The Employment Relation in a Free Enterprise System

The discussion of the central contracting agent in the classical capitalist firm explains how the authority relation can prevent input providers from exploiting the firm by forcing contract renegotiations when unforeseen contingencies arise, but it does not explain what prevents the firm from exploiting input providers.

Another reason for addressing this question at the systemic level is that the employment relation in a free enterprise system is fundamentally dif­ferent from that relation in a market socialist system because of differences in the structures of the predominant organizational forms. For these rea­sons, a comparative evaluation of the two types of systems on the question of exploitation requires an examination of this relation in both types of sys­tems. This section discusses the question of whether or not firms in a free enterprise system systematically exploit labor; the second section of chapter 6 will discuss the same question in reference to a market socialist system.

It has been a staple of socialist and leftist thought for the past century and a half that the workers are systematically exploited in a free enterprise sys­tem.

The framework for discussing exploitation provided in the last chapter will prove useful in this regard. Let us begin by identifying the general conditions that would have to be met for a worker to be exploited by a firm in a market economy generally. It will then be possible to address the ques­tion of whether or not these conditions are standardly met in the employ­ment relation in a free enterprise system.

A worker’s asset is his or her capacity to labor. Either that asset has appropriable quasi-rents attached to it or it does not. Suppose it does not. This implies one of two things: either his assets have no quasi-rents at all or the quasi-rents are not appropriable. If the former is the case, the worker would earn about the same wage he would get if that asset were deployed in its next best use; this would be the case with unskilled labor in a tight labor market. If there are quasi-rents associated with his labor assets but they are not appropriable, that means that there are competitors willing to offer the worker about the value of what he is selling. For example, doctors have highly specialized skills, and the next best use for their capacity to labor is generally not nearly as attractive; nevertheless, the quasi-rent value of their assets is well protected because demand for their services as doctors is high, and alternative employers are relatively easy to find.

Notice that given the way the concepts ‘quasi-rent’ and ‘appropriable quasi-rent’ are defined (Klein, Crawford, and Alchian 1978, 298), there is no implication that the asset owner in question knows about the other potential ρurchaser(s) of what he is selling. All that matters is that there are other potential purchasers out there, so to speak. There does seem to be a tacit supposition to the effect that if an asset owner’s quasi-rents are nonap- propriable, then both the asset owner and the purchaser of the asset’s ser­vices know this and act accordingly. If this were the case, then if the work­er’s assets have no appropriable quasi-rents, then ipso facto he gets the full value of his assets. However, suppose that the quasi-rents meet the defini­tion of nonappropriability in the sense that there are other purchasers out there willing to pay the full value of the asset, but the asset owner does not know about these other purchasers out there waiting to be found at a rea­sonable search (and asset redeployment) cost. Such an asset owner has sim­ply failed to search out these other purchasers. This means that while the quasi-rent value of his asset may not be appropriable according to the defi­nition, some of that asset’s value may get appropriated anyway, since the asset owner did not avail himself of his next best opportunity.

In the case of a worker, this means that a worker may fail to get the full value of his services because he does not know about, or has not sought out, better alternative employment opportunities. As in the case described in chapter 3 of the woman who sold the heirloom, this worker does not get a “fair” price for what he is selling, since part of the value of his labor has been appropriated by the purchaser. However, even though he is not get­ting the value of his contribution, he is not being exploited in this situation because there are available alternatives that he has simply failed to investi­gate or discover. Recall that there are two necessary conditions for exploita­tion: the person is not getting the value of what he contributes, and he effectively has nowhere else to go.

The upshot of all this is that if the workers are systematically exploited in the employment relation, part of the value of their labor assets must be appropriable quasi-rents. If the quasi-rents of their assets are nonexistent, then they are getting the full value of their contribution. On the other hand, if their assets have associated quasi-rents but they are not appropri­able in the sense implied by the definition, then whether or not they are paid the full value of their respective contributions, the workers are not being exploited. Conversely, if a worker’s assets have quasi-rents and they are appropriable, then some of the value of those assets can be appropriat­ed without his being in a position to do anything about it; that is, he is in a position to be exploited.

