The firm: a hierarchical organization

The first conception, which is issued from Hobbesian thought – and recalled by Williamson in a 1990 article dedicated to the contribution of Barnard to organization theory – considers that authority comes from the top of society so that individuals abdicate their freedom in support of an external authority granting them the external security of the city of man. This is Hobbes's Leviathan. As regards organization theory, this conception is used by authors such as Taylor (Reference Taylor1911), Fayol (Reference Fayol1916), or Weber (Reference Weber, Gerth and Mills1946) to show that the top hierarchy holds a statutory and vertical authority resulting from its rank and position. More precisely, hierarchies have certain standard features (Ricketts, Reference Ricketts2002: 172–173). Firstly, the firm is made up of people assigned to the various layers in the hierarchy, and these layers are for Williamson defined in terms of authority: those assigned to a higher rank supervise those in the subordinate layer.Footnote ⁴ In the same view held by Coase (Reference Coase1937),Footnote ⁵ Williamson (Reference Williamson1975: XV) notes that ‘hierarchy usually implies a superior–subordinate relationship’. For him (1975: 54), this ‘supervisor's constitutional authority’ to perform auditing and experience-rating is central to understanding the employment relationship and differences between market and firm. Secondly, positions higher up a hierarchy are predominantly filled by promoting people from lower positions in the hierarchy. The internal promotion system is strictly economic and has two advantages. On the one hand, this system develops incentive features because the employee knows that his ‘talent and degrees of cooperativeness’ (Williamson, Reference Williamson1975: 78) will be rewarded in the long term by promotions. On the other hand, thanks to evaluation and control procedures, the employer is able to adjust the employee's productivity and wages on a long-term basis. Additionally, this adjustment does not operate in the ports of entry because wages are linked to job status. Joining an organization entails that ‘the worker now is subject to much more detailed supervision’ (Williamson, Reference Williamson1985: 219). Finally, hierarchy is characterized by the fact that its associated rules are similarly applied to all people located at a given hierarchical level. For example, departing from the work of Doeringer and Piore (Reference Doeringer and Piore1971), Williamson (Reference Williamson1975: 74) notes that ‘the internal labor market achieves a fundamental transformation by shifting to a system where wage rates are attached mainly to jobs rather than to workers’. This argument is close to the Weberian model of bureaucracy, which considers that individuals are similarly and impersonally treated when they are assigned to a hierarchical position (e.g., Weber, Reference Weber, Gerth and Mills1946).

This vertical conception of hierarchical authority is strengthened in his 1985 book, in which Williamson defines what he calls the ‘unified governance structure’, ‘where the transaction is removed from the market and organized within the firm subject to an authority relation’ (1985: 75–76). And, in the 1991 article, the importance of authority in the distinction between the firm and the market is again confirmed: ‘bilateral adaptation effected through fiat is a distinguishing feature of internal organization’ (1991: 274).

The consensual nature of authority

The second conception, which is issued from Locke's philosophy, considers that society is based on the consent of citizens’. It is this consensus that constitutes the common good. Barnard is particularly representative of this stream of thought in organization theory. He accepts a consensual conception of authority and, more generally, of contracts, by arguing that individuals adhere to organizational objectives under the condition that they get a net income. According to Barnard, ‘authority is the character of a communication (order) in a formal organization by virtue of which it is accepted by a contributor to or member of the organization as governing the action he contributes; that is, as governing or determining what he does or is not to do so far as the organization is concerned’ (1938: 163). This consensual appreciation of authority can be understood with regard to the fact that, in an employment relationship, individuals accept authority in a ‘zone of indifference’ (1938: 167), the extension of this zone varying with the membership of individuals in an internal organization. Thus, the focus is here on the horizontal dimension of the authority-based coordination. Williamson (Reference Williamson and Williamson1990) also takes up this conception of authority and employment relationship proposed by Barnard in his 1938 book. For example, he notes in his 1975 book that ‘the employment relation is commonly associated with voluntary subordination’ (1975: XV).

Concerning the question of the acceptance of authority by employees, the argument of Williamson is relatively trivial insofar as he does not develop and discuss the reasons that drive the individual to accept the authority relationship. However, Williamson is also interested by the comparison between what he calls the ‘peer group’ – akin to a self-managed firm – and the ‘simple hierarchy’, notably in his 1975 book and in a 1980 article. From this comparison, Williamson finds three main complementary hypotheses.

Firstly, individuals would prefer working in a capitalist firm than in a self-managed firm because being a member of this latter type of firms leads to a salary sacrifice (Williamson, Reference Williamson1980). This choice of economic organization is less efficient than the choice of the capitalist firm. This argument seems to be paradoxical because, for Williamson, work satisfaction would be stronger in self-managed firms, notably because the supervisory control is weaker (Williamson, Reference Williamson1975). Williamson supports his theory by evoking the concept of revealed preferences. Therefore, individuals make a trade-off between income and satisfaction and would often prefer the first branch of the alternative. Secondly, as wages are the main motivation of employees, the opportunities of advancement offered by the internal labour market are capable of encouraging the acceptance of authority. Thirdly, Williamson quotes Barnard in arguing that group discipline strengthens this acceptance because an employee who would refuse to obey would penalize other organization members who get a net income from their participation in organizational objectives.

