2. Protectionism and subsidies in evolutionary biology

In her book The Ant and the Peacock, Helena Cronin observes:

Chapter nine of Darwin's (Reference Darwin and Burrow1859) Origin of Species was ‘On the Imperfection of the Geological Record’. Ever since Darwin, this ‘imperfection’ has been used to reconcile the missing steps in phylogeny with the gradual nature of the adaptation predicted by the Darwinian theory. The fundamental riddle in the Darwinian theory seemed to be that evolution should be continuous and related to the intensity of natural selection. Unlike the smooth changes of simple bacteria evolving in a laboratory, fossils of complex organisms provided increasingly compelling evidence of long periods of stasis and phases of rapid evolutionary change. In the case of complex organisms, missing evolutionary links seemed to pose a major problem for Darwinian theory.

However, recent research has shown that, for complex organisms, the puzzle may have been rather different and, indeed, of almost opposite nature. Even when ‘complementary’ mutations could improve the fitness of the genotype, the epistatic interactions among the gene loci imply that the genotype is characterised by a built-in inertia.Footnote ³ If the blind nature of evolution involves a succession of single mutations, natural selection would eliminate those genotypes that undergo only one of the complementary changes required for the adaptation of the species.

If natural selection has operated for a sufficient length of time, each evolutionary unit is likely to be optimally adjusted to the other complementary units. In these conditions, single mutations are likely to be less fit and be eliminated by ‘normalising’ selection. Natural selection would stabilise the species and keep it at a ‘local’ fitness peak (even when higher peaks exist). Because of epistatic interactions (or, in the economist's language, complementarities), long periods of stasis may characterise the evolution of complex organisms.

In the case of complex organisms, what must be explained is how the stabilising force of natural selection can be, eventually, overcome. There are two main routes through which major evolutionary change may take place. The first route is allopatric speciation, which is analogous to a form of unintended economic protectionism. The second is selection complementarity, which finds its counterpart in the economist's language as unintended subsidies coming from different selection domains.

3. The emergence of new organisational species

A simple definition of an organisation of production can be based on two factors. The first is its technology and, in particular, the technological characteristics of the resources used in production. The second is the set of rights (which may be legal rights and/or customary rights supported by social norms) on the resources employed in the organisation and on the organisation itself.

The relationship between these two factors has traditionally been a controversial issue in social sciences: if causation exists, it can operate in two directions. On the one hand, property rights can be seen as factors shaping the nature and the characteristics of the resources used in production. On the other hand, the technological characteristics of resources employed in production can be considered to be the causes of changes in the system of property rights.

This two-way relationship was at the very root of Marx's theory of history and of his view of the firm. It was the source of interesting problems and contradictions within this theory. Marxist analysis has often oscillated between ‘technological determinism’ (technology invariably gives rise to a unique set of property rights) and ‘property rights romanticism’ (alternative property rights can invariably bring about an alternative technology).Footnote ⁶ Moreover, as Hirschman (Reference Hirschman1981) observes, Marx ‘oscillated between the grand generalisation characterising an entire epoch or process and the discriminating analysis of events which made differences between countries and subperiods stand out in richly textured detail’.

In spite of these contradictions and limitations, the two-way relationship considered by Marx is still an important key for the understanding of alternative organisations, and it is difficult to disagree with John Hicks when he maintains that when we consider ‘theories of history’ ‘there is so little in the way of an alternative vision which is available’ (Hicks, Reference Hicks1969: 3).

The relationship between property rights and the characteristics of productive forces, which created so many interesting problems and contradictions (as well as so many wrong ‘predictions’) in the Marxian approach, became a non-issue in neoclassical theory. In a market economy, ownership by workers or capitalists would have had no effect on the characteristics of the resources (or of the productive forces) employed by the firm. At the same time, the characteristics of the resources employed in the firm would have no impact whatever on the form of ownership which was going to characterise the firm.

Both new institutional and radical economists have reconsidered the interaction between rights and technology. However, the relationship between rights and technology is still very controversial. In these two streams of literature the causality runs in opposite directions. In new institutional economics, rights are endogenously and efficiently determined by the characteristics of the resources employed by organisations:Footnote ⁷ namely their degree of specificity and their monitoring requirements. By contrast, in the radical literature the characteristics of the resources employed in the firm are in their turn determined by the rights which owners of different factors exercise on the organisation.

