August 30, 2016 | Backgrounder on Labor
Only 6 percent of current unionized workers ever voted for their own union. Most U.S. union members are represented by unions they had no say in electing—they are represented by a union that other people chose, often decades earlier.
Workers are legally represented by unions in workplace negotiations. In theory, this legal representation is legitimized through a democratic process in which workers vote in favor of union representation. In practice, only 6 percent of those covered by unions under the National Labor Relations Act (NLRA) have ever voted for union representation. In some cases, National Labor Relations Board (NLRB) rules allow unions to organize workplaces without majority support or even an election. Unions do not have to stand for re-election after being formed, and it is difficult for workers to request a decertification election to remove an unwanted union. As a result, most current U.S. union members are represented by unions they had no say in electing. The NLRA election process, while nominally democratic, does not measure the preferences of current American workers.
The preamble to the NLRA declares its goal as “protecting the exercise by workers of full freedom of association, self-organization, and designation of representatives of their own choosing.” On paper, the NLRA fulfills this goal by allowing workers to democratically elect bargaining representatives. If workers vote for union representation, their union then negotiates working conditions on their behalf. Theoretically, the election process makes unions accountable to the workers they represent. In reality, very few union members ever voted for union representation.
Chart 1 shows the total number of workers who voted for union representation in National Labor Relations Board (NLRB) elections between 1973 and 2015. Unions represent 8 million workers under the NLRA. Only 478,000 of these—6 percent—voted for union representation at some point in their careers and remain employed by the company at which they voted. The remaining 94 percent are represented by a union they either voted against or never voted on.
Four aspects of NLRB election procedures cause unions to primarily represent workers who never voted for their representation:
These polices systematically reduce union accountability to their members.
The NLRA allows—but does not require—employers to request a secret ballot election before it requires them to recognize and bargain with a union. The Supreme Court has ruled that an employer may voluntarily recognize a union if a majority of its workers publicly sign membership cards (card check). The right to request an election resides with the employer, not with employees.
Unions find organizing under card check much easier than with secret ballot elections. If a worker decides not to sign the card, organizers can come back again and again to pressure him to change his mind. In some cases, organizers do not even explain that the cards represent a binding commitment to support the union. As a result, once an employer assents to card check, unions almost always become workers’ representatives.
Employers typically prefer secret ballot elections, but unions have become adept at pressuring them into accepting card check. Unions will wage public relations campaigns against corporations designed to hurt their reputation and drive away customers. Unions threaten to continue these attacks until the company agrees to forgo a secret ballot vote. As former UNITE-HERE President Bruce Raynor explained, unions win these campaigns because: “We’re not businessmen, and at the end of the day, they are. If we’re willing to cost them enough, they’ll give in.”
The AFL–CIO claims that 80 percent of its newly organized workers join through card check. Independent estimates put the figure closer to one-third. Either way, large numbers of union members are organized without an election. Many of these workers would not choose representation if they could cast a private ballot.
The United Auto Workers (UAW) demonstrated this recently in a Volkswagen plant in Chattanooga, Tennessee. A majority of workers at the plant publicly signed union membership cards. Volkswagen then requested a secret ballot election. Both the union and the company expected the union to win the vote. However, in the privacy of the voting booth, a majority of the plant’s workers voted not to join the UAW.
Once a union is certified, either through card check or an election, it remains certified indefinitely. This stands in stark contrast to political representation. Political representatives in America serve for a fixed term of office. Their election gives them a mandate to represent their constituents, but that mandate lasts only a fixed period of time. To continue in office, political representatives must run for re-election.
Union representation does not operate on this principle. The NLRB presumes that workers continue to support their union absent strong evidence to the contrary. This “presumption of majority support” means that most unions remain certified indefinitely. Unless workers successfully petition for decertification, the union remains in place. New hires at a unionized firm inherit the representatives for which their predecessors voted.
For example, between 1936 and 1941, the UAW organized the Detroit automakers—General Motors, Ford, and Chrysler. The automakers’ current employees were all hired well after 1941, and they never voted for UAW representation themselves. But the “presumption of majority support” means that the UAW never had to win their support in a new election.
Most U.S. union members inherited representation this way. Chart 2 shows the number of pro-union votes in NLRB organizing elections unions won between 1973 and 2015, and the number of those workers still working at their company. The overwhelming majority of workers who voted for union representation have either retired or switched employers. Their decisions nonetheless bind the employees who replaced them.
When employers request a vote, unions often win without the support of a majority of workers. On paper, the law seems to prohibit this. The NLRA specifies that:
Representatives designated or selected for the purposes of collective bargaining by the majority of the employees in a unit appropriate for such purposes, shall be the exclusive representatives of all the employees in such unit for the purposes of collective bargaining.
