New Data: NLRB Process Fails to Ensure a Fair Vote

John Logan, Erin Johanssonand Ryan Lamare

Press Coverage

The National Labor Relations Board (NLRB) is mandated by the National Labor Relations Act (NLRA) to protect the rights of employees to form unions and collectively bargain. One key role of the NLRB is to oversee representation elections, where workers decide whether or not to form unions. Several recent studies have indicated that the current NLRB election process fails to ensure that workers can freely form unions.[1]

Our new analysis of NLRB election data reveals how current NLRB procedures, which grant employers significant control over the timing of the election process, can prevent workers from fairly choosing whether or not to have union representation. Our key findings include the following:

  • The current NLRB election process allows significant delays in the vote. Employers can delay a vote once workers file a petition by objecting to the proposed bargaining unit and forcing a pre-election hearing. In cases where a hearing is held, the election occurs an average of 124 days after the petition is filed.
  • Delayed elections exacerbate the problem of unfair labor practices. Controlling for factors like industry, size of the unit, and location, there is a considerable causal relationship between the length of election delay and the number of NLRB complaints filed against employers. The longer the delay between the filing of the petition and the election date, the more likely it is that the NLRB will charge employers with illegal activity.
  • Anti-union campaigns begin early in the election process. If workers believe an employer is violating the law after they file a petition for an election, they can choose to file an unfair labor practice charge that “blocks” the vote from proceeding. Of these elections blocked by unfair labor practice charges against employers, 39 percent were blocked by charges filed within ten days of the petition being filed, while more than half of cases were blocked within 20 days of the petition.[2]

The NLRA does not require a lengthy election period, nor does it mandate any procedures on the requisite length of elections, but the NLRB’s current administrative process is subject to manipulation in a manner contrary to the law’s intent (see Appendix A). While the NLRB reports that the median time between the petition and election is 39 days, our analysis reveals that a significant percentage of elections experience far longer delays due to the hearing process.[3] There is no limit on a party’s ability to demand a pre-election hearing on a range of issues, including objecting to the eligibility of the employees in the proposed unit or the scope of the unit. In cases where a pre-election hearing was held, the election occurred an average of 124 days after the petition was filed. Ten percent of elections with a hearing took place more than 193 days after the petition was filed. When an election case involves a decision by the Labor Board, the vote is delayed by an average of 198 days.

The majority of petitions proceed to an election without holding a hearing. However, the mere fact that the NLRB allows parties the ability to delay cases for an extended period simply by forcing a hearing skews the process in employers’ favor. For example, in order to avoid a hearing and the resulting delay, workers often agree to employers’ demands to change the description of the bargaining unit. Influencing the size and composition of the bargaining unit can advantage one party to the detriment of the other (e.g., excluding known union supporters). Indeed, anti-union consultants include in their services advice to employers on the importance of manipulating the composition of the bargaining unit during the election process.[4]

Delayed elections expose workers to more illegal employer campaigning

Our analysis reveals that the NLRB election process is conflict-ridden, and that delayed elections exacerbate this conflict.[5] We examined NLRB elections and matched them with corresponding unfair labor practice charges filed no earlier than three months prior to the petition date, and no later than three months after the election date. Unfair labor practice charges against employers were filed in 46 percent of elections. In more than half of the elections with charges filed, the NLRB found merit to at least one charge.[6] Of all the charges filed in connection with these elections, the NLRB found merit to 37.2 percent, including 19.5 percent where they issued a complaint charging the employer with violations of the law. This percentage of election-related charges resulting in complaints is nearly double the percentage (8.3 percent) of all charges filed against employers during a similar time period that resulted in complaints.[7]

What is the significance of this data? We conducted a series of multivariate regressions and found the longer the delay between the filing of the petition an d election date, the more likely it is that the NLRB will issue complaints charging employers with illegal activity. Controlling for factors like industry, size of the unit, and location, we found a considerable causal relationship between the length of election delay and the number of NLRB complaints issued.[8] The amount of time between the petition filing and election date was the most substantial predictor of NLRB complaint increases.[9] In addition to the above regression, we conducted a number of other tests that also demonstrated a substantial connection between election delays and NLRB complaints.[10]

