Viability and competition

Back to main page Law and Society Viability and competition ------- Study of two opposing modes of complex regulation in economics and in law Serge DIEBOLT - Jean-Marc DURRIEU 1 General introduction Recent contributions of complexity theories have enabled legal theorists to renew the paradigm of a legality all too enclosed in conceptions of the hierarchical type and to open up systems which until then there had been a tendency to close rather hastily. The aim was to account not only for the present state of the law, which is limited to mere technique, but also for its changes, which implies going beyond the purely technical framework. This awareness is not new, but has not always had the same importance among jurists. For them, human activity is a perpetual oscillation between two contradictory tendencies: consuming or conserving. This oscillation is explained by the fact that the realisation of one implies the destruction of the other, and that the one nevertheless needs the other. No resource can be exploited to excess without exhausting it (history is only too well supplied with examples of this kind of counter-effect: forests, oil, ores, etc.). On the other hand, conservation, if it is merely an act of pure altruism towards future generations as regards non-renewable resources, is closer to development management (balance between extraction and renewal) for the others. Economic theory, for its part, asserts that the competitive organisation of the market constitutes the rule serving as a reference for the other forms of organisation, which are deviations from it made possible by regulation and more generally by the intervention of an external third party. If the legal sphere can then intervene, it is as a regulator, from the standpoint of viability of development. Indeed the law, by virtue of its vocation to bend human behaviour towards cooperation, presents itself as the nodal point of any action aimed at establishing or preserving an action in any other branch of human activity. Caught between the sometimes divergent interests of the Economy and of Nature, the Law presents itself as a whole with multiple complex dimensions. The first is constituted by the reciprocal actions of the environment on the law (1.1) and of the law on the environment (1.2). It will then be a matter of arbitrating a conflict of interests arising from the competition between the human species and the biosphere (1), but also between individuals themselves (2). We shall be interested in the particular case of the interaction between the productive system and its financial environment, by trying to identify what the conditions of viability are of a financial system based on the competitive organisation of the market, and what the zone of economic viability is that is determined by this competitive organisation of financial markets. 1 The law, for viability in competition When it is necessary to reconcile the irreconcilable... the law intervenes. Conceived initially as a mere regulator, this instrument of power subsequently became a sort of director, a guide. Perceived nowadays as a phenomenon that is always already there, its different modes of operation have become interwoven and more complex over the course of historical developments. The result is a somewhat heterogeneous whole made of overlaps and entanglements: a dynamic system between order and disorder. In fact, if one extends the legal model to its whole sphere of influence, one ends up encompassing the biosphere itself. The butterfly effect then takes on a quite different dimension. Not only can it set off a storm over New York, but it can also bring down Wall Street or trigger a revolution in Russia... Admittedly we sense intuitively that the conditions for setting them off are almost never met, but it can be demonstrated that they could occur, and in a perfectly deterministic manner: any evolution of a system (closed or not) is sensitive not only to its initial conditions, but also to tiny exogenous variations. The interweaving of relations between the law and the environment is due simply to the fact that individuals interact daily with the nature surrounding them. Other, more complex interweavings occur with the economic sphere, which the legal sphere must in part account for. This double entanglement leads the law to manage the contradictory interests which emerge from the confrontation of economic and ecological interests in the broad sense on the plane of competition. 1 The environment makes the law... The idea that environmental data must be taken into account within the framework of a morphogenetic theory of law dates in fact from Montesquieu. This political thinker was a great traveller and took care, while travelling the world, to study carefully the peoples whom he had the opportunity to meet (1.1.1). He was followed in this approach by many historians, but it is the evolutionist models inspired by those of biology that suggest to us most, nowadays, how environmental data can influence the way in which men build their systems of rules (1.1.2). 1 The environment in the political thought of Montesquieu The environment plays a central role in the political thought of Montesquieu (The Spirit of the Laws, books XIV-XVIII): differing according to the regions of the globe where one is situated, it influences people, and this change has repercussions on the forms of government, and the means of governing. For example, the hot and humid climate slackens the fibres of the body and encourages peoples to nonchalance, making them little enterprising in commerce. On the other hand, heat heats the spirits, and these peoples will show an inclination all the greater to quarrel, all the more so as a favourable environment will meet their needs more quickly, leaving more room for desires, the cradles of passions. The tendencies to war will be more marked, and laws based on greater firmness, even a genuine despotism, will ensue. If this somewhat mechanistic and superficial view may prompt a smile, the fact remains that these conclusions were based on methodical and systematic observations, which nothing today authorises us to contest. Even if nowadays the universality of communications and the cultural transfers which resulted from it have largely erased the trace that the environment may have produced, certain remarks seem to us singularly pertinent, and it is not excluded that in their time, when groups were often very autonomous and self-sufficient and communications slow and uncertain, the cause-and-effect relationships noted by Montesquieu were not on the whole accurate. 1 The environment, a fundamental co-actor of every normative system 1 Genesis The emergence of a legal system is a phenomenon that is still largely poorly understood. Presenting itself to litigants born and educated within modern societies, it presents itself as a phenomenon that is always already there, and its genesis is often difficult to reconstruct, all the more so as modern and new systems of thought are nonetheless marked by their origins. One can easily describe a movement such as Marxism as an evolution in the face of another system which pre-exists it. It is therefore not really a creation. So much so that it is difficult to conceive what a pure creation ex nihilo could have been, so true is it that just as we are beings of language, we are social beings, and it becomes problematic to show which of man or society created the other. However, by integrating the environment as a virtual third party, one can sketch the outline of the emergence of a social consciousness. The influences which Montesquieu had noticed of climates on the morphology of individuals will be disregarded, on account of their uncertain character and the difficulty of generalising them. If his remarks seem to us generally sensible, they nevertheless suffer from formidable exceptions. Northern peoples do not always behave in a gloomy and tendentiously alcoholic manner, those of the South are not always indolent and/or passionate, etc. Nowadays, the globalisation of communications has in any case made any empirical verification of these assertions impossible. Perhaps they were well founded two centuries ago, when each society formed a quasi-closed system. It will therefore be considered that the actors in the game summarily described are not necessarily a product of their climate. One obtains a variant of the prisoner's dilemma game. Take two individuals. Each of them possesses a living space, which is defined as the space which is necessary for them to satisfy their vital needs. The environment presents itself in the form of two exogenous factors, one static, the other dynamic (typically, fauna and flora). Man's place in the food chain places him in a global relationship of predation (flora constitutes more a resource than a threat for him). Fauna is on the whole not very hostile to him, but may constitute a danger depending on the species. In this context, the first task of an autonomous system such as the individual is to ensure his survival. Resources being by definition limited, any other similar being constitutes, if not a danger, a potential competitor. Three strategies are then available: indifference, cooperation and competition. The choice of these strategies will be made by means of a "calculation" of the evaluative-associative type. It consists in evaluating the advantages or drawbacks of the foreign presence nearby. One obtains the following matrix: CALCULATION ESTIMATION ACTION 1+1<2 Proximity of the other = - resources or + danger Competition 1+1=2 Maintenance Coexistence 1+1>2 Proximity of the other = + resources or - danger Cooperation Initially, the evaluative calculation may be considered objective, that is to say that the choice of strategies translates into an immediate effect on the environment. This evaluation follows an inverse curve according to the variations in environmental data: the decrease in resources (man-predator pattern) or the increase in danger (man-prey pattern) translates into the conclusion that competition is necessary, and vice versa. 