Copyright (c) Henry H. Perritt, Jr
Kansas Law Review November, 2000

Introduction:

The National Labor Relations Act (NLRA) ¹ is the basic federal statute governing the relationship between labor unions and employers, other than railroads and airlines, ² in the private sector. The NLRA dates from 1935 ³ and its principal provisions were last amended in 1959. ⁴ It is premised on  [*2]  a workplace with the following characteristics: a fixed location or, in the case of mobile employees such as truck drivers, at least a fixed central location to which employees physically report; where communication among employees or between employees and management is person-to-person, either face-to-face or through written instrument or telephone; and commerce with consumers is largely through face-to-face transactions. In sixty-five years' experience with such traditional workplaces, the National Labor Relations Board (NLRB or the "Board") and the courts have applied the statute to develop a legal regime governing groupings of employees that are appropriate units for collective bargaining; ⁵ the rights of employees to solicit coworkers for union activity and the privilege of management to curtail such solicitation; ⁶ the rights of unions to have access to employees on employer property; ⁷ limitations on employer access to employees during nonworking time for purposes of campaigning against union representation; ⁸ and subjects that employers and unions are required to negotiate. ⁹ The NLRB and the courts also have developed a legal regime regulating each side's resort to economic weapons to pressure the other side. ¹⁰ The economic weapon which has received the most regulatory attention is picketing, ¹¹ characterized by physical confrontation between picketers and customers, employees and other service providers. ¹²

Today's workplace is becoming increasingly electronic. A growing number of employees telecommute or otherwise report electronically, instead of reporting physically to a fixed location. ¹³ Communication via electronic mail and other systems whereby the recipient controls when he or she receives and responds to the message is commonplace. An in-creasing amount of consumer commerce is taking place over the Internet. ¹⁴ Although most Internet commerce currently consists of orders placed electronically but filled via traditional delivery methods, thousands of consumers now order computer software, books and music compact discs and have the orders filled electronically over the Internet.



Only recently has the NLRB begun to confront issues requiring it to adapt the NLRA's legal regime to electronic workplaces. The NLRB has faced questions of the degree of statutory protection to give employee use of e-mail, ¹⁵ the appropriate bargaining unit where no employees have a fixed work location and all report electronically, ¹⁶ and union access to the employees in a virtual workplace. ¹⁷ In the future, the NLRB will have to  [*4]  consider such questions as whether an employer may approach employees in their homes when their homes are also their workplaces, the scope of bargaining over such electronic issues as telecommuting, ¹⁸ and the types of economic pressure that may be applied to completely electronic companies.



This Article takes the first steps in developing a framework for adapting the National Labor Relations Act to electronic workplaces. It focuses on the earliest stages of the employer-union relationship, when a union seeks to organize an employer's employees and establish itself as the employees' exclusive bargaining representative. The Article explores the continued utility in electronic workplaces of established NLRA doctrines, developed in traditional workplaces, governing the units of employees appropriate for collective bargaining and access to employees to solicit their support for, or opposition to, union representation. It also examines the developing potential for the use of electronic media to assess employee preferences regarding union representation.



Adapting the NLRA to electronic workplaces requires an under-standing of the basic policies embodied in the statute. Therefore, Part II sets forth the sometimes conflicting policies that underlie the National Labor Relations Act. Adapting the NLRA also requires an appreciation of the differences between electronic and traditional workplaces. Part III discusses these differences, deriving them from a comparison of traditional and electronic communities. Part IV draws on the analysis presented in Parts II and III to develop an approach to defining appropriate bargaining units in electronic workplaces. Part V develops an approach to employer, employee, and union solicitation of employees in electronic workplaces. Part VI considers the use of electronic union authorization cards and electronic ballots in representation elections. Part VII concludes with some observations concerning the adaptation of the NLRA to electronic workplaces.  

II. Policies Underlying the NLRA



Many consider the New Deal as the beginning of the era of "big government," i.e., major government intervention in the economy and regulation of business. Indeed, the economic philosophy of the admin-istration of President Franklin Delano Roosevelt was to avoid another depression by spreading the wealth among a greater number of people. This philosophy was manifested in the Social Security Act of 1935, ¹⁹ which established old age insurance and provided incentives for states to establish unemployment insurance systems, and the Fair Labor Standards Act of 1938, ²⁰ which established minimum wage levels and required premium pay for hours worked in excess of a stated maximum.



