All Commentary
Friday, July 1, 1960

Union Power and Government Aid

Dr. Petro is Professor of Law at New York University School of Law. For an elaboration of his suggestions on putting the principles of freedom to work in labor relations, see The Labor Policy of the Free Society (New York: Ronald Press, 1957).

Union power over the last sixty years has varied in accordance with the privileges and immunities which governments in this coun­try have given trade unionists. One is tempted to go further and say that government has responded to more or less clear shifts in pub­lic opinion. But the situation is more complicated than that.

The problem of ends and means is universal in political economy. Though there may be agreement concerning ends, policy judgments all involve choices among means. A proper choice of means requires a degree of knowledge and sophis­tication far beyond that which the general public possesses. These limitations rule any discussion of the role of public opinion. One may say that opinion rules all govern­mental action, but the problem is—whose opinion?—and how?

Union Membership: 1897 – 1959

As a logical matter, union mem­bership and union power need not necessarily vary in direct propor­tion with each other. Quite the contrary is possible—union power might increase while membership declined, or vice versa. There are other problems of definition. In the abstract, union power might be de­fined as the capacity of unions to gain their ends. More concretely and realistically, it means the ability of unions to extract im­mediately from employers and ultimately from society and the consumers a greater return for the efforts of union leaders and mem­bers than a free market would ac­cord them. Power so defined will vary from union to union, so that gross membership of all unions does not necessarily correspond to the total power of all unions. Yet, as will appear more clearly later, there is a definite relationship between union membership and union power. Growth of union member­ship and increase of union power come from the same source: special privileges and immunities granted by governments.

Union membership has grown in this country from 440,000 in 1897 to 17,024,000 in 1959. The 1959 figure includes all AFL-CIO unions and all independent national and international unions with reported U.S. membership in excess of 100,­000. Canadian members are ex­cluded.

It will not do, however, to notice only the extremes. Significant in­sights are to be derived from look­ing at what happened in between:


2,116,000 members



































Not much has happened since 1947. While unions today claim, as has been noted, a total member­ship of 17,000,000, even if that claim is credited the figure would suggest that union membership as a percentage of the labor force has lost ground in the last thirteen years.

Certainly there has been a great preoccupation among the top men of the AFL-CIO over the fact that unions have at best, from their point of view, stood still, and at worst seriously lost ground since World War II ended. If newspaper stories are to be credited, there was much head-shaking amid the palms at Bal Harbour, Florida, in February, where the dedicated men of the AFL-CIO met to ponder their problems. A story by Mr. A. H. Raskin in The New York Times (Feb. 10, 1960, p. 16) quotes John W. Livingston, the federation’s organizing director, as attributing the sad situation to “ferocious attacks” by employer groups and “labor’s” inability to overcome internal feuds. The unionized sector is supposed to represent only 39 per cent of the organizable potential today—as against 40 per cent in 1955. Plans to combat the decline include stepped-up organizing and a vigor­ous approach to the resolution of interunion rivalries.

Governmental Action or Inaction: Relationship to Union Growth

If the conclusions I have derived from study of our labor law his­tory are accurate, these measures are not likely to prove effective without positive government as­sistance and connivance. For noth­ing emerges more clearly and con­vincingly from that history, in my opinion, than the conclusion that unions unaided by government cannot induce substantial numbers of the working force to become and to remain members. They need compulsion and coercion in order to get and keep great numbers of members, and these they can exer­cise only when government does not do its basic duty to society.

I could be wrong on this, al­though I think I am not. But even if unions can induce great num­bers to become members, without special privileges of compulsion and violence they can get no better wages and working conditions than the free market would provide. If that is true, I cannot see why they should be successful in keeping great numbers, if we as a nation deny unions the privilege of com­pulsion.

The essential basis of the gen­eral conclusion can be stated briefly and simply: during the period when the laws of the land were ap­plied with some rigor to trade-union action, unions made little or no net progress in enlisting and keeping members; they about kept pace with the growth of the labor force. But when all branches of government came to the assistance of the unions, their membership growth was dramatic. Then, when law and its administration became a little more even-handed, unions were once again unable to achieve any material growth.

The key period in union growth covered the years 1935-1945, and more particularly 1936-1940 and 1942-1943. But before exploring the relevant events of those periods it will be well to broaden the perspective. We shall not be able properly to evaluate the gov­ernmental conduct of those key periods unless we have some basis of comparison.