1 Complex evolution Many examples can be found of natural systems using this type of strategy. The symbiotic relationship is one of the most sophisticated examples, which leads heterogeneous species to cooperate. But most of the time, cooperation remains intra-species. The evaluative calculation follows the O.I.D. (organisation-information-decision) decision pattern of the recursive S.T.I. type. Its decision will therefore most of the time follow an operation of memorisation, and the initial experience event will be of contingent origin. It is naturally stabilising because the objective is originally fixed towards a reduction of differences (homeostatic maintenance of the individual). The environment may on this occasion take on, in the imagination of the actors, the features of a virtual third party, which unites against it the initially competing interests. Complexity intervenes from this level, despite the simplicity of the reasoning, because individuals are genetically all different, although possessing an important common heritage. This characteristic causes at the cognitive and decision-making level significant interferences, which grow over time, like two deterministic dynamic systems which diverge because of their sensitivity to initial conditions. Except that living systems feed back on one another, each time with a share of the Same and of the Other. This means that these partially positive and negative feedbacks make individual reactions unknowable a priori, although by experience partially predictable. In this case, Reason presents itself as a negative feedback stabilising decision-making processes towards a lasting equilibrium. We then tend to describe as rational what makes us react in the same manner to the same stimuli. 1 Stability and virtual third party Without going into the details, one can summarise legal morphogenesis as a lasting development of behavioural habits: when a lasting cooperation is established between several individuals, a phenomenon of emergence of a collective consciousness occurs. If the phenomenon extends across generations, individuals then feel the phenomenon as always already there, and enrich it with their own experience towards better viability. The collective phenomenon then acquires de facto an exteriority which brings it almost to autonomy. The law presents itself at this level as a set of norms generally accepted and applied by individuals which guarantee the autonomy of the collective by maintaining its homeostatic equilibrium. Transmitted orally, the law is called customary, and may be accompanied by institutions responsible for having it applied. Writing then comes to perpetuate the normative system, and to facilitate its transmission in time and space. In these conditions viability presents itself as an adequacy of the evolution of the law with its environmental constraints. These premises introduce the intervention of the environment within a complex factor: the emergence of norms is in fact dependent on the interindividual and therefore subjective calculation which triggers the strategy of cooperation. This calculation is carried out according to the perception that individuals have of their environmental data. The norm then presents itself as emanating from the stabilisation of a cooperation strategy. The memorisation function here comes to reinforce the natural stabilisation. The non-monotone becomes monotone. So much so that individuals may continue behaviours without really knowing the reason for them. This, in conjunction with the ideas of Montesquieu, shows that it is possible to consider that, laws being relative to the complex personality of those who create them and to the environmental constraints which pre-exist them, it cannot be denied that each legal system is in a certain manner determined by the data of an environment, support but also object of the law. 1 ... but does the law make the environment, and how? Legal viability between management and direction The reciprocal action poses less of a problem nowadays. Environmental law is henceforth a branch in its own right. Governed by the penal code, the rural code and by an important corpus of regulatory texts in France, the environment can rapidly change appearance at the will of a legislator. As a result, the multiplication of degrees of complexity refocuses the problem of a viable development towards an arbitration between several antagonistic interests orchestrated by a whole perceived as complex with varied modes of action. 1 Difficulties in defining a legal viability While it is relatively easy to situate the influence of environmental variables in a perspective of legal morphogenesis, their a posteriori intervention remains more uncertain. The temptation was long great to wall oneself in behind technique in order to consecrate customary practices while merely limiting abuses. Generously granted permits, administrative tolerances, it took the birth of protest movements and the formation of pressure groups for a new drafting of legal texts to emerge. A real legislative and regulatory awareness came to light at a high hierarchical level, giving concrete form to a real intervention by the public authorities. From vaguely protective, the law turned towards a will of preservation and rehabilitation. It is on this occasion that the latent question of a legal viability arose. 1 A vague, relative and unstable notion Defining what a legal viability might be is not simple, because the law as it presents itself to us in France nowadays can be described as a complex set of norms using numerous vague notions or notions with variable content. It is an instrument of social peace, but also of power. This duality of function is due to the fact that the law, faithful to the old maxim prevention is better than cure, presents itself as an instrument for resolving problems ex post and also ex ante. It is this latter point that makes it an instrument of power. The viability of the law as such is therefore in fact merged with that of its object (the social group that it regulates). Moreover a notion of viability of the law itself seems capable of being set aside. Indeed, as a regulator, the law has no other object than its own destruction. If society regulated itself under the effect of an invisible hand or of a sudden improvement in the behaviour of individuals, the legal system would disappear of its own accord, its object being achieved. But if it is conceivable, this object immediately reveals its utopian nature: how to prevent men from coveting the same things, the same persons? For the law, competition is not compatible with viability. On the one hand, viability means for the law its disappearance for the benefit of the development of the social group that is the object of its regulation, and this viability is called into question because of the persistence of the phenomenon of competition. If one reverses the argument, one can then assert that the viability of legal development would be, so to speak, immanent in competition within the social group that is the object of its regulation, although it would then be more appropriate to speak of persistence rather than of viability. This phenomenon is inherent in the entanglement of hierarchies which will be developed in the following paragraph. On the other hand, the viable development of a social group that is the object of legal regulation consists in adjusting the values of this regulation to counter the concentrating and destructive effects of competition taken to excess. To this end, the law frames with strict rules the interindividual relations relating to persons and property. Civil law, for example, manages family relations to avoid competition between generations, competition of powers within the family, it also organises interfamily relations: marriage blocks competition in matrimonial relations, successions in inheritances, etc. In the economic field, this regulation of competition is even more obvious, despite the flexibility of the rule of consensualism. The law, far from suppressing them, therefore tends to channel competitive tendencies. How then to organise a viable development of the social group, insofar as one does not suppress one of its greatest potential causes of destabilisation? It will be seen that in fact, this solution is not so paradoxical given the entangled structure and the overlaps of the different dynamic systems present. 