The NLRA, however, was a relatively conservative piece of legislation. It did not radically alter the capitalistic nature of business by having the government dictate terms and conditions of employment. ²¹ Instead, the underlying philosophy of the NLRA was that guaranteeing workers the right to organize and bargain collectively, and giving their collective representative the exclusive right to negotiate for all employees in the bargaining unit, would alleviate the imbalance in bargaining power between workers and their employers. It was expected that the resulting freely negotiated agreements would improve wages and working conditions, thereby spreading the wealth further and improving the economy. The goal was private ordering, rather than substantive government regulation.



The NLRA's conservatism also was reflected in the absence of organization of employees along class lines. Rather, the private ordering of workplaces that the statute embraces occurs at relatively local levels. In this representation system, workers are organized based on their  [*6]  community of shared interests. Bargaining units defined in terms of relatively small communities of interest mitigate concerns that unionization is the precursor of class warfare. The workers of the world are not uniting

-only the members of a relatively small group within one workplace are collaborating.



Another major goal of the NLRA was to reduce industrial strife substantially. The statute was designed to accomplish this goal in several ways. First, the statute created a procedure whereby workers could compel employers to recognize and bargain with their unions, thereby eliminating the need to resort to strikes and other tools of economic warfare to compel recognition. Second, through unionization, workers would achieve a voice in the workplace. This coming of industrial democracy was envisioned as itself reducing industrial strife. Third, the mandate that employers bargain exclusively and in good faith with unions selected by a majority of employees, and the backing of that mandate with the legal ability to resort to economic weapons, was seen as most likely to lead to peaceful agreements. Thus, the protection afforded the right to strike is somewhat paradoxical. The statute protects the right to strike on the assumption that the desire to avoid strikes and other economic warfare will provide the incentive for both sides to reach agreement peacefully.



The NLRA was enacted to improve the capitalist system. Consequently, capitalism's basic assumptions concerning private ownership of property became an implicit part of the statute. As a result, several rights of employers, although not expressly provided for in the Act, have been read as essential ingredients in the statutory scheme.



Chief among these are the employer's property rights. The Supreme Court articulated the statute's foundation of respect for property rights in NLRB v. Babcock & Wilcox Co.: ²² "Organization rights are granted to workers by the same authority, the National Government, that preserves property rights. Accommodation between the two must be obtained with as little destruction of one as is consistent with the maintenance of the other." ²³

The clash between employer property rights and employee rights to organize manifests itself most vividly in cases involving union access to employees on employer property. When the individuals seeking access are not employees of the employer, they are seeking to trespass on the employer's property. The employer's basic property right to exclude trespassers outweighs employee organizational rights unless the locations of the employer's property and the employees' living quarters place the employees beyond the reach of reasonable efforts by the union to communicate with them. ²⁴



Employees seeking to communicate with coworkers concerning union activity are invitees on the employer's property. Consequently, the employer's property right to exclude trespassers is not at issue. Other aspects of the employer's property rights, however, are at issue. As the Court has characterized it, employee access to coworkers on employer property requires "an adjustment between the undisputed right of self-organization assured to employees . . . and the equally undisputed right of employers to maintain discipline in their establishments." ²⁵ Consequently, employers may prohibit solicitation during working time but may not prohibit such conduct during nonworking time, absent special circumstances which justify the prohibition as necessary for the efficient operation of the business. ²⁶



Closely related to employer property rights are what we shall term employer entrepreneurial rights, i.e., rights to engage in basic business decisionmaking. Entrepreneurial rights recognize an employer's interest in deciding what uses to make of its property. The most fundamental example of such rights is the right to decide whether to be in business at all. In Textile Workers Union v. Darlington Manufacturing Co., ²⁷ the Court made clear that the NLRA was premised on respect for such a right. The Court unequivocally rejected the union's contention that the NLRA rendered illegal a complete closing of a business motivated by an anti- union animus: "A proposition that a single businessman cannot choose to go out of business if he wants to would represent such a startling  [*8]  innovation that it should not be entertained without the clearest manifestation of legislative intent or unequivocal judicial precedent so construing the Labor Relations Act. We find neither." ²⁸



Consequently, an employer has an absolute right to go completely out of business that always outweighs employee rights to organize and bargain collectively. In less extreme circumstances, the NLRB and the courts have had to balance employee organizational rights against employer entrepreneurial rights. Such balancing characterizes, inter alia, cases involving the degree to which a successor employer is bound by the relationships between the predecessor employer and its employees' union, ²⁹ and the degree to which the statute obligates an employer to negotiate over basic business decisions which necessarily and directly affect job security. ³⁰



Thus, the NLRB and the courts have had to resolve conflicts between the NLRA's express policy of guaranteeing workers' rights to organize and bargain collectively and the implied policy recognizing employer property and entrepreneurial rights. Policy conflicts also have arisen as a result of the manner in which the statute has been enacted and amended.



