Union Organizing

By Budd, J.W.

Edited by Paul Ducham


Most union organizing drives—that is, campaigns to organize nonunion workers into unions to gain recognition of the unions by their employers—follow a common sequence of steps (see Figure 6.1 ).

Initiating an Organizing Drive

The first step is initiation. In theory, there are three possible initiators of an organizing drive: one or more employees, a union, or an employer. The last possibility might seem odd, but as an example, in the 1960s and 1970s some agricultural companies signed contracts with the Teamsters rather than risk having their employees choose the more militant United Farm Workers. Note that this interferes with employee free choice, and therefore employer-initiated organizing drives are illegal [Section 8(a)(1) unfair labor practices under the NLRA]. Thus our concern is with employee- or union-initiated campaigns.

An employee-initiated organizing drive is perhaps the classic scenario. Low pay, excessive overtime, harsh supervision, lack of respect and voice, or numerous other factors cause dissatisfied employees to talk with each other about forming a union to increase their collective strength. These employees might try to form a union on their own or, more frequently, contact a union organizer who works for an existing union in their area, industry, or occupation. In either case this is an employee-initiated campaign.

Alternatively, unions may initiate organizing campaigns by advertising, distributing information, and trying to contact employees to show the benefits of Unionization. There are two types of union-initiated campaigns: strategic and opportunistic. Strategic campaigns are those in which organizing a particular workplace will enhance a union’s ability to effectively represent existing employees. For example, the United Auto Workers (UAW) frequently tries to organize the nonunion U.S. manufacturing plants of the Japanese auto companies such as Honda to maintain its Bargaining Power in that industry. On the other hand, opportunistic campaigns attempt to increase a union’s membership by organizing dissatisfied workers who are “ripe” for unionization but who do not have a strategic fit with the existing membership, such as when the United Steelworkers of America tries to organize nursing home workers.

Building and Documenting Support

In any case, once an organizing drive has been initiated, the next step in the union organizing time line is building support (see Figure 6.1 ). This generally involves meeting interested employees outside the workplace and distributing information. If an organizing drive is to continue, sooner or later employees must concretely express support for having a union represent them. The most important method in U.S. labor relations for showing this support is to sign an authorization card . An authorization card is a preprinted form containing something like “I authorize [union name] to represent me for the purposes of collective bargaining” which employees fill out and sign (see Figure 6.2 ). Collecting signed authorization cards is a crucial part of the union organizing process because the cards demonstrate the interest in unionization to three important players: union organizers, the employer, and if necessary, the NLRB. Some union organizers will not continue to devote their time and resources to a campaign if they cannot collect a certain number of signatures in a certain time period.

More importantly, authorization cards show the employer that a majority of employees want union representation. Both employers and unions violate the NLRA by agreeing to recognize a union when the union lacks majority support. But if a union gathers signed cards from more than 50 percent of the employees, it can ask the employer to recognize the union as the bargaining agent of the employees. If the employer is agreeable, a neutral party can examine the cards and determine if truly more than 50 percent of the employees signed cards. If so, the employer can recognize the union and is then obligated to bargain with the union. This is called voluntary recognition, and when this occurs, the union organizing process is over (see Figure 6.1 ). Recognition based on authorization cards is called card check recognition or a majority sign-up procedure.

Unions are aggressively pushing for card check recognition procedures, both through legislative action (via the Employee Free Choice Act) that would require employers to accept card check recognition and through agreements with specific companies to voluntarily accept card check recognition. But unless the Employee Free Choice Act or similar legislation is enacted, card check recognition will probably continue to be the exception rather than the norm because employers typically refuse to recognize unions voluntarily.

Alternatives to Voluntary Recognition

What happens after an employer declines a request for voluntary recognition? First consider the pre-NLRA era before 1935. In this era employees had only one option: Step up the pressure on the company to force it to change its stance. As a result employees would launch a recognition strike —a strike to try to compel the employer to recognize their union. The Ludlow Massacre, the 1934 general strikes in San Francisco and Minneapolis, the General Motors sit-down strike, the Memorial Day Massacre, and the Memphis sanitation strike were all recognition strikes. And recall further how costly these strikes were—in loss of human life, human suffering, lost profits, and economic disruption. A major goal of the NLRA is to replace these costly recognition strikes with an orderly alternative. What is the natural, orderly mechanism in a democratic society for determining the wishes of the majority? An election.

The 1935 passage of the NLRA therefore created a certification procedure in which employees can petition the NLRB for an election to determine if a union has the support of a majority of the employees. The NLRA does not ban recognition strikes per se (though The Landrum–Griffin Act restricts recognition picketing to 30 days) but rather tries to make them obsolete by giving employees a safer alternative that does not involve lost wages and the risk of being replaced by new employees during a strike. Most recognition questions today are settled through NLRB representation elections rather than through strikes.

