or for worse, American political science defines the overall contours of political science per se, not only in Canada but to a very large extent in the western world as a whole.)
The shift within the discipline toward policy studies had not been motivated originally by a deliberate intent to re-think the role of the state in politics. Rather it was a reflection of the emergence of a number of new or pressing issues, from the civil rights movement in the United States to environmental questions in the whole of North America and beyond.1 But very logically, the study of public policy led to a re-focusing on political institutions. Questions pertaining to political economy invariably are posed in terms of power relations and of conflicting interests within contested institutional contexts. Certainly, there were other reasons than a new policy orientation for the re-discovery of institutions. The internal contradictions and inconclusive results of models designed to explain political behaviour in terms of cultural dispositions2 probably contributed to a refocusing on a more manageable, less systemic or abstract level of analysis.
An important first steps toward that goal were taken by Marxist theorists preoccupied with the problem of the relative autonomy of state institutions in capitalist societies. Although these authors concluded that the state could only enjoy a degree of relative autonomy, they initiated a debate that was soon joined by non-Marxist political theorists who proposed a somewhat wider range of alternative explanations of the centrality of the state in political life.3 Two authors, in particular, played a crucial role in that respect: Eric Nordlinger and Theda Skocpol.1 Space lacks here to discuss the contributions made by these authors and those whom they inspired. There is, however, one point that needs to be underlined: despite their liberal democratic orientation, these authors share with Marxist theorists a concern with macro-structural processes, often associated with a historical perspective.
It is precisely this structural orientation which is largely absent from the rational choice approach to political institutions (and to a lesser extent perhaps, the same could be said of the historical perspective). As I explain below, the first models of the political process which were derived from economics ignored institutions. But this situation has changed very dramatically. The contributions made by rational choice theorists to the study of institutions stand out as some of the most intellectually original and the most relevant to a series of very pressing policy problems, from environmental management to educational reform or constitutional renewal.2
Thanks in large measure to the efforts of Elinor Ostrom and all those whom she influenced through the activities of the Indiana University Workshop in Political Theory and Policy Analysis, rational choice has now become one of the pillars of the new institutionalism. Rather than focusing on structures, institutional rational choice analyzes the influence of norms and conventions on the decisions made by the members of an institution. It posits that bounded rationality rather than utility-maximization calculations guides the actions of political actors.
C/ Parallels Trends in Other Disciplines
Institutionalism is by no means confined to political science. Sociologists have always been concerned with institutions–a term which they sometimes define in a very open ended way–and societal structures. What is worth noting, however, is that in recent years there has been a movement toward the study of meso-level networks and organizational structures.1 The works of Michael Thompson and Aaron Wildavsky on cultural theory must also be cited in this context.2
The prominence achieved in recent years by the new institutionalism in economics, however, bears a more direct relationship to my subject. This phenomenon is rather surprising if one considers that mainstream economics has largely shunted the institutionalist tradition originated by Thorstein Veblen and John R. Commons;3 economists who have chosen to continue along that path have been marginalized. Mainstream economics continues to rely on models of utility maximization which treat institutional factor as being exogenous and thus not requiring any special attention. This is beginning to change, however. In part because it is much more theoretical and more sophisticated from a formal point of view, the “new institutionalism” which has asserted itself during the last ten to fifteen years, building on the seminal work of Ronald Coase in earlier decades, has had a much more noticeable impact on contemporary economic thought.1 To wit, two recent Nobel Prize winners have come from this school of thought: Coase himself and Douglas North. Transaction cost economics, which constitutes one of the most important building blocks of the new institutionalism, opens up interesting perspectives on problems of institutional design, as I explain further below.
If new trends in the social sciences suggest that institutions of all sorts form the background of most social interactions, then perhaps ordered social relations can be created or maintained through a variety of means. Indeed even the seemingly anarchistic order of the market reflects complex historical and legal constraints on the self-serving calculations of homo economicus. But neither the market nor the hierarchical structures of the modern state account for more than a fraction of the actual or potential institutional arrangements that a combination of positive and normative perspectives can bring to light. This is precisely the point I wish to emphasize hereafter.
II. FROM SIMPLE TO COMPLEX MODELS OF CHOICE
A number of factors have contributed to the spectacularly successful development of a rational choice approach to the study of politics. In the early stages of its development, however, rational choice theorists were not motivated by a desire to understand better the place of institutions in political life. It is only during the last decade or so that this situation has changed significantly.
