Freedom and American Labor Relations Law; C. Baird

THEFREEMAN IDEASON UBERTY Freedom and American Labor Relations Law: 1946-1996 by Charles W. Baird A t the close of World War II, large seg ments of the American economy were in the iron grip of forced unionism. The Norris-LaGuardia Act (1932) together with the National Labor Relations Act (NLRA, 1935)had effectively exempted labor unions from the ordinary rule of law to which all individuals and other institutions were sub ject. John L. Lewis, the notorious leader of the United Mine Workers, actually was allowed to endanger the war effort in 1943by a massive strike in the coal mines. Only the passage of the War Labor Disputes Act of that year, over President Roosevelt's veto, stopped the strike.! That Act expired after the war, and there followed an "unprece dented wave of strikes that were to almost overwhelm the nation during the winter of 1945 and the first half of 1946.,,2 Accord ing to Congressman Fred Hartley, the co sponsor of the 1947 Taft-Hartley Act, labor racketeering was so pervasive that the press almost ignored it as commonplace.3 Largely because of union-based chaos, terror, and corruption, and the widespread perception that the Democratic party was in thrall to union bosses, Republicans gained control of both houses of Congress in the off-year elections of 1946 (as they did again in 1994). The 80th Congress, which was Dr. Baird, a contributing editor ofThe Freeman, is a professor of economics and the director of The Smith Center for Private Enterprise Studies at California State University, Hayward.

seated in January 1947,immediately entered into a battle with President Truman on the issue of taming the unions. The result was the Taft-Hartley Act which became law on June 23, 1947, by a congressional override of Truman's veto. This essay first outlines some basic prin ciples of the common law of contract, prop erty, and tort and explains what Norris LaGuardia and the NLRA substituted for it. It then explains the ineffectiveness of the Taft-Hartley to ameliorate the worst ex cesses of Norris-LaGuardia and the NLRA despite its authors' hopes. The third section discusses the phenomenon of government sector unionism and why it is the only form of unionism that is prospering in 1996. In conclusion, on the horizon may be changes in American labor relations law that are consistent with the principles of a free society articulated by FEE in the past fifty years. Common Law, Norris LaGuardia, and the NLRA Common Law and Voluntary Exchange The definitive common law critique of American labor relations law is Richard Epstein's "A Common Law for Labor Re lations: A Critique of the New Deal Labor Legislation.,,4 Prior to the 1930s, the em ployment relationship was simply one of voluntary exchange contract between a will299 300 THE FREEMAN • MAY 1996 ing employer and a willing employee. The proper role of government was "to prevent the private use of force and violence and to enforce promises, except when those promises are induced by duress or misrep resentation, or made by persons of manifest incompetence such as infants and insane persons. ,,5 The common law develops over time as successive judges discover the moral and logical implications of the principles of vol untary exchange in many different kinds of disputes in many different kinds of circum stances. Voluntary exchange contracts meet four criteria: 1. Entitlement. All parties to the contract must either own that which they offer to exchange, or they must be acting as the authorized agent of the owner(s). In employ ment contracts, workers own their labor and employers own the job (in the sense that they own or lease the plant and equipment and site at which the job is done). Workers and employers are free to hire or not hire agents to represent them in the labor mar ket.

2. Consent. All parties to the contract must agree to enter into the contracting re lationship-Le., to bargain with each oth er-and to the terms at which any actual ex change takes place-i.e., the final outcome of the bargaining. No forced bargaining can result in a voluntary exchange contract. 3. Escape. All negotiating parties must be able to tum down any offers they do not like and walk away from the bargaining process without losing anything to which they are entitled. There is no requirement that bargaining continue until a satisfactory deal is made or that either side must make concessions. 4. No misrepresentation. No party to the contracting may defraud any other parties. That is, no one can tell a lie. This leaves room for honest error. Someone can make a claim that he believes to be true when made, even if it turns out later to be incor rect. Moreover, this criterion does not re quire the parties to tell all they know. It merely proscribes any person saying some thing he knows to be false.

