How should we understand and explain individual acts of racism? Despite extensive debate about the broader place and importance of racism in America, there is surprisingly little theoretical or empirical analysis of what leads individuals to commit racist acts. Although scholars differ on racism's particularities, extensiveness, and significance, most political scientists are in agreement that racism is at its root an individual psychological attitude, an irrational prejudice, that stems from feelings of resentment and animus toward others or from a desire to create group status hierarchies (e.g., Bobo 1988; Kinder and Sanders 1996; Sears 1988; Sidanius and Pratto 1999; Sniderman and Piazza 1993). Only a few studies explore what leads people to act on their attitudes and those that do tend to similarly emphasize individual factors such as the racist actor's feelings of resentment, insecurity, and anxiety (e.g., Gourevitch 1999; Green et al. 1998; Olzak 1992). Because scholars see racism as an irrational prejudice, they tend not to examine the role that rules, institutions, and politics play in determining why some individuals act on their racist attitudes and why others do not. Certainly, in comparison to the wealth of studies documenting the importance of rules and institutions in guiding the behavior of individual legislators, party elites, judges, and public administrators, there has been little examination of why certain individuals commit racist acts.
This model of racism is not confined to public opinion scholars. From de Tocqueville (2001) to Du Bois (1935) to Myrdal (1944) to Roediger (1991), race theorists have consistently conceived of racism as both an individual and psychological phenomenon. Racism is treated as antithetical to ideological traditions of tolerance and equality and is attributed to people wanting to maintain emotionally based hierarchies. Even those who highlight the relationship between racism and power tend to fall back on definitions that see it as a “deformity of rationality” (e.g., Appiah 1990, 8; Rogin 1988; Takaki 2000). For instance, although Albert Memmi (1999, 38, 27) argues that the “machinery of racism” enables elites to exercise power and privilege, he claims that racism arises from an individual's “mistrust, if not repulsion and fear” of something—or someone—who is different, “like an unfamiliar plant growing by the side of the road, whose odor itself may be noxious.” Rogers Smith (1997, 38) sees racism as intrinsic to elite efforts at state building; at the same time he argues that racism both derives and sustains itself from individual resentments and the desire, particularly among those less powerful, “to feel part of a larger, more enduring whole of intrinsic worth.” Similarly, many critical race theorists who claim that law and institutions enable and legitimate racist activity maintain that racism stems from individual-driven irrational behavior that is unconscious or an “attenuated” psychological predisposition (e.g., Haney López 2000, 1730; Krieger 1995; Lawrence 1987). At base, then, all of these analyses believe individuals buy into the appeal of racism because of psychological needs, not because they are motivated by broader institutional dynamics.
Viewing racism through the lens of psychology is not so much wrong as incomplete. By understanding racial conflict as irrational acts conducted outside the confines of political actors and institutions, these works depoliticize racist activity and ignore important dynamics of power and incentives that shape individual behavior. Racism, writes Adolph Reed (2000), “is not an affliction…. Nor is it a thing that can act on its own; it exists only as it is reproduced in specific social arrangements in specific societies under historically specific conditions of law, state, and class power.” Racist manifestations by individuals are the result of a complex set of factors, and latent psychology, I argue, is less helpful for understanding it than are the maneuverings and behavior of strategic actors following rules and incentives provided by institutions. Moreover, racism is not politically problematic simply because some or even many individuals hold racist attitudes; it becomes problematic when institutional dynamics legitimate and promote racist behavior in a concentrated and systematic manner. As such, we need to examine the ways in which institutions encourage racist acts by providing rules and procedures that motivate people to behave in a racist manner or behave in a manner that motivates others to do so.
The institutional analysis of racism I put forth in this paper draws from and expands on the work of a multidisciplinary group of race scholars. Many of these scholars have examined how racial cleavages intersect with institutional dynamics, leading to racism's continuing importance in America even as societal attitudes seemingly change (e.g., Bonilla-Silva 1996; Frymer 1999; Hochschild 1984; Lieberman 1998). Others have focused on the role of state elites in configuring racial and racist understandings through bureaucracies and political institutions (e.g., Brown et al. 2003; Katznelson 1973; King 1997; Marx 1998; Skrentny 2002; Walton 1997) and through the dissemination of ideologies that either promote racial hierarchies or attempt to reconcile public aspirations for freedom with widespread racial inequality (e.g., Du Bois [1903] 1999; Fields 1982; King and Smith 2005; Smith 1997). Claire Kim (2000, 9) argues for a notion of “racial power” that is not “something that an individual or group exercises directly and intentionally over another individual or group but rather as a systemic property, permeating, circulating throughout and continuously constituting society.” Still others have examined the confluence between the state and the market, arguing in different ways that racism is related to class hegemony (e.g., Bobo 1988; Cox 1948; Goldfield 1997; Reed 2002). All of these works intersect with and have been influenced by the scholarship of those who contend that race is continually being formed and re-formed in the context of political struggle (e.g., Gilroy 1984; Hall 2000; Kim 1999; Omi and Winant 1994). Race and racism can take many forms—as Anthony Appiah (1990) argues, there are distinct “racisms”—that are constructed by elites who fight over the meaning of the terms in an effort to maintain power and promote differently racialized agendas.
This previous work almost exclusively focuses on race and racism at a theoretical and/or macro level (but see Haney López 2000; Kim 2000). In doing so, it has offered compelling arguments for how racism is a part of our national identity and embedded in and shaped by “the State,” but it has not examined how concrete institutional dynamics work to motivate individual behavior. Even studies of “institutional racism” emphasize less how institutions motivate individuals than simply making the important point that racism is located within places of power (e.g., Carmichael and Hamilton 1967; Knowles and Prewitt 1969). By contrast, the institutional understanding of individual racism I put forth emphasizes the following features, all of which are fairly common to institutional studies of politics but have not been applied to understandings of individual racism. First, institutions do not merely provide avenues for racist actors to operate, but can independently encourage racist acts by influencing individual preferences and rewarding certain types of behavior over others—they are, as March and Olsen (1984, 738) argue, “more than simple mirrors of social forces.” Second, institutions enhance the power of those actors who are well situated within them, providing these actors with coercive power, the ability to set the agenda, and the ability to anticipate and in turn shape the behavior of those with whom they interact (e.g., Gaventa 1980; Lowi 1969). Third, not all forms of political and individual behavior result in the same opportunities and outcomes; collective action problems in particular disadvantage some interests and forms of political action while promoting others (Mansbridge 1986; Olson 1971). To understand manifestations of individual racism, we must recognize that institutional structure and organizational dynamics influence whether racist actors express themselves or whether they remain silent. Fourth, the use of race and racism is not, by itself, politically problematic, nor are all racist expressions equal in significance. Only by placing the manifestations of individual racism in a broader context can we understand how the act acquires importance and meaning (Baliber 1992; Omi and Winant 1994; Said 1990). By deemphasizing the importance of individual prejudice—although by no means denying its existence—as a determinative feature of racist manifestations, I wish to locate the act within the context of institutional combat. Such an understanding, in turn, sees racism in society less as intractable and innate than as something malleable and politically determined.