In spite of their differences, the lines of inquiry pursued by new institutional and radicals are not necessarily incompatible, and they can be integrated into a framework that considers the two-way relationship between property rights and technology.

The ‘radical direction’Footnote ⁸ of causation runs from property rights to technology. It is argued that the specificity and monitoring characteristics of the resources are due to the nature of the property rights under which they are employed. For instance, individuals working in organisations where they do not have rights are likely to be characterised by a relative underinvestment in organisation-specific skills and by an unfavourable distribution of asymmetric information attributes that makes them ‘easy to monitor’. Specific and difficult-to-monitor workers are high-agency-cost resources that are expensive for the present owner, who has an incentive to replace them with low-agency-cost resources. By contrast, no similar replacement occurs for the individuals with rights on the organisation: the alignment of their objectives with that of the organisation allows a considerable saving on the high agency costs that would have been paid if they were employed by other agents.

When we leave the neoclassical world with zero agency costs, the ‘radical direction’ of causation can be justified by using a fundamental principle of economic theory: that profit-maximising employers tend to replace high-cost factors with low-cost ones. This point becomes evident when one considers that a change in property rights from one factor to the other also changes the relative costs of employing those factors. The new owning factors will save on their own agency costs while they will pay the agency costs of employing the former owning factor (while this cost was saved in the former ownership arrangement). Thus profit-maximising entails that property rights influence the combinations of productive forces that are going to be adopted: the optimal technology is bound to change with different assignments of property rights.

Also the ‘new institutionalist’ directionFootnote ⁹ of causality, running from the nature of technology to property rights, can be easily understood by using another fundamental principle of economic theory: namely that, like other economic goods, organisations tend to be owned by those individuals in whose hands they are more valuable. This implies that, for each combination of resources employed in production, property rights should go to those individuals able to save the most on agency costs when they own the organisation: these are the high-agency-cost factors that involve higher monitoring and specificity insurance costs when they are employed by other people. Thus for each combination of resources employed by the organisation there is an optimal assignment of the ownership rights on the organisation.

Consequently, in a world of positive agency costs there is an optimal technology for given ownership rights on the organisation and an optimal set of ownership rights for a given technology that is employed by the organisation. Using biological terminology, we may say that organisations are characterised by epistatic interactions between rights and technology. Complementarities do not only characterise the world of nature; they also shape the world of institutions.

Like the frozen part of a genotype, the interactions between technology and property rights have a built-in inertia.Footnote ¹⁰

An organisational equilibrium is defined by the fact that technology is optimal relatively to property rights, and property rights are optimal given the technology that is employed. The self-sustaining nature of organisational equilibria derives from the fact that owning factors saving on their own agency costs tend to choose a technology characterised by high intensity of their own high-agency-cost factors – that is, a technology under which their ownership is optimal.

The analogy between the epistatic relations characterising natural species and the characteristics defining an organisational equilibrium must, of course, be taken with some caution. Human learning may allow patterns that are not permitted to genes. On the other hand, the concept of organisational equilibria already entails a considerable degree of rationality. The optimality of technology given property rights, and that of property rights given technology, defines a Nash equilibrium; this is tantamount to assuming that financiers are able to choose the optimal owners for each firm characterised by a certain given technology and that production managers are able to choose the optimal technology for a certain given ownership structure. Indeed, it is reasonable to assume that this equilibrium is achieved by an evolutionary process by which firms that have suboptimal technologies given the ownership structure, as well as those that have suboptimal ownership structures given technology, are gradually eliminated by competitive forces.

In natural selection, the pressure of competition helps select the best members of a given species; however, we have seen that the effects of natural selection on speciation are much more controversial. Our question is related to the case of speciation: we are not asking whether competition can select the best member of a given species of organisation but whether it can help the formation of a new more efficient species of organisation characterised by different technology-property rights genotypes.

We have seen that complementarities imply that complex species are characterised by important development constraints: the fitness of each mutation is constrained by the other characteristics of the species. This implies that many evolutionary paths may be blocked. Unfortunately, in the case of organisational equilibria, these obstacles may work exactly against those changes that may otherwise lead to the formation of a superior species of organisation. Suppose that there are some efficient alternative potential owners that could obtain a higher ownership rent than the present owners. These alternative owners are efficient because their employment by the present owners involves very high agency costs that could be saved if they own the organisation. For this reason, the factors of the potential alternative owners are promptly replaced by factors that are cheaper for the present owners. In other words, an anti-speciation mechanism is embodied in each species of organisational equilibrium and it has the unfortunate characteristic that its strength is related to the efficiency of the alternative potential species.