As first enacted in 1935, the NLRA established a policy to encourage collective bargaining. The original Wagner Act ³¹ was markedly pro-union. Indeed, one of its most controversial elements was its regulation of employer conduct with no comparable regulation of union conduct. ³² In  [*9]  1947, however, the tide turned against organized labor with the enactment of the Taft-Hartley Act amendments to the NLRA. ³³ The Taft-Hartley Act, however, did not repeal the policies of the Wagner Act. Instead, it layered a new level of restrictions on NLRB encouragement of unionization and on union activity. Consequently, the NLRB and the courts have had to reconcile Wagner Act language designed to encourage collective bargaining with Taft-Hartley Act language designed to restore balance to national labor policy.



For example, the Wagner Act provided that representation of employees would occur in bargaining units, i.e., portions of an employer's workforce who shared common interests in wages and working conditions. The statute directed the Board to decide in each case which bargaining unit was appropriate for collective bargaining in a manner that would "assure to employees the fullest freedom in exercising the rights guaranteed by this subchapter." ³⁴ The Board took this mandate seriously and, in defining a bargaining unit, placed great weight on the extent to which the union had succeeded in organizing the employees. ³⁵



In the Taft-Hartley Act, Congress reacted to this approach, but did not alter the language of the Board's mandate to define bargaining units to assure employees the fullest freedom under the Act. Instead, Congress  [*10]  layered on top of the statutory bargaining unit considerations a requirement that "the extent to which the employees have organized shall not be controlling." ³⁶ Consequently, the Board and the courts have had to reconcile these two seemingly conflicting mandates concerning bargaining units.



The Board and the courts have reconciled these competing policies in the context of workplaces dominated by fixedlocations and direct person-to-person communication. As more and more communication in the workplace becomes electronic and as more employees punch in remotely to virtual workplaces, the continuing validity of the policy resolutions developed in the context of traditional workplaces must be re- examined. Before such a re-examination can occur, however, it is necessary to examine the differences between traditional and electronic workplaces. Key differences arise from the different nature of traditional and electronic communities. Therefore, the next part explores the nature of electronic communities in general and electronic workplaces in particular, and how they differ from traditional communities and workplaces.



III. Traditional vs. Electronic Workplaces



Although the concept of community is important in labor law, its relevance to the workplace antedates formal legal regulation of labor-management relations. Bargaining and unions themselves arose from communities of workers long before the law responded by creating frameworks for both.



Community is a practical rather than a purely legal concept. Communities exist when their participants are interdependent, when the communities address important participant needs, when there is a psychological or ideological commitment, and when the attachment of participants is not completely transitory. Collective bargaining works only when a group of employees participating in collective bargaining experience social and economic forces that make them stick together. No amount of legal or institutional structure will make collective bargaining work if these social and economic forces are absent.

 [*11] 

Whether a community exists is tested at two major points in the representation and collective bargaining processes: first, when a union seeks to organize a group of workers for purposes of collective bargaining; and second, when a union seeks to maintain a strike. Both events require individual employees to decide if they are better off acting as a part of a group or acting alone. The practical concept of community has been determined largely by forces operating in relatively compact physical workplaces, with some notable exceptions in the transportation industry. Now, information technology makes it possible to organize work across formerly immutable physical boundaries, substantially decreasing the relevance of physical space as a consideration in the organization of work. Community may mean something different in these new geographically dispersed workplaces.



Communities play several important roles in American labor law and economics. Students of the sociology of workplaces long have observed that social networks (communities) exert powerful forces, sometimes at odds with formal rules, in defining and enforcing norms of behavior in the workplace. Erving Goffman studied the behavior of worker communities in emphasizing common problems and building a sense of solidarity by critical remarks about customers and bosses. ³⁷ This behavior is a spontaneous social phenomenon originating in a sense of community and reinforcing the community sense at the same time.



