Approaches to Institutional Analysis

12.3.1. The ‘new political economy’ and surroundings

Under the term ‘new political economy’ it is usual to include a mixed group of areas of study, from public choice to new institutional economics and from behavioural economics to the economics of property rights.

Conventional theory tries to explain the choices of economic agents, their interaction, and the ensuing aggregate results under a double order of assumptions. The first is that the ends and motivations of human actions are exogenous and given a priori and take the form of a function to be maximized under constraints. The second concerns the legal-institutional structure in which individuals make their choices. The basic hypothesis here is that even such a structure is exogenous, that is, a datum which conditions the choices but is not conditioned by them. It is true that some variants have been formulated which loosen the rigour of these hypotheses. In ‘search theory’, for example, the assumption that the set of alternatives is given a priori is replaced by one for which new alternatives can be generated by a search process, the cost of which is, however, known a priori. In still other variants, it is assumed that the consequences of the alternatives delimiting the range of choices are not known with certainty; yet, the decision-maker possesses a joint probability distribution of the results, so that his problem becomes one of maximizing expected utilities. These, however, are attenua­tions which do not modify the nature of the basic hypotheses about agents’ behaviour.

The new political economy endeavours to study the properties of altern­ative legal-institutional configurations.

The new political economy can be seen as a resumption, in modern guise, of an old Smithian project, for which the analysis of market functioning was only a necessary stage towards a much more general goal: to demonstrate that market efficiency constitutes a normative argument in favour of a given institutional structure. According to the interpretation of James Buchanan, Gordon Tullock, and Friederich von Hayek, the main exponents of this stream of thought, Adam Smith was basically concerned with comparing different institutional structures. His proposal of a ‘minimal State’ emerged as a solution from the comparison of the advantages and disadvantages of each alternative. Thus, according to these scholars, the hegemony of neo­classicism is responsible for a discontinuity produced in economic science. The creation of welfare economics as a branch of study with a certain autonomy meant that the economic study of institutions was relegated to that field, in which it was undertaken, not in terms of comparative analysis, but in terms of efficiency. So, even the normative reaction against excessive diffusion of laissez faire was carried out in terms of ‘market failures’ rather than of comparison of different institutions.

12.3.2. Contractarian neo-institutionalism

The foundation of the Journal of Law and Economics in 1958 sanctioned the birth of a fruitful association between the faculty of law and the faculty of economics at the University of Chicago. It was the beginning of American neo­institutionalism. The new stream of thought was stimulated by a simple observation: that relationships among economic agents, in modern capitalist societies, are regulated by an interweaving of institutional mechanisms that are much more complex and articulated than those considered by the tradi­tional model of perfect competition. Society is controlled by sophisticated legal systems which give rise to property rights; to criteria for the allocation of common-ownership goods and the provision of public goods; and to long-term contractual relationships capable of encouraging the maintenance through time of monopolistic or collusive structures. The aim of this branch of research was to analyse such a dense intermeshing of institutional facts, to study its conditions of efficiency, and to give a microeconomic justification to it.

On the basis of these objectives, contractarian neo-institutionalism developed two approaches: ‘process oriented’ and ‘end-state’. The former claims that in expressing valuations, attention should be paid to the delib­erative process and not to the end result. Correct application of a procedure is considered sufficient to achieve a correct result. This is tantamount to saying that the process justifies the result and not vice versa, as it is the case with the end-state approach. Many ‘minimal State’ theorists took this stance, the most famous of whom was American philosopher Robert Nozick. In Anarchy, State and Utopia, of 1974, Nozick held that the market is the sole justifiable mechanism for allocating resources because the market alone is compatible with protecting negative freedom, i.e. freedom intended as absence of restraints. On the other hand, negative freedom is the only con­ception of freedom that liberal society can and must uphold.

Nozick used the following example to illustrate his views in debates on the economic theories of justice. If a vast number of individuals decide of their own free will to pay a given amount of money to see an artist’s show, nobody can rightly object to the ensuing distribution of income, even if the result is that the artist becomes immensely rich. This is a modern variation of Locke’s celebrated claim that ‘by agreeing to use money’ individuals end up by agreeing to any ensuing distribution of income. Clearly, such a position is anything but sound from a philosophical viewpoint, for even if market agents cannot be held responsible for market results, this does not mean that they cannot be held responsible for not remedying the situation. Nozick’s example merely shows that some subjects prefer to pay a given price rather than miss the show. But it would be a non sequitur to assert that they prefer the ensuing distribution of income; and this for the obvious reason that the latter is an option that is not part of the individuals’ choice set. Since one is concerned here with an aggregate result, a given income distribution can be ‘chosen’ only through a collective decision.

The second approach (end-state) developed within the philosophical perspective of neo-contractualism, which is associated, as recalled in chapter 10, with the pioneering contribution made by John Rawls in 1971. Here the main idea is that it is the final outcomes of allocative mechanisms, rather than the procedures, which must be put to trial on the basis of pre­established evaluation criteria.

The aim of the end-state approach is to explain the institutional set-up of a society by means of a model which justifies the constitution and its operation, not only in terms of economic efficiency, but also in terms of a consensus based on individual rationality.

Another important result of the end-state approach was the introduction into economics of problems that neoclassical economists had dismissed from their field of study. Institutions such as the moral rules of social life, long­term contracts, authority relationships and reputation had long been con­fined to the research fields of other disciplines, moral philosophy, sociology, law, and political science. Neo-institutionalism helped to put these questions back on the economists’ agenda.

When facing a situation of ‘market failure’, the contractarian attempts a solution by assuming that agents are able to organize their own social life according to a conscious plan. The ‘planners’ have the job of ‘redesigning’ society and its institutions in such a way that all the actions are directed towards known ends.

In this context important contributions have been made by L. Hurwicz in The Design of Mechanisms for Resource Allocations and J. Geanakoplos and H. Polemarchakis in Existence, Regularity and Constrained Suboptimality of Competitive Allocations when Asset Market is Incomplete. The interesting ‘discovery’ made in the latter is that in economies with incomplete markets the general equilibrium is as a rule not only suboptimal but also constrained suboptimal. This implies that a ‘planner’ is able to ‘do better’ than the market. That an omniscient planner can ‘do better’ than the market, when the latter is incomplete, had already been known for some time. But the new point of view suggests that in economies where the absence of markets is the cause of serious inefficiencies, the State can make good the deficiency by creating an economic structure that is less incomplete.

It does not of course follow that wherever laisser faire falls short governmental interference is expedient; since the inevitable drawbacks and disadvantages of the latter may, in any particular case, be worse than the shortcomings of private enterprise. (1883, p. 414).

The school’s declared intention is to demonstrate that while the invisible hand of the market fails to transform private vices into public virtues, the visible hand of the State does not necessarily succeed in transforming the government’s vices into public virtues.

Now, what are these ‘government vices’? There are several. One is bureaucracy: the bureaucrat, whom we can never do without, is interposed between the citizen and the ruler, and succeeds in advoking part of the surplus to himself. Another is rent-seeking costs which, since they are sustained to maintain rent positions, are not productive. Then there are the so-called internalities: the punishment and reward structures within non­market organizations enter the government agency’s utility function gener­ating distortive effects on resource allocation. Lastly, there is the essentially problematic nature of each collective choice rule: just as the various impossibility theorems demonstrated, any collective decision mechanism is either efficient but dictatorial or democratic but inefficient.