Employers' Common Law Rights Prior to Norris-LaGuardia and the NLRA, an employer had a common law right to resist the unionization of his firm. For ex ample, as part of a job description, he was free to include a requirement that a pro spective employee refrain from union related activity. An employee who accepted a job offer that included that union-free (or so-called "yellow dog") requirement was bound by his promise. A union that subse quently tried to get the employee to join a union and also to continue the employment relationship would be guilty of the tort of inducement of breach of contract. 6 Note, the union-free agreement did not prevent workers from joining unions. They simply required that if a worker chose to join a union he had to sever the employment relationship. Prior to 1932 government en forced these union-free contracts like any other contracts. 7 In 1932,Norris-LaGuardia made them unenforceable, and three years later the NLRA made them illegal.

Another way that an employer could le gally resist unionization under the common law was to hold meetings with employees, on his time and premises, to try to convince them, without misrepresentation, that unions were a bad choice for employees. The NLRA imposed a gag order on employ ers. The NLRA forced employers to remain silent on the question of unionization. It was none of their business. It was up to the employees and the unions, without inter vention from the employers, to decide the issue of unionization. Section 8(1) made it an "unfair labor practice" for an employer "to interfere with" the employee's decision to unionize. While interference by means of force or fraud are properly enjoined, "interference" in the form of honest ex pressions of opinion to the effect that unions are bad for workers are not. Another wayan employer could legally resist unionization under the common law was to refuse to hire a worker who was known to want to unionize his firm. Workers in unionized firms are faced with conflicting demands for loyalty. If I run a union-free shop my employees will be more loyal to FREEDOM AND AMERICAN LABOR RELATIONS LAW: 1946-1996 301 the firm than if they are unionized. As an ordinary measure of self defense, the com mon law would allow me to refuse to hire anyone who would reduce the loyalty of my employees to my firm. Section 8(1) of the NLRA made it an unfair labor practice for an employer to discriminate for or against a worker based on his membership or non membership in a union.

Employers also had a common law right to choose to promote unionism. For exam pIe, an employer was free to agree with a union to hire only workers who had chosen to be its members. So long as the employer and the union were not forced to bargain over the issue, and so long as both con sented to the arrangement without coercion or misrepresentation, it was a voluntary exchange contract which the government was supposed to enforce. The NLRA forced employers to bargain with unions to set up such" closed shops. " Moreover, the unions with which the employers were forced to bargain were not made up of voluntary members. The NLRA forced all individual workers to be represented by the union that was "selected" by a majority of workers. Employers had a common law right to set up their own unions. These were called company unions, and they were very com mon during the 1920s. They were programs by which employers sought to bring employ ees into some decision-making. They were early examples of what today are called worker-management cooperation schemes.

They were also used by some employers to deflect the organizing efforts of independent unions. The NLRA made company unions illegal. Other Special Privileges for Unions The common law of property entitles an owner to "possess, use, and dispose" of that which he owns as he sees fit so long as in so doing he does not engage any other person in any involuntary exchange. Jones who seeks to enter, remain on, or use the property of Smith may do so only with Smith's permission. Absent that permis sion, Jones is guilty of trespass. Moreover, if Jones blocks access of others to Smith's property, through force or threat of force, Jones is guilty of trespass. What then of union organizers, strikes, and picketing? As we saw above, Section 8(1) of the NLRA made it an unfair labor practice for an employer' 'to interfere with" an employ ee's decision to unionize. In addition to restrictions on employer speech, this was interpreted to prevent an employer from denying union organizers access to his plant. Suppose my employees are interested in signing up for computer instruction. A representative of a computer training firm shows up at my door asking to speak to my employees. I have a common law right to prevent that representative from entering my plant at any time and talking with my employees during working hours. But, un der the NLRA, if a union organizer showed up at my door asking to speak with my employees, I could not keep him out of the plant. I could prevent him from talking with my employees while they were working, but I could not prevent him from talking with them, on my premises, during breaks.

A strike is more than just a collective withholding of labor by workers who each regard the employer's offerof compensation (or some other proposal) to be inadequate. A strike is that, but it is also, through picket lines, an attempt to prevent replacement workers, suppliers, and customers from doing business with the struck firm. Mass picketing, even if it is peaceful, is intimidat ing. It is an implicit (and most often an explicit) threat of harm to anyone who crosses the line. As such, picketing is tres pass against the common law property right of the strike target to do business with willingreplacement workers, suppliers, and customers. The U.S. Supreme Court upheld that common law right in American Steel Foundries v. TriCity Central Trades Coun cil (257 U.S. 184 [1921]).The Court limited pickets to one per entrance, and stated that all pickets had to be actual employees of the firm being picketed. Strangers, nonstrikers bussed in by the union to create formidable picket lines, were forbidden to picket in strikes. The Court allowed one picket per entrance on free speech grounds, but disal302 THE FREEMAN • MAY 1996 lowed mass picketing and any picketing by strangers on grounds of the common law of property and trespass. Norris-LaGuardia repealed the Court's TriCity decision. It permitted mass picketing, including picket ing by strangers.

