The Cato Review of Business & Government
The Cato Review of Business & Government
Leo Troy is a professor of economics at Rutgers University
In March 1993 Secretary of Labor Robert Reich and Secretary of Commerce Ronald Brown established the Commission on the Future of Worker-Management Relations. This commission, more commonly known as the Dunlop Commission, after John T. Dunlop, its chairman, was to make policy recommendations that would bring the American workplace into the 21st century.
Instead, the recommendations issued by the Dunlop Commission in its report of January 1995 look more to the past than the future. The commission recommended reforms of the American workplace that, with few exceptions, look backward to the 1930s and 1940s, an era when the hand of government became conspicuously visible in labor markets, and not toward the 21st century, an era that will depend more on the invisible hand of worldwide competition. Ironically, the commission’s look backward parallels the direction in which unions are moving. If present trends continue, by the year 2000 the private sector union movement will have slipped back a century. As the new millennium begins, unions’ share of private employment will likely fall to 7 percent, about the same proportion it was at the beginning of the 20th century.
The Dunlop Commission’s recommendations focus on rejuvenating the trade union, an institution that requires shelter from competitive forces if it is to survive and prosper. The labor policies of the 1930s and 1940s provided such shelter for a time. These policies of the past became the model for the commission’s recommendations.
But markets are dynamic, not static, and for all practical purposes, by the mid-1950s they had "repealed" the basic labor law of the land, the National Labor Relations Act of 1935 (NLRA). Since the mid-1950s private unions’ share of the labor market has steadily declined. For this reason, the commission’s revisitation of the New Deal era is out of step with the times. After November’s historic elections, the commission’s recommendations are unlikely to be implemented. But even if they were, these antiquated policies would eventually succumb to the competitive forces that have already weakened unionism.
A balanced assessment of the Dunlop Commission’s work should begin by recognizing the distinction between organized labor-management relations in the private and those in the public workplace. In contrast to organized relations in the private sector, those in the public sector are virtually immune to competition. Consequently, organized labor in the private sector has declined across all industrial sectors and classes of workers since the mid-1950s. Meanwhile, public sector unions have gained record numbers of members and an increasingly large market share. Unions’ share of the private labor market peaked in 1953 at 36 percent, but as of 1994 it stands at only about 11 percent—less than in 1929. The number of union members in the private sector peaked in 1970 at 17 million. Even after a minute gain in 1994, private sector unions have dropped to about 7.25 million members.
Internationally, private sector unions are in retreat across all major industrial economies: Britain, Canada, France, Germany, Italy, and Japan. The Canadian example is the most relevant. For the better part of a decade, many members of the Dunlop Commission had advocated that the U.S. government emulate Canadian policies in labor-management relations because they mistakenly believed that Canada had avoided the "American disease" of decline. For example, in a landmark article in early 1980 decrying the NLRA, Paul Weiler, who served as counsel to the Dunlop Commission, contrasted the apparent growth of Canadian unionism with the decline of U.S. unionism. By the time the Dunlop Commission had been formed, however, the evidence of Canadian unions’ decline was overwhelming. Nonetheless, several Dunlop Commission recommendations, such as expedited representation elections, government intervention in first contract settlements, and union access to shopping malls during the run-up to an election, were borrowed from the Canadian model.
Formation of the Commission
When the Carter administration attempted labor reform in 1977, it followed the traditional route: it proposed a bill. Instead of using this direct method, the Clinton administration established the Dunlop Commission. This select panel was to
generate recommendations that would be wrapped in the aura of disinterested expertise. However, the commission was anything but a balanced, disinterested group of experts, despite the claims of Labor Secretary Robert Reich. The known positions of the commission’s counsel and most commission members were so blatantly prounion that they mocked the secretary’s claim of balance. For example, well before the reports were issued, commission member Paula Voos claimed that current labor law required rigid job distinctions that hindered the flexible use of the workforce. To remedy that alleged inflexibility, she proposed that supervisors and mid-level management be given organized representation. This turned out to be one of the commission’s recommendations.