To further explore and compare institutional and individual approaches, I examine manifestations of racism in labor union elections. I analyze more than 150 cases in which the National Labor Relations Board (NLRB) and federal appellate courts have formally responded to reported instances of racism during a union election. Racist acts in union elections are considered an unfair labor practice under national labor law and either the NLRB or a federal appellate court can overturn an election's outcome if it finds that those acts unduly influenced the voters' decisions. The holdings of the NLRB and federal courts, as well as extensive detail of the facts and context of the racist acts, are publicly available and the data set that I have compiled is the universe of reported cases between 1935 (the year that the National Labor Relations Act was passed) and 2000. The data set is unique and relevant for many reasons. Most important, the thick discriptions of each case provide an opportunity to analyze racism within a broader legal and institutional context of workers and employers strategically vying for power over the workplace. Moreover, it offers an excellent opportunity to compare the theoretical leverage provided by individual and institutional models of racism: as we shall see, federal courts typically respond to reports of racism in a manner parallel to the individual-psychological model endorsed by most political scientists whereas the NLRB consistently treats the same factual events as engrained in, and as a product of, institutions. Analyzing how two different legal bodies come to often entirely different interpretations of the same incident allows us to see both the advantages and disadvantages of each theoretical approach and to highlight the assumptions that underlie both. The data set, it should be noted, is also limited in important ways. It relies entirely on published cases and cannot include the untold number of situations that either went unreported or did not warrant a response in the Board's estimation. The available data illustrate how contrasting theories of racism explain individual acts but do not provide a “test” of the models in any way nor, because of inherent problems with sampling bias, an empirically conclusive comparison of Board-court behavior. The goal of this paper, then, is at the level of theory: to illustrate both the limits of an individual approach to understanding racism and the theoretical contributions of an institutional approach.
TWO VIEWS OF RACISM: ANTI-DISCRIMINATION AND LABOR LAW
American employers, workers, and labor unions have a long history of participating in workplace racism, producing racially segregated and unequal workforces and unions (see, e.g., Nelson 2001; Roediger 1991). Federal labor law in the first few decades of the twentieth century gave unions and employers the opportunity to sign collective bargaining agreements that provided for white-only workforces, leading to the effective removal of African Americans and other racial minorities from whole spheres of employment (Arnesen 2001; King 1997). By the mid-twentieth century a significant number of national and local unions participated in systematic and widespread discrimination, particularly against African American workers, by denying them employment and union representation, committing unfair labor practices, and participating in explicit and often violent racial conflict (Gould 1977; Hill 1985).1
In claiming that unions had significant race problems, I am dramatically simplifying a complex story, overlooking those unions whose leaders and members worked actively for civil rights (see, e.g., Draper 1994; Kelley 1990).
Both the federal courts and the NLRB have actively intervened, but the two institutions have consistently responded to union racism in fundamentally different ways. Courts focus on the wrongdoing of the actor and view racist acts as outside the confines of rational politics, and the NLRB, although finding racism virulent and wrong, has tended to treat it as part and parcel of a broader set of institutional and political dynamics.
Federal courts have confronted union racism primarily through Title VII of the 1964 Civil Rights Act and the Equal Protection clause of the Fourteenth Amendment. Both legal instruments have been used in different ways at different times by judges of different political persuasions and scholars have thoroughly argued that multiple traditions of antidiscrimination law have had moments of resonance (e.g., Balkin and Siegel 2003; Forbath 1999; Kersch 2004). Despite this variety, one assumption remains quite persistent, particularly in the post–Civil Rights era: in a democratic and liberal society that values individual tolerance, racism is wrong, irrational, and should always be outside of politics and law. Racism is “obviously irrelevant and invidious,” declared the Supreme Court in perhaps its most famous union discrimination case (Steele v. Louisville & N.R. Co., 323 U.S. 192, 203 [1944]). “A racial classification, regardless of motivation, is presumptively invalid” (Personnel Administrator of Massachusetts v. Feeney, 442 U.S. 256, 272 [1979]) and places a “brand upon” those who are its targets (Strauder v. West Virginia, 100 U.S. 303, 308 [1880]). The Constitution, after all, is “color-blind” (Plessy v. Ferguson, 163 U.S. 537, 559 [1896], Harlan dissenting), and race and racism should have no place in political dialogue and involvement—indeed, the racial classification itself constitutes a prima facie indicator of discrimination and prejudice (e.g., Anderson v. Martin, 375 U.S. 399 [1964]; Loving v. Virginia, 388 U.S. 1 [1967]; Shaw v. Reno, 509 U.S. 630 [1993]). In contrast to some forms of government classifications on the basis of sex, sexuality, disability, and age that the Supreme Court has argued have at least the potential of being rational and legitimate considerations, racial classifications are given the highest level of scrutiny. The Court strongly suspects racial classifications are motivated by invidious and irrational goals and assumptions and will only allow the classification if the government can provide a “compelling” reason (Brest 1976; Post 2000). When a broader political or social context has been introduced as a rationale for the government's racial classification, the Supreme Court has been fairly consistent in rejecting the broader context unless specific individual-level racism can be proven (e.g., McKleskey v. Kemp, 481 U.S. 279 [1977]; Milliken v. Bradley, 418 U.S. 717 [1974]; Wygant v. Jackson Board of Educ., 476 US 267 [1986]).
The Court's interpretation of Title VII has followed similar assumptions to the aforementioned Equal Protection cases. In Title VII cases, the issue is less about a racial classification than about the motivations behind a purported incident of discrimination in the workplace, usually by an employer. A typical case involves a judge attempting to locate (often with the help of psychologists) the precise moment when an individual is specifically and directly motivated by racism (or sexism or other forms of animus), and to separate this moment from other moments when the individual is presumably motivated by rational pursuits (e.g., St. Mary's Honor Ctr. v. Hicks, 509 U.S. 502 [1993]; Price Waterhouse v. Hopkins, 490 U.S. 228 [1989]; Alexander v. Sandoval, 532 U.S. 275 [2001]). Sometimes, this precise moment and motive is difficult to prove and courts have used a variety of measures to make it easier for a lawsuit to go forth without a “smoking gun” to find whether animus was the pretext for the actor's otherwise seemingly rational decision, whether by making it relatively easy for a plaintiff to establish a prima facie (McDonnell Douglas Corp. v. Green, 411 U.S. 792 [1973]) or by allowing for statistical evidence to show a “disparate impact” (Griggs v. Duke Power Co., 401 U.S. 424 [1971]) and a “pattern and practice” of discrimination absent a finding of individual intent (e.g., Teamsters v. United States, 431 U.S. 324 [1975]). Critical to current antidiscrimination law, however, is the sense that an individual or group of individuals is responsible for the racist act, that they must be punished, and that such behavior must be removed from the sphere of rational decision making (see Freeman 1978; Krieger 1995). Moreover, similar to the assumptions behind the Supreme Court's Equal Protection decisions, Title VII law declares that race can never be a factor in rational decision making, even when the law explicitly makes an exception for other forms of discrimination on grounds that gender or disability can, under certain conditions, be objective considerations for employment.2
Under Title VII law, there is a “bona-fide occupational qualification” exception that enables an employer to discriminate on the basis of gender and age, but not race, if the discrimination is deemed essential to the job criteria
.
Premised on an entirely different understanding of individuals and power, labor law provides a sharp and interesting contrast to the courts' approach to racism. Congress passed the National Labor Relations Act in 1935 to promote peaceful resolution of workplace conflicts by giving unions the opportunity to engage in collective bargaining with employers. In passing the Act, legislators recognized that workers and management were fundamentally at odds, and that the NLRB was given regulating power to negotiate contracts and bargaining agreements between the two sides, all the while recognizing the potential dangers of discrimination and intimidation by both employers and unions during the context of union activity. Unlike most government institutions that claim to operate in an environment of more or less equally powerful actors, and unlike federal courts which are dominated by an anticlassification model of discrimination that looks only for the mere mention of race, the NLRA uniquely recognizes the fundamentally unequal power relations of the workplace and the “relative weakness of the isolated wage earner” (Senate Report No. 573, 74th Cong., 1st Sess., 1935, 3). Interpreting the Act in NLRB v. Jones & Laughlin Steel Corp. (301 U.S. 1, 33 [1937]), the Supreme Court acknowledged “that a single employee was helpless in dealing with an employer; that he was dependent ordinarily on his daily wage for the maintenance of himself and family…that a union was essential to give laborers opportunity to deal on an equality with their employer.” More than three decades later, the Court reiterated in NLRB v. Gissel Packing Co. (395 U.S. 575, 617 [1969]) that any rights of the employer to promote its position to the workers during a union drive must be balanced by “the economic dependence of the employees on their employers, and the necessary tendency of the former, because of that relationship, to pick up intended implications of the latter that might be more readily dismissed by a more disinterested ear.” Although the NLRB is an agency that continually changes based on the preferences of its politically appointed members, its statutory obligations and regulatory nature have led it to consistently scrutinize individual actions in the context of NLRA rules and the broader goals of the Act.