The general framework of the Public Choice school is in line with the European (Continental) tradition of public finance of the last decade of the nineteenth century, a tradition which numbered among its most important members the Italian economists Pantaleoni, De Viti de Marco, Mazzola, and Montemartini and the Swedes Wicksell and Lindahl. In Limits of Liberty: Between Anarchy and Leviathan (1975), one of his most wide-ranging works, Buchanan studied the economic organization in a society of free individuals. His aim was to arrive at an economic constitution based on individualistic principles, where the ‘constitution’ is a set of rules agreed upon beforehand and according to which all the actions in the post-constitutional phase would be undertaken. Buchanan’s economic constitutionalism is contractarian in the sense that the rules on which it is based presuppose consensus. It is important to note that, from the constitutionalist point of view, welfare economics should be (in the words of Schotter, The Economic Theory of Social Institutions, p. 5) ‘the study of the welfare aspects of comparative social institutions’, and not the discipline which studies the conditions for the optimal allocation of resources in a given institutional setting.

12.3.3. Utilitarian neo-institutionalism

The version of neo-institutionalism which we call ‘utilitarian’ is substantially different from the contractarian one. It is still firmly linked to the utilitarian philosophical approach, even though it has rid itself of much of the ingenuousness in J. Bentham’s original formulation. Douglas North initiated this theoretical project, to which he assigned the task of correcting and extending the explanatory capacity of neoclassical theory to explain the need of economic institutions. In what way can this theoretical system be said to move within the neoclassical domain? North replies to the query in ‘Institutions and Economic Theory’, where he says that the approach

begins with the scarcity hence competition postulate; views economics as a theory of choice subject to constraints; employs price theory as an essential part of the analysis of institutions; and sees changes in relative prices as a major force inducing change in institutions (p. 4).

And in which sense does it go beyond neoclassical theory? In the sense that in addition to a modification of the rationality postulate, it adds institutions as a critical constraint and analyzes the role of transaction costs [... ] It extends economic theory by incorporating [agents’] ideas and ideologies into the analysis, modelling the political process as a critical factor in the performance of economies, as the source of the diverse performance of economies, and as the explanation for the ‘inefficient’ markets. (p. 4)

A field of study where this version of neo-institutionalism has been widely applied in the last fifteen years is that of the theory of the firm and markets. Why do firms exist? And why is there such a great variety of types of firm, hierarchical structure, size, product diversification, and ownership structure? This is not an idle question, given that, conceptually speaking, a market economy could exist with well-developed product specialization even without firms. In fact, the division of labour does not necessarily imply the existence of firms. Why does the firm appear as an institution if, according to the model of perfect competition, the market is able to ensure allocative efficiency?

The problem was posed explicitly for the first time by Ronald Coase in ‘The Nature of the Firm’ (1937), where he observed that the market has use costs which must be taken into consideration together with the costs of production. When the former rise above a certain level, the market enters into crisis and is replaced by the firm. In other words, the firm is an alternative to the market if information acquires value, as the market is an inefficient means of collecting and controlling information.

Coase’s initial ideas have subsequentively been taken up by Oliver Williamson, who has used them to construct a transaction-cost approach to the theories of markets and firms. The transaction-cost approach was first laid out in Markets and Hierarchies: A Study in the Economics of Internal Organizations (1975) and later generalized in The Economic Institutions of Capitalism (1985). Williamson distinguished between ex ante and ex post transaction costs. The former are identified by the traditional category of the use costs of the market, while the latter are those arising in the phase of enforcement of a transaction and are caused by circumstances which are not regulated in advance by the contract. The level of costs is determined by the transaction characteristics, for which Williamson has identified three main dimensions. The first is specificity: a specific investment is required by one or both parties to the contract in order for the transaction to take place. The second dimension is frequency: for example, the repeated use of a habitual supplier allows for a marked reduction in costs such as those due to quality control. The third dimension is uncertainty: the more uncertain the exchange, the more detailed will the contract have to be.

In conclusion, transaction costs cannot be reduced to market use costs, nor exchanges to market exchanges. It is possible, for example, to speak of transaction costs in regard to the resources used to regulate the enforcement of employment contracts. Transaction costs are classified by Williamson into those required for co-ordination (necessary to reach an agreement among the contracting parties) and those required to cope with contract uncom­pleteness. With this conceptual framework Williamson has been able to explain in detail various organizational stages the firm has passed through, from the classical Marshallian form to the large modern conglomerates.

An internal criticism of the theory of transaction costs has been put forward by Harold Demsetz in Economic, Legal and Political Dimensions of Competition (1982). Demsetz’s central question was: why do firms sometimes produce their own inputs and at other times find it convenient to buy them from other firms? The question has to do not so much with the substitution between markets and firms as with the level of centralization of managerial co-ordination within and among firms. The more a firm produces internally the inputs it needs, the more centralized is its managerial co-ordination system. Demsetz observed that an increase in transaction costs does not lead to a substitution from market to managerial co-ordination, as the theory of transaction costs would imply. What happens instead is the replacement of managerial co-ordination in numerous but small firms by managerial co-ordination within a few large-scale firms.

Firms buy inputs when it is more convenient than producing them themselves. Although transaction costs are an element of the acquisition cost, it is only one element. According to Demsetz, emphasis on transaction costs has contributed to obscure the picture by implicitly maintaining that all firms are able to produce goods and services equally well. On the contrary, it is a fact that different firms are not perfect substitutes in the production of goods; so that a firm may find it convenient to produce its own inputs even if transaction costs are zero and administration costs positive. Basically, the confusion derives from the fact that, while it is assumed that the information for transaction purposes has a price, it is taken to become free when used for productive purposes.

Demsetz’s approach can be considered as a natural development of the idea of the firm as a ‘team production function’, an idea Alchian and Demsetz had put forward in ‘Production, Information Costs and Economic Organization’ (1972). In this article it is argued that the firm originates from a particular type of ‘market failure’: markets are not able efficiently to organize ‘team production’ because they are not able to supply sufficient information to evaluate the contribution of the single factors involved in production. If one worker’s contribution cannot be distinguished from another’s, it is difficult to establish to whom any variation in output is attributable. The remuneration of factors can therefore no longer be based on the marginal productivity of single inputs. It must be based on the joint product. However, in this case, each worker is encouraged to act opportunistically, as a free rider, by putting a suboptimal amount of effort into his work. The end result is productive inefficiency. Alchian and Demsetz suggested that a possible solution to the problem is to appoint one member of the team to act as supervisor, entrusting him with the task of controlling the individual subjects and fixing their remuneration. But then another problem arises: how can the supervisor be prevented from conniving with other members of the team, neglecting his own duties? The two scholars have an answer to this dilemma too; the supervisor should be entitled to dispose of the residual income, in other words, of what remains when all the inputs have been remunerated. Since opportunism reduces profits, it is in the supervisor’s interest to fight it. In conclusion, since the owner is the subject entitled to receive the residual income, the most effective controller is the owner of the firm.

