Unions are complex organizations, have many facets, and serve many purposes. Efficiency aspects are mainly emphasized here. They are examined from the standpoint of private ordering. Other aspects, however, are also considered, including a discussion of power.
1. Private Ordering
Katherine Stone has recently examined the development of American labor law in the period following 1945, with special attention to the “industrial pluralism” model, which holds that “collective bargaining is self-government by management and labor” (1981, p. 1511). She contends that this model distorts rather than clarifies issues because it is “based upon a false assumption: the assumption that management and labor have equal power in the workplace” (1981, p. 1511).
I submit that examining the development of labor law as a private ordering exercise is instructive. Although the industrial pluralists with whom Stone takes exception—especially professors Schulman and Cox among labor law scholars, Justice Douglas on the Supreme Court, and Arthur Goldberg as a labor lawyer— treat labor in a more aggregative way than I think adviseable, all appear to recognize the benefits of a private ordering orientation. Inasmuch, moreover, as an “equal power in the workplace” criterion is nowhere defined, I am unable to ascertain how labor markets would have been organized if the leading figures had adopted her preferred standard.
Schulman interprets the Wagner Act as a “bare legal framework” within which private ordering by labor and management can proceed (1955, p. 1000). To be sure, the Act made express reference to the desirability of creating “equality of bargaining power between employers and employees.”11 Stone views this as a call to political reform, whereby the law intervenes actively “to alter the definition of property rights in order to create true equality” (1981, p. 1580). Power equalization being an ambiguous criterion, the industrial pluralists concentrated instead on attainable goals: fashioning efficient and feasible reforms.
The institution of arbitration lies at the core of industrial pluralism. As Stone puts it:
In the industrial pluralist model, disputes over breaches of collective agreements are not submitted to an administrative or judicial tribunal. Rather, they are submitted to the dispute resoitition mechanism that the parties in this minidemocracy have established for themselves—private arbitration. . . .
A corollary of this description of the industrial world is the prescription that the processes of the state—the courts and administrative tribunals—should keep out. The workplace . . . becomes … an island of self-rule whose self-regulating mechanisms must not be disrupted by judicial intervention or other scrutiny by outsiders. [Stone, 1981, p. 1515]
Stone goes on to observe, “Judicial resolution of labor disputes… was unacceptable because it imposed a noncontractual solution upon the parties” (1981, p. 1524). She contends, moreover, that the purported advantages of voluntary arbitration—special expertise of arbitrators, the informality of the arbitration procedures, the arbitrators’ flexibility of remedy—could all have been realized by assigning this responsibility not to an arbitrator mutually chosen by the parties but to the National Labor Relations Board (Stone, 1981, p. 1531).
To be sure, the parties would lose control over the arbitrator in the process. But Stone does not regard that as a liability. She projects no adverse consequences because- of her confidence in the NLRB bureaucracy and because of her preference for a judicial result: “If arbitration serves the function of a judiciary in a mini-democracy, then in theory, the arbitrator is to interpret the language of the written agreement, not please the parties” (Stone, 1981, p. 1552, n. 238; emphasis in original).
But suppose that both the immediate parties and society at large benefit from a more efficient exchange relation. If the pleasure of the parties is promoted by restoring a position on the shifting contract curve, and if that is better realized by voluntary arbitration than by. a legalistic (or black letter) approach to contract, then little wonder that the parties Opted for private ordering.
Indeed, had the judicial model favored by Stone been adopted, the parties would presumably have recognized the problems and would have redefined their relationship appropriately. Rather than submit to the inefficiencies of legal centralism, the parties would attempt to craft private orderings. And if such efforts are resisted and defeated by the judiciary, the parties will then predictably respond by avoiding technologies and working relations in which continuity is valued.
The simple contracting schema set out in Chapter 1 is again relevant. Recall that the parties first have to reach a decision on whether or not to make specific investments. If k = 0, a discrete market contracting relation obtains. If instead a k > 0 decision is made, then a further decision has to be made on whether or not to safeguard the transaction with protective governance structure. The industrial pluralists believed that it was useful to “please the parties,” hence a specialized governance structure designed to harmonize interests and promote continuity was devised. Stone rejects that in favor of litigation. If, however, insistence on the latter places an inefficiency burden on the parties, which it arguably does, efforts to void that result can be anticipated. Reverting to a nonspecific (k = 0) labor relation is one possibility.