Further undermining the Dunlop Commission’s claim to impartiality is its inappropriate focus on the organized labor sector at the expense of the majority of workers. The emphasis of the commission’s two reports, its fact-finding report of May 1994 and the report on its recommendations, released January 1995, is on the interests of the organized sector, which accounts for little more than a tenth of the private labor force. With some exceptions, the term worker-management is a euphemism for union-management in the commission’s two reports.
The Dunlop Commission believes that unionism contributes to efficiency, and this basic assumption is reflected in its legislative recommendations. As the commission sees it, current labor law undermines American competitiveness by insufficiently protecting U.S. workers’ right to organize.
Sacred Cows Untouched
The Dunlop Commission’s prounion bias becomes apparent when one considers what is missing from the commission’s report. The commission seems to have gone out of its way to leave organized labor’s sacred cows untouched. Consideration of the ban on the permanent replacement of strikers was omitted from the commission’s purview because the Clinton administration decided to move immediately on that issue, a matter of the highest urgency to organized labor. Early in 1993 a bill was introduced and passed in the House, but it died in the Senate when a Senate majority could not muster the votes to terminate debate. After the bill’s demise, the commission did not take up the issue, although it could and should have done so.
A review of the Davis-Bacon Act, which governs how pay scales are determined in government-funded construction, is also missing from the Dunlop Commission’s report. The Davis-Bacon Act’s procedures call for the Labor Department to set the prevailing wage scales in the relevant labor market, and, unfailingly, the Labor Department finds those rates to be the union scale. This measure, which dates from 1931, has been extremely costly to taxpayers, and for the commission to pass it by in its analysis of government regulation and its effects on efficiency demonstrates just how sacred a cow this law remains.
The Commission’s Recommendations
"Quickie" Elections. The Dunlop Commission set forth a number of proposals that, if enacted, would tend to foster unionization. Among them are expedited, or "quickie," representation elections. The commission contends that petitioning unions are disadvantaged by the delay between filing a petition and holding contested elections. Specifically, its report argues that the time now required for the National Labor Relations Board (NLRB) to determine the bargaining unit and the employees who will be eligible to vote delays elections and thereby weakens unions’ prospects of winning.
To overcome this handicap, the commission recommends that elections be held as rapidly as possible—typically within two weeks. As the NLRB’s general counsel testified before the commission, under current law even the speediest elections take about seven weeks. To attain the goal of two-week proceedings, the commission recommends that the NLRB eliminate pre-election hearings on contested matters, such as scope of unit and employee eligibility, until after the election, with disputed ballots sealed in the interim. The loser would be permitted to challenge the validity of the outcome.
This proposal bears examination. The effort to speed up elections may be intended to restrict employer speech. The labor boards under the original National Labor Relations Act of 1935 severely restricted what employers and their "agents" could say during the course of representation elections. The Taft-Hartley Act of 1947 overrode those board practices by specifically protecting employer speech, so long as it contained no promises of reward or threats of reprisal. Quickie elections can be used as a way around the Taft-Hartley protections. Instead of curtailing what employers and their agents can legally say, as in the past, organized labor can use the law to limit the time in which employers may respond.
The Dunlop Commission sees this measure as a means to rejuvenate private sector unionism. However, the Canadian experience with expedited elections shows that they have failed to arrest the decline of private union membership and market share in that country. Commissioners who had previously contended that Canadian unions are more successful in organizing than American unions compounded that error by claiming that Canadian unionism had not declined. In both instances they failed to distinguish between the private and public sectors. Like the commission’s official figures on unions, the official Canadian data combine public and private results on representation elections. Since unions in the public domain always win a significantly higher proportion of representation elections than unions in the private labor market, the Canadian data, which combine the two sectors, are misleading. In contrast, in the U.S., NLRB elections apply almost entirely to private labor unions. The only major group of U.S. government workers subject to the NLRA is the postal workers, and they have very few occasions to utilize its election procedures, since the U.S. Postal Service is already almost completely organized.
