Hayek Vs Keynes

Jean-Pierre Dupuy
Éole Polytechnique, Paris
& Stanford University

The logic of imitation

   I will concentrate on Friedrich Hayek's notion of imitation which, like the concept of spontaneous social order, links him to the great tradition of the Scottish Enlightenment, the fertile soil from which political economy developed. I propose to compare Hayek and his celebrated adversary, Keynes, on this question of imitation. It is well known that the two economists were on opposite sides of the economic policy debate of the thirties over the respective roles of market dynamics and state intervention in the occurrence of disequilibria. Never to my knowledge has it occurred to anyone to compare their conceptions of the rôle of imitation in market functioning. It is not hard to understand why not. Hayek himself became aware of its importance only gradually, bringing it to the fore in his final book. As for Keynes, his remarks on the relations between imitation and rationality are found in his theory of financial speculation, a chapter of the General Theory that until recently was not taken seriously. The past several years, however, have seen an impressive blossoming of studies more or less directly inspired by Keynes's intuitions. The work of the French school of the "economics of conventions" has a prominent place in this trend, and I will refer in particular to the research of André Orléan.

   What seems to be a fundamental contradiction should strike any reader of Hayek from the outset. All through his writings, the Austrian economist marvels at that wonder of social self-regulation that is the market. It automatically finds the path of its equilibrium, and this equilibrium is an efficient social state. What gives the market its capacity for self-organisation? The answer lies in the negative feedback mechanisms that automatically go into play as soon as an agent departs from equilibrium behaviour. The penalty that he incurs (falling revenues, bankruptcy, etc.) obliges him either to quit the "catallactic" game or to respect its rules. Hayek bases himself precisely on the need to leave these mechanisms freedom of action in order to turn back against the partisans of social justice the accusation of conservatism often made against him. The State that intends, in the name of this ideal, to oppose the sanctions of the market by nibbling away at successes and compensating failures, freezes accumulations of wealth and stabilises differences in income at the same time that it derails the economic engine.

   Simultaneously, Hayek emphasises the rôle of imitation in market competition. Now, it is well known that imitation is eminently productive of positive feedback, a major source of dynamic instability. The vast majority of market theorists are oblivious to imitation. There are profound reasons for this. What is at stake here is both the conception of the modern individual and that of the social order. The independent and self-sufficient individual posited by economic theory is not supposed to be subject to the influence of his peers. The collective phenomena of which the market is the framework are not supposed to have anything in common with the crowd phenomena and the contagion of sentiments and acts of which they are the theatre. And yet, Hayek gives imitation a central rôle. To be sure, he is not alone: Smith and Keynes do the same. The company is not negligible: these are the finest economists of all time who see eye to eye on this point. Nevertheless, the question remains: how can Hayek reconcile his unshakeable faith in the market's capacity for self-regulation and his acute perception of the importance of imitation? Indeed, the problem is much broader since it concerns not only his theory of the market, but also his theory of cultural evolution, which likewise associates competition, imitation and efficiency. In what follows, I will not always distinguish these two levels of analysis. The formal properties are essentially independent of the level. Moreover, for Hayek the market pursues the work of cultural evolution by other means.

   In order to grasp clearly the nature of the problem that imitation poses for any theory of social self-regulation, consider the following elementary model. Two subjects A and B reciprocally imitate each other. The object of their mutual imitation is of indeterminate nature. But suppose that a rumour leads A to believe that B desires (seeks, wants to buy, places confidence in, hopes for, etc.) an object O. A now knows what he needs to desire (respectively: to seek, etc.): he then takes the initiative himself in such a way that his own action brings the object O to B's attention, and when B manifests in turn his interest in O, A has proof his initial hypothesis was correct. His representation, as implausible as it may have been a priori, has been self-realised. This phenomenon where an objectivity or exteriority emerges through the closure upon itself of a system of actions all imitating one another gathers strength as the number of actions rises. The most absurd rumours can polarise a unanimous crowd upon the most unexpected object, everyone finding proof of its value in the eyes or gestures of all the others. The process unfolds in two stages: the first is a mirror game, specular or speculative, in which everyone watches for signs of the coveted knowledge in everyone else, until they all end up being propelled in the same direction; the second stage is the stabilisation of the object that has emerged as the arbitrariness inherent in the conditions of its genesis is forgotten. The unanimity that presided at its birth projects it for a time outside the system of actors, who, all looking in the direction that it indicates, stop searching for clues in one another's gaze.

   This phenomenological description of the world of imitation can be sharpened and confirmed by mathematical modelling. A very active branch of formal economics is today exploring the rôle of what it calls interpersonal influences in economic activity. We know enough about this subject, however, to appreciate how far removed this mimetic universe is from the ideal market. Contrary to what one might have thought a priori, and to what many authors have indeed thought, generalised imitation produces something rather than nothing. It creates self-reinforcing dynamics that converge so resolutely on their target that it is difficult to believe that this convergence is not the manifestation of an underlying necessity, in the manner of a mechanical or thermodynamic system returning invariably to its equilibrium state after straying from it under the effect of some perturbation. Yet one sees that the concept of equilibrium, which the theory of the market imported from rational mechanics, is absolutely unsuited to characterise the "attractors" of mimetic dynamics. Far from expressing an implicit order, they spring from the amplification of an initial disorder, and their appearance of pre-established harmony is a mere effect of unanimous polarisation. They are condensations of order and disorder. The mimetic dynamic seems to be guided by an end that pre-exists it— and that is how it is experienced from the inside— but it is in reality the dynamic itself that brings forth its own end. Perfectly arbitrary and indeterminate a priori, it acquires a quality of self-evidence as the vice of collective opinion tightens. If there is a social process that illustrates to the highest degree the notion of "pure procedure"— even more than do economic exchange or the drawing of lots— it is the mimetic dynamic. There is no other way to determine its result than to let it proceed to its conclusion. It is a random procedure that takes on an aura of necessity.