A/ The Simple World of Rational Actors
The pioneers of rational choice intended to show that the basic assumptions of neo-classical economics can explain how individuals choose between different political options. They started from the assumption that there is nothing unique about politics, and the rules of the economic game which reflect the working of a rather loosely structured institution of sort, i.e., the competitive market economy, apply also to apparently more structured and specific settings. (While sociologists may want to describe the market as an institution, orthodox neo-classical models of economic equilibrium, as exemplified by K. Arrow and G. Debreu famous 1954 article, are completely silent about “norms as mechanisms constituting social order; in these models, each agent maximizes exogenously defined goals under the assumption of perfect information.”1) The early works of Anthony Downs exemplified this relative lack of concern for the institutional setting of political events.2 But perhaps the most radical attempt to extend the logic of microeconomic analysis to the whole range of societal interactions can be found in the writings of Gary Becker.3
In spite of this narrow focus, rational actor models4 were not totally irrelevant to the study of institutions. They served as a useful reminder that the members of any organization or societal system respond to incentives which are not always consistent with the formal objectives of the structures within which they act. To ignore this obvious fact is tantamount to reifying organizational goals and values and makes it much harder to identify organizational pathologies like goal displacement due to rent-seeking strategies. At a more macro-level, hypotheses like Arrow’s much debated “impossibility theorem” had the merit, at the very least, of provoking stimulating reflections on the limits of the democratic process.1 However, we must look beyond these somewhat outdated models for insight about contemporary debates about the role of the state in society and the specific policy problems that derive from the crisis of the administrative state.
B/ A Game Theoretic Perspective
Whereas rational actor models, such as were used by Downs or Arrow, only considered the choices made by isolated actors, game theoretical models take into account the complex relations of interdependence that emerge in social situations. Starting from a set of assumptions chosen very parsimoniously, game theory progresses through logical steps toward the analysis of very complex problems like the evolution of moral systems or the origin of institutional norms, as I explain hereafter.
The “games” under consideration can be zero-sum or not and, in the latter case, they can be either cooperative or non-cooperative. At the moment, the games which have generated some of the most interesting theoretical advances are non-cooperative non zero-sum games; perhaps the best known of such games is the prisoners’ dilemma (PD). In the somewhat contrived scenario usually given to illustrate this game situation, the police has arrested two suspects for a minor violation but there are reasons to believe the two were involved in a more serious crime; there is insufficient evidence, however, for pressing charges related to the more serious offense, so the two suspects are told that if one of them confesses to the crime, while the other remains silent, he will go free and his silent accomplice will get at least a ten-year sentence, if both remain silent they get only about 6 months in jail, and if both confess they are likely to get more than 6 months but less than ten years. It is rather easy to show that both prisoners invariably confess, although it would have been better for them to remain silent, which is supposedly what rational actors would do if they tried to maximize their welfare. This seemingly trivial example shows that the calculations performed by the rational players are constrained by the institutional setting in which they find themselves (i.e., the rules of the game1). Moreover, they have to imagine and interpret the possible moves that their alter ego–their accomplice in crime or, for that matter, anyone with whom a player shares a meaningful social attributes–is likely to settle for. Indeed, it is worth noting in passing game theory also supposes that not only do the players know what each other knows, but also that they know that they know, and so on (i.e., the condition of “common knowledge”); now common knowledge, which supposes a high degree of socialization, could not be achieved outside of a complex cultural and institutional setting.2 (Some critics have indeed argued that this condition is too restrictive and should be relaxed.)
In PD games, there exists a strictly dominant strategy (i.e., confess) in the event that the game is played only once or a limited number of times which is known in advance to the players. It is this aspect of the dilemma which first caught the attention of social theorists. For example, Garrett Hardin’s well-known scenario–the so-called “tragedy of the commons”–describes how common pool resources are depleted through over-exploitation because no one finds it profitable to cooperate (i.e., to refrain from taking more than is sustainable) as long as every one else refuses to cooperate. Most commentators argue that this leaves policy-makers with only two options for avoiding the catastrophic end toward which this “tragedy of the commons” seems to run ineluctably.1 (A tragedy of this kind is unfolding off Newfoundland at present…) The choice is between a Hobbesian solution which consists in the transformation of the commons into a publicly owned resource managed by the state, on the one hand, and the parceling of the commonly held resource followed by the creation of property rights. Hardin favours the first alternative and there are indeed many examples of publicly controlled natural resources, in Canada and elsewhere, but there also are historical examples of the second alternative, as well as a substantial literature providing arguments in support of this option.2
What Elinor Ostrom and others who share her outlook have done is to show, both in formal terms and through a meticulous examination of many empirical examples, that there are many ways out of this dilemma. In other words, it is not an either or situation: one can observe a variety of ingenuous institutional innovations that have evolved in many different geographical and cultural contexts to avert the tragic outcome anticipated by Garrett Hardin. That is, in none of theses cases success has been achieved by the creation of discrete property rights or by centralized bureaucratic controls.