While the Dunlop Commission’s call for quickie elections would curtail the time in which employers could counter union rhetoric or misrepresentation, it would expand opportunities for unions to get the ear of unorganized workers. In a related recommendation, the commission would give union organizers access to shopping malls and similar public places, now off-limits because of the Supreme Court’s decision in Lechermer v. NLRB (1992). The commission calls on Congress to reverse that decision with legislation.
Determining the First Contract. Also designed to foster unionism are the Dunlop Commission’s recommendations on first contracts between a union newly chosen to represent workers in a given enterprise and the enterprise’s management. Once a group of workers has chosen a representative, negotiating an agreement has been thwarted by employers at least one-third of the time, according to the commission. The commission’s remedies are early access to either the Federal Mediation and Conciliation Service or private mediation and the establishment of a new agency, a tripartite "First Contract Advisory Board." The advisory board would be empowered to employ several procedures, culminating in compulsory arbitration, to settle disputes. The commission argues that the arbitration option is crucial to the overall representation system.
The proposed advisory board is another example of a return to the past. It is fashioned after the National War Labor Board of World War II, of which Chairman Dunlop had been a member. The National War Labor Board was tripartite and, because of the war, invested with the power to make binding awards. The exercise of that power probably did more to foster unionism than the original NLRA. A return to compulsory arbitration now is intended to achieve similar results. Nevertheless, if the Canadian experience is any indication, compulsory arbitration is unlikely to foster unionism on a significant scale. Moreover, the Canadian experience with first contract arbitration indicates that such "shotgun agreements" do not necessarily lead to durable bargaining relationships.
Pre-Hire Agreements. The Dunlop Commission also proposes pre-hire collective bargaining agreements, agreements that would materially assist unions in the organization of workers. This measure is aimed at the employees of a new plant or a plant that the employer has moved to a new location. Under its terms, employers could recognize unions interested in representing the employees at the new location. Workers’ support for the bargaining representative would be postponed for one year, after which the union would be required to demonstrate worker support by a check of membership cards or a representation election. Card checks were used as a substitute for elections under the original NLRA, but because they are so open to abuse they became infrequently used. Moreover, as political elections show, many voters who declare one way in pre-election preferences, akin to signing membership cards in the run-up to an NLRB election, do change their minds. If polls cannot consistently predict the true preferences of voters in a political election, why should they do any better in a union election?
To grant a union representative status without prior evidence of workers’ support for that union would be analogous to the electorate’s agreement to be represented by an interested person on a trial basis, before holding an election. To soften its stance a little, the commission states that a contract negotiated under its pre-hire proposal would not bar a challenge by a competing union or a decertification election. Normally the NLRB bars a challenge to the incumbent union for at least one year after its initialÊrecognition as the majority representative. The proposed pre-hire agreement apparently comes from the construction industry, where the law allows employers to enter into agreements with unions under which employers may recruit their workers from the unions’ hiring hall. This was the Taft-Hartley Act’s way of recognizing the closed shop. De facto recognition of the closed shop in construction enabled the building trades unions to maintain their strong position in the industry for a long time, increasing costs of construction in both the private and public sectors. However, market forces steadily reduced the construction unions’ market share, which as of 1994 was down to 18 percent of employment, a rate far below that even in the depths of the Great Depression.
Mandated Injunctions. To enforce its proposals for expedited elections, resolution of first contract disputes, and other employer violations, the commission recommends that the NLRB be mandated to seek court injunctions to put a stop to unfair labor practices on the part of employers. The Taft-Hartley Act introduced mandated injunctions against unions but limited them to intra-union battles involving jurisdictional disputes, secondary boycotts, and sympathy strikes. In these instances, the mandatory injunction stopped a union from coercing an employer with whom it had no dispute. Mandatory injunctions are a powerful weapon; for this reason, current labor law limits their use to seriously disruptive intra-union disputes in which the employer is an innocent bystander. Since unfair labor practices on the part of employers do not disrupt commerce, the commission is comparing apples and oranges.