However, suppose that this anti-speciation factor is overcome and one of the characteristics of the old species mutates into a new species of organisation that is potentially more efficient. For instance, some organisations are characterised by new property rights that, if they were coupled with the associated optimal new technology, could form a new more efficient organisational equilibrium. Until this new technological combination is developed and employed, we will have a situation of organisational disequilibrium or, in other words, an inferior hybrid between the new property rights and the old technology. If the pressure of competition by the members of the old species is strong, the hybrid is likely to be wiped out before it has any chance of turning into the new superior species. Or, in other words, the epistatic interactions between property rights and technology imply the existence of a rugged multi-peaked fitness landscape; in these conditions the pressure of competition will act to keep the firms at the local peaks.

However, even if speciation is successful, the survival of the new species may be put at risk by tough competition raised by the old species.

In the first place, if there are few members of the new organisational species, interbreeding with the many members of the old species will be very frequent and will produce numerous inferior hybrids. In these conditions interbreeding may lead to the extinction of both mutations. When the new technology is imitated and run under the old property rights system, it turns out to be inferior and, vice versa, when the new rights are influenced by the old technology they also turn out to be inferior.

Secondly, in nature, the efficiency of each species depends on its frequency. Also, organisations have this characteristic. For instance, network externalities in property rights and in technologies may imply that few firms characterised by different organisational equilibria are not viable: they would be outcompeted by firms that, even if inferior when they exist with the same frequency, can benefit better from network externalities because of their present large number.

Thirdly, as Darwin pointed out, more numerous species may enjoy more mutations. Also organisations that are more numerous will have this advantage for the non-frozen part of their characteristics, and this will again constrain the possibility of major evolutionary changes.

In conclusion, because of interlocking complementarities, we should expect the formation of new organisational species to require allopatric conditions or major external subsidies. Major evolutionary change requires that the members of the new species must somehow be protected against competition by the members of the incumbent species. Alternatively, institutional and technological shocks should be strong enough to overcome the inertia of complementarities and rely on some external subsidy originating from a different selection domain. In the following two sections, we will give historical examples of each of these two possible cases.

4. The allopatric emergence of managerial capitalism

In the last half of the nineteenth century there ‘came into being a new economic institution, the managerial business enterprise, and a new subspecies of economic man, the salaried manager.Footnote ¹¹ With their coming, the world received a new type of capitalism – one in which the decisions about current operations, employment, output, and the allocation of resources for future operations were made by salaried managers who were not owners of the enterprise’ (Chandler, Reference Ostrom1990: 2).

According to Chandler, the advent of the new institutions and the ‘new subspecies’ of economic man were strictly related to the building and operating of rail and telegraph systems. The complexities of their operations required firm-specific organisational capabilities that could not have been developed within the members of the family owning the firm, nor efficiently monitored and controlled by them. The new firms required a managerial hierarchy whereby to a great extent salaried managers controlled other managers. In other words, the new technology required the employment of high-agency-cost managerial skills. In turn, this required that rights, incentives and safeguards were to be given to these managers. In particular, it was vitally important for managerial effort, as well as for the efficiency of the firm, to know that promotions from the low to the high positions in the managerial hierarchy would be related to achievements and unrelated to family and other social ties.

The new system first came into being in the rail and telegraph industry, but it exhibited greater efficiency in many of the industries characterised by economies of scale and scope that could be efficiently exploited by the use of managerial hierarchies. Indeed, the diffusion of the new organisational model characterised the advent of a new species of capitalism: managerial capitalism. The new species had two local varieties: competitive managerial capitalism in the USA and cooperative managerial capitalism in Germany. While in German industries family control lasted longer than in the USA, in both countries salaried managers with little or no equity in the enterprises for which they worked participated in decisions concerning current production and distribution, as well as in the planning and allocation of resources for future production. ‘The greatest difference, however, came in interfirm and intrafirm relationships’ (1990: 395). In the USA the new managerial firms competed aggressively for market share and profits, and the anti-monopolist legislation reflected a shared belief in the value of competition. By contrast, in Germany many firms preferred to cooperate, and trade associations played a much larger role in Germany than in the USA.