The question now becomes, Does this happen on a regular basis and for systematic reasons in a free enterprise system? What is needed to sustain a positive answer to this question is some structural feature of free enterprise systems that leaves all or nearly all workers’ quasi-rents permanently exposed. One apparently promising candidate for that feature is the alleged fact that firms exercise disproportionate bargaining power in con­tract negotiations with prospective workers. The way this is usually explained is that it is easier for employers to find workers to accept a low wage (i.e., one that is less than the value of the worker’s labor) than it is for a worker to turn that offer down. Unlike the firm, the worker has nowhere else to go. This is why workers are in general exploited by the firms that hire them. The problem with this is that as an explanation, it is a non­starter: it simply restates the allegation to be proved. The reason is that to say that an employer actually exercises (and not merely has) disproportion­ate bargaining power in contract negotiations with worker X is to say three things: (1) the employer’s offer is less than the value of worker X’s labor, (2) the employer’s next best alternative to hiring worker X at the offered rate is as good or better than hiring worker X (suppose it is to hire a compara­ble worker at the same wage); and (3) worker X’s next best alternative is about the same or significantly worse than accepting that offer.

One phenomenon that might do some explanatory work in this con­text (and thus serve to justify the claim about exploitation) is that free enterprise systems produce downturns in the business cycle that bottom out in recession or depression. It might be argued that when macroeco­nomic conditions take a turn for the worse, workers are not in a position to flee if and when firms make a grab for their quasi-rents, which they could do by, for example, asking for give-backs or by not giving raises to match the rate of inflation. It might be further argued that continued erratic economic performance, including high rates of unemployment, makes it possible for firms to persist in offering workers less than the value of their contributions.

There are two problems with this line of argument. One is that it assumes that free enterprise systems are inherently unstable at the macro level. Among economists who favor the free enterprise system, there is a long tradition of disputing this. They have argued that depressions and recessions in existing societies with free enterprise systems are state- induced destabilizations brought on by the latter’s inflation or deflation of the money supply.³⁰ If their analysis is correct, free enterprise systems are not inherently unstable at the macro level, so that even if the workers are exploited in the way suggested, it is not the free enterprise system that is responsible. It is instead the state or the political system that is to blame.

But even if that is not true and a free enterprise system (i.e., any free enterprise system) is inherently unstable at the macro level, the conditions that accompany depressions and recessions do not allow the workers to be exploited by the firms that hire them. Vulnerable though they may be, the workers’ problem under these circumstances is not exploitation. To see why, recall from chapter 3 that the value of an asset is determined by what it would fetch in a competitively efficient market or on the leading edge of a market in transition, where a competitively efficient market is defined in terms of a local equilibrium in the determinants of the scarcity value (e.g., natural resources, labor skills, technology, distribution of wealth and income) of the asset in question.³¹ In times of changing economic condi­tions, the value of assets, including labor assets, is changing; in a depres­sion or recession, that change is usually in the direction of a decline.

This means that if an asset earns a lower return than it used to, it does not follow that its owner is having some of its value appropriated by the purchaser of that asset or the asset’s service. The terms of the exchange may be on the leading edge of a market in transition. For example, if aggregate demand has fallen precipitously in a depression, workers may find they have no choice but to take a cut in wages. Since the underlying value of their assets has declined, however, this cut reflects a revaluation of their assets and not the appropriation of quasi-rents by their employer. Indeed, if their wages are contractually guaranteed for a period that extends into the new macroeconomic environment, they are in a position to exploit their employers; they would be getting more than the value of what they are contributing, and their employer would have nowhere else to go. To take another example, population growth or immigration may make the workforce grow faster than new workers can be absorbed by the economy; once again, workers may find that they have no choice but to take a cut in wages. Although the workers are suffering a decline in wages, they are nonetheless getting the full value, including the quasi-rent value, of their assets. It may be unfortunate—indeed, it may even be unjust according to some conceptions of justice — that workers undergo the accompanying hardships while (one may suppose) their bosses or the firm’s owners do not. But the former are not being exploited by the latter. The firm’s owners or managers are simply bringing the workers the bad news about the declin­ing value of what the workers are selling, namely, their labor. Employers get this bad news when they sell the firm’s products in the market; they then pass the news on to workers as quickly as possible. As has been repeat­edly urged in this book, markets serve an important information-transmit­ting function in a world of changing economic conditions.