1 A complex notion: entangled hierarchies and group conflicts The complexity of legal viable development comes from its situation, at the confluence of several heterogeneous fields: natural sciences, economics, sociology... Each system of representation of real phenomena is connected to the others by links which are often bilateral. The entangled hierarchy that ensues then produces its classic effects: counter-effect, complexity and reversal of values. Interindividual competition is therefore balanced between two poles, one positive, the other negative. On one side, competition leads men to surpass themselves, on the other side, the idea of sacrificing the weakest seems incompatible with the most elementary notions of humanism. The notion becomes more complex in passing from one field to another. Each model has its own values which sometimes come into conflict with those of other systems (the case of regulator-regulated relations for law and economics, cf. infra). One therefore obtains for one and the same phenomenon a multitude of multicomplex systems possessing self-similar but not quite fractal structures, in homeostatic equilibrium. If one considers each country as a legally regulated system, one can note that the introduction of monoculture into a self-sufficient economy is of a kind to reinforce the link between client country and producer country, because specialisation increases yield for the common profit, but competition plays a destabilising role. If a client country has the choice between several producer countries, it will choose the one with the most attractive prices, to the detriment of its former partners (treaties and other international conventions have demonstrated their notorious ineffectiveness in this area when it came to stabilising situations). It follows that the loser is doubly a loser: he loses the advantage of his specialisation (he now has to import his basic necessities), but the loss of his know-how in the other fields obliges him to make additional efforts to rebuild it, depriving him of almost all his chances of returning to the race: competition destabilises what cooperation stabilises. In these conditions, viability presents itself as an extremely unstable notion and eminently linked to the notion of power, which is exercised primarily through the exercise of the legal order of direction. 1 Viability of a legal order of direction Depending on its modes of action, the law will experience varying performances in terms of viability. 1 The law is a set of binding norms The Romano-Germanic conception which underlies the conception of the State on the old continent classically presents the law as a set of norms of constraint to which all individuals are subject. It is the State which ensures the durability and application of these rights and obligations. This conception gives a pyramidal representation of norms which ranks the norms according to criteria of superiority, which makes it possible to establish their validity. One obeys a norm because it conforms to a higher norm, and so on. The problem is knowing what is right at the top of this pyramid, the fundamental norm being, in the absence of transcendence, difficult to materialise. This difficulty has not prevented the construction of extremely solid and efficient systems which we still see at work today. The institutional hierarchy comes to complete, by reinforcing it, the legal hierarchy, for the greater security of the system. 1 A powerful whole but a fragile viability This type of construction is not without drawbacks. It suffices to realise this to analyse the circulation of information flows within this type of structure. The hierarchy by definition favours a top-down operation of data, most of the time in the form of orders or directives. The upward flows are most of the time merely the vectors of the corollary of this type of information, namely responsibilities (if the structure is transparent). For agents situated at the lower levels, decisions emanating "from above" have a fairly marked character of opacity. The result is a phenomenon of partial blindness of the agents, who as a result are led to decision-making of limited scope, on the one hand, in conformity with the "higher" directives, on the other. These structures, inspired by military models taking up the metaphor of the head and the legs, have demonstrated their great efficiency, when it is a matter of a few men taking decisions relating to the entirety of the community. It has been shown that, depending on the size of the groups considered, this type of structure was almost optimal. However, certain dangers have also been pointed out. Let us cite among the most important that, resolved in his own way by Hobbes, of investing one individual with a strong power of action over the whole of a group. Arbitrariness is a bad companion of absolutism. This makes the structure heavily dependent on the achievement of the goals for which it was created. Indeed, individuals commonly consider that their freedom is reduced, which they accept as long as they consider that this reduction is necessary or profitable (cf. table in 1.1.2.1). Structural stability is therefore dependent on an interindividual calculation of relevance: the appraisal of the adequacy of the social system. But this stability is made fragile, because of the continual individual-group confrontation. As long as the information flow is reduced or non-existent from the base towards the top of the structure, any exogenous shock must be compensated, either by an increase in repression (blind compensation for the inadequacy), or by an adaptation (spontaneous self-modification of the structure because of meta-rules or hierarchical decision, supposing that the impact of the shock is sufficient to travel back up to the "top"). This type of regulation can therefore be presented as durable (examples exist of millennia-old organisations), but of low viability. The latter depends indeed only on the capacity of the preponderant decision-making bodies to take the measure of the importance of the changes they have to face. Each exogenous shock presents itself, facing a rigid artefact, as a potential risk of destabilisation. To try to compensate for these problems, other types of government are conceivable. But a greater potential viability does not thereby signify an increased durability. 1 Viability of a self-regulated legal order Legal orders tending towards self-regulation are for the most part of Anglo-Saxon inspiration. They oppose the notion of hierarchy by a philosophy inspired by economic liberalism: in England, it was political liberalism. 1 Popular sovereignty tends towards self-management In France, the ideas leading to the idea of a sort of self-organisation are found indirectly in the thought of Rousseau and Voltaire. From a systemic point of view, the notion of popular sovereignty is nothing other than a pure feedback of a social whole on itself by means of a political system which controls it and which it controls. An entangled hierarchy par excellence, the Rousseauist system is the prototype of the global and complex autonomous being, State-synthesis of heterogeneous individuals. British thought, the most representative of which is that of Adam Smith, and all that which succeeded it, considers not so much the structure which feeds back as that which emerges, and considers self-organisation as an end in itself more than as an active end. One may nevertheless stress that, as J.-P. Dupuy pointed out (Dupuy 1994), the principle of sympathy is a form of feedback of the unit on the totality, disguised insofar as it is only the product of direct individual will. One harmonises voluntarily with others in Adam Smith when one is harmonised (also voluntarily, thanks to the social contract) in Rousseau. In fact, whether direct or indirect, this individual will to subordinate oneself to a group or a supra-individual structure has as its direct consequence to engender a complex global being which will exist against other beings, other complex communities. New conflicts or cooperations will come to light according to the same type of estimation as that of individuals, and so on. However, each structure possesses a capacity for autonomy, which confers on it powers of self-management. In these conditions, the viability of groups, by definition excellent since it emanates from a totality or a stable majority of the ensembles, will assure them great stability: legal viability entails an endogenous durability. However, this durability may be considered intersubjectively counterproductive. 1 A good viability which is paid for by a strong legal insecurity Indeed, we have just seen that property, also a condition of legal viability, is the flexibility of the structure which supports institutions, particularly State institutions. This flexibility makes it possible for information to travel back up according to a bottom-up pattern which ensures a permanent adequacy of the structure to individual or exogenous changes. But this adaptation also means a mutation, by its essence prejudicial to the very notion of regulation, which generally presupposes that of the regulator. Concretely, legal stability means that agreements are subject to a stable legal framework which ensures their durability. Now, it emerges from the analysis of self-regulating patterns that these tend to be unable to ensure their stability and that of the groups they manage, for at least two reasons: • on the one hand, the notion of implicit self-organisation as defined by the liberals the premises of its own functioning. To postulate that one assumes that sympathy will come to make the effects of competition beneficial supposes that one implicitly admits that competition contains within itself its own causes of limitation. Now competition, like any mechanism of elimination by selection, is concentrating. It is therefore utopian to consider that agents moved by competitive considerations alone will spontaneously adopt, in the conditions we currently know, a self-limiting attitude. • on the other hand, legal stability as we have just evoked it loses in the case of a self-regulated system its character of fixed point. If the rules change as a function of interindividual variations, the agreements subject to the legal system therefore begin to feed back on it. The law of contracts becomes what contracts make of it, and it becomes necessary to include in each contract clauses which contemplate the changes of its environment. These forms of management based on the maintenance of spontaneous orders have nevertheless shown their durable character in the face of other types of groups: exogenous durability harms legal viability. 