While the advent of these two subspecies of managerial capitalism made the USA and Germany the two most important actors of the second industrial revolution, Britain – the main actor of the first industrial revolution – lagged behind in many of the new industries. Britain continued to be committed to the species of personal capitalism that had been so successful at the time of the first industrial revolution. Whereas long-term profits based on long-term growth were a goal on which the managers and the major investors of the American and German managerial firms could agree, the families owning the British firms often preferred to pay out earnings as dividends rather than using them to make the extensive investments required to move into foreign markets or to develop new products in related industries. ‘Because their firms grew slowly and because they hired only a small numbers of managers, the founders and their families remained influential in the affairs of the enterprise and so affected dividend policy.’ (1990: 595). By contrast, the long-term growth of American firms helped the managers to gain strong job rights in their firms. ‘Such a goal not only helped to assure tenure for the senior executives, but it also enhanced the opportunity for advancement for the more junior managers.’ (ibid.).

British firms did not provide similar opportunities to non-owning managers. The key managerial positions were usually reserved for the owning family. Social and family ties were more important than competence in advancing up the managerial ladder. There were few opportunities for junior managers, while no job security similar to that in German and American firms could be given to senior executives. It is hardly surprising that organisational capabilities of such importance for the firms of the second industrial revolution stagnated. As a result, Britain lost world economic leadership to the countries that had speciated the new form of managerial capitalism.

Using the terminology of the preceding sections, the advent of competitive and cooperative managerial capitalism can be seen as a form of allopatric speciation: some form of protectionism helped the infant industry of managerial capitalism. The speciation did not occur in Britain, where the competitive strength of personal capitalism was the strongest, but in countries that had not participated in the first industrial revolution. In the USA and Germany the nature of the productive forces required by railways and the telegraph industry first changed the structure of rights that characterised these sectors and then, after a few years, that of the other industries where the growth of productive forces could benefit from the change. Soon, the new rights favoured the employment of specific and difficult-to-monitor (and, therefore, high-agency-cost) managers, while their employment had a self-reinforcement feedback that made the new rights of managerial capitalism an irreversible choice.

Thus, the speciation of new organisational equilibria occurred in the USA and Germany where managerial rights and managerial skills were fundamental characteristics of new epistatic interactions between production relations and productive forces. On the one hand, given tenure rights for senior executives and fair promotion opportunities for junior managers, productive forces were best characterised by a relative high intensity of high-agency-cost managerial skills. On the other hand, given the employment of these skills, only a system of strong managerial rights could ensure the commitment of managers to the organisations and save the high agency costs that should have otherwise been paid in case of pronounced incongruence between their goals and those of the organisation.

In Britain, by contrast, the second industrial revolution was unable to break the complementarities between the structure of rights and the nature of productive forces. By that time, the managerial skills accumulated by the British families had already made some of their members high-agency-cost factors. This reinforced the stability of their ownership rights: their ownership saved on their own high agency and had a feedback effect on technology involving the continuation of the intensive employment of the high-agency-cost skills of the owning family members.

The vicious circle between managerial skills and rights mirrored the self-reinforcing interactions that we have just considered. Since managers were not involved in the organisation, their agency costs could not be saved like those of the committed family members. For this reason, the former were often replaced by the latter; but this caused even further damage to managers’ commitment to the organisation and their accumulation of organisational skills. At the same time, the low intensity of high-agency-cost managerial skills implied that the (partial) saving of these agency costs (that would have occurred under managerial capitalism) could not favourably compare with the saving of agency costs of family members (that characterised British personal capitalism).

Thus, in spite of the technological shocks induced by the second industrial revolution, the frozen part of the genotype of the British firms did not melt. The fairly strong competition which characterised the British environment favoured the organisational ‘stasis’ of British capitalism. The emergence of managerial capitalism required allopatric conditions or some protection against the competitive pressure of family capitalism. The major role of Britain in the first industrial revolution became a paradoxical disadvantage in the environment of the second industrial revolution.

5. Post-war Japan and Italy: the role of unintended subsidies

In pre-war Japan, large firms were organised in zaibatsu such as Mitsubishi, Mitsui, Sumitomo and Yasuda. The zaibatsu were mostly family controlled and originated from a family firm. For instance, the Mitsui and the Sumitomo were two of the most powerful merchant families in Tokugawa Japan. However, in the tradition of the Japanese family system, the entire emphasis was placed on perpetuation of the family ‘name’ (kamei) rather than on continuity of the blood lineage. Good managers could be adopted by means of arranged marriages and given the family name.