The general point here has other applications. Suppose that the market for teachers in a free enterprise system is competitively efficient, as that notion was defined in chapter 3. Someone might nevertheless believe that teachers are exploited because their pay is so low relative to jobs that are, or seem to be, much less socially valuable. According to this socialist, this is a classic example of a “social irrationality” thrown up by the operation of the market. However, whether or not it is a social irrationality, it is not an instance of exploitation. Why not? It may be true that if the distribution of wealth and income were different, teachers would make much more—and others Withjobsjudged less important, much less—than under the existing system. It may even be that at some level, justice requires a different distri­bution of wealth and income. However, neither proposition changes the fact that the teachers are getting the value of their contributions and thus are not exploited.

Indeed, the complaint against this society can be usefully rephrased as a complaint against its values, as those values have emerged in and through the market. The free enterprise system is the messenger which brings the news about the value of people’s contributions. Sometimes the news is not good. A paradoxical way of making this point is to say that what is wrong with this society, according to our hypothetical socialist teacher, is that its values are so twisted that teachers can make paltry salaries and yet not be exploited!³² This is true even if a free enterprise system is responsible for creating those values, which (supposing all of this to be true) constitutes yet further grounds for criticism of this type of system. A general lesson here, as in chapter 3, is that it is important that the charge of exploitation not be conflated with other charges a social critic might want to make against a free enterprise system.

Is there any other way that workers might have their quasi-rents sys­tematically appropriated? Perhaps the most promising place to look is cases in which employees make relatively specific investments in developing their skills or knowledge; then, once the fundamental transformation has taken place and they have entered into a long-term employment relation, they may have no comparable alternative employment opportunities. Under these circumstances, their quasi-rents would be vulnerable to appropriation by opportunistic employers.

However, a number of policies or procedures have evolved in existing free enterprise systems to protect workers’ quasi-rents under these circum­stances. One form of protection is the collective bargaining agreement. Some features of collective bargaining serve to limit the exposure of the quasi-rents of employees’ relatively specific assets, whether they are intend­ed to have this effect or not. Union contracts typically require that the com­pany observe due process before firing workers. This serves not only to protect workers who have built up firm-specific assets but also as a counter­weight to the fact that dismissal imposes disproportionate costs on workers (of whatever skill level) because of its disruptive effects on family and social life. Unions also negotiate other personnel practices and evaluate wage and benefit offers (Reid and Faith 1980; Freeman and Medoff 1979). In this way, knowledgeable negotiators can help to ensure that workers are receiv­ing the approximate value of their labor assets; if they are getting that value, then despite the fundamental transformation that locks them into that employment relation, they are not being exploited.

Indeed, it is possible to give a transactions cost explanation for the exis­tence (or persistence) of unions.³³ What unions do is bring large amounts of human capital under one decision-making roof. When workers are unionized, firms bargain not with individual workers but with the larger collective entity that represents them. Unions effect a kind of horizontal integration of human capital. This can serve to protect the quasi-rent val­ues of their members’ assets in a manner that parallels the protection pro­vided by vertical integration of the quasi-rent values of nonhuman assets. Of course, unions own neither their members nor their members’ human capital, so they are not the asset owners but their agents. This raises the possibility of principal-agent problems, problems that have been serious in cases where the union leadership has used the assets at its disposal to fur­ther a personal or political agenda not endorsed by the membership. In some cases, their only real monitor has been the Mafia (an organization known for its effectiveness in monitoring).

These problems to one side, the efficiency functions that unions serve are, of course, not their only functions; they also serve a monopoly function and a more political, “voice” function (Hirschman 1970)—which leads, among other things, to reduced income inequality among workers (Freeman and Medoff 1979). Arguably, both functions are, at bottom, exploitative. The coercive and quasi-coercive power that unions wield effectively closes off alternatives for employers and prospective nonunion employees, and it allows unions and their members to appropriate the quasi-rents of the firm’s equity owners.³⁴ If unions or their members are exploiters, however, it would not be the first time that an organization that serves to mitigate some social vice in one respect aggravates that very prob­lem in some other respect.

