Labor Unions in a Free Market

Ernest van den Haag
Professor of Social Philosophy, New York University

Ernest van den HaagDr. Van den Haag is professor of social philosophy at New York University and the New School for Social Research and was a participant in the fourth seminar of the Center for Constructive Alternatives, "Education in America: Democratic Triumph or Egalitarian Disaster?"

Dr. van den Haag delivered this presentation at Hillsdale during the Center for Constructive Alternatives seminar on "The Future of American Labor Unions."

I shall try to deal with four topics: I. Why do labor unions exist? What do they do? II. What are the effects of their activities? III. Strikes and restrictive practices. IV. How could the law control labor unions and their activities in the public interest?


G. F. W. Hegel once said, “Everything that is, is rational.” Surely that is not true if rational means “desirable.” But it is true, if rational means explainable as the effect of causes. For, actions are undertaken, and associations continue, only as long as they foster the interests—material or psychological—of at least some persons. That goes for the Mafia, and for the police department, for Hillsdale College—and for labor unions.

Unions yield power, prestige, and income to their officers; but they would not be supported by the members if their interests were not served too. They are: unions often succeed in raising wages and other benefits. Because they can obtain these advantages, unions are supported by many employees. In the U.S. one-fourth of the labor force is unionized; in other countries the proportion is even higher.

Unions also give workers a feeling, so often lacking in our life, of participation, of solidarity with fellow employees and of community. This feeling is not generated by simply working with others. It is created when unions succeed in providing fringe benefits (e.g., health services, or pensions, or life insurance) tied to the job and common leisure activities (outings, vacations, etc.) for their members. These things could be provided by employers too (they are in Japan) regardless of unions. But in the U.S. employers have acted in this area only for executives. This failure has helped unions to give workers a proprietary stake in their jobs while fostering solidarity among members and antagonism to employers. Unions have been able to create solidarity among members by helping them to perceive one another as brothers standing together against a hostile and exploitative father figure on whom they depend: the boss. Few things strengthen solidarity more than a common enemy. To be sure, economic life actually rests on cooperation. Without employers there can be no employees. Yet there is ambivalence and occasional conflict about how the products of the common work of employees and management, of labor and capital, are to be shared, and about who is to make decisions. Unions capitalize on the feelings of members that they ought to get more power and money and promise to get both for them.

Unions do not always benefit members. Some workers feel they are better off on their own. Others resent specific union practices or policies, or the cost of membership. Wherefore quite a number of workers resist unionization; they are often pressured and sometimes legally compelled to join unions if they want to keep their jobs. Despite these practices, union members are now a smaller proportion of all employees than they were in the past. Unions retain high membership mainly where there is a strong union tradition, or where workers are compelled to join. They have gained new members chiefly among government workers: the employer—the government—has been politically unwilling or unable to resist them. Public, unlike private, employers have a nearly bottomless purse, fed by the printing press, or by taxation. Thus, New York City, as employer, hovers near bankruptcy because of too many overpaid employees. The unions prevent the necessary firings and wage and pension decreases.


When unions benefit their members by obtaining higher wages, the money cannot come out of the profits of employers, although members are encouraged to believe that it does. All the employer’s yearly profit hardly ever suffices to pay for a single hefty pay raise. (If he were deprived of all profit, the employer would have no reason to stay in business.) On the average employers make a profit of roughly .05 per dollar of sales. Labor costs usually make up more than half of the sales dollar. Hence an increase cannot come out of profits. Thus the money for raises must come:

1. out of higher productivity—more value produced per unit of input. But productivity increases only 3% a year on the average (less in the last few years). Unions attempt to raise wage rates far beyond that.

2. from higher prices for whatever is produced by the workers whose wage rate has risen. But when prices rise, less is being sold than would be sold at a lower price—else sellers would have charged more to begin with. When less is sold, less is produced and fewer workers are employed in the industries that had to raise prices. Or, sometimes, prices do not rise despite higher wage costs. The market may not permit price hikes without excessive sales losses. In that case the least efficient producers lose money and drop out. Once more less is produced and fewer workers are employed.

Either way, union members get higher wages only at the expense of workers who do not get, or do not keep, jobs because of lower sales and less production. These workers remain unemployed or have to go into lower paying non-unionized jobs. (Recently some of the losses have been shifted to taxpayers. Those who become unemployed, because of the higher wage rates union demand, receive unemployment and other benefits. As a result they may not be willing to work in low paying occupations.)