Besides the transaction-cost approach, the other large research area in which utilitarian neo-institutionalist has developed is that of property rights theory. The double aim of this theory, as Alchian and Demsetz pointed out in ‘The Property Rights Paradigm’ (1972), is, on the one hand, to compare, in regard to efficiency, the consequences that alternative property structures can have on social allocations and, on the other, and in a rather more demanding way, to explain, on the basis of a criterion of efficiency, which structure of property rights is endogenously determined.

The basic idea underlying this research is that, in the exchange relation­ships of goods and services, it is not the goods and services themselves that procure satisfaction (or utility) and that give meaning to the exchange. What really counts is what the subjects have the right to do once they have entered into possession of the goods and services. This leads to a view of exchange as exchange of property rights: the value a subject attributes to a resource depends on the property rights it enables him to command, taking account of the fact that a property right includes a right to residual earning and a right to residual control. Thus, the research aims at explaining the development of different types of property rights through time, and at answering questions such as: How does the nature of the rights an individual has at his disposal influence his behaviour? What is the explanatory value of alternative models of property rights? Finally, and more specifically, how is it possible to explain the emergence of the firm as an alternative institution to the market by resorting to the category of property rights? In Demsetz’s original formulation, the structure of rights that exist in the capitalist firm is a reflection of the transaction costs caused by information asymmetries and the idiosyncratic nature of the actions through which individual perform­ances materialize. The most recent theory of property rights has tackled the problem of the factors on which the optimal structure of property rights in the firm depends.

S. Grossman and O. Hart and O. Hart and J. Moore started from a different premiss: the firm is no longer seen as a nexus of contracts and a way of solving the problem of transaction costs, but rather as a set of activities, capital goods and competences that are essential to the production process. The point is that individuals and capital goods are heterogeneous and cannot as a rule be substituted for each other. Complementarity among individuals and capital goods is therefore frequent and gives rise to ‘technological interdependence networks’. A subject may be endowed with such vast know­how and skills as to be indispensable to the firm. It follows that this subject should be the owner of the firm. One particular merit of the theory of Grossman, Hart and Moore lies in its ability to explain the emergence and sustainability of forms of business other than capitalist firms, as for example, co-operatives; which is tantamount to saying that the capitalist form of business is not the ‘natural’ form.

Along the same line of thought, H. Hansmann put forward a theory of business ownership based on minimizing the total costs of bargaining and ownership. Hansmann’s theory is grounded on the observation that trans­actions between economic agents can be regulated either by contract or by authority relations. Both methods of regulation entail specific costs: bar­gaining costs in one case, ownership costs in the other. In addition, there are different classes of stakeholders: workers, suppliers, customers, managers, shareholders, public authorities. Hansmann’s theory is that the ownership of a firm should be attributed to the class of stakeholder that demonstrates its ability to minimize the sum of bargaining and ownership costs.

One area of research where the utilitarian neo-institutionalism approach has produced copious results is the so-called ‘agency’ theory. Reflecting on the consequences of the separation between ownership and control of the modern firm, M. H. Jensen and W. Meckling brought to light a new type of relationship between owners and managers: the ‘agency’ relation. Here, a subject, the agent, undertakes, against payment, to carry out actions in the interests of another subject, the principal. In the specific case of a firm, the ‘principals’ are the owners, the ‘agents’ are the managers. The latter might well be tempted to maximize their objective function, which as a rule does not coincide with the owners’. Since it is impossible for the latter to fix all the managers’ obligations contractually and to control their actions, the only way out of this moral hazard problem is to define suitable incentive con­tracts. It is therefore a question of identifying a bonus and punishment scheme that will stimulate the agent to pursue the principal’s interests. Obviously an incentive contract will entail specific costs.

‘Agency’ costs come under a new category that differs entirely from the transaction costs category. As A. Shleifer and R. Vishny demonstrated, ‘agency’ theory has made business scholars pose the problem of designing an appropriate corporate governance structure, intended as a set of organiza­tional, legal and cultural instruments, on the basis of which the owners of the firm can secure the highest possible return on their investment. Beginning from the important distinction between formal authority (exercised by those holding rights of ownership) and substantial authority (held by those controlling strategic decisions and, consequently, utilization of resources), P. Aghion and J. Tirole drew attention to the fact that the constraints imposed by incentives are just as essential as those imposed by resources. In fact, if information and/or individual actions are neither knowable nor verifiable, paying attention to available resources will not suffice: it is necessary to determine which incentive scheme should be applied to induce individuals to reveal private information in their possession and refrain from acting opportunistically.

12.3.4. The new ‘old’ institutionalism

In the deluge of contemporary neo-institutionalist approaches, it should not be overlooked that there is an ‘old’ institutionalism which, indeed, is still very much alive and kicking: we refer to the institutionalism predicated by the heirs of Veblen and Commons.

When Walton Hamilton, in 1918, submitted a report to the American Economic Association entitled The Institutionalist Approach to Economic Theory, he was taking a gamble: that institutionalism, because of its greater capacity to reflect reality, would supplant neoclassicism as the dominating economic ideology, at least in America. To support his optimism, besides the intellectual prestige of social philosophers of the calibre of Dewey, Veblen, and Commons, there was also a rapid spread of institutionalist research in the American university system in those years. And perhaps, too, there was a growing hold of institutionalist thought on public opinion and political leaders. The New Deal appeared to confirm Hamilton’s optimism. In that ambitious experiment in economic and social reform, America tried to find the force of its frontier spirit to sort out apparently unsolvable problems; it endeavoured to overcome the shallows of depression, poverty and unem­ployment with an effort of will that appeared as a rebellion against the ‘natural laws’ of the market. But those problems were unsolvable only for neoclassical orthodox science. The institutionalists had the answers, put them forward, and the government adopted them. Institutionalist thought, as J. Fagg Foster theorized, was the incarnation of the true frontier spirit, the ‘can do’ spirit, the public’s ability to influence historical events positively, by rupturing institutional continuity.

At that time, Hamilton appeared to have won his wager, except that in the second post-war period things changed drastically. Within a few years the neoclassical approach, after assimilating the Keynesian revolution, took on a dominant role and the institutionalists were isolated. While the neoclassics won over all most prestigious universities and were awarded one Nobel prize after another, the institutionalists were relegated to a few universities in the provinces; they appeared destined to die a slow and inexorable death or to spend a wretched existence in the underworld of heretics.

But the institutionalist approach survived and a significant number of unconventional intellectuals continued their work waiting for better times to come. Heretics like J. Fagg Foster, Dudley Dillard, and Allan Gruchy set up small but valorous schools; John Galbraith, Warren Samuels, Marc Tools, Kenneth Boulding, and several others persisted in their criticism of neoclassical metaphysics and elaborated an alternative theoretical system. For many years, their words fell on deaf ears, sometimes they even found it difficult to publish their works, but still they continued to build. Until, at the end of the 1970s, the neo-institutionalist revolution broke out. Many eco­nomists, including some of neoclassical formation, in their endeavour to escape from the deadlock in which the official economic science of those years was floundering, uncovered the institutions. Thus good old American institutionalism was ‘rediscovered’; and many of the research studies, achievements and lectures of the heirs of Veblen and Commons, who had gone unheeded in the 1950s and 1960s, found new audiences not only among the public at large but also in the universities. Institutionalism has not died out in the long run and Hamilton may well win his wager after all.