1 Viability as a compromise for durability In fact, the two "pure" types of regulation have in the long run proved non-viable. The British legal system is compelled to have recourse in an indirect manner to principles of interpretation and management of contractual relations, to cite only those. Whether it is the age-old habit of a certain form of authoritarianism of government, or the inability of rational agents to push their rationality to its limits (these two points may combine), one notes that most so-called developed societies propose compromises alternating bottom-up regulation by means of regulations and top-down regulation by means of customs (the market can partake of both at once). Each sphere (generally economic, political, social) is isolated according to three elements: internal constraints, external constraints, and external conflicts which serve as a fixed point. Competition plays the driving role which ensures an interested cohesion. What viability can be found in a system based not on the resolution, but on the pursuit of conflicts, what is more when these conflicts are motivated by divergent interests? 1 Viable development: the emerging paradigm of a tension The relative positions of the different systems and the complexity of their regulator-regulated relations bring out the existence of antagonistic interests, hence, failing conflicts, of structural tensions always putting back into play an equilibrium that is conceived and perceived as precarious. The stakes of a viable development therefore imply an in-depth analysis of these tensions, insofar as a parameter perceived as small is liable to produce significant stochastic effects. 1 The entangled hierarchy of the economic and legal spheres... When one places oneself from the standpoint of the development of an emerging social group, the social parameters (interindividual relations) are rapidly doubled by self-similar economic relations (individual-object associations). The theory of conventions provides an excellent complement to the model of social emergence, by introducing a global dimension to property relations. 1 Cooperation and convention: a structural instability The theory of conventions suggests the necessity of the intervention of an external third party. We shall show that this being can be found in a hypostatised environment or a self-transcended collective phenomenon. To realise this, one can use a sequential pattern inspired by that of Kreps (1990): A B 0 0 5 5 -10 10 In the first situation (where the agents do not cooperate), the agents' gains are nil. If they both cooperate, they derive a benefit from it. On the other hand, if only agent A decides to cooperate, his gains will be negative, whereas those of B will be higher than those expected in the event of cooperation. This example is applicable to any type of exchange of the onerous type and characterises the relations which the law describes as synallagmatic. These are typically exchanges where the operations of delivery and payment are not simultaneous, so that the cooperation/defection decisions are doubled and non-monotone. A common example is that of the restaurant: if a customer enters a restaurant, consumes a dish and it turns out that he has no money to pay for it, he will have benefited from a free service, while the restaurateur will suffer on the one hand the loss of his property (the meal) and the loss of earnings that the transaction constitutes (he could have resold it to another customer, a solvent one). The "socially" optimal situation proposed by this pattern is that where both agents cooperate since the transaction brings a gain to each of the agents (let us recall that the sine qua non condition of the maintenance of an autonomous social structure is the affectio societatis of each of its individuals, cf. our model of emergence supra), and is in any case higher than the others (10 against 0). However, according to this pattern, if the agents are rational, cooperation cannot be reached, because if A cooperates, B has every interest in not cooperating. Thus, the only "equilibrium" situation is that where neither of the two agents cooperates, that is to say a socially prejudicial solution (except in the absence of exogenous constraints on the group). 1 Stability and regulation One finds here the problem of the "incompleteness of pure market logic" (Orléan, 1994a), that is to say the result of the strict horizontality of the relation in the sense that the individuals present share nothing, except the desire to maximise their personal utility. Obtaining the optimal solution (the most unstable) presupposes the existence of an external third party. This may be either the virtual third party arising from an environmental constraint, possibly hypostatised, or constructions arising from the perception of a self-transcended social fact. One can, with A. Orléan (1994b), distinguish several forms of cooperation, each requiring the intervention of a third party whose effectiveness rests on its perceived exteriority in relation to the strategic game, to individuals and to the market world. These may be: - trust based on reputation (itself based on trust), negative feedback based on the repetition of past successes which suggests the probability of future successes, which leads A to collaborate ab initio, - the oath based on the sacred (rites, sacred texts and traditions), - the contract, guaranteed by customs (or an invisible hand) or a judicial apparatus. Anthropologists know that in primitive societies, barter does not answer to economic concerns but to a logic orchestrated by custom, magic or religion. The study of the circulation of objects shows us its subordination to an order which pre-exists it: the sacred, respect for the dead, lineage etc... Goods circulate horizontally along networks of kinship or alliance, and vertically by prestation, redistribution or gift. Exchange does not obey an interindividual relation; it is founded on values of the collectivity, answers to rites and customs and in that respect must be considered only as a momentary episode forming part of continuous social relations. The judicial apparatus may play the role of the third party if the protagonists enter into a contract insofar as the judges have no interest in the exchange and cannot be bought. A's gains will however be reduced by the costs of justice which must therefore not be greater than 5. Repetition of the game does not necessarily solve the problem. The potential gains of each being 5n if they cooperate n times, one might think that they will cooperate at least once since n > 2 => 5n > 10. B's gain if he decides to play the game will therefore be higher than what he would obtain if he decided not to play the game at the first stage. However, if A and B know n, since at the nth stage B no longer has any interest in cooperating, A will cut the relationship short at n-1 stages. If B is rational, he will anticipate A's behaviour and cheat at the n-1th stage etc... Finally, by proceeding by recursive reasoning, one finds that the agents have no interest in cooperating. For cooperation to be possible, either the number of stages must be unknown (random) to the participants, or it must be infinite. In any event, it is necessary that the system in which these interactions take place be of infinite duration or a priori unknowable. This condition will be considered as a condition of viability of the system. A contrario, a stabilisation will subsequently be established as a function of the constraints posed by the environment, normally tending towards an equilibrium, since insofar as respect for conventions yields a gain, it is profitable for everyone to ensure their continuity. A relationship of dependence is built up (the common gain becomes a minimum threshold of well-being by adjustment) which reduces for each of the agents the potential gain from defection: in an iterated game process, defection yields an immediate gain but a long-term loss that is by definition infinite (repetition of loss of earnings). The agents must therefore see their relations as an infinite gain. As regulator of property relations, the law is de facto an economic regulator. The notions of public policy of direction and protection, of equity or expediency, of consensualism and of good faith are so many conceptual instruments serving to bend commercial relations towards a quasi-ethics. The most flagrant example is found in the emergence of competition and consumer law. A judge-made creation of the last twenty years, this new branch of law has seen the judge turn into an economic arbiter taking more and more into consideration the