   In coming to an equilibrium, the economists' ideal market is supposed to reflect an external reality. The prices express objective, "fundamental" values that synthesise information as diverse as the availability of techniques, the scarcity of resources or the preferences of consumers. The mimetic dynamic for its part is completely closed upon itself. The attractors that it generates are not in any relationship of correspondence with an external reality, they simply reflect a condition of internal consistency: the correspondence between a priori beliefs and a posteriori results. The mimetic attractors are self-realising representations.

   Generalised imitation has the power to create worlds that are perfectly disconnected from reality: at once orderly, stable, and totally illusory. It is this "mythopoetic" capacity that makes it so fascinating. If there are hidden truths somewhere to be discovered, one must not count on mimetic dynamics to disclose them. If it is real-world efficiency one is looking for, it is again better not to have to depend on them. Efficiency and the capacity to reveal hidden information: those are two properties that economists readily attribute to the ideal market. The distance between the latter and the mimetic process seems insuperable.

   The clinical picture of the imitative logic is in its essentials already present at the stage of a very simple model in which the mimetic connections between agents are given and remain fixed throughout the whole process: the probability that a given agent imitates another given agent is a constant, possibly null. Phenomenologically, we know that this hypothesis is too restrictive and that the mimetic dynamic has the ability to modify the structure of its own connections: one subject has all the more chances of being imitated by another given subject if he is already imitated by many other subjects. An opinion's power of attraction increases with the number of individuals who share it. One can see that if this is the case, the effects of mimetic polarisation are accentuated accordingly. It may seem, however, that such hypotheses depend too much on the irrationality of crowd phenomena. In fact, research in recent years has shown that they correspond to forms of behaviour that are individually rational. Several possibilities may be envisaged. There are cases where the personal advantage that an individual derives from joining the mass grows objectively with the size of the latter. This hypothesis is today a commonplace in the economic literature that deals with the choice of techniques. As a technique spreads, more is learned about it and it develops and improves; the more users there are, the richer and more diversified the selection of products becomes; production costs diminish, and so does the risk of failure. In these conditions the competition between rival techniques displays features which distinguish it markedly from the "perfect competition" of economists. The first is the multiplicity of "equilibria" (the term is still used by historians of technology but, as we have seen, it is completely improper: it would be better to speak of "attractors"). The "selection" of one among them cannot be determined by deduction from the formal structure of the problem; it is the actual history of events, with its contingencies, fluctuations, and random turns, especially those affecting the system's first steps, which are responsible. One concept plays a crucial rôle here, that of "path-dependence." Fundamentally, it expresses the same idea as that of "pure procedure." We are poles away from Le Chatelier's principle, a thermodynamic reference still popular with theorists of the market who want to laud the latter's capacity to neutralise perturbations which affect it. The evolution of such a dynamic is highly unpredictable. There is obviously no reason for the selection that it accomplishes to be the most efficient one. If a certain technique is favoured by chance at the outset, it will benefit from a "selective advantage" that it will maintain and amplify as the number of users grows. It may end up dominating the market even though another technique would have shown itself to be more advantageous for everyone if only chance had selected it from the start. Technological evolution thus has a strong propensity to get locked into undesirable paths from which it is harder and harder to remove it. Chance, selection, "order through fluctuation," self-organising process: all of these terms used today by historians of technology define a theory of evolution that has only the remotest kinship with neo-Darwinism. The same troubling question thus recurs here in a new form: apart from his ideological biases or his possible lack of knowledge, what right has Hayek to ignore a type of evolutionary process that rests on imitation and mobilises the whole gamut of concepts that he himself champions? Is it simply that he would have been forced to give up the conclusion that cultural evolution has any kind of optimising function?

   The foregoing hypothesis may seem too restrictive. In many cases, it is not true that the objective advantage derived from joining the mass increases with the size of latter. In the case of a culture or tradition in Hayek's sense, for example, one does not see why that should be the case. There exists however a more general reason to go with the majority: uncertainty. If one does not know what is good or what is true, it is rational to imitate others: there is a chance that they know, and by following their lead, one will benefit from their knowledge. Many economic models of what are called "rational expectations" confirm and elaborate this intuition. If the imitated agents really do know what's what, the uninformed agents can accede to their knowledge, even if only indirectly through the intermediary of prices, whose rôle becomes that of disclosing hidden information. The rationality of speculative behaviour— the behaviour of the speculator, but also of those who take him as their model— rests on this mechanism. But what happens if the agents who serve as models are not themselves in possession of the coveted information? Or, a case at once subtler and perhaps more frequent, what if they know without being sure of their knowledge? They will themselves be encouraged to imitate those who imitate them. We have here, at a formal level, a general law of imitative logic: it inevitably leads to these mirror games and other dizzying mises en abyme. What remains to be understood is how or why Hayek seems to avoid them.

   André Orléan has shown that the same models of rational expectations that serve to formalise the progressive unveiling of accurate information, if it exists, and the unanimous convergence of the agents on the optimal behaviour, can just as easily depict an effect of polarisation on any single arbitrary value when the agents, in the grip of uncertainty, seek a remedy in reciprocal imitation. Between the equilibrium and the self-realising representation, there is no formal difference. Yet an abyss separates these two worlds as far as their meaning is concerned.