A community has shared norms, ideals, and forms of social action and understanding. ³⁸ These shared attitudes play, and are seen by members of the community as playing, a significant formative role in the historical development of a community: we are the way we are now because we were more or less this way in the past. Communities tend to be self-sustaining regularities in thought and action. When, as a member of a community, one shares its norms and values, one persists as a member of the community because one is already a member-i.e., because the norms and values I share make me continue to identify with the other members and conform to the community's standards. This approach explains why  [*12]  communities are characterized by the four factors previously identified: interdependence, fulfillment of participant needs, psychological or ideological commitment, and non-transitory attachment. ³⁹ Commitment is the key. Commitment to shared norms and values is explicitly related to self-conscious community membership. We do not simply share norms and values, we see ourselves and define ourselves by doing so. Communities so conceived fulfill member needs and desires in part because they also instill those very needs and desires. Member needs and desires are functions of the norms and values the member shares as a member of the community. Given that the community realizes a self-sustaining regularity in thoughts and actions centered around those norms and values, it must sufficiently meet the needs and desires it creates. Otherwise, the community would not continue to exist. Communities so conceived are obviously characterized by non-transitory attachment.



Leadership, rhetoric, and art are important techniques for raising community consciousness and increasing community solidarity. ⁴⁰ Solidarity refers to the willingness of members of a community to act for a common purpose, even if individual sacrifice is required to do so. Communities distinguish sharply between members and nonmembers, and they typically have mechanisms for expelling members who violate community norms. Individuals are more tightly bound to communities when they depend upon community affirmation of their behavior and characteristics to enhance individual self-esteem. Often, community solidarity is increased by leaders who emphasize the differences between members of the led community from other communities, typically arguing that other communities are inferior. At the limit, this ethnocentric rhetoric leads to violent conflict. Derision of "scabs" (strike breakers) during a strike is an example.



Attachment to a community is transitory when the transaction costs of withdrawing are low. Tourists do not have strong attachments to the  [*13]  communities through which they pass even though while they are there, the communities may meet many of their physical and social needs. The reason is that their attachment is entirely transitory. ⁴¹ Exercise gyms often are not communities because the transaction costs of withdrawing are low. Conversely, prison communities are not transitory because of the high transaction costs of withdrawing. ⁴² A workplace community may be transitory when labor market conditions make it easy to get another job and when the practice is for workers to move around from one employer to another in physically separated locations. Conversely, when labor market conditions or industry practices make it unusual for workers to quit and change jobs frequently, the workplace community is less transitory.



A. Communities in Physical Workplaces



Collective bargaining links sociology with law. Fundamentally, collective bargaining originated in voluntary communities of workers, not in law. The earliest American trade unions were craft organizations, organizing bodies of coopers, cobblers, and printers, for the purpose of regulating their labor markets. Members of these guilds and craft unions were united by their common skills defining their crafts. Members were distinguished from all others because practicing the craft required a skill set not widely possessed. Formal mechanisms such as membership cards, constitutions, rules, procedures for expulsion, and the economic reality that only members could practice the trade and earn higher wage levels held the members together.



Before the Industrial Revolution, trade unions were communities of entrepreneurs who banded together to mitigate the effects of market forces. Market forces were delimited by physical factors, encouraging localization  [*14]  of guilds. Early trade unions limited competition by agreeing on prices to be charged for their work and, as the geographic scope of markets increased, by regulating market entry by traveling members of the craft. Their geographic scope expanded as the geographic scope of their markets expanded.



The Industrial Revolution began to threaten established mechanisms for organizing work by breaking down skills components so they could be performed by less-skilled workers. Trade organizations reacted, transforming themselves into labor unions that sought to establish monopolies on labor and increase bargaining power vis-a-vis purchasers of that labor. ⁴³ While trade organizations regulated all facets of work, including the entrepreneurial policy component, labor unions drew sharp boundaries between labor markets and product markets. In other words, entrepreneurial policy was outside the scope of matters addressed by the trade unions through collective bargaining. ⁴⁴



Industrial unionism presented new challenges to which labor law responded. For one thing, the sense of community was more diffuse when the only unifying factor was employment in a particular enterprise, as contrasted with the practice of a craft. Second, the potential existed for jurisdictional conflicts between newer industrial unions and older craft unions. Third, while craft unions had regulated labor markets, benefiting employers, there was no history of industrial union regulation of the labor market. There was much greater hostility by employers to unionization. Class warfare rhetoric used by labor union advocates to build a sense of community among industrial workers increased this hostility. ⁴⁵



The challenges posed by industrial unionism were mitigated by the characteristics of traditional physical workplaces. Physical workplaces reinforce the sense of community in groups of workers performing the same or interdependent tasks. In a traditional workplace, workers are present in the same physical area for eight hours a day. They work for a  [*15]  common boss. Everything that happens in those eight hours, whether on the assembly line or in the locker room, lunch room, break room, or parking lot, is necessarily shared because of physical proximity.