criteria of economic power and dominant position in order to assess the equality of commercial relations. Consensualism is henceforth bent by taking account of the economic positions of the parties and of the weight they de facto possess in a negotiation. This mechanism presents itself as an adaptation of an earlier law which had not taken this type of parameter into account. Using the same methods, competition law has tended to rethink legal relations between companies in the light of greater economic realism. The objective of this approach was to soften a corpus of rules that were too heavy and too easy to circumvent. The result is a penetration of the legal sphere into the economic sphere. No economic phenomenon is nowadays totally isolable from its legal (and by extension, political) environment. The examples are numerous, the most striking being the confusion which was long maintained on the occasion of the tensions of East-West relations between, on the one hand, liberal democracies and, on the other, regimes of communist inspiration. The various communisms were moreover economic doctrines, but one could just as well have said legal economies. But it is not deniable either that the law, by means of bottom-up influences, undergoes from the economic sphere shocks which durably affect its structure: the emergence of competition and consumer law is the result of socio-economic changes which called into question mechanisms as theoretically powerfully intangible as the general principles of civil law. The law influences the economy, and vice versa. The result is very often a conflict of interests engendered by the very notion of the regulation of one system by another. 1 ... reverses the models of optimal regulation Competition, because of entangled hierarchies, presents itself either for the economy as the engine of innovation and of the adjustment of prices on the market, or for the law as a concentrating and dangerous mechanism, which must be controlled failing being suppressed. In fact, each discipline considers this notion from an angle of its own. Schematically, the law places itself at the extreme, and the economy at the centre of a circular phenomenon. That is why the consequences which are envisaged are radically different: generalised competition leads to monopoly, hence to its own elimination. This property, characteristic of a circular hierarchy, leads each discipline to give its own definition of the notion of competition. One can therefore distinguish an "economic competition" from a "legal competition", itself different from "social competition" or "biological competition". The natural sciences consider "biological competition" as the convergent will of two autonomous systems towards a common object. For example, the will to seize the same prey or the same portion of territory. As we have seen, this convergence may be resolved by competition (trial of strength), or in certain cases, by cooperation. "Social competition" is a similar phenomenon. It is the common desire that individuals have for material or non-material objects emanating from the group in which they live. This competition in the broad sense targets either the objects that the community makes it possible to produce, or intangible values but inherent in this same community. For example, consumer goods or services emanating from institutions, or valued virtualities such as prestige, power, social rank. Psychologically, the foundations of this notion rapidly gain in complexity, as the object becomes more and more intangible. Faced with this mechanism, "economic competition" finally appears simpler. Schematically, it resembles more its ecological sister, with as a difference the object of the agents' covetousness: it will not be so much resources as customers. What will make the difference will therefore not be so much linked to the idea of force as to that of competitiveness. Each notion fluctuates according to the articulation of these different systems. Biological competition places man in a relationship of competition with his environment, and insofar as his influence now extends over the whole earth and even beyond, one may consider that humanity competes with the rest of the earth's species. By his faculty to exploit the mechanisms of cooperation and memorisation of collective experiences, man has managed to raise himself to the rank of a "super-predator", whom his technology has placed beyond the real reach of any competitor. He moves faster, further and strikes harder than any other being. In fact, there therefore remains a priori to man only a social competition which relegates biological competition to the background, without however annihilating it. The latter simply passes to the background of priorities, since it is the most dangerous "predator" that one considers first. Now, any imbalance in one system having immediate repercussions on the others in a counter-effective manner because of entangled hierarchies, viability may be considered as the fragile maintenance of an acceptable compromise between short- and long-term interests, between individuals and groups of the four systems. But this implies a cognitive openness, that is to say a constant information about the potentialities and risks of each action, and about its short- and long-term repercussions. Let us say it: this faculty of estimation is not given to everyone, far from it. It is in fact more pertinent to speak for the agents of the partial blindness that we mentioned earlier. It should be noted that this blindness may be endured but also willed. Here too an estimation is made at the interindividual level between the advantages of an action in the long or short term, the short term often meaning the same action but spread over a shorter period of time. Now these two blindnesses are not regulated in the same manner. Against endured blindness, there is information, but against willed blindness, there are only more or less elaborate forms of repression. 1 Competitive organisation of financial systems and economic viability 1 Systemic introduction to financial integration The notion of development introduces a qualitative dimension to that of change because it takes into account changes of structure following an irreversible process (J.L. Gaffard, M. Amendola; 1995). To be interested in economic development is therefore in particular to look at the structures for financing productive activity, at their interaction with the productive system and at the dynamics that result from it insofar as one considers, in the continuity of the Credit View, that one is dealing with open systems, that is to say productive systems sensitive to changes in financing conditions, or more precisely a entangled hierarchy (in the sense of Hofstadter (1986) cited by O. Godard; 1996): Not only does the identity of the environment of the productive system - here the organisation of the financial system and more precisely the form of bank-firm relations - proceed from its relation to the system (self-referential logic) but also the environment as an encompassing reality from which coherences, regulations but also threats and imbalances emanate imposes itself on the system as a condition of its existence and survival (hetero-referential logic). Let us note indeed that for standard economic theory (the Money View), there exists only a relation of causality from the environment towards the system via an exogenous shock on technical progress, consumers' tastes or the money supply, without any possible feedback effect since these causes are assumed exogenous by hypothesis. In these conditions, nothing makes it possible to explain why the economy finds itself durably outside the equilibrium situation. Any deviation from the equilibrium growth rate can only be momentary. Furthermore Stiglitz and Weiss (1981) show that there may exist credit rationing equilibrium situations due to the non-linearity of credit supply: While one may consider a priori that for any credit granted, the bank's profit increases with the interest rate, on looking more closely one must also consider that, with equal expected income from projects, when the interest rate rises, the least risky projects are not taken up by borrowers; which increases the risk of bankruptcy and therefore the bankruptcy costs which, beyond a certain interest rate r*, will be greater than the additional gains resulting from a rise in the interest rate (see graph below). One consequence of the existence of credit rationing equilibria lies in the fact that not only does a variation in the money supply not necessarily entail a variation in the interest rate and therefore a variation in the level of economic activity, but the level of economic activity may vary without there necessarily being a variation in the interest rate, if the occurrence of rationing equilibrium increases. For the Credit View, in the continuity of Kalecki and Keynes, insofar as the influence of the financial sphere on the productive system is exerted via the banks' assets, money constituting only the counterpart of the credits granted, there exists a feedback effect of the financial system on the financial sphere via the profits which determine the effective repayments and the risks of default of the borrowers, and hence the conditions of future credit supply (availability and cost of credit). Information asymmetry models show indeed that without information asymmetry distinguishing lenders and borrowers, the economy