   It cannot be emphasised too often— and Hayek is the first to do so— that in the face of the social world's complexity imitation is the rational form of access to knowledge. But imitation is also, simultaneously, the source of all illusions— and this is the tragic side of the human social condition that Hayek does not want to see. A very simple model from André Orléan will suffice to illustrate this fundamental dilemma. A key variable of the social system under consideration (for example, the future value of a security) is assigned an objective uncertainty— of the "meteorological" type. In a world where the representations were adequate to the external reality, the diversity of opinions concerning the value of this variable should correspond to the objective distribution of probabilities characterising it. However, each agent is able to base his opinion on two sources of information: his knowledge of the probabilities, but also observation of the distribution of opinions (such as it is reflected, for example, in the current market price of the security). Except in the case where he has absolute confidence in his apprehension of the objective probability, it would be irrational on the agent's part not to take into account the opinion of others: their opinion is based on observations that the agent has not directly made, but from which he may benefit indirectly by re-evaluating the probability (for example, using Bayes's formula) on the basis of existing opinions. The agent's reasons for proceeding in this manner are all the more sound in that he knows what we are in the process of demonstrating: namely that the very process of seeking information about a probability modifies it. If many opinions lean in a certain direction, they will pull the probability in the same direction. A key element in the agent's evaluation procedure is the relative confidence that he puts in his knowledge of objective factors compared to that which he places in the opinion of others. What we have just seen implies that this opinion will carry more weight the closer it is to unanimity. In these conditions, the distribution of opinions evolves in the course of time in recursive fashion: the evolutionary process perpetually feeds off its previous results, its determination is self-referentially closed. Question: can its limiting states be characterised? André Orléan has shown that there exists a critical value of the relative confidence that agents place in their knowledge of the objective probability such that, below this value, the distribution of opinions becomes totally disconnected from that objective probability. Average opinion becomes in a sense the principal cause of itself and it can converge on values that are stable but completely arbitrary. This collective narcissism is a source of aberrant behaviour even though it rests on individual strategies of information seeking and use that are, it should be emphasised, perfectly rational.

   Most of the intuitions that guide these models come from Keynes. In his study of financial speculation, Keynes, like Hayek, sees the fundamental rôle of imitation. In a situation of radical uncertainty, such as the one prevailing in a financial market in crisis, the only rational form of conduct is to imitate others. A first reason brings into play the cognitive mechanisms that André Orléan's model formalises: "Knowing that our own individual judgement is worthless, we endeavour to fall back on the judgement of the rest of the world which is perhaps better informed. That is, we endeavour to conform with the behaviour of the majority or the average." However, Keynes is sensitive to an aspect of imitation as a procedure for the discovery of information that Hayek obstinately ignores: its ambivalence. Take the case of an expert who, for his part, knows a given security's objective worth as a function of the probable dividends it will bring. Say that he observes its going rate to be seriously undervalued. Can he ignore the opinion of the ignorant? No, replies Keynes, for our expert cannot be assured that life's vicissitudes will not one day oblige him to liquidate his portfolio. And in that case it is at the market price that, volens nolens, he will have to do so: "For it is not sensible to pay 25 for an investment of which you believe the prospective yield to justify a value of 30, if you also believe that the market will value it at 20 three months hence." It is too risky to depart from the majority evaluation. As Orléan remarks, "One cannot be right against the crowd." This second reason to resort to imitation puts the expert in the same boat as the ignorant. Note that it brings into play a mimetic rationality based on "objective" considerations and not just "cognitive" ones, to use the distinction we introduced earlier: an individual's interest is better served when he falls into line with the mass.

   Hayek forgets all of this in his purely optimistic vision of the rôle of speculation. The speculator will only fulfil his social function as a seeker after accurate information if he is encouraged to do so. But he is only encouraged to do so if the market adopts, somewhat later of course, the same evaluations that he does. Suppose that the speculator judges a certain piece of information not to be relevant to the estimation of the "underlying" or objective value of a security, but that he thinks the market, for its part, is going to consider it relevant. The information in question will be the basis for speculative behaviour. This is what André Orléan has called the "Reagan effect," with reference to the following anecdote. On a December day in 1987, President Reagan declared that, in his view, the dollar had fallen too low. No broker gave the least credence to the president's economic judgements. And yet, most of them bought dollars as soon as they heard the news. Irrational? No, because they expected that others would do the same, which would push up the price of the currency. Since these expectations were effectively fulfilled, we are justified in speaking of "rational expectations," even though the movement of the market reflected no external reality.

   It follows that the smart speculator is not the one who is the first to discern relevant information concerning the fundamentals of the market. Speculation becomes, in Keynes' words, "the activity of forecasting the psychology of the market." The smart speculator is the one who is able to "guess better than the crowd how the crowd will behave." The speculator is like the snob: he wants to be the leader, the beacon of the masses; he is in their tow. Far from being the model imitated by all, he is the model of all the imitators. Result: the speculative market is like a crowd, Keynes writes, in which each "is endeavouring to copy the others."

   I will now come back to my original question: how can Hayek avoid such cheerless conclusions, given that his social philosophy is based on a philosophy of mind in which imitation plays the starring rôle? Quite simply. Consider a universe where everybody imitates everybody else, with the exception of a single individual who imitates nobody. It is easy to demonstrate that this individual will become the keystone of the system in that everyone will end up imitating him and him alone. Let us make one more assumption: this individual imitates nobody because he knows he is right. Then we have an evolutionary process that acts as a very efficient discoverer and propagator of information. We encounter once more this troubling property of imitation that we have noted so many times, namely its ambivalence. It is efficient if the correct information is present somewhere and recognised and such, but otherwise it becomes a source of illusions and waste. The problem is that it is impossible from inside the system to know in which of the two cases one finds oneself. To overcome this undecidability, it is necessary to resort to an exteriority. When the evolving path arrives at "truth" or "efficiency," a buzzer must go off signalling "look no further"— in other words, "stop imitating." The self-exteriorisation produced by generalised imitation will only manifest its optimising virtue within the framework of a genuine exteriority. Without an authentic transcendence to guide it, self-transcendence is liable to take a wrong turn and get thoroughly lost. If we are talking about the cultural evolution of humanity, the question is obviously what status to give to this transcendence and who may speak in its name. There is no way to avoid prophets here and the foremost prophet, of course, goes by the name of Hayek.