At the theoretical level, we know already from the work of Robert Axelrod that in 2-person games iterated over an undetermined length of time cooperation will develop between the two players; it can be demonstrated that they will settle for a “tit-for-tat” strategy whereby they cooperate (e.g., both prisoners choose to remain silent) unless one of them defects, in which case the other player retaliates in the next round by refusing to cooperate.3 N-person games are more complex and the equilibrium strategies which are reached cannot be specified in such simple terms. But Ostrom and her colleagues have also used laboratory experiments to test the hypothesis that rational players have strong incentives to cooperate over time. They have found that this is in fact the case. They summarize their findings in the following terms:
When substantial benefits can be gained by arriving at a joint plan of action for a series of future interactions, individuals may have in their repertoire of heuristics simple sharing rules to propose, backed up by a presumption that others will use something like a measured response. If in addition, individuals have learned how a monitoring and sanctioning system enhances the likelihood that agreements will be sustained, they are capable of setting up their own enforcement mechanism.1
Not only is this true in a formal sense and in the psychology lab, but it is also an empirical fact that local communities have been able to devise complex procedures for managing common pool resources like alpine meadows, irrigation systems, underground water basins, forests and (although with a higher rate of failures) coastal fisheries.2 When one examines how these decentralized institutions are run one is often impressed with their democratic character, even under rather inauspicious historical and political circumstances. The case of the huerta irrigation works in Spain, which are controlled by democratically elected officials according to customs that date back to medieval times, illustrate this point quite remarkably when one contrasts these institutions, first, with the centralized and absolutist Spanish monarchy and then with Franco’s regime.
Mark Sproule-Jones starts from very similar premises, but instead of considering how formal rules and procedures evolve to cope with the management of common-pool resources, he looks at the ways in which both formal and informal rules (what he calls “rules-in-use”) interact, or, as he puts it, become “nested” within each other.1 In addition, Sproule-Jones differentiates between different policies and how these differences are reflected in the application of the rules. The empirical cases he uses concern commercial shipping, pleasure boating, and water quality in Hamilton Harbour. In his conclusion, he explains that he has found that
Public policies become neither the pronouncements of of a government nor the rules promulgated by a government, but rather the end-products of social decision making. There are non-governmental “market-like” processes that result in public policies in the field. The scope of these processes or of more traditionally conceived governmental processes is simply a matter of empirical investigation.2
C/ Transaction Cost Analysis
From game theoretic models we gain new perspectives on the evolution of societal norms through cooperation. These models provide interesting insights on the origins of economic, social and political institutions. However, because they continue to treat institutional rules as a given, they are ill suited to the analysis of the workings of institutional arrangements and cannot significantly contribute to questions concerning institutional reform. The approach to which I turn now, which is derived from institutional economics, integrates institutional realities into its conceptual apparatus to a greater extent.