Government and Unions: 1800-1917

From the beginning of our his­tory as a nation till World War I, trade unionism was viewed by law and government as no different from other forms of voluntary as­sociation. Today’s orthodoxy, al­though it is losing some of its dog­matic confidence, continues to in­sist that trade unions were dealt with unfairly by the common law, and by state and federal govern­ments. But this position rests upon mere assertion. The fact is that government, even with the best will in the world, has always had trouble applying the laws of the land to trade union action. At com­mon law, unions were not con­sidered suable entities and could therefore not be reached by or­dinary suits at law. Violence has always been a feature of trade union organization and collective bargaining, and local police have always had an extremely difficult time keeping the peace in labor disputes.

Perhaps unions should have been outlawed as criminal conspiracies, but, current orthodoxy to the con­trary notwithstanding, neither the common law nor legislation ac­tually did so. Hence unions as such were allowed to survive in spite of their having committed the most viciously antisocial kinds of acts. Indeed, Eugene Debs was virtually canonized even in his own day, al­though he was guilty of the worst kind of defiance of law and order.

Still, it was pretty well under­stood, at least among the responsi­ble members of society and of gov­ernment, that trade unions had no special privilege to violate either the basic laws of civilization or the rules and principles of the com­mon law. There was a good deal of uncertainty among common-law courts as to just where the line ought to be drawn as regards such forms of monopolistic coercion as the closed shop, secondary boy­cotts, and stranger picketing. But that uncertainty existed, I be­lieve, because of the essential dif­ficulty of the legal problem—not because the courts were intent upon providing special privileges for union coercion. The important thing was that there was no un­certainty at all as regards the im­permissibility of outright violence. Courts were firm on that issue, and even the police and the politicians were of the view that a man was not entitled to use violence merely because he was a trade union officer or agent.

Almost equally important, the common-law rights of employers were still intact. They could refuse to hire a man because he was a union member; hire him only on condition that he refrain from union membership; and fire any employee for joining a union. Em­ployers could also refuse to bar­gain with unions, just as they could refuse to bargain with any other person or agency.

Beginning in the 1890′s and con­tinuing into the 1930′s, trade-union expansionism also had to contend with the antitrust laws. These laws did not explicitly or directly limit monopoly-type coer­cive organizing techniques, but in putting limits upon certain kinds of secondary boycotts they did have an indirect effect of that kind, for secondary boycotts have always been used mainly as or­ganizing devices. While the anti­trust laws did not apply to union violence, they did constitute a limit upon industry-wide strikes where it could be shown that the inten­tion was to affect market prices.

To sum up on the conditions ex­isting till roughly World War I, one may say that unions were in the same position, legally, as all other self-interest groups. We might call this the period of free competition. Unions had no special privileges or immunities. They could use some forms of monopo­listic coercion in order to compel membership and bargaining. But the government did not do their organizing for them, and it did not force any employer to bargain with them. Perhaps even more im­portant, employers still had intact the freedom of contract which was so essential a feature of the com­mon law scheme of things.

Today we consider it vital that as purchasers we remain free to break off relations with any seller who does not suit us. This, we recognize is what keeps business­men serving the public, rather than exploiting it. In those days it was considered equally vital that the purchasers of labor have aright to break off relationships with unions when they proved un­reliable or exploitative suppliers of labor. The famous Hitchman Coal case, which held that em­ployers had a constitutional right to condition employment on a promise by employees not to join unions, was perhaps the clearest expression of the then prevailing policy.

In this period of what I have called “free competition,” trade unions survived and even grew, but their growth was anything but spectacular. By 1917 they had gained a membership of less than three million (2,976,000).

Government and Unions During World War I

From 1917 to 1920 unions grew to a little over five million mem­bers. In 1914, however, a signifi­cant event had occurred. Devoting serious effort to political action, unionists achieved a substantial success when Congress passed the Clayton Act, in 1914. The Clayton Act did not, in fact, free unions from the strictures of either the Sherman Act or the equity powers of the federal courts. As regards both, the legal situation changed not at all, for the Supreme Court in the Duplex and the Tri-City cases held that the Clayton Act was merely declaratory of existing law. And yet merely securing the Clayton Act was a substantial suc­cess. Moreover, it originated a pat­tern of thought and action, espe­cially in Congress, which continues to this day: a predisposition to deal very gently with union peti­tions and complaints.