Dissolving contradictions in Absolute knowledge

   The exegetes and critics, even the most favourably disposed among them, have discovered in Hayek's writings what they believe to be major contradictions. We are now in a position to view these in a new light. To tell the truth, the contradictions vanish, but only to make way for what turns out to be a perfectly arbitrary act of faith.

   Certain commentators see a contradiction at the very heart of the theory of cultural evolution. On the one hand, Hayek presents it as a self-organising process, unfolding beyond human consciousness and will; but he also asserts that the mechanism that selects the systems of abstract rules rests mainly on imitation. Does not imitation imply, on the part of the groups that imitate a tradition which originally is not their own, an awareness of the latter's superiority and a desire to adopt it for that reason? Are not this awareness and this desire the condition for the efficiency of the evolutionary process? Hence the supposed contradiction. If such were really Hayek's conception, his inconsistency would indeed be serious. For his entire philosophy of mind is opposed to this interpretation of imitation as the result of a conscious calculation and a deliberate choice. The imitation of rules of conduct is fundamentally blind: "Most people can, after all, recognise and adapt themselves to several different patterns of conduct without being able to explain or describe them." The condition for the efficiency of cultural evolution is therefore not that those who imitate another tradition are aware of what they are doing. It is that those who, by chance, stumble onto the "right" tradition stick with it and stop imitating others. It is thus not those who imitate, but those who do not imitate, who embody the consciousness of evolution. The problem is that this consciousness must come to them from the outside. The tragedy of the West, according to Hayek, is that having discovered the extended order of the market and the liberal principles that govern it, it was unable to recognise their superiority and immediately abandoned them in favour of the constructivist illusions. There is no contradiction; there is rather a very great consistency on Hayek's part in suggesting that the rejection of liberalism and the rise of interventionist governments are an "error" that has spread by contagion. Like Keynes, Hayek recognises here that generalised imitation can be the best or the worst of things. For it to be the best, it needs a guide on the outside to bring it to a halt once it has finally, if blindly, found the truth. One could not imagine a more depressing spectacle for Hayek than that of a progressive Western intellectual (Michel Foucault) seeking salvation from a Middle-Eastern theocracy; there was probably nothing crueller for him than the supposed current abandonment by American universities of the values of the West in favour of third-world or minority cultures... In the face of these renunciations, Hayek stubbornly asserted the necessity of defending liberty dogmatically.

   Nearly all the commentators are alert to what seems to be the major contradiction in Hayek's social philosophy. I mean the status of the demonstration establishing the absolute superiority of the market. This demonstration rests in principle on the theory of cultural evolution. Only the abstract orders that pass through the filter of evolution can lay claim to the loftiest rank, whatever the criterion of the competition: efficiency, justice, liberty, utility, the reproduction and expansion of life, etc. In particular, never could the human mind or reason conceive orders as complex as those selected by evolution. The problem is obviously that Hayek can hardly claim that the market has passed the test since all his work presents itself as a radical and, one wants to say, "rational" critique of modern civilisation, guilty of letting itself be seduced by the sirens of constructivism. His critics therefore conclude that Hayek has to do one of two things. Either he should give up his theory of cultural evolution and found the superiority of the market on rationalist arguments, or else, if he maintains that theory, he needs to admit that the extended order of the market is not the best. This supposed contradiction, or at any rate this tension, colours the entirety of his output and gives it its peculiar tonality, a mélange of traditionalist conservatism and critical radicalism. On one hand, it is asserted that critical rationalism quickly reaches its limits for it can only be exercised within a tradition that remains beyond criticism; on the other hand, condemnations are showered down upon contemporary civilisation in the name of an ideal of liberty which, as Hayek is the first to say, may have existed in thought but never in fact. John Gray rightly notes that the most diametrically opposite ideologies can find support in Hayek's writings: traditionalists who defend the existing social conventions and are prepared to sacrifice individual liberty to the "bourgeois" values, transmitted by the family, of virtue, merit, honesty, morality, work, etc.; but also rationalist libertarians and "anarcho-capitalists" who, not hesitating to wipe the slate clean of established moral values, promote a heroic ethic of individual autonomy and unbridled competition, where only success or failure count in the end, independently of any notion of merit or virtue. Conversely, criticism arrives from every camp. For example, such American neo-conservatives as Daniel Bell or Irving Kristol serve warning on Hayek: the liberal order that he promotes is founded on a moral capital of bourgeois values that its unlimited development tends to destroy irrecoverably.

   Our analysis of the properties of generalised imitation allows us to dissolve this apparent contradiction. There is not, over here, an evolution that unfailingly finds by itself the right path, this right path turning out precisely to be other than that of the market; and, over there, Hayek who asserts the superiority of the market. Without the intervention of a knowledge that transcends it, there is no guarantee that cultural evolution based on imitation will converge towards a satisfactory order, and even less an optimal one. It is in the name of such transcendent knowledge that Hayek speaks, without of course being able to found this knowledge. Hayek was an "engaged" intellectual who wrote books, gave lectures, mobilised his epigones on a world scale and organised them into powerful pressure groups: he sought to influence his contemporaries and, beyond them, the course of evolution. There is nothing here that is not perfectly consistent with a theory of evolution based precisely on reciprocal influences. But he could not, without imposture, speak in the name of evolution. Nor could he, without contradicting his philosophy of mind, speak in the name of mere human reason. There remains, of course, the vantage point of Absolute knowledge. It is understandable that an avowed anti-rationalist like Hayek hesitated to occupy this spot.

From the incentive system to the termite-colony market

    Let us come to the question of social justice. We are now ready to see a contradiction, or at least a tension, in Hayek's celebrated criticism of this notion, which might go unnoticed by anyone who has not followed the road that we have here.