Contemporary examples reinforce the NLRA’s logic. For example, graduate student teaching assistants at Yale University have been trying to form a union since the early 1990s. 3 A major complicating factor, however, has been whether in the eyes of the law these individuals are employees or students. Except for a period between 2000 and 2004, the NLRB has ruled that graduate assistants at private universities are students, not employees, and are therefore not covered under the NLRA. Absent access to the NLRB certification election process, the only way for the graduate assistants to force Yale to recognize their union is through economic pressure tactics. In fact, at the end of the 1995 fall semester, the Graduate Employees and Students Organization (GESO) led a grade strike—the graduate assistants refused to hand in grades for the classes they were teaching—to try to force the university to recognize the GESO as their union. The strike failed, and the conflict between the graduate assistants and the university continues to fester. Yale graduate assistants struck again for a week in 2005 to try to win recognition for their union. This time they were joined by Columbia University teaching assistants also striking for recognition. Whether in the form of sit-down strikes, grade strikes, or traditional strikes, U.S. labor law tries to prevent this type of disruptive activity by using secret ballot elections to decide questions of representation.

Figure 6.1

Figure 6.2


Types of Elections

The several types of NLRB representation elections correspond to different questions of representation, but all have the same goal: To determine the wishes of the majority of employees. The most frequent type is the certification election , which is used in a non-union location to ascertain if a majority of employees want to become unionized—that is, to designate a specific union as their bargaining agent. If so, the NLRB uses the election results to certify this union as the bargaining agent. Most certification elections have just one union on the ballot (so the choice is between a specific union and no union), but some have multiple unions vying for representation rights. In these cases runoff elections between the top vote getters might be needed to determine the wishes of the majority.

The opposite of a certification election is a decertification election . This type of election is used to determine whether a majority of unionized employees no longer wish to be represented by their union. If so, this union is decertified and loses the right to represent and bargain for these employees. A small number of decertification elections decertify the existing union and certify a new union; this is a raid election in which employees can choose between their existing union, a challenging (raiding) union, and no union. Most decertification elections result in a workplace going from union to nonunion. The NLRB conducts around 2,000 representation elections each year; approximately 85 percent are certification elections (see Table 6.1). Unions win roughly half of certification elections and lose about two-thirds of decertification elections. This article largely focuses on certification elections. Most of the issues discussed are similar for decertification elections.

Getting the NLRB to Conduct an Election

To request that the NLRB conduct a certification election, a group of employees or a union must file a petition form with the NLRB (you can see a blank form at www.nlrb.gov/nlrb/ shared_files/forms/nlrbform502.pdf). This petition must be supported by a demonstration of sufficient interest among the employees for such an election. “Sufficient interest” is defined by the NLRB as 30 percent. This is perhaps the most important use of signed authorization cards—the typical way to demonstrate sufficient interest is to provide signed authorization cards from at least 30 percent of the employees. Many unions wait until they have cards from more than 50 percent, but 30 percent is the legal minimum. Authorization cards are presumed valid for one year. Thirty percent is also the threshold for demonstrating that there is sufficient interest for holding a decertification election, though this would be demonstrated through signatures on an employee petition or other means, not by authorization cards.

In addition to verifying sufficient interest, the NLRB must deal with several other details before scheduling a representation election. First, unless there are unusual circumstances, the NLRB will not allow more than one election in a 12-month period. Second, elections will not be authorized within 12 months of any union certification. Third, a decertification election cannot be held when there is a valid collective bargaining agreement in place (up to a limit of three years); this is called the contract bar doctrine.

Finally, the NLRB must handle the most contentious aspect of the petition— defining the occupations and geographical locations included in a certification election. A union will be certified as the exclusive bargaining representative for employees when a majority of them support the union, but to what set of employees does this refer? For example, in a grocery store, does this mean just full-time cashiers, all cashiers, all hourly employees, all employees including managers, or some other group? For a grocery chain with multiple stores in a single city, is the unit limited to a single store or does it include multiple locations?

Determining the Appropriate Bargaining Unit

When a petition for a certification election is filed with the NLRB, a definition of the relevant jobs and locations is proposed by the party filing the petition, such as this:

All full-time and regular part-time licensed dealers employed by the Employer at its Connecticut Casino, including poker dealers, table game dealers, and dual rate dealers; but excluding all other employees, office clerical employees, and guards, professional employees and supervisors as defined in the Act.