Absent unions, what institutional structures are available to employees to protect the value of the relatively specific assets they bring to the job? As a way of enticing prospective employees, employers will often voluntarily adopt personnel policies and procedures (such as due process rights) to ensure that workers will not be Bred or have their pay cut unless it is war­ranted by changes in economic conditions and thus in the value of their services. Another way that Rrms can bond their promises not to act oppor­tunistically toward employees is to pay substantial severance pay to those whom it lets go. This serves the double function of providing the dismissed employee with a cushion against the blow of temporary unemployment and of binding the employer to a policy of laying off employees only if business conditions require it. These and other “up-front” policies are usually more effective than a firm’s general reputation, which may be distorted or not widely known among prospective employees. Reputation effects do, how­ever, have their role to play in purely local markets.

Sometimes the ways in which employees’ transaction-specific assets are protected are not directly evident; the up-or-out system in corporate law firms (also common in other professions), which was discussed in the last chapter, is a case in point. To put this in terms of the protection of quasi­rents, recall that associate attorneys develop substantial firm-specific assets during the probationary period leading up to the partnership decision. Toward the end of that probationary period, the associate is being billed out at a relatively high rate and yet is not making anything close to what she is earning for the firm (even taking into account overhead, etc.). Because some of her assets are so firm-specific, her next best alternative employment opportunity might be significantly worse than staying on as an associate at that particular firm. Clearly, no one would want to leave her quasi-rents exposed for too long, especially to lawyers. Commitment to the up-or-out policy prevents this from happening. By publicly adopting this policy, the firm effectively commits itself to making the partnership deci­sion on the merits of the case, since it cannot opportunistically deny the associate promotion and then turn around and offer to keep her on indefi­nitely at her current rank and pay, thereby appropriating the considerable quasi-rents that she has built up over the years.

Notice that just as employees are potentially vulnerable to opportunism by firms, so, too, are firms potentially vulnerable to having the quasi-rents of their assets opportunistically appropriated by employees. The firm may make a capital investment in equipment especially designed for a particular employee, or, more commonly, it may spend resources to train an employ­ee for his or her job. This training, which the employee gets to keep, as it were, may consist wholly or in part of knowledge and skills that can be redeployed elsewhere. It is as if the Hrm gives the employee a tool that it cannot take back when the employee leaves the firm. How do firms protect investments like this from being opportunistically appropriated by their employees?

One way is by not vesting workers in a pension until they have accumu­lated many years on the job (Mortenson 1978; Becker 1962). This gives workers a strong incentive not to quit before retirement. Other ways include tying pay to positions, instead of to individuals, and promoting from within on the basis of merit (Putterman 1984, 176). This encourages individuals to stay with the Hrm and use the skills they have acquired to benefit the firm. Finally, a firm may involve the employee in a process of social conditioning so that he or she identifies with the firm and its goals. This is especially important in team production, which requires a high degree of cooperation among its members. Foulkes contends that the fiercely egalitarian treatment of managers and workers in many Japanese firms and some American Hrms is an instance of this type of conditioning and serves to protect the firm’s investments in human capital and the value of other specialized assets employed by the team (Foulkes 1981).

Alternatively, when these values or investments cannot be protected, the firm simply pays the employee less than the value of her services for as long as it can as a way of recouping its investment in the employee that will be lost when the employee quits. For example, the career path in major public accounting firms provides opportunities for staff accountants to acquire extremely valuable knowledge of the workings of various industries (finan­cial institutions, real estate, manufacturing, health care, insurance, etc.), as well as knowledge of how particular client firms operate. This involves not just learning by doing but also specialized and costly training sessions paid for by the firm.