    A first series of arguments advanced by Hayek can be found in Rawls as well. The results of the market are in themselves devoid of any moral value. They are ethically blind. All that is asked of the market is to be efficient. That entails letting it recompense everybody's activity, labour, efforts and strategic choices as it sees fit, if one may put it that way. Now, the remunerations and valuations effected by the market are utterly indifferent to the merit, moral worth, or needs of the agents. Take a hardworking doctor who is deserving and needy, but incompetent: he will be swept away by the competition. Is that unjust? Justice has nothing to do with it. The rules are the same for everyone; the process is anonymous, bereft of intention, without a subject. It is the intervention of a central authority acting as a protection against failure that would be immoral and destructive of freedom. Above all, it would break the engine that makes the market efficient by mucking up its system of penalties and rewards. The market gives to each not what he needs or deserves, but the equivalent of the services that he renders others. It encourages the subjects, even if unbeknownst to them, to step out of their subjectivity and to anticipate others and their needs. That is its own moral value.

    For Hayek these arguments are so many blows struck against the notion of distributive justice. He pretends to believe that a theory of social justice is necessarily meritocratic and requires that everyone receive in accordance with his moral value or, possibly, his needs. The simple existence of the Rawlsian theory, of which Hayek is not unaware, is enough to ruin this postulate.

    In order to contrast it with what is to come, let us emphasise from this first series of arguments the following point: the market establishes a visible link between subjects' actions and the penalties or rewards they receive from it, but this link is not meritocratic. A policy of "social justice" breaks this link and, distorting the system of incentives, destroys the beautiful efficiency of the economic machine. Egalitarianism, for example, renders subjects irresponsible: since they are assured of obtaining the same material advantages and the same esteem whatever they do, they have no incentive to make the effort to adopt the behaviours or to make the choices that are the most rational and, therefore, the most beneficial for everyone.

    In the second series of arguments presented by Hayek, it is not only on the ethical plane that the results of the market are blind, it is also and above all on the cognitive plane. There is a lot of suffering in the market according to Hayek: people can't find work or lose their job, firms go bankrupt, distributors are abandoned by long-time clients, speculators go for broke and lose everything, new products sink without a trace, researchers who toil long and hard discover nothing, etc. Now, these sanctions fall out of the blue like strokes of fate, unjustified, unpredictable, incomprehensible. The argument against a policy of "social justice" is no longer that it unnecessarily upsets the incentive system since there clearly is no possibility of an incentive system here. It is that such a policy can only be blind itself for one does not dictate to a spontaneous order the results it must attain.

    The foregoing argument is obviously founded on the thesis of complexity. Nobody can determine a priori the value of a job, an effort, or a product, for the market alone will decide and its verdicts cannot be anticipated. This argument in terms of "pure procedure" demonstrates how far Hayek is from the neoclassical model of general economic equilibrium. The latter's "non-complex" character is discernible in the fact that it is indifferent whether it is read "bottom-up" or "top down." In the first case, the thesis is that, the agents considering the prices as given and therefore known, their actions are automatically coordinated in an equilibrium, which is a (Pareto) optimum; in the second, the thesis is that any optimum can be reached in decentralised fashion, by communicating to the agents a set of adequate prices. That this is an "anti-Hayekian" world is evident when one recalls the following delectable detail: not so long ago, in what is now a bygone era, one could still hear it affirmed that the best illustration of the Walrasian model was the planned economies of the Eastern bloc. The central planner who communicates to the decentralised agents the values that will allow them to make their theoretically autonomous decisions had to solve, in order to calculate those values, an overall program that gave him, not only the values, but also the decisions of the agents. What hypocrisy not to impose these decisions directly upon them! Hayek could only reject with horror this serfdom disguised as liberty.

    Hayek's theory of market coordination has neither the clarity nor the precision of the Walrasian model, which has by now been totally axiomatised and formalised. Yet one may assert that Hayek's theory refuses to give itself an ultimate level of explanation, whether it be that of individuals or that of the collectivity, because it insists on the recursive codetermination between the two levels. As a consequence, the agents do not have access to the collective |knowledge represented by prices before these are established in the marketplace. Let's take an extreme example: a French worker suddenly learns that the value to the collectivity of his services and qualifications has become null at the moment that he is laid off, some multinational having decided to close the factory that employed him— the reason being that world economic conditions make it more profitable to set up shop in Singapore or Brazil. To be sure, in the world of perfect information that economists are fond of imagining, our man could have anticipated the process of factory migration and either migrated himself or trained for a new job. The complexity of the social process rules this out, Hayek declares. To make himself understood, he often uses terms like chance or good or bad luck rather than complexity. There is, however, a deep kinship between these categories. At this stage, it is not only merit or moral value that may receive blind retribution from the market; it is also effort, talent, skill, or studied strategic choices: no recompense is guaranteed in the face of the vicissitudes and contingencies of social life.

    The right attitude is then— in a sort of Gelassenheit that is to the concern to "repair social injustices" what laissez-faire is to the trial-and-error interventionism of the Welfare State— to abandon oneself to "the blind forces of the social process." To let oneself go in this way is the essence of liberty according to Hayek. It is also the condition for efficiency. The "invisible hand" of the market takes on a very different meaning here from the one it has in the Walrasian model. We are no longer dealing with a spontaneously harmonious composition of behaviours that, even though they do not have the common interest for their object, remain nevertheless, at the individual level, guided by a spirit of calculation and a concern for rationality. What we have instead is the kind of self-organisation found in the termite colony— a termite colony, however, where imitation has supplanted instinct. Individual behaviours that are blind succeed nonetheless in forming an efficient system thanks to a "selection" that eliminates what must be eliminated. But, as we have analysed at length, there is no guarantee that the competitive and mimetic market will conduct itself in reasonable fashion if its self-exteriorisation is not in communication with a true exteriority, that of Absolute knowledge. Yet nothing in what Hayek says, obviously, provides this guarantee.