3. If real wages for all workers—their actual purchasing power—could be increased by union activity (or by government decree), Italian or English wages, or Soviet wages, would be higher than ours. They are not. Or, our government could double all wages by law and get itself reelected forever. This sort of thing often has been tried (minimum wage laws are a remnant of such attempts) but it does not work.’ Actually, if all wages are increased—by concerted union action or by governments—one of two things happens:

(a) If the government does not allow the volume of monetary circulation to increase, inflation is prevented: prices cannot rise as wages do. When wages increase and prices (or productivity) do not (or not as much), profits decline and become losses for some; many firms stop, or reduce production, and hardly any expand. There is unemployment, and the real income of the population as a whole is reduced, although the workers who keep their jobs do well. They are not likely to enjoy it though. They lose overtime. Also they may fear for their jobs as others are fired and production is reduced. And among the fired, they may have relatives and friends.

(b) If the government allows the volume of monetary circulation to increase as wages do, there is inflation. Prices rise, and real wages (purchasing power) do not go up. Workers get more and have to pay more. People on fixed incomes suffer, and so do creditors who get the money due them but can buy less for it. All kinds of undesirable things happen, and there either is hyperinflation—rapid devaluation of the currency (rapid price level increases) and a return to barter—or the government deflates after all, i.e., it finally reduces the volume of monetary circulation. The government can do so by reducing its own expenditure financed by creating money, by increasing interest rates, and by other means which shrink the volume of money in circulation. In a major deflation prices fall and so do profits, employment and production. This is what Chile had to do after President Allende decreed wage increases financed by inflation, and to some extent we are doing so now (1978-9). In a minor deflation, such as we are bringing about, prices merely stop rising, or rise at a lower rate.

That much if all wages rise. The rise does not help workers, whether they have unions or not, because the real wage level (the wages of workers as a group, as distinguished from the wages of groups of workers) does not depend on unions, or laws, but on productivity. However, as indicated, unions can benefit their members by increasing not the wage level but the wages of groups of members, or of unionized industries and firms, at the expense of causing unemployment and an oversupply of workers, or depressed wages, in non-unionized industries.

Thus, unions do not foster the interest of the working class as a whole—but members may find that unions can increase their own wages and benefits. Even if inflation occurs, union members usually do not understand the role of unions in causing it: greedy speculators and businessmen are blamed by union leaders. The members are led to feel that the union continues to serve their interests and to protect them from the effects of the inflation for which “speculators” or “business” are blamed.


With or without unions wage rates are determined by the demand for, and the supply of, different kinds of labor. But with unions the market is not free. The function of unions is to attempt to monopolize and restrict the supply of labor—at least of labor in specific places and of specific kinds—and to obtain a monopoly price for it, higher than the free market price would be.’ This is done in two ways:

(1) By restricting union membership while compelling employers to hire only union members. This is how craft unions prevent employers from availing themselves of the full labor pool. Thus, the drivers of newspaper delivery trucks in New York get high wages. They allow very few people to join the union and prevent the newspapers from hiring non-union members. They can bargain from the monopoly position they create because New York law, and the practices of the New York police department, would make it impossible for the newspapers to hire non-union drivers: they might be killed, and their trucks would be sabotaged.

(2) By threatening to strike, unions may be able to compel employers to confine their hiring to union members and to pay them higher than free market wages. Even when industrial (as distinguished from craft) unions permit employers to hire non-union members, or allow anyone to join, the strike threat can be used to raise wages whenever enough employees can be persuaded or intimidated into refraining from work in a concerted effort to support union demands.

To strike is to refrain from working while keeping one’s job and discouraging, or preventing, others from doing it.3 In economic terms, to strike is to withhold the supply of labor, to prevent its being replaced, and to attempt to compel employers to pay the price the union wishes to charge. Unless the employer can break the strike, by hiring non-strikers, or by getting strikers to return, he has to accept the union’s terms. He has to pay a wage higher than the free market wage—the wage just high enough to attract the workers he needs.

Strikes—and all union imposed monopoly prices for labor—are contrary to the public interest. If lost by the union, strikes leave the status quo ante in force—after a loss of production and income by employers and employees and, often, considerable losses to third parties. If won by the union, strikes, in addition to losses of income and production, lead to a redistribution of income from non-union to union members; this increases the inequality of incomes while not increasing the income of workers as a whole. Finally, if they lead to general wage increases, strikes generate inflation. Strikes cannot redistribute income from employers to employees.