The NLRB embraced the concept of workplace communities in evaluating the appropriateness of proposed bargaining units. Although the legal determinants of an appropriate bargaining unit have developed a mechanical character, their conceptual foundation is premised on the existence of communities of workers. This foundation is captured in the term "community of interest." As Bok and Dunlop noted more than forty years ago, the need to define bargaining units arises from the principle of exclusive representation. ⁴⁶ Other systems of collective bargaining lacking the exclusive representation principle rely altogether on informal communities of workers to assert their demands through collective bargaining. ⁴⁷ Limiting the exclusive representative to a group with a community of interest mitigates criticisms that workers are being forced to accede to bargains struck against their will. Members of relatively homogeneous workgroups are more likely to have similar preference or utility functions. ⁴⁸



Consequently, the community of interest concept developed thus far in labor law focuses on economic considerations, especially on wage competition among members of the putative community. Labor law, as it has been expressed in cases and commentary, does not concern itself with purely social and psychological dimensions of association. Collective bargaining, however, extends beyond wages to subjects such as work rules and seniority systems which make up the framework of a complex social system in the workplace. The community of interest concept encompasses inquiry into these aspects of the social order for work. Moreover, social attachment to a workplace group-an aspect of community in the more than purely economic sense-is an important determinant of a union's  [*16]  ability to organize and to apply economic pressure by withholding services. As John Dunlop observed over half a century ago:



The informal organization and social pressure of a group of wage earners, even in the absence of a trade union, substantially influences the amount of labor supplied under piece rates. Customary standards of the number of pieces that constitute a day's work develop and are typically enforced by informal and spontaneous pressures. The significant fact is that a working force is composed of much more than isolated and discrete individuals. A social community in miniature develops its own leadership, mores, and standards of proficiency and output. The individual's choice between income and leisure must be placed in this social context. ⁴⁹



Collective bargaining is a political, as well as an economic, process. ⁵⁰ Accordingly, an explicitly political view of community is useful in supplementing an economic assessment. To be sure, if members of a group of workers have the same preference or utility functions, they will have little difficulty in formulating a bargaining strategy; they all will want the same thing. Conversely, if they have significantly different preferences, it will be much harder to agree on a group strategy. But if other determinants of solidarity are lacking, members are less likely to agree to act in concert, even if they all individually desire the same things. Whereas if other determinants of solidarity are present, members may be willing to rely on community institutions - the trade unions representing them ⁵¹ - to mediate individual differences. The ideology of solidarity may induce individual members of the community to sacrifice selfish preferences to support pursuit of goals important to others.



Political solidarity in this sense is important to a system of collective bargaining in three ways: (1) A union cannot win a representation election or maintain a strike without political solidarity. (2) A bargaining representative cannot effectively mediate differences over bargaining  [*17]  priorities without the legitimacy that comes from political solidarity and a sense of community in the social sense. Effective trade unionism depends on social norms regarding dispute resolution and institutional legitimacy. (3) Identifying subgroups of workers forming communities is necessary to limit the scope of economic conflict which otherwise might extend to the entire working class.



As long as workplaces were physically determined, no one had to define community of interest in social or political terms because the physical features of plants, reporting locations, and employer organization provided useful mechanical tests for assessing community. Board and court cases evidence little interest in the underlying theory of community as it relates to bargaining. Douglas Leslie did develop an underlying theory, but focused mainly on economic forces operating in employee communities. ⁵²



Shifting from physical to virtual workplaces does not ordinarily change the purely economic factors of competition among workers. However, it potentially changes social forces and therefore impacts seniority and work rule considerations in the conventional community of interest inquiry, as well as the practical ability of unions to organize and apply economic pressure. Therefore, with electronic workplaces we have to probe more deeply into the concept of community, unless the policies of labor law are to become inoperative in virtual workplaces because they cannot meet the physical tests for community of interest. A purely economic analysis is insufficient, in part because it would not limit sufficiently the scope of community of interest in labor markets that approach global extent when they operate through electronic rather than face-to-face channels. Merely assessing economic competition is not enough. Workers who compete in offering their services may nevertheless lack other ties that give a group of  [*18]  workers sufficient solidarity to choose an exclusive representative, to develop a coherent bargaining agenda, and to engage in concerted action in support of their agenda.