Initially, transaction cost economics was not developed for the purpose of dealing with the study of political institutions; economic organizations (e.g., large integrated firms) were its main objects of analysis. No real obstacle, however, stand in the way of its application to social and political problems. Indeed one can expect that it will help to pose these problems in a new way. As one of the best known institutional economists, Oliver Williamson, puts it, the “focus [of transaction cost economics] runs the gamut from discrete market exchanges at the one extreme to centralized hierarchical organization at the other, with myriad mixed or intermediate modes filling the range in between.”1
Space is lacking here for a complete and detailed presentation of all the conceptual tools used by transaction cost economics. I wish to zero in on its central core, however.2 First, we need a definition of a “transaction.” According to Williamson,
A transaction occurs when a good or service is transferred across a technologically separable interface. One stage of activity terminate and another begins. With a well-working interface, as with a well-working machine, the transfers occur smoothly. In mechanical systems we look for friction….The economic counterpart of friction is transaction cost: Do the parties to the exchange operate harmoniously, or are there frequent misunderstandings and conflict that lead to delays, breakdowns, and other malfunctions?3
The concept of transaction can be analyzed in terms of several dimensions. From the point of view of political economy and policy analysis, the most significant of these dimension has to do with the uncertainty which surrounds the options facing the decision-makers, and their concomitant pursuit of bounded rationality rather than utility maximization. What renders the basic assumption of neoclassical equilibrium models largely irrelevant to political situations is that they assume perfect or close to perfect information about prices and the consequences of making a particular choice. But in many economic, not to mention political situations, uncertainty prevails; this is precisely what transaction cost economics recognizes as evident. From the standpoint of political behaviour, opportunism is another very relevant dimension of analysis. Opportunism–defined by Williamson as “self-interest seeking with guile,”1 e.g., lying, cheating, and more subtle forms of deceit–results from uncertainty: when the contracting parties have unequal access to information about the consequences of doing things in a certain way, and also in cases where one or both parties is/are unsure about the reliability of the other party, the likelihood that agreement can be reached on a mutually advantageous transaction diminishes. Opportunism takes two forms: ex ante opportunism affects the chance of a satisfactory agreement being reached; ex post opportunism concerns execution problems.
The other dimensions are more technical in nature. Asset specificity describes the extent to which an asset is specialized in the performance of a specific task or dependent on other facilities in the same geographical area, and so on. High asset specificity can induce parties to act opportunistically. Transaction frequency is self-explanatory but is an important dimension: dramatic changes in the frequency of transactions over time can create new costs. Finally, the small number condition refers to the (sometimes very limited) number of potential partners once assets have been dedicated to a set of transactions; this can obviously lead to transaction failure.
In the case of transactions involving public goods, two other dimensions must be taken into account: first, the extent to which users of a service can (or cannot) be excluded from having access to it; second, jointness of consumption refers to the effect that one consumer of the good or user of the service has on the other consumers or users (some goods can easily be jointly consumed, e.g., air, other cannot, e.g., a scarce resource).
Because the combination of all these dimensions increases the likelihood of transaction failure, i.e., the possibility that a pure market transaction will not spontaneously take place, certain institutional mechanisms must be implemented to prevent this eventuality. Hierarchical organizations are one example of such mechanisms, but there are many others–including, of course, unregulated markets in situations where the likelihood of transaction failure is low or moderate. For example, Bryson and Ring mention: direct provision of service by a government agency, regulation, intergovernmental agreement, subsidy, voucher, categorical grant, tax incentive, voluntary arrangements, self-service (this is not an exhaustive list).2 Of course, we encounter here concepts that overlap with those used by the literature on policy instruments.3 But it could be argued that they make more sense within the more coherent framework of transaction cost economics.
Now economists are primarily concerned with the search for efficient governance mechanisms. Political scientists, on the other hand, will want to add, or perhaps even substitute, other criteria. Byrson and Ring, for example, use both efficiency and equity. They apply these criteria to a comparative assessment of what they regard as the six possible alternative open to policy-makers embarking on a reform of the education system: self-service (e.g., home schooling, learning on one’s own, etc.), voluntary arrangements, regulations, markets (i.e., private schools), government services (i.e., the traditional public school system), and new ideas about “open enrollment” with the use of vouchers. One can quibble with the particulars of that list, or with they way in which they rank these alternatives in terms of efficiency (i.e., the degree to which they are efficient in dealing with problems caused by uncertainty, asset specificity, etc.) and equity. But their analysis is exemplary as far the ease with which they use concepts derived from transaction cost economics to identify a very wide range of possibilities for offering education services. With new technologies, some of the options which were ignored in the past for reasons of efficiency (e.g., self-service) might become more attractive in the future.
In the context of a democratic regime, all institutional arrangements are, at least potentially, (re)negotiable. They can be compared to (quasi)contracts between the citizens–often acting through interest groups–and public authorities (or sometimes among the citizens themselves). A reading of the French economist Eric Brousseau’s latest book, L’Economie des contrats, suggests that we have so far paid little attention to a number of ways in which these contracts could be renegotiated in ways that fall somewhere in between pure market transactions and hierarchical organizations (direct government controls).4 Here again, we have to look at a number of dimensions–some of which are too frequently ignored–which are inherent in contractual relations. These relations aim at three complementary objectives: coordination to produce pay-offs, distribution of the pay-offs, and sanctioning procedures to insure compliance. Real or quasi-contracts include clauses covering these three dimensions; in addition, one also finds another type of clause dealing with the duration of the contract. In turn, various mechanisms contribute to the pursuit of these goals: routine procedures often based on customs, conventions, etc.; authoritative allocation; monitoring; and mechanisms generating incentives for preventing opportunistic behaviours. Brousseau’s main objective is to study the ways in which corporations respond to technological challenges and opportunities by moving toward organizational structures that vary in numerous ways along these dimensions. Indeed the private sector is undergoing, as it has become quite obvious, massive transformations; but so is the public sector, and so are the relationships between the two sectors–indeed the dividing line between them is becoming increasingly blurred.