2. The Many Facets of Unions
The dominant view of unions is that they are organizations whose main purpose is to raise wages—what Richard Freeman and James Medoff (1979) refer to as the monopoly face of unionism. There is a growing appreciation, however, that (1) unions serve other important purposes and (2) the functions served by unions vary systematically with the nature of the task. These are the matters of special interest here.
One of the reasons why the monopoly face of unionism has received so much attention in relation to other aspects is that monopoly analysis is congenial to neoclassical economics. A second reason is that confrontational wage bargaining is much more newsworthy than is reporting on humdrum, day-to-day governance. Neither, however, justifies the neglect of the other two faces of unionism: efficiency and voice. Efficiency raises governance issues of a transaction cost kind (Williamson, Wachter, and Harris, 1975), while voice has been addressed and elaborated by Freeman (1976) and Freeman and Medoff (1979). All three faces of unionism—monopoly, efficiency, and voice _ have to be recognized if the organization of labor is to be accu rately assessed. Furthermore, there are distinguishable facets within each of the three faces, the separate treatment of which will further contribute to a more informed assessment of unionism.
2.1. THE MONOPOLY FACE
Price discrimination aside, monopoly manifests itself as a condition of contrived scarcity. At least three distinguishable types of unions that aspired to raise wages, presumably by controlling supply, can be identified: class unions, craft unions, and industrial unions.
The first represented workers as a group in relation to employers as a group. Attempts in the United Stales-iir the nineteenth century to organize along those lines (National Ltf6or Union, Knights of Labor) were not successful. A substantial problem with such an approach to unionism is that economic differences among workers and jobs are ignored or suppressed. Craft unions, by contrast, are organized along much narrower lines. The organizational features and wage bargaining attributes of the union can be adjusted to the nature of the job much more fully if the craft union model is adopted. Glenn Porter’s examination of the early history of labor unions is instructive:
[R]ailroad workers were the first to achieve genuine collective bargaining and grievance channels through their national unions, the railroad brotherhoods. Initially these unions, like many other early American labor organizations, were social and mutual benefit societies. By the 1870s, though, they were evolving into modem unions. Like many of the craft unions which formed the American Federation of Labor in the 1880s, the railway brotherhoods derived economic ‘ strength from the fact that their members had scarce and hard to replace skills. A strike by such a union was a real threat to employers, because it was extremely difficult to break „the strike by bringing in outside workers (“scabs” in union parlance). Furthermore, the railway workers were additionally vital because they controlled the use and maintenance of expensive equipment. The unhappy history of unions that tried to include all the nation’s working people, such as the National Labor Union and the Knights of Labor, indicated that it was very difficult, if not impossible, to create and maintain unions unless the members had scarce economic skills like the railroad workers and the members of the craft unions that made up the American Federation of Labor in the 1880s. The all- inclusive unions faced other difficulties as well. Gerald Grob’s Workers and Utopia (1961) convincingly argued that the members and the leaders of such noncraft unions shared an ideological reluctance to accept the wage system. [Porter, 1973, pp. 34-55]
Industrial unions in the United States made their appearance after the passage of the Wagner Act. Requiring as they did the assistance of the political process, they evidently lacked natural advantages of either contrived scarcity or efficiency kinds. Unlike craft unions, industrial unions were or- ganizing workers whose jobs required little skill. Controlling entry—-e.g. through licensure—was for that reason difficult. Accordingly, incumbent workers in those industries were unable, without political assistance, to reach and enforce supracompetitive wage bargains—because “potential entrants” for their jobs would undo them. Additionally, industrial unions had relatively less to offer in the way of efficiency benefits, since simple governance structures are adequate to service the efficiency- needs of the employment relation for jobs that entail little human asset specificity.
The monopoly face of unionism is thus one in which control over entry is emphasized. It can be effected selectively by licensure among craft unions or more generally with the support of the political process, as with industrial unions. Class unionism appears to be feasible, however, only by effecting more massive political change in which capitalism is supplanted by socialism. Both Stone (1981, pp. 1579—80) and Branko Horvat (1982) appear to be in agreement on this last.
Unions can serve efficiency purposes in at least two respects. For one thing, unions can serve certain basic agency functions. In addition, and more important, unions can serve important governance purposes.
- Agency. The union as agent argument has been set out by Joseph Reid and Roger Faith (1980) and has been discussed by Freeman and Medoff under the heading of personnel practices and employee benefits (1979, pp. 82-84). Unions can both serve as a source of information regarding employee needs and preferences (with respect, for example, to fringe benefits) and assist employees in evaluating complex wage and benefit offers/Thus whereas it is unlikely, not to say inefficient, for individual workers to evaluate alternative compensation packages, unions can and do “hire the lawyers, actuaries, and other experts necessary to perform these analyses” (Freeman and Medoff, 1979, p. 83).