As regards the last paragraph, it seems that Williamson clearly adheres to the theory of the organizational equilibrium developed by Barnard in his 1938 book. For Barnard, individuals accept authority because there is a balance between workers’ contribution and workers’ compensation. According to March and Simon (Reference March and Simon1958), the theory of the organizational equilibrium is the proof that the organization succeeds in granting its members sufficient remuneration to motivate them and promote their cooperation through time. It seems evident here that the behavioural theory of the firm has had certain influence on the use of organizational literature by Williamson.Footnote ⁶ Thus Williamson cites their 1958 book that argues that ‘within some limits (defined both explicitly and implicitly by the terms of the employment contract), the employee will accept as premises of his behaviour orders and instructions supplied to him by the organization’ (March and Simon, Reference March and Simon1958: 90). However, in this particular case, March and Simon go much further than Williamson in the definition of the organizational equilibrium. They are interested in the ‘participation decision’ and question the necessary conditions for an individual to be inclined to accept authority and go on to participate in organizational objectives. The contribution– compensation equilibrium has for them two main components: (i) the wish to leave the organization linked both to work satisfaction and the perception of opportunities in the organization, and (ii) the perceived capability to leave the organization linked to the perceived outside options. However, Williamson does not really address this question, merely noting that ‘it is possible for workers to choose between general purpose and firm-specific skills’ (1985: 259).

The institutional environment: the role of legal rules

Even though Williamson adheres in part to the consensual conception of authority proposed by Barnard, he nevertheless emphasizes that this conception cannot really be stretched because of State authority (Williamson, Reference Williamson and Williamson1990). Transaction cost economics is a part of the New Institutional Economics that differs from orthodox economics because it is informed by both organization and law (Williamson, Reference Williamson1996a). Interestingly, it is important to show in what ways law affects organization in Williamson's view. The legal rules are issued from the institutional environment, but mainly constitute resources complementary to the governance structure. For Williamson, these legal rules form the ‘rules of the game’ (Williamson, Reference Williamson1993a: 113). In fact, these legal rules are a parameter that contractors take into account when they have to choose an appropriate governance structure, as these rules have an influence on transaction costs. For instance, in his 1991 article, Williamson shows that a change in the ‘forbearance doctrine’ – which is an implicit contractual duty peculiar to the firm – leads to a modification of the governance costs of a hierarchy. Indeed, the interference of courts to sort out intra-firm technical disputes would increase hierarchy costs, which discriminates against this governance structure compared to the market or the hybrid form. Finally, contracts and property rights enable the delimiting of the positions of involved parties in the presence of conflicts and persistent disputes – with recourse to vertical authority.

In this view, the contribution of Masten (Reference Masten1988) gives crucial insights on this question and suitably clarifies the Williamsonian conception of authority. In this article, Masten seeks to establish whether any rights or authority of the employer can be replicated in commercial contracts. His main conclusion is that ‘the investigation reveals that the law does in fact recognize substantial differences in the obligations, sanctions and procedures governing the two types of exchange’ (1988: 196). Concerning authority, he argues that, inside the firm, individuals engaged in an employment relationship accept ‘an implied duty to yield obedience to all reasonable rules, orders and instructions of the employer’ (ibid.). For this, Masten adds that ‘the traditional emphasis in economics on the authority of management to direct the efforts of employees is at least nominally supported by the law governing employment transactions’ (1988: 186).

Williamson's analysis is, however, more complex, insofar as the legal rules, embedded in an institutional environment, shape only a legal framework, and contractors have the latitude to arrange this framework in accordance with their preferences and goals in order to implement the best governance structure. Hence, Williamson refers to the works of Llewellyn (Reference Llewellyn1931) and MacNeil (Reference MacNeil1978) to affirm that the contract constitutes only a simple legal framework, which is incomplete, owing to the bounded rationality of individuals. Indeed, transaction cost economics analyses organizations from the ‘contractual man’ view based on the assumptions of opportunism and bounded rationality. In other words, the governance structure of a ‘firm’ is a private arrangement embedded in an institutional environment (Williamson, Reference Williamson2002).Footnote ⁷ Henceforth, for Williamson, authority and fiat do not exclusively result from the legal system of employment contracts (as we have shown above) but also result from the hierarchical structure and the acceptance of orders by the employees.

In this view, the TCE (transaction cost economics) differs sensibly from neoclassical theories of the firm that consider that the firm is simply a continuation of the market (e.g., Alchian and Demsetz, Reference Alchian and Demsetz1972; Jensen and Meckling, Reference Jensen and Meckling1976). Such advancement needs to give strong attention to the internal organization of the firm. This implies the need to identify the emergence conditions of cooperation and also to take into account intra-firm conflicts regardless of market considerations. To do so, Williamson heavily uses the arguments of organization theory.