B. Communities in Electronic Workplaces



The degree to which workplaces have become electronic varies. At one extreme are workplaces that are completely virtual. In virtual workplaces, employees never report to fixed locations and transact all business electronically. ⁵³ Less extreme are employees who have fixed work locations but telecommute some work days. ⁵⁴ Even employees who report to fixed work locations every day have seen their work environments evolve to a point where they interact to an ever- increasing degree electronically, rather than face-to-face. The discussion by the water cooler is in the process of being replaced by the discussion via e- mail. ⁵⁵



Information technology is creating a revolution in the organization of production as profound as the Industrial Revolution more than a century ago. Entrepreneurial policy is being reintegrated with work, as information technology makes it possible for work to be performed outside physical workplaces and in circumstances in which the worker has greater independence. The boundary between labor markets and product markets is becoming less clear, and the exclusion of entrepreneurial policy from the collective bargaining process is a less satisfactory boundary for defining communities for purposes of labor law.



Electronic community is a much talked-about concept. Some observers of Internet newsgroups have seen in them new sources of social ties and the emergence of strong new communities. Skepticism is appropriate, however, with respect to virtual communities. Typically,  [*19]  membership in virtual communities is a unidimensional phenomenon. Little ties an individual member to a virtual place except for the possibility of conversation. When other tying factors exist, they are attributable to a pre-existing traditional community rather than to the virtual community itself. A religious group may meet online, but its character as a religious group has nothing to do with electronic communications in the first instance. Electronic communities tend to be transitory because the costs of withdrawal are low.



The viability and strength of a virtual community may increase significantly, however, when the electronic media make possible other kinds of relationships. Technology may reduce the costs of communities which otherwise would not be feasible to meet important needs.



This is particularly important in the labor context. The same electronic channels that may permit the formation of virtual communities also enable work relationships. When the work relationship is overlaid with other attributes of community, a stronger virtual community may be enabled than might be the case for a purely social community. Indeed, the better analytical model might be to postulate the same features and behavior of a virtual work community that one would find for the same group of people functioning in the same labor market in conventional ways, asking only, "What is missing, because the work community is virtual rather than real?"



In contrast to physical communities, which tend to be multidimensional, electronic communities tend to be tied together by specialized common experiences or interests, much as professional relationships tie professors, physicians, and lawyers together across the boundaries defining universities, hospitals, and law firms. The common experiences and interests are multidimensional, much as geographic proximity, community schools, and local work opportunities tie neighbors together, but they are more narrowly focused. The shift from multidimensional to specialized communities is a phenomenon of twentieth-century technology, characterized first by automobile transportation and more recently by communications and information technology.



Alienation may result from the replacement of multidimensional communities by specialized communities. The common boss may become relatively less important as a force tying a community of subordinates together defensively. As technology reduces transaction costs for communities that may exist across large physical spaces, it becomes easier  [*20]  to organize work of greater scope and raise the level of supervisory decisionmaking far above the former level. The scope of foreman control in a traditional physical workplace is almost always defined in terms of relatively compact physical groupings of employees. When workers are tied together by e-mail, it is as easy for a boss located halfway around the world to give instructions to workers as it is for one at the end of the assembly line or down the hall. While the number of workers with a common boss may be larger, increasing the possibility of a larger scale community motivated by potential threats from the common boss, it also increases the diversity of experiences, thus reducing commonalities that define communities. In other words, the set of experiences that are important to an individual include a relatively smaller proportion of shared experiences.