is in the standard framework described above. Whereas in the event of information asymmetry allowing borrowers to develop opportunistic behaviours, the guarantees demanded by lenders play an essential role in the conditions of credit supply, introducing a feedback effect between the system and its environment. Thus one can account for cumulative phenomena characteristic of systemic risk calling into question economic viability: On the one hand profits are determined by the availability of credit via the level of investment, and on the other hand, the cost of credit constitutes a burden weighing on profits, the latter affecting in return the conditions of future credit supply. Thus when borrowers' profits fall, repayment defaults increase and banks adopt more restrictive conditions for selecting borrowers (in cost and in quantity) which in turn affect their profits. To summarise, the feedback effect liable to generate cumulative phenomena characteristic of economic and financial cycles and crises therefore follows two possible paths, which are: - The banks' assets. If their claims are of poor quality, banks must increase their provisions and harden their conditions of credit supply in order to maintain their profit margin. - The borrowers' balance sheet. Rather than seeking to determine a state-contingent contract which raises problems of monitoring costs, the bank may have an interest in writing a very simple contract by selecting borrowers according to the guarantees provided. If the borrowers' financial situation deteriorates, the value of their guarantees falls and their access to credit is restricted. The importance of this feedback effect depends on the level of information asymmetry existing between lender and borrowers which may generate opportunistic behaviours on the part of the latter justifying the collateralisation of credits and the hardening of financing supply conditions. It therefore appears that the nature of lender-borrower relations (or bank-firm relations) plays an essential role as a condition of economic viability according to whether opportunistic behaviours on the part of borrowers can (or cannot) develop there, as the Kreps model set out in the first part showed us. It is on the one hand to the determination of the conditions for the development of opportunistic behaviours linked to competitive conditions and on the other hand to their influence on economic viability that we shall devote the second part of this paper. To this end, we will first use the contributions of transaction cost theory to establish an interpretive framework enabling us to analyse the specific features of financial systems with regard to competitive organisation and their role in terms of viability. 1 The contribution of transaction cost theory Transaction cost theory provides us with the first elements of an interpretive framework for determining the appropriate coordination structures according to the characteristics of the transactions envisaged. 1 The attributes of transactions For O. Williamson (1986), the attributes of transactions are: - the uncertainty of the environment - the frequency of transactions - asset specificity Asset specificity refers to the geographical and temporal ordering of transactions: they may be transferable (high reversibility) or dedicated, that is to say designed for a particular use in terms of location, human resources or, more particularly, when they are intangible assets corresponding to know-how resulting from learning phenomena. A specific investment is intended for a certain type of transaction and would lose value for other uses. The producer, like the customer, is therefore locked to a certain extent into the pattern of the transaction, and the parties to the exchange have an interest in the transaction having a certain continuity. One can therefore speak of commitment, with which the notion of irreversibility is associated. The stronger the specificity, the less the market is the efficient space of exchange and the more the organisation makes it possible to reduce transaction costs. When assets are reserved for a particular use (or a particular customer), the firm fully realises savings if it produces "in-house". There is therefore a link between the frequency of transactions, asset specificity and uncertainty, represented by the diagram below: Specificity determines an irreversibility that favours the renewal of exchanges and reduces uncertainty: in the model presented above, if cooperation between the agents requires investment in specific assets (in the sense that they can only serve cooperation between these agents), each agent's incentive not to cooperate decreases as its investment grows. In this sense, investment in specific assets promotes cooperation, in the same way as traditional ties, justice or reputation. Conversely, the frequency of exchanges reduces uncertainty and encourages the actors to commit to acquiring irreversible assets. Whereas uncertainty does not encourage investment in specific assets. 1 Alternative forms of coordination The theorists of the transaction cost approach (Coase, O. Williamson) consider two alternative forms of coordination between agents corresponding to the attributes of transactions we have just described: the market and the organisation. The organisation responds to the need to save transaction costs resulting from the incompleteness of contracts and the opportunism of behaviour, in that it makes it possible to introduce an institutional structure for monitoring and following up contracts. To prevent opportunism, agents set up systems of surveillance and conflict regulation (punishments) that are inseparable from surveillance. The quality of supervision depends on the qualification of the supervisor: the specialised supervisor appears more efficient than the judicial system, since the latter's intervention presupposes that the proof of any deviance is not only observable but also verifiable, that is to say that it takes an objective form. Thus, for O. Williamson, the choice of the optimal mode of coordination rests on a trade-off between the degree of uncertainty and the greater or lesser reversibility. Markets are more appropriate when information is well structured or defined in terms of risks; that is to say, when it is possible to establish complete contracts. The organisation, by contrast, has an advantage for ex ante evaluation and for the forms of subsequent adjustment that require ex post follow-up and control resulting from uncertainty. 1 Juxtaposition of a second interpretive framework We take the view that the Walrasian market is the most centralised form of coordination in existence. Because of the existence of the Smithian "invisible hand" or the omniscient Walrasian auctioneer, the competitive market structure constitutes a particularly elaborate organisational form. This is why it seems useful to us to superimpose on the first interpretive framework a second one borrowed from Aoki (1986). Aoki contrasts two models of organisation: - the American-style hierarchical model H, which rests on a hierarchical separation between design and execution - the J model, characterised by horizontal coordination between operating units and the sharing of ex post information gathered from experience acquired: this model corresponds to the decentralised Japanese-style firm, which operates through neighbouring workshops in an upstream-downstream chain of supplier-user relations. Horizontal organisation in lasting temporary ties favours learning by making it possible to better identify unforeseen events, and allows a faster reaction by avoiding the hierarchical route. Hierarchical centralisation is coupled with market-based management, which constitutes the firm's environment; this prevents the accumulation of learning, since resources are not stable. Whereas the environment of the J firm, made up of subcontractors and a customer base, is more stable and favours learning. In the case of the hierarchical model, efficiency gains are drawn from specialisation. Any unforeseen event is handled by means defined a priori. Whereas, owing to the strong flexibility of the organisation in the context of horizontal coordination, Aoki shows the superiority of the J model when the environment is changing. In summary, horizontal coordination proves more efficient than vertical coordination in the face of uncertainty, owing on the one hand to its capacity to stabilise the environment and on the other hand to its capacity to adapt to that environment. If one considers that the activity of firms is made up of current production and innovative activity, current production is favoured by the hierarchical structure, which favours static efficiency. By contrast, the efficiency of creation and innovation requires a more horizontal organisational structure in order to increase the possible new combinations of development. To summarise, the choice of the mode of coordination can be represented on the following graph: Let us note, to complete the picture, that the choice of the organisational structure (or of the J model) results from the fact that it reduces uncertainty by introducing an irreversible process; which is not the case of the competitive market structure (or H model). This introduces a dynamic dimension to this interpretive framework. 