Worse than that, it allowed violent pick eting to prevent replacement workers, sup pliers, and customers from crossing the line. Norris-LaGuardia prohibited federal judges from issuing injunctions against picketing, or other strike activity, even if violence was involved. (It was left to the usually outnumbered local police officials to keep the peace.) Specifically, Section 7(c) of Norris-LaGuardia said no strike activity could be enjoined by a federal court unless testimony, subject to cross examination and accompanied by rebuttal testimony, had been taken which leads the judge to think "that as to each item ofreliefgranted greater injury will be inflicted upon complainant [the employer seeking the injunction] by the denial of relief than will be inflicted upon defendants [the violent union picketers] by the granting of relief." Suppose a gang is raiding an office building. If Section 7(c) of Norris-LaGuardia applied, ajudge could not enjoin the raid unless competing testimony led him to think that the damage to the building owner if the raid continues would be greater than the damage to the gang if the raid were stopped.

There is a common-law proscription of "combinations in restraint of trade." The 1890 Sherman Antitrust Act and the 1914 Clayton Act were codifications of that com mon law proscription. Prior to Norris LaGuardia, labor unions were subject to the antitrust laws. Some secondary strikes (e.g., a strike of the employees of firm A against firmB that isa customer or a supplier of firm A) were enjoined as impermissible restraints of trade. Violent primary strikes were also sometimes enjoined on antitrust grounds. But Norris-LaGuardia gave unions immunity to the antitrust laws.8 Strikes, even violent strikes, and even secondary strikes, could not be enjoined on any grounds, whether trespass or antitrust, in federal courts. The common law includes the doctrine of respondeat superior, or vicarious responsi bility. If I am driving a delivery truck for my employer and I hit a pedestrian, I am liable for damages, and so, too, is my employer.

As a driver for him, I am his agent. He is responsible for any damages I create. Under the common law, the same rule would apply to labor unions. Under the NLRA, if I am walkinga picket line at the behest of a union, and I hit a replacement worker over the head with a hammer, I am liable for damages (and even criminal prosecution by local authori ties), but the union that placed me on that picket line is not. Norris-LaGuardia gave unions complete freedom from vicarious responsibility. No union can be prosecuted for any of the acts of its strikers, no matter how violent or even ifunion bosses order the violence. The heinous results of the special privi leges granted to unions by Norris-La Guardia are well illustrated in the case of Apex Hosiery Co. v. Leader (310 U.S. 409 [1940]). The employer was operating on an open-shop basis. Th~ union wanted to force all 2,500 employees to unionize. Eight em ployees who were union members, joined by members of the same union who were employed by other firms (Le., strangers to Apex), undertook a sitdown strike. In other words, they occupied the premises of the employer, prevented willing employees from working, and proceeded to destroy machinery on the shop floor. The company applied for an injunction against the union on Sherman Act grounds of a violent com bination in restraint of trade including tres pass on private property. In the end the U.S.

Supreme Court denied the company any relief by claiming that Norris-LaGuardia protected unions from any antitrust prose cution. In the words of the Court, "Re straints not in the [Sherman] Act when achieved by peaceful means, are not brought within its sweep merely because, without other differences, they are attended by violence." So much for the basic respon sibility of government "to prevent the pri vate use offorce and violence" and maintain the peace.

FREEDOM AND AMERICAN LABOR RELATIONS LAW: 1946-1996 303 Workers' Common Law Rights Workers also had a common law right to resist unionization. This was part of their freedom of association as guaranteed by the Bill of Rights in the U. S. Constitution. Each individual worker could decide for himself, notwithstanding what a majority of his col leagues might choose, whether to be repre sented by a union in the sale of his labor. A worker who wanted such representation joined the union and paid its dues. A worker who wanted to speak for himself in the sale of his labor neither joined a union nor paid its dues. The NLRA destroyed the freedom of association of individual workers who wanted to remain union-free although a majority of their colleagues wanted to unionize. Specifically, Section 9(a) states that, "Representatives designated or se lected for the purposes ofcollective bargain ing by the majority of the employees in a unit appropriate for such purposes, shall be the exclusive representatives of all the em ployees in such unit for the purposes of collective bargaining in respect to rates of pay, wages, hours of employment, or other conditions of employment. " Individuals are forbidden to represent themselves. They are forbidden to enter individual contracts with their employers.