Conflicts, opportunism and organization theory

The behavioural hypothesis of opportunism accepted by Williamson has been the subject of a large literature. It seems clear that this choice is a key piece of the global reasoning he has developed. For him, individuals do not behave opportunistically all of the time but may behave opportunistically some of the time so that ‘plainly, were it not for opportunism, all behaviour could be rule-governed’ (Williamson, Reference Williamson1985: 48). However, this concept – and its variants – is in fact largely and previously used in organization theory. Interestingly, the strategic approach to the organization (Crozier, Reference Crozier1964; Crozier and Friedberg, Reference Crozier and Friedberg1977) is very close to Williamson's theory regarding its appreciation of individuals’ behaviour. Indeed, this organizational approach rests on the following three postulates, which are also mobilized by Williamson. Firstly, in his 1964 book, Crozier takes up the concept of bounded rationality as it is defined by March and Simon (Reference March and Simon1958) to analyse the behaviour of economic actors. Secondly, for Crozier, each actor has, in an organization, a margin of relative freedom. Arguably, it seems possible to parallel the strategic approach to the organization and Williamson's theory of the firm. It is notably the case when Williamson (Reference Williamson1975) affirms that each employee is able to choose between two types of cooperation: the perfunctory (minimal) cooperation and the consummate (maximal) cooperation.Footnote ⁹ Thirdly, even though Crozier does not use the term ‘opportunism’, actors are opportunistic in the strategic approach to the organization – similarly to Williamson.

Anyway, Williamson assesses in a 1990 article that ‘organization theorists were familiar with and had an extensive literature dealing with opportunism long before economists got around to it’ (1990: 290). For instance, the term ‘subgoal pursuit’ used by March and Simon (Reference March and Simon1958) is for Williamson intimately tied to opportunism: ‘members of an organizational unit tend to evaluate action only in terms of subgoals, even when these are in conflict with the goals of the larger organization’ (quoted by Williamson, Reference Williamson and Williamson1990: 181). However, Williamson refers also to Crozier (Reference Crozier1964) and indicates some important examples Crozier gives to appreciate opportunistic behaviours, particularly when he emphasizes ‘the active tendency of the human agent to take advantage, in any circumstances, of all available means to further his own privileges’ (quoted by Williamson, Reference Williamson1993a: 115). Finally, for Williamson as for Crozier, individuals can be opportunistic because they have a real margin of freedom in organizations.

Each actor then attempts to take advantage of his situation in order to satisfy his rational objectives, which do not necessarily coincide with the organization's objectives. According to Crozier, the organization consequently has to implement specific governance modes to contain conflicts and secure its stability and durability in spite of divergent actors’ goals. The parallel between Williamson and Crozier is here very striking. For Williamson, the employer has to ensure that employees do not act in their own interests – for example, by acquiring and exploiting monopolistic positions issued from their specific assets – but act for the firm's common goal. Hence, Williamson shows that the creation of an internal labour market allows the bypassing of transactional difficulties For example, by shifting to a system where wage rates are attached mainly to jobs rather than to workers, the possibility of individual bargaining is strongly diminished and the incentive to act opportunistically decreases during contract renegotiation (1975: chapter 4). Nevertheless, it is also indispensable to analyse the question of employer opportunism, i.e., to admit that ‘individuals in positions of authority may exploit authority relations opportunistically’ (Moschandreas, Reference Moschandreas1997: 39). Numerous analysts reproach Williamson for considering that only employees are potentially opportunistic, whereas the employer would never be. Indeed, the ‘universal’ character of opportunism seems clearly to be in contradiction with the fact that employers would not behave opportunistically. For example, Dow (Reference Dow1987: 20) notes that ‘transaction cost theorists tend to see authority primarily as a remedy for opportunism, rather than as a device which might be abused in an opportunistic fashion’. Dow adds that the authority relationship creates the conditions for the emergence of employer opportunism. Included in these conditions are asymmetrical information, small number case, and commanding decision making. In organization theory, the works of Crozier have shown that people at the top of the hierarchy can also act opportunistically. However, it is right that the internal labour market has the main mission to fight against employees’ opportunism.

This critique seems nevertheless to be severe because Williamson does not deny employer opportunism, even though he does not analyse this situation in detail. In his 1985 book, he explains that, in certain circumstances, the employer could exploit the specific investments of incumbent employees. Put differently, Williamson has introduced – contrary to what Dow argues – what economists call the ‘double moral hazard’. Williamson agrees with Crozier on this point. Indeed, for the strategic approach to the organization, top managers have a large margin of freedom due to their privileged position, which allows them to favour their own interests by manipulating information they have (Crozier, Reference Crozier1964).

What relation between opportunism and power?

For the theorists of the strategic approach to the organization, power is a crucial phenomenon in organizations. Crozier (Reference Crozier1973) defines power as a non-transitive reciprocal relation where one of the actors can win more than others and where nobody is powerless. Hence, power is defined as the ability of an entity A to get a favourable situation to that of an entity B during the negotiation process. Thus, power results from the discretionary scope economic actors have on each other. Each actor has a power, and the degree to which that actor controls the sources of uncertainty parallels his level of power (Crozier and Friedberg, Reference Crozier and Friedberg1977). However, Williamson is very critical towards the concept of power and considers that ‘the main problem with power is that the concept is so poorly defined that power can be and is invoked to explain virtually anything’ (1985: 237–238). To reinforce his position, Williamson does not hesitate to quote March (Reference March and Easton1966), who thinks that ‘power is a disappointing concept’ (1966: 70).