    The tension between the two series of arguments sometimes reaches a peak, when they are led to meet at the same point. The question of "equal chances" is a fine example. Those who proclaim this slogan have in mind the sporting ideal of justice: may the best contender win! For the competition to be equitable, the "initial" positions must be equalised, for example by a system of handicaps— "setting the counters back to zero," in the elegant expression of French meritocrats. But in the "continuous process" of social life, "this initial position of any person will always be a result of preceding phases," Hayek remarks. If parents knew that their children's start in life would be independent of their own success because a central authority put everybody back on the same level by, among other things, abolishing inheritance, one of the most essential motors of market competition would be broken! Construed in this way, equality of chances also has totalitarian implications. As the enemy of efficiency and liberty, the equalisation of chances is truly a detestable ideal. Conclusion? "We must resolve to consider the starting situation of each individual to be an 'accident'," writes Philippe Nemo, commenting on this point in Hayek— and here is how the latter finishes the sentence we began to quote above: "In a continuous process this initial position of any person will always be a result of preceding phases, and therefore be as much an undesigned fact and dependent on chance as the future development."

    Here the tension is at its peak. On the one hand, as Nemo writes, "these initial positions will be... the fruit of previous efforts and merits, in particular within families." We won't comment on "merit," a slip which reveals that meritocracy dies hard. But on the other hand, one must ask of children to see in their initial position, and thus in the greater or lesser success of their parents— in the degree of effort and sacrifice that the parents invested in the children— a simple "accident," a pure "fluke." That amounts, if not to a logical contradiction, then to an important psychological one. The incentive for parents to save on behalf of their children rests precisely on their hope that the latter will not treat the fruit of their labour as an "unintended fact" or as manna falling arbitrarily from heaven. Here the argument of "social complexity" does not hold: social inheritance or "reproduction" is not realised anonymously at the abstract level of the "great society," but in the most emotionally charged of face-to-face relationships, those in the family.

    How could our author let his theoretical system get into such a fix? The market is in Hayek's view a source of great suffering for those who consent to live by its heroic morality. It is not difficult to understand why. According to the first line of argument, the supreme value of the market, which justifies all the sacrifices, is its efficiency, and this rests on a system of incentives whose principal signals are personal success and failure. To be sure, it is not their "merit" or moral value that the agents put into play, but their ability to be useful to others. Those who fail receive the signal that others do not appreciate their enterprises, their efforts, their talents. They observe that others succeed and are encouraged to follow in their footsteps. The losers imitate the winners because they envy them.

    Why aren't people happy in the modern world? Because they envy one another. This answer which can be found in many analysts of modernity is also Hayek's. It is well however to reformulate this assertion in a more precise and nuanced manner. Envy as a theme is certainly present in the work of our author, but he quite obviously attributes only minor importance to it. Many are the notations that go as follows: some succeed, others fail even though they have expended just as much effort. Why? Perhaps because they are less gifted or less competent (first line of argument), perhaps because they had less luck (second line of argument). Be that as it may, the "envy of those who have tried just as hard [as those who have succeeded], although fully understandable, works against the common interest." Hayek apparently belongs to that large set of authors who recognise the existence of envy, to be sure, but who do not take it very seriously. It is merely a "nasty defect", as popular wisdom would have it, a benign affliction chiefly affecting children. Hayek adds: envy does not make for happiness, but happiness, after all, is but a rationalist philosopher's invention.

    I claim on the contrary that it is the threat that envy poses for his system that leads Hayek to give the latter that bipolar and ultimately contradictory structure that I have analysed. People suffer from envy in a competitive market, where only success and failure count (this is the market as a system of incentives). Failure signifying the incapacity to serve others adequately, it ought inevitably to entail the loss of others' esteem and, as a result, the loss of self-esteem: "It is because there is a relationship between what I do and the esteem in which I am held that it is rational for me to seek to comport myself in the best way possible," writes Nemo. Hayek will seek a remedy for this threat in the most traditional of solutions: the appeal to an exteriority. It is the "complexity" of the social which provides the requisite exteriority— or more precisely, as I have shown, it is the articulation between the self-exteriorising movement of the social and a genuine exteriority, that of Absolute knowledge (which pushes us over into the second conception of the market, that of the "termite colony"). The following quotation from Hayek is eloquent enough: "Inequality is undoubtedly more readily borne, and affects the dignity of the person much less, if it is determined by impersonal forces than when it is due to design. In a competitive society it is no slight to a person, no offence to his dignity, to be told by any particular firm that it has no need for his services or that it cannot offer him a better job. ...[T]he unemployment or the loss of income which will always affect some in any society is certainly less degrading if it is the result of misfortune and not deliberately imposed by authority."

    The problem is that this solution can only work on condition that everyone shares the philosopher's conviction, namely that they are dealing with a true exteriority, which is very doubtful. Nobody will abandon himself to these "impersonal forces" if he has good reason to doubt that they will take the world in the right direction, or simply in a viable direction. But the "extended order of the market" can easily get stuck in a dead end or even plunge into the abyss, as we have shown theoretically and as is proved by historical experience. Obviously nobody can believe in the guarantee that Hayek's system requires in order that such misadventures be impossible: access to Absolute knowledge.

Market and tradition

    To close this critical analysis, I would like to return to the question of tradition and to a closely related one, that of imitation. By placing imitation at the heart of his philosophy of mind, Hayek is led to make it play a central rôle, both in his theory of tradition, which is not unusual, and his theory of the market, which is much more so. The strength of his social philosophy is revealed here, but also its blindness. To put it somewhat schematically, Hayek's reconstruction of the cultural history of humanity falls into three stages. In the primitive phase, an instinctual morality is made up of solidarity and altruism, and animism and anthropomorphism dominate the conception of the world: every form of order is attributed to a design, a will. Then comes the invention of tradition and of the many human cultures. Spontaneous social orders deploy themselves and compete with one another; cultural evolution is off and running. It discovers that peculiar tradition constituted by the extended order of the market, founded on the abstract rules of liberalism. But, far from recognising its superiority, it abandons it in favour of constructivist and rationalist illusions: that is the modern phase, which represents a return to primitive anthropomorphism and a rejection of tradition.