J. K. Galbraith was undoubtedly the most important exponent of con­temporary institutionalist thought, a stream of thought which has in the Journal of Economic Issues its most prestigious critical platform. In the groove cut by Veblen, and on the basis of a very particular reading of Keynesian thought, Galbraith has explored the organizational nature and the planning methods of the company system as well as the influence of what he considers ‘the technological imperatives’; but he has also concerned himself with the social formation of individual preferences, interaction between private and public spheres, and the forces that influence the formation of opinion in the public sector.

In American Capitalism (1961), Galbraith put forward the theory of counter-vailing power. According to this theory, one way to keep a social system in equilibrium, whilst reducing inequalities, injustices, and exploita­tion, is that of balancing the excess power held by certain socioeconomic groups (large-scale companies, oligopolistic cartels, ownership associations, etc.) by allowing the constitution of other power groups with opposing interests. This is a theory full of realism and wisdom, but one which could find no place within the neoclassical theoretical system. In the trilogy made up by The Affluent Society (1958), The New Industrial State (1967), and Economic and Public Purpose (1973), Galbraith raised the argument that the ‘invisible hand’ is a long way from having the beneficial effects the laissez-faire theorists had attributed to it. On the contrary, it leads to a sharpening of inequality in the distribution of income, to the predominance of private over public interests, to the ‘squalor’ of the public economy, and finally to a low level of research and development activities. This last point plays a fundamental role in the process of economic growth; and it is a fact, according to Galbraith, that a large part of research and development is undertaken by the large-scale companies. It is also for this reason that Galbraith has been rather sceptical about the effectiveness and utility of anti­trust policies. He believes that strategic planning is a more useful kind of public intervention in the economic sphere, and that it should not aim at coercing private activity, but at co-ordinating it and turning it to the service of the public interest. In his most recent essays, The Nature of Mass Poverty (1979) and The Anatomy of Power (1983), Galbraith has moved so far along this road as to reach the point of calling for State intervention systematically directed to redistributing income in favour of the poorest strata of society.

As to J. Fagg Foster, who taught at Denver University from 1946 to 1976, although he published only a few works he contributed to keeping alive an oral tradition which helped to form several generations of heterodox eco­nomists. He was deeply committed to his endeavour to assimilate Keynes’s thought to the institutionalist approach, by emphasizing, for example, the endogenous nature of structural and institutional change. Keynesian theories and policy methods are not valid in the short run alone, as the neoclassical economists asserted, but especially in the long run. Public pol­icies are necessary to cope with the problem of unemployment and inefficient use of resources, which is not a short-run disequilibrium phenomenon, but rather a chronic problem of capitalist economies. For this reason, political and institutional adjustments, as, for example, the socialization of invest­ments and euthanasia of rentiers, must assume the form of structural reforms. These reforms must satisfy a valuation criterion which Foster called the principle of ‘instrumental efficiency’.

The criterion refers to the distinction between instrumental and ceremo­nial judgements. It was developed by Marc R. Tool when he elaborated the institutionalist theory of instrumental value. The evolution of an economy is seen as an endless series of problems, whose solutions call for continuous social restructuring through institutional adjustment. In this process, the principle of instrumental efficiency calls for regressive ceremonial institu­tions to be abandoned in favour of progressive and socially advantageous instrumental change.

To return to Foster, his theory of capital formation was developed along the lines of Keynes’s theory of the savings-investments equality. Foster interpreted this equality as an accounting identity and used it to ground a theory of long-run accumulation. From this equality he deduced that it was impossible for capital formation to be curbed by lack of savings. Indeed, in his opinion, it is the weakness in consumption that slows down growth. Savings are adjusted passively to investments, so that the process of accumulation is stimulated by technical progress and investment activity. Financing takes place through debt formation and subsequent reimburse­ment with income generated by accumulation. Political and economic institutions, particularly the Treasury and the banking system, play a fun­damental role in guiding and sustaining capital formation. Again in this case, Keynes has clearly been re-interpreted as a doctrine of economic policy that is valid above all in the long run.

Foster was an active member of the Wardman Group of institutionalist economists who met at the American Economic Association at the end of the 1950s. They later set up the Association for Evolutionary Economics (AFEE), an academic association that united institutionalist-oriented American economists. One of the most influential founder members of AFEE was Allan Garfield Gruchy, a leader of contemporary American institutionalism. Gruchy, like Foster, was largely influenced by the Keynesian revolution. He held that, far from being an efficient resource allocator, the market is the source of widespread failure and inefficiency. Unemployment, income inequality, inflation, inadequacy of social services, the excessive power of big companies and the alienation and oppression of workers and consumers, are all problems which, instead of being automatically adjusted by the market, are caused by it. In this perspective, it is clear that although the aggregate demand control policies proposed by the neo-Keynesians may possibly be beneficial, they are nonetheless quite inadequate, since they do not correct the basic malfunction. Gruchy held that more courageous policies were called for to influence the structure of the economic system. He did not hesitate to propose an active incomes policy, strict regulation of joint stock companies, a resolute anti-monopolistic policy, development of progressive social programmes and, above all, planning of the national economy— neither Soviet-type centralized planning nor, however, a mild indicative type of planning as adopted in France. Gruchy was, instead, inclined to identify with the social-democratic type of planning experimented in Scandinavian countries.

On a more abstract level, Gruchy proposed an approach which he defined as ‘holistic economics’, by which he intended to disrupt the conceptual and academic practices that isolated and separated the various social disciplines. Since the nature of man and society, that is, the object focused in the study of economics, is multi-dimensional, this discipline should be a science of culture and develop in symbiosis with sociology, anthropology, politics and his­toriography. It should reject the mechanistic paradigm of Newton’s physics and follow the Darwinian approach proposed by Veblen. Gruchy preferred to think of himself as an ‘evolutionary’ rather than an ‘institutionalist’. Yet he spent much of his career in the company of institutionalists, at Maryland University, where the Department of Economics became a den of heretical economists in the 1950s.

Dudley Dillard was the undisputed leader and ingenious moving spirit of the ‘institutionalist school of Maryland’. He developed what he called the ‘Trinitarian’ approach to economics, that is, an approach founded on three disciplines: microeconomics, macroeconomics and the theory of economic systems. The latter was to become the most important of the three. The economic categories are in fact historically and institutionally determined. Therefore, in trying to understand how real economies work, the study of the capitalist system should take priority over micro and macroeconomics. For Dillard, understanding contemporary capitalism means understanding money and technical progress, in other words the driving force behind accumulation. Capitalism is seen as a system of monetary production driven by innovative investments.

Warren J. Samuels is an economist who broke away from the approach of other institutionalists of this generation. He went back to a current of thought followed by Commons and Hale (rather than that of Veblen and Ayres) and chose law and economics as his field of study. He made an in-depth study of the State’s role in social construction. By reversing the classic liberal approach whereby the State is set up by mandate of social subjects, Samuels brought to light the constitutive role played by the State in respect of the economy and civil society. In fact the State defines rights and liberties, attributes power and enforces law. The State intervenes in the economy because it is legitimated to issue and change laws. At the same time, however, the economy intervenes in the State, by turning it into a territory of conflict and bargaining for the distribution of power. Thus, politics and economics influence each other; in reality they interact by building, modi­fying and rebuilding each other in a continual process of change. Samuels in fact rejected the neo-institutionalist approaches of neoclassical origin because of their tendency to juxtapose State and market as though they were two separate and independent things. In reality, this separation is quite pointless for the simple reason that the market is created by the State through the definition of rules and the attribution of rights. Moreover, the real political problem, in Samuels’s opinion, is not that of deciding whether the State should or should not intervene in the market and the economy, but that of understanding whose interests it serves by intervening.