Expanding the NLRA. The Dunlop Commission recommends several changes in statutory definitions that would expand the coverage of the NLRA. It calls for narrower definitions of "independent contractors" and "contingent workers" and an expanded definition of "single employers." These measures would increase the number of persons subject to the NLRA. The proposed change of the definition of "employer" would end construction contractors’ ability to operate separate companies, unionized and nonunion. This would eliminate what is known in the industry as "double breasting"—a goal long sought by the building trades unions.
Employment Law. The Dunlop Commission also makes a number of recommendations in employment law, which, like its labor relations recommendations, would enlarge the government’s role in the workplace. For example, it recommended that every workplace be required to have a health and safety program and stressed that there should be substantial worker participation in the design and administration of these programs. Under the commission’s proposal, in nonunion establishments, employees would have the legal right to seek outside opinions and advice. Moreover, the Occupational Safety and Health Administration (OSHA) would have oversight of the structure and performance of these programs. No one is against health and safety in the workplace, but surely it is in the interests of employees and employers to address these problems on their own initiative. Employees in most workplaces already play a direct role in setting safety standards, according to a survey sanctioned by the commission. Besides, OSHA already monitors health and safety in the workplace. The commission’s recommendations would simply impose more red tape on the private sector.
The commission also recommended structures for dispute resolution, including arbitration in nonunion establishments. Employers in nonunion workplaces have experimented with a variety of means to resolve disputes fairly, but most would view a system "encouraged" by government as an opening to unionization.
New Oversight Bodies. To monitor and evaluate future developments in the workplace, the Dunlop Commission advocates the establishment of a National Forum on the Workplace. This forum would consist of representatives from business, organized labor, women’s and civil rights groups, and government. Its agenda would cover such topics as the growing disparity of income in the workplace, the status of low-wage workers, the interests of working women, government regulation of the workplace, and the impact of the global economy. Quite a broad agenda, and quite a lot of hungry interest groups to feed.
Workplace issues would be handled separately by a so-called National Labor-Management Committee consisting of the labor and management members of the National Forum on the Workplace. It would focus on the same issues addressed by the commission in its two reports. The commission also believes that forums and labor-management committees should be set up in various industries and localities. Although these bodies would have no enforcement powers, by their very nature they would promote meddling in worker-management relations. One wonders how the commission can reconcile the establishment of these proposed bodies with its finding that regulatory burdens on employers are one of the forces preventing the shift to a competitive, 21st-century workplace.
Employee Participation Programs. Legalizing employee participation programs is the Dunlop Commission’s principal forward-looking recommendation. These groups—short- and long-term work teams, quality circles, and the like—are designed to stimulate efficiency in production. Because of employers’ role in establishing these structures in nonunion settings, the NLRB has ruled that these programs violate the law. Currently, the lead case, the Electromation case, decided by the NLRB in 1992, is before the Seventh Circuit Court but will very likely end up on the docket of the Supreme Court. Meanwhile, Sen. Nancy Kassebaum (R-Kans.), the chair of the Senate Committee on Labor and Human Resources, has already expressed support for congressional action to legalize these employee groups.
But the commission’s endorsement of these programs is conditional; it wants the NLRA extended to cover workers who have supervisory and managerial authority over these work teams. Current law excludes supervisory employees from the jurisdiction of the NLRA. According to the commission, the traditional hierarchy of authority in worker-supervisor relationships has become diffused, and labor law has not caught up with these changes. The commission therefore called for new legal definitions to make some supervisory and managerial personnel eligible for coverage under the law’s protection of workers’ right to organize.
The impact of the Committee’s proposed changes would be severe; supervisory personnel are in the forefront of managerial communication with employees, so to agree to their coverage under the law would seriously dilute management’s authority. Even in what is called the new "horizontal," or "flat," structure, in contrast to the traditional "vertical," or hierarchical, one, supervisory and managerial personnel must still supply workplace leadership. For nonunion employers to agree to coverage of supervisory and lower managerial staff in exchange for legalization of employee participation programs would be a Faustian pact.