Unionists excuse this tyranny of the ma jority on the grounds that it is democratic. But "democracy" is a form of government designed to give the governed some say over what the government does. Democracy is an institution originally conceived as a way of preventing government from trespassing on the protected private domain of human action. The Founders of the American re public never intended that the will of a majority should control anything except their short list of constitutionally authorized governmental activities. The sale of a pri vate citizen's labor is not a governmental matter. It is a private matter. Under com mon law, government's only legitimate role in the employment relationship is to enforce voluntary exchange contracts. Moreover, the "selection" made by a majority in Section 9(a) was not really democratic. The National Labor Relations Board (NLRB) was given the authority to deter mine what a majority wants. There was no mandatory secret ballot election. Union "organizers, " could solicit signatures of workers on an eyeball-to-eyeball basis.

Workers who refused to sign often did so at their own risk. The NLRB could order an election, but it didn't have to. Inasmuch as members of the NLRB were selected from the ranks of union sympathizers, they usu ally certified exclusive bargaining agents without benefit of a vote. The scheme was undemocratic in yet another respect. The NLRA did not require any regularly scheduled re-elections. Once certified as an exclusive bargaining agent, a union was presumed to continue to have majority support forever, even if all the workers that originally" selected" it left the firm and were replaced with new workers. The new workers never had any kind of say in the question. To make matters worse, the NLRA forced employers to bargain with unions over whether to force all the employees represented by an exclusive bargaining agent to be, or become, dues-paying union members. Section 8(3) of the NLRA makes it an unfair labor practice for an employer "to encourage or discourage membership in any labor organization: Provided, that noth ing in this Act . . . shall preclude an em ployer from making an agreement with a labor organization . . . to require as a con dition of employment membership therein"

[emphasis in original]. Note the duplicity: An employer cannot encourage or discour age membership in a union, he can only require it. Forced union membership is called "union security. " Unions were granted security against workers who want to be union-free. Thus workers who didn't want to associate with unions were coerced in two ways. They were forced to have a union "selected" by a majority of their colleagues represent them, and they were forced to pay tribute (dues) to those unions as a condition of continuing their employment relationship with their employers.

304 THE FREEMAN • MAY 1996 Unions defended their union security (forcedmembership)schemes on the grounds that since Section 9(a) forced unions "se lected" by a majority to represent all work ers, it was only fair for all· workers to be members. Otherwise a minority would be "free-riders." They would get the benefits of union representation for free. Of course, if unions represented only their voluntary members, only those who individually wanted union representation, there could be no free-riders. Unions, and the politicians in their thrall, were not embarrassed by the fact that they fought long and hard to get the privilege of exclusive representation and then claimed that exclusive representation "forced" them to represent all workers who therefore must be forced to join and pay dues. The common law, as we have seen, is based on voluntary exchange contracting between individuals. The preambles of both Norris-LaGuardia and the NLRA asserted that workers do not have "actual liberty of contract, " thus the common law was inad equate to protect the rights of workers.

Employees were, the preamble to the NLRA asserted, on the short end of an "inequality of bargaining power," with em ployers. This idea that workers on their own are helpless in the labor market so they need unions for self-defense is a hoary myth. As W. H. Hutt9 and, later, Morgan Reynolds1o have demonstrated, data falsify the myth. In the nineteenth century, long before the existence of significant unionization in the United States, real wages were on a strong upward trend, and worker-initiated job switching was frequent and became increas ingly common. Contrary to the view that says large-scale employers exploited unor ganized workers, large firms paid workers more than small firms. Contrary to the claim that employers had unfair bargaining power because unorganized workers could not af ford to turn down even poor job offers, workers with savings weren't able to bar gain for better wages than workers without.