An agency role for unions is a purely instrumental one that permits the parties to reach and enforce preferred bargains. Virtually all types of labor can benefit from use of an agent to perform those functions. Workers in non- unionized firms will ordinarily recognize the benefits and will often develop a machinery (including a collectiveltieans by which to cover the costs) so as to realize them.
- Governance. Whereas the agency benefits of unions apply quite generally, the governance benefits apply in a more selective fashion. Indeed, this is the source of much of the predictive content of the transaction cost approach. The basic argument is by now familiar: Continuity of the em- ployment relation is valued by both employer and employee for tasks that involve the acquisition of significant transaction-specific skills, while tasks for which skill acquisition is insubstantial and/or general purpose do not create the same continuity interests.
Gary Becker, to whom much of the pioneering work on human capital is due, recognized that continuity interests would be manifested in incentive schemes. As he put it:
A pension plan with incomplete vesting privileges penalizes employees who quit before retirement and thus provides an incentive—often an extremely powerful one—not to quit. At the same time pension plans “insure’” firms against quits, for they are given a lump sum—the nonvested portion of payments—whenever a worker quits. Insurance is needed for specifically trained employees because their turnover would impose capital losses on the firm. [Becker, 1965, p. 18]
The issues here have subsequerjtfy’^been elaborated by Dale Mortensen (1978).
It was not, however, until 1975 that the collective organization ramifica- tions were recognized. The basic argument is this:
[A]lthough it is in the interest of each worker, bargaining individually or as a part of a small team, to acquire and exploit monopoly positions, it is plainly not in the interest of the system that employees should behave in this way. Opportunistic bargaining not only itself absorbs real resources, but efficient adaptations will be delayed and possibly forgone altogether. What this suggests, accordingly, is that the employment relation be transformed in such a way that systems concerns arc made more fully to prevail and the following objectives are realized: (1) bargaining costs are made lower, (2) the internal wage structure is rationalized in terms of objective task characteristics, (3) consummate rather than perfunctory cooperation is encouraged, and (4) investments of idiosyncratic types, which constitute a potential source of monopoly, are undertaken without risk of exploitation. [Williamson, Wachter, and Harris, 1975, p. 270]
Collective organization can be made to serve each of those purposes.
The mutual interest between workers and firm in protecting the em- ployment relation against exploitation by the other should have given rise to “company unions” in the pre-Wagner Act era. Although there evidently were some developments along those lines, they were scarcely widespread. Whether that reflected lack of knowledge of the benefits, apprehension over the potential monopoly uses of collective organization, or the fact that the efficiency benefits were rarely great is unclear.13 In any event, if the potential benefits of collective organization vary directly with human asset specificity, then the firms in which human asset specificity were greater were those that confronted a tradeoff in contemplating company union formation. By con-trast, firms in which human asset specificity was negligible experienced no such tradeoff: There were few gains to be realized, while the monopoly hazards were clear.
Two testable propositions thus emerge from examining collective organi- zation in governance terms: (1) The incentive to organize production workers within a collective governance structure increases with the degree of human asset specificity, and (2) the degree to which an internal governance structure is elaborated will vary directly with the degree of human asset specificity. Transaction cost analysis thus predicts that unions will arise early in such industries as railroads, where the skills are highly specific, and will arise late in such industries as migrant farm labor, where skills are nonspecific. It further predicts that the governance structure (job ladders, grievance procedures, pay scales) will be more fully elaborated in industries with greater specificity than in those with less (steel versus autos is an example). The preliminary data appear to support both propositions. The quotation from Porter under 3.2a above is germane in the first of these respects. Cox’s remarks about activating the arbitration machinery when disputes arise have a bearing on the second:
[G]iving the union control over all claims arising under the collective agreement comports so much better with the functional nature of a collective bargaining agreement … Allowing an individual to carry a claim to arbitration whenever he is dissatisfied with the adjustment worked out by the company and the union . . . discourages the kind of day-to-day cooperation between company and union which is normally themark of sound industrial relations—a relationship in which grievances are treated as problems to be solved and contracts are only guideposts in a dynamic human relationship. When the individual’s claim endangers group interests, the union’s function is to resolve the competition by reaching an accommodation or striking a balance. [Cox, 1958, p. 24]
Another aspect of arbitration, less widely remarked but also germane, is that arbitration “gives management a low-cost method of ascertaining when low-level supervisors are failing to follow the wishes of upper management. If the [labor rules] to which the management has agreed do in fact increase productivity, it is important to the firm that they be followed” (Vogel, 1981, p. 24). Arbitrary and capricious behavior by foremen is at the expense of the firm’s long-run interests. Practices that deter suboptimization will naturally be favored. More generally, “Employers who know that their actions are subject to arbitral review will seek to avoid unjustified discipline in the first instance by articulation of rules, instruction of foremen, careful investigation and other management controls” (Summers, 1976, pp. 507-8). Creating a governance unit to check myopic abuses is thus in the mutual interests of both labor and management.