The NLRB's handling of union elections typifies this approach to the individual actor. Labor unions are most commonly formed through employee elections. These elections are unlike other democratic elections in the United States in that the free speech of all the participants involved—the workers who are voting, the union organizers, and the employers—is severely restricted. Labor law gives voters in union elections “the opportunity of exercising a reasoned, untrammeled choice for or against labor organizations seeking representation rights” (Sewell Manufacturing, 138 NLRB 66 [1962]). The NLRB has consistently worried that workers will treat speech during elections as a threat that can be carried out, usually by employers because they control hiring and firing, pay, benefits, and promotions. Thus, the Board scrutinizes acts and speech during the election campaign that might be conceived as a threat—explicit or implicit—and, as a result, lead workers to vote in a manner that does not reflect their true preferences. Over the years, the NLRB has devised a number of rules governing the speech and conduct of the relevant actors during union elections, which are arranged and directly supervised by regional boards around the country (for an overview of NLRB election policy, see Becker 1993; Getman et al. 1976). Employers, for instance, are allowed to state their opposition to unions but are not allowed to threaten reprisals or promise benefits. No unlawful firings are allowed, no threats of plant closure or loss of jobs or denial of future benefits can be made, no promises of specific future benefits may be suggested, nor are bribes by the employer or by the union to employees allowed. The employer is not allowed to interrogate employees about their union sympathies, nor is it allowed to improperly survey union activities. The employer may not make campaign speeches to the employees within 24 hours of the election. At the same time, the union cannot pass out literature and campaign on workplace grounds (though the employer can), cannot participate in “captive audience speeches” held by the employer (unless the employer chooses otherwise), and, similar to the employer, cannot threaten the workers in any way. If a violation of these rules occurs during the election period, the Board has the power to overturn election results or, in more extreme cases, to issue injunctions and/or bargaining orders that force the employer to sign a contract with the union even if the union has lost the election.
Racism by one of the parties during the campaign is a violation that can lead the Board to overturn an election. Unlike the federal court decisions under antidiscrimination law, racism in labor law is regulated for its potentially damaging political consequence, not because it is considered reprehensible and unacceptable in any context. The Board approaches racist acts through Section 8(a) of the NLRA, which prohibits threats to workers made in the context of union election campaigns. The Board has feared that, in the effort to dissuade employees from voting in favor of the union, the employer may resort to a variety of tactics designed to inflame racial prejudice among employees and, as a result, convince workers that voting for a union will hurt their workplace environment. Alternatively, the union may attempt to arouse feelings of racial pride in employees either by attempting to create an exclusive racial hierarchy among one group of workers at the expense of another, or to attack the employer on racial grounds, or to convince the employees that they need a union for protection from a racist employer. The individual worker is clearly the object of these appeals and the fact that workers respond to these tactics is no doubt the result of at least implicit support—whether psychological or ideological—for such ideas. But the ideas in the abstract are not what the Board sees as problematic, rather the Board focuses on two things; which side prompted the racist act and if the act was consequential. And here, unlike the view of courts and much of political science, the causation lies not in individuals wishing to benefit from an emotive or psychological wage, but from institutional dynamics that promote this type of behavior among employers and union leaders to obtain their broader goals.
The Board perceives racism, as the following pages show, as a politically rational strategy employed by both sides during union elections. One side wants to use race and racism to divide workers, the other side to unify workers. The institutional context in which union elections are waged allows race-based strategies to be used in many different ways, by many different groups, and often in a less than explicit manner. The Board's response is multifaceted. First, it has defined racism in the context of whether it is threatening or not. If racism is seen as economically or politically coercive, it is regulated; if not, it is usually allowed. Second, there is an institutional dimension that is relevant, unbalanced, and an important influence on individual behavior. Employers must bargain with the union once it is elected and, thus, are not allowed to intimidate or coerce during the campaign. At the same time, the employer retains certain powers, most notably the power to hire new workers, to shut down or move the company out of the United States, to play hardball in union negotiations, and to hire replacement workers if a strike were to ensue. The Board is highly attuned to an employer's power to set agendas and manipulate institutional rules and disallows racism to the extent that it contributes to that power, particularly when its introduction to an election campaign seems a product of employer efforts to defeat the union. Third, the explicit assumption of labor law is that workers cannot succeed without collective action, and so the Board tends to discount individual preferences in favor of majority rule. Because the Board focuses more on majorities than individuals, it usually differentiates between leaders who are deemed accountable to their constituencies for their actions and individuals who are not. This means that the Board examines the identity of the perpetrator of the racist act—whether he or she is an authorized leader or merely an individual—to determine whether the act is politically significant and worth regulating.
UNION ELECTIONS AND LABOR REGULATIONS AGAINST RACISM
The NLRB Cases: Agenda Setting, Power, and Collective Action
The three aforementioned institutional dynamics broadly explain the Board's handling and understanding of union racism. In this section, I examine more than six decades of decisions in which the Board confronts racism in the context of union elections. Cases were compiled from the use of two legal search engines, Lexis and Westlaw. There are a total of 115 cases decided by the NLRB. (In the following section, I will examine the 43 cases that were decided by federal appellate courts.) As mentioned earlier, the cases are the universe of Board decisions; however, they are not the universe of racist incidents in union election campaigns as I examine only those cases that reach the Board's review and are published.3
The cases examined are also not the universe of incidents of racism to which the Board has officially responded. The Board has also dealt with union and employer racism in the context of hiring discrimination and a “duty of fair representation.” These matters represent separate realms of cases beyond the scope of this paper and have frequently involved not only federal courts, but other administrative agencies such as the Fair Employment Practices Committee, the Equal Employment Opportunity Commission, and the Department of Labor. (For further discussion of these matters, and the contrasting ways in which these different agencies handled union racism, see Frymer 2004; Kryder 2000.)
An untold number of cases are either handled only by an administrative law judge or are never officially brought forth by a party as a violation of labor law. But, for the purposes of this paper, these cases allow us to systematically examine the Board's decision making and the principles it uses to resolve cases.
The earliest reported cases (1935–66) took place almost exclusively in the South, and dealt with clear manifestations of what political scientists label “traditional racism” where employers used racially explicit epithets to scare white workers, and on occasion black workers, from voting for the union. With the exception of a few conflicts between unions and Jewish employers and lawyers, all of the cases during this time involve divisions between whites and blacks. The Board handed down decisions in 71 cases during this time period, only two of which responded to an accusation by the employer that the union instigated the racist activity. In some cases, the racism went beyond words to physical violence, leading to the physical assault of black employees and suspected union organizers, the brandishing of weapons, and even the shooting of union members. Shortly after the Civil Rights Act of 1964, the nature of the cases changed dramatically. Out of the 44 cases decided by the Board between 1967 and 2000, only 25% involved accusations that the employer committed the racist act. Instead, the overwhelming number of cases involved situations where the employer accused the union of racism, either through union member epithets against other workers or the employer, or in situations where unions used race-specific appeals to mobilize African American, Asian American, or Latino workers. By the 1980s, an increasing number of these cases involved conflicts between racial and ethnic minorities—blacks and Latinos, Filipinos and Japanese, Jews and Protestants, and so forth. But, despite these differences in the parties involved, differences in which groups were the target of racist actions, and differences in the Board's political composition, the Board has consistently understood these acts as a product of institutional combat.