Another familiar pattern was set in the early days of World War I: Unions utilized the emer­gency in order to exact concessions from the nation.

Professors Harry A. Millis and Royal E. Montgomery have long been identified as friends of the “labor movement.” We may as­sume that their account of the policies adopted during World War I was not distorted by a bias against unions. In their carefully researched book, Organized Labor (McGraw-Hill, 1945), they re­ported quite candidly that the AFL leaders conditioned their support of the war effort on the govern­ment’s recognition of “the organ­ized labor movement as the agency through which it must cooperate with the wage-earners.” (p. 136) The government agreed. When it began organizing the various bu­reaucracies which were to run the country during the war, the gov­ernment accorded suitable recogni­tion to Mr. Samuel Gompers and his associates in the “labor move­ment.” They were key figures in the Council of National Defense, the Emergency Construction

Board, the Fuel Administration, the powerful War Industries Board and other such agencies. (p. 138)

In this period the basic features of the Railway Labor Act and the Wagner Act were conceived: com­pulsory collective bargaining and protection of workers against dis­charge for union membership. Compulsory arbitration was not used during World War I, but the government’s mediation agencies, counting trade-union leaders among their members, paid due regard to union demands. Millis and Montgomery sum up the situ­ation: “Organized labor made only one real concession—relinquish­ment of the right to strike—but it made this concession in general terms, with no penalties attached. In such a setting, the trade union growth that has already been sum­marized was almost inevitable.” (p. 139)

Incidentally, the unions did not honor their no-strike pledge dur­ing World War I any more faith­fully than they did during World War II.¹

Lest accurate perspective be lost, special emphasis should be laid upon the fact that during World War I, while government lent administrative aid and pres­tige to the unions, the basic legal structure remained unchanged: neither statutory nor common-law principles were modified. When the war ended, the basic rules of the competitive society once more be­came applicable. Probably for this reason, union membership declined after the war ended, and remained down for the next fifteen years.

Government and Unions: 1920-1935

Total union membership fell from 5,034,000 in 1920 to 3,728,­000 in 1935. Some have suggested that the depression caused this drop, but that explanation is unac­ceptable. Actually, union member­ship remained steady at about 3,500,000 from 1923 on, having fallen from 5 million in 1920 to 3,629,000 in 1923. The real explanation, I feel, lies in the fact that government was giving unions no great assistance during the years 1920-1935. The implication is that without such special assistance unions could not keep more than about three and a half million members enrolled.

While government aid to unions was limited in this period, it will not do to neglect the steps which were taken in the direction of spe­cial privilege. Two were especially significant. The first of these was the enactment of the Railway Labor Act in 1926, establishing for the first time the statutory principles of protection of union membership and of compulsory collective bargaining between em­ployers and the majority repre­sentatives of their employees. The second was the Norris-LaGuardia Act of 1932, which made it virtu­ally impossible for employers to secure injunctive relief in the fed­eral courts against monopolistic­ally coercive union boycotts, and extremely difficult to secure such relief from even outright union violence.

Both statutes contained latent features which were in the suc­ceeding years to transform condi­tions in labor relations. The ele­mental compulsory bargaining principles of the Railway Labor Act were to become the framework of the Wagner Act. The restrictive features of the Norris Act were to produce antitrust exemp­tions for unions. Perhaps even more important, in taking away the jurisdiction of the federal courts the Norris Act established the pattern which was to make labor relations an administrative law field, with the National Labor Relations Board the dominant agency. From this a great many serious and untoward consequences flowed. But these were to be real­ized only in the years after 1935. Till then, neither the Railway Labor Act nor the Norris Act pro­vided enough special privilege for unions to achieve any dramatic growth.

Government and Unions: 1935-1947

Between 1935 and 1947 unions achieved the goal of every special interest group: a full, even over-full, complement of legal rights and privileges—with no corres­ponding legal duties. As already noted, they did very well in terms of membership gains. From a membership of 3,728,000 in 1935 they acquired a membership of over 15,000,000 in 1947. The dra­matic character of this increment can be appreciated only when one sets it against a near-stable mem­bership of under four million for the preceding thirty-five years.