Finally, contrary to the conventional wis dom that unions were necessary to offset employer combinations designed to keep wages low, most employer associations were formed in self-defense against unions that had already been formed to attempt to take wages out of competition. The question of bargaining power in vol untary exchange contracting is one of alter natives. The labor market is like any other market. Buyers (employers) compete with other buyers, and sellers (employees) com pete with other sellers. When a buyer· and seller come together to bargain on a mutu ally beneficial exchange, their bargaining power depends on those two types of com petition. Other things being equal, the em ployee has more bargaining power when there is strong competition among employ ers to hire his type of labor and when there is weak competition among other workers trying to sell his type of labor. Other things equal, the employer has more bargaining power when there is weak competition among employers seeking to hire similar labor and strong competition among work ers seeking to sell similar labor. Obviously, insofar as a worker is not responsible for denying an employer access to other work ers selling similar labor, there are no moral grounds for government to favor the em ployer over the worker. Why then, insofar as an employer is not responsible for deny ing a worker access to alternative employ ment opportunities, should government favor the worker over the employer? The Fourteenth Amendment to the U.S. Consti tution is supposed to guarantee the em ployer and the employee "equal protection of the laws."

The Taft-Hartley Act The Taft-Hartley Act is named after its principal sponsors: Senator Robert A. Taft and Representative Fred A. Hartley, Jr. The purpose of the legislation was to "re store some balance" between unions and employers, by curbing the power of unions. Whereas the preamble of the NLRA blamed employers for the ills the legislation was supposed to cure, the preamble of Taft Hartley assigned equal blame to unions and employers. Whereas the announced intent FREEDOM AND AMERICAN LABOR RELATIONS LAW: 1946-1996 305 of the NLRA was to promote and assist unions, the announced intent of Taft Hartley was to protect the rights of workers, unions, and employers. Whereas under the NLRA the officialjob of the NLRB was to get workers into unions, under Taft-Hartley its official job was to be a neutral umpire in labor disputes. ll Whereas the NLRA assured the right of workers to unionize, Taft-Hartley added a right of workers to refrain from organizing. Whereas the NLRA listed only employer unfair labor practices, Taft-Hartley added a list of union unfair labor practices. There is no doubt that Taft-Hartley did tip the scales toward more balance. However, it fellfar short of achiev ing that balance.

The union movement labeled Taft-Hart ley the "Slave Labor Bill" when it tried, unsuccessfully, to defeat it in Congress in 1947. The union movement referred to Taft Hartley as the "Slave Labor Act" in its successful attempt to reelect President Tru man and restore both houses of Congress to the Democrats in the elections of November 1948.12 Notwithstanding that victory, the union movement failed subsequently to re peal Taft-Hartley. It is still the law of the land. Far from a slave labor act, I think it is better labeled the "Continued Forced Unionism Act of 1947." First, Taft-Hartley didn't reach many of the points raised in the previous section at all. Union-free (or "yellow dog") contracts are still illegal. Employers are still not free to refuse to hire union sympathizers. Em ployers are still not free to form company unions and offer them to their employees as alternatives to independent unions. In fact, in 1992 the NLRB ruled that a worker management cooperation program in a non union firm was an illegal company union and was used by the employer in an illegal way to discourage unionization. 13 Now, it seems, labor-management cooperation that is not union-management cooperation is illegal. Mass picketing by strangers is still legal. Unions are still immune to the anti trust laws. They are still exempt from the common-law principle of vicarious respon sibility; and, in primary strikes, they are still immune to injunctions against any, includ ing violent, strike activity. In United States v. Enmons (410 U.S. 396 [1973]) the Su preme Court explicitly granted unions im munity to the Hobbs Anti-Extortion Act. 14 As long as their activities are related to their legitimate purposes in a primary strike, they can be as violent as they like. Individual perpetrators of violence are liable to pros ecution by local authorities, but the unions themselves are not. Taft-Hartley did affect some points raised in the first section, but inadequately. It restricted union secondary strikes, but the NLRB found ways around the restrictions, so in 1959 the Congress had to strengthen those restrictions in Title VII of the Landrum-GriffinAct. The restrictions are still inadequate.

Taft-Hartley did not directly affect the access of union organizers to employers' property, but it did so indirectly. In 1956 the Supreme Court, in NLRB v. Babcock & Wilcox (351 U.S. 105), apparently inspired by the intent of Taft-Hartley to restore balance, made a distinction between union organizers who are already employees and those who are not. The former were granted unrestricted access, the latter were granted access if they had no other means of com municating with the workers they sought to organize. This was called the principle of accommodation. In 1992, the Court, in Lechmere v. NLRB (502 U.S. 527), the first majority opinion written by Justice Clarence Thomas, greatly restricted this principle of accommodation to those (very few) cases of worker isolation such as in residential log ging camps. The authors of Taft-Hartley tried to ad dress the issue of employer free speech.