The position of Williamson concerning power is astonishing regarding the behavioural assumptions he uses, namely the bounded rationality and the opportunism of economic agents. Indeed, we have shown that these hypotheses theoretically lead to potential conflicts and thus to power relationships. Thus, power seems not to be far from Williamson's analysis. In fact, for Williamson, power rests on ex post situations. On the contrary, he argues that, on the one hand, transaction should be studied in its entirety – i.e., both ex ante and ex post – and, on the other hand, that economic agents are not myopicFootnote ¹⁰ but anticipate possible power positions. In other words, economic agents could protect themselves against power situations, as the following arguments demonstrate.

As it is explained above, an employer is able to act opportunistically and to ‘exploit’ the incumbent employees who own specific assets by using his power. However, for Williamson, employer opportunism can be thwarted by employees’ actions.Footnote ¹¹ In fact, employees are not really without recourse and can act in a way that penalizes their employer by choosing perfunctory cooperation. Williamson argues that ‘incumbent employees who are forced to accept inferior terms can adjust quality to the disadvantage of a predatory employer’ (1985: 262). This last argument Williamson uses to show that employer opportunism can be restrained is very interesting and relevant. As regards this question, Williamson is very close to organization theory. Indeed, Williamson's argument seems to indicate that the protective governance structure does not totally control the worker in the execution of his work, an argument previously developed by Crozier (Reference Crozier1964). Thanks to their margin of freedom, employees have a certain ‘retaliation power’ on employers. The final outcome of this confrontation depends on the types of resources involved actors have. Here, we are close to a resource dependence analysis.Footnote ¹²

Additionally, Williamson quite rightly observes a positive relation between the degree of specificity of assets and the opportunism of individuals. For example, the specificity of human assets can be used by employees in the negotiation process. As regards the definition of power made by Crozier, it appears that Williamson is again very close to this approach when engaging the question of the distribution of the surplus of the employment relationship. The two actors have critical resources: the employer can fire the opportunistic employees, whereas employees can retaliate against their employer by selecting perfunctory cooperation. Put differently, there is potentially a place for power in Williamson's theoretical framework. Interestingly, Williamsons departs from the fact that the employer can fire employees, whereas the reverse is not true, and claims there to be ‘a strategic assessment of the employment relation in which asymmetries are recognized and explicitly taken into account’ (1985: 260). However, again, Williamson seems to make ‘shallow’ propositions without theoretical deepening, which distinguishes transaction cost economics from the strategic approach to the organization. Thus, for Williamson, it is always possible to circumscribe conflicts insofar as the appropriate economic institutions are implemented, such as arm's length contracts, the internal labour market, or reputation.

Finally, even though Williamson (Reference Williamson1995) argues that power could be relevant for transaction costs economics, he also considers it as a shifty concept often used to rationalize ex post actors’ behaviours. But, is power really a useless concept for transaction cost economics?

A critical comment on Williamson's analysis of efficiency and power

In opposition to the strategic approach to the organization, Williamson ignores the strong interest of power for economics. However, it seems that power is a strong vehicle of cohesion and thus a necessary condition for firms to function (Chassagnon, Reference Chassagnon2009). Besides, as Williamson recognizes himself, a large organizational literature has been interested in the concept of power in organizations. But Williamson has never seemed to be convinced by organization theorists on the issue of power, even though some economists have been critical of the neglect of power considerations in transaction cost economics (see, e.g., Hodgson, Reference Hodgson and Pitelis1993; Palermo, Reference Palermo2000).

The Williamson's methodological framework can be invoked to justify his controversial position. For him (2009), the theoretical success of transaction cost economics is based on the operationalization of both Coase (Reference Coase1937) and Commons (Reference Commons1931). He adds that his theory of the firm yields numerous refutable implications so that he has ‘no hesitation in declaring that transaction cost economics is an empirical success story’ (2000: 607). Hence, a methodological concern of Williamson is surely about the operationalization of the insights and concepts he brings from other perspectives and disciplines. Needless to add that the concept of power such as it is used in organizational theory cannot be easily adapted with the Williamson's methodological commitments.Footnote ¹³ It would be unsuitable to introduce power in the economics of the firm because it would be impossible to operationalize this vague concept. This can explain the fact that Williamson does not benefit from the important organizational contributions to power analysis.

Notably, a curious point in Williamson's analysis of power rests on the poor consideration for resource dependency theory. Williamson contents himself with quoting some important authors in organization theory such as Selznick (Reference Selznick1957), Emerson (Reference Emerson1962), Pfeffer (Reference Pfeffer1981), Perrow (Reference Perrow1986), or Scott (Reference Scott1987) and conceding that it suffices for his purpose ‘to remark that power, for a long time, has been the congenial organization theory perspective’ (Williamson, Reference Williamson and Williamson1990: 182). Arguably, it is interesting to understand more precisely why Williamson does not give great attention to the concept of power as it is used by resource dependency theory. To do this, it is necessary to recall the contribution of organization theory to the concept of power.