Traditional Racism, 1935–66: Employer Threats and Agenda Setting
Two types of scenarios were particularly common during the cases in the first three decades of the Act, almost all of which predate the 1964 Civil Rights Act. More than two-thirds of these cases were situations where the employer attempted to defeat the union by appealing to white worker racism with suggestions that electing a union would lead to a racially integrated workforce. The cases involved a variety of types of ways in which employers race-baited. In one, the employer called the union organizer “a communist, an agitator, and generally a ‘no-good nigger,’” (California Cotton Oil, 20 NLRB 540, 549 [1940]); in another, the employer told white workers that if a union came in, the factory would “be fulla Negroes” (S.K. Wellman Co., 53 NLRB 214, 215 [1943]); while a third case involved an employer hiring five African Americans to pass out leaflets supporting the union in an effort to push the company's white workers to vote in the opposite direction because of a perception that the union was associated with racial diversity (Heintz Division, 126 NLRB 151 [1960]). The events of Bush Hog, Inc. (161 NLRB 1575, 1592 [1961]) typified the period. The company president told workers during the election campaign in Selma, Alabama that “if the Union went in and all that, that we would have to work with Negroes” and if the union were not elected the employer “would keep them out.” The Board described the election campaign as “charged…with the atmosphere of the Negro revolution for equality and the march from Selma to Montgomery” and both sides were accused of playing up race issues. The company president was a member of the local White Citizen's Council and a chief part of his campaign against the union was to portray it as pro–civil rights (a charge the Teamsters consistently and vehemently denied to the workers). After referring to the Teamsters' donation of $25,000 to Martin Luther King, the president “went on to say that we are not going to hire any Negroes. [But] if the Union comes in you will be working beside Negroes.” Later in the campaign, he called the Teamsters “nothing but nigger-loving gangsters.” A poster was put up by the employer that showed a “fat” African American man smoking a cigar with the caption, “Us and that Union are going to change things around here.”
A second set, representing roughly 20% of the cases during this time, involved situations that paralleled the aforementioned cases but involved employers attempting to scare African American workers, either by intimating that whites would be hired to replace them, by firing African American workers and replacing them with white workers to scare others (Bess F. Young, 91 NLRB 1430 [1950]), or through straightforward physical intimidation. Typical was Fred A. Snow Co. (41 NLRB 1288 [1942]), where the employer told his black employees that if the union organizing movement were to succeed, he would turn his business over to his son who “didn't like colored people.” Sometimes the employer's actions were fairly subtle, as when an employer brought in roughly 100 white job applicants in full view of the majority black employees just before the employees were to vote on whether to endorse a union as their representative (Associated Grocers Port Arthur, Inc., 134 NLRB 468 [1961]). Others were more explicit. In Taylor Colquitt Co. (47 NLRB 225 [1943]) the employer's actions, as well as the actions of white employees opposed to the union, involved direct acts of racism and intimidation. In this case, the employer repeatedly threatened physical harm toward her black workers, telling them early in the campaign that “All you boys will be out of a job; you won't have nothing to do; you will be going around hungry. Furthermore…there is going to be some trouble around here if you don't stop this Union. There [will] be some blood shed.” Later, after she confronted a union supporter (he had told her, “Mrs. LaBoone, I don't want to talk to you on this…. You are a white lady. I am a colored boy. I couldn't talk to you on nothing like that”), she warned him that “If you vote for it I will kill you.” Later, she encouraged white employees, all brandishing rifles, to confront a group of black workers who were told they would not be allowed to go on “organizin' agin the whites.”
The NLRB members were initially unsure how to respond to these cases and many of the decisions included dissents and concurrences. The NLRA had been passed by Congress with no specific provision that defines racism as illegal or an “unfair labor practice.” A “duty of fair representation,” while endorsed by the Supreme Court in a case involving the parallel National Railway Act (Steele v. Louisville & Nashville R. Co., 323 U.S. 192 [1944]), would not be enforced by the NLRB until 1964 and only involved representation issues for those who were already union members. In some of these early cases, the Board held that the employer's comments, and even acts of physical intimidation and violence, did not influence the election results. The Board argued that as long as the comments were not combined with a clear threat and as long as they did not misrepresent the facts, the statements themselves, while unpleasant and undesirable, were not coercive (e.g., Happ Brothers Co., 90 NLRB 1513 [1950]; Sharnay Hosiery Mills, 120 NLRB 750 [1958]). Particularly in the cases involving white workers who responded to race-baiting, the Board was sympathetic to employer arguments that they were doing nothing more than making factual representations about the union and what would happen if the union were to win. The problem, employers argued, was not their actions but the workers' racist reactions.
But even in these cases, Board members were frequently in disagreement with each other. One concurring member in Westinghouse Electric Co. (119 NLRB 117 [1957]) disagreed that employers were just providing factual representations: “The more subtle problem, however, arises when the reference to job retention or job loss is tied to the fact that the Union has a policy, at odds with that of the Employer, which calls for disregarding racial lines in the allocation of jobs, the implication being that, if the Union wins the election, union policy will probably prevail thereafter in the plant.” The fear was that employers were playing on worker psychology to either threaten or distract them from whether or not they wanted a union. In other cases, the Board found employer actions to be “calculated to feed upon the employees' latent prejudices and to arouse resentment and antagonism against the Union, [to distort] the Union's policy of equality into a threat” (Pittsburgh Steamship Co., 69 NLRB 1395, 1414 [1946]). In this case, the employer used both physical and verbal threats throughout the campaign to play on the workers' opposition to integration. One manager told the workers, “I'm going to hire a big nigger to be your partner and the blacker the better.” A second manager told workers, “The CIO isn't going to last always, President Roosevelt isn't going to live always, and when he dies all the Jews, the God damned Jews are going to be out and we will have a different set-up.” Still later, the same official told workers, “if you do win the election, you are going to bring up a lot of goddam niggers from the coast, and they are going to put one in every room…. How would you like to eat and sleep with a nigger?” In this case and others like it, the Board found the comments to be threats by the employers that working conditions would worsen if a union came in. Employers were succeeding because their words resonated with the workforce, but the Board focused on the employer's prompting—and recognition that it was designed to defeat the union—and not the individual attitudes.
By the 1960s, Board members were coming to some consensus on how to interpret and respond to racism in union elections and in Sewell Manufacturing (138 NLRB 66, 67–68 [1962]) the Board set out a number of broad principles that have since become a standard or “doctrine” by which subsequent union racism would be judged and compared. Sewell involved an election in two Georgia towns where the union was soundly defeated because the employer seemingly linked the union organizers to the civil rights movement. Two weeks prior to the election, the employer showed workers a picture of the union president dancing with an African American woman, with a caption referring to it as “race mixing.” He pointed out that the union used membership funds to support various civil rights groups and told the workers “the unions…have tried to force (integration) down the throats of the people living in the South.” In overturning the election results, the Board made two statements about how the institutional context could impact the manifestation of union racism. First, it emphasized that a union election was different than a regular political election. Not all the participants were equal and the Board believed it had a responsibility to scrutinize speech and behavior by all of those involved to “insure that the voters have the opportunity of exercising a reasoned, untrammeled choice” regarding unionization. Because the employer controls the employment relationship and, in almost all circumstances, possesses more economic power than does the individual employee, the employer's words were to be considered imbued with a “force independent of persuasion” (NLRB v. Federbush Co., Inc, 121 F.2d 954, 957 [2nd Cir., 1941]). Second, the Board made a distinction between when racism was an acceptable and rational part of a campaign and when it was not. Racist language, race-baiting, or other forms of racist speech were potentially allowable and legitimate in union elections. “Some appeal to prejudice of one kind or another is an inevitable part of electoral campaigning.” It is only when the racist speech “can have no purpose except to inflame the racial feelings of voters in the election,” and particularly when the speech is being imposed in a manner to infringe upon the institutional mandate of the Board—to ensure that elections are independent of coercion—that it would find the speech actionable and overturn the election results. In this case, “it seems obvious from the kind and extent of propaganda material distributed that the Employer calculatedly embarked on a campaign so to inflame racial prejudice of its employees that they would reject the Petitioner out of hand on racial grounds alone.”
