The Organization of Labor: An Abstract Approach

Although it is sometimes argued that there is a single preferred way by which to organize labor, casual inspection discloses that labor organization is highly varied. An explanation for the differences would plainly add to our under- standing. The transaction cost approach rests on the proposition that governance Structures for labor must be matched with the attributes of labor transactions in a discriminating way if transaction cost economizing is to be accomplished. To use a simple structure to govern a complex transaction will predictably have disruptive consequences—and possibly fracture the relationship—while to use a complex structure to govern a simple transaction is to incur excessive costs. The efficiency orientation employed here is broadly consistent with Frank knight’s view that

. . men in general, and within limits, wish to behave economically, to make their activities and their organization “efficient” rather than wasteful. This fact does deserve the utmost emphasis; and an adequate definition of the science of economics . . . might well make it explicit that the main relevance of the discus-sion is found in its relation to social policy, assumed to be directed toward the end indicated, of increasing economic efficiency, of reducing waste. [Knight, 1941, p. 252; emphasis added]

1. Dimensions


Recall that the principal dimensions for describing transactions are fre-quency, uncertainty, and asset specificity. The transactions of interest here ‘are ones of a recurring kind. Accordingly, frequency will be set aside and emphasis placed on uncertainty and asset specificity.

For the reasons given previously, the labor market transactions for which continuity between firm and worker are valued are those for which a firm- specific human asset condition develops. Note in this connection that skill acquisition is a necessary but not a sufficient condition for asset specificity features to appear. The nature of the skills also matters. Thus physicians, engineers, lawyers, and the like possess valued skills for which they expect to be compensated, but such skills do not by themselves pose a governance issue. Unless those skills are deepened and specialized to a particular employer, neither employer nor employee has a productive interest in maintaining a continuing employment relation. The employer can easily hire a substitute and the employee can move to alternative employment without loss of productive value.

Mere deepening of skills through job experience does not by itself pose a problem either. Thus typing skills may be enhanced by practice, but if they are equally valued by current and potential employers there is no need to devise special protection for an ongoing employment relation. Knowledge of a particular firm’s filing system, by contrast, may be highly specific (non- transferable). Continuity of the employment relation in the latter case is a source of added value.

Thus whereas neoclassical reasoning links skills to productivity and compensation, transaction cost reasoning introduces organizational considera- tions. Specifically, skills that are acquired in a leaming-by-doing fashion and that are imperfectly transferable across employers have to he embedded in a protective governance structure lest productive values be sacrificed if the employment relation is unwittingly severed. The argument here is related to but is usefully distinguished from Gary Becker’s (1962) argument that compensation structures vary systematically with human asset specificity. That is correct as far as it goes, but it ignores important organizational features of the employment relation. Rules governing ports of entry, job ladders, bumping, grievance procedures, and the like are all part of the employment relation broadly conceived (Doeringer and Piore, 1971) but are not treated by Becker. Transaction cost economics maintains that governance structures must be crafted more carefully as The degree of human asset specificity increases.

The second dimension on which the study of labor market transactions turns is that of uncertainty. As with intermediate product market transactions, an increase in parametric uncertainty is troublesome mainly for transactions in which human asset specificity is great. Possible transition costs aside,5 neither party has a continuity interest in transactions in which asset specificity is negligible. Presumably, therefore, increases in parametric uncertainty will be reflected in the original terms upon which such bargains are struck. But the governance of transactions in which asset specificity is slight will otherwise be unaffected.

Transactions in which continuity is valued, by contrast, are placed under greater stress as parametric uncertainty increases. Such transactions will have to be adapted more frequently or extensively to restore a position on the shifting contract curve. Such contracts will therefore be negotiated more carefully. Also, as will be discussed in section 3 below, the contract renewal interval may be shortened. Whatever the length of the contract interval, management will have greater latitude of adjustment (e.g. be less encumbered by work rules) where uncertainty is great than would otherwise be the case. Lest such latitude be exercised irresponsibly, procedural safeguards to provide for the early and equitable disposition of grievances become more impor-tant.

The argument thus far mainly’repeats that made earlier in conjunction with intermediate product markets. The organization of labor, however, poses additional complications that have not previously been encountered. They arise in conjunction with the measurement of work product.