First of all, we have to keep in mind that because no economic organization is strictly formal – i.e., regulated by instituted rules of law – power is present in all economic organizations. It is impossible to study organizational behaviours without investigating the relationships of power that emerge in organizations (Clegg et al., Reference Clegg, Courpasson and Phillips2006). Organization theory clarifies the concept of power and proposes to use it with the understanding of organizational behaviours. The starting point of this argument is social exchange, from which we can observe the intrinsic origin of power: resource dependence (Emerson, Reference Emerson1962). Thus, the scarcity of resources and the position in an exchange network determine the relative power of an individual (Pfeffer, Reference Pfeffer1981). Organization theory analyses the relationships of power from a relational perspective, looking at the dependence of an actor vis-à-vis another and shows, both theoretically and empirically, that the more the alternatives of an actor increase, the greater the available resources are and the weaker the dependence of actors on another's resources – and thus power – is (Cook and Emerson, Reference Cook and Emerson1978). Similarly, Jacobs (Reference Jacobs1974) argues that the essentiality of resources is the main determinant of power-dependence relations between organizations. In general terms, the knowledge of a social system is, in its own right, an internal base of power (Krackhardt Reference Krackhardt1990).Footnote ¹⁴

This theory of power based on resource dependence seems at a first sight to echo back to the hold-up theory of Williamson,Footnote ¹⁵ according to which the economic agents who invest in specific assets have potential power. However, this conclusion is hasty, as in organization theory relational cohesionFootnote ¹⁶ and informal commitment take the place of opportunism and contractual arrangement recommended by transaction cost economics. Therefore, a strong difference between organization theory and transaction cost economics lies in the theoretical scope they give to power. For Williamson, hierarchy is a source of efficiency for all participants, owing to its coordinating role of economic activities. On the contrary, Williamson and Ouchi (Reference Williamson, Ouchi, Francis, Turk and Williams1983) assign power a role just as soon as efficiency is sacrificed in support of personal interests – that is, as soon as the market fails.

While authority serves efficiency, power is, according to Williamson, the manifestation of inefficiency. Here, authority should be clearly distinguished from power. They are two distinct concepts. Authority is a distinct moment of power. More precisely, authority is the institutionalization of power, especially through the law that recognizes hierarchical subordination. Hence, authority refers to a mutual arrangement that stops when a contract expires (Palermo, Reference Palermo2000). On the contrary, power comes from insubordination, i.e., outside this instituted area of obedience (Chassagnon, Reference Chassagnon2009). In this area of insubordination, resources dependence is a determinant in power distribution. However, efficiency considerations seem to explain the poor attention Williamson gives to resource dependency theory. It appears evident in his 1990 article, when he affirms that ‘many dependency issues can be addressed in efficiency terms, whereupon power considerations largely vanish’ (1990: 182). Although Williamson (Reference Williamson1985) takes into account the informal and relational elementsFootnote ¹⁷ of the firm, power ultimately has no place in his analysis. Indeed, Williamson (Reference Williamson1985: 247) quotes the work of Ouchi (Reference Ouchi1980) on ‘clans’ and introduces what he calls the ‘relational team’ (see also Macneil, 1978) to designate the firm engaged ‘in considerable social conditioning to help assure that employees understand and are dedicated to the purposes of the firm’. However, even from this perspective, he makes no reference to the distribution of power between the members of the relational team. In sum, ‘the concept of fiat, used by Williamson to criticize the nexus of contracts stream and reinforce his claim on the relevance of authority, however, seems yet to leave him within the neoclassical sphere of free individual choices in any situation as long as he ignores power’ (Pessali and Fernandez, Reference Pessali and Fernandez1999: 267).

If Williamson's theory of the firm casts doubt on the theoretical scope of power, organization theory has otherwise shown the utility of power concept. These two theories are indeed very complementary, and we believe that economists have to take into account the analysis of power based on resource dependence proposed by organization theory.Footnote ¹⁸

To conclude on this section, even though there are some important divergences, the work of organization theorists – notably the strategic approach to organizations – are complementary to Williamson's work on the internal organization of the firm. This allows the clarification of some crucial questions on the functioning rules of the firm, notably from the key issue of intra-firm conflict. However, the neglect of power issues and the focus on opportunismFootnote ¹⁹ seem to show that Williamson's theory ‘is vulnerable to the reproach that production is neglected in favour of allocation’ (Hodgson, Reference Hodgson and Medema1998a: 31). Otherwise, it may be easier both to understand the positive role of power in the emergence of intra-firm cooperation and to relax the hypothesis of omnipresent opportunism in favour of a strengthening analysis of organizational atmosphere.

Atmosphere: satisfying exchange relation and calculativeness

According to Williamson, it is crucial to find some organizational regularities to apprehend the internal organization of the firm. The neglect of these regularities leads organizations to have dysfunctions. In this view, Williamson makes a critique of neoclassical economists who compare the firm to a productive function, as such a ‘machine model of organization’ (1993a: 116) focuses on the intended consequences to the detriment of unintended consequences. Atmosphere and bureaucratic dysfunctions are two important applications of Williamson's theory.Footnote ²⁰

Atmosphere is the first organizational regularity but is rarely analysed in the literature devoted to transaction cost economics. Indeed, Williamson casts doubt on the standard economic model by emphasizing that this theory assumes that ‘individuals regard transactions in a strictly neutral, instrumental manner’, whereas, for Williamson, it is essential ‘to regard the exchange process itself as an object of value’ (1975: 38). Thus, internal organization has to supply a ‘satisfying exchange relation’ to individuals thanks to atmosphere (ibid.). In a subsequent paper, Williamson adds that ‘atmosphere refers to interactions between transactions that are technologically separable but are joined attitudinally and have systems consequences’ (1993a: 135). Williamson clearly aims to go beyond the calculative orientation of the standard economic model and, for him, three elements are susceptible to have an influence on the atmosphere of internal organization: the control of work, the incentive system, and the settlement of disputes.