Union Racism—Focusing on a Harm Independent of the Act
Although the Board's early decisions dealt almost exclusively with employers' race-baiting in an effort to divide unions, 75% of the cases after 1966 would deal with accusations by the employer that the union had race-baited or that a union member committed a racist act. This immediately suggests an institutional dynamic at work. The Board's decision in Sewell was widely discussed in employer manuals; anti-union consultants told employers specifically that they could not race bait. And seemingly, employers stopped race-baiting, at least in the manner of the pre-Sewell cases. Interestingly, while the Board found the employer guilty of more than 80% of the cases in the first time period examined, it would only find the union guilty in 4 of the 33 cases that came after 1966. Instead of finding the racist act harmful and actionable as in many of the aforementioned cases, the Board repeatedly dismisses the union acts of racism in the latter cases as either incidental, harmless, or essential and rational to union mobilizing efforts. As the Board stated in Maple Shade Nursing Home, Inc. (223 NLRB 1475, 1483 [1976]), a case in which union members frequently belittled the employer's heavy “Jewish” accent, “union activities (are) not any form of tea party. The Union did nothing here to inflame irrational prejudices, and the employees laughed at their own jokes.” In Bancroft Manufacturing (210 NLRB 1007 [1974]), it dismissed the relevance of a union comment that a black worker who had been given a car by the employer was a “sold out soul brother,” whereas in another case where union supporters repeatedly called a man a “house nigger,” the Board found the comments to be “obvious ribbing” and an effort by the union to get the man to “abandon his servant type mentality” (Vitek Electronics, 268 NLRB 522 [1984]). When the Board in Beatrice Grocery Products (287 NLRB 302 [1987]) looked at a statement by a union representative that a supervisor had called the employees “dumb niggers,” it argued that the statement by the union member was made in order to confront racism, not create it.4
A dissenting Board member wrote in response, “The remark… was such that the employees were not likely soon to forget it…. The history of the term ‘nigger’ has rendered the use of it so opprobrious that it triggers instanter a whole complex of memories and resentments. We may as well ignore the devastating effects of a discharged firearm by describing the pull of the trigger as ‘isolated’ as pass silently by the effects the use of this single word is capable of causing.”
As long as the topic of discussion is “whether employees have been unfairly treated,” it is legitimate regardless of the racist content (
Coca-Cola, Inc., 273 NLRB 444 [1984]).
By focusing on the impact of racist words, the Board has argued that words themselves are not, by definition, harmful—they can be potentially neutralized by political or institutional context. In Foundry Div. of Alcon Indus (328 NLRB 129 [1999]), the Board argued that workers calling each other “nigger” while waiting in line to vote did not have an impact on their behavior because other workers immediately countered the comments. Whereas a federal court (United Packinghouse v. NLRB, 416 F.2d 1126 (DC Cir, 1969) held that racism in union campaigns inevitably led to docility and a demobilization of worker protest, the Board disagreed: racism and discrimination were political categories which could be mobilized and manipulated in a myriad of ways, some of which could be for the good. “A continued practice of discrimination may in fact cause minority groups to coalesce, and it is possible that this could lead to collective action with nonminority group union members” (Jubilee Manufacturing Co., 202 NLRB 272 [1973]). It reiterated this argument in Handy Andy (228 NLRB 447 [1977]), claiming that union racism may serve multiple purposes: “employers faced with the prospect of unionization will be provided and have been provided…an incentive to inject charges of union discrimination…as a delaying tactic in order to avoid collective bargaining altogether rather than to attack racial discrimination.” Moreover, the Board consistently looked at whether the statements were made as a part of electoral strategy or in isolation. In DID Building Services (291 NLRB 37 [1988]), the Board found that comments made in the heat of the moment were probably “discounted” by workers “as impulsively made.” Comments that were “vile and seething with prejudice” were considered isolated and irrelevant to the election campaign. The point here is not to defend or legitimate racist acts and practices, nor to disagree with those that argue words alone can “wound” (Feagin 1991; Mackinnon 1993; Matsuda 1993). It is to argue that such an act is fundamentally situated within a broader set of politics and can only be understood within this context before it is deemed actionable.
When the Board has found union racism to be worthy of overturning an election victory, it has involved situations where the racism replaced the political confrontation as the primary focus of debate and where union leaders clearly acted strategically in placing the race issue on the agenda. Two contrasting cases provide an example. For instance, in YKK (USA) (269 NLRB 82 [1984]), the Board confronted a situation where the comments were clearly made by union leaders and became a centerpiece of the union's campaign. Union leaders at a Japanese-owned zipper company passed out campaign literature that made repeated derogatory references aimed at the owner's nationality. At a union meeting shortly before the vote, the union's national representative told the workers to stick together against the “Japs,” ending his speech with words to the effect that “we beat the Japs after Pearl Harbor and we can beat them again.” Later, this same union officer shouted at a Japanese engineer of the company, “[t]here goes one of those damn Japs. Go back where you came from, you damn Jap,” while the union vice president wore a t-shirt with the phrases “Japs go home,” and “slant eyes.” The Board held that the racism in this case was distinguishable from past cases because “[t]here is no conceivable way that a reference to beating ‘Japs’ at Pearl Harbor could be relevant to a legitimate campaign issue.” Because the union made race the center of the campaign, the Board viewed the racism as unconnected to legitimate worker concerns and held that it served no purpose beyond being inflammatory and illicit and an effort to mobilize workers around their racism towards the Japanese. In contrast, the Board allowed an election to stand in KI (USA) (309 NLRB 1063 [1992]), a case where union members again attacked the Japanese company owners, both with private jokes among employees and by disseminating and attacking a letter that it claimed to be from the company's president as an example of the Japanese “screwing us over.” The disseminated letter stated: “I am appalled at the typical lazy, uneducated American worker…. I suggest the Americans start developing a healthy respect for Japan because one of my colleagues will eventually become your boss.” The Board distinguished this case from YKK, arguing that “notwithstanding any racial overtones, the topic of how American workers were regarded by management was a relevant campaign issue.”
[T]he context was that the employees were concerned about the impact of the attitudes of the Japanese owners on their workplace. Thus, (dissemination of the letter) appears to be an attempt at least to pose the question of whether there is some connection between the two. It does not automatically follow that this communication is inherently objectionable. Although such claims raise the specter that some voters may overreact and respond in an equally prejudicial manner…the Board has not equated the broaching of such topics to opening a Pandora's box.