1 Application to the study of the specific features of financial systems With regard to credit relations, uncertainty can be equated with the information asymmetry existing between lenders and borrowers. The higher the information asymmetry, the more the borrower is encouraged to adopt opportunistic behaviour, because on the one hand the verification costs for the lender are high, and on the other hand the probability that the lender will be punished decreases. Selection methods therefore matter, as does the nature of the relations that may be established between lenders and borrowers outside the credit relationship. We will see moreover that the methods of selecting borrowers and the nature of these relations are interdependent. 1 The forms of borrower selection Insofar as the information asymmetry between lenders and borrowers influences the latter's opportunistic behaviour and thereby their risk of default, the information held by lenders about borrowers constitutes an essential asset of their patrimony. Yet this information depends on the method used by banks to select potential borrowers. There are two specific forms of borrower selection: - The scoring technique, which consists in selecting borrowers by mass computer processing on the basis of ratios that are significant as to their present and future solvency. It therefore applies indiscriminately to all potential borrowers, and thereby limits investment in specific assets. A bank of this type uses public information and forms an opinion with the help of previously established criteria of financial analysis based on the law of large numbers, which it applies identically to every potential borrower. Uncertainty is assumed to be given (exogenous), and therefore probabilisable. Information asymmetry being assumed to be given, the aim is not to remedy it but to guard against it. In this framework, since banks hold no private information, they play an intermediation role identical to that exercised by the capital market. This method limits the irreversibility of the credit relationship: if the bank decides not to renew a loan, it loses nothing since it holds no specific information on the customer. - The risk analysis technique, which consists on the contrary in seeking to obtain specific information on each customer. The decision is no longer solely a function of a pre-established calculation, but also rests on the assessment of the quality of the managers, on an analysis of the future of the industry, and on the validity of the firm's strategy. Unlike the first method, the aim is to produce information by means of an investment that will be lost if the relationship is interrupted. Renewal of the relationship attenuates this irreversible cost, since it is specific to each customer. This cost will only be incurred if the borrower's opportunism, which generates uncertainty, is assumed to be limited. And conversely, because it makes it possible to produce information on the borrower and its project, this investment in information reduces the borrower's incentive to adopt opportunistic behaviour, and thereby the systemic risk at the macroeconomic level if one refers to the models presented above. By reference to transaction cost theory, when the bank-firm relationship corresponds to the first case, we will equate this relationship with a market relationship (or with the H model). Whereas we will equate the second case with an organisational-type relationship (or with the J model). 1 Competitive specificities of the American, Japanese and German banking systems The borrower selection technique is generally used in Anglo-Saxon countries. Whereas the risk analysis technique is particularly used in Germany, and even in Japan. German banks have very high-performing economic forecasting departments, engineers specialised by industrial sector, and management consultants who mobilise on each file. Let us note moreover that it is in Anglo-Saxon countries that the practice of collateralisation is most widely used. Yet we pointed out above that the greater the information asymmetry between borrowers and lenders, the more important a role guarantees play in the conditions of credit supply. To what extent do these financial systems differ in competitive terms, and to what extent do these competitive specificities interact with the borrower selection techniques set out above? It is these questions that we will seek to answer, by insisting on certain characteristics that differentiate the competitive organisation of financial systems. 1 Banks' equity stakes in the capital of industrial firms In a context of information asymmetry, equity stakes have several effects: - On the one hand, they constitute a commitment on the part of the bank in specific assets corresponding to a lengthening of the time horizon of the actors: because of the existence of these costs that cannot be recovered (in the short term), the bank tends to help the firm in difficult times, and the firm can envisage long-term strategies. - On the other hand, they in turn allow banks better control of the moral hazard problem. Their access to information is made easier by the fact that they are shareholders. J. Stiglitz (1985) shows that a participation by banks in the capital of firms makes it possible to increase the reliability of the control system, and therefore to improve the behaviour of firms with regard to the collective interest. - Lastly, being shareholders, they can voice their opinion by taking part in the borrower's industrial decisions, bringing their skills, rather than voting with their feet. In the United States, the members of the FED may not hold equity stakes in the capital of firms. Financial institutions hold 24% of the shares of firms against 43.5% in Japan, and commercial banks 0% against 20.5% in Japan. In Japan, the influence of banks over industrial companies runs through the channel of a participation in their capital. The large closed groups (Zaibatsu) re-formed unofficially after their dismantling in 1945 in the form of conglomerates of almost identical structures: the "keiretsu". Although Japanese anti-monopoly law limits the quota of shares held by banks to 5% of the capital, the organisation in pools makes it possible to circumvent this law, the so-called secondary banks delegating their power to the principal bank, the "Main Bank", which plays the role of Lender of Last Resort to customers in a situation of temporary illiquidity. When a firm is in difficulty, the reputation of the main bank being at stake, rescue, reorganisation or liquidation occur more quickly in the event of difficulties. Hoshi, Kashyap and Scharfstein (1990) observe that main banks have the ability to discipline firms, in exchange for which they provide them with "insurance" during periods of crisis. The influence of the main bank is thus based on long-term, multidimensional financial ties and on advisory activities for affiliated companies. Similarly in Germany, although banks hold only 10% of stock market capitalisation in direct holdings, small shareholders delegate their power to them in order to be represented on boards of directors. German banks sit on the Board of Directors of 318 of the 400 largest German firms and they also sit on the Supervisory Boards. This gives them an important power of monitoring and control. 1 The principle of the universal bank Whereas in the United States the Glass-Steagall Act maintained the compartmentalisation of banking activities, the German banking system is characterised by the predominance of universal banks that carry out all kinds of financial operations: credit, deposit collection, securities transactions, equity stakes... This is the "Hausbank" (house bank) model, which assumes that the lender can satisfy the borrower across a complete range of financial services. By achieving economies of scope on activities joint to credit activities, banks secure a competitive advantage by acquiring information on the potential borrower. Thus the provision of means of payment enables the bank to obtain information on the cash position. Thus the Anglo-Saxon competitive market model is contrasted by the German and Japanese models, which are distinguished by customer relations of commitment and cooperation that are long-lasting. One can speak of commitment relations when the bank holds a share of the borrower's capital, of customer relations when the lender retains the borrower's loyalty by offering it joint services, and of cooperation when the lender takes part in the borrower's decisions by bringing it know-how. In these three cases, an organisational irreversibility that reduces uncertainty is introduced within the lender-borrower relationship, corresponding to an investment in specific assets. The bank draws its gains from the ex post transaction cost savings resulting from the circulation of information, learning and the organisational flexibility characteristic of the J model of horizontal coordination. Whereas in the first case the bank draws its gains from the standardisation of procedures, which makes it save ex ante transaction costs through mass processing, in accordance with the H model of vertical coordination. There is no longer any question of investing in assets whose benefits will only be obtained in the long term. Banks are all the less encouraged to acquire information on borrowers because they know that the latter risk subsequently turning to another institution, and, on the other hand, they are all the less encouraged to prolong the relationship as the unrecoverable costs (the investment in information) are low. 1 Competition and viability The choice between the two series of models we have mentioned joins the debate on the efficiency/stability dilemma. While some consider that competitive market organisation may generate instability, it is nevertheless supposed to promote efficiency in the allocation of resources. We will see that this is not necessarily the case; in other words, that the conditions of banking competition determine a zone of economic viability which, contrary to what economic theory asserts, is all the narrower the closer one gets to the competitive framework of the market. 