Samuels distanced himself from the other institutionalists of his genera­tion also in that his studies of the history of economic thought and method­ology developed the pragmatist tradition of American institutionalism, which he reinforced with contributions from hermeneutics and post-modern thought. By so doing, he formulated a philosophical approach that got rid of the positivist base of economic science and brought to light its discursive, rhetorical, ideological, but above all, constructive nature. Economic reality is socially constructed whereas economic thought is socially conditioned. Social interests determine the ideological positions of the fighting groups. These, through the game of economic discourse, create political positions which, in turn, determine State action and construct economic reality.

The constructive nature of theories and ideas was also developed by Kenneth Ewart Boulding, an English economist who emigrated to America. Although he cannot be classed as a member of any particular school, he was very close to institutionalist and evolutionary thought. Among the various contributions he made, it is worth recalling the ‘image’ concept he proposed in 1956 to describe the mental constructs on which people base their choices, take decisions, undertake actions and, in so doing, construct social reality. These ‘images’ are particularly strong when shared socially; in this case they contribute to forming groups and determining collective action.

Boulding’s contribution to the evolutionary approach consisted in the elaboration of an impressive general theory of evolution in which all pro­duction, biological, physical, economic and technological processes, are seen as mechanisms for the creation of phenotypes, starting from genotypes. The latter are defined as genetic structures containing information and know-how; for example, the DNA in a chromosome, the egg from which a chick hatches, or the design from which a motor car is manufactured. In this perspective the differences between the creative activities of man and natural creation tend to fade. There are no substantial differences, except that the genotype of arti­ficial productions is contained in man’s mind rather than in the products of creation. Know-how is creative potential. But it needs to capture energy in order to transform the materials with which it gives rise to a phenotypical structure. Here, however, the production process comes up against some ‘limiting factors’, since energy, materials, space and time are scarce.

Lastly, we should like to recall Boulding’s contribution to constructing the post-Keynesian theory of distribution. For this, the English economist developed the ‘widow’s cruse’ parable to explain how, in the aggregate, incomes earned are the result of spending rather than the cause of it. Experience tells us that spending necessitates having some money and that spending decumulates our liquid balances. We therefore appear to spend as much as we earn. From the macroeconomic point of view, however, there is no decumulation of liquid assets in the goods exchange process; money is merely transferred from one hand to another. This series of transfers ends with a profit for those selling the goods, and will be the higher, the higher is the expenditure for consumption and investments. Thus it can be seen that, in the aggregate, expenditure generates profits rather than the reverse.

12.3.5 Evolutionary neo-institutionalism

This is how we define an area of research which some call ‘New European Institutionalism’ and which has recently been developed in the European Association for Evolutionary Political Economy (EAEPE). The economists who follow this approach refuse to be associated with neoclassical neo­institutionalists, whether of a contractarian or utilitarian credo. They rather appear to accept some form of relationship with the new ‘old’ American institutionalists, to the extent that EAEPE is sometimes introduced as AFEE’s European twin. It must however be remembered that European institutionalist traditions are far more heterogeneous and complex than American tradition, so that this ‘relationship’ may seem rather cursory.

Evolutionary neo-institutionalism links itself back to the critical tradition of Veblen, at the same time paying much attention to various subjects which are dear to Marxist economics. There is also a marked interest in Schumpeter and (among the more recent economists) in authors such as Michel Aglietta, J. K. Galbraith, Nicholas Georgescu-Roegen, and the post-Keynesians.

This stream of thought is like a river that numbers among its main tri­butaries the French ‘regulation’ school, various Schumpeterian approaches to economic change, including some recent ‘evolutionary’ lines of research on technical progress, and, finally, some of the attempts to analyse structural change and long waves. Polemic against the neoclassical tradition is strong, and the rejection of methodological individualism explicit—sometimes to the point of considering conditioned and unconscious action more important than rational choice. The approach stresses the role played by institutions, intended as social devices created to deal with uncertainty, as mechanisms of acquisition and elaboration of information, and also as frameworks capable of moulding individual actions and producing collective behaviour. There is a certain emphasis on the indeterminism of human choices and, in relation to this, a tendency to study economic change in terms of ‘evolutionary’ pro­cesses of the cumulative type, in a view reminiscent of the theories of Schumpeter and Kaldor, but, above all, Veblen.

The adjective ‘evolutionary’ is important in describing this type of institutionalism and in understanding what distinguishes it from the con­tractarian and utilitarian approaches. Here, the neoclassical theoretical system is dismissed ‘en bloc’ because the hypotheses that traditionally define Homo oeconomicus are rejected: hedonism, individualism, substantial rationality, completeness of information, exogeneity of preferences. This gives rise to two problems: how can the structure and dynamics of society be determined if one does not start from the rational choices of individuals? If individual behaviours are subject to strong externalities, what guarantees that social relations have a structure? And what guarantees that this struc­ture evolves in an orderly way, so as to satisfy human needs? The answer to all these questions is: Darwinian evolution. Organizations, productive apparatuses, regulations, habits, institutions, are all instruments used in the fight for human survival. For this reason, they change with the evolution of the species. There is no need to assume that institutions are built rationally in order to consider them humanly beneficial. It suffices to assume that a sort of ‘natural’ selection operates also among human artefacts. There will be a tendency towards the establishment of those institutions that best serve the needs of individuals, even if the latter have not chosen them intentionally. Society will therefore evolve in a positive way; yet not an optimal one, since constrained maximization is not one of the institution selection criteria. Nor is evolution teleological, since it is not decided theoretically by a central planner or a decision-maker outside the selection process. Nor, lastly, is it one-way, being continually exposed both to innovation and error. The state of society at a given time in history is path-dependent, that is, it depends of the path it has taken to arrive there; and at all times, history is faced with bifurcations: the past is a determining factor, but the future is open.

Lack of space prevents us from doing justice to all the original contribu­tions that emerge from the jumble of evolutionary neo-institutionalism. We shall therefore mention only a few of the more prominent authors.

First, Geoffrey M. Hodgson, who wrote two manifestos on new evolu­tionary institutionalism. He built upon the ‘old’ institutionalist approach as a base for a thorough criticism and a radical alternative to mainstream economics. He rejuvenated that approach by taking account of many achievements of contemporary research in evolutionary, post-Keynesian and post-Marxian theories. He relinquished the traditional notions of equilib­rium and Homo economicus and made instead wide resort to such new conceptions as open systems and cumulative causation. More than any other evolutionary institutionalist, Hodgson is contributing to the construction of a novel theoretical system which is completely free of any neoclassical rem­nant. Then we will mention Bart Nootebom and Ulrike Witt: the former distinguished himself in an attempt to overcome the economics of transac­tion costs through the study of organizational learning and trust-building. The latter dealt with the natural selection processes of firms in the presence of incomplete information, as well as with the problem of self-organization and spontaneous coordination of individual activities. In addition to these, we should mention Robert Sugden, who, after starting from a Buchanan- oriented contractarian approach, gradually left this behind and studied the role played by sentiments of equity in providing public and merit goods, as well as that played by spontaneous social conventions in attempts to sort out problems of multiple equilibria. Sugden, with J. Loomes, also made an interesting contribution on regret theory, where he tackled the problem of choices in the presence of uncertainty by allowing non-transitive preferences.