Contrasting Racism as Mobilizing versus Racism as Dividing
The Board has responded in an entirely different manner to cases where employers have accused minority workers of using race as a mobilizing tool in their election campaigns. In Aristocrat Linen Supply Co. (150 NLRB 1448 [1965]), African American workers used civil rights appeals to promote worker solidarity against the employer. The union passed out a flier to a predominantly black workforce that ended with the statement “this is why the labor hater is always a twin-headed creature spewing anti-Negro talk from one mouth and anti-propaganda from the other.” Union leaders later exhorted workers not to be a “Handkerchief head Uncle Tom.” The Board, while finding the campaign rhetoric “undeniably based upon a racial issue,” argued that “a distinction must be drawn between racial propaganda designed to inflame racial hatred and set the tone of a union campaign as a battle of one race against another as in Sewell, and racial propaganda designed to encourage racial pride and concerted action.” The same year, in a case of similar circumstances, the Board again contextualized race and racism within the political battle, allowing for multiple ways in which race-specific campaigns can be used: “An appeal to racial self-consciousness may produce a variety of emotions, depending upon the context. In some cases, such appeals may result in vicious race hatred. In another circumstance, such appeals may promote reasoned and admirable ambition in an unfortunate race of people” (Archer Laundry, 150 NLRB 1427 [1965]). In a later case where the union mobilized around race issues, Baltimore Luggage (162 NLRB 1230, 1233 [1967]), union organizers told black workers that they received lower pay than white workers because of their race. Again, the Board distinguished between the irrational use of race language in Sewell and the arguably rational way in which it was presented here: “In Sewell, we did not lay down the rule that parties would be forbidden to discuss race in representation elections.” The Board argued that unions could make race-specific appeals when they are used to promote the rights of disadvantaged groups in their quest for economic empowerment: “campaign material of this type is directed at undoing disadvantages historically imposed [generally unlawfully] upon Negroes because of their race, through an appeal to collective action of the disadvantaged. The choice of racial basis for concerted action has been made, not by the victims who organize to seek redress, but by those who use race as a basis to impose the disadvantage.”
In Carrington South Health Care Center (1994 NLRB Lexis 397 [1994]), a largely African American workforce was given three cartoons by union leaders designed to encourage their support for the union. Two cartoons showed a clearly white owner either exploiting or enslaving a clearly black employee, while the third showed a white “boss” directing a nervous looking black employee to an electric chair, stating “You don't need your union rep. Just have a seat and we'll discuss your grievance like two rational human beings.” The Board found these race-specific cartoons to be appropriate for an election because they reflected the benefits of being in a union—in fact, the Board argued that these were not race-specific appeals at all, but simply a form of contestation over economic concerns. In Bancroft Manufacturing Co. (210 NLRB 1007, 1008 [1974]) the Board dealt with a case in which a black union organizer told workers to stay in solidarity because individual black workers were being bribed with new cars, and because if the union lost, “all blacks would be fired.” Here the Board doubted whether the union would make strategic use of such racially specific comments because the workforce was nearly 60% white. Since it would be “suicidal” to play the race card in this way, and since the use of race during the campaign stressed “black pride, the past history of discrimination against blacks in American society or the present disadvantaged status of blacks as a class,” it found the comments to be a legitimate part of the campaign discourse. In 1998, as race mobilization cases became more and more frequent, and as employers continually objected to their use by referring to the Sewell doctrine, the Board's General Counsel, William Gould, proposed a new doctrine that would be used to distinguish the racial mobilization cases from other racist acts in union campaigning:
Because the employer controls the employment relationship and…possesses more economic power than does the individual employee, the Board's concerns about racial appeals expressed in Sewell…have peculiar applicability to remarks of employers as opposed to those of unions and their representatives…. Union organizational efforts aimed at blacks and other racial minorities and women must necessarily focus, in part, upon grievances peculiar and unique to such groups, i.e., employment conditions which are attributable to racial inequities or what appear to be racial inequities and other forms of arbitrary treatment (Shepherd Tissue, 326 NLRB 369 [1998]).
As we will see, although this may seem an unsurprising interpretation by the Board given the comparison with the cases that it has found objectionable, the race mobilization cases will provide one of the most dramatic discrepancies between the Board's understanding of race and the understanding of federal courts.
Collective Action
We saw previously how the Board scrutinized the institutional dynamics that led both sides to attempt to shift the focus of a union campaign from politics and economics to race. Another way in which the Board scrutinizes the institutional context of racist acts involves its examination of collective action concerns within a union organizing drive. Collective action problems in unions are one of the most fundamental ways in which labor law is different from the types of cases that ordinarily appear before federal courts. As mentioned earlier, unlike most other realms of law, the individual is not at the center of labor law. Unions exist because workers agree to limit their individual opportunities in the effort to benefit as a group. As a result, unions consistently confront the difficulty of maintaining the support of potential “free riders” who may choose to reap the benefits of the union without participating in the costs of its formation and maintenance (Olson 1971). The NLRA is cognizant of this and there are numerous statutory ways that unions can discipline individuals. As the Supreme Court wrote in NLRB v. Allis-Chambers Mfg. (388 U.S. 177 [1967]):
National labor policy has been built on the premise that by pooling their economic strength and acting through a labor organization freely chosen by the majority, the employees…have the most effective means of bargaining for improvements in wages, hours, and working conditions. The policy therefore extinguishes the individual employee's power to order his own relations with his employer and creates a power vested in the chosen representative to act in the interests of all employees.
The nature of collective action, and the fact that unions are designed to promote a “public good” to overcome economic inequalities, creates further problems for the organization in maintaining internal hierarchies and leadership (Levi
2003; Mansbridge
1986; Polletta
2002). Unlike a company that is run by a “boss” or “CEO,” union hierarchies are relatively fluid and democratic. The power of union leaders to keep members disciplined is less fundamental than that of a CEO, and as a result members are not controlled by leaders when they wish to “speak” on behalf of the union. Particularly in union organizing battles where many of the union supporters are not official union members until after a certifying election, unions face significant problems in maintaining coordination and a unified message. As a result, labor law has provided unions with different opportunities to promote a group identity in spite of collective action problems. Labor law allows unions (at times) to impose closed shops, to punish members (within limits) for refusing to follow majoritarian decisions, and generally to prevent its members from dissenting and abstracting themselves from the union decision-making process. Today, although some of these opportunities have been weakened or even taken away, the NLRB continues to recognize the collective action problems that underlie union leadership and action.
More specific to questions of handling and understanding racism, the Board has argued that it will only find a labor union accountable for a racist act of a union supporter or member if the actor is deemed an official leader and authorized to comment on the union's behalf. Other individuals, even strong union supporters, are generally deemed beyond the union's control and responsibility. As a result, the Board applies a less rigorous “third party” standard where the union leadership cannot be connected to the racial statements made by non-union leaders—even if they are employees who are close to the organizing campaign. Similar to the other examples mentioned earlier, then, the Board's separation between leadership and members reflects that not all racist acts are equal as well as recognition that institutional relationships will otherwise make it more likely that a union member is involved in a racist act than a member of the employer's staff. That the cases post-1966 have been so heavily dominated by accusations against the union is arguably reflective of this dynamic. Since the late 1960s, employers have increasingly relied on “union-busting” consultants that specify what employers and their managers can and cannot say and do during a union campaign. Employers simply have more hierarchy and discipline over their managers and supervisors—they can fire managers, for instance, and suppress dissent—than unions who by law must protect employee speech and dissent, are limited in the forms in which they can discipline their members, and powerless against union “supporters” who are not currently members.
The cases reflect recognition of this political and institutional inequality. For instance, in Zartic, Inc. (315 NLRB 495, 500–508 [1994]), the Board found that statements by a union member linking the employer to the Ku Klux Klan(KKK), while baseless and designed intentionally to “exploit the ethnic fears of the Hispanic employees by making a visceral connection between the KKK and working conditions,” were nonetheless not liable under the Sewell doctrine because the union lacked control over the individual. The employer had a “stricter burden of proof…to establish that the conduct of third parties was of so serious a nature.” The Board similarly discounted comments by African American organizers (“Boy, you white sons-of-bitches, you are all the same, you're scared to take a stand”) as being outside of the authority of the union leadership (Herbert Halperin Distributing Co., 1968 NLRB 247 [1986]) and, in Air Express Int'l Corp. (289 NLRB 608 [1988]), argued that when a pro-union employee told others of the employer's dislike of Cubans, that the comment was not a “systematic attempt (on the part of the union leadership) to inject the ‘racial’ issue into the campaign; but that the employees probably had blown the statement out of proportion in the retelling of it.” Many other cases, meanwhile, dealt with rumors that workers spread during the course of the election campaign. In one, a disagreement between a labor board member and the employer's attorney was misrepresented by workers as a disagreement over whether workers would be allowed to speak Spanish at work, and thus on the eve of the election, workers discussed widely the belief that the employer was anti-Latino (Singer Co., 191 NLRB 179 [1971]). Though it injected racial animus into the campaign, it was deemed outside the responsibility of any union actor. More typical of the rumor cases was Information Magnetics (227 NLRB 1493 [1977]), where workers spread rumors that the employer had brought in the Immigration and Naturalization Services to deport union supporters who were illegal immigrants. Again, although the rumor clearly had an impact on the election, the Board held that the rumors were not controllable by union leaders. In a contrasting case, when a union leader threatened a worker with deportation, the election was overturned (Professional Research, Inc., 218 NLRB 96 [1975]).