Every branch of government came to the assistance of the unions during this period. Legis­latures, state and federal, gave them favorable laws. Administra­tions, especially those of the na­tional government, enforced these favorable laws well beyond the hilt. And the Supreme Court of the United States not only went along with distorted interpreta­tions of already unduly favorable laws, but also established consti­tutional privileges for such coer­cive union action as picketing. The favoritism did not diminish during the emergency years of World War II. On the contrary, it tended to expand. Unions did little to hide their intentions to profit from the emergency. I think it is not a dis­tortion to say that, although union leaders talked a great deal about patriotism, they did not act the part of patriots.

Fundamentally what happened in the years 1935-1947 is that most of the effective free market checks to forced union growth were de­stroyed at the same time that the most effective legal restraints upon aggressive and compulsory union­ism were removed. To put it an­other way, peaceful and lawful resistance to unionization was pro­hibited in one way or another, while violent and unlawful action by unions went substantially un­checked.

Free market checks to expansive unionism can come from two sources: (a) employees who prefer not to join unions, and (b) em­ployers who find that dealing with unions is neither an effective nor economically feasible method of solving their personnel problems. The Wagner Act of 1935 did not completely abolish either of these checks, but it impaired them con­siderably. If a majority of em­ployees in an appropriate bargain­ing unit voted in favor of union representation, the minority was left with no choice other than to accept the union as exclusive bar­gaining representative. When one realizes that the selection of the appropriate bargaining unit was left to the almost unhampered dis­cretion of the National Labor Re­lations Board—and that the Board thought its duty was to carve out the bargaining unit which was most likely to result in the election of a union—one is likely to conclude that a great deal of gerrymandering went on. That conclusion is affirmed by ex­amination of the cases.

After a union was certified as exclusive bargaining representa­tive the employer was under a duty to bargain with that union—and with no other—on all matters re­lating to wages, hours, and other terms and conditions of employ­ment. The employer did not have the choice which a free market makes available to all other pur­chasers; he could not shop around; he had to bargain with that one agency. Thus it is proper to refer to the union’s position as a monop­olistic one. And the results to be expected from all such monopolies were forthcoming in labor rela­tions. Unions abused their posi­tion.

Violence Tolerated

Unfortunately, the monopolistic privileges of unions did not end there. Legally, an employer could refuse to make the concessions sought by the exclusive bargain­ing agent. Moreover, if the bar­gaining agent called a strike in order to reinforce its demands, the employer had a legal right to at­tempt to keep his plant operating by hiring replacements for the strikers. The law has always pro­vided, too, that a union could not use violence during strikes as a means of blocking the access of struck employers to the labor market. As a practical matter, however, owing to faulty and in­adequate enforcement of the laws, unions had a virtual privilege to commit violence. The sit-down strikes are a memorial to some of the blackest days for law enforce­ment in the history of the coun­try. The tenor of the thirties is nowhere more clearly symbolized, in my opinion, than in the history of Frank Murphy’s career. As Governor of Michigan it was his sworn duty to prevent the violence and the sit-down strikes of the UAW. He flouted that duty. He was re­warded, not punished, for that dereliction. Franklin Delano Roosevelt appointed him a justice of the highest court of law in the nation after he had been guilty of cynically abusing the law of the land. As a justice of the Supreme Court, one of Mr. Murphy’s most notable opinions was the one in Thornhill v. Alabama (1940), where he held in effect that a coercive union act, picketing, was entitled to the protection of the Constitution of the United States as a form of freedom of speech.

The favorable-to-unions legal climate was not limited to the re­moval of the checks which em­ployers and nonunion employees might pose to expansive and ag­gressive unionism. The Norris Act had been on the federal statute books since 1932, preventing fed­eral courts from granting injunc­tive relief to beleaguered em­ployers and employees. There were similar statutes in most of the states, especially the more indus­trialized states. Toward the end of the thirties and early in the forties the Supreme Court inter­preted the Norris Act as in effect canceling the application of the antitrust laws to monopolistically coercive union activities. The picketing-free speech doctrine in a rough sort of way tended to free the unions from state laws pro­hibiting picketing and boycotts. Thus, at the same time that the Wagner Act preached in terms of majority rule, the absence of all checks upon picketing and sec­ondary boycotts operated to give unions an unimpeded right to force themselves upon unwilling em­ployers and employees even where none of the employees desired rep­resentation. The same immunity made it possible for unions to have their way more often than not in disputes with employers over sub­stantive terms and conditions of employment.