12.3.6. Irreversibilities, increasing returns, and complexity

The notion of increasing returns to scale has long been present in economic literature, but has only been satisfactorily dealt with in recent years. Already in The Wealth of Nations, Adam Smith had placed the emphasis on increasing returns and division of labour to explain economic growth; later, Alfred Marshall, introducing the concepts of externalities and irreversibilities of supply curves, identified increasing returns as the cause of multiple equilibria.

Cumulative causation, dynamic economies of scale, virtuous cycles, are all terms appearing in contributions made in the 1920s and 1930s to theories on international trade, regional and industrial economy and, later, from the 1960s onwards, to growth theories. As already recalled in section 8.5.1, Young had emphasized the role played by increasing returns and dynamic economies of scale in market expansion; Kaldor, realizing that there was a close link between increasing returns and technical progress, had introduced the notion of ‘cumulative causation’ to explain the differences in the industrial devel­opment of various countries; lastly, Myrdal had examined various mechan­isms of cumulative causation and ‘virtuous and vicious cycles’. More recently, Krugman applied these ideas to the study of spatial localization and modern theories of intra-industrial differentiation in international trade.

The pioneering contributions of Brian Arthur and Paul David come into this context. Both gave a clear explanation of the role of increasing returns in economic development. The dynamics of economic systems are history­dependent, in other words, they depend on the initial conditions and the course of the economy in previous periods. Another essential characteristic brought to light by their works is stochasticity: contingent deviations from the path of equilibrium influence the long-run trend. In particular, Polya type non-linear stochastic processes, studied by Arthur and the statisticians of Santa Fe school, have shown various interesting properties: for example, they are compatible with a plurality of possible long-run states, each of which is dependent on the initial conditions and casual fluctuations; fur­thermore, the structure of each state represents the outcome of a process that assumes the property of ‘self-organization’.

There are various types of increasing returns. Learning-by-doing and learning-by-using, co-ordination effects and network externalities generate increasing returns, depending on the adoption of certain technologies, because the more a technological standard is used, the greater are the advantages that derive from its use. The transmission of information based on experience and learning may also strengthen the initial position and consequently act in a similar way to more usual forms of increasing returns. In all the above cases, investment decisions appear to be irreversible and this explains why hysteresis prevails, i.e. effects that persist even after the causes that determined them have disappeared.

Increasing returns may give rise to various problems, as for example the existence of multiple equilibria and forms of Pareto inefficiency. Economists have long been aware of these and Arthur (1988; 1989) identified others:

(1) Lock-in effects: once a state has been achieved it is difficult to change; the dynamics become structurally rigid, with marked inflexibility and irreversibility.

(2) Path-dependence phenomena: the dynamics are influenced by initial conditions, so that history is of relevance.

(3) Symmetry-breaking processes: although starting from symmet­rical initial conditions, the final state may show asymmetrical characteristics.

Take the case of competition between a plurality of technologies initially present on the market. The result may be a technological monopoly: the market ends up trapped in one of the available technologies, not necessarily the most efficient. In this case, increasing returns generate positive feedback (the effects of certain actions strengthen the reason for persisting in those actions), which, combined with historical ‘accidents’ and casual fluctuations, determine lock-in effects.

Nevertheless, there are also cases of history-dependent systems which do not present lock-in effects, situtations where the agents may renegotiate and revise their choices. The effects of decision-making reviews can be studied using the concept of ‘probability of transition’ from one state to another. History is important in these models because of the path-dependence of the probabilities of transition. For example, in a market competition model, the probabilities of transition from one market structure to another depend on the existing market shares in a given state or the number of previous choices of a particular group of agents; these processes show convergence in the distribution of probabilities rather than point-convergence, and are parti­cularly suitable for analysing ‘social network’ contexts, or structures where different groups of individuals exchange goods and information. In these cases social conventions, rules and institutions may emerge to deal with the casual receptivity of information or the unpredictability of interactions among agents.

According to David, the conventions and rules that regulate the workings of organizations and institutions are ‘carriers of history’. They may arise from the evolution of spontaneously and endogenously originated structures or may be the result of conscious rules of interaction. Greif wrote (1994, p. 918): ‘The capacity of society’s organizational order to change is a func­tion of history, because institutions are a combination of organizations and cultural beliefs [... ] On the other hand, the organizations and beliefs of the past affect subsequent organizations and equilibria’. Recent developments in the study of the co-evolution of technologies, organizations and institutions are founded on this kind of considerations. The idea is now being established that the economy is a complex system.

Some systems, though evolving casually, tend to converge to a similar configuration to the initial one. These are moderately complex systems. Conversely, the historical evolution of very complex systems is typically one-way. History never repeats itself. Very complex systems are highly sensitive to mild disorder of initial conditions. The problem then is to assess whether fluctuations are controllable through regulation mechanisms. If they are not, they may, in the long run, give rise to catastrophic evolutions.

As stated above, systems with increasing returns are characterized by marked non-linearity. Non-linear systems may have many attractors, that is, areas in the space of the states by which the system is attracted, which represent a kind of ‘emerging solution’. The simplest type of attractor is a fixed point, that corresponds to a stationary solution. For non-linear dynamic solutions there is a vast range of behaviours: periodic attractors, quasi-periodic, non periodic, etc. Then there are deterministic systems, again rather simple in structure, which may behave unpredictably. And it may be essential unpredictability, in that it cannot be eliminated by acquiring further information. This phenomenon is known as ‘chaos’. To be more precise, it is ‘deterministic chaos’, a trend generated by laws which do not themselves entail anything stochastic, but which may produce erratic trajectories depending on slight disorders in the initial conditions. Dynamic systems in which the variables fluctuate unperiodically and trajectories meander chaotically are examples of complex dynamics. In general, for continuous time dynamic systems, chaotic temporal evolution takes place in spaces of a dimension greater than two. Furthermore, the interaction between inde­pendent systems makes chaos more likely, especially if the interactions are sufficiently high.

The discovery of chaotic phenomena has involved many scientific dis­ciplines. In economics it led to a new vision. On the one side chaos produces limitations to the models’ previsional capacities, on the other, the deter­minism inherent in chaos ensures that many seemingly casual phenomena are in reality less so than might be thought. Technological development and growth models with increasing returns are particularly suitable for repres­enting economic systems showing chaotic behaviour. Since the concepts on which chaos theories are grounded have developed in physical sciences, a parallel between the evolution of a physical system and the development of an economic system may serve to shed light on some interesting theoretical implications. At low levels of technological development, the economy remains stationary (corresponding to the stationary state of a fluid subjected to a moderate heat source). Higher levels of technological development (respectively, of temperature) may give rise to periodic oscillations: economic cycles. At more advanced levels of technological development, overlapping of two or more periodicities may be observed. Lastly, in an even more advanced stage, the economy may become turbulent (as in the case of boiling liquid), with irregular variations and strong dependency on initial condi­tions. This may be our present-day economic scenario.