If the employer objects to this definition, the NLRB must make a determination. Section 9(b) of the NLRA states that the NLRB “shall decide in each case whether, in order to assure to employees the fullest freedom in exercising the rights guaranteed by this Act, the unit appropriate for the purposes of collective bargaining shall be the employer unit, craft unit, plant unit, or subdivision thereof.” Thus the group of occupations and locations relevant to the certification election is referred to as the appropriate bargaining unit ; employees in these occupations can vote in the election and will be represented by the union if the union wins the election.

But how should the appropriate unit be determined? Security guards cannot be in the same unit with other employees. Individuals who are not defined as employees by the NLRA, such as supervisors and managers, are excluded from NLRB-defined bargaining units. In some cases this can be a critical issue and can even end an organizing drive if enough employees are excluded. College professors can be seen as managerial because of their control over curricula, courses, and admissions; registered nurses can be seen as supervisors because they direct the work of less skilled health care workers (more about this soon); and graduate students have been ruled to be students rather than employees.

Once the supervisory, managerial, and other exclusions are resolved, the NLRB’s determination of the appropriate bargaining unit is generally based on grouping together the jobs that share a community of interest. In manufacturing it is common to include all production and maintenance employees in a single facility; this is called an industrial unit because it follows the industrial unionism model. Drivers, clerical employees, and workers at other locations of the same employer may or may not be included based on specific circumstances, such as whether there are common human resource policies or significant similarities and interactions between employee groups. In contrast, under the Railway Labor Act, bargaining units are narrow in occupation but broad in geography and include all an employer’s locations. But under the NLRA there is great diversity: Some units span many occupations and diverse locations, whereas under other circumstances the NLRB might rule that a single occupation in one location is appropriate (a craft rather than industrial unit). In some cases a bargaining unit consists of only two employees. The NLRB’s determination of the appropriate bargaining unit can be contentious because each side— the union and the employer—wants the unit defined to maximize their chances of winning the election. In fact, unions fare significantly better in smaller rather than larger elections, and in elections with more homogeneous rather than heterogeneous skill groups.

Ultimately determining the appropriate bargaining unit, including who is a supervisor or a managerial employee and which employees share a community of interest, must be done case by case because the details can vary from one workplace to another. The hearing for one 2004 case involving college professors took 19 days and resulted in a 112-page decision. The resulting delays in some of these cases can weaken or end an organizing drive. In fact, once the appropriate bargaining unit is determined, the petition for a certification election might no longer be supported by signed authorization cards from at least 30 percent of the relevant employees. For example, in the Harvard University case discussed in the accompanying “Labor Relations Application,” when the NLRB expanded the bargaining unit in 1984 to include all Harvard University clerical and technical workers, not just those in the medical area, the size of the bargaining unit tripled from 1,200 to 3,600 employees. The union had been collecting cards only in the medical area, so with this expansion it no longer had 30 percent. In such a situation the union must return to the start of the time line in Figure 6.1 —building support and collecting signed authorization cards. In the Harvard University case the union spent nearly four additional years collecting enough cards to demonstrate sufficient interest for a certification election based on the new unit. But not all unions are successful in this endeavor; and because of changed unit definitions, antiunion campaigns, delays litigating unfair labor practice allegations, and other events that erode support for unionization, fewer than 70 percent of all petition filings result in elections.

The Supervisor Controversy

The exclusion of supervisors from NLRB-determined bargaining units, and more generally from the NLRA protections, has become particularly important as the skill requirements for many jobs have increased: If everyone who occasionally directs another worker to do something is deemed to be a supervisor, countless workers will find themselves outside the NLRA protections. Unions also fear that companies are intentionally giving employees just enough responsibility to make them legally seen as supervisors with the express intent of removing them from the NLRA protections and thus making it harder for employees to unionize.

Each time this issue is brought before the NLRB, it must determine who is and is not a supervisor by applying the statutory definition contained in the NLRA:

2(11) The term “supervisor” means any individual having authority, in the interest of the employer, to hire, transfer, suspend, lay off, recall, promote, discharge, assign, reward, or discipline other employees, or responsibly to direct them, or to adjust their grievances, or effectively to recommend such action, if in connection with the foregoing the exercise of such authority is not of a merely routine or clerical nature, but requires the use of independent judgment.

The key legal difficulties are how to interpret these functions (for example, what do “assign,” “responsibly direct,” and “independent judgment” mean?) and how to weight them (for example, what if only 10 percent of a worker’s tasks are supervisory?). After much anticipation by unions and employers, the NLRB simultaneously issued three key decisions in 2006. These are called the Kentucky River cases after the name of the Supreme Court ruling that prompted the NLRB to rethink its approach to determining supervisory status. In these cases the NLRB crafted new definitions for assign, responsibly direct, and independent judgment; and in three specific settings it found that permanent charge nurses who frequently assign employees to specific patients are supervisors, rotating charge nurses who occasionally assign employees to patients are not supervisors, other charge nurses who are not accountable for the performance of other nurses are not supervisors, and lead employees in a manufacturing facility whose direction of other employees is largely routine are not supervisors.