Section 8(c) of the Act states that "The expressing of any views, argument, or opin ion, or the dissemination thereof, whether in written, printed, graphic, or visual form, shall not constitute or be evidence of an unfair labor practice under and of the pro visions of this Act, if such expression con tains no threat of reprisal or force or promise of benefit." It is supposed to apply equally to unions and to employers, but, in practice, it does not. For example, employers are 306 THE FREEMAN • MAY 1996 forbidden to promise a pay raise in exchange for a worker's vote against a union in a certification (or representation) election. But unions can promise a pay raise to a worker in exchange for a vote in favor of a union in the same election. The fig leaf that covers that unequal protection of the law is that unlike the employer, the union doesn't promise the pay raise out of its own pocket.

Worse, in NLRB v. Gissel Packing Co. (395 U.S. 575 [1969])the Supreme Court said that the employer was guilty of an unfair labor practice because, during a certification elec tion, the employer claimed that it was not financially strong so that if the union were certified and then called a strike the plant may have to be closed. The Court said that an employer may express general views about unions, but any predictions of conse quences of any specific unionization must be limited to consequence beyond the em ployer's control. So much for meaningful free speech. While the NLRA imposed a duty on employers to bargain with certified unions, .the unions had no such duty; they could refuse to bargain with impunity. Taft Hartley imposed a duty to bargain on unions as well as employers. Moreover the bargain inghad to be in "good faith." In practice this means that employers (and, to a smaller extent, unions) have to grant concessions during the bargaining. For example, in NLRB v. General Electric Co. (418 F. (2d) 736 [1969]), the employer was found guilty of an unfair labor practice because its rep resentative placed a proposal before the union negotiators and, in effect, said take it or leave it. The employer refused to grant concessions. Note that this, too, is a restric tion on employer free speech. In this case the employer arrived at his offer by polling the workers to see what they would consider reasonable. The employer was chastised for dealing with the union through the employ ees rather than, as Taft-Hartley requires, dealing with the employees through the union.

Taft-Hartley did make some significant changes with regard to the "selection" of unions to be exclusive bargaining agents. First, it mandates secret ballot elections unless the employer agrees to waive an election. The employer cannot recognize a union as the exclusive bargaining agent unless the union has majority support. Most of the time employers insist on an election as the only way a union can demonstrate that majority support. Sometimes, however, an employer willrecognize a union on the basis of signedauthorization cards. If a union gets at least 30 percent of the workers it seeks to unionize to sign authorization cards, it can petition the NLRB to hold a secret ballot certification election. Taft-Hartley also added a decertification election process to the law. Once certified, a union is still presumed to have majority support indefinitely, even if all the workers that voted for the union are no longer employed by the firm; but disgruntled em ployees may attempt to decertify the union.

Employers must keep hands offthe process, but if individual employees are able to collect the signatures of at least 30 percent of the relevant workers on a petition re questing a decertification election, the NLRB will order such an election. It is as if a member of Congress could hold his office indefinitelyunless some voters in his district got at least 30 percent of the eligible voters in his district to sign a petition requesting a recall election. More importantly, Taft-Hartley did noth ing to exclusive representation itself. It is still true that if a majority of workers who vote in a certification election vote in favor of a union, that union is the exclusive bar gaining agent for all workers in the unit. It represents the workers who voted for it, the workers who voted for another union, the workers who voted to be union-free, and the workers who didn't vote. It still is a winner-take-all system subject to the same objections raised in the first section of this essay.

Taft-Hartley also made some significant changes with regard to union security. It outlawed the closed shop, but put the union shop and the agency shop in its place. In the former, a worker doesn't have to be a union member to be hired, but after a probationary FREEDOM AND AMERICAN LABOR RELATIONS LAW: 1946-1996 307 period he must join the union as a condition of continued employment. The British call the union shop a post-entry closed shop. In the agency shop version of union security, workers don't have to join unions to keep their jobs; they just have to pay union dues. It is still the case that employers are forbid den to encourage or discourage workers from affiliating with unions, they can only compel workers to do so. However, Section 14(b) of Taft-Hartley does affect union se curity in a significant way. It allows the individual state governments to pass right to work laws within their own jurisdictions.

