After years of decline in U.S. private sector unionization, we may be at a turning point in which the long-run retreat ends and public policies more actively encourage union organizing. Support for unions among Democrats in Congress and President Obama have coalesced around the Employee Free Choice Act (EFCA), which revises the National Labor Relations Act (NLRA) to ease the path to union organizing and first contracts. 

But is a policy shift intended to substantially ease union organizing a good thing? Not necessarily. While it is appropriate to encourage enhanced workplace voice and participation, traditional unions are not the only source for enhanced voice and not always the best one. Rather, alternative forms of nonunion voice, some currently prohibited under the NLRA, should be encouraged.

Whether traditional unions should receive a boost through the EFCA (or other policies) requires that one first understand why private sector unionization has gone through a more than 30-year decline. Shifts in employment from what were traditionally highly unionized industries, occupations, and locales (e.g., production workers in manufacturing in the Midwest) to less unionized sectors played a role, explaining perhaps a quarter of the decline. Strong opposition from management combined with weak penalties for unfair labor practices have also played a role, although management resistance in part reflects competitive pressures.  

The most fundamental reason for union decline in the private sector has been an increasingly competitive and dynamic global economy.

But the most fundamental reason for union decline in the private sector has been an increasingly competitive and dynamic global economy. Unionization typically places companies at a competitive disadvantage and, thus, union governance is not surviving the market test. The wages of workers represented by unions are 15 percent higher than those of nonunion workers doing similar work. The union advantage in benefits exceeds that for wages. While substantially raising worker compensation, union effects on productivity are on average small (but probably positive) and highly variable across U.S. workplaces. Thus, higher labor compensation is not fully offset by higher output per worker, leading to higher per union costs and lower profitability in union than in nonunion businesses, as well as lower rates of capital investment and slower growth. 

In a less competitive and dynamic environment, slow-moving union governance coupled with restricted management discretion might not be a bad thing, if efficiency losses were more than offset by benefits from workplace democracy and enhanced worker voice. But in a competitive and dynamic global environment, private sector unionization is self-limiting and likely to remain a strictly minority form of workplace governance. Unions that during good times place a high “tax” on firms’ profits risk shrinking their employers, possibly through bankruptcy, since relatively sluggish union governance might not allow a quick response to sharp, unanticipated downturns in business. Although an oversimplification, this is not too bad a characterization of what necessitated the bankruptcies of major airlines in the early 2000s and automotive bankruptcies more recently. Unions have flourished in the public sector, but here competitive forces and dynamism are far more limited than in the private sector, with workplace governance being determined as much through political as through market forces. 

I take as given that we should encourage enhanced worker voice and participation in the workplace and that our society accepts the right of workers to freely and democratically choose to be represented by a collective agent. The NLRA establishes procedures to do just that, and the relevant question is whether the NLRA process might better satisfy employee choice while at the same time lead to good societal outcomes. I fault the current process on (at least) two broad fronts. First, workers form their decision on whether to unionize in an environment that is often highly contentious and may provide poor-quality information to workers. Second, worker choice is largely restricted to either a traditional union or no formal collective agent. 

It is hard to believe that a secret ballot gives a less accurate representation of a worker’s true preferences than a worker’s signing or not signing a union petition.

The information problem of union campaigns is a difficult one given the access that employers have to their workers and the legal limits placed on outside union presence and organizing efforts within the workplace. Ideally, I would like to see union campaign issues discussed among workers in forums in which both union organizers and employer representatives participated. This at least would limit blatantly false and unreliable assertions because parties would have the opportunity for rebuttal. Such two-sided information forums will not arise in any significant number voluntarily, but might be encouraged if not legally mandated under the auspices and supervision of the National Labor Relations Board (NLRB). The Internet has provided unions with a low-cost vehicle for communication, although they are constrained in communicating to workers via workplace email addresses. Modest changes in the law (or in interpretation of current law) might ease this restriction.

Concerns regarding worker choice in the presence of a contentious election process and the lack of choice of nonunion collective voice agents arise with the proposed EFCA as well as with the current regime. Under current law, if at least 30 percent of workers sign a “card check” supporting a union, then the NLRB schedules a secret ballot representation election requiring majority support for union recognition. If a majority of workers sign a card check, the employer has the option to recognize the union without an election.

The EFCA has three provisions revising the organization process. The “card check” provision states that if a majority of workers sign cards, then the NLRB shall recognize the union, absent the secret ballot election. Second, once a union is recognized, failure by the employer and new union to reach agreement on a collective bargaining contract within 120 days (90 days followed by a 30-day mediation period) leads to binding arbitration on a two-year contract. Thus, the EFCA means that a majority card check is sufficient for both union recognition and a first contract. Third (and not so controversial) are higher penalties for unfair labor practices. 

Unions have flourished in the public sector, but here competitive forces and dynamism are far more limited than in the private sector, with workplace governance being determined as much through political as though market forces.