Day illustrated the chaotic properties of some models of classical growth theories. Albin examined a disaggregated model in which interaction between firms gives rise, at aggregate level, to the chaotic behaviours studied by Day. Baumol and Wolff demonstrated the outbreak of chaos in a technological innovation model, while Deneckere and Judd studied chaos in markets with patented innovations. Interesting applications were also developed by Woodford for the financial market, where chaotic and unpredictable dynamics are actually quite frequent. Incidentally, the field of finance, on account of the ready availability and reliability of data, is an ideal terrain for the application of chaos theories.

Twentieth-century science witnessed the fall of Laplace’s rigid determin­ism. Heisenberg’s principle of indetermination, emphasizing the impossibil­ity of obtaining, simultaneously, accurate measurements for two different magnitudes (as for example, the position and speed of a particle), placed a strong limitation on the possibility of making forecasts. On the other hand, on a macroscopic scale, unpredictability may stem from the turbulence of a phenomenon, resulting in errors being amplified due to complex dynamics. These two aspects explain the failure of classical deterministic theories. The theory of chaos has brought us up against an even more dramatic limitation: the impossibility of foreseeing the future. The result is that equilibrium analysis, so central to economic discourse, is now disputed by a great number of economists. In its stead approaches are being developed to study how the preferences, expectations and actions of economic subjects react systemat­ically to the results which they themselves contribute to generate.

It is interesting to observe how these approaches are radically changing the traditional perception of the role of economic policy. From these studies it emerges that government authority, at any level, should watch out for both extreme positions of dirigism and laissez-faire. Rather, it should contrive to ‘gently’ push the system towards adopting behaviour structures that grow spontaneously. In the words of Brian Arthur, we do not need an invisible hand nor a heavy hand, but just a ‘nudging hand’.

12.3.7. Von Hayek and the neo-Austrian school

The expression ‘neo-Austrian school’ denotes a variegated group of scholars who, while endorsing Carl Menger’s line of thought, developed different research programmes, arriving at different levels of analytical rigour and scientific innovation. A small group of these authors argued that the neo­Austrian approach referred not so much to a specific economic doctrine, but, more importantly, to the way in which economic discourse can be used to demonstrate the superiority of liberalist ideology. It is to Fritz Machlup and his interpretation of the work of von Mises that we owe the spread of this stance—a stance which has recently been pursued by von Mises’ staunchest American follower, Murray Rothbard.

It is, however, in von Hayek’s scientific contribution that we can recognise the quality of a real system of thought. During the last quarter of the century, particularly in the USA, this system was the subject of a branch of research that can be defined as neo-Austrian proper. The main exponents are I. Kirzner, M. Rizzo, G. O’Driscoll, and L. Lachmann. To understand the guidelines to this approach, it is necessary to go back to the foundations of Hayek’s theoretical system. In previous chapters we have come across various specific contributions by von Hayek: the monetary cycle theory, the inter-temporal equilibrium theory, the critique of socialist planning. In this section we aim to provide an overview of his thought to emphasize the radically heterodox nature of the neo-Austrian approach when compared with neoclassical mainstream.

Von Hayek was an outstandingly prolific author, with far-reaching interests in research ranging from epistemological to economic problems, not to mention philosophy and psychology. It is, therefore, not surprising that although many scholars did not identify with the neo-Austrian school, they were nonetheless intellectually indebted to it. There is a unitarity in von Hayek thought that makes the view according to which there is an ‘early Hayek’, an orthodox neoclassical economist, and a ‘late Hayek’, a neo­Austrian theorist, hardly plausible. It is, however, possible to identify a first phase in his vast scientific production—in which he was interested mainly in the analytical aspects of economic theory—and a second phase, which commenced with the publication of his celebrated work The Road to Serf­dom, in 1944, where he dealt with political and moral philosophy.

An underlying theme that makes for coherency in Hayek’s theoretical system is the evolutionary perspective, which on its own is the true altern­ative to the equilibrium paradigm. Hayek attributed little explanatory capacity to the equilibrium category and the method of analysis based on it: ‘The main difficulty in the traditional approach lies in its complete abstraction from time. A notion of equilibrium that is essentially applicable only to an economic system conceived as atemporal can be of no great use’ (1939, p. 139). But the reasons why the equilibrium methodology is unable to produce a useful economic science were set out above all in his essay of 1937, ‘Economics and Knowledge’. The thesis put forward by the theorists of the microfoundations of macroeconomics—one of whom was Robert Lucas, who pretended that the first inspiration of the rational expectations counter­revolution derived from Hayek—is therefore unacceptable. Hayek sub­stituted the notion of ‘equilibrium’ with the more vague, but certainly more realistic one of ‘spontaneous order’. Order does indeed entail some form of regularity at the aggregate level, but is compatible with different config­urations of microeconomic structures. There is no contradiction between the existence of a macroeconomic order and the action of chaotic processes of individual adjustment. On the other hand, Hayek’s notion of order displays more than a few points in common with the notion of ‘complexity’ as used in the modern theory of systems.

The Austrian economist’s reflections focused on the problem of know­ledge, which he tackled from the point of view of the single agent—how he gets to know what he knows—and society as a whole—how knowledge is accumulated and used collectively. Already at the time of his theory of the market as a spontaneous order, the problem of knowledge was central to Hayek’s thought. The market enables the use of knowledge that is distributed among all the subjects; it is essentially a process of discovery, an evolutionary process which, through trial and error, succeeds in systematically adjusting the choices of economic agents and their results. Market order is spontan­eous and not artificially induced, least of all imposed by someone’s will. In this sense, even before the market is a system for exchanging goods and services, it is a system for producing and exchanging signals. The most sig­nificant of these signals are prices, which should not be conceived as indices of utility or remuneration of productive factors, but rather as means of information. A market economy enables a spontaneous order to emerge through individual learning processes associated with the information con­tent of prices:

The sum of information reflected in prices is entirely the product of competition, or at least of market receptiveness to whoever may have important information on the supply and demand of the good in question [... ] through channelling of information in coded form, the efforts of competition in the market game ensure the use of widely distributed knowledge [... ] Competition operates as a discovery process not only in the sense that it gives anyone the possibility of exploiting particular circumstances [... ], but also in the sense that it provides other parties with information that a certain opportunity exists (1991, p. 301).

Behind Hayek’s view of the market as a discovery and learning process, is an original conception of cultural evolution, which should not be confused with the theory of natural evolution. It is a process of competitive selection that gives rise to the rules of individual conduct and the social institutions that deal with problems of human interaction. Individuals take decisions based on social rules and this implies that their violation is always socially sanctioned. Such conception sharply contrasts with rational choice theory, where the decision-making subject, on the basis of a set of beliefs and a well-defined order of preferences, acts so as to maximize an objective function under constraint. For Hayek, the Homo oeconomicus of rational choice theory is a fetish devoid of sense. On the other hand, Hayek assumed that the agent is not fully aware of the social rules on the basis of which he takes his decisions. This was justified by the theory that rules materialize from an evolutionary process that is not governed by any one agent. The dynamics of the process are not intelligible on account of the complexity of the interactions between indivi­duals. In Law, Legislation and Liberty (1973-9, p. 534) he wrote:

Man has undoubtedly learned more often to do the right things without under­standing why, and still today is better served by his habits than by his ability to understand [... ] It was a set of learned rules that told man the right or wrong way to act in different circumstances and gave him an increasing ability to adapt to changing conditions and, in particular, to cooperate with other members of his group.