Twenty-one states have done so. In those states, unions and employers are forbidden to include any union security clauses in their collective bargaining agreements. In the 29 states that have failed to pass right-to-work laws, the objections raised to union security in the first part of this essay still apply. Congress is currently debating a National Right to Work Bill, which would make all union security schemes in the private sector illegal, but President Clinton is sure to veto it. It is likely that there are insufficientvotes to override the veto. Government-SectorUnionism Government employees were exempted from coverage under both the NLRA and Taft-Hartley. Until 1962 government em ployee unionism was widely regarded as unthinkable, even by union-friendly politi cians such as Franklin Roosevelt and Harry Truman. However, in that year President Kennedy issued Executive Order 10988 which authorized limited, but mandatory, collective bargainingby unions representing federal employees. In Title VII of the Civil Service Reform Act of 1978,the principle of exclusive representation was imposed on federal employees, but they were spared the principle of union security. There is still a limited scope of collective bargaining. For example, federal employee unions cannot bargain over wages. Nevertheless, in 1993 President Clinton appointed the National Partnership Council whose charge it is to promote the imposition of the full burdens of private sector unionism on federal em ployees. With the present (104th)Congress, there is little likelihood that will happen.

After President Kennedy's executive or der, state after state imposed the full bur dens of private-sector unionism on state and local government employees. At present 24 states have done so. In 1994 President Clinton appointed the Task Force on Ex cellence in State and Local Government through Labor-Management Cooperation to study the possibility of enacting a national government employees labor relations law to force all states to adopt Taft-Hartley-style unionism for their state and local govern ment employees. Again, the 104thCongress is unlikely to cooperate. However, the government sector is the only place where unionism flourishes today. In the private sector only 10 percent of the labor force are unionized. In the govern ment sector, 39 percent are. 15 The peak year of private-sector density was 1953, when it was 36 percent. In that year figures on government-sector density weren't even collected. Government-sector unionism was almost nonexistent, and where it did exist it was not officiallyrecognized.

According to Leo Troy, private-sector density will be below seven percent in the year 2000,about where it was in 1900,before Norris-LaGuardia and the NLRA. Troy calls this the symmetry of history.16 The primary reasons for the decline of private sector unionism are the globalization of economic competition and technological changes. Competitive pressures have made it virtually impossible for employers to pass union-based cost increases forward to con sumers so the employers are more resistant to unionism than they were in the past. Advances in technology have greatly de creased the market shares of blue-collar industries in which private-sector unionism had its strength. Private-sector unionism is declining in all major industrial countries. The real threat of unionism to freedom is now in the government sector. The new president of the AFL-CIO is John Sweeney, the erstwhile president of the Service Em ployees International Union, which consists 308 THE FREEMAN • MAY 1996 primarily of government employees. There are three reasons that government-sector unionism is flourishing. First, government agencies are usually monopoly providers of their products and services so it is easy for government employers to pass union-based cost increases on to customers (taxpayers).

Second, government unions and govern ment agencies that employ government workers are actually on the same side of the bargaining table. It is in the interest of both groups to pick the pockets of taxpayers and to have their budgets and responsibilities grow. Troy calls government-sector union ism the "new socialism." It is primarily an attempt to redistribute more and more of the national income to government. Third, to a large extent the government-sector unions have organized the already organized. Union organizers, aided by favorable legis lation, merely converted already estab lished public-employee associations into unions. Government-sector unionism is inher ently anti-democracy. 17 Unionists have long argued that employment in the government sector is the same as employment in the private sector. If we allow unionism in the latter, we must do so in the former. But this argument doesn't work. Think how collec tive bargaining is done in the private sector.

First, the employer is forced to bargain in good faith with a certified union. The em ployer must make concessions to the union. Second, the employer is forbidden to deal with workers directly. The union's approval is necessary before the employer decides anything that comes under the scope of collective bargaining. Third, the bargaining is done behind closed doors with both sides legally bound to keep the negotiation con fidential until either an impasse or an agree ment is reached. In sum, the employer is forced to share decision-making power with the union, and the general public is excluded from the process.