As our numbers reveal, delaying elections is clearly a strategy that advantages anti-union employers. Gordon Lafer of the University of Oregon notes how consultants counsel employers to challenge everything, in order to delay the vote to buy more time to engage in anti-union campaigning.[11] In an article titled “Time Is On Your Side,” the law firm Jackson Lewis has advised clients that pre-election legal proceedings should be considered “an opportunity for the heat of the union’s message to chill prior to the election.”[12] Chicago consultant John Sheridan explained that “If a [certification] petition is filed today, and the election is held in two weeks, we’ll lose it.”[13] By manipulating the process to delay the election, employers extend the exposure of employees to an anti-union campaign and thereby suppress union support. Even if the workers vote to form a union, the employer has communicated to workers that exercising their rights as a union will come at a high cost.

Employers begin campaigning before the NLRB petition is filed

As our analysis reveals, reducing the period of time between the petition and election would likely decrease the number of NLRB complaints issued during an organizing campaign. Yet a shortened election period does not equate to an insufficient period of campaigning. The day that workers file a petition for union representation is not the beginning of the campaign. An election petition can only be filed with the NLRB after workers have demonstrated that at least 30 percent of the workforce wants to be represented by a union. Such a demonstration of interest must be in writing, on signed petitions or signed cards in which workers authorize a union to represent them. The signing of petitions or authorization cards is preceded by much discussion among employees about their working conditions and whether to form a union, a decision to contact a union and the selection of which union to be contacted for assistance with the process, and the solicitation of co-workers to sign the petition/cards. Only after all of these steps are completed, and the requisite number of signatures obtained, can a petition be filed with the NLRB. The filing of an NLRB election petition is one of the last steps in the process, but it currently triggers a lengthy delay.

Employers can and do communicate their views about unions to employees well before the petition is filed, and many do so during inductions when employees are first hired. In fact, anti-union consultants advise employers to communicate their views on unions well before the union arrives at the doorstep.[14] Employers are encouraged to include their opposition to “third-party intervention” in orientation materials, as employees are considered more impressionable when they start a new job.[15] Employers communicate their viewpoints even in the absence of any organizing effort through posters in the breakroom, company email communications, and one-on-one discussions between supervisors and employees. Walmart, for example, has communicated its views on unions to employees at a company-wide level, even though organizing efforts were underway at only a few locations.[16]

In addition to employers sharing their views about unions, some anti-union employers begin actively campaigning against unions prior to a petition being filed. If workers believe their employer is violating labor laws after their petition has been filed, they can choose to file an unfair labor practice charge that blocks an election from proceeding. Significantly, when these allegations are made they are made early in the election process, implying that employers are responding quickly with anti-union tactics. Between fiscal years 2006 and 2009, 6 percent of representation petitions filed were blocked from proceeding due to unfair labor practice charges alleging employer violations.[17] Though this practice is uncommon, the timing of the charges is informative. Of these blocked elections, 39 percent were blocked by charges filed within ten days of the petition being filed, while more than half of cases were blocked within 20 days of the petition. Thus, these cases demonstrate that anti-union tactics are taking place within days of the petition being filed, if not earlier.

Forthcoming research from Kate Bronfenbrenner of Cornell University and Dorian Warren of Columbia University further reveals evidence of early employer communication. They found that employees file serious unfair labor practice charges with the NLRB alleging illegal employer tactics well before any petition for an NLRB election has been filed: “30 percent of serious violations against workers by employers occurred 30 days before the petition was filed and 47 percent of all serious allegations against employers occurred before the petition was filed.”[18] In 2009 Bronfenbrenner reported that 23 percent of all unfair labor practice charges against employers during NLRB election campaigns were filed before the petition was filed, and 16 percent were filed more than 30 days before the petition was filed.[19]

Without a fair and even election process, workers lose hope

Those who argue that the NLRB election process works well point to the fact that workers still vote to form unions in a majority of NLRB elections. Yet this argument ignores the fact that a large percentage of NLRB organizing efforts fail once workers file a petition and the employer engages in an anti-union campaign. A 2005 study by researchers at the University of Illinois at Chicago found that in 91 percent of petitions filed with the NLRB, a majority of workers signed cards indicating they wanted a union before the petition was filed.[20] However, only 69 percent of those election petitions actually resulted in elections. In the other 31 percent of cases, the workers decided to abandon and withdraw their petition before the election was even held. The authors’ data “shows that when employers use a broad set of tactics—legal or illegal—they substantially increase the likelihood that the union will lose its majority status.”

Similarly, a 2008 analysis by John-Paul Ferguson of Stanford Business School found that the 14,615 NLRB representation elections held between 1999 and 2004 represented only two-thirds of all petitions filed—a full 35 percent of petitions were withdrawn before the election was held.[21] Based on the largest single longitudinal analysis of NLRB elections available, Ferguson revealed that unfair labor practice charges against employers are associated with an average 25 percent higher likelihood of petition withdrawal. Without a fair and swift election process to mitigate employer resistance to their efforts to form unions, workers are prone to give up.

Case studies

As the cases below illustrate, by forcing parties to resolve legal issues prior to the vote, the NLRB election process can be manipulated, ridden with conflict, and can require employees to wait years before they are able to vote.

Avante at Wilson

In August 2002, a group of licensed practical nurses and registered nurses at the Avante at Wilson nursing home in Wilson, NC, filed a petition in support of forming a union with the United Food and Commercial Workers union. The nursing home objected to the petition on the basis of eligibility to vote, arguing the nurses were supervisors, not employees, and therefore ineligible to participate in an NLRB election. The NLRB ruled in 2006, over 1,500 days after the petition, that the nurses were employees, and that the employer had “failed to support with the requisite facts its position that staff nurses” held the authority of supervisors.[22]

The elections for the two units of nurses were finally held in July 2008. After waiting 2,148 days (more than five years) since first petitioning for a union, the nurses, not surprisingly, voted against forming a union.

Durham School Services

In late April 2010, school bus drivers with Durham School Services in Antioch, IL, filed a petition to have union representation with the International Brotherhood of Teamsters. The workers wanted to hold an election at the end of May, before the school year was over. The company argued that May was too soon, proposed the election be held in Octo ber, and objected to the proposed unit of drivers based on the eligibility of drivers it claimed were supervisors, thus forcing the NLRB to hold a pre-election hearing. The Regional Director eventually ordered that the election be held in late August. Durham again objected to this proposal and suggested the election be held at a later date, and the NLRB agreed to do so.

The election was eventually held September 10, 2010, more than four months after the original petition was filed. With the advantage of a four-month delay, Durham had run an extensive anti- union campaign, where the company held captive audience meetings, distributed anti-union literature, and called the police on a small group of community leaders who were peacefully delivering a letter to the company in support of the bus drivers. Despite these delays and anti-union tactics, the workers voted in favor of union representation, but the lengthy delay meant they were exposed to almost daily anti-union campaigning by Durham.[23]

Ash Grove Cement

On May 11, 2005, 87 percent of the line workers at the Ash Grove Cement plant in Seattle, WA, petitioned for representation with the International Longshore and Warehouse Union. The company objected to the proposed unit, alleging it included both supervisors not eligible to organize and workers who did not share a community of interest with the rest of the plant. After a hearing, the Regional Director rejected the employer’s position and issued a decision on June 6 directing the election to proceed with the initially proposed unit.[24] The election was held July 6, 56 days after the petition was filed. By forcing a hearing, the employer had gained more time, in which it ran an aggressive anti-union campaign. High-level company executives flew in and forced employees into captive audience and one-on-one meetings.[25] Despite the overwhelming support workers initially demonstrated for the union, after weeks of anti-union tactics by the employer, workers voted against union representation.

The following cases illustrate how anti-union employers act immediately to shut down an organizing effort as soon as they find out about it, or communicate their views on unions where there isn’t even an organizing effort underway:

Green Valley Manor

In the span of three weeks in 2007, employees of this nursing home in St. Louis, MO, attempted to form a union and were forced to stop in the face of severe retaliation by their employer. Below is a timeline of events:

Green Valley Manor Summer 2007 Timeline of Events

The Board eventually ruled the firings and retaliation tactics were illegal activity,[26] but the workers had already postponed the election indefinitely.

Vi-Jon Laboratories

Workers for this Smyrna, TN, manufacturer began organizing to form a union in April 2010. In response, Vi-Jon held captive audience meetings, gave out raises and a “tool allowance,” and in December fired 18 employees. All of this occurred prior to the petition for an election that was filed with the NLRB on December 20, 2010.[27]

Sunrise Senior Living

Despite a lack of any ongoing effort by their employees to form a union, a Sunrise Living nursing home facility outside of Philadelphia, PA, forces employees to sit through a mandatory, one-hour anti-union meeting every March. Employees report to SEIU that patients are left with minimal care during this meeting.[28] The company shows an anti-union video and distributes a power point that asks: “Who pays the bills, the union or the company?” and asserts that “joining a union has consequences…you and your family could be living with the facts for a very long time.”[29]

Conclusion: The NLRB must act to reform the current election process

When the NLRA was signed into law 75 years ago, Congress granted the NLRB a mandate to promote and protect the right of workers to collectively bargain for wages and working conditions. The intent of the Act was to address the inequality in bargaining power between employers and employees. From a very early period, lawmakers and NLRB officials worried that employer delays would undermine the integrity of the union election process. Just five years after the enactment of the NLRA, one senior official expressed concern that mandatory NLRB elections may “furnish to employers…a means of sabotaging the bargaining process through dilatory tactics.”[30] These concerns have proved justified. Our new findings suggest that the NLRB has not lived up to its original mandate in its implementation of the law. To the detriment of workers, the NLRB’s election procedures have created inequalities and unfairness over the timing of elections, and allowed the process to be plagued with unfair labor practices


[1] Kate Bronfenbrenner (May 2009). No Holds Barred: The Intensification of Employer Opposition to Organizing, Washington, D.C., Economic Policy Institute and American Rights at Work Education Fund; John Schmitt and Ben Zipperer (March 2009). “Dropping the Ax: Illegal Firings During Union Election Campaigns, 1951-2007,” Center for Economic and Policy Research.

[2] We obtained data on all RC elections blocked between fiscal years 2006 and 2009 through a FOIA request fulfilled by the NLRB 30 September 2010.

[3] Data from a FOIA request of the National Labor Relations Board, fulfilled December 2009.

[4] John Logan (2002). “Consultants, lawyers, and the ‘union free’ movement in the USA since the 1970s,” Industrial Relations Journal , 33:3. pp 200-201.

[5] Source, cited herein as “matched data:” Erin Johansson obtained data on all unfair labor practice (ULP) cases and all election petitions closed between 1999 and 2007 from a FOIA request fulfilled by the NLRB. In order to determine how many ULPs were associated with each election petition, we selected one year, 2003, and manually matched all representation petitions filed in 2003 with corresponding ULPs. For further information on methodology, email Erin Johansson:

[6] Following the NLRB’s definition of “merit,” we included unfair labor practice charges where there were pre-complaint settlements or adjustments, in addition to those where complaints were issued.

[7] Source for percentage of all charges resulting in complaints: Annual Report of the National Labor Relations Board for the Fiscal Year Ended in 30th September 2003, pages 8-9, Annual_Reports/NLRB2003.pdf.

[8] In our regression analysis, we used the company’s NAICS code to determine its industry. We grouped location by official Census Bureau designations (i.e., New England, Mid-Atlantic, etc.). Our reference categories for these variables were manufacturing (for industry) and New England (for location).