The political side of unionism is described by Freeman and Medoff (1979) under the heading of “voice”—which has its origins in Albert Hirschman’s (1970) book in which he distinguishes, exit and voice as alternative means of organizing economic activity. Hirschman regards exit as the usual economic means for expressing preferences, while voice is the relatively neglected political process for influencing outcomes. Consumers, workers, voters and the like vote with their pocketbook or with their feet in the former case. Voice, by contrast, involves dialog, persuasion, and sustained organizational effort.
Freeman and Medoff impute efficiency, distributional, and social organi- zation effects to the collective-voice view of unionism. The efficiency benefits they ascribe to voice, however, are essentially those described above in conjunction with agency and governance. But there is a difference. Whereas the voice view of unionism attributes beneficial governance features to union organization quite generally, the transaction cost (or governance) approach predicts that they will vary with the continuity needs of the parties. As set out in section 2 above, those continuity needs are greatest where human assets are more highly specific. Spot market contracting will continue to be efficacious, however, where human asset skills are nonspecific and inputs are separable. The voice view of unionism evidently holds otherwise, in that it ascribes efficiency benefits to unions in all circumstances—spot market contracting included. In principle, a discriminating test of the governance versus voice approaches to unionism can thus be had by examining the efficiency benefits of collective organization associated with the k0, S0 cell (which is the case where assets are nonspecific and separable).
The merits of the voice view of unionism do not, however, turn entirely on efficiency ramifications. Distributional considerations also warrant attention. Effects of two kinds have been ascribed to unions. The first is the conventional monopoly distortion associated with union wage gains. The relatively neglected feature to which Freeman and Medoff call attention, which is distinctively associated with the voice view of unionism, is that income inequality—within a firm or within an industry—is reduced among organized workers.
The social effects of unions, according to the collective-voice approach, is that unions are political institutions that represent both the will’of their members and the political interests of lower-income and disadvantaged persons. The latter effects might be disputed, but unions plainly are important to the political process—in health and safety as well as in other societal respects.
Those effects would be difficult to actualize under the agency view of unions, which locates union activity at a plant or firm level. The joining of union interests at political levels is apt to benefit from composite (or hierarchical) union organization, which the collective-voice view of unionism arguably supports. ,
The claim or suggestion that power rather than efficiency is responsible for decisions to organize exchange relations one way rather than another runs through much of the social science commentary on labor organization. Rarely is power defined, however. Partly this is because it is widely believed that while power “may be tricky to define it is not that difficult to recognize” (Pfeffer, 1981, p. 3). I submit, however, that much of what is “recognized” as power is the result of looking at individual contracts in an ex post state rather than, as comparative institutional analysis requires, considering the set of relevant contracts in their entirety.
Sometimes what is referred to as power reduces to a preference for an alternative distribution of income. Those who have fewer resources would have greater purchasing “power” if this were accomplished. But the organization of labor need not be affected on that account. To be sure, the mix of goods and services would probably change. But the way in which work is organized need not. Indeed, i?efficiency is driving organizational outcomes, modes that are efficient under one distribution of income will normally remain efficient under another.93 Since mutual gains are potentially available whenever a move from a less to a more efficient configuration is accomplished, the incentive to choose more efficient modes is transparent.
Issues that require clarification in evaluating the power literature include the following: (1) nondiversifiable risk, (2) reputation effects, <3) competitive process, and (4) worker control over the intensity and quality of the labor input. A common misconception that runs through much of the power literature is that aggregate power can be inferred by ascertaining which of two contestants will win in an isolated confrontation.