After the end of the Second World War, the Americans interpreted faithfulness to the family name as a feudal relationship that formed the basis of Japanese militarism. They expropriated the zaibatsu families, nationalised their property and tried to sell all the stock to small shareholders. Concentrations of ownership and power were considered to be dangerous. Bank ownership and cross-shareholding were forbidden. By contrast, union activities were (initially) encouraged.

The war period and expropriation of the zaibatsu families reinforced insiders’ feelings of job security together with their specific investments in companies. The contradiction between these conditions and the system that the Americans intended to create became evident with the 1949 stock market crash. This made take-overs very easy: new owners, unconstrained by implicit contracts, could jeopardise workers’ job security and their firm-specific investments. As a result, cross-shareholding and bank ownership of stock were tacitly allowed, creating the conditions for delivery of the keiretsu system based on main banking, cross-shareholding and job security.

It is useful to compare the Japanese and Italian experiences in the years following the defeat.Footnote ¹² The comparison enables some sort of natural experiment to be conducted on the role of political subsidies in shaping new organisational structures.

The crisis of the 1930s was particularly severe in Italy. In order to save banks, the fascist regime had to buy out their industrial holdings and transfer them to a new agency created specifically for this purpose in 1933, known as the Istituto per la Ricostruzione Industriale (‘Institute for Industrial Reconstruction’) (Pagano and Trento, Reference Pagano, Trento, Di Matteo and Piacentini2003). Many large companies were owned by the state. Under the Banking Law of 1936, banks were prohibited from holding equity in industrial companies. No liquid stock exchange had been developed. Those firms that were not controlled by the state were family-owned. However, unlike the case of the Japanese zaibatsu, Italian family membership was a matter of blood, not the sharing of a kamei.

In Italy, it would have been much easier to introduce the American model based on separation of ownership and control. Many assets were already owned by the state. Moreover, banks and industry had been separated since 1936. But the political situation was different. Italy had not suddenly collapsed; rather, the liberation from German occupation had taken two years. In the meantime, the Italian counterpart of the zaibatsu families had all the time necessary to make deals with the Anglo-American forces. In the end, the Italian corporate governance system was not changed by the Anglo-American occupation.

The comparison with Italy confirms the role that political subsidies can perform in engendering irreversible institutional change. The main steps in the speciation of the post-war Japanese model can be summarised as follows:

1. (1) Before the war, under the zaibatsu system, a class of managers loyal to the firm already existed. The war enhanced the autonomy of management from the zaibatsu families.

2. (2) During the quasi-nationalisation of the zaibatsu companies, the rights of junior managers and workers were enhanced and had time to favour the associated technology. A return to the zaibatsu was not politically feasible: in other words, the new rights enjoyed a political subsidy from the domain of political selection. The inferior hybrids between the new rights and the old technology could not be wiped out. By contrast, they had an opportunity to move towards the speciation of a new organisational equilibrium.

3. On the other hand, while the insiders were becoming high-agency-cost factors, the kind of privatisation promoted by the Americans was based on legislation whereby shareholders could exercise hiring and firing rights, and on the separation between commercial banking and industry. This created a situation of potential ‘organisational disequilibrium’ between formal rights and technology. The threat of take-overs, following the 1949 stock crash, made this disequilibrium evident and dramatic.

4. (3) The property rights system that emerged from the crisis was based on cross-shareholding and the main bank system. The combination of this property rights system with technology based on specific and difficult-to-monitor labour allowed the definitive speciation of the new organisational equilibrium. Observe that speciation was favoured not only by an initial protection of the new system of informal rights but also by the fact that all companies were involved in this change. Hence the network externalities among property rights standards and technology did not inhibit the change. Indeed, they even favoured it because all the firms were simultaneously involved in the process. The network externalities in property rights were particularly evident in the cases of cross-shareholding and the main bank system, whose emergence required that a plurality of companies should be involved in the institutional change.

As the comparison with the case of Italy shows, a high political subsidy was necessary for change in the Japanese system of corporate governance. In Italy, the situation was riper for change, but because no political input came from the occupying forces, no change occurred. By contrast, in Japan, the role of policies implemented by the occupying forces was analogous to that of sexual selection in the development of the human brain. It created the conditions for the emergence of a more powerful organism whose evolution was inhibited by numerous interlocking complementarities. The dramatic events in the political domain produced a decisive subsidy for the emergence of a new organisational species.