The preconditions of community stated at the beginning of this Part identified four important factors: interdependence, fulfillment of participant needs, physical or ideological commitment, and attachment that is not transitory. ⁵⁶ Interdependence on workplace tasks may be increased by technology in an electronic workplace, but interdependence on other important social or psychological factors may decrease because physical separation leads to greater autonomy in meeting these needs. Because participants in an electronic community are physically separated from each other, they may-and must-meet many of their needs independently of each other. Members of electronic communities are far less likely to live in the same or contiguous neighborhoods; thus, their children are less likely to go to the same schools, and their families are less likely to shop in the same stores. The weather is less likely to be an interesting topic of conversation because members of electronic communities spread across wide geographic areas are less likely to experience the same weather phenomena. Members no longer park their cars in the same parking lot. They do not get dressed in the same locker room. There is little reason for them to take lunch or coffee breaks together. Reduced costs of withdrawal make the electronic community more transitory. Breaking attachment to  [*21]  one electronic community and forming an attachment to another can be as easy as clicking a mouse button. ⁵⁷



The differences between communities of workers in traditional and electronic workplaces have significant implications for NLRA doctrine as it relates to appropriate bargaining units and to rules governing access to employees. Parts IV and V explore these concerns.



IV. Bargaining Units in Electronic Workplaces



The NLRB typically is called upon to determine the appropriateness of a bargaining unit when a union files a representation petition asking the Board to conduct an election to determine whether a majority of the employees wishes to be represented for purposes of collective bargaining. The employer may contest the appropriateness of the bargaining unit in which the union is seeking the election. In resolving such contests, the NLRB applies its community of interest criteria. The factors the Board considers include: "(1) similarity in skills, interests, duties, and working conditions; (2) functional integration of the plant, including interchange and contact among the employees; (3) the employer's organizational and supervisory structure; (4) the bargaining history; and, (5) the extent of union organization among the employees." ⁵⁸ In physical workplaces, the  [*22]  Board also has developed a presumption that a bargaining unit limited to a single physical location is appropriate. The Board, however, has not articulated a coherent theory explaining how these factors relate to the concept of community in the workplace. ⁵⁹



The absence of such a coherent theory of community in the workplace was evident in the Board's first encounter with bargaining units in a completely electronic workplace. In Technology Services Solutions, ⁶⁰ the employees were customer service representatives (CSRs) who installed, serviced, and repaired computer equipment for the employer's clients on a nationwide basis. The CSRs had no fixed work location. They worked out of their homes or trucks and spent most of their time at the employer's customers' sites. The CSRs reported to customer service managers (CSMs) who were assigned territories. The CSMs also worked out of their homes. The employer did have a regional office that serviced a multistate, multi-CSM territory region. The union petitioned for an election in a unit consisting of CSRs in Colorado. The NLRB Regional Director directed that elections be held in two bargaining units, consisting of the territories of two CSMs in Colorado. The NLRB reversed and held that the smallest appropriate unit consisted of all CSRs in the region covered by the central regional office. ⁶¹



Both the Regional Director and the Board attempted to fit these employees' virtual workplaces into the traditional community of interest criteria as developed in physically discrete workplaces. The Regional Director applied the presumption favoring a single location bargaining unit and concluded that each CSM was a single location in cyberspace. The Board rejected this analogy because of the absence of a discrete physical location. The Board focused its attention on the degree of centralized  [*23]  control exercised by the central regional office and concluded that the central office's region was the smallest appropriate unit, emphasizing the third community of interest criterion (organizational and supervisory structure). ⁶² Neither the Regional Director nor the Board attempted to analyze the concepts of community underlying bargaining unit determinations and how those concepts might differ in electronic workplaces.



As developed previously, a community exists when its participants are interdependent, the community addresses important participant needs, there is a psychological or ideological commitment, and the participants' attachments are not transitory. ⁶³ The Board's community of interest criteria for testing the appropriateness of a proposed bargaining unit reflect these characteristics that define the concept of community. The first factor, similarity in skills, interests, duties and working conditions, is crucial to determining whether the employees in a proposed bargaining unit are interdependent and whether the proposed bargaining unit is capable of addressing important participant needs. Furthermore, to the extent that employees have similar skills, interests, duties and working conditions, they are likely to share a psychological commitment to the community necessary for a community to exist. The commitment, or sense of attachment, is stronger to the extent that the employees' common skills, interests, duties and working conditions differentiate them from other employees.



In traditional workplaces, functional integration of the plant and contact among employees are physically determined, in large part. This factor also differentiates a particular group of employees from others, thereby increasing their sense of attachment to the proposed bargaining unit. It also reflects the employees' interdependence. Moreover, to the extent that employees excluded from the unit are functionally integrated with the employees in the proposed unit, and to the extent that there is considerable contact and transfer between the two groups, membership in the proposed unit may be transitory as employees' attachments fluctuate from group to group.