Brousseau’s “grammar of contracts”5 suggests that depending on the way in which the coordinating, distributive, and monitoring functions are structured, institutional arrangements of several types emerge. Of course, we are all familiar with hierarchical public bureaucracies which used centralized authority to coordinate strategic plans, as well as organizational and operational-technical matters, unilateral implementation of guaranties and incentives, the judicial system for monitoring and sanctioning performance, re-distributes the pay-offs according to some pre-determined schedule, and is supposed to provide goods or services for an indefinite period of time. (A hierarchical industrial organization like firms at the beginning of the industrial revolution would use a contracting supervisor rather than the judicial system for monitoring; presumably, the distribution of the pay offs would be far more unequal but would also be based on some ex ante remuneration schedule). By contrast, pure markets transactions are characterized by the use of routine procedure for all types of coordination, provide no guaranties (“buyers beware”), may (or may not) use a third party to monitor performance, distribute pay-offs according to the criterion of marginal utility, and relies on contracts of limited duration (often a one-shot deal). But it is easy to use this analytical grid for describing, or even designing de novo, alternative types of institutional arrangements: self-management, voluntary organizations, joint ventures, and so on. As I already mentioned, accelerating technological change is bound to free up may avenues for exploring these various combinations.
III. RATIONALITY AND POLITICAL REASON
There are, of course, all sorts of objections that can be raised against the use of rational choice models to analyze how political institutions evolve or to propose reforms of their core principles. Some of these objections are based on ontological or epistemological presuppositions that are radically at variance with those that sustain the models I outline. I do not intend to deal with such objections here. To engage in a battle of the paradigms would require much more space than I have at my disposal here. I intend rather to raise a few points that do not necessarily challenge the very foundations of rational choice but force us to think critically about the limits of its reach.
The first objection has to do with the very notion of contract. As Atkinson notes, “It does not make intuitive sense to think that participants in the policy process are constantly contracting for new institutions to produce the policy outcomes they want.”6 To take that idea one step further, one can ask: Can institutional design be perfected to the point where it could become merely a technical skill? Can we locate exactly where on the continuum between centralization and decentralization new institutions can be created and with what consequences? Or, on the contrary, are the values and normative assumptions of the political actors so intrinsically intertwined with their social and political roles that the analytical distinction between the rules of the political game and the game itself becomes difficult to sustain? If the latter is the case, doesn’t it mean that not every social arrangement can be renegotiated at will?
The models presented above suggest that there are several options open to policy-makers and interest groups. It would seem that if some form of agreement can be reached on the goals, the choice of the means would be relatively easy. And agreement on the goals is probably easier to reach when the participants in the debate know that they do not have to choose only between strictly laissez-faire or purely collectivist solutions but can meet somewhere in between. But does not this reassuring conclusion rest on the modernist illusion that everything can be measured and calculated?7 Are not the technically ingenuous methods developed by the authors previously cited for dealing with uncertainty and complexity not themselves too clever by half? In the end, can institutional reformers be confident that they know what they are doing? In particular, would the decentralizing reforms proposed by Ostrom and company produce the promised results or are these reforms not themselves premised on conditions that are the product of unique historical circumstances?
Of course, one might expect that questions of this kind would be raised by critics who are inimical to liberal individualism and argue in favour of some version or other of a communitarian or radical democratic vision. As I indicated, to confront these arguments would be the subject of another paper. But what is worth emphasizing is that critics whose ideological orientation does not place them completely at odds with rational choice have voiced similar objections.