The potential impact of this measure can be gauged by examining our past experience with subjecting supervisory personnel to the coverage of labor law. During World War II the NLRB decided that supervisors were employees within the meaning of the NLRA and therefore legally protected in organizing unions. As a result, unions of supervisors mushroomed during the 1940s. But just as rapidly as supervisor unions grew, they disappeared after the Taft-Hartley Act of 1947 overruled the NLRB’s decision and excluded supervisors from coverage. The commission’s recommendation would take us back to the pre-Taft-Hartley status quo.
The Dunlop Commission’s endorsement of employee participation programs in nonunion workplaces brought a dissent from one commissioner, Douglas A. Fraser, a former president of the United Auto Workers. Organized labor has regularly regarded these structures in nonunion settings as barriers to future union organization. In an effort to allay organized labor’s concerns, the commission emphasized that its support of employee participation was not to be construed to diminish its support for American workers’ right to organize, and that it did not intend to encourage employer-dominated unions. While organized labor worries that employee participation programs may be a bar to organizing, the academic members of the commission see them from the opposite perspective: as bridges, not bars to future unionism—in short, as a Trojan horse.
Widening Earnings Differentials
The Dunlop Commission argues that its recommendations are given a special urgency by the widening of earnings differentials since the 1970s. It attributes the widening differentials to the decline of collective bargaining in the U.S., pointing out that European countries, which are more heavily unionized, have less of an "earnings gap." In its recommendations, the commission cited a comment by the Council of Economic Advisers in its 1994 Annual Report which attributed some of the widening wage gap to the decline of collective bargaining in the U.S.
However, the Council of Economic Advisers identified market forces—increasing demand relative to supply of educated and skilled workers during the period 1974-1992—as the principal factor responsible for widening earnings differentials. As for the comparison with Europe, the commission members should take note of a 1994 Organization for Economic Cooperation and Development study indicating another significant reason for Europe’s smaller earnings gap: extravagant social spending. European workers’ benefits programs have reduced the number of employed low-wage workers, which creates a narrower gap between employed higher-paid and lower-paid workers—a statistical artifact. Higher social spending engenders a higher reservation wage for low-paid workers, taking them out of the market, while contributing to a European unemployment rate that is higher than the U.S. rate.
Reasons for Organized Labor’s Decline
What of the Dunlop Commission’s contention that employer opposition to unions is a major reason for organized labor’s decline? The evidence suggests that this is not the case. A 1994 survey authorized by the commission polled nonunion workers who had been at a company when unions tried to organize. Twenty-five percent reported that their employer made no attempt to stop the effort to unionize, while 43 percent reported that their employer opposed the union with information only. Only 23 percent of the workers surveyed said that their employer threatened or harassed some union supporters. These figures suggest that the virulence of employer opposition is overstated.
Economic Change. Employer opposition is not the reason unions have lost over 7 million members since 1970; economic change is. During the past few decades, we have seen intense domestic and international competition and structural shifts in the labor market. The American economy has undergone changes that have downsized employment in unionized industries, occupations, and locations, while stimulating the growth of nonunion jobs. In its fact-finding report, the commission documented these changes in the labor market and concluded that they affect nearly all Americans and American firms and pose a major challenge to worker-management relations. Yet the commission made no connection between these facts and the decline of unionism. Instead, it focused its attention and emphasis on employer actions that deny workers their rights under the NLRA.
Worker Opposition. Another significant contribution to the ebb in unionism has been the opposition of nonunion workers. In several polls of nonunion workers’ attitudes towards voting for a union, approximately two-thirds rejected unions. The most noteworthy poll was one conducted for the AFL-CIO in 1984 by Lou Harris Associates. Not only did the Harris study find about two-thirds of workers rejecting union representation, it also found that fear of employer retribution ranked near the bottom of the list of workers’ reasons for rejecting unions. A similar poll of Canadian nonunion workers in 1990 showed them rejecting union representation by an almost identical margin. This is particularly stunning because one of the dogmas of those championing the Canadian model of labor-management relations was the assumption that the Canadian workforce had a different attitude towards unions.