3.1. NONDIVERSIFIABLE RISK
Branko Horvat contrasts owners with workers in risk-bearing terms as follows: “The owner can spread the risks by acquiring a diversified portfolio of shares, while the worker has just one labor power and one job” (1982, p. 447). Several observations are relevant. First, there is a narrow technical sense in which the notion that labor is nondiversifiable is correct. Legal prohibitions against indentured servitude aside, a market in human capital in which risks are diversified through the buying and selling of shares in individual income streams is not viable. “One cannot, for example, sell a piece of oneself if one is a lawyer in Cincinnati and buy a portion of a carpenter in San Diego” (Gordon, 1974, p. 447).
But second, it is possible for workers to choose between general purpose and firm-specific skills. Workers who choose the former will be qualified to work for a large number of employers. Only those who invest very heavily in ‘firm- specific skill acquisition are accurately described as ‘ ‘one labor power and one job.” Even here, alternative employment, albeit at reduced levels of productivity, is ordinarily feasible. Furthermore, and more important, workers who accept employment of a firm-specific kind will presumably recognize the risks and insist upon surrounding suchjobs- with protective governance structures. One labor power and one^jdbregarded nakedly and one labor power and one job embedded in a protective governance structure have very different connotations.
3.2. REPUTATION EFFECTS
The snapshot view of worker versus firm in power terms suppresses future consequences. Each confrontation is regarded separately, and the worker is inevitably the loser in each. The worker is terminated and has to find another job. The firm hires a replacement from the reserve army of the unemployed.
That scenario is defective in two respects. First, there is an implicit assumption that the firm experiences no dislocation costs when an employee is terminated. That is true, however, only if human asset specificity is negligible and any team effects are of the primitive team kind. Second, it assumes that workers have no employment options, hence do not choose among alter- »tives with an eye on different practices. But as Arthur Okun observes, “in the absence of an explicit contract, applicants will seek information from other workers about the employers’ past performance. Applicants are obliged to judge the employer, in part, by reputation” (1981, p. 51). Those firms with better reputations will presumably be able to hire workers on better-terms, ceteris paribus. (Again, the p versus p comparison of Chapter 1 is germane.)
Reputation effects are subtle matters, however. The possibility that the firm will use reputation in a strategic way presumably warrants consideration. Thus the argument that both workers and employers have a mutual interest in the continuity of the employment relation where investments in transaction- specific human assets are great might be disputed on the grounds that the worker is one of many and that the employer can and will realize strategic advantages by making an example of one or a few workers, thereby teaching a lesson to the many. The symmetry argument is thus mistaken, because it ignores this fundamental disparity. Put differently, the proposition that both parties stand to lose if an employee with significant transaction-specific skills quits or is terminated—the employee loses because he cannot turn to other employment without loss of productive value; the employer loses because of the costs of disruption and training that replacement entails—is correct as far as it goes, but it does not go far enough. It ignores the strategic aspects and for that reason must be supplanted by a strategic assessment of the employment relation in which asymmetries are recognized and explicitly taken into account.
The importance of reputation effects has been discussed by Christian von Weizsacker in conjunction with what he refers to as the extrapolation principle:
One of the most effective mechanisms available to society for the reduction of information production cost is the principle of extrapolation. By this I mean the phenomenon that people extrapolate the behavior of others from past observations and that this extrapolation is self-stabilizing, because it provides an incentive for others to live up to these expectations By observing others’ behavior in the past, one can fairly confidently predict their behavior in the future without incurring further costs. . . .
[This] extrapolation principle is deeply rooted in the structure of human behavior. Indeed it is also available in animal societies The fight between two chickens does not only produce information about relative strengths in the present, but also about relative strength in the future. [Weizsacker, 1980b, pp. 72- 73]
The issues have been developed more formally by David Kreps and Robert Wilson, who address the credibility of predatory threats. They show that where there is uncertainty about the dominant firm’s payoffs and where the dominant firm is engaged in repeated play with a sequence of opponents, “none of whom have the ability to foster a reputation” (Kreps and Wilson, 1980, p. 58), punitive behavior becomes a much more attractive policy.
A series of parallels between dominant firm and small rivals on the one hand and the employer and numerous individual employees on the other invite the transfer of that reasoning to the employment context. Thus (1) the em- ployer’s resources are much more extensive than are those of the typical employee, (2) individual employees may well have difficulty in assessing the payoffs to employers of alternative bargains, and (3) individual employees maybe thought of as a sequence of opponents, each of whom is unable to develop a countervailing reputation. The parallels, however, are incomplete. Specifically, whereas the dominant firm is dealing with rivals and hopes to have no further dealings with the would-be entrant or his ilk, the employer is dealing with suppliers and has continuing needs to hire workers. That difference can be decisive.