Although these specific outcomes mostly favored the employees, the door was opened for employers to win future cases, which would exclude more employees from NLRB-defined bargaining units and from the NLRA protections. Remember that being excluded from NLRA protection does not mean it is illegal to form a union. But by splitting workers deemed to be supervisors from other employees, unions can be weakened, and without NLRA protection for the excluded employees, they can be fired for trying to organize. Increased supervisory exclusions from the NLRA are therefore another example of the continued erosion of employee rights to form labor unions. Consequently, the labor movement champions proposed the Re-Empowerment of Skilled and Professional Employees and Construction Tradesworkers Act (the RESPECT Act), which would remove “assign” and “responsibly” direct from the definition of supervisor and would require the other supervisory functions to be a majority of an individual’s job in order for him or her to be considered a supervisor. Whether this bill passes or the NLRB narrows its definition of a supervisor will be important developments to watch for under the Obama administration.

Scheduling the Election

Once all the details are ironed out—sufficient interest, timeliness, and unit definition—the NLRB will schedule an election. Elections are supervised and monitored by NLRB officials and usually take place at the employees’ worksite—aboard ships and in factories, warehouses, offices, restaurants, movie studios, sports stadiums, and train yards. In special circumstances mail ballots are allowed. Each eligible worker can vote using a secret ballot. Voter turnout is often quite high—around 80 percent on average. Half of all elections are held within 39 days of a petition being filed, and over 90 percent are held within 56 days. Challenges to a unit definition can drastically lengthen the election timetable, and a few elections might not occur until a year or more after a petition is filed. In this period between filing the petition and conducting the election, the most intense campaigning by both unions and employers takes place—all with the intent of shaping how each individual worker votes in the election.

Labor Relations Application The Appropriate Bargaining Unit for Harvard University Clerical and Technical Workers

Harvard University’s medical area consists of its medical, dental, and public health schools and is located about three miles from the main Harvard campus. The medical area employs approximately 1,200 clerical and technical workers. These employees are similar to the 2,400 clerical and technical workers throughout the rest of the university in typing manuscripts, handling correspondence, and performing technical duties, though in the medical area the work is more medically oriented and perhaps includes more health hazards. There are relatively more technical workers than clerical workers in the medical area compared to the rest of Harvard. Of the Harvard clerical and technical employees in the medical area that transfer between jobs, approximately 80 percent stay within the medical area. Human resource policies are established centrally for all of Harvard; the medical area is the only campus unit to have its own human resources office. This office is responsible for hiring new employees but must comply with the centrally established job classification and wage system.

In the early 1970s a handful of women employed in Harvard University’s medical area started a group to bring attention to discriminatory treatment against female workers. Out of frustration with Harvard’s unresponsiveness, this movement grew into a campaign to unionize, and in 1975 a petition was filed with the NLRB to conduct an election for clerical and technical employees in the medical area.


1. As the human resources director for Harvard University, do you object to the definition of the bargaining unit? In other words, would you rather have an election for all clerical and technical workers at Harvard, not just those in the medical area? What are the risks and benefits?

2. As the NLRB, how would you decide on the appropriate bargaining unit if Harvard objected?

Table 6.1


In the days and weeks leading up to NLRB representation elections, employers and unions typically conduct intense campaigns to bolster support for their sides. Unions try to create a social climate in the workplace that supports collective rather than individual action while highlighting reasons why employees should be dissatisfied with their jobs, how the union will be effective in improving their jobs (union instrumentality), and that unions in general are a positive societal force. On the other hand, employers try to make employees feel satisfied with their jobs, question the effectiveness of and need for unions in improving the workplace, negatively portray unions as dues-hungry or corrupt, and emphasize the value of individual rather than collective action. In short, employer and union campaigns try to affect the key determinants of how individuals vote in representation elections. Various campaign tactics will be discussed in the next two sections; but first ask yourself whether there should be any limits or restrictions on employer and union campaigning. Remember that NLRB representation elections are not about employers or unions. Rather, the NLRA authorizes these elections to determine whether the workers want union representation.

The key legal standard for NLRB representation elections is therefore employee free choice. Campaigning by employers and unions is permitted by the NLRA’s employer’s free speech provision [Section 8(c)]:

The expressing of any views, argument, or opinion, or the dissemination thereof, whether in written, printed, graphic, or visual form, shall not constitute or be evidence of an unfair labor practice under any of the provisions of this Act, if such expression contains no threat of reprisal or force or promise of benefit.