With Government Aid

If the strong unions were in a position to extend their organiza­tions to the limit, nobody should be surprised. And if they had the strength to impose almost any terms and conditions they pleased, nobody should be surprised, either. They had acquired a privileged monopoly and they did not intend to let it go unused. Let me remind the reader that union membership grew from 3,728,000 in 1935 to 10,489,000 in 1940. The surprising thing is that these highly privi­leged organizations did not man­age to unionize every employee during that period.

Enrolling union members is one thing; keeping them on the mem­bership rolls is another. Ap­parently this was the great prob­lem of the union leaders during the war years. Their efforts during 1942-1945 seemed to be directed most vigorously toward inducing the National War Labor Board to help them organize employees and then, by way of the maintenance of membership device which the WLB evolved, to keep them paying dues. Millis and Montgomery re­port in Organized Labor that as time wore on the WLB granted “union security” to all unions which requested it, except where the “requesting union failed to demonstrate its responsibility in adhering to the no-strike pledge.” (p. 764) In blunter language, the government bought the loyalty of the union leaders by compelling employees to maintain their union membership.

It seems permissible to infer that the WLB proved exceedingly serviceable to the unions during the war, for there is a remarkable jump in union membership from 1942 to 1943. In the years pre­ceding as well as in those follow­ing 1942-1943, the growth is noth­ing like what occurred in those years. From 10,489,000 in 1941, unions grew to only 10,762,000 in 1942. But by 1943 the figure jumped to 13,642,000. Then the growth rate, while still significant, slipped a little from 1943 to 1944, when the total membership has been counted at 14,621,000.

There has been no really marked union growth since then. In that fact lies the basis for some really interesting reflection. Why have unions stopped growing since roughly the end of World War II?

Government and Unions: 1947 to Date

The easy answer would be that the Taft-Hartley Act, enacted in 1947, did the job. But I am dis­satisfied with such an answer. At the very least it is necessary to note that there has been other legislation, especially on the state level, which like the Taft-Hartley Act has taken a stand against com­pulsory unionism. The right-to-work laws are perhaps the most significant. But even when one adds to Taft-Hartley all the state and local laws designed to cut down the privileged coercion which unions enjoyed well into the for­ties, the fact remains that both national and state legislation has been by no means vigorously en­forced against unions. My book, entitled How the NLRB Repealed Taft-Hartley, records my opinion that the Taft-Hartley Act was by no means faithfully applied. The opinion is common, moreover, that the right-to-work laws have not been widely respected. Finally, a good many of the special privileges which unions enjoyed prior to Taft-Hartley they enjoy equally today—as, for example, exemption from the antitrust laws, and the privilege of stranger picketing which the Supreme Court’s pre­emption doctrine affords.

If unions have stopped growing even though they still enjoy spe­cial privileges from government, does it follow that the special priv­ileges did not account for their great growth in the thirties and early forties? I think not. The most important fact to bear in mind here is that government has not since 1947—and earlier in some states—been single-mindedly on the side of the unions as it was during the preceding period. State and federal law now takes the po­sition, subject to some qualifica­tion, that unions are no more justi­fied in coercing people into unions than employers are in coercing them out of unions. Union re­straint or coercion of free em­ployee choice is just as illegiti­mate as employer coercion of that choice.

A Change in Attitude

This limitation on aggressive union organizing is supplemented in national and state policies by measures which reinvigorate the most effective free-market checks to expansive unionism: those posed by employees and employers un­willing to submit to union domina­tion. Employees are declared to have the right to refuse to join unions or to participate in con­certed activities; employers have had restored their right to combat unionism with statements of anti­union opinion—so long as they do not contain threats of reprisal or force or promises of benefit.