Unlike Walrasian general equilibrium, spontaneous order is not stationary. It evolves because the social rules of behaviour change in time depending on the results of human actions and interactions. Here Hayek anticipated the present-day notion of co-evolution between the rules of behaviour and spontaneous orders, a notion on whose basis the theory of evolutionary games has been constructed.

Hayek’s political liberalism is founded on his highly individualistic social theory. One of the Austrian economist’s proudest successes was the foundation in 1947 of the ‘Mont Pelerin Society’, a veritable seed-bed of contemporary liberal thought. The market is the best guarantor of political liberty; the very idea of social justice is unfounded: these are just two of the most brilliant gems of Hayek’s liberalism. They were of course exploited by the ‘new Right’, who used them as a battle horse to oppose all forms of economic intervention by the State. It should, however, be stressed that Hayek was first and foremost a liberalist, not a superficialist. One episode in his life is worthy of note: the Austrian economist taught at Chicago Uni­versity from 1950 to 1962, and for this reason is often, but erroneously, associated with the so-called ‘Chicago School’. In point of fact, he never taught at the Economics Department of that university, which unwaveringly refused him access.

Hayek’s main reason for insisting on the concept of the market as a spontaneous order traces back to his rejection of ‘rationalist constructivism’, whose paternity he ascribed to Descartes. Referring to Adam Ferguson, who had already dubbed social orders that evolve spontaneously as orders ‘of human action but not of human design’, Hayek invariably stigmatized the basic fault of constructivism: that ‘synoptic illusion’ which makes us believe that we can add up the knowledge of different individuals and, worse still, that this summation can be consigned to some intellectually well-endowed subject or committee of sages. In Hayek’s opinion, no centralized planning can ever work and no system of rules of conduct can ever be a work of social engineering. Where rationalist constructivism ‘makes us feel we have unlimited power to achieve our desires’, the evolutionary social theory ‘makes us realize there are limits to what we can try to achieve and acknowledge that some of our hopes are merely illusions’ (1973-9, p. 13).

Hayek is preoccupied, in a rather post-modern way, with putting econo­mists and philosophers on their guard against the Faustian aspiration to achieve perfect rationality. The old story of Faust can be traced as far back as the sixteenth century in almost all languages. But the most famous version, by Goethe, makes significant innovations over the previous ones. In fact, while Faust was traditionally described as such an egoist as to arrive at selling his soul to the devil in exchange for advantages and money, in Goethe’s version he is driven by noble and altruistic intentions: to free man from suffering and slavery. To succeed in his intent, Faust believes it essential to build a new social organization and feels the need to acquire

perfect knowledge. Thus, he makes his pact with the devil: soul versus knowledge. This is the Faust that Hayek had in mind when he railed against rationalist constructivism.

And what can be said of his peremptory judgement of social justice? Referring back to Kantian liberal tradition, Hayek observed that justice is a virtue and, as such, its field of application is individual action. An action is just if it conforms to the rules in force in the society to which an individual belongs. Therefore, since collective bodies do not have, nor can they have, virtues, to say that a social order is unjust is to commit an error of anthropomorphism. Hayek reached the conclusion that any attempt to achieve a plan of social justice can only lead to a ‘totalist’ State which destroys the freedom of individuals and social groups. In putting forward this argument Hayek brought to its natural conclusion an idea that had already been elaborated in the eighteenth century: the idea that social order emerges as an ‘unintended consequence of intentional action’. Nowadays it is referred to as ‘heterogenesis of goals’: there are social situations in which several individuals, moved by particular intentions, interact with each other by complying with the rules of socially shared conduct, thus giving rise to beneficial consequences that were not originally intended.

Hayek did not, however, subscribe to the idea that the systems of social rules in force in a given context should be left alone. Action in favour of an institutional design can and must take place every time the general condi­tions can be improved to increase the use and discovery of new knowledge, but not to increase the welfare of human beings. In his Studies in Philosophy, Politics and Economics (p. 150) we read:

The systems of rules of conduct develop as sets [... ]; and the fact that a new rule, combined with all the other rules in the group [... ] increases or reduces the efficiency of the entire group will depend on the order to which individual conduct leads, which in one situation may be detrimental, and in another beneficial.

What must be avoided at all costs—according to Hayek—is intervention with dirigistic-type social engineering plans. He believed these may generate adverse effects on the existing spontaneous order and undermine overall efficiency. Several critics have, however, observed that once the neoclassical theories of equilibrium and Pareto efficiency are rejected, the meaning of efficiency of a ‘spontaneous order’ is difficult to comprehend.

In conclusion, Hayek’s reflections are a point of arrival and an exaspera­tion of that invisible hand philosophy so authoritatively advocated by Smith.

As already mentioned, in the neo-Austrian school, Kirzner was undoubtedly Hayek’s most faithful follower. His most important work, Competition and Entrepreneurship, of 1973, deals expressly with an analysis of the entrepreneurial function. For Kirzner, it is impossible to know the market structure a priori. This emerges from the competitive process and therefore depends on the path taken to reach it. The orthodox theory is not acceptable, since it maintains that the market structure—perfect competition, oligopoly or monopoly—is a datum of the problem. Moreover, the market is characterized by permanent disequilibrium; and it is pointless to try to assess, like welfare economics tries to do, the advantages and disadvantages of a position of equilibrium which exists solely in the mind of the researcher. On the other hand, disequilibrium is the natural context of entrepreneurial affairs. The effects of the entrepreneur’s work are felt only in situations where individual plans fail to harmonize with each other. It follows that the entrepreneur’s raison d’etre lies in his ability to launch co-ordination processes that end up by satisfying consumers’ needs. In a certain sense, Kirzner’s approach is specular to Schumpeter’s. Whereas the latter starts from equilibrium to highlight the specific role of the entrepreneur-innovator, which is to break the harmony of the Walrasian circular flow, Kirzner starts from disequilibrium to exalt the entrepreneur’s equilibrating function.

In Discovery and the Capitalist Process, of 1985, the American economist dealt with essentially methodological issues, developing the theory that it is impossible to explain social objects without ‘‘understanding’’ them first. The researcher’s involvement with the object studied is a necessary condition rather than an impediment to scientific knowledge. This point was further pursued by L. Lachman in The Market as an Economic Process, of 1986. Lachman surpassed Kirzner in his radical subjectivism. It is to him, however, that we owe the distinction between ‘information as incorporation of a flow of messages which are the object of exchanges’, and ‘knowledge as the composition of thoughts that an individual is able to recall when he prepares and plans actions at a given moment in time’ (p. 49). It is precisely the difference between information and knowledge that makes it impossible to approximate, even asymptotically, to any informational equilibrium.

12.4.