This right is not unlimited—threats and promises are not allowed. Why? Because threats and promises can interfere with employee free choice. Union supporters might not vote for a union if they are afraid they will lose their jobs if the union wins; employees who do not favor unionization might not vote against a union if union organizers have threatened them with physical violence. More generally, to promote the underlying standard of employee free choice, the NLRB has used the NLRA to establish boundaries for employer and union campaigning: Campaign tactics that hamper, restrict, or interfere with employee free choice in deciding whether to have union representation are prohibited. This is similar to the familiar standards for political elections for government offices: Well-informed voters are important for democracy, but buying votes through bribes or violence undermines democracy and is not allowed.

In particular, the NLRB evaluates election conduct using its laboratory conditions doctrine , also known as the General Shoe doctrine (after the name of the NLRB decision that created this doctrine):

In election proceedings, it is the [NLRB’s] function to provide a laboratory in which an experiment may be conducted, under conditions as nearly ideal as possible, to determine the uninhibited desires of the employees. It is our duty to establish those conditions; it is also our duty to determine whether they have been fulfilled. When, in the rare extreme case, the standard drops too low, because of our fault or that of others, the requisite laboratory conditions are not present and the experiment must be conducted over again.

When the NLRB feels that election campaigning or conduct causes employees to vote differently from their true preferences, the election results will be thrown out and a new election conducted. The clearest violations of laboratory conditions are the NLRA’s employer and union unfair labor practices: interfering with or coercing employees [Sections 8(a)(1) and 8(b)(1)], establishing a company-dominated sham union [8(a)(2)], and discriminating against employees on the basis of union support [8(a)(3)]. The classic example is firing union activists, which can affect the outcome of an election in several ways. Pro-union votes are kept from the election. Individuals who are key in spreading information about the union and creating a workplace climate supportive of collective action are removed. And other employees may be afraid to support the union for fear of being fired next. Firing (or otherwise discriminating against) union activists because of their union support therefore interferes with employee free choice and is illegal. Threatening to close a plant if a union wins an election or promising wage increases if the union loses amounts to buying votes and interferes with free choice. These actions are therefore prohibited. Improving wages, benefits, and working conditions before an election in order to defeat a union is also illegal interference.

But the laboratory conditions standard goes beyond unfair labor practices. In other words, election conduct need not rise to the severity of a clear unfair labor practice in order to violate laboratory standards and therefore trigger a new election. For example, employer or union campaign tactics that inflame racial prejudices are not necessarily unfair labor practices, but they can create a distorted environment in which employee free choice is affected. Lies and other distortions or misrepresentations of facts are also not unfair labor practices but can undermine free choice. Even though the NLRB has largely abandoned the view that lies and misrepresentations by themselves distort laboratory conditions—its logic is that workers can see through such propaganda—in extreme cases where the “pervasiveness of misrepresentation or the artfulness of deception during an election campaign renders employees so unable to separate truth from untruth that their free and fair choice is affected,” a new election will be held. Finally, the actions of a third party—that is, individuals not under the direction of the employer or the union, such as individual employees, the mayor or chief of police, or business and labor union leaders from other companies and unions—cannot be unfair labor practices but nevertheless can negatively impact laboratory conditions.

Figure 6.3


Employer campaign tactics try to influence the four key determinants of individual voting decisions: job dissatisfaction, union instrumentality, general union attitudes, and collective social identity. At a minimum these tactics usually include providing pro-company and antiunion information and opinions to the employees. This is generally legal. Some companies are more aggressive and supplement this information with the manipulation of wages, benefits, working conditions, and job assignments as well as with threats, promises, rumors, layoffs, and firings. Antiunion committees might be formed, supervisors might be reassigned, and parties might be thrown. Many of these more aggressive activities can distort laboratory conditions and are therefore prohibited. These efforts might be led by full-time antiunion consultants.

Communicating with Employees

Information and opinions are shared with employees using a variety of methods. Supervisors might meet with employees individually or in small groups. Letters or e-mail messages might be sent to workers. The letters in Figure 6.4 are representative of common employer messages: The first letter tries to weaken workers’ perceptions of union instrumentality, and the second portrays unions more generally in negative terms. Strikes and paying dues are also frequent themes. Another important employer tactic for sharing information and opinions is the captive audience meeting —a group meeting held in the workplace during working hours in which employees are forced to listen to management’s antiunion and pro-company presentations (captive audience speeches). Such meetings are legal as long as they are not within 24 hours of the election. Captive audience meetings are hotly debated—employers justify them on the basis of property and free speech rights; unions criticize them as giving employers an unfair advantage in communicating with, and perhaps pressuring, the employees.