Firstly, for Williamson and contrary to the analysis of Alchian and Demsetz in their 1972 article, the control of work by the measurement of individual productivity leads to a calculative atmosphere which generates negative effects – externalities – on the employment relationship. On the one hand, this control produces direct costs. On the other hand, it negatively affects the work involvement of employees; employees opting for perfunctory cooperation.

However, the incentive system of the organization also exerts an effect on atmosphere. Contrary to individualistic contracting modes, where rewards usually depend on performance in a transaction – as on a piece-rate basis – ‘workers in internal labor markets are metered less intensively’ (Williamson, Reference Williamson1975: 79), on a long-term basis, which increases their involvement. Similarly, the advancement system based on seniority creates a satisfying atmosphere, and, finally, the negotiation procedures of the internal labour market contribute to create a ‘sense of justice’ (ibid.).

The third and last device that has an influence on the organizational atmosphere is the settlement of disputes. When disputes are settled within the organization from the acceptance logic – i.e., in private manner regardless of legal structures – these ones are dynamically reduced. Williamson (Reference Williamson1975: 81) explains that ‘a grievance procedure, with impartial arbitration as the usual final step allow the firm and the workers to deal with continually changing conditions in a relatively non-litigious manner’ and concludes that ‘contract revision and renewal take place in an atmosphere of mutual restraint in which the parties are committed to continuing accommodation’.

Atmosphere and organization theory: the contribution of the human relations school

The contribution of organization theory to Williamson's work on the organizational atmosphere is very clear and suitable. Emphasizing the importance of human relations in the organization and the perverse effects of control, Williamson matches against the mechanical conception of the individual and so criticizes the classical rationalism of Taylor's one best way. Putterman (Reference Putterman and Putterman1986) considers that thanks to the concept of atmosphere we understand better the important gap between Williamson's conception of the firm and other theories of the firm. This is particularly manifest concerning agency theory, which is based on a mechanical and reductionist model of the individual insofar as salary is the only variable able to affect employees’ productivity (see Jensen and Meckling, Reference Jensen and Meckling1994). Finally, the study of organizational atmosphere allows Williamson to cast doubt on the mechanical approaches to the economic organization. For Williamson (Reference Williamson1993a), these calculative approaches to organizational phenomena can lead – as an excessive control shows – to some important internal dysfunctions. As a result, Williamson's analysis of atmosphere sheds light on the weight of non-monetary satisfactionsFootnote ²¹ as the human relations school does.

The analysis of the organizational atmosphere proposed by Williamson rests largely on the human relations school, notably on the works of Barnard (Reference Barnard1938) and Mayo (Reference Mayo1945). The human relations school has emphasized the role of human capital in organizations, which has involved a profound reconsideration of the rationalist classical thought of Taylor (Reference Taylor1911) and Fayol (Reference Fayol1916). More precisely, for this organizational approach, the firm has two main functions: (i) an economic function linked to cost, benefit, and technical efficiency considerations and (ii) a collective satisfaction function. The firm has to reconcile these two purposes.

The human relations school – and notably Barnard – distinguishes the formal organization from the informal organization. The formal organization is defined by Barnard (Reference Barnard1938: 81) as ‘a system of consciously coordinated activities or forces of two or more persons’. This definition is based on the explicit functioning rules of intra-firm cooperation. Therefore, the formal organization differs from the informal organization, which is ‘the aggregate of the personal contacts and interactions and the associated groupings of people’ (ibid.: 115). The industrial firm is considered as a real social system and must be treated as a whole – as an entity – because each composing part is interdependent on the others. According to the human relations school, these two distinct forms of organization correspond closely to two different views on an organization's members, namely a logic of productivity and a logic of feelings. Ultimately, for this school of thought, the employees of a firm are motivated both by economic interests and by feelings – non-economic motivations. In this sense, Williamson's theory is close to the human relations school.Footnote ²²

In his 1990 article devoted to Barnard, Williamson confirms the dichotomy between the formal organization and the informal organization and argues that this latter ‘may be a manifestation of a more general condition of atmosphere’ (1990: 184). He adds that ‘informal organization arguably helps to safeguard the security and integrity needs of employees’ (ibid.). Additionally, the informal organization appears in his comparative analysis of governance structures as being better than the hybrid form, notably regarding the settlement of contract disputes (Williamson, Reference Williamson1991). The organizational atmosphere is a part of internal organization, but is distinct from authority and clearly rests on a logic of feelings. Thanks to the human relations school, the crucial role Williamson devotes to the organizational atmosphere is better understood. As we have mentioned above, one of the more important functions of internal organization is to give its members a satisfying exchange relation by considering their feelings, which are intrinsic to individuals and are thus variable. For example, in his study of peer groups and simple hierarchy, Williamson underlines that these forms of internal organization ‘will presumably be organized by those individuals who find market transactions less satisfying than a nonmarket relationship’ (1975: 45).