These accountants are very often hired away in two or three years by the very firms and industries they have audited. Accounting firms compen­sate themselves for this lost investment in their employees by underpaying and overworking everyone on the professional staff below the rank of part­ner. They also benefit by having former staff accountants in client firms because their former employees can shape the client firm’s accounting poli­cies and procedures to facilitate the yearly audit, thereby making it possible for the accounting firm to underbid its rivals. Staff accountants who stay on to become partners in the public accounting firm recoup this underpay­ment, as do those who leave the firm in a few years with their newly acquired human capital for more lucrative opportunities outside of public accounting. Very few stay on the career path in public accountancy for more than few years but less than what it takes to be nominated for part­ner. Though these employees are not being paid the value of their services, they are not, in general, being exploited because they have alternative employment opportunities. Up to a point, the longer they stay on, the bet­ter those alternatives get. This way of dealing with the problem of unse­cured human capital must be carefully handled, however, if the firm is not to create a bad reputation and, as a result, suffer an adverse selection prob­lem when they go to recruit new staff accountants.³⁵

Policies and procedures such as those just described protect the quasi­rent values that both firms and workers bring to the exchange that consti­tutes the employment relation. The real question now becomes: Are poli­cies and procedures of this sort the exception or the rule? In other words, are the quasi-rents of firms and their employees in general protected from each other in free enterprise systems? The answer in large measure depends on whether or not the evolutionary hypothesis that underlies transactions cost analysis is true, at least as it applies to the employment relation. This hypothesis starts with the observation that in a free enter­prise system, people are free to craft whatever contractual arrangements they find mutually beneficial. Competitive pressures in such a system will tend to select out those policies and procedures that minimize transactions costs. One the main types of transactions costs to be minimized are the potential expropriation hazards that each side faces from the other side. If employment policies and procedures evolve that minimize these costs, then workers (and the firms that hire them) will tend to get the value of what they bring to the employment transaction. In such circumstances, the quasi­rent values of workers’ specialized assets will tend to be protected, as will the quasi-rent values of the assets that support the wage offers that employ­ers make.

While it is likely that there is a tendency for this sort of thing to happen, it is not obvious how strong it is or whether there are any countertenden­cies or other factors at work in the other direction. In other words, while there may be a tendency for workers to get the value of their labor contri­bution and thus not be exploited, no evidence has been offered that this tendency always or usually works itself out rather than being held up or counteracted by some countertendency or some other endogenous factor operative in any free enterprise system. It is at least possible that this hap­pens and that those affected have nowhere else to go. In short, it still seems at least possible that most or nearly all workers are systematically exploited in a free enterprise system.

Though this remains a possibility, it can be argued that it is nevertheless quite unlikely. The argument proceeds by reductio ad absurdum. Suppose that most or nearly all workers (the “most or nearly all” qualifier will hence­forth be suppressed) are exploited by their employers. According to the definition of exploitation developed in chapter 3, this means that they are not getting the value of their contributions and have no real alternative but to accept the wage they have been offered. If they are not getting the value of their respective contributions, it is either because they are in a stagnant market (i.e., a market that is neither competitively efficient, nor in the process of becoming competitively efficient, nor highly volatile) or because they are in a market in transition but the exchange they are making is not on the leading edge of that market. Suppose the latter. Given that the mar­ket is in transition, it is on the way to becoming competitively efficient. Therefore the workers will be getting the value of their contributions and thus will not be exploited. Moreover, if they are not on the leading edge of a market in transition, an exchange on the leading edge will often be avail­able or accessible at a reasonable search and redeployment cost, in which case they do have a real alternative to their current situation and thus are not being exploited.