But even when union leaders are involved in the racist action, the Board has scrutinized their participation and the degree to which they officially spoke for the union and which other union leaders had control over the individual. In Benjamin Coal (294 NLRB 572 [1989]), where the Board dealt with comments made by members of the union organizing committee, it argued that because the committee was made up of volunteers and was open to any employee who wished to join, the union leadership could be distinguished from its organizers. It held that the union did not “echo or condone these highly offensive sentiments.” When union leaders heard antisemitic statements during the campaign coming from a union organizer, “the organizers immediately quieted (him) and told the audience that such comments were irrelevant to the campaign.” Although the Board made clear that racist statements had no place in a campaign, “To hold that the election was tainted by such prejudice would be to hold that no election could ever be held in any plant with a prejudiced work force unless the union attempting the campaign were able to accomplish what management itself had been unable to do before the union came on the scene, namely, eliminate all expressions of racial, ethnic, or religious bias.” In Pacific Micronesia Corp. (326 NLRB 458 [1998]), meanwhile, the Board, in refusing to overturn the election results, pointed to the fact that the president of the union refuted statements made by a union organizer to the effect that the company was hiring Nepalese workers to weaken the strength of the company's Filipino workers.
Federal Courts: the Inherent Damage and Irrationality of the Racist Act
After the Board makes the initial decision on election conduct, either side can appeal to a federal appellate court. Sometimes there is no appeal and sometimes the appeal centers on a different accusation of unfair labor practices—in many of the cases discussed here, racism is just one of many changes that the Board and courts are asked to deal with. When a Board ruling on race is appealed, more often than not, federal courts defer to Board decisions, reflecting the deference they generally give to administrative agencies. Nonetheless, federal courts have differed in significant ways with the Board's interpretation of racism in union elections, and the manner of this clash is theoretically illuminating, as it has reflected a very different understanding of racism, one that parallels the individual-prejudice model so widely endorsed in political science. When federal judges object to NLRB decisions, it is consistently on the same grounds; that racism is itself the harmful act and that it is irrelevant whether the appeal to racial passions was made by the employer with the goal of dividing the workforce, or by the union with the goal of enhancing solidarity, or whether it was made by a worker who was not a member of the union leadership. To these judges, racist acts are never to be tolerated and are always the responsibility of the individual who carried them out. Federal judges have argued that racism is intrinsically irrational and thus can never be understood differently regardless of the context and institutional dynamic in which the racist act was situated. Prejudice trumps institutional considerations.
An emblematic example of federal court interpretation of union racism is when the Sixth Circuit (76 F.3d 802, 807 [1996]) reviewed and overturned the Carrington case discussed earlier. The Board had found the acts to be a way of mobilizing disadvantaged workers around issues that intersected race, class, and power, and argued that they were “devoid…of appeals to racial bigotry.” The federal court disagreed, holding that the racially specific cartoons that identified African American workers being exploited by their employer were used deliberately by the union to exacerbate racial feelings with irrelevant and inflammatory appeals. Although two of the cartoons made a passing reference to legitimate campaign issues, the judge found the imagery to be “quite troubling” and a “graphic appeal to racial prejudice.”
Each cartoon uses obvious images of bondage or violence visited upon racial minorities by a white majority: a white man purchases a group of black (or mostly black) workers; a group of workers labor as beasts of burden, pulling their superiors in a wagon while being whipped; a black worker is to be summarily executed by a white overlord….the cartoons could therefore be construed as a deliberate exacerbation of racial feelings by irrelevant and inflammatory appeals.
Other federal courts have reacted in like fashion, responding to the inherent irrationality of the act and not the political context in which the manifestation occurred. The Sixth Circuit overturned
KI Corp., as discussed earlier (where the union pointed out a racist letter written by a Japanese business owner intended toward the employees), because the “negative stereotyping…has [no] legitimate place” regardless of context and the use of the letter “exceeds the bounds of legitimate discussion” (
KI Corp v.
NLRB, 35 F.3d 256 [1994]). The Fourth Circuit in
NLRB v.
Schapiro & Whitehouse, Inc. (356 F.2d 675 [1966]) found that a union leaflet that pointed out that a union would help solve racial discrimination in employment was “deplorable” and “highly inflammatory” speech. The court wrote that the “equality of race [was] not presently an issue. That the majority of the employees were Negroes did not make it so.”
Union antisemitism has been the issue of a series of cases where federal courts have overturned Board decisions. In one, where the employer was compared to Hitler, the Fourth Circuit overturned the election on grounds that the union had “interjected into the election one of the most sordid episodes in modern history” (Schneider Mills, Inc. v. NLRB, 390 F.2d 375 [4th Cir., 1968]). The Third Circuit (NLRB v. Silverman's Men's Wear, Inc., 656 F.2d 53 [3rd Cir., 1981]) overturned the Board in a case involving anti-Jewish rhetoric when a union representative called the company's vice president a “stingy Jew.” The court held that appeals to racial prejudice constituted a prima facie warrant for setting aside an election. Unlike the Board, then, the explicit assumption of the court was that racism had no economic or political component that in any way could be associated with an organizing drive. In this case, the court found “such a remark has no purpose except blatantly to exploit religious prejudices of the voters. It deliberately injects ‘an element which is destructive of the very purpose of an election.’…We can see no reason for the remark except to inflame and incite religious or racial tensions.” Similarly, in Katz v. NLRB (701 F.2d. 703 [7th Cir., 1983]), the court overturned the Board for not finding antisemitic statements to be problematic: “There is no conceivable way in which either the movie “Holocaust” or the Nazis' treatment of Jewish people during World War II could be relevant to a legitimate campaign issue. To the extent that the priest's comment regarding the wealth of Jewish people, as juxtaposed to the poverty of the employees, might be relevant to the campaign, the point could have been made without resort to a religious slur.”
Federal courts have also been far more likely to discount consideration of collective action issues as has been done by the Board. In the aforementioned Katz decision, the court dismissed the Board's determination that the statement had been made by a priest with no direct affiliation with the union. The court held that the collective action questions raised by the Board were not determinative; all that mattered was whether the statements were racist and provocative and hence overturned the union victory. The Eleventh Circuit (M&M Supermarkets v. NLRB, 818 F.2d 1567 [1987]) made no distinction between the union and its leadership and an “outspoken union supporter and advocate” who said in support of the union: “the damn Jews who run this Company are all alike. They pay us pennies out here in the warehouse, and take all their money to the bank…. Us blacks were out in the cotton field while…the damned Jews, took their money from the poor hardworking people.” The Board had concluded in contrast that he was not a union agent, nor that there was any evidence that the union authorized or was even aware of his actions. The court however stated simply that “such feelings [have] no place in our system of justice.” Similarly, in NLRB v. Eurodrive, Inc. (724 F.2d. 556 [6th Cir., 1984]), the court overturned a Board's decision that had found a union not to be responsible for an organizer's statements, nor responsible for what it perceived to be racial tensions that existed in the workplace before the beginning of union organizing. While the Board found the comments made by the employee to be “intemperate, abusive, and inaccurate,” the court argued that the organizer's attenuated link to the union was less relevant than the racial statement itself, and that the statement had exacerbated racist tensions and “had an appreciable effect upon the employees' freedom of choice.”