Austrian economics is a school of thought which plays a noted role within the new institutionalism practiced by economists. It is a conceptual orientation which is derived from the seminal works of Carl Menger and Ludwig von Mises, and to which authors like Friedrich Hayek, Israel Kirzner, Ludwig Lachmann or G.L.S Shackle have further contributed in a more or less recent past.8 One can glean from the literature associated with this movement two interdependent arguments that are related to my present concern. Austrian economists are fond of repeating that economic and social institutions are spontaneously evolved.9 They are far too complex to have been voluntarily designed, and when politicians or social activists attempt to redesign them, the results are usually catastrophic because all sorts of unintended and unexpected consequences follow from what was originally thought to be worthwile objectives. This argument can be pushed too far, and Hayek himself, who has been the most vocal defender of this view, is at times ambiguous about the extent to which institutional rules can and should be adapted to changing circumstances.10 One does not also have to draw exactly the same conclusion as “Austrians” do about the virtues of the spontaneous order of the market vs. the supposedly coercive and severely disruptive interventions of the administrative state in economic and social affairs. But as a general principle of prudence, it makes sense to remind would be reformers that they probably know less than they should about the consequences of re-designing seemingly inadequate institutional arrangements. The paradox of both game theoretic and transaction cost models of the type discussed above is that, on the one hand, they provide evidence of the spontaneous emergence of economic and political institutions to resolve coordination problems, but, on the other hand, their sophisticated logical structures provide analysts who have mastered these techniques with what appears to be reliable mathematical tools for predicting the outcomes of alternative designs and thus for lifting the veil of uncertainty which had shrouded spontaneous order processes until now. Austrians see there the example of a dangerous fallacy. They are, as a rule very sceptical about the uses of mathematical modeling and accuse mainstream economics of having become almost totally irrelevant now that it is almost entirely quantitative! Political scientists can only gain from investing more intellectual resources in game theory, but they must also guard themselves against he scientistic temptation of pretending that because they use formal models they have achieved a sure grasp of the complexities of political life. Game theory ought rather to be conceived as an instrument for exploring paradoxes, not for “resolving” them. All the more so because rational choice uses concepts and methods that frame social realities within rather narrow constructs.
I should add that some rational choice theorists are well aware of the need to complement their models with concepts and methods derived from other approaches. Jon Elster, for one, has done much to illustrate the merits of this eclectic methodology.11 I would point out also that some of the most perceptive arguments about the limits of instrumental rationality, and the need to take into account the symbolic dimension of communication to move from an exclusive preoccupation with ways of doing things to a concern with the intrinsic value of rationality itself, can be found in Robert Nozick’s latest book The Nature of Rationality12 (and in choosing Nozick as a critic of instrumental rationality, I am certainly are not stepping outside of the tradition of thought from which rational choice originates.)
As for the Austrian economists, what kind of methodological posture do they propose? Usually implicitly, but sometimes explicitly,13 they recommend a hermeneutical approach. Certainly, the exploration of alternative institutional design must begin with the articulation of bare bone frameworks constructed by means of the models I presented, but it cannot end there. The meaning of the “strategies” or “contracts” revealed by such models can and must be read at several levels. Now, most social scientists are familiar with arguments of this sort but often as a result of having read works in the tradition of critical theory, i.e., a tradition that moves in a direction opposite to Austrian economics since the latter advocates institutional reforms that would restore a completely unhindered market economy, whereas the former pushes toward more collectivist solutions. What the interpretive approach developed by the Austrians suggest is that there is plenty of room for re-interpreting the interpretations! But I am not advocating infinite regress.
In the end, analysis must issue into a political debate in a concrete setting; such a debate ought no doubt be informed by contributions from the social sciences, but it must remain open-ended and will point in the direction of institutional reforms that embody both theoretical knowledge and practical experience.
1For an attempt to rectify this imbalance see Charles Wolf, Jr., Markets Or Governments: Choosing between Imperfect Alternatives (Cambridge, MA: MIT Press, 1991).
2One of the classical analyses of this phenomenon is Karl Wittfogel, Oriental Despotism (New York: Yale University Press, 1957).
1Theodore Lowi, “The State and Political Science,” American Political Science Review 86 (1992), 1-7.
2Even a cursory glance at the table of contents of a few issues of the American Political Science review and of the Canadian Journal of Political Science suffices to get a sense of this difference.
1For an example of this kind of attack, see Jane J. Mansbridge (ed.), Beyond Self-Interest (Chicago: University of Chicago Press, 1990).
1J.G. March and J.P. Olsen, “The New Institutionalism: Organizational Factors in Political Life,” American Political Science Review 78 (1984), 734-49.
2As Michael Atkinson puts it, “[t]he study of public policy has become something of a growth industry in the last twenty years.” M. Atkinson, “Public Policy and the New Institutionalism,” in M. Atkinson (ed.), Governing Canada: Institutions and Public Policy (Toronto: Harcourt Brace Jovanovich, 1993), 17.