The September-October 1994 survey authorized by the commission showed a smaller proportion of nonunion workers, 55 percent, rejecting union representation in an election. Virtually the same percentage as reported in previous polls, 32 percent, would vote for a union, and 13 percent did not know or refused to answer. But it is very likely that the poll overstates workers’ desire for unions because the population of nonunion workers was underrepresented. In the survey, 86 percent of employees were nonunion, but the proportion of nonunion workers in the total population is closer to 91 percent. The industrial distribution was also skewed, with manufacturing, for example, being overrepresented.
Compared to that poll and previous ones, unions do better in official NLRB representation elections than would be indicated by the attitudinal polls. Over the past two decades, unions won just over one-half of elections and about the same proportion of votes in previously unorganized units. According to the attitudinal polls, unions should have lost between 55 percent and 66 percent of representation elections in previously unorganized units. Why have they done better? First, unions usually select only those election opportunities that they are likely to win. Second, employer opposition is not as effective as union supporters have claimed. If employer opposition were as formidable as it is portrayed in the commission’s reports, unions should have lost more elections than they actually did. Clearly, there is a limited demand for union representation. Employee opposition may be a more significant factor than employer opposition for the unions’ record in organizing.
Curiously, the commission bases its argument that there is an unmet worker demand for organized representation on the minority proportion of respondents in surveys who would vote for a union. In its fact-finding report, the commission translated this into 15 million workers, a number about 50 percent larger than the actual number of union members currently in the private sector. Yet in the commission’s own survey, 55 percent of all workers reported that they preferred to deal with management on their own, rather than through a collective body.
In its fact-finding report the Dunlop Commission referred to several measures of unmet demand for some form of collective representation other than a union. At the time, prior to November 8, 1994, the reason was the commission’s apparent interest in a version of German works councils for the U.S. To support this idea, the commission stimulated surveys on the unfilled demand for collective representation other than union representation. The fact-finding report provided several measures of the dimension of that demand. One indicated that 40 to 50 million workers would like to participate in decisions affecting their jobs but lacked the opportunity. The survey source for those figures also mentioned an unfilled demand of 80 million workers who want some form of collective voice. In the report on its recommendations, the commission refers to the 32 percent of nonunion workers who in the survey it endorsed said they would like to be represented. It used this figure to argue for expedited representation elections.
It would take far more than tinkering with labor law for unions to regain their peak membership. Private sector unions would have to make up a deficit of more than 7 million rank-and-file members lost from 1970 to 1994—and without losing a single member. Even further beyond the reach of organized labor are the 21 million workers who would have to be organized for unions to regain their all-time market penetration high of 36 percent of the private labor market, reached in 1953. Therefore, one may justifiably ask: are we witnessing the twilight of private sector unionism?
The Workplace of the Future
The Dunlop Commission report focused more on the past than the future. Its recommendations seek to revitalize the decaying private sector union movement. But even if its additions to the NLRA were to be enacted, they would not save the unions from decline. Private sector unionism is in the grip of market forces across all major industrial nations. No private sector union movement in any major industrial country has escaped the American disease of decline.
Given its sponsorship, its composition, and the known positions of most commissioners, the commission’s proposals are less interventionist than one might expect. A review of the commission’s fact-finding report of May 1994 suggests that the commission was poised to recommend an Americanized version of German works councils, a much more interventionist proposal than any it did make. In my judgment, the commission considered the idea because it had come to the conclusion that there was little it could do to rejuvenate private sector unionism. The change in the political climate as it was preparing its policy recommendations may have led the commission to recommend legalization of programs of employee participation in the production process. This had the merit of recommending a current practice in the workplace that looks to the future, but at the same time could, like works councils, serve as a springboard for unionism. But that is most unlikely in the workplace of the future. If, as seems likely, employee participation programs become legal, they will reinforce the continuing expansion of the nonunion workplace and the twilight of private sector unionism.
Regulation is published four times a year by the Cato Institute. Editorial and business offices are located at 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001. For subscription information, please write to Circulation Department, Cato Institute, same address, or call (202) 842-0200. Send email inquiries to subscriptions@cato.org, or subscribe online via the World Wide Web at: http://www.cato.org/pubs/regulation/reg-ordr.html
| Regulation | Home | Publications |