Although measurement issues of both ex ante screening and ex post execution kinds are germane to the organization of labor, the latter are given principal emphasis here. The team organization problems to which Alchian and Demsetz (1972) refer and the issues dealt with by Ouchi in distinguishing between “behavior control” and “output control” (1978, pp. 174-75) are the matters “ot Interest. The compfiEaiionlvfth which Alchian and Demsetz are concerned has its origins in a condition of technological nonseparability. The condition was earlier recognized by Marx (1967, vol. 1, chap. 13) and is illustrated by Alchian and Demsetz with the manual freight loading example:

Two men jointly lift cargo into trucks. Solely by observing the total weight loaded per day, it is impossible to determine each person’s marginal productivity… The output is yielded by a team, by definition, and it is not a sum of separable outputs of each of its members. [1972, p. 779]

Where tasks are nonseparable in that sense, individual productivity cannot be assessed by measuring output. An assessment of inputs is needed. Sometimes it may be inferred by observing the intensity with which an individual works, the aspect emphasized by Alchian and Demsetz. A monitor is created so as to discourage shirking. Often, however, the assessment of inputs is much more subtle than eTlorf accounting. Does the employee cooperate in helping to devise and implement complex responses to unanticipated circumstances, or does he attend to own or local goals at the expense of the others?6 Those concerns are especially great where the members of the team develop idiosyncratic working relationships with one another, in which case no single member can be replaced without having disruptive effects on the productivity of the unit. More complex teams in which mutual motivation and internal monitoring are encouraged are apt to take shape in such circumstances.


Letting k0 and it, represent low and high degrees of human asset specificity and S0 and S, represent separable and nonseparable work relations, respectively, the following four-way classification of internal governance structures is tentatively proposed:7

  1. kQ, S0: internal spot market

Human assets that are nonspecific and separable are meeting market tests continuously for their jobs’. Neither workers nor firms have an efficiency interest in maintaining the association. Workers can move between employers without loss of productivity, and firms can secure replacements without incurring start-up costs. No special governance structure is thus devised to sustain the relation. Instead, the employment relation is terminated when dissatisfaction by either party occurs. An internal spot market labor relation may be said to exist. Examples include migrant farm workers and custodial employees. Professional employees whose skills are nonspecific (certain drafts- men and engineers) also fall in this category. Such jobs appear to be of the kind that Arthur Okun had in mind in his reference to the use of “brokers” to help supply labor where “the jobs at stake . . . require unskilled workers (like farm workers) or transitory workers (like office fill-ins) or involve formally graded skills (as is the case when unions certify craftsmen in construction, longshoring, and printing)” (1981, p. 63).

  1. k0, S,: primitive team

Although the human assets here are nonspecific, individual output cannot be fnetered easily. This is the team organization to which Alchian and Demsetz refer (1972). Although the membership of such teams can be altered without loss of productivity, compensation cannot easily be determined on an individual basis. The simple brokerage role described above is thus extended to include supervision. As John Pencavel indicates, “the Italian padrone at the turn of the century not only advanced credit to poorly-informed immigrants and occupied the role of the go between for laborers and employers, but also sometimes acted as foreman and paymaster” and the jamadar in India both “enlists and supervises workers for firms in the construction industry” (1977, pp. 251-52, n. 5). More generally, the coupling of employment with an oversight assignment is involved. The structure is referred to as a primitive team to distinguish it from the relational team described under 4, below.

  1. kt, S0: obligational market

There is a considerable amount of firm-specific learning here, but tasks are easy to meter.9 Idiosyncratic technological experience (as described, for example, by Doeringerand Piore, 1971, pp. 15-16), and idiosyncratic organizational experience (accounting and data processing conventions, internalization of other complex rules and procedures, and the like) both contribute to asset specificity. Okun’s “toll model” of employment applies, in that sunk costs are incurred in qualifying a worker for productive employment in the firm (Okun, 1982, pp. 49- 77). Both firm and workers have an interest in maintaining the continuity of such employment relations. Procedural safeguards and monetary penalties, such as severance pay, will thus be devised to discourage arbitrary dismissal. And nonvested retirement and other benefits will accrue to such workers so as to discourage unwanted quits (Mortenson, 1978).

  1. ky, 5,: relational team

The human assets here arc specific to the firm and involve a significant team aspect. This approximates Ouchi’s (1980b) “clan” form of organiza-tion. The firm here will engage in considerable social conditioning to help assure that employees understand and are dedicated to the purposes of the firm, and employees will be provided with considerable job security, which gives them assurance against exploitation. Effective adaptation in a cooperative team context is especially difficult and important to achieve. A sense that management and workers are “in this together” furthers all of those purposes.

The Japanese corporation is said to have those attributes (Dore, 1973; Clark, 1979; Gibney, 1982) and both Ouchi (1981) and Fred Foulkes (1981) contend that some large American corporations have crafted the same. In addition to a governance structure in which employment safeguards are provided and respected, Foulkes contends that “fiercely egalitarian” practices— same parking lot, medical benefits, cafeteria, and spartan offices—for both management and labor contribute to “a high degree of employee, loyalty, a low rate of turnover and absenteeism, and low degree of resistance to technological change” (1981, p. 90).