1On the insistent demand for relevant research on pressing issues by critics of behaviouralism and its impact on American political science at that time, see David Easton, “The New Revolution in Political Science,” American Political Science Review 68 (1969), 1051-61.
2For an exemplary application of this approach see Richard Simeon and David Elkins, “Provincial Political Cultures in Canada,” Canadian Journal of Political Science 7, no. 3 (1974), 397-437.
3See Nicos Poulantzas, Political Power and Social Class (London: Verso, 1978); Ralph Miliband, The State in Capitalist Society (London: Weidenfeld & Niccolson, 1969), and Class Power and State Power (London: Verso, 1983); and on the special case of the Canadian state, Leo Panitch (ed.), The Canadian State: Political Economy and Political Power (Toronto: University of Toronto Press, 1977).
1See E. Nordlinger, On the Autonomy of the Democratic State (Cambridge, Mass.: Harvard University Press, 1981); P. Evans, D. Rueschmeyer and T. Skocpol (eds.), Bringing the State Back In (London: Cambridge University Press, 1985).
2the emergence of rational choice as a significant aspect of the new institutionalism has been noted by the following authors, among others: Atkinson, “Public Policy and the New Institutionalism,” 27-31; David Brian Robertson, “The Return to History and the New Institutionalism in American Political Science,” Social Science History 17 (1993), 1-36; Elinor Ostrom, “New Horizons in Institutional Analysis,” American Political Science Review 89 (1995), 174-78; and Keith Dowding, “The Compatibility of Behaviouralism, Rational Choice and ‘New Institutionalism’,” Journal of Theoretical Politics 6 (1994), 105-17.
1See, for example, R.S. Burt, Toward a Structural Theory of Action: Network Models of Social Stucture, Perception and Action (New York: Academic Press, 1982); and the following is very often cited in the literature on the sociology of organizations: Walter Powell and Paul J. DiMaggio (eds.), The New Institutionalism in Organization Theory (Chicago: University of Chicago Press, 1991).
2See Michael Thomspon and Aaron Wildavsky, “A Cultural Theory of Information Bias in Organization,” Journal of Management Studies 23 (1986), 273-86; M. Thompson, R. Ellis, and A. Wildasky, Cultural Theory (Boulder, CO: Wetview Press, 1990); and A. Wildavsky, “Why Self-Interest Means Less Outside of a Social Context: Cultural Contributions to a Theory of Rational Choice,” Journal of Theoretical Politics 6 (1994), 131-59; see also Gunnar Grendstad and Per Selle, “Cultural Theory and the New Institutionalism,” Journal of Theoretical Politics 7 (1995), 5-27.
3Although Veblen and Commons are generally regarded as the pillars of the “old institutional economics,” their respective approaches differed in some significant respects, notably the extent to which they accepted (i.e., Commons) or rejected (i.e., Veblen) the tenets of neo-classical equilibrium analysis.
1See Malcom Rutherford, Institutions in Economics: The Old and New Institutionalism (Cambridge: Cambridge University Press, 1994).
1Review of Jon Elster’s Nuts and Bolts for the Social Sciences by Mark Gould in The American Journal of Sociology (1991),
2Downs concedes today, however, that the model he proposed in his An Economic Theory of Democracy was too narrowly focused and one-dimensional; see his “Social Values and Democracy” in K.R. Monroe (ed.), An Economic Approach to Politics (New York: Harper Collins, 1991).
3See, for example, G. Becker The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976).
4Strictly speaking, these are models of expected utility maximization under certainty (e.g., consumer theory) where the conditions of reflexivity, completeness, transitivity and continuity apply. For a technical definition of these terms see, among other texts, Shaun Hargreaves Heap et al., The Theory of Choice: A Practical Guide (Oxford: Blackwell, 1992), 6.
1The theorem was first proposed by Kenneth Arrow in his Social Choice and Individual Values (New Haven: Yale University Press, 1951); it is is too complex and technical to be briefly summarized here but good explanatory accounts can be found in Heap et al., The Theory of Choice, 209-12, or in Robert Abrams “Rationality and Collective Choice Theory” in S.L. Long (ed.), The Handbook of Political Behavior ( ), 230-37; for an analysis of the implications of Arrow’s theorem for the study of political institutions see William H. Riker, Liberalism against Populism (New York: W.H. Freeman, 1982).