From this conceptual framework, the theorists of the human relations school have been interested by the role of management. Management aims to preserve the equilibrium of the social system ‘firm’ and to locate and resolve potential dysfunctions. To do this, it is necessary to keep in mind the two logics of individuals respectively based on productivity and feelings so that individuals are satisfied and willing to cooperate. Hence, one of the more important roles of managers is to sustain and maintain cohesion and satisfying conditions and another is to avoid internal dysfunctions.

Some extended remarks on atmosphere

What should we keep from the concept of atmosphere developed by Williamson? We think that atmosphere is one of the main original insights of the work of Williamson, notably because it allows one to bypass the neoclassical theoretical framework and to propose a new vision of the firm. Thus, Pessali considers that the concept of atmosphere ‘gives room for factors to which other social scientists can more easily relate’ (2006: 57). The concept of atmosphere is important because it permits Williamson to include some non-calculative considerations in the traditional economic reasoning. In addition, Williamson gives then the possibility to strongly distinguish – from a discrete alternative perspective – the firm from the market. For Williamson, the market exacerbates calculative attitudes and thus the risks of opportunism. On the contrary, the internal organization of the firm encourages ‘quasi-moral’ commitments (Williamson, Reference Williamson1975: 38), which gives Williamson the opportunity to cite the work of Gouldner (Reference Gouldner1960) on reciprocity.Footnote ²³ It is thus surprising that Williamson has not continued using the concept of atmosphere after his 1975 book. How can one explain this? Three reasons can be advanced.

The first, given by Williamson himself, is that he has not made ‘any headway with it’ after 1975 (See Swedberg, Reference Swedberg1990). The second, notably developed by Pessali (Reference Pessali2006), is linked to the difficulty of reconciling the concept of atmosphere and the concept of opportunism. Pessali notes (2006: 57) that ‘atmosphere entails human features – reciprocity, dignity, trust, and loyalty – that arguably compensate for opportunism’ so that ‘atmosphere has become more of a burden than a help’. In fact, the assumption of universal opportunism has led to ‘the adoption of administrative lines of authority and subordination that may have serious effects on the work atmosphere’ (Moschandreas, Reference Moschandreas1997: 42). On the contrary, the concept of organizational atmosphere is more in line with the idea that trust can be real ‘in the sense that it goes beyond calculative self-interest’ (Klos and Nooteboom, Reference Klos and Nooteboom2001: 507). In other words, it would be the difficulty of reconciling calculative elements and non-calculative considerations that would have led Williamson to remove the concept of atmosphere. However, this argument can be counterbalanced by considering that the ‘degree’ of opportunism varies as a function of the transaction context. In short, the market creates competitive pressures that would result in calculative attitudes, whereas, inside the firm, the internal labour market – and the fact that all of the members of a given hierarchical level are similarly treated – is free from competition and thus is able to reduce the propensity of employees to act opportunistically and to be calculative. The third reason rests on the fact that the concept of atmosphere has progressively been overtaken by the idea of ‘relational contracting’.Footnote ²⁴ Such a choice puts Williamson in the position to base his analysis on a more contractual approach than in his 1975 book.Footnote ²⁵ Similarly, the focus on ‘private ordering’ is strengthened (Williamson, Reference Williamson2002). As a result, it has become difficult to include atmosphere in a more contractual theoretical framework.Footnote ²⁶

In any case, an abundant literature has underlined that the concept of opportunism used by Williamson needs more attention to both the details of the process and its practical implications (see, e.g., Hodgson, Reference Hodgson1988; Noorderhaven, Reference Noorderhaven1995; Ghoshal and Moran, Reference Ghoshal and Moran1996; Nooteboom, Reference Nooteboom1993, Reference Nooteboom2002). Arguably, particularly in his 1975 book, the main objective of Williamson seems mainly to introduce the concept of atmosphere in response to orthodox models of the firm that reduce it to an impersonal workplace (function of production).

Bureaucratic dysfunctions and farsighted contracting

Merton and Gouldner have been interested by the unexpected reactions of an organization's members to the supervision of tasks. The model of Taylor (Reference Taylor1911) explains that controlling mechanisms must be used for securing the behavioural compliance of individuals to rigid organizational patterns. In response to this opinion, Merton and Gouldner have shown that there are some important dysfunctional consequences in bureaucratic organizations. These dysfunctions are generated by organizational procedures based on a mechanical – Taylorian – representation of human behaviour. If these procedures have some consequences intended by the management, there are also some unintended consequences, such as the growth of interpersonal tensions and the weakness of motivation in the organization. Then, the unintended consequences strengthen the need for control and supervision and so on, which leads the organization in a vicious circle. The argument of Williamson thus meets the argument of Crozier (Reference Crozier1964), who concludes that dysfunctions are the result of the employees’ opposition to management orders, which are mechanically given. The analysis of organizational regularities thus allows Williamson to again criticize the calculative approach to economic organization that ‘can and sometimes does result in obsessive demands for control’ (Williamson, Reference Williamson1993c: 459).