Even though the new restraints upon union aggression have not been enforced as vigorously as they might have been in all cases, they have still had a substantial effect. One need only review the decisions of the National Labor Relations Board and of the state and federal courts over the last thirteen years in order to see that aggressive unionism has fre­quently encountered legal restric­tions. In thousands of cases during that period unions have been pre­vented by the law from imposing their will upon unwilling em­ployees and resisting employers. The Taft-Hartley Act and similar state legislation have not broken unions or reduced them signifi­cantly in size and numbers, but it seems clear that they have had a braking effect. Many types of picketing and boycotting have re­mained privileged; much violence has continued. But numerous in­stances of each kind of aggression have been hindered or completely prevented. And this fact, I conclude, has mainly accounted for the observable halt in union growth.

There have been other, closely integrated causes. As a general rule it seems to be extremely diffi­cult for large organizations to maintain great growth rates after a certain point has been reached. Accretions seem to become pro­gressively more difficult. The easiest segments of the working force were organized first; now the difficult ones are left. If one must organize the more difficult ones with fewer effective instru­ments than were available when the easier ones were brought into the fold, the results are not likely to be so good.

Public Opinion Modified

Finally—concurrently with law changes and the phenomenon just noted—public opinion at every level has changed. The common working men, the professional men, political figures, even aca­demic intellectuals have come to a more realistic opinion concern­ing unions. Many continue to feel that unionization is “good for the country,” that unions are needed in order to keep employers from abusing workers. But very few people today are of the opinion that unions can do no wrong, per­haps workingmen least of all. With “public opinion” so oriented, it is too much to expect that all the special privileges which unions enjoy will be repealed in the near future. However, courts and ad­ministrators are more likely under these conditions to apply the exist­ing law fairly and accurately to unions; and in those cases where legislation is not entirely clear, one may reasonably expect that they will not favor an interpreta­tion which adds to union privi­leges, as they did so often in the past.

If public opinion, speaking gen­erally, at once holds unionism a “good thing” but distrusts union leaders, the immediate legislative results are likely to be neither clear-cut nor healthy. One may expect the same kind of fragmen­tary, incoherent legislative ap­proach which has occurred in con­nection with the regulation of business. In fact this process has already begun. I contended in an article in National Review (March 26, 1960) that the Landrum-Griffin Law is bad legislation, bad in de­tail and unwholesome in general approach. It takes the government further along the interventionist path, and its detail is so complex as to be incomprehensible at points. The same thing can be said of much business regulation—most notably, perhaps, the Robin­son-Patman Act.

If we continue along the route marked by such legislation as Robinson-Patman and Landrum-Griffin, we shall, I believe, even­tually break down in one way or another. We shall either strangle ourselves in bureaucratic red tape, corrupt our bureaucracy so that we can get something done, or so hamper the activity of our private associations that full socialism will seem the only reasonable way out.

Return to a Free Market

The realistic alternative is to rid ourselves of special privilege and the companion welfare-state idea that government is an all-purpose device fit to solve all our problems. In order to do this it is necessary to refute all totalitarian ideas, whether of the Marxian or Keynesian varieties, and to take up again the development of free-market principles with a full un­derstanding of the theory and practice of the free society.

Although Marxism and New Dealism have enjoyed great vic­tories in the past generation, the strange fact is that the theory of the free society has, although very quietly, made great strides during the same period. Those who wish to promote free enterprise will do well to acquaint themselves with the great literature of the free so­ciety, not only of past centuries, but of this one as well. Some of the best of the current literature is to be found in the writings of Fried­rich Hayek, Henry Hazlitt, Ludwig von Mises, and Wilhelm Roepke. But there has been much more, as perusal of Henry Hazlitt’s biblio­graphy, The Free Man’s Library, will demonstrate. Further im­provement in the climate of labor relations can come only as indi­viduals better understand, explain, and practice in their daily living the economic and moral principles of the free society.




Ideas on Liberty

A Study in Numbers

E. I. DU PONT DE NEMOURS & COMPANY, INC., was owned by 219,375 stockholders as of March 31,1960.

The company has approximately 86,000 em­ployees of whom about 46,600 are stockholders.

News release of April 13, 1960

  • Sylvester Petro (1917–2007) was a professor of law and the author of several books on the history of labor policy in the United States, including The Labor Policy of a Free Society, The Kohler Strike, and The Kingsport Press Strike.
    As professsor and director of the Wake Forest University Institute of Law and Policy Analysis, he taught generations of students about the history of labor unions, while defending free association and free contract as essential to the free and prosperous commonwealth.