1For an analysis of the several dimensions of the rules specifying a game in game theoretic models, see E. Ostrom, R. Gardner and J. Walker, Rules, Games, and Common-Pool Resources (Ann Arbor: University of Michigan Press, 1994), chap. 4.
2Hans Christian Andersen’s charming tale of the naked emperor illustrates beautifully the potentially powerful transformative impact of common knowledge: once the innocent child says out loud what everybody knew already, but without knowing that everyone else knew it too, thereby realizing the condition of common knowledge, the emperor immediately loses face…
1See G. Hardin, “The Tragedy of the Commons,” Science 162 (1968), 1243-48.
2See for example Walter Block (ed.), Economics and the Environmant: A Reconciliation (Vancouver: The Fraser Institute, 1990); numerous references to similar studies are listed in this volume.
3See Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984). Readers who are already familiar with the subtleties of game theory might be tempted to object that it is known that “tit-for-tat” is not always the best strategy in itereated 2-person games, but I have no space here to deal with objections of this type.
1Ostrom, Gardner and Walker, Rules, Games and Common-Pool Resources, 220.
2See ibid., part 3; E. Ostrom, Governing the Commons (New York: Cambridge University Press, 1990), chaps. 2 to 4; Margaret A. McKeans, “Success on the Commons: A Comparative Examination of Institutions for Common Property Resource management,” Journal of Theoretical Politics 4 (July 1992), 247-81; Daniel F. Bromley (ed.), Making the Commons Work: Theory, Practice, and Policy (San Francisco: Institute for Contemporary Studies, 1992); and William Blomquist, Dividing the Waters: Governing Groundwater in Southern California (San Francisco: Institute for Contemporary Studies, 1992).
1M. Sproule-Jones, Governments at Work: Canadian Parliamentary Federalism and its Public Policy Effects (Toronto: University of Toronto Press, 1993).
1O.E. Williamson, The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting (New York: The Free Press, 1985), 16.
2I follow closely here the presentation of the fundamentals of transaction cost economics given by John M. Bryson and Peter Smith Ring in their article “A Transaction-Based Approach to Policy Intervention,” Policy Sciences 23 (1990), 205-29; these authors themselves rely extensively on Williamson’works.
3Williamson, The Economic Institutions of Capitalism, 1-2.
1Bryson and Ring, “A Transaction-Based Approach to Policy Intervention,” 208.
3See, for example, Christopher Hood, The Tools of Government (London: Macmillan, 1983); M.J. Trebilock et al., The Choice of Governing Instruments (Ottawa: Supply and Services, 1982); and Michael Howlett, “Policy Instruments, Policy Styles, and Policy Implementation: National Approaches to Theories of Instrument Choice,” Policy Studies Journal 19 (1991), 1-21.
4See Eric Brousseau, L’Economie des contrats (Paris: Presses Universitaires de France, 1993); this work is aptly summarized and analyzed by RŽjean Landry, see R. Landry, “Rational Choice and Canadian Policy Studies,” in L. Dobuzinskis, D. Laycock, and M. Howlett (eds.), Policy Studies in Canada: The State of the Art (Toronto: University of Toronto Press, forthcoming).
5Brousseau, L’Economie des contrats, 126.
6Atkinson, “Public Policy and the New Institutionalism,” 29.
7Note, however, that neo-classical economists have criticized transaction cost economics precisely because they claim that transaction costs cannot be measured; they are, after all, opportunity costs, not tangible ones.
8There is now a new generation of “Austrian” economists in North America and in Europe; I cannot list all of them here, but certainly Professor Don Lavoie of George Mason University deserves to be cited.
9See, for example, F.A. Hayek, Law, Legislation and Liberty vol. 1 Rules and Order (Chicago: University of Chicago Press, 1973).
10On this point, see L. Dobuzinkis, “The Complexities of Spontaneous Order,” Critical Review 3 (1989), 251-2.
11See, for example, J. Elster, Cement of Society: A Study of Social Order (Cambridge: Cambridge University Press, 1989).
12R. Nozick, The Nature of Rationality (Princeton: Princeton University Press, 1994), chap. 5.
13See D. Lavoie (ed.), Hermeneutics and Economics (London: Routledge, 1991); and L. Lachmann, “Austrian Economics: A Hermeneutic Approach” in D. Lavoie (ed.), Expectations and the Meaning of Institutions: Essays in Economics by Ludwig Lachmann (London: Routledge, 1994).