That is not to say that an employer cannot successfully teach a lesson to many employees by making an example of one or a few. Inasmuch as all employees who have made transaction-specific investments are vulnerable to exploitation, an employer may, through selective but conspicuous punitive measures, induce all incumbents to accept inferior terms. Reference to incumbents, however, flags an important limit on that type of behavior. Among employers who plan to be in business on a continuing basis, successor generation employees may also learn—and the lesson here is very different from that of the incumbents.
Specifically, employers who have a reputation for exploiting incumbent employees will not thereafter be abletp-induce new employees to accept employment on the same terms. A wage premium may have to be paid; or tasks may have to be redefined to eliminate the transaction-specific features; or contractual guarantees against future abuses may have to be granted. In consideration of those possibilities, the strategy of exploiting the specific investments of incumbent employees is effectively restricted to circumstances where (1) firms are of a fly-by-night kind, (2) firms are playing end games, and (3) intergenerational learning is negligible. In circumstances, however, where firms are continuously in the employment market and successor generations learn, efforts to exploit incumbent employees are myopic and will predictably elicit protective reactions.
There are good reasons, nonetheless, for employees to organize in such a way as to forestall even one-time or mistaken efforts to exploit the transaction- specific investments of labor. For one thing, exposure to myopic or fly- by-night operators and to end games can thereby be reduced. Second, intergenerational reputation effects may more assuredly be brought to bear on continuing firms by creating an institutional machinery to record and communicate incidents of expropriation. Third, employers may recognize the merits of fair dealing with employees but be unable to impress them on first-line supervisors. The collective organization of workers (unions) has advantages in each of the three respects.
3.3. COMPETITIVE PROCESS
The suggestion that employers rather than employees or society at large is the gainer whenever more efficient work practices are implemented as-sumes that workers lack bargaining power and neglects the competitive process. Work rule changes made during a contract are normally subject to arbitration, however. And those made during contract renewal negotiations are part of a much larger package in which tradeoffs are worked out.
The neglect of the competitive process is especially regretable. Thus assume that a more efficient practice can be identified and suppose that the employer initially appropriates the whole of the efficiency gain. Even though workers are no better off (indeed, depending on the particulars, some may be released and need to find new employment), society stands to gain in two respects. First, the resources saved by the reorganization of work can be productively reemployed in alternative uses. Second, the immediate profits that accrue to the firm will rarely be durable. Instead, the scenario described earlier in conjunction with the steel industry15 will normally obtain: significant reforms will be detected and imitated by others, and prices will fall as margins are restored to earlier levels.16 Further allocative efficiency gains are realized as a consequence.
3.4. WORKER DISCRETION
The suggestion that workers’ bargaining power is limited to one dimen- sion—report for work or strike—is common but mistaken. Thus Stone observes, “The availability of injunctive relief to force a union to cease striking . . . meant that [the] union’s alternative economic weapons were withdrawn” (1981, p. 1539). In fact, however, what “the firm wants when it hires an employee is productive performance It wishes to buy quality of work rather than merely time on the job” (Okun, 1981, p. 63).17 Accordingly, exploited incumbent employees are not totally without recourse. Incumbent employees who are “forced” to accept inferior terms can adjust quality to the disadvantage of a predatory employer. The issues here have been addressed previously in distinguishing between consummate and perfunctory cooperation (Williamson, 1975, p. 69). Of necessity, the employment contract is an incomplete agreement, and performance varies with the way in which it is executed. Consummate cooperation is an affirmative job attitude whereby gaps are filled, initiative is taken, and judgment is exercised in an instrumental way. Perfunctory cooperation involves working to rules and in other respects performing in a minimally acceptable way. As Peter Blau and Richard Scott observe:
[T]he contract obligates employees to perform only a set of duties in accordance with minimum standards and docs not assure their striving to achieve optimum performance… [L]egal authority does not and cannot command the employee’s willingness to devote his ingenuity and energy to performing his tasks to the best of his ability… It promotes compliance with directives and discipline, but does not encourage employees to exert effort, to accept responsibilities, or to exercise initiative. [Blau and Scott, 1962, p. 140]
In fact, most contracts—labor, intermediate goods, and other—are incomplete in significant respects, on which account suppliers enjoy discretion. Buyers thus “abuse” suppliers by demanding exacting performance in accordance with the letter of the contract only at hazard.
Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.