CONCLUSION
I have argued that the Board's approach to racism in union elections provides theoretical insights into a more complicated and complete understanding of the dynamics that lead individuals to commit racist acts. The NLRB's approach, by scrutinizing context, differentials in power, and the particular use of agenda setting and collective action issues, places the individual act within a broader political and institutional setting. It highlights how institutions impact actors, biasing the ways in which racism is manifested and the ways in which it is hidden. To have a fuller understanding of racism, then, we need to place it in a broader political-institutional context in which individuals act strategically to pursue goals.
I do not wish to argue that the Board's approach is the only explanation of racism, nor that its holdings in each case are always “correctly decided.” The Board's approach is both novel and an important addendum to those who too often limit their understandings to individual psychology and attitudes. But the Board's particular application is not the only way to interpret racism within a political and institutional context. The Board members themselves are institutionally limited in the ways that they can interpret and address discrimination in unions. The development of labor institutions in the United States occurred at a time when the dominant national unions of the American Federation of Labor were overwhelmingly white and male. Key union leaders pushed successfully for provisions of the NLRA that enabled them to maintain racial hierarchies and these provisions were supported by southern Democrats in Congress (Frymer 2003). It took three decades after its creation, moreover, for the NLRB to sanction a union on grounds that it was discriminating in the hiring and representation of minority workers, in part because the Board believed it had previously lacked the statutory power to do so. The NLRB's “tardy assumption of jurisdiction” over fair representation issues involving race led the Supreme Court to refuse to allow the Board to maintain sole control over labor civil rights matters, particularly involving racial discrimination under Title VII (Vaca v. Sipes, 386 U.S. 171, 183 [1967]). The Board has also been largely absent from handling widespread discrimination on the basis of nonracial characteristics such as gender and disability, again in part because it does not see itself as an institutional actor designed to handle such matters (indeed, with the exception of a broad duty of “fair representation” to all workers, there are no specific provisions in labor law on topic).
Perhaps symbolic of the Board's institutional limits in handling race-specific discrimination is a case where it defended the principle of a union's right to have exclusive representation in the workplace. In doing so, it allowed the firing of a group of African American workers who protested their company's discriminatory employment practices and what they perceived to be the unwillingness by their union to adequately represent them (The Emporium and Western Addition Community Organization, 192 NLRB 173 [1971]). The Board thus confronted a conflict between principles of civil rights and union authority and since the NLRA emphasizes the latter over the former, it is perhaps not surprising that the Board found on behalf of the union. The Board held that the union was the only representative that could bargain with the employer under labor law; because the workers protested the company's civil rights policy independent of the union's support, they were legitimately fired from their jobs for participating in an illegal work stoppage. The Board was initially overturned by the D.C. Court of Appeals on grounds that labor laws should not intervene when principles of racial equality are at work. Civil rights, the federal court argued, trumps considerations of union power. But Thurgood Marshall, writing for the majority of the Supreme Court (Emporium Capwell v. Western Addition Community Organization, 420 U.S. 50, 67 [1975]), sided with the Board. Whereas the appeals court had argued that confronting racism should not be obstructed by labor law statutes, Marshall countered that questions of race and labor power were inseparable: “Competing claims on the employer's ability to accommodate each group's demands…could only set one group against the other. Having divided themselves, the minority employees will not be in position to advance their cause unless it be by recourse seriatim to economic coercion, which can only have the effect of further dividing them along racial or other lines.” Limiting union rights, Marshall argued, would undermine the ability of unions and their workers to fight for civil rights and thus hurt their efforts at ending racial discrimination in the work-place.
Emporium Capwell was in many ways emblematic of the debate between the individual and institutional approaches, as well as suggestive of the problems of both the federal court and the NLRB approach to understanding racism. As developed, labor law and civil rights law in America have suggested either/or alternatives, and both have had great difficulty in incorporating racial and class inequality into one regulatory body and one legal understanding (Forbath 1999; Frymer 2004; Iglesias 1993; Katznelson 1989; Klare 1982). The election cases discussed earlier provide some suggestive ways in which this intersection might be accomplished—but this is only meant as a starting point, and future work in applying a political approach to these questions must go further in intersecting questions not just of race and class, but of gender, sexuality, and other existing dynamics of inequality (for very good starts at this, see Cohen 1999; Warren 2004; Young 1990). In particular, it is at the intersection of issues involving the contestation between marginal groups that the individual model of prejudice becomes most wanting. Although historically varied, institutions have quite often created opportunities for powerful actors to benefit from those less powerful being pitted against each other in sites where these intersectional conflicts are most visible. It is when conflict between less powerful groups is most intense that an institutional explanation can allow us to step back and see such motivations and behavior in a broader context of power, rationality, and structure (Piven and Cloward 1978).
To see racism institutionally, in turn, suggests a further research agenda for political scientists beyond the sphere of labor union dynamics. Political-institutional arguments have been made extensively in non-race-specific spheres of politics, where rational choice and new institutional scholars have examined the motivations behind the decision making of individual actors. The assumption in these studies is that psychology is more or less irrelevant to understanding individual behavior. All actors are assumed to be rational, informed, and to act according to the incentives that institutions provide. But in the realm of behavior deemed “irrational”—violence, prejudice, collective action—scholars far too often attempt to explain the phenomenon solely in psychological ways, ignoring how individuals and their leaders are often motivated by the same political and institutional understandings that motivate members of Congress, executives, and interest groups (for notable exceptions to this, see Chong 1991; Robin 2004). The consequence is that racism becomes understood as an innate ‘evil’ that works in the underbelly of society—it is removed from politics and we thus lose focus of the myriad ways in which strategic politicians and institutions can both promote and prevent racist activity.
A limited number of studies have begun to examine how institutional dynamics influence elite handling of race and racism, focusing on the incentives followed by elected officials that lead them to employ specific types of race strategies (e.g., Fraga and Leal 2004; Frymer 1999; Kim 2001; Walters 1988). This is particularly the case where race intersects with congressional representation, in which scholars interested in strategic behavior view the manipulations and treatment of race as part of normal politics (e.g., Canon 1999; Tate 2004). Among racial attitude scholars, there has been an increasing effort to situate racist acts within institutions. Both James Glaser (2002) and Tali Mendelberg (2001), for example, argue that white attitudes are often ambivalent and individuals are heavily influenced by ballot structures, party leader decisions, and other institutional dynamics. Yet other scholars have attempted to look at broader socioeconomic context to understand the moments of violence and racial protests of the post–Civil Rights era (e.g., Olzak 1992; Sears 1994). And one need not exclude psychological dynamics to incorporate an institutional-political understanding. Psychologists have provided nuanced evidence that shows how almost anyone can be induced to follow orders or change their behavior given a specific institutional context, whether through simple peer pressure or an effort to follow structured rules (e.g., Asch 1958; Milgram 1974; Zimbardo 1973). It is hoped then that by recognizing that racism, like other behaviors in society, can be analyzed as a political act, we will begin to provide a more complete account of why it remains a far too meaningful and widely used form of combat.
Thanks go to Amy Bridges, Michael Brown, David Canon, Kerstin Carlson, John Coleman, Zoli Hajnal, Kim Johnson, Tom Kim, Ari Krantz, Cathie Lee, Michael McCann, Eileen McDonagh, Tali Mendelberg, Shehzad Nadeem, Corey Robin, Mark Sawyer, Pam Singh, John Skrentny, Rogers Smith, Julie Suk, Rick Valelly, Dorian Warren, and Lee Sigelman. The Law and Public Affairs Program at Princeton University and the University of California's Institute for Labor and Employment provided generous financial assistance.
%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
Comments
No Comments have been published for this article.