Nonetheless, it would be a mistake to blame unions for all strikes. Strikes can happen without unions, as history shows, and against unions, as “wild cat” strikes show. Indeed, unionization may prevent strikes as well as bring them about. As much may be said about most other union activity. It is true that much union activity is harmful to the public interest. Feather-bedding is; so are work rules which require employment of unproductive workers; and jurisdictional disputes: all of these reduce productivity. But the free market, though efficient, is also uncomfortable. Groups always have tried to protect themselves against some of the most uncomfortable effects of the free market, such is loss of jobs in any industry because of technological change, or competition. (Employment elsewhere may )e available, but the transition is difficult.) Groups of employers have tried to protect themselves by tariffs, or laws, or licensing requirements. Groups of workers, too, will protect themselves spontaneously, and it is futile merely to point out that their actions are against he public interest. People tend to protect their own. At best we can minimize the harmful effects of special interest actions, including those of unions. To this I shall turn now.


In the U.S. unions and their relations with employers and with the community at large are to a great extent regulated by law. This regulation is a mixed blessing. But other countries, such as England, with much less legal regulation, are not better off. Thus the issue is how much and what kind of regulation is needed, and how it can be made effective in the public interest. In the U.S. a deliberate attempt has been made to legally protect unions against what was thought of as the overwhelming power of employers. But the power distribution has changed, partly as a result of these laws, and today employers, individuals, and the public, do need protections from unions more than unions need protection from employers. Nonetheless, present law still attempts

  1. to protect and foster union activities;
  2. to compel employers to bargain;
  3. to compel employers not to replace strikers;
  4. to give unions a monopoly in representing workers to employers, in fixing the price of labor and the conditions of employment;
  5. to compel the government to pay union wages (in practice more than union wages) to all its employees and to force government contractors to do so (the Davis-Bacon Act).
  6. Very little regulation restricts the power of unions to interfere with the processes of production, to determine how many people and who will do a particular job, to strike at will, and to deprive society of needed services.

As long as employers have a chance to resist strikes by replacing workers or by moving away, union demands can be restrained. But there is little chance of doing so. Usually employers have to yield or give up their businesses. The latter possibility, which would deprive union members of employment, may or may not restrain union demands. It seldom does, for union leaders depend on the votes of their members. They cannot easily moderate their demands on employers, for other more militant leaders may be elected in their stead. Hence, when unions have enough power to use the strike threat effectively, the wages of members tend to rise, making unemployment of non-members, or inflation, hard to avoid.

A union must be bargained with by law, if the majority of employees join. The union then will bargain for non-members as well—whether or not they like it; usually it can compel the minority to join as a condition of continued employment. Employers must bargain not only for wages, but also for working conditions, promotions, seniority, manning, hiring and firing, the use of technology, the amount of production per worker, and fringe benefits.’ Unions are allowed to picket to advertise disputes. And although pickets legally are not allowed to intimidate or to prevent workers or others who wish to work, or to enter the plant, from doing so, in practice pickets can prevent non-strikers from working. Unions also are allowed to boycott products and producers and to harass customers and third parties not involved in a labor conflict so as to get indirectly at their employers.

One peculiar result is that unions have driven the price of labor so high in some industries (e.g., maritime and housing) that government subsidies become necessary if the industry is to continue. Unions have used political means to get these subsidies. Thus, in effect, the higher wages are paid by taxpayers who pay subsidies for housing and maritime activity.

What could regulation do to protect the public interest while being fair to all parties involved? It will be convenient to deal with occupations and industries according to the degree of interest the public has in avoiding strikes.

1. Strikes should not be permitted in public service industries or by public employees—e.g., police and firemen, garbage men, teachers, hospital or nursing home personnel, and workers in major utilities, such as electricity producers or mass transportation facilities, whether publicly or privately owned.

Here strikes should be altogether prohibited because quite obviously the harm they cause to innocent bystanders—electricity consumers, hospital patients, school children—is far greater than any conceivable benefit to the parties in conflict could be. Workers volunteer to work in the public service industries. In accepting employment they should be notified that strikes are prohibited by law. Indeed there are such prohibitions in many states. But they are not enforceable because the penalties have been kept so mild as not to deter strikes. Appropriate penalties would have to include:

(a) Fines of sufficient size levied on the union: e.g., for the first day of strike (or slowdown) an amount equal to one year’s dues for each member; for the second and third an amount equal to two years’ dues, etc.

(b) Each striking union member would lose three years’ seniority for each day of striking.

(c) Each striker would lose a part of his pension benefits for each day of striking so that on the fifth day all his pension benefits would be lost (except for any amount he himself paid in).

(d) A similar reduction in other accumulated benefits.

All of the above should be mandatory: neither courts, nor employers, nor public authorities, should be allowed to reduce these penalties once the existence of an illegal strike and the participation of the union and of the individual strikers involved has been established.

It is most unlikely that under these conditions strikes would take place—they would be suicidal for the unions and for their members. This is the point. To repeat, strikes in public service industries do harm to innocent bystanders far in excess of any benefit the strikers can hope to receive.

