In Information and the Market for Union Representation, Professor Matthew Bodie provides an instructive framework for addressing information deficiencies in union elections.1 His consumer or "purchase of services" paradigm is apt and well illustrates the shortcomings of the more dominant approaches to elections. The extent to which this paradigm should drive the National Labor Relations Board's (NLRB) regulation of union elections is less obvious. The best fit with Bodie's consumer paradigm appears to be a system in which employees can easily designate a union as their representative, yet can just as easily get rid of the union. In other words, employees, like many other consumers, could purchase union services with the knowledge that they can easily change their mind later. It is not clear, however, whether the benefits associated with that model are worth its costs.
Bodie rightly decries the NLRB's failure to ensure that employees have access to the information needed to make a fully informed decision whether to unionize. In particular, his criticism of the NLRB's "laboratory conditions" and "political elections" models provides valuable insights into the nature of the union election process and raises important questions about the governance of elections. Yet his consumer paradigm is no panacea. For example, employees' collective decision whether to unionize differs from a typical consumer decision, which generally does not face concerns from a group of dissenting members. Moreover, even if employees are acting as consumers when voting on union representation, there is a real danger of taking that model too far. The informational concerns that Bodie emphasizes may lead to changes with significant costs of their own. Thus, I am not convinced that the gains from a consumer approach to union elections are large enough to warrant the regulatory response it demands.
Bodie describes numerous problems with the NLRB's current approaches to election governance, particularly their inability to provide employees with the information needed to make a choice about unionization. NLRB-run elections today often leave employees unable to predict a union's performance, as they fail to provide access to information that could help employees determine whether a union could improve work conditions and whether such improvements would outweigh the costs of dues, possible strikes, and bitterness within the workplace. Although these problems are important, they should be viewed in context. Information asymmetries are not limited to union elections, as employees generally have an abysmal understanding of their legal work conditions.2 Indeed, Bodie has focused on a situation where employees are privy to vastly more information than in most other circumstances. Nevertheless, his observations are salient and could very well serve as the prism through which the Board regulates union campaigns in the future. The question, then is how best to address the informational concerns that the consumer paradigm identifies.
The problem of unions and employers providing inaccurate or insufficient information to voting employees has no quick fix. As Bodie notes, one option is some form of mandatory disclosure rule. Such a rule could not realistically provide the most valuable type of information. Voting employees want to know how well a union will be able to bargain on their behalf, as well as information about their employer's financial situation and ability to make concessions. Yet few unions or employers would be willing or able to provide such information.
A better option would be to focus more directly on employees' desire to predict the effectiveness of the union. Questions about the union's ability to negotiate with the employer and the impact of unionization on the workplace go to the heart of the decision whether to unionize. Although Bodie rightly points out some difficulties in employees' ability to judge these questions, there are useful avenues for giving employees the information they need. In particular, providing employees real experience with a union as their representative would eliminate many of the information deficiencies, as there is no better way to predict future performance than to observe actual performance. This is what consumers do all the time. They are allowed, and often encouraged, to try out a product and given the option to return it if they are not satisfied. If Bodie is correct that voting employees act like consumers purchasing services, why not treat them the same way? We could create a system that allows employees to easily choose union representation, while also allowing them to easily jettison the union if its performance is lacking. Employees could essentially rent union services for a period of time, then decide either to end the relationship or to "own" the union by continuing to purchase its services indefinitely.
This approach is one example of what Professor Sam Estreicher has called an "easy in, easy out" policy.3 Under such a scheme, employees could easily designate a union as their representative—for example, through a "card-check" designation in which a majority of employees sign cards stating that they want the union as their representative.4 In conjunction with this "easy in" rule, employees would have a much easier time ridding themselves of a union. Currently, employees' ability to remove an incumbent union is significantly limited by the "decertification bar," which prohibits questioning a union's majority status for up to a year after a union is certified as a bargaining representative, and the "contract bar," which applies for up to three years while a contract with the employer is in force. An "easy out" rule would reduce these impediments. For example, the one year recognition bar against decertifying a newly recognized union could be reduced by several months, which would make it much easier for employees to end their relationship with a union that did not live up to expectations. This change would also lower the risk of an initial decision to unionize. If choosing a union no longer involves a one year lock-in period, some employees may be more willing to take the risk of unionization. The risk of lock-in is also heightened by the contract bar. An easy out policy could expand the current thirty day window that is now employees' only chance to decertify a union that has entered into a contract with an employer.5 Lowering the current three year maximum on the contract bar would also provide employees more opportunities to change their mind about unionization.
Despite its faults, the current "hard in, hard out" system has its justifications. The risk of union intimidation and lack of secrecy are often cited as a reason to require a "hard in" secret ballot election rather than a card-check designation. Moreover, the need for stability and more time for unions to develop new bargaining relationships free from threats of decertification support the hard out rule. Eliminating these safeguards could impose severe costs to employees' freedom to choose whether to unionize and to unions' ability to maintain support after they are initially designated. Indeed, I am far from certain that those costs would not overwhelm the informational benefits of an easy in, easy out approach. Nevertheless, a well-designed consumer model could minimize some of these costs.
Completely eliminating the recognition bar, for example, could undermine the goals of the consumer paradigm, as a union may be removed before it had any real chance to represent employees. Thus, an easy out rule should give unions some type of buffer—perhaps six months—during which they cannot be decertified. After that period, employees would be able to remove a union in the same ways that they are able to designate a union, including a card majority. Six months should be enough time for employees to see the union at work and decide if it is worth the dues and other associated costs. If not, employees could easily decide not to continue buying the union's services. One could similarly limit the contract bar by making it effective for only one year or by greatly expanding the thirty day window that may now occur as little as once every three years. It makes sense to provide some time during which a contractual relationship is protected from decertification threats, but three years seems excessive under a regime that prizes employees' ability to choose whether to purchase union services above all else.
In spite of these benefits, the costs of instability from an easy out scheme would still be considerable. Some unions may be forced into a semi-permanent campaign, which could interfere with their ability to effectively represent employees. Compounding this problem is the reality that employers would have a strong incentive to resist meaningful bargaining, because delay is likely to foster discontent among employees and could lead to the union's removal. Thus, an important facet of any easy out system would be a significant expansion of the Board's ability to punish employer unfair labor practices, especially failures to bargain in good faith. Fines, in particular, could counteract employers' incentive to delay, although no remedial measure is likely to completely eliminate this problem.
I will confess to being agnostic as to whether an easy in, easy out policy would do more harm than good. At a minimum, it would provide a promising option to a system that seems to be broken. Other proposals, such as the easy in, hard out approach of the Employee Free Choice Act are more about the NLRB's inability and unwillingness to enforce rules against employers' unlawful interference in the election process than anything else. As such, it does little to alleviate the information deficiencies that plague today's NLRB-run elections. In contrast, a major advantage of the easy in, easy out model is that it seems to be one of the best means for providing employees with the information that Bodie rightly identifies as crucial to employees' decisionmaking process. Employees would be able see how a union actually performs in their workplace with the knowledge that they have a relatively easy option of removing the union if they are unsatisfied. The conditions surrounding this "rent-to-own" approach are not ideal—the union would be new and the employer would have strong disincentives to cooperate. However, a union would be able inform employees of the situation with the hope that it would survive this tryout period and enjoy a more stable relationship in the future. Indeed, I suspect that an easy in, easy out system, especially one accompanied by increased enforcement powers for the NLRB, would result in a higher rate of unionism. An incumbent union is likely to have far more success in surviving a challenge to its majority status than a outside union trying to gain such support for the first time.
Bodie's consumer paradigm provides a valuable insight into the information deficiencies implicated by union elections. At a minimum, this model should inform the NLRB's regulation of union elections. Yet relying on the model too much could lead to more problems than it is worth. Identifying the most effective middle ground between these concerns is difficult; indeed, such ground may not exist. Despite this uncertainty, Bodie's consumer paradigm serves an important role, as it provides further evidence that the NLRB's current governance of elections is in dire need of reform and establishes itself as a essential factor in any future attempts at reform.
Preferred citation: Jeffrey M. Hirsch, Rent-to-Own Unionism?, 94 Va. L. Rev. In Brief 9 (2008), http://www.virginialawreview.org/inbrief/2008/03/24/hirsch.pdf.
[Back to this issue]