However, NLRB regulations require unions only to obtain a majority from those voting in organizing elections—not necessarily a majority of the overall workforce. The NLRB presumes that non-voting workers assent to their voting colleagues’ choices.
For example, on August 19, 2015, workers at the LINQ Hotel and Casino in Las Vegas voted on whether to join the UAW. The UAW got 74 votes, 42 workers voted against it, and 131 did not vote. The UAW won the election despite the fact that fewer than one-third of all employees voted for representation by the UAW.
This system causes unions to organize many workplaces with support from only a minority of workers. Figure 1 shows the results of all NLRB certification elections between fiscal year (FY) 2012 and FY 2015. One-third of all employees organized in these elections work in firms where the union lacked majority support. If unions needed majority support in the workplace—not just a majority of voters—they would have organized 60,000 fewer workers between 2012 and 2015. Unions represent many workplaces where most workers did not vote for them. They continue to represent these workers indefinitely.
Dissatisfied workers can petition for a decertification election to remove their union representative. However, NLRB policies and union opposition make requesting a decertification vote difficult. Workers must collect signatures from 30 percent of the employees in their bargaining unit. Under the NLRB’s “contract bar” doctrine workers cannot collect or submit these signatures during the course of a collective bargaining agreement. This typically leaves workers with just one month every three years to petition for a new election. Furthermore, workers may not collect these signatures while on the clock or in work areas. This limited time window and restrictive policies make gathering enough signatures to trigger an election challenging.
Unions heavily pressure workers not to sign or distribute these petitions; they do not want to lose members and dues. Most unions also expel workers who support decertification. The UAW constitution, for example, calls for the expulsion of any member “affirmatively engaged in efforts to decertify the International Union or any subordinate body thereof.”
Being expelled from the union means that a worker cannot vote in elections for union officers or vote on ratifying a new contract. Signing a decertification petition means that workers lose the little say they have in how their union operates—even though it will still represent them if the decertification attempt fails.
These barriers make decertification votes extremely rare. In FY 2015, fewer than 10,000 private-sector union members voted in a decertification election. That amounts to just 0.1 percent of the 8 million workers unionized under the NLRA. Unions lost three-fifths of these decertification elections. But the remaining 99.9 percent of unionized workers never even had the opportunity to vote.
The combination of these policies undermines the NLRA’s goal of allowing workers to designate representatives of their own choosing. Fully 94 percent of union members did not vote for union representation. They are represented by a union that someone else chose, often decades earlier. Workers who want a different union representative—or no union—face considerable difficulty in implementing these preferences. Current NLRB election processes fail to reflect the preferences of American workers.—James Sherk is Research Fellow in Labor Economics in the Center for Data Analysis, of the Institute for Economic Freedom and Opportunity, at The Heritage Foundation. The author is grateful to CDA intern Max Lies for his invaluable research.
Charts 1 and 2 show estimates of the number of union-represented workers who voted for a union and remain represented by the union for which they voted. The Heritage Foundation’s Center for Data (CDA) analysis calculated these figures as follows:
Unionstats.com contains data on union coverage by industry from 1983 to 2015. For prior years, the CDA estimated airline and railway union coverage by calculating the ratio of railway and airline union coverage to total private-sector coverage from 1983 to 1985. The CDA then multiplied this ratio by total private-sector-coverage estimates for 1973 to 1982 to obtain annual estimates of union coverage in the airline and railway industries.
 The NLRA covers most private-sector employees. It does not cover private-sector workers in the railway and airline industries (who are regulated under the Railway Labor Act) or the agricultural sector, nor does it apply to government employees.
 The National Labor Relations Act, 29 U.S. Code § 151.
 The National Labor Relations Act is principally enforced by the National Labor Relations Board.
 Heritage Foundation calculations using data on union election votes from the NLRB, and data on the job tenure of unionized employees and total union membership from the CPS. See appendix for details.
 29 U.S. Code § 159(9)(c)(B).
 NLRB vs. Gissel Packing Co., 395 U.S. 575, 596–597.
 James Sherk and Ryan O’Donnell, “EFCA: High Pressure Spin-Selling and Creating Organizing for Unions,” Heritage Foundation WebMemo No. 2335, March 11, 2009, http://www.heritage.org/research/reports/2009/03/efca-high-pressure-spin-selling-and-creative-organizing-for-labor-unions.
 Jarol Manheim, The Death of a Thousand Cuts: Corporate Campaigns and the Attack on the Corporation (New York: Routledge, 2000).