1 The question of efficiency In the case represented by the Anglo-Saxon model, firms play banks off against one another, which does not encourage banks to invest in dedicated assets by acquiring specific information, and opportunism prevails. Whereas in the case represented by the Japanese or German models, borrowers are less encouraged to take advantage of their private information, or even to undertake projects with too high a probability of default, because they cannot play on competition to obtain a loan if it is rationed. Moreover, by producing information, banks have a greater capacity to distinguish good borrowers from bad ones and, in the latter case, they can bring their know-how to bear to turn the situation around. Thus, for a given interest rate, the profits of banks operating within the second type of model will be higher than the profits of banks operating within the competitive market framework. On the graph below, borrowed from Stiglitz and Weiss (1981), if the solid curve in the south-east quadrant represents the profit curve of competitive banks, the profit curve of banks operating within a commitment or customer relationship will be further from the x-axis (the dotted curve). As a result, there will also be an upward shift of the credit supply curve (north-east quadrant). This means that the conditions of credit supply will be less strict: the occurrence of credit rationing will be reduced and, for a given amount of credit granted, the interest rate will be lower than in a competitive market situation. This situation, which may seem paradoxical, results from the fact that banks have invested in specific assets. Moreover, as we saw in the introduction, the more the economy is in a competitive market situation, the greater the feedback effect. The role of guarantees increases and the conditions of financing supply become more drastic, reducing the zone of economic viability. 1 The question of stability In the case of the Anglo-Saxon-type model, competition pushes firms - in this instance financial institutions - to reduce their profit margin in order to retain, or even increase, their market share. This is the principle of marginal cost pricing, which must lead to zero profit. The result is greater sensitivity to cyclical shocks and a shortening of the time horizon. Yet the apparatus of financial analysis rests on very subjective criteria, as M. Aglietta (1991) remarks: "The evaluation of a debtor's solvency by a particular bank depends on competing banks. It is an irreversible process." R. Rajan (1990) builds a model showing that the managers of banking firms, if they have a short-term horizon, have credit policies that influence and are influenced by the other banks. Competition risks generating an overexpansion of the banking system and risky behaviour to retain market share (destructive competition). To cope with the fall in their profit margins resulting from the intensification of competition, American banks engaged in real estate lending, which rose from 15.7% of their assets in 1985 to 23.5% in 1990. Between 1984 and 1992, 1,296 American commercial banks had to merge or be liquidated, at a cost of more than $300 billion. As D. Peguin (1993) remarks, "the very restrictive legislation in the United States is no doubt not unrelated to the characteristic often attributed to American firms, of having short-term management incompatible with the pursuit of a profit-maximisation objective... In these conditions, there are strong grounds for observing in the United States the existence of a quasi-structural link between the level of indebtedness of firms and the level of their output." D. Rivaut-Danset (1993) observes in France a standardisation of credit-granting practices since the deregulation that took place in the 1980s: "Many firms have turned, as we know, towards financing procedures providing greater flexibility and towards the least costly sources, banks choosing to multiply customers and per-transaction operations." SOFARIS has calculated that, statistically, in recent years the annual losses of all French credit institutions on the SME sector have risen sharply to represent on average 2.5% of their outstanding credit. Yet in France, borrower selection techniques differ depending on whether the customer is an SME or a Large Enterprise. In the second case, the bank seeks to retain its customers' loyalty and relations are personalised. Whereas it generally practises scoring with an SME. Japanese deregulation seems to have had similar effects by intensifying competition and loosening the privileged ties that existed between banks and industrial firms. In their 1989 study, Hoshi, Kashyap and Scharfstein observe that deregulation in Japan, having increased competition between banks, weakened their financial health and reduced their capacity to support firms in their industrial development, making them more sensitive to the economic situation. To cope with the fall in their profit margins, Japanese banks turned to speculation. In 1994-1995, the twenty-one largest national banks had to write off 300 billion francs of irrecoverable debts. These results suggest that banks subject to the law of competition have shown "myopia". The myopia of banks is characterised by discontinuities of behaviour (Minsky 1982, 1985; Aglietta 1991): when banks are below their perception threshold, they compete fiercely to offer credit on very favourable terms and by taking reckless risks, which lead to a situation of financial fragility that increases the system's sensitivity to shocks. The myopia of banks has therefore led to a reduction of the zone of viability by increasing the instability of markets. 1 Conclusion As we have been able to observe, the specificity of the Anglo-Saxon model rests on the existence of regulations. The banking system based on competition can only function and develop thanks to legal restrictions on banking activity. This joins the criticism that can be levelled at transaction cost theory: if the organisation constitutes an alternative mode of coordination to the market, the market remains the reference as a mode of coordination. The organisation exists insofar as the market has defects, in particular if it is not a competitive market. Yet we have seen that for there to be a competitive market, a third-party outside intervention is required, as we showed with Kreps's model. The post-war period was marked on the one hand by the success of German and Japanese industrial development and on the other hand by Anglo-Saxon economic and financial instability, and there are therefore reasons to think that the conditions of banking competition are not unrelated to the determination of the zones of economic viability of these countries. It must be stressed, however, that economic development is not an end in itself, far from it. It has value only to the extent that we grant it value, and it is highly likely that things would be quite different if growing abundance were no longer for individuals an objective towards which they turned in a competitive manner. The legal system, ideologically neutral (it is a vector, not a component), is the only tool that makes it possible to arbitrate the various possible conflicts of influence between the political, economic and social spheres. Often poorly perceived because it generates constraints, it is nevertheless the only tool enabling actors informed at the global level to influence actors driven by competition and trusting in the spontaneous regulation of the market. The conflict that then ensues is comparable to a conflict of ideologies. Social constraints are set against economic constraints and environmental constraints, and it is brought out that local regulators (market, invisible hand, predation) produce global counter-effects. Within this permanent tension, the legal sphere alone must ensure a balance by proposing to everyone goals that seem to them sustainable in terms of development, that is to say objectives that appear attainable. As an example of a principle generating its opposite, one can imagine an ideology based on personal profit without harm to others or to the natural environment, which organises a kind of sympathetic competition such as Adam Smith might have conceived it, based on specular self-control. A fixed point of this type is conceivable, viable and probably durable, but hardly sustainable at present. How can it be globalised? Through incentive, education, or constraint? What threshold will have to be crossed for agents to emerge from their blindness and project themselves specularly from the short to the long term? Doubtless when the future feeds back on the past, but for that we would have to lose our linear vision of time and rediscover our ancient cyclical conception... And what will become of the viability of modes of coordination alternative to the market in the face of worldwide globalisation? Contents