 [*24] 

The Board often relies on the employer's supervisory and organiza-tional structure as a basis for finding a union's proposed bargaining unit to be inappropriately small. For example, in Technology Services Solutions, ⁶⁴ the Board rejected the Regional Director's direction of an election limited to CSRs in particular CSMs' territories because of the degree of control exercised by the central regional office. Similarly, when the Board decides that single physical location bargaining units are not appropriate, it relies on the degree of centralization of employer personnel functions. Such analysis is incomplete. Rather, it is necessary to determine whether the degree of centralization in the employer's supervisory and organizational structure affects the nature of the community of employees. The relevant inquiry should be whether centralized administration causes the employees to identify with employees outside the proposed bargaining unit in defining the needs that the community should address, disperses the employees' sense of commitment or attachment, or otherwise undermines the existence of a community within the proposed smaller bargaining unit.



The Board lists bargaining history as a separate factor considered in unit determinations. Unlike the other factors which reflect the likelihood that a group of employees form a community that will effectuate collective bargaining, bargaining history provides direct empirical evidence of whether such a community actually exists. A history of successful collective bargaining within a particular unit is strong evidence that the employees comprising that unit are interdependent, that the unit addresses important needs of the members, that the members of the unit share a psychological attachment to the unit, and that their membership is not transitory. On the other hand, a history of troubled collective bargaining may be evidence that the unit does not form a true community; or that a part of the unit, such as skilled craft employees within a larger industrial unit, lacks sufficient identity with the other members of the unit in terms of interdependence, participant needs, and psychological attachment such that it really is not part of the same community.



The relevance of the extent of union organization must be considered on two levels. On one level, the extent of organization may provide direct evidence of whether a community exists. If a particular group of  [*25]  employees has organized to a greater extent than any other group of employees, the disparity in organization may be due to the interdependence of the organized group of employees, their particular needs that can be met by collective bargaining, and their particular attachment to each other. Indeed, the fact that a particular group of employees is organized while most others are not may, in and of itself, differentiate that group from the rest of the workforce strongly enough to mark that group as a separate community.



On a second level, however, consideration of the extent of organization represents a statutory policy that favors unions in bargaining unit determinations. Appreciation for this policy requires consideration of the context in which NLRB unit determinations are made.



When the Board determines the appropriateness of a bargaining unit, its decision determines the election unit, i.e., the unit in which a representation election will be held. The positions of the parties to a dispute over the definition of the bargaining unit usually are determined by their election strategies. Unions usually seek the largest unit in which they believe they can win the support of a majority of employees. Employers seek larger units because such units are more difficult to organize and, therefore, are less prone to union victory in the election. A common employer strategy is to seek a unit so diverse that the employees within the unit will reject union representation because they are not a true community.



For example, during the NLRB's rule-making proceeding for hospital bargaining units, evidence showed that hospitals insisted on including non-nurse professionals in the same bargaining unit as registered nurses. ⁶⁵ When their demands were met, the same employers urged the non-nurse professionals to vote against the union, arguing that they would be a minority in the bargaining unit and that a nurses union could not represent their interests adequately. The evidence further showed that when the union won elections in bargaining units combining nurses and non-nurse professionals, employers sometimes proposed to sever the non-nurse  [*26]  professionals into a separate unit because of the difficulties of negotiating in such a broad unit. ⁶⁶



The NLRA provides that the Board determine unit appropriateness in such a manner as will "assure to employees the fullest freedom in exercising the rights guaranteed by this subchapter." ⁶⁷ In the early years of the NLRA, the Board interpreted this mandate by allowing unions to carve out of an employer's workforce the subset of employees with whom it enjoyed support and obtain an election limited to that group. For example, in Jacoby-Bender, Inc., ⁶⁸ the union had tried unsuccessfully to organize all production employees of the employer, a manufacturer of metal watch bands. It then petitioned for an election in a unit limited to the employer's polishing department. The record revealed that all production departments had uniform hours and vacation policies, as well as similar working conditions. ⁶⁹ Although permanent transfers between departments were rare, there was temporary interchange when necessary to equalize workloads. ⁷⁰ Nevertheless, the Board held the petitioned-for unit to be appropriate, emphasizing the importance of the extent of the union's organization. ⁷¹



In Botany Worsted Mills, ⁷² the union sought a bargaining unit limