? Threatening employees with loss of jobs or benefits if they should join or vote for a union.
? Threatening to close down the plant if a union should be organized in it.
? Questioning employees about their union activities or membership in such circumstances as will tend to restrain or coerce the employees.
? Spying on union gatherings, or pretending to spy.
? Granting wage increases deliberately timed to discourage employees from forming or joining a union. (NLRB 1997, 14)
Section 8(a)(2) prohibits an employer from “dominating” or illegally assisting a labor union. An employer who interferes in the formation or control of a union may be attempting to gain an advantage in bargaining or to discourage support of an actual union that is trying to organize the employees (NLRB 1997, 14).
Section 8(a)(3) prohibits an employer from discriminating against any worker because of union activity. Section 8(a)(3) charges are identified elsewhere in this report under the label “discipline, firings, changes in work terms.”
Examples of illegal discrimination under Section 8(a)(3) include:
? Discharging employees because they urged other employees to join a union.
? Refusing to reinstate employees when jobs they are qualified for are open because they took part in a union’s lawful strike.
? Granting of “superseniority” to those hired to replace employees engaged in a lawful strike.
? Demoting employees because they circulated a union petition among other employees asking the employer for an increase in pay.
? Discontinuing an operation at one plant and discharging the employees involved followed by opening the same operation at another plant with new employees because the employees at the first plant joined a union.
? Refusing to hire qualified applicants for jobs because they belong to a union. It would also be a violation if the qualified applicants were refused employment because they did not belong to a union, or because they belonged to one union rather than another. (NLRB 1997, 16)
Section 8(a)(4) prohibits an employer from punishing a worker for filing charges with the National Labor Relations Board.
Examples of violations of Section 8(a)(4) are:
? Refusing to reinstate employees…because they filed charges with the NLRB….
? Demoting employees because they testified at an NLRB hearing. (NLRB 1997, 17)
Section 8(a)(5) requires the employer to bargain in good faith with the union (NLRB 1997, 17). Section 8(a)(5) charges are identified elsewhere in this report under the label “refusal to bargain, repudiation of contract.”
While the NLRA guarantees workers the right to organize and collectively bargain, it in fact fails to provide penalties sufficient to deter employer violations of these protections. The Act does not include civil penalties or punitive damages. Employers who illegally fire employees for union support are liable only for back pay—minus any wages the worker has earned while waiting to be reinstated—and the worker is required to mitigate damages by searching for new employment. Other typical remedies for employer violations include the issuance of an order directing an employer to cease and desist from conduct found to be unlawful and an order directing that an employer post a notice informing employees of their rights under the NLRA.6
Further, the NLRA fails to prohibit a wide range of employer conduct that frustrates workers’ right to a free and fair union election. For example, employers are permitted to require employees attend “captive audience meetings,” mandatory meetings at which the employer is free to deliver anti-union messaging. Further, employers may flood the workplace with anti-union communications while banning union organizers from communicating with workers at the worksite. Employers are also free to have supervisors meet one on one with each of their direct subordinates to deliver anti-union messages. These anti-union activities are often aided by “union avoidance” consultants—who represent a multi-million-dollar industry.
The NLRA’s weak enforcement provisions and insufficient penalties, combined with its failure to outlaw certain coercive employer behaviors that make it impossible for workers to freely exercise their right to unionize and collectively bargain, has led more and more employers to adopt aggressive anti-union tactics in the face of worker organizing.
Our comprehensive analysis of NLRB election and ULP filings demonstrates that aggressive employer opposition is a common feature of union election campaigns. Using NLRB election data from 2016 and 2017, we find that employers were frequently alleged to have engaged in unfair labor practices around the time of elections. Employers were charged with violating federal law in 41.5% of all elections, with at least one ULP charge filed in each case. Between 19.9% and 29.6% of election filings were associated with a ULP charge that claimed employees were illegally fired for union activity. Compared with studies of NLRB data from the early 2000s, our more recent analysis shows higher rates of ULP charges filed overall and for firings in particular. Additionally, these estimates likely understate the full extent of anti-union activity because workers do not always file charges with the NLRB. In this section, we first explain how we use the data to calculate ULP charge rates associated with elections. Then we review our findings and describe how they compare with previous research.
To quantify employer interference during union election campaigns, we match NLRB election filings data to NLRB data on ULP charges. The election filings data is the complete set of 3,620 NLRB elections that either were filed for or occurred in calendar year 2016 or 2017.7 We chose not to examine election filings data for prior years (2015 or earlier), because new rules governing the election process took effect partway through 2015 (see NLRB 2015).8
About two-thirds (69.4%) of the 2016–2017 election filings data we analyzed were for elections that had actually been held by the time we collected the data in 2018. The remaining third (30.6%) were election petitions that were ultimately withdrawn or petitions for elections that had not yet occurred when we collected the data. For simplicity, and because we are interested in both attempted and completed elections, we refer to the complete set of all petitions and completed elections as “elections” throughout this report.
We matched the calendar year 2016–2017 election filings data to the complete universe of 49,396 cases of ULP charges filed against employers between fiscal year 2015 and fiscal year 2018 (October 1, 2014–September 30, 2018). We include ULP charges prior to and after the election filings data because ULP charges may be filed in the lead-up to or aftermath of an election. This time period excludes election-related ULP cases filed after September 2018, but it was the most recent data available at the time of the FOIA requests.9
The NLRB does not systematically track which ULP charges occur in the context of an election campaign. However, both the data on elections and the separate data on ULP charges contain the employer name, city, and state. We assume that any election and ULP charge sharing all three of these values are related. Since our ULP sample covers a brief time period, we are confident in our assumption that the actions alleged in these cases would have been in reaction to or would have affected any organizing effort at that employer. (See the methodological appendix for details about the matching process.)
From the matched election and ULP data, we estimate that 41.5% of all elections experienced a ULP allegation (see Figure A). In other words, four out of ten elections to form a union involved employer behavior that generated ULP charges. Figure A further disaggregates these charges into specific alleged violations of the NLRA, according to the relevant NLRA 8(a) sections described above. The major categories of election-related ULP charges against employers were 8(a)(1) charges for threats, coercion, retaliation; 8(a)(3) charges for firings, unlawful disciplinary action, or changing terms of work; and 8(a)(5) charges for refusing to bargain in good faith. The shares of elections with an 8(a)(1), (3), or (5) charge ranged from 27.0% to 29.3%. (More detailed allegation information is available in Appendix Table 1.)
|ULP type||Share of elections|
|8(a)(1): Threatened, coerced, retaliated||29.2%|
|8(a)(3): Disciplined, fired, changed work terms||29.3%|
|8(a)(5): Refused to bargain, repudiated contract||27.0%|
Notes: ULP charges are charges that an employer violated Section 8(a) of the labor code by interfering with workers’ rights to form a union and bargain collectively. Specific charge types 8(a)(1), 8(a)(3), and 8(a)(5) refer to sections of the labor code governing these rights. “Any charge” refers to any violation of Section 8(a) of the labor code (parts 1–5).
Source: Authors’ analysis of National Labor Relations Board election data for calendar years 2016–2017 and ULP filings from fiscal years 2015–2018
While these rates may seem high, it is also important to recognize they likely understate the extent of employer aggression against unions, as they cannot capture the full extent of all illegal or coercive behavior by employers in opposing worker organizing efforts. For example, many union organizing efforts are thwarted by employers before making it to the filing stage. In addition, many anti-union violations go unreported. The NLRB processes for resolving violations often involve unnecessary delay as the result of employer manipulation of the legal process. Further, the NLRA provides weak remedies when employers are found to have violated the law. The best estimates suggest that the incidence of underreporting may be very large: Bronfenbrenner (2009) finds that whereas 30% to 40% of elections had a ULP charge filed with the NLRB, a detailed survey revealed that unions claimed employers committed ULPs in 89% of elections.
For elections that were completed,10 the number of eligible voters or the size of the potential bargaining unit varied across elections. Figure B shows how the ULP charge rates vary by unit size for completed elections. Each of the size groupings in Figure B represents roughly one-quarter of the total number of units.11 In roughly the bottom half of the size distribution, where there were 25 or fewer eligible voters, 36.4% of completed elections had a ULP charge. In roughly the top quarter of eligible voter sizes, where there are more than 60 potential unit members, completed elections are much more likely to experience ULP charges (54.4% of elections).
|Bargaining unit size||Share of elections|
|10 or fewer employees||35.5%|
Notes:?ULP charges are charges that an employer violated the terms of the National Labor Relations Act (Section 8(a) of the labor code) by interfering with workers’ rights to form a union and bargain collectively. Data by bargaining unit size are available for only for those elections that were completed by the time we obtained the data in 2018.
Source: Authors’ analysis of National Labor Relations Board election data for calendar years 2016–2017 and ULP filings from fiscal years 2015–2018
One of the most severe forms of employer aggression and intimidation is firing or discharging workers for union activity. Because of the aggregated nature of NLRB charge classifications in our data, it is difficult to determine whether a given allegation type implies that a worker was fired. A narrow definition of firings includes only the subset of 8(a)(3) allegations classified as “Discharge (including layoff and refusal to hire (not salting)).” Because this category likely fails to account for all election-related discharges and similar retaliation, we also calculate a second estimate, based on a broader definition of firings that includes two other subcategories of charges: 8(a)(1) “Concerted activities (retaliation, discharge, discipline); and 8(a)(3) and 8(a)(4) “Changes in terms and conditions of employment.”?(See Appendix Table 1 for detailed charge categories.)
Using these narrower and broader definitions, Figure C shows our best range of estimates for the share of union elections with a ULP charge for firing. A disturbingly high share of elections involved an allegation that an employer fired a worker for union activity: Between 19.9% and 29.6% of all union elections in our sample were associated with a firing charge, depending on whether the narrower or broader definition of firing charges is used. For larger potential bargaining units, the shares were even larger. When the potential bargaining unit was more than 60 employees, between 27.2% and 41.3% of completed12 union elections were associated with a ULP for firing or discharging a worker.
|Firing, narrowly defined||Firing, broadly defined|
|All sizes (all elections)||19.9%||9.8%|
|61+ employees (completed elections only)||27.2%||14.1%|
Notes: “All sizes” estimates are based on all election data from calendar years 2016–2017, whether the elections were already completed or not by the time we obtained the data, while “61+ employees” estimates are based on data for elections that had been completed by the time we obtained the data in 2018 (data by bargaining unit size are available only for completed elections). “Narrowly defined” refers only to unfair labor practice charges in the category “Discharge (including layoff and refusal to hire (not salting))” in Section 8(a)(3) of the labor code. “Broadly defined” also includes ULP charges in the categories “Concerted activities (retaliation, discharge, discipline)” (from Section 8(a)(1)) and “Changes in terms and conditions of employment” (from Sections 8(a)(3) and 8(a)(4)).
Source: Authors’ analysis of National Labor Relations Board election data for calendar years 2016–2017 and ULP filings from fiscal years 2015–2018
Our findings build on existing research into employer opposition to organizing efforts. Figure D presents our estimates alongside estimates from earlier studies. Our estimates confirm earlier evidence that employers regularly engage in aggressive anti-union activity—and suggest that the problem may be getting worse.
The first three bars in Figure D compare the findings of our study regarding overall ULP charge rates with the findings of Bronfenbrenner’s 2009 study. In that study, Bronfenbrenner compiles a sample of 1,004 NLRB elections between 1999 and 2003; the units represented by these elections contained at least 50 potential unit members. Our finding that 41.5% of union elections (for units of all sizes) had associated ULP charges is similar to Bronfenbrenner’s high-end estimate of 40% and about 10 percentage points higher than her low-end estimate of 30%. When we restrict our sample to units of at least 50 employees to make it more comparable to Bronfenbrenner’s sample, we find an even higher rate of ULP charges during elections: 54.2%. Our more recent data from 2016 and 2017 therefore suggest that ULP activity during elections is at least as pervasive—and likely more pervasive—than it was approximately 15 years ago.
|EPI, NLRB, all sizes||41.5%|
|EPI, NLRB, 50+||54.2%|
|Bronfenbrenner, NLRB, 50+||30.0%||10.0%|
|EPI, NLRB, all sizes||19.9%||9.8%|
|EPI, NLRB, 50+||27.2%||13.4%|
|Bronfenbrenner, NLRB, 50+||17.0%|
|Bronfenbrenner, survey, 50+||34.0%|
|Mehta & Theodore, survey||30.0%|
|Schmitt & Zipperer, NLRB||26.0%|
Notes: The rows “EPI, NLRB, all sizes” and “EPI, NLRB, 50+” refer to our current analysis of National Labor Relations Board (NLRB) data from union elections for which a petition was filed or the election was completed in 2016–2017, for?potential bargaining units of any size (“all sizes”), and for larger units of 50 or more employees (“50+”). For the latter group, data are available only for those elections that had been completed by the time we obtained the data in 2018. For each group, we estimate a range based on broad and narrow definitions of “firing.” “Bronfenbrenner, NLRB, 50+” refers to Bronfenbrenner’s 2009 analysis of NLRB data for larger units (“50+”) that had elections in 1999–2003, which includes two estimates, while “Bronfenbrenner, survey, 50+” refers to her analysis of survey data for these larger units. Estimates from Mehta and Theodore’s 2005 analysis of survey data for elections that were petitioned for in 2002 and Schmitt and Zipperer’s 2009 analysis of 2001–2007 NLRB data are also shown.
Sources: Authors’ analysis of National Labor Relations Board (NLRB) election data for calendar years 2016–2017 and ULP filings from fiscal years 2015–2018; Bronfenbrenner, No Holds Barred–The Intensification of Employer Opposition to Organizing (2009); Mehta and Theodore, Undermining the Right to Organize: Employer Behavior During Union Representation Campaigns (2005); Schmitt and Zipperer, Dropping the Ax: Illegal Firings During Union Election Campaigns, 1951–2007 (2009)
The second set of bars in Figure D focuses specifically on illegal firings charges associated with union elections, with our findings presented alongside earlier findings by Bronfenbrenner (2009), Mehta and Theodore (2005), and Schmitt and Zipperer (2009). Our findings are strictly based on NLRB data, as described above. Bronfenbrenner uses NLRB election data between 1999 and 2003 (the sample of 1,004 elections noted above), but also uses survey data from 562 of those campaigns to gather a more complete picture of employer aggression. Bronfenbrenner’s analysis of NLRB data shows that there were illegal firing charges in 17.0% of the elections, while her survey data indicate that illegal firings occurred in 34.0% of the elections. Mehta and Theodore (2005) find a similar estimate (30.0%) in their survey of 62 union representation campaigns that began in 2002 in Chicago. Schmitt and Zipperer (2009) use aggregate data on all elections and discharges from NLRB annual reports to extend the analysis of LaLonde and Meltzer (1991) and find that employers discharged workers for union activity in 26.0% of NLRB elections between 2001 and 2007.
Compared with previous findings, our analysis of 2016–2017 NLRB elections finds similar rates of illegal firing charges. The estimates in this paper find rates between 19.9% and 40.6%, depending on the definition of the charge and the size of the potential unit, and the previous studies find rates between 17.0% and 34.0%. Our findings are therefore consistent with earlier findings that employer aggression against workers is a prominent feature of unionization campaigns.
Over the past few decades, employers’ attempts to thwart organizing have become more prevalent, with more employers turning to the scorched-earth tactics of “union avoidance” consultants. Bronfenbrenner (2009) shows that, by the early 2000s, three-quarters of all employers involved in union elections with 50 or more voters hired union avoidance consultants.13 Based on the most recent available data, we estimate that employers are now spending nearly $340 million per year on such consultants.14 This work is well compensated—consultants often report being paid $350-plus hourly rates or $2,500-plus daily rates for their work defeating union organizing efforts.15
The main goal of union avoidance consulting firms is to prevent a union election from taking place—and if that fails, to ensure that workers vote against the union. The firm Sparta Solutions—a prominent player in the industry—urges employers to “let SPARTA show you how not only to win your election, but also teach your staff advanced techniques for union avoidance to ensure your company never goes through a union election again.”16 Labor Relations Institute (LRI), another of the nation’s largest union avoidance firms, was contracted by an employer to “prevent NLRB [union election] petitions from being filed at all company locations.”17 Indeed, such strategies are often explicitly written into the financial terms of consultant contracts. A 2016 contract between Sparta Solutions and an employer states that Sparta will receive a $25,000 bonus if it succeeds in getting the workers’ petition withdrawn and thereby preventing an election.18 LRI offers employers standard contract language stipulating that 50% of the firm’s fee is contingent on a management “win,” defined as “withdrawal of the petition or a win at the ballot box.”19
Over the past five years, employers using union avoidance consultants have included FedEx, Bed Bath & Beyond, and LabCorp, among others.20 Table 1 lists just a few of these employers, along with the reported financial investments they made to thwart union organizing during the specified years.21 Each of these firms employs thousands of American workers.
|Laboratory Corporation of America||$4,300,000||2014–2018|
|Albert Einstein Medical Center||$1,100,000||2014–2017|
|Simmons Bedding Co.||$848,000||2015–2017|
|Trump International Hotel Las Vegas||$569,000||2015–2016|
|Bed Bath & Beyond||$506,000||2014, 2018|
|J.B. Hunt Transport||$354,000||2016–2018|
|Hilton Grand Vacations||$340,000||2014–2015|
|Archer Daniels Midland||$324,000||2016–2017|
|Robert Wood Johnson University Hospital||$316,000||2014–2016|
|Associated Grocers of New England||$190,000||2014–2017|
|Pier 1 Imports||$169,000||2015–2016|
Source: Lafer and Loustaunau’s analysis of LM-20 and LM-21 forms filed by consultants with the U.S. Department of Labor (DOL) Office of Labor-Management Standards (OLMS), 2014–2018
Current labor law’s ineffectiveness in addressing employer opposition to unions has created a situation where fewer and fewer workers are able to organize and collectively bargain. Further, in recent decades, union suppression and avoidance has become a primary goal of corporate interests and a focus of Republican lawmakers. As a result, workers’ rights have been attacked through legislation, executive rulemaking, and the judiciary. Indeed, corporations have seen their interests advanced by policymakers, while working people’s interests have been largely neglected. Instead of advancing reforms that would rebalance an increasingly rigged system, lawmakers have repeatedly failed to prioritize measures that advance workers’ rights.
The result of this political negligence is evidenced by the extreme inequality that marks the American economy—the highest ever in U.S. history, according to Census Bureau data.22 Chief executive officer compensation has grown 940% since 1978, while typical worker compensation has risen only 12% during that time (Mishel and Wolfe 2019). From 1979 to 2016, the wages of the top 1% grew nearly 150%, whereas the wages of the bottom 90% combined grew just 21.3%, roughly one-seventh as fast (Mishel and Wolfe 2018).?Even today’s very low unemployment rate has not been enough to spur truly robust wage growth for most workers.
The effects of declining union representation are not only economic. The last few decades have seen an erosion of basic workplace protections. The federal minimum wage has not been raised for over 10 years—the longest period in history without an increase (Cooper, Gould, and Zipperer 2019). Fewer workers have access to overtime protections (Shierholz 2019). Workers are increasingly less safe on the job (Zoorob 2018). Workers are increasingly subject to forced arbitration clauses that require them to resolve workplace disputes in a process that favors the employer (Hamaji et al. 2019).
However, workers are demanding change. From the teachers’ strikes in several states to the General Motors strike to the Google workers’ walkout, workers have demonstrated the power of collective action and their willingness to challenge a legal framework that no longer lives up to its promise of promoting and encouraging the practice of collective bargaining.
Policymakers are following their lead and have introduced legislation, the Protecting the Right to Organize (PRO) Act, that would significantly reform current labor law. Some of the most damaging tactics employers frequently engage in to oppose workers’ union organizing efforts would be restricted under the legislation, and meaningful penalties would be imposed when employers violate the law. Table 2 outlines the most significant reforms of the PRO Act that address employer opposition to unions. Table 3 describes additional important reforms included in the legislation in response to deficiencies and loopholes in the current law.
The PRO Act will help ensure that workers have a meaningful right to organize and bargain collectively by streamlining the process when workers form a union, bolstering workers’ chances of success at negotiating a first agreement, and holding employers accountable when they violate the law. Indeed, the PRO Act addresses many of the tactics of aggressive employer opposition. This type of legislative reform is needed to restore workers’ rights to join together and bargain for a better life. However, policymakers must do more. They must prioritize a workers’ rights agenda and hold agencies responsible for enforcing worker protections accountable.
|Employer coercive activity||Share of union elections with ULP charge||Penalty under the PRO Act|
|Firing workers for union activity||19.9%||The PRO Act imposes a civil penalty up to $50,000 per incident for illegal firings in retaliation for union activity; the penalty may be doubled (up to $100,000) if it is a repeat violation, that is, if the employer was previously found to have committed a violation causing economic harm to a worker at any time in the prior five years.
In addition, any worker who is illegally fired for engaging in union activity shall be awarded “back pay without any reduction (including any reduction based on the employee’s interim earnings or failure to earn interim earnings), front pay (when appropriate), consequential damages, and an additional amount as liquidated damages equal to two times the amount of damages awarded.”
|Threatening to cut benefits and wages||18.2%||The PRO Act imposes a civil penalty up to $50,000 per incident; the penalty may be doubled (up to $100,000) for repeat violations causing economic harm to workers.|
|Spying on workers engaged in union activities or creating the impression of spying||13.9%||The PRO Act imposes a civil penalty up to $50,000 per incident; the penalty may be doubled (up to $100,000) for repeat violations causing economic harm to workers.|
|Questioning workers about union activity or support||7.3%||The PRO Act imposes a civil penalty up to $50,000 per incident; the penalty may be doubled (up to $100,000) for repeat violations causing economic harm to workers.|
|Refusal to bargain in good faith||18.6%|
Notes: Second column refers to the share of all union elections in our analysis that had an unfair labor practice (ULP) charge filed with the National Labor Relations Board (NLRB) alleging the specified violation.
Source: Authors’ analysis of the Protecting the Right to Organize Act of 2019, H.R. 2474, 116th Cong. (2019) (for penalties); and of National Labor Relations Board election data for calendar years 2016–2017 and ULP filings from fiscal years 2015–2018 (for ULP charges filed)
|Deficiency in current labor law||Policy reform under the PRO Act|
|Allows employers to hold captive audience meetings.||Prohibits employers from forcing workers to attend captive audience meetings.|
|No civil penalties for employers who violate workers’ rights under the National Labor Relations Act (NLRA).||Creates civil penalties for violations under the NLRA.|
|Allows employers to misclassify workers as independent contractors without violating the NLRA.||Institutes an “ABC” test to determine employee status and makes employee misclassification a violation under the NLRA.|
|Allows multiple employers to dictate workers’ terms of employment while also refusing to recognize the workers as their employees.||Codifies a strong joint employer standard.|
|Allows employers to permanently replace workers who go on strike.||Prohibits employers from permanently replacing striking workers.|
|Allows employers to lock out workers, prior to a strike, to influence the collective bargaining process.||Bans the use of offensive lockouts.|
|Prohibits workers from engaging in picketing or striking in solidarity with another company’s workers.||Removes prohibitions on secondary strikes.|
|Does not require employers to inform employees of their rights under the NLRA.||Requires employers to post a notice of workers’ rights under the NLRA. Failure to post notice results in a civil penalty up to $500 for each violation.|
|Allows employers to withhold or fail to provide accurate lists of eligible voters to the bargaining unit.||Requires employers to provide the bargaining unit a credible list of eligible voters in an election within two business days. Failure to provide list in a timely manner results in a civil penalty up to $500 for each violation.|
|Allows workers to vote in union elections only by certified mail, at the worksite, or off the worksite.||Allows workers to vote electronically in union elections in addition to certified mail, at the worksite, or off the worksite.|
|Allows employers to force workers to sign arbitration agreements that waive the right to collective or class action litigation.||Bans the use of collective and class action waivers.|
|Prevents workers from bringing civil lawsuits against their employer.||Provides workers a private right to civil action.|
Source: Authors’ analysis of current labor law and the Protecting the Right to Organize Act of 2019, H.R. 2474, 116th Cong. (2019)
The authors would like to thank our summer interns, Karla Martinez and Carl Romer, for their diligent and meticulous work. Without them, we might still be hand-matching the ULP and election filings!
We hand-matched the complete set of 3,620 NLRB election filings from calendar years 2016 and 2017 to all 49,396 cases of ULP charges filed against employers between fiscal year 2015 and fiscal year 2018. We assumed that any election and ULP charge were related if they shared an employer name, city, and state. To our knowledge the NLRB does not systematically track which ULP charges occur in the context of an election campaign, and it was cost- and time-prohibitive to FOIA-request and manual review in each ULP charge filing.23 The main text describes our results and Appendix Table 1 shows additional detailed results by charge type.
|Charge type||Share of elections with the specified charge|
|Overall (any ULP charge)||41.5%|
|Illegal firing, narrowly defined||19.9%|
|Illegal firing, broadly defined||29.6%|
|Coercive statements (threats, promises of benefits, etc.)||18.2%|
|Concerted activities (retaliation, discharge, discipline)||16.2%|
|Coercive actions (surveillance, etc.)||13.9%|
|Interrogation (including polling)||7.3%|
|Other 8(a)(1) allegations||5.8%|
|Changes in terms and conditions of employment||17.3%|
|Discharge (including layoff and refusal to hire (not salting))||19.9%|
|Other 8(a)(3) allegations||2.1%|
|Refusal to bargain/bad faith bargaining including surface bargaining/direct dealing||18.6%|
|Repudiation/modification of contract (Section 8(d)/unilateral changes)||18.5%|
|Refusal to furnish information||12.5%|
|Other 8(a)(5) allegations||5.6%|
Notes: ULP charges are charges that an employer violated Section 8(a) of the labor code by interfering with workers’ rights to form a union and bargain collectively. Specific charge types 8(a)(1), 8(a)(2), etc.,?refer to sections of the labor code governing these rights. “Overall (any ULP charge)” refers to any violation of?Section 8(a) of the labor code (parts 1–5). “Narrowly defined” illegal firings refers only to ULP charges in the category “Discharge (including layoff and refusal to hire (not salting))” from 8(a)(3). “Broadly defined” also includes ULP charges in the categories “Concerted activities (retaliation, discharge, discipline)” from 8(a)(1) and “Changes in terms and conditions of employment” from 8(a)(3) and 8(a)(4).
Source: Authors’ analysis of National Labor Relations Board election data for calendar years 2016–2017 and ULP?filings from fiscal years 2015–2018
To begin the hand-matching, we used Google’s Open Refine tool24 to mitigate any misspellings or inconsistencies in city names (for example, “St. Louis” versus “Saint Louis”). The refining tool is imperfect and will have left a few inconsistencies unchecked, leading to an undercount of matches. We noticed that many establishments located in the New York City boroughs had their city listed as “New York City” some of the time and as the borough other times. To ensure that we did not miss a large number of potential matches, we considered the city for all boroughs to be “New York City.” This could potentially result in an overcount of matches, but we saw enough clear-cut cases of this inconsistency to convince us that it was worth correcting. Establishments in other large cities, such as Los Angeles or Chicago, did not seem to have this problem to the same extent, so we corrected for it only in New York City.
Using these refined city names, we generated a list of potential matches for each election consisting of all ULP charges in our sample that occurred in the same city and state as the election. We then linked election filings to potential ULP matches based on employer name. Since there were inconsistencies in recording of employer names that could not be entirely overcome using matching tools, we reviewed these lists of potential matches by hand. For example, the same employer may be listed alternately as “United Parcel Service,” “UPS,” or “UPS Inc.” The potential matches for each election were reviewed by at least two people, and we additionally reviewed and resolved any cases where the two reviewers disagreed. If the employer names were the same for an election and any ULP charges within the same city and state, we considered them to be a match.
Our analysis of the union avoidance industry is built on an examination of publicly available forms LM-20 and LM-21 filed with the U.S. Department of Labor’s Office of Labor-Management Standards (OLMS). We downloaded all LM-20s and LM-21s for the years in question via OLMS’s Online Public Disclosure Room.25
LM-20s are filed by consultants for each individual employer they contract with. These forms identify the employer worked for, the consultants working on the project, and the rate of pay and terms of consultant employment, but not the total amount received for the contract. Our analysis included all LM-20s filed between January 1, 2015, and December 31, 2018. In addition, we also examined those LM-20s filed in 2012–2014 that had actual employer-consultant contracts attached to the LM-20 filing. We used these forms to gather background information on consultant rates and contract terms, as well as to help us cross-check information in the LM-21 filings, as described below.
To develop our overall estimate of the annual value of the union avoidance industry, we began by looking at LM-21 filings. LM-21s are filed once a year by consultant firms, identifying all clients and subcontractors, with the total amounts paid by each employer and for each subcontractor. We looked at all LM-21s filed between January 1, 2014, and December 31, 2018, to collect data on the total industry-wide amounts received for work performed in 2014–2017. For the purposes of looking at industry-wide totals, we did not look at data on amounts received for work performed in 2018; the 2018 data would be incomplete as of December 31, 2018, and we needed full years of data to estimate an annual average for union avoidance industry earnings. (However, in constructing Table 1, we did use all available data including the partial data for work performed in 2018, to calculate the amounts spent by specific employers over the five-year period from 2014 to 2018.)
Estimating the size of the industry of union-busting consultants and attorneys has long been an elusive matter, both because much work is not reported and because reports are often filed with incomplete information. Because reporting is so haphazard, it is not unusual for contractors and subcontractors on a given campaign to each independently submit their own LM-21, which might easily result in double-counting and producing an inflated estimate of the total money spent. For instance, if an employer hires Consultant A for $100,000, and Consultant A subcontracts with Consultant B for $50,000, and each of these firms independently files an LM-21, it might appear that a total of $150,000 had been spent, when the actual total was only $100,000. To avoid potential double-counting, we match each LM-21 with the employer, year, and location at which the work was performed (cross-checking with LM-20 forms as needed), identifying all LM-21s that were part of a single campaign.
Totaling up all the reported campaigns, we find that employers spent $95.7 million on anti-union campaigns in the years 2014–2017. In addition, there is $9.5 million in consultant earnings reported during these years that, because of insufficient information, we could not identify as being part of a known campaign. Since there is no way to control for potential double-counting in these reports, we have conservatively estimated actual total spending for these reports at half that reported, i.e., $4.75 million. Adding this to the campaigns total shows that employers spent a reported total of just over $100 million on anti-union consultants in 2014–2017, or an average of just over $25 million per year.
However, the total reported in LM-21s is only a small part of total industry activity. In part, this is because the U.S. Department of Labor requires reporting only by those engaged in direct communication with employees, meaning that a huge number of law firms and consultants who craft and oversee anti-union campaigns by training supervisors to deliver scripted messages to their subordinates—but do not themselves talk directly to workers—have no obligation to report their work. But even those consultants who are legally required to report their work face no significant penalty for ignoring this requirement. Those who fail to file reports are not fined; the Department of Labor simply reminds them once again to comply.26 As long ago as 1980, the U.S. House of Representatives concluded that consultant reporting under the Labor-Management Reporting and Disclosure Act was a “virtual dead letter, ignored by employers and consultants and unenforced by the Department of Labor.”27
The most comprehensive review of research on the anti-union industry was conducted by the U.S. Department of Labor in 2011.28 Based on that research, Labor Department analysts estimated that the amount of work actually reported by consultants (as measured by LM-20 forms that are theoretically required to be filed each time a consultant contracts to work for a given employer) represented only 7.4% of the work actually performed. In the absence of any more precise measure, we assumed that the share of all work reported on LM-21s (showing the dollar value of work performed) was the same as that measured by the DOL using LM-20s. To calculate the total value of this industry, we assume that the volume of business contained in LM-21 reports over this period—averaging $25 million per year—represents 7.4% of the actual overall volume. Thus, the total annual value of the union avoidance consultant industry would be $25 million divided by 0.074, or $338 million.
There is good reason to believe that this is an underestimate, because the share of work reported on LM-21s may be even smaller than 7.4%, following the Department of Labor OLMS’s 2016 declaration that it would no longer enforce consultants’ requirement to report financial receipts or disbursements related to anti-union work.29 Following this declaration, the number of LM-21s filed fell substantially—much more than would be expected given the decline in NLRB union elections. In 2017, the number of NLRB union elections was 8% lower than in 2016, but the number of LM-21s filed in this year was 38% lower than in 2016. The difference likely largely reflects the impact of OLMS’s 2016 declaration noted above. Even a small decrease in the share of work reported on LM-21s would increase our estimates substantially.
Click the links below to see examples of LM-20 and LM-21 forms filed by consultants.
1. See Zoorob 2018 for how unions create safer workplaces. See Bivens et al. 2017 for other information about unions.
2. ULP charges filed with the NLRB are investigated by agency staff to determine if the charge has merit. If the charge is found to have merit, the agency will pursue a complaint against the employer. The filing of a charge itself is not a determination of a violation of the National Labor Relations Act.
3. This represents charges that employers violated Section 8(a)(1) of the National Labor Relations Act.
4. This represents charges that employers violated Section 8(a)(3) of the National Labor Relations Act.
5. When private-sector workers decide to join together in a union, they can ask their employers to voluntarily recognize that union, or they can file a petition for a union election with the NLRB. Workers can petition the NLRB for a union election when they can demonstrate they have the support of at least 30% of the employees in the potential bargaining unit. If a majority of the potential bargaining unit votes for union representation, then the NLRA requires that the employer recognize that union and bargain a first contract in good faith (NLRB 2015).
6. See Gold?2009.
7. Election filings data for calendar years 2016–2017 were obtained from the NLRB through FOIA requests NLRB-2018-001366 (submitted 9/26/2018, completed 10/26/2018) and NLRB-2019-000178 (submitted 11/28/2018, completed 10/26/2018).
8. Representation—Case Procedures; Final Rule, 79 Fed. Reg. 74308–74489 (December 15, 2014).
9. ULP filings for fiscal years 2015–2018 were obtained from the NLRB through FOIA requests NLRB-2018-001322 (submitted 9/13/2018, records released 10/17/2018) and NLRB-2019-000491 (submitted 2/21/19, records released 4/15/2019).
10. Note that data on the prevalence of ULPs by bargaining unit size are available for completed elections only, not for petitions. Completed elections include all elections that had been completed by the time we obtained the data in 2018.
11. The 25th-percentile unit size is 10 employees; the 50th-percentile unit size is 24 employees; and the 75th-percentile unit size is 59 employees.
12. As noted above, data by bargaining unit size is available only for elections that had been completed by the time we obtained the data in 2018.
13. Bronfenbrenner analyzed NLRB election data from 1999–2003.
14. Estimate based on Gordon Lafer and Lola Loustaunau’s analysis of LM-20 and LM-21 reports filed with the U.S. Department of Labor Office of Labor-Management Standards between 2014 and 2018 and the findings of a 2011 DOL report on underreporting by consultants. See the methodological appendix for more details.
15. Gordon Lafer and Lola Loustaunau’s analysis of LM-20 reports filed with the U.S. Department of Labor Office of Labor-Management Standards between 2015 and 2018. A total of 2,139 LM-20 reports were filed during this period, of which 773 reported either hourly or daily pay rates. (The remainder reported total contract amounts, without breaking down hourly or daily pay rates.)
16. “About Sparta Solutions: Core Values” (web page), accessed November 11, 2019.
17. Quoted in von Wilpert?2017.
18. Sparta’s 2016 contract with Biery Cheese stipulates that “the fee for a day rate for 4 consultant [sic] is $375 per hour per calendar day worked by each Consultant totaling $3000 a day per consultant × 10 days plus travel expenses with a 50% Guarantee at risk. There will be a[n] additional $25,000 withdrawl [sic]?bonus.” The 50% “guarantee at risk” means that the firm is only paid half the total amount if the union wins. And if the petition is withdrawn, they are paid their full fee plus $25k. Contract terms are reported in LM20_C66578_11_22_2016_631604, accessed via the DOL OLMS website.
19. From contract terms reported in LM20_C525_01_28_2013_525631 (accessed via the DOL OLMS website). LRI’s contract stipulates that the firm will be paid $155,000 plus expenses, of which “$75,000 of the fee is an incentive fee guaranteeing an election win…defined as a withdrawal of the petition or a win at the ballot box.” “Should your company lose the NLRB election,” LRI promises, “you will not owe the remainder of the project price.” This is standard language offered by the firm to employers. A 2012 LRI contract explains that “for the ‘Partially Guaranteed Option’ we require a 50% retainer due upon acceptance of the proposal. We will apply that retainer to the project price. Based on the vote count and in the event of a ‘win’ you agree to pay the balance of the project price within 7 days of the NLRB election. Should your company lose the NLRB election, your company will not owe the balance of the project price” (from LM20_C525_9_30_2012_507012).
20. Firms recorded in LM-20 or LM-21 reports filed between 2014 and 2018, accessed via the DOL OLMS website.
21. It is worth noting that, due to loopholes in reporting requirements, these amounts may represent only a fraction of these companies’ anti-union investments; see the methodological appendix for more information.
22. United States Census Bureau, “American Community Survey Provides New State and Local Income, Poverty and Health Insurance Statistics” (press release), September 26, 2019.
23. While the NLRB does track “related cases” in their internal database (which is where we requested information from through FOIA), the data are used only for internal tracking during the investigation and proceedings. What is considered to be a “related” case varies from region to region and even from case to case, so it would not be at all useful to rely on that for our purposes.
24. See https://openrefine.org.
25. OLMS’s Online Public Disclosure Room is at https://www.dol.gov/olms/regs/compliance/rrlo/lmrda.htm.
26. John Lund, former Assistant Secretary of Labor for Labor-Management Standards, conversation with Gordon Lafer, July 2, 2019.
27. U.S. House of Representatives, Committee on Education and Labor, Subcommittee on Labor-Management Relations, Pressures in Today’s Workplace, 1980, quoted in Notice of Proposed Rulemaking: Labor-Management Reporting and Disclosure Act; Interpretation of “Advice” Exemption, 76 Fed. Reg. 36177 (June 21, 2011).
28. Notice of Proposed Rulemaking: Labor-Management Reporting and Disclosure Act; Interpretation of “Advice” Exemption, 76 Fed. Reg. 36177 (June 21, 2011).
29. U.S. Department of Labor, Office of Labor-Management Standards, “Form LM-21 Special Enforcement Policy” (web page), posted April 13, 2016; updated July 18, 2018.
Bivens,?Josh, Lora Engdahl,?Elise Gould,?Teresa Kroeger,?Celine McNicholas,?Lawrence Mishel,?Zane Mokhiber,?Heidi Shierholz,?Marni von Wilpert,?Valerie Wilson, and?Ben Zipperer. 2017.?How Today’s Unions Help Working People: Giving Workers the Power to Improve Their Jobs and Unrig the Economy. Economic Policy Institute, August 2017.
Bronfenbrenner, Kate. 2009. No Holds Barred–The Intensification of Employer Opposition to Organizing. Economic Policy Institute and the American Rights at Work Education Fund, May 2009.
Bureau of Labor Statistics (BLS). 2019. “Union Members Summary” (news release). January 18, 2019.
Cooper, David, Elise Gould, and Ben Zipperer. 2019. Low-Wage Workers Are Suffering from a Decline in the Real Value of the Federal Minimum Wage. Economic Policy Institute, August 2019.
Gold, Wayne. 2009. “NLRB Remedies Today.” Paper prepared for the American Bar Association Section of Labor and Employment Law Third Annual CLE Conference, September 2009.
Hamaji, Kate,?Rachel Deutsch,?Elizabeth Nicolas,?Celine McNicholas,?Heidi Shierholz, and?Margaret Poydock. 2019. Unchecked Corporate Power: Forced Arbitration, the Enforcement Crisis, and How Workers Are Fighting Back. Economic Policy Institute and the Center for Popular Democracy, May 2019.
Kochan, Thomas, William T. Kimball, Duanyi Yang, and Erin L. Kelly. 2018. “Voice Gaps at Work, Options for Closing Them, and Challenges for Future Actions and Research.” MIT Sloan School of Management, Institute for Work and Employment Research working paper, June 2018.
LaLonde, Robert J., and Bernard D. Meltzer. 1991. “Hard Times for Unions: Another Look at the Significance of Employer Illegalities.” University of Chicago Law Review?58, no. 3: 953–1014.
Mehta, Chirag, and Nik Theodore. 2005. Undermining the Right to Organize: Employer Behavior During Union Representation Campaigns. A Report for American Rights at Work, December 2005.
Mishel, Lawrence, and Julia Wolfe. 2018.?“Top 1.0 Percent Reaches Highest Wages Ever—Up 157 Percent Since 1979.”?Working Economics?(Economics Policy Institute blog), October 18, 2018.
Mishel, Lawrence, and Julia Wolfe. 2019.?CEO Compensation Has Grown 940% Since 1978: Typical Worker Compensation Has Risen Only 12% During That Time. Economic Policy Institute, August 2019.
National Labor Relations Board (NLRB). 1997. Basic Guide to the National Labor Relations Act: General Principles of Law Under the Statute and Procedures of the National Labor Relations Board. Washington, D.C.: U.S. Government Printing Office, 1997.
National Labor Relations Board (NLRB). 2015. “What We Do: Conduct Elections” (web page). Description of the election petition process effective April 14, 2015.
Schmitt, John, and Ben Zipperer. 2009. Dropping the Ax: Illegal Firings During Union Election Campaigns, 1951–2007. Center for Economic and Policy Research, March 2009.
Shierholz, Heidi. 2019. “More Than Eight Million Workers Will Be Left Behind by the Trump Overtime Rule.” Working Economics?(Economic Policy Institute blog), September 25, 2019.
Von Wilpert, Marni. 2017. “Union Busters Are More Prevalent Than They Seem, and May Soon Even Be at the NLRB.” Working Economics Blog (Economic Policy Institute), May 1, 2017.
Zoorob, Michael. 2018. “Does ‘Right to Work’ Imperil the Right to Health? The Effect of Labour Unions on Workplace Fatalities.” Occupational & Environmental Medicine 75, no. 10 (June 2018): 736–738, http://dx.doi.org/10.1136/oemed-2017-104747.