Realistically, the only way that the workers could be systematically ex­ploited is if labor markets are and always have been stagnant, that is, neither competitively efficient nor in transition to becoming competitively effi­cient.³⁶ Suppose, then, that this is the case. Further suppose that the workers have no real alternative but to accept the terms they do accept. This means that all of the alternatives are about the same or much worse. Suppose, first, that they are about the same. This means that, in general, workers are get­ting less than the value of their contributions and that everywhere they turn, the prospects are about the same. How could this happen? One way is through collusion among employers.³⁷ They could have all got together and agreed to hold down wages. The problem with that scenario, aside from the lack of any empirical evidence to support it, is that it ignores the fact that a cartel-like arrangement of this sort creates obvious incentives to violate it. An entrepreneur who offered workers slightly higher wages and charged slightly lower prices could increase sales, market share, and net profits. A self-reinforcing process would get underway, which means that the market would be in transition and on its way to ensuring that workers got the full value of their contributions. This is not to say that collusion never happens, but it is unreasonable to posit it as a permanent and uni­versal feature of free enterprise systems. What makes it unreasonable is the internal dynamics of the quest for entrepreneurial profit in this type of sys­tem. Admittedly, collusion is not the only possibility. Nevertheless, whatever explanation might be offered, the fact remains that if labor markets are consistently stagnant, there are profit opportunities there to be seized— opportunities that entrepreneurs are consistently unwilling or unable to seize. On the face of it, this seems quite implausible.

Still operating under the supposition that labor markets are stagnant, consider now the possibility that the workers’ alternatives to their present situation are much worse. Perhaps their only alternative is to accept public relief, which is much worse than continuing to work at their jobs for less than the value of their labor. But if this were true of most workers most of the time, one would observe very little labor mobility throughout the histo­ry of free enterprise systems. (Indeed, the same phenomenon would be observed if their alternatives were all about the same.) Having no real alter­native but to accept the positions they are currently in, most workers would not change jobs, except insofar as such changes are occasioned by the gyra­tions of the business cycle. In point of fact, however, this has not been the case. Historically, free enterprise systems have been remarkably dynamic, especially when government involvement has been minimal. People change jobs often even when macroeconomic conditions are relatively stable, and such systems have set off some of the largest and most successful peaceful migrations in human history. Labor mobility is a fact of life for large seg­ments of the population who do not have the security of tenure or the civil service.³⁸ The fact of labor mobility refutes the idea that all, or nearly all, workers in a free enterprise system are usually in a situation where all the alternatives are considerably worse than continuing to do what they have been doing.

Whatever the alternatives, the possibility that workers are consistently being paid less than the value of their labor presupposes that there are per­sistent profit opportunities that innovative entrepreneurs are unwilling or unable to seize. This sits uneasily with a widely shared belief, even among socialists, that a distinctive feature of free enterprise systems is a prodigious capacity for innovation on the part of entrepreneurs. It is hard to believe that at least some of the creative, clever, and highly self-interested people who control productive resources in these systems are not innovative enough to discover a way to exploit workers a little less than they are being exploited, thereby setting in motion a self-reinforcing process that ends up with workers getting about the value of their contributions. Socialists have yet to come up with a plausible explanation of what would repeatedly stop this process short of equilibrium.

If the picture of the employment relation sketched in this section is accurate, it does not follow that no worker is ever exploited. What is at issue is the claim that all or nearly all workers are consistently or persistent­ly exploited in a free enterprise system. The denial of this claim is consis­tent with the proposition that some workers are sometimes exploited in such systems. Indeed, the latter would be difficult to deny. Exploitation of workers can and does take place. Labor markets are sometimes stagnant, which results in workers’ not getting the value of their labor contribution and some of these workers effectively have nowhere else to go. In addition, exogenous changes, or shocks, can uncover previously protected quasi­rents, which employers are then in a position to appropriate. However, as employees in a firm, industry, or region have their quasi-rents appropriat­ed by their employers, they and others eventually “wise up” and develop their talents and abilities in other directions or else migrate, as the case may be. If a clever entrepreneur can develop a governance structure to protect these quasi-rents, that firm will be able to prosper at the expense of its “greedier” rivals who cannot resist grabbing for their workers’ quasi-rents. Workers will develop specialized talents and abilities and mutually benefi­cial exchanges will go through that otherwise would not, or else exchanges will take place at a lower rate than they would if they had to include a pre­mium to compensate workers for placing firm-specific assets at risk. Though exploitation will take place in any free enterprise system and, indeed, in any market economy, it will not be a pervasive and systemic phe­nomenon afflicting most workers for most of their working lives, as socialist critics of the free enterprise have maintained down through the years.