These forms of communication are acceptable unless the employer is too aggressive and makes threats or promises. Sometimes, however, there is a fine line between predictions or opinions on one hand and threats on the other. Suppose an employer tells employees that if a union increases labor costs too much, the plant will have to close. Is this an opinion, a prediction, or a threat? Or suppose an employer tells employees that similar facilities have closed after being organized by this same union. Is this a factual statement or an implied threat? It depends. Predictions based on objective facts that some events will likely occur because of forces beyond the employer’s control (like competitive forces) are acceptable; statements that convey the impression that these events are inevitable or at the discretion of the employer are prohibited threats. An acceptable conversation between a supervisor and an employee also becomes objectionable if it turns into an interrogation about the employee’s views of unionization. A visit to an employee’s home by a management official is grounds for invalidating an election—because the employer controls an employee’s job, visiting him or her at home is viewed as intimidating and coercive.

Employer Restrictions on Employees and Union Organizers

Another employer campaign tactic is to use no solicitation rules. The employer can use its private property rights to prohibit outside organizations from entering the workplace and interacting with workers. As long as these rules are equally enforced for all types of outside organizations, union organizers can be denied access to employees in the workplace, except in extreme circumstances when unions have no other access to employees (such as in remote mining camps). Union organizers can even be banned from shopping mall parking lots. Note carefully that this applies to outside union organizers, not the employer’s employees.

Employees can discuss unionization in the workplace, but such conversations can be restricted to nonwork hours, and if they interfere with production or customers, to nonwork locations such as an employee cafeteria. As such, some unions, especially in the construction industry, have tried a tactic called salting in which paid union organizers try to get hired as regular employees (and therefore “salt” or enrich the workplace) for the purpose of organizing the workers. In a controversial ruling, the Supreme Court has decided that salts are employees under the NLRA and therefore protected against discrimination in hiring and firing. But to be protected, a salt must also have a genuine interest in working for the employer.

Restrictions on both employees and nonemployee union organizers must be equally enforced. Consider an organization that allows its employees to sell Girl Scout cookies in working areas but restricts employee discussions of unions to nonwork areas, or allows Boy Scouts into the workplace to sell holiday wreaths but prohibits union organizers from the premises. Isolated charitable acts will be overlooked. But if employees are allowed to repeatedly sell Girl Scout cookies to other workers during work time, such activities are likely not disruptive. To then restrict discussions of unionization by claiming that they are disruptive is disingenuous and amounts to discriminating against union activity. For the same reason, discriminatory enforcement of no solicitation rules is objectionable interference with employee rights to organize unions.

Currently the sharpest debates are over company policies that restrict employees’ use of company e-mail systems. It is well established in labor law that employees do not have a statutory right to use their employer’s equipment for union organizing or other Section 7 rights, whether it be bulletin boards, telephones, copiers, or e-mail systems. As with other restrictions, company policies on e-mail usage cannot be discriminatory; but in a controversial 2007 decision, the Bush-appointed NLRB ruled that employers can distinguish between various types of e-mail solicitations. Under this ruling employers are not viewed as discriminatory if they allow charitable and personal e-mail solicitations but not solicitations for noncharitable organizations, such as labor unions, as long as the policy is not motivated by antiunion reasons. Allowing antiunion e-mail messages while banning pro-union messages is discriminatory and is therefore prohibited by the NLRA. Allowing employees to send messages asking people to join a local club but not allowing messages soliciting coworkers to join a union is also discriminatory. But allowing employees to use e-mail to invite coworkers to a birthday party or to sell Girl Scout cookies while prohibiting solicitations to support a union is not discriminatory under this ruling. This controversial ruling is likely to be revisited by the Obama-appointed NLRB. Moreover, some argue that because of the sheer importance of e-mail for daily communication, employees should be allowed to use e-mail for union-related activities unless a company can show a valid business reason for restricting it.

Union Avoidance Consultants

The use of outside union avoidance consultants and lawyers is a prominent component of employers’ campaigns. Consultants can help managers take advantage of tactics such as captive audience meetings and train supervisors in union avoidance methods. Unions, however, view union avoidance consultants as another element of a deck stacked in favor of employers who can hold captive audience meetings, ban union organizers from the workplace, restrict employee use of e-mail, and hire expensive consultants to lead sophisticated communications and public relations campaigns. Moreover, at least some union avoidance consultants aggressively seek to break unions by resorting to almost any means necessary—a less publicly visible descendant of the aggressive strikebreaking agencies like the Pinkertons in the 19th century. This is labor relations at its worst: lies, threats, promises, manipulation, harassment, espionage, abuse, and firings. In one extreme case in South Carolina in 1999, a 17-year employee with no record of violence challenged his plant manager to let the employees talk with a union organizer. The following day the employer notified the local sheriff that this employee was threatening workers, so the sheriff’s deputies surrounded him at gunpoint on his way to work and forcibly took him to the local hospital. Based on what the employer told the sheriff, the employee was involuntarily committed to a mental hospital. He was held for two weeks against his will and forcibly injected with antipsychotic drugs until a lawyer could obtain his release.