[9] In an OLS regression, using NLRB complaints as a continuous dependent variable and time between petition and election as a continuous independent variable, we found a significant and positive relationship at the .01 level of analysis. We used OLS regressions, ordinal regressions, bivariate correlations, and chi-square tests to determine the relationship between election delays and NLRB complaints. In our regressions, we tested for multicollinearity of the independent variables, and used ANOVA tests to determine that the model was effective in predicting the dependent variable. All of our statistical tests and analyses are available on request.

[10] For the ordinal regressions, we ran election delay as a continuous variable in one model, and as a categorical variable (with 60+ days as the reference point) in another model. We also ran NLRB complaint as both a continuous and categorical variable, to ensure sufficient N values for the highest number of charges. Our ordinal regression passed the parallel lines test, and the model fit the data well, with a Nagelkerke R 2 of .383. The full regressions are available on request.

[11] Gordon Lafer (July 2007). “Neither Free Nor Fair: The Subversion of Democracy Under NLRB Elections,” American Rights at Work, fair.html/

[12] UnionkNOw, September 2001, quoted in John Logan (2004), “The Fine Art of Union Busting,” New Labor Forum , 13:22, pp. 76-91.

[13] David Moberg (1992). ‘Union Buster,’ Chicago Reader , 24 July, quoted in Logan, 2002.

[14] Logan, 2002, 198.

[15] Lafer, 2007.

[16] Erin Johansson (2007). “Checking Out: The Rise of Wal-Mart and the Fall of Middle Class Retailing Jobs,” Connecticut Law Review , Vol 39, No. 4,

[17] We obtained data on all RC elections blocked between fiscal years 2006 and 2009 through a FOIA request fulfilled by the NLRB 30 Sept. 2010.

[18] Kate Bronfenbrenner and Dorian Warren (2011). The Empirical Case for Streamlining the NLRB Certification Process: The Role of Date of Unfair Labor Practice Occurrence . Forthcoming.

[19] Bronfenbrenner, 2009

[20] Chirag Mehta and Nik Theodore (December 2005). “Undermining the Right to Organize: Employer Behavior During Union Representation Campaigns,” Center for Urban Economic Development, University of Illinois at Chicago, http://www.americanrightsatwor organize-employer-behavior-during-union-representation-campaigns.html.

[21] John-Paul Ferguson (October 2008). “The Eyes of the Needles: A Sequential Model of Union Organizing Drives, 1999–2004,” Industrial and Labor Relations Review , Vol. 62, No. 1.

[22] Avante at Wilson, 348 NLRB No. 31 (2006)

[23] Kim Keller, Deputy Director of Organizing, International Brotherhood of Teamsters. Personal interview by Erin Johansson, 7 September 2010.

[24] Ash Grove Cement Company , 19-RC-14734, Regional Director Decision and Direction of Election, 6 June 2005.

[25] “Fear Campaign: Union votes to take on worker safety at cement plant,” Real Change News, 6 July 2005.; Brier, Jon, International Organizer, ILWU. Personal interview by Erin Johansson, 3 December 2010.

[26] Green Valley Manor , 353 NLRB No. 92 (2009).

[27] Laura Yancey, Organizer, United Food and Commercial Workers Union. Email interview by Erin Johansson, 13 April 2011

[28] Eric Noyes, Assistant Organizing Director, SEIU Healthcare Pennsylvania. Personal interview by Erin Johansson, 13 April 2011.

[29] Copy of Sunrise powerpoint on hand with Erin Johansson at IRLE.

[30] Joseph Rosenfarb (1940). National Labor Policy and How It Works. New York and London: Harper and Brothers, p. 314.

About the authors

John Logan is Director and Professor and Director of Labor Studies at San Francisco State University and
Senior Labor Policy Specialist at the UC Berkeley Labor Center. Erin Johansson is Research Director of
American Rights at Work and Visiting Scholar at the Institute for Research on Labor and Employment at UC
Berkeley. Ryan Lamare is a Lecturer at the Manchester Business School at the University of Manchester in
Manchester, England.