    Not only does this scenario involve certain naiveties (the irenic conception of primitive society); it is highly implausible and threatens the coherence of the evolutionist theory of culture. Hayek is certainly right to characterise modernity by the rejection of tradition. That is not very original, although one may dispute the notion that Descartes was responsible for this radical change with regard to tradition. But this is not the essential point. To restore coherence and plausibility to his scenario, it would be enough for Hayek to make two hypotheses: first, that the expansion of the market order is not unrelated to the break with tradition, since the market is anything but a tradition; second, that rationalist constructivism itself has profound affinities with the market. Hayek sometimes seems inclined to admit the second hypothesis— without being able to go very far in this direction, which makes the justification of liberalism in the terms of the theory of evolution extremely problematic. The idea is that the extension of the market to every realm of personal and social life generates an engineer's mentality, a spirit of calculation that tends to erode what are known as "traditional" values. This criticism of the market, which one is not surprised to find in the neoconservative writings of Irving Kristol and Daniel Bell, sometimes meets with Hayek's approval. For example, The Road to Serfdom tells us that the material advances made possible by liberalism have given men a "new sense of power over their own fate, the belief in the unbounded possibilities of improving their own lot" and that as a consequence "the very success of liberalism became the cause of its decline."

    Hayek is the furthest from the first hypothesis. It is essential to his system that the market be a tradition, springing from cultural evolution. This is to remain blind to the radical break separating the market order from the traditional order where imitation is concerned. In the traditional order, the models are so distant from those who imitate them that no rivalry can arise between them; in market society, on the contrary, the models are equals and competitors: one copies one's more fortunate rival and one is in rivalry with one's model. Here Hayek does not measure up to his own model, Adam Smith. He does not see what the latter demonstrated: that this new regime of imitation gives rise to envy, and that this envy has highly ambivalent effects: it is on the one hand a cause of great suffering (the "corruption of moral sentiments"), but on the other hand, it gives market society an unprecedented energy, a potentially unlimited force of expansion. Moreover, far from being a product of the "constructivist illusion," the idea of social justice is inherent in the propensity of the modern individual always to put himself in the place of the other, to live permanently under the gaze of the other.

    It is painful to have to imitate a rival who is successful. It means openly recognising what one would prefer to hide, from one's own eyes and those of the world: that one holds the other to be superior. In the world of business, and in the economy in general, it is nonetheless rational to act in this manner, as Keynes well saw, and Hayek himself in a lesser measure. Post-war world economic history provides a hundred examples of imitators become innovators in their turn. What is startling about the economy is that this truth is to be found in the public square. The reason for this doubtless lies in the market's remarkable capacity to transform envy into a source of creative energy, instead of letting it sink into resentment and impotent hatred. The same cannot be said for other areas of social life. Let us turn to those who are Hayek's bêtes noires: the "intellectuals" (progressives, it goes without saying). In the area of culture, for example, the credo is: only what is new is interesting. Radical innovation, understood as creation ex nihilo, surpassing oneself rather than others, has the dual advantage of denying both imitation and competition. It takes a lot of naivete for Hayek to think that these intellectuals are nostalgic for the altruism and solidarity proper to primitive society. The rejection of competition typical of these "critical demystifiers" springs from an excess, and not a lack, of competitive spirit.

    The ambivalence of envy also manifests itself within the world of the economy. Despite the headlong flight forward that characterises it at every moment, it is no stranger to the torments of failure and resentment. Hayek at least deserves credit for recognising this, even if he doesn't see that the demand for social justice springs first from this suffering and not from what he terms the principal obstacle to the market: critical rationalism. We have studied the solution that he proposes: it is what is responsible for the tension that runs through his work, between an incentive-system market and a termite colony market. Since competitive imitation is the source of the evils, let's return to the traditional modality of imitation; let's make a tradition out of the market. The imitation of a transcendent model, in the manner of that of Jesus Christ, keeps people from taking each other as models and sinking into the torments of envy. The transcendence here is however only a self-transcendence that, in order to aspire to be a trustworthy guide, requires access to that genuine transcendence that is Absolute knowledge. Hayek's solution can only work if liberalism is made into a religion.

    What has just been said about the market applies in the same manner to cultural evolution. Hayek wants to see in it nothing but a peaceable contest among traditions so that the best may win. It would be more consistent with his theory to present it as a war of religions. Hayek's philosophy of knowledge entails that people become attached to their tradition for reasons other than the benefits that they derive from it— benefits which they are incapable of measuring exactly. There is therefore a misrecognition on the part of the agents, that is the individual counterpart of the opacity and complexity of the collective. Hence the problem of the stability of tradition. It is rational for the agents to conform to tradition, but they do not know that. Only irrational motives can impel them to do so. For Hayek, it is religious, magical or mythical beliefs that fill this rôle. The animism and anthropomorphism of primitive thought, so criticised elsewhere, are at the source: what is the meaning of "God" for a given society, asks Hayek, more Durkheimian than ever, if not the self-transcendence of the social that is not seen as such? In short, the rôle of religion in evolution is fundamental. The necessity for a dogmatic attachment to the principles of liberalism is but a particular instance. "The only religions that have survived," Hayek notes elsewhere, "are those which support property and the family."