 Jarol B. Manheim, George Washington University, professor, before the Subcommittee on Workforce Protections, Committee on Education and the Workforce, U.S. House of Representatives, July 23, 2002.
 Rafael Gely and Timothy Chandler, “Card Check Recognition: New House Rules for Union Organizing,” Fordham Urban Law Journal, Vol. 35 (2008), p. 247. Note that Table 2 of that report shows twice as many workers organized through NLRB elections than card check recognition. AFL-CIO membership gains due to “mergers and affiliation” are not newly organized workers.
 Lydia DePillis, “Auto Union Loses Historic Election at Volkswagen Plant in Tennessee,” The Washington Post, February 14, 2014, https://www.washingtonpost.com/news/wonk/wp/2014/02/14/united-auto-workers-lose-historic-election-at-chattanooga-volkswagen-plant/ (accessed April 5, 2016).
 Levitz Furniture Company of the Pacific, 333 N.L.R.B No. 105.
 29 U.S. Code § 159(a) (Emphasis added).
 National Labor Relations Board, “Cases Closed–August 2015,” Election Reports, Case No. 28-RC-156969, https://www.nlrb.gov/sites/default/files/attachments/basic-page/node-4416/August%202015.pdf (accessed April 5, 2016).
 National Labor Relations Board, “Basic Guide to the National Labor Relations Act,” 1997, pp. 8–10, https://www.nlrb.gov/sites/default/files/attachments/basic-page/node-3024/basicguide.pdf (accessed April 5, 2016).
 National Right to Work Legal Defense Foundation, “Decertification Election,” http://www.nrtw.org/decertification-election (accessed April 5, 2016).
 Article XXXI, Section 24 of the constitution of the United Automobile, Aerospace, and Agricultural Implement Workers of America, http://uaw.org/uaw-constitution/ (accessed April 12, 2016).
 More precisely 9,799 employees worked in bargaining units in which decertification elections were held in FY 2015. Heritage Foundation calculations using data on FY 2015 union election votes from: National Labor Relations Board, “Election Reports–FY 2015,” https://www.nlrb.gov/reports-guidance/reports/election-reports/election-reports-fy-2015 (accessed April 5, 2016).
 Ibid. Unions lost 101 of the 174 decertification elections held in FY 2015.
 This includes RC, RM, and RD elections. This pro-union voters twice in cases of decertification elections: first when they vote for the union, then when the same workers vote to retain that union as their representative. This situation is rare and only slightly affects these results. To the extent it does, the estimates in Charts 1 and 3 slightly overstate the number of workers who voted for their current bargaining representative.
 National Labor Relations Board, “Annual Reports,” https://www.nlrb.gov/reports-guidance/reports/annual-reports (accessed April 5, 2016).
 National Labor Relations Board, “Statistical Tables–FY 2010,” https://www.nlrb.gov/reports-guidance/reports/annual-reports/statistical-tables-fy-2010 (accessed April 5, 2016).
 In elections featuring more than one labor organization, the CDA counted a vote for either union as a vote for union representation. See also National Labor Relations Board, “Election Reports,” https://www.nlrb.gov/reports-guidance/reports/election-reports (accessed April 5, 2016).
 The CDA included both union members and non-union members covered by a collective bargaining agreement in these calculations.
 This analysis assumes that the job-tenure distribution does not significantly differ between the average union member and newly organized workers.
 To understand the reason for this adjustment, one should note that many unionized firms have gone out of business. No one who voted for union representation at those firms remains with his original employer. Job-tenure surveys of workers currently represented by a union do not account for this contraction in overall union coverage. To see this mathematically, note that the CPS job-tenure supplement provides data on the proportion of workers currently represented by unions and employed by the same firm at time t and t–n, which can be represented as P1 = Number of Workers Represented by Unions Employed with the Same Firm in time t and t-n / Number of Workers Represented by Unions at time t. The goal is to find the proportion of workers represented by unions at time t-n who are still employed at time t. This can be represented as P2 = Number of Workers Represented by Unions Employed with the Same Firm in time t and t-n) / Number of Workers Represented by Unions at time t-n. One can derive P2 from P1 through the identity P2 = P1 * Number of Workers Represented by Unions at time t / Number of Workers Represented by Unions at time t-n.
 Barry T. Hirsch and David A. Macpherson, “Union Membership and Coverage Database from the Current Population Survey: Note,” Industrial and Labor Relations Review, Vol. 56, No. 2 (January 2003), http://www.unionstats.com/ (accessed April 5, 2016).
 This ratio is 1.085.
 This ratio is 0.044.