During a union campaign, workers may receive undue pressure from the employer, coworkers, and union organizers. Either signing or not signing a card is a “public” statement by a worker and can be influenced, perhaps heavily, by peer pressure and a desire not to stick out or behave differently from a norm. It is hard to believe that a secret ballot gives a less accurate representation of a worker’s true preference than signing a petition. Unions typically receive more votes in card checks than in subsequent secret ballot elections. It is unknown how much of the difference reflects the implicit pressure to sign a union card and how much indicates a changing of minds following what can be a contentious campaign with strong management opposition. One can argue for shorter campaigns and reforms that might improve the quality and balance of information received by workers. Reducing the role of secret ballot elections is not the way to more accurately measure workers’ preferences. 

Passage of the EFCA will lead to more organizing, more first contracts, and more formalized union governance in the American workplace. The magnitude of EFCA effects cannot be known in advance and may be difficult to assess after the fact, but it is unlikely to be minor. Apart from the issue of what procedures most accurately elicit worker preferences, one can question whether policy should tilt toward encouraging the growth of unions. Buttressing traditional unionism comes with considerable costs as well as benefits. I would argue that we should instead work to develop alternative avenues for workplace voice in addition to traditional unions. Ideally, alternative institutions would provide greater voice and participation while creating value-added to the workplace and not placing companies at a competitive disadvantage. If such new institutions end up as substitutes for traditional unions based on the choices of workers, so be it. Given the low rates of union coverage in the private sector, a more likely scenario is that alternative voice options for workers will energize and complement traditional unions.

Less clear are the alternatives to traditional unions, in part because American workers have limited options. How might this change? One way is to modify NLRA restrictions on employer-sponsored work organizations. Sections 8(a)(2) and 2(5) of the NLRA affect worker participation within nonunion firms. The former prohibits employer domination or support for any labor organization. The latter defines a labor organization as one in which employees participate and which has the purpose, in whole or in part, to deal with employers over grievances, disputes, wages, rates of pay, hours of employment, or conditions of work. The legitimate goals of the provisions are to prevent employer-dominated worker groups that would stop workers from choosing an independent union and to restrain employer interference with a union. But such provisions also retard development of nonunion vehicles for employer-employee cooperation and productivity-enhancing worker voice. Although unions are concerned that such worker groups might become a substitute for traditional unions, the process of electing worker representatives and the exercise of voice in nonunion companies could complement the organization of traditional unions. Other developed countries, most notably Canada, bar company-dominated unions, but do not foreclose worker groups initiated or supported by employers that might engage in discussion over compensation and working conditions.

Public policy should be designed to enhance long-term worker well-being in a competitive, dynamic world.

I would modify NLRA restrictions on employer-sponsored work organizations by changing Section 2(5)’s definition of labor organization to include only those entities that have been certified by the NLRB, or recognized by an employer, as an exclusive collective-bargaining representative (that is, traditional unions). This would permit employers to create or maintain employee groups that discuss terms and conditions of employment, so long as those groups are not unions. It gives the nonunion employers of more than 90 percent of all private sector workers the opportunity to promote worker voice and participation without the specter of a Section 8(a)(2) violation, while maintaining the aim of that provision. Employees who want representation by an independent union may pursue that goal. Such a policy change would enhance employee choice and encourage union and employer competition in responding to employee demands.

The proposed change would expand opportunities for worker input, while preserving the path to an independent union. Although such a reform should be welfare enhancing, it is unlikely to bring about large-scale change. Current law is weakly enforced and may not provide an overwhelming barrier to nonunion worker participation programs. Such programs may be more limited by management reluctance to increase worker participation than by legal barriers. Relaxation of NLRA restrictions would be a change in the right direction, however, encouraging and publicly sanctioning participation and employee-employer cooperation in nonunion companies. It is difficult to see such a move damaging unions, although our goal should be to improve the well-being of American workers and not to bolster traditional unionism. Competition and complementarity between union and nonunion vehicles for worker voice and participation are likely to pull unions in a direction aimed more at value creation and less at taxing company earnings. 

There exist less modest proposals to enhance workplace voice. One might adopt a workplace governance default other than the current one of non-union management discretion—for example, one that allows workers to opt in to something akin to German works councils. Discussion of such alternatives extends well beyond this article’s scope. What I wish to emphasize is that public policy should be designed to enhance long-term worker well-being in a competitive, dynamic world. If this goal were best achieved through promotion of traditional unions, then enhancing organizing (in ways other than the EFCA) would be appropriate. If that is not the case, we should look to alternative paths toward worker voice. Unions could evolve to help provide or complement such paths.

Barry Hirsch is W.J. Usery Chair of the American Workplace and professor of economics at the Andrew Young School of Policy Studies Georgia State University.

Image by Darren Wamboldt/The Bergman Group.