    But at the same time, as we saw, evolution can only fulfil its selective function effectively insofar as individuals and groups are ever willing to abandon one tradition for another, by siding with whichever ones muster the greatest numbers. Like the generalised market, which extends its action ten times more effectively, cultural evolution sets up a competition among ways of being, thinking and acting that are mutually fascinated by one another even while pretending to struggle jealously, "dogmatically," to maintain their identity. One should therefore find at its level that same state of being torn between self and other that Smith called "sympathy" at the level of individuals. And that is indeed what one observes. A dominant culture that hates itself, and others, fascinated by it, that hate it while imitating it: there is no shortage of contemporary examples, which we could develop if we wished. Thus, as will have been understood, our criticism of Hayek does not bear on this point: to have placed competitive imitation at the heart of his system is his stroke of genius. The problem is that, a prisoner of his resolutely optimistic ideology, he remains blind to the irreducible ambivalence of this conflictual mimesis. Unless one gives the goddess Evolution a rôle equivalent to the one he attributes to the god of the Market, there is no guarantee, if the very idea has a meaning, that human history, with its sound and fury, will go spontaneously in the "right direction."

    Hayek's œuvre doubtless constitutes the most grandiose attempt to found a theory of the good and just society on the contingence of human affairs and on social complexity. Its failure weighs heavily.

See the model presented by André Orléan in "Money and Mimetic Speculation," in P. Dumouchel (ed.), Violence and Truth, Stanford University Press, 1988. Under certain conditions, it is demonstrated that the imitative dynamic converges toward unanimity of the group. These conditions reflect the fact that there is an effective interdependence among all the agents; in other words, very few probabilities pij are null, pij being the probability that the agent i imitates the agent j.
Cf. W.B. Arthur, "Competing Technologies: An Overview," in R. Dosi et al. (ed.), Technical Change and Economic Theory, London, Pinter Publishers.
A general demonstration of this result can be found in S. Grossman, "On the Efficiency of Competitive Stock Markets where Traders Have Diverse Information," The Journal of Finance, vol. 21, no. 2, May 1976.
Cf. André Orléan, "Mimétisme et anticipations rationnelles: une perspective keynésienne," Recherches économiques de Louvain, vol. 52, no. 1, 1986.
André Orléan, "Le rôle des influences interpersonnelles dans la détermination des cours boursiers," Revue économique, no. 5, September 1990. Cf. J.M. Keynes, The General Theory of Employment, Interest and Money, London, Macmillan, 1936.
J.M. Keynes, "The General Theory of Employment," Quarterly Journal of Economics, vol. 51, no. 2, February 1937, p. 214.
General Theory, op. cit., p. 155.
A. Orléan, "Mimétisme et anticipations rationnelles: une perspective keynésienne," loc. cit.
A. Orléan, "L'autoréférence dans la théorie keynésienne de la spéculation," Cahiers d'économie politique, no. 14-15, 1988.
General Theory, op. cit., pp. 157-158.
J.M. Keynes, "The General Theory of Employment," loc. cit.
This property is a trivial consequence of the theorem demonstrated by A. Orléan in "Money and Mimetic Speculation," loc. cit.
Cf., for example, Bernard Manin, "F.A. Hayek et la question du libéralisme," Revue française de science politique, vol. 33, no. 1, February 1983.
F. Hayek, The Fatal Conceit, op. cit., p. 78.
See for example F. Hayek, The Road to Serfdom, Chicago, Chicago University Press, 1944. von Hayek," loc. cit.
Cf., for example, I. Kristol, "Capitalism, Socialism and Nihilism," in Two Cheers for Capitalism, New York, Basic Books, 1978, chap. 7; quoted by J. Gray, Hayek on Liberty, op. cit.
In Law, Legislation and Liberty, vol. 2, op. cit., pp. xiii and 100, Hayek says he is in basic agreement with Rawls. In The Fatal Conceit, his position has changed radically: on p. 74, he is able to write that a "Rawlsian world could never have become civilised" because of the difference principle's "repressing differentiation due to luck" (Hayek thus falls into the same misreading as Nozick). This new attitude is less surprising than the first one, but the wavering thus made manifest reveals that the Rawlsian theory poses a problem for Hayek.
In fact, imitation plays a rôle in the self-constitution of the termite colony, but the process is much more mechanical and therefore less reflexive than in a human society. See J.-L. Deneubourg, "Application de l'ordre par fluctuation à la description de certaines étapes de la construction du nid chez les termites," Insectes sociaux, vol. 24, no. 2, 1977; cited by Ilya Prigogine and Isabelle Stengers, Order out of Chaos, New York, Bantam, 1984, p. 181.
F. Hayek, Law, Legislation and Liberty, vol. 2, op. cit., p. 130; cf. P. Nemo, La Société de droit selon F.A. Hayek, op. cit., p. 239.
P. Nemo, La Société de droit, op. cit., p. 239.
F. Hayek, Law, Legislation and Liberty, vol. 2, op. cit., p. 130 (emphasis added).
P. Nemo, La société de droit, op. cit., p. 239 (emphasis added).
F. Hayek, The Fatal Conceit, op. cit., p. 118.
The opposition that I employ here between envy as the apparent theme of a work, and the structuring but hidden effect of envy on a work, is analogous to the distinction made by René Girard between the victimage mechanism as theme and as structure. See for example his book The Scapegoat, Baltimore, Johns Hopkins University Press, 1986.
P. Nemo, La Société de droit, op. cit., p. 304.
F. Hayek, The Road to Serfdom, op. cit., p. 106; cf. P. Nemo, op. cit., p. 272.
F. Hayek, Ibid., pp. 17, 19.
This distinction corresponds to the one that René Girard establishes between "external mediation" and "internal mediation." See Deceit, Desire, and the Novel, Baltimore, Johns Hopkins University Press, 1965.
With Paul Dumouchel, I analysed what gives the market economy this singular capacity. See our Enfer des choses, Paris, Seuil, 1979. See also René Girard, "Innovation and Repetition," Substance, 62/63, 1990.
F. Hayek, The Fatal Conceit, op. cit., p. 136.
Ibid., pp. 139-140.
Ibid., p. 137.

Formatted in HTML by Peter Y. Chou,