Concerning the question of control – more precisely, the unintended consequences of control – Williamson introduces a managerial and sociological approach to the firm, even in terms of efficiency. For him, thanks to their ‘farsighted propensity’ and ‘rational spirit’ (1993a: 116), economic actors are able to remedy dysfunctions. Such a conclusion seems to be in contradiction with the concept of bounded rationality (Pessali, Reference Pessali2006). This is why Williamson has in recent times tried to clarify the place he gives to bounded rationality in his theory of the firm. He writes (2009: 150): ‘transaction cost economics describes both cognition and self-interest in a two-part way. Specifically, cognition combines bounded rationality with feasible foresight while self-interest joins benign behavior with opportunism. Thus all complex contracts are unavoidably incomplete (by reason of bounded rationality) yet human actors are assumed to have the capacity to look ahead, recognize hazards, work out the mechanisms, and, albeit imperfectly, factor the ramifications back into the ex ante contractual design (by reason of feasible foresight).’ When dysfunctions are observed and understood, the unintended consequences are anticipated, and the organizational design is consequently revised. This results in an improvement of economic performance.

In this sense, the analysis of atmosphere and bureaucratic dysfunctions gives Williamson the ability to develop an approach that appears to be very appropriate for assessing transaction cost economics. For Williamson, contracts are generally incomplete but transactions are not always linked to the hazards, because economic agents use farsighted contracting. It is for this reason that transaction cost economics is not static but developed from an inter-temporal perspective, in which organizational learning is essential. Economic agents are not omniscient, their rationality is bounded and they are capable of being mistaken, but the experience of past transactions leads them, from an efficiency perspective, to regularly improve the contractual framework. As a proof of this argument, three applications used in Williamson's work can be employed.

Three applications of farsighted contracting

First of all, as we have mentioned above, Williamson considers that the control and metering of inputs encourage a calculative atmosphere, which affects the involvement of workers who select perfunctory cooperation. However, his analysis is more subtle as he clarifies that ‘the more general questions to be addressed, however, are not non-metering or full-metering’ because, for him, ‘individuals differ with respect to their preferences for metering’ (Williamson, Reference Williamson1975: 55). He thus concludes that the organization has to supply ‘the requisite metering mixture of structures’ (ibid.). As individual feelings rest on innate dispositions, it is possible to introduce this consideration in organizational objectives because there is a positive relation between satisfying exchange relation and efficiency. This argument claims the existence of farsighted contracting and reinforces the conclusions of the human relations school.

A second example can be found in the 1972 article of Alchian and Demsetz, who discuss the metering problem through the example of a university where ‘the faculty use office telephones, paper and mail for personal uses beyond strict university productivity’ (ibid.: 780). For Alchian and Demsetz, organizational control over work can eliminate such practices, thus enabling the increase of pecuniary rewards. However, Williamson criticizes this conclusion and underlines that there are interactions between the attitudes of workers with regard to transactions. As a result, Williamson asserts that ‘it seems at least plausible that extending metering with respect to such peccadilloes as appear to be of concern to Alchian and Demsetz until the costs and benefits are equalized at the margin will be regarded as picayune and will elicit resentment’ (Williamson, Reference Williamson1975: 56). He adds that ‘cooperative attitudes will be impaired with the result that tasks such as teaching effectiveness – which can be metered only with difficulty, because information is deeply impacted but for which consummate cooperation is important – will be discharged in a more perfunctory way’ (ibid.). Again, the role of the organizational atmosphere and the logic of feelings are crucial, which supports Williamson's argument and justifies the implementation of a farsighted governance structure.

Finally, let us recall the example of the bureaucratic vicious circle illuminated by Merton (Reference Merton1957) and Gouldner (Reference Gouldner1960). According to Williamson, once the unintended consequences are known, it is possible to break this vicious circle ‘by factoring these into the original organizational design’ (1993a: 117), which improves efficiency. This example is clearly a relevant application in the spirit of farsighted contracting.

For all this, it is important to emphasize that the implementation of farsighted contracting prohibits Pareto efficiency. Thus, Williamson (Reference Williamson2005b) argues that it is not accurate reasoning in terms of hypothetical ideals. On the contrary, it is needed to focus on comparing feasible alternatives. Indeed, for him, ‘public policy analysis becomes more complicated when analysts can no longer condemn extant modes because they deviate from a hypothetical ideal, full stop’ (Williamson, Reference Williamson2000: 601). Hence Williamson uses the concept of ‘remediableness’, which means that economic agents, who are rationally bounded and have asymmetrical information, do not succeed in achieving Pareto efficiency but do succeed in selecting the best organizational design for a given period. More precisely, remediableness is a criterion that ‘holds that an extant mode of organization for which no superior feasible alternative can be described and implemented with expected net gains is presumed to be efficient’ (ibid.). In conclusion, ‘absent rebuttal, the remediableness criterion stands as a reminder of the obvious: it is impossible to do better than one's best’ (ibid.). The concept of remediableness shows clearly that the economic theory of Williamson differs substantially from the standard economic model.