The above-described match of internal governance structures with internal transactional attributes is summarized in Figure 10-1. Admittedly, describing internal transactions in bivariate, binary terms simplifies considerably. The overall framework is nevertheless in place and refinements can be made as needed. (Thus mixed internal governance structures will presumably arise to service transactions that take on  intermediate, rather than extreme, k and S values.)

The foregoing discussion makes no explicit reference to union organiza-tion. It nevertheless has significant ramifications for both the attitude firms display toward unions and the structure of the employment relation among unionized employees. The issues here are developed in Section 3. Before turning to that, however, consider first whether the hazards of expropriation posed by workers vitiate the argument that market organization and internal organization differ in appropriability respects.

FIGURE 10-1. Efficient Work Organization

2. Expropriation by Workers

The mistaken argument is this: Shifting a transaction out of the market into the firm merely relocates an expropriation hazard but otherwise leaves the trading hazard intact. The reasoning here is that appropriable assets that are not exacted by trading opposites in an intermediate product market transaction are exposed to an equivalent expropriation hazard by employees of the firm should integration be attempted. There being no safe haven for specific assets, the transaction dost rationale for vertical integration collapses.

The issues will be clarified by breaking the argument down into parts:

  1. Which assets are subject to expropriation?
  2. In what ways, if any, do employment contracts differ from commercial contracts?
  3. What are the ramifications for corporate governance? Consider these seriatim.


It is presumably beyond dispute that employees will bargain over their share of the quasi-rents that are embodied in human capital. That has a well- defined upper limit. The expropriation argument has reference not to that but to investments made by suppliers of capital (owners of equity; long-term debt holders). Site-specific investments in plant and equipment and idiosyncratic physical capital are both candidates.

The magnitude of those equity and debt supported assets could greatly exceed the amount of firm-specific human capital. Absent property rights to jobs, however, the upper limit of worker expropriation is the amount of firm- specific human capital. Employees risk—indeed, invite—termination should their demands exceed that limit.10 Property rights considerations aside, therefore, the potentially appropriable quasi-rents to which employees haye access are those reflected in human assets. Plainly this is a smaller magnitude than the total quasi-rent exposure, which includes transaction-specific physical capital.


Employment contracts and commercial contracts differ in the following fundamental respect:

Even where the collective agreement lists certain offenses or the parties negotiate plant rules, management may normally supplement the listed offenses or negotiated rules. Rules prescribed by management are subject to arbitrator review, but they carry a presumptive validity and will be upheld so long as they are reasonably related to achieving efficient operation  and maintaining order and are not manifestly unfair or do not unnecessarily burden employees’ rights.

Management also is entitled to have its orders obeyed and may discipline employees for refusing , improper orders. Arbitrators almost uniformly hold that an employee must obey first and then seek recourse through the grievance procedure, except where obeying would expose him to substantial risks of health and safety. [Summers, 1976, p. 502]

Procurement of the same good or service from an autonomous supplier or- dinarily lacks that command and control aspect but requires mutual consent before adaptations can be affected.

Internal organization is thus able to adapt more effectively than can inter- firm trading to changing market and technological circumstances. Not only do employment contracts contemplate such flexibility by providing for “zones of acceptance” within which orders will be implemented without resistance, but orders that exceed the scope of the authority relation can be implemented in extreme circumstances. Commercial agreements lack the same degree of re- sponsiveness, since to write a commercial contract that awards buyers with effective control over the assets of a seller is to risk asset abuse if not expro-priation. The issues here are akin to those addressed in Chapter 6.

The point can be made somewhat differently as follows: “When a number of persons are participating in a decision-making process, and these individuals have the same operational goals, differences in opinion will be resolved by predominately analytic processes. . . . But when goals are not shared . . . the decision will be reached by predominately bargaining processes (March and Simon, 1958, p. 156). Acquiescence, if not goal sharing, is presumed for activities that fall within the zone of acceptance, so that analytical processes predominate. Bargaining is more prevalent between autonomous trading parties whenever efforts are made to restore the parties to a position on the shifting contract curve.

The upshot is that the opportunities to haggle rather than adapt are vastly greater for commercial contracts than they are for employment contracts. As a consequence the asserted advantages of internal organization as compared with market organization survive.


Issues of corporate governance are discussed in detail in Chapter 12. Suffice it to observe here that corporate governance has to be sensitive to expropriation hazards. Property rights over jobs, the uses of corporate debt and the institution of bankruptcy, and control over the board of directors all have a bearing.

Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.

Leave a Reply

Your email address will not be published. Required fields are marked *