FREEDOM AND AMERICAN LABOR RELATIONS LAW: 1946-1996 309 Now, wages and salaries and other terms and conditions of employment in the gov ernment sector are matters of public policy. Collective bargaining in government, on the private-sector model, means that govern ment is forced to share its making of public policy with a private organization, on an exclusive basis. The general public is for bidden to participate in the process. The Constitution established three branches of government: the executive, the legislature, and the judiciary. There is no fourth branch of government called unions of government employees. Government employee unions don't just lobby government on matters of public policy like other special interest groups, the Sierra Club, for example. They actually co-determine public policy with the government. This is the model from Mus solini's Italy. Government workers are citizens like any other citizens. Therefore, they should have the same influence on public policy as any other group of citizens. But they shouldn't be given the disproportionate influence that mandatory, closed-door bargaining with ex clusive bargaining agents in the government sector gives them. It is undemocratic for government employees to conspire with government agencies, in rooms from which taxpayers are excluded, on the size of the government budget.

In the private sector the optimal amount of unionism is whatever would emerge un der neutral legislation. Government should neither support nor inhibit private-sector unionism. However, the optimal amount of unionism in the government sector is zero. It amounts to taxation of the people, by the unions, for the unions. In Conclusion The fiftieth anniversary year of the Foun dation for Economic Education could be a watershed year in labor relations law. If the 105th Congress, which will be elected this November, is a bit more sympathetic to individual liberty than the incumbent 104th Congress (especially the Senate), and if the office of the President is filled by a similarly inclined person, all existing labor relations law could be scrapped. Abolishing exclusive representation would protect the right of each individual to select representatives "of his own choos ing." If unions bargained for their voluntary members and no one else, union security would be moot. Moreover, returning to the common law of contract would mean mak ing collective bargaining voluntary instead of compulsory, and restoring the right of workers and employers to resist unioniza tion. We could return to the common law of property, by allowing owners themselves, rather than the NLRB, to decide who has access to their property. In sum, the only "unfair labor practices" would be those that are inconsistent with the principles of vol untaryexchange. D 1. Fred A. Hartley, Jr., Our New National Labor Policy (New York: Funk & Wagnalls Co., 1948), p. 16.

2. Ibid., p. 18. 3. Ibid., p. 41. 4. Richard A. Epstein, ••A Common Law for Labor Relations: A Critique of the New Deal Labor Legislation," The Yale Law Journal, Volume 92, Number 8, July 1983, pp. 1357-1408. 5. Ibid., p. 1359. 6. Ibid., p. 1374. 7. See, for example, Hitchman Coal & Coke Co. v. Mitchell, 245 U.S. 229 (1917). Interestingly, Morgan Reynolds, Power and Privilege, (New York: Universe Books, 1984), pp. 98-100, showed that at least in this case the union-free agreements were initiated by employees to inhibit the United Mine Workers Union organizers. 8. Many free-market economists say that there should be no antitrust laws at all. D. T. Armentano says they have been misused to protect specificcompetitors rather than the process of competition. My view is that if we have antitrust laws they ought to be applied equally to employers and to unions. 9. W. H. Hutt, The Strike Threat System (New Rochelle, N.Y.: Arlington House, 1973), Chapter 16.

10. Morgan O. Reynolds, Power and Privilege: Labor Unions in America, op. cit., Chapter 3. 11. R. Alton Lee, Truman and Taft-Hartley (Lexington, Ky.: University of Kentucky Press, 1966), p. 184. 12. Ibid., Chapters 4-7. 13. Charles W. Baird, "Are Quality Circles Illegal? Global Competition Meets the New Deal," Cato Briefing Papers No. 18, The Cato Institute, Washington, D.C., February 10, 1993. The case was Electromation 309 NLRB No. 163. 14. For an exhaustive record of union violence from 1947 through 1983, see Armand J. Thieblot, Jr., and Thomas R. Haggard, Union Violence: The Record and the Response by Courts, Legislatures, and the NLRB, The Wharton School, Industrial Research Unit, University of Pennsylvania, 1983. 15. Leo Troy, "Private-Sector Unionism Weakens," The Wall Street Journal, September 1, 1995, op-ed page. 16. For an excellent discussion ofthe history and prospects of government-sector unionism and the decline of private sector unionism, see Leo Troy, The New Unionism in the New Society (Fairfax, Va.: George Mason University Press, 1994).

17. Robert S. Summers, Collective Bargaining and Public Benefit Conferral: A Jurisprudential Critique (Ithaca, N.Y.: Institute of Public Employment, Cornell University, 1976).