With the help of attorneys and consultants, another employer tactic is delay. By challenging the proposed bargaining unit definition in an election petition, an employer can slow down the election time line and delay the election date for a couple of months or more. The 1975 unit determination case for Harvard University clerical and technical workers was stretched out by the university’s lawyers to include more than 20 days of hearings; and in a similar case at Yale University, the university’s lawyers submitted a witness list of over 300 individuals. Why is delay a significant tactic in the employer’s favor? The employer gains more time to campaign against the union; employee turnover may result in the loss of union supporters; perceptions of union instrumentality might be weakened as the union appears helpless to counter the employer’s legal maneuverings; and critical union momentum is lost.

The Controversies over Employer Campaign Tactics

Employer campaigning during NLRB representation elections is a controversial topic in U.S. labor relations. Some argue that union representation is a question solely for workers, and employers should not be granted rights as formal participants in the process—free speech should be allowed, but not the right to object to the definition of the unit, to use special campaign tactics like captive audience speeches, and to challenge the results. The usual counterarguments are that the employers’ property rights give them the right to participate in the process, that unions are too powerful without a fair counterweight of employer campaigning, and that employers need to speak for the antiunion employees and deliver their message.

More pragmatically, there is also significant debate over the practical questions of the prevalence and significance of employer campaigning. Some campaign tactics are hard to observe or measure (such as informal supervisor conversations with employees), and most are undertaken quietly (especially the illegal ones). Using NLRB data on illegal discharge unfair labor practices, it has been estimated that workers were illegally fired in 25 percent of representation election campaigns during the 2000s and that there is nearly a 2 percent chance that any pro-union worker involved in an organizing drive will be illegally fired. If union activists represent 10 percent of pro-union workers, this means that there is a 20 percent chance that union activists will be fired. Moreover, employees perceive the chances of retribution for organizing activity to be high: In one survey 41 percent of nonunion respondents agreed that “it is likely I will lose my job if I try to form a union,” and 79 percent of all respondents said it is “very” or “somewhat” likely that “nonunion workers will get fired if they try to organize a union.”

More generally, research seems to reveal a pattern of broad-based employer campaigning during NLRB representation elections. This research often relies on the reports of union organizers, so it might overstate (if organizers or their inside sources intentionally or unintentionally inflate the amount of employer resistance) or understate (if organizers do not observe all the campaigning) the true level of employer campaigning. With this in mind, Table 6.2 presents some typical statistics. Note that a large fraction of employers reportedly use outside consultants, one-on-one meetings between workers and supervisors, and multiple captive audience meetings.

But do these tactics matter? This is the most debated question. Relative to their peers in other industrialized countries, American managers appear to be exceptionally hostile toward unions and have significantly stronger traditions of using union avoidance tactics. Various studies find that employer antiunion campaign tactics reduce the likelihood that employees will vote for unions in NLRB representation elections and that unions are less likely to win these elections. In contrast, a famous study in the 1970s found that most workers already had their minds made up before employer campaigning began and therefore that employer tactics do not matter. On the other hand, a reanalysis of this study’s data revealed that because many elections are decided by thin margins, captive audience meetings as well as objectionable threats and actions toward union supporters can affect enough votes to shift the outcome of the election. Yet another study found that employer unfair labor practices did not reduce the probability of unions winning an election, but they did reduce overall union success rates by decreasing the likelihood that a petition filing would result in an election being held. And the debate continues.

In sum, by many accounts employers devote a lot of time and money to campaigning against unions in NLRB representation elections. Some campaign tactics are legal, though still criticized by labor supporters; other tactics are illegal, and union busting can be “a very dirty business.” Given the resources spent on campaigning, managers must perceive them as effective in reducing the likelihood that a union organizing drive will be successful. Moreover, the evidence points toward the presence of significant incentives to avoid unions at all costs. One-third of nonunion managers believe that their careers will be harmed if their employees unionize. And they are probably right: In one study, managers in establishments without any union organizing activity had a 21 percent chance of being promoted and a 2 percent chance of being fired; in similar establishments that experienced an organizing drive, not a single manager was promoted and 15 percent were fired.