In the absence of effective strike threats, what would prevent employers, public or private, from driving down the wages of employees? If these wages (including fringe benefits) were to fall below what employees can earn in other occupations, a shortage of labor would develop. The industry that pays low wages would be compelled to raise them to get the workers it needs. On the other hand, as long as there are more eligible people who want to be teachers, or policemen, than there are jobs for them, teachers and policemen are not underpaid. If the supply greatly exceeds the demand, they are overpaid, if we use not moral but market standards. (The medieval moral notion of a “fair wage” pretium justum—ignores market forces and makes no economic sense.)

There remain a host of questions about which employees may wish to bargain through a union. They can do so without strike threats. Grievance and obligatory arbitration procedures may be set up and even binding arbitration may be required. However, wages should not be subject to arbitration—for they will be regulated satisfactorily by the market in the absence of arbitration. And arbitrators could do no more than guess what the market wage is likely to be. (Moreover they have a tendency to be generous with other people’s money.)

2. What about strikes in privately owned industries less directly affecting the public interest—e.g., newspapers, steel mills, coal mines, textile firms…? There are gradations. Present law gives considerable power to the President to take action to prevent or delay strikes if he finds there is a national emergency. Such strikes may be a menace even when they do not paralyze essential services. A steel strike may harm automobile workers who are not involved and even make them unemployed if it lasts long enough. However, except in time of war, such a strike need not create an emergency.

I suggest the following:

(a) Strikes should be legal only if not contrary to contracts.

(b) Union members should vote on authorizing strikes only after governmental mediation and nonbinding arbitration have failed.

(c) Strikers, who, after all, voluntarily refrain from working and earning, should not be entitled to receive unemployment or welfare benefits.

(d) Nor should strikers be allowed to intimidate others or to prevent them from working.

(e) Employers should be free to keep employees hired during a strike and to fire strikers for being strikers. (Thus employers will have a chance to win as well as lose. Strikes will become more risky for unions.)

(f) While unions should be free to bargain on working conditions, wages and benefits, they should not be allowed to compel employers to hire or keep workers they feel are not needed. Featherbedding should be outlawed, and manning should not be subject to bargaining. Featherbedding increases the cost of production (which ultimately has to be paid for by the public) and keeps workers paid for something unproductive, preventing them from doing something productive. (Featherbedding has greatly contributed to ruining our railroads.)

(g) Unions should be able to bargain on pensions (and on severance pay). But, by law, pensions should never be permitted to exceed two-thirds of the pay received by employees while on the job. Else, the employer might have to pay the same whether a worker is working or not—which would lead to a reduction in social productivity. (These rules must be enforced by effective penalties. Hefty mandatory fines for violators might be enough—unions don’t like to lose their treasure.)

The proposed restrictions on union activity may appear one-sided and, therefore, unfair. Actually they restore the balance that has been tilted in favor of unions by legislation. Moreover, employer-employee relations are not like a football match in which both teams should have an equal chance under fair conditions.

Strikes harm third persons, not just employers and employees. Unions and employers fight, win or lose, at the expense of the public interest. The public loses more, in all major strikes, than either party can hope to win. And if the union wins and compels the employer to pay wages in excess of those that demand and supply in a free market would have brought about, union members benefit at the expense of other workers and of the public at large. Thus the problem is not of arranging a fair fight between unions and employers, but rather of avoiding a fight altogether, and of limiting the results of union victories in the interest of other workers and of the public at large. The very existence of unions and of strikes, however unavoidable, tilts the mechanism of the free market in an unfair manner in favor of the union members and against other workers and the public. Legal restrictions on union power, far from being unfair, merely reduce the invasion of the labor market by monopoly power.


[1] Minimum wage laws, where the government rather than unions raises the wage, result in high unemployment among the least productive workers—mainly black adolescents. Employers do not hire those to whom they have to pay more than the market value of their contribution to production. Many of those unemployed because of minimum wage laws turn to crime—an uncovered occupation. Hence the crime rate among black adolescents is particularly high. They have been excluded from legitimate occupations owing to the efforts, presumably on their behalf, of their liberal and union friends.

[2] Unions are legally exempted from anti-monopoly legislation.

A “lock-out”—an attempt by an employer to pressure workers to accept his wages and working conditions by refusing to employ them otherwise and dismissing those who are employed—apart from occurring quite rarely, is not parallel to a strike: employers do not attempt, by pickets or otherwise, to prevent their employees from finding other jobs.

To bargain is defined as yielding on some of one’s initial positions.

Jailing of strikers or strike leaders has been shown to be wholly ineffective. So are fines, unless massive size is mandated.