Ethics in Action A Union Avoidance Consultant Tells

All but into the bedroom if necessary. The buster is not only a terrorist; he is also a spy. My team and I routinely pried into workers’ police records, personnel files, credit histories, medical records, and family lives in search of a weakness that we could use to discredit union activists.

Once in a while, a worker is impeccable. So some consultants resort to lies. To fell the sturdiest union supporters in the 1970s, I frequently launched rumors that the targeted worker was gay or was cheating on his wife. . . . If even the nasty stories failed to muzzle an effective union proponent, the busters might get the worker fired. . . .

Not only were working people crushed by the cruelty of the union busters, but the companies themselves were raped, as consultants and attorneys conspired to wring as much as they could out of their clients. The executives paid whatever they were asked, the consultants having convinced them that a union organizing effort amounted to the worst crisis of their business lives. In the end I understood that a union busting campaign left a company financially devastated and hopelessly divided and almost invariably created an even more intolerable work environment than before. . . .


1. Show how the type of union busting described here is unethical in all six ethical theories.

2. Given that this type of union busting is unethical in all six perspectives, why do some managers hire such consultants? Why do some people become union busting consultants? Is it simply because some people are unethical, or are there more complex reasons?

Marty Levitt spent more than 20 years as a union avoidance consultant, aggressively directing campaigns to prevent unions from winning representation elections. Then he had a change of heart and publicly revealed his past activities, offering a glimpse into the big business of labor relations consulting. Here are some of his revelations:

There are many forms of union busting. Some labor consultants and attorneys take on unions that already represent a workforce, squeezing negotiators At the Bargaining Table, forcing workers out on strike, harassing union officers. My career took another path. I refined the specialty . . . called “counterorganizing drives,” battling nonunion employees as they struggled to win union representation. The enemy was the collective spirit. I got hold of that spirit while it was still a seedling; I poisoned it, choked it, bludgeoned it if I had to, anything to be sure that it would never blossom into a united workforce, the dreaded foe of any corporate tyrant.

For my campaigns I identified two key targets: the rank-and-file workers and their immediate supervisors. The supervisors served as my front line. I took them hostage on the first day and sent them to anti-union boot camp. I knew that people who didn’t feel threatened wouldn’t fight. So through hours of seminars, rallies, and one-on-one encounters, I taught the supervisors to despise and fear the union. I persuaded them that a union organizing drive was a personal attack on them, a referendum on their leadership skills, and an attempt to humiliate them. I was friendly, even jovial at times, but always unforgiving as I compelled each supervisor to feel he was somehow to blame for the union push and consequently obliged to defeat it. . . .

Although I took on the supervisors face-to-face, my war on union activists was covert. To stop a union proponent—a “pusher,” in the antiunion lexicon—the buster will go anywhere, not just to the lunchroom,

Labor Relations Application The Case against Employer Campaigning

The major counterargument [to proposals to limit employer campaigning] made by employers— especially the majority who do not engage in this most egregious breaking of labor law tradition—is that they are denied what should be their equal right to campaign for the allegiance of employee voters. This is said to be the equivalent of a political election that allows the Democratic party, but not the Republicans, to campaign effectively for voter support in political elections.

The fallacy in that analogy and argument is that it mistakenly assumes that an affirmative vote for the union . . . means that the union is now governing employers (or even employees). All that a successful employee verdict does is give the union the mandate to represent employees in negotiations with the employer under labor law. Under Employment Law, we would never dream of suggesting that the employer should have an affirmative right and opportunity to campaign against the employee’s decision about whether to hire a law firm (and if so, which one) when challenging employer policies regarding occupational safety or sexual harassment, for example. . . .

The more apt political analog of the role of the employer in a representation election is the role of a foreign government in an American election. Canada, for example, has a significant interest in which party is elected to govern the United States; selection of one party rather than the other may make life considerably easier or more difficult for the Canadian government in negotiations over defense, trade, natural resources, energy, foreign investment, and so on. Yet no one would argue that Canadian government agencies should therefore have a right to participate in an American election campaign in order to try to persuade United States citizens to vote for a party that would be favorable to Canadian interests. After all, it is the job of the United States government to advance the interests of its own citizens when those interests conflict with the interests of Canadians; Canadians have their own government to defend their interests irrespective of the electoral verdict in the United States. . . .

The law should not restrain the employer’s freedom to say what it will about collective bargaining—censorship in the representation campaign has the same offensive flavor that it has in politics or in the arts—but there is no principle of fairness that requires that the representation process be structured to facilitate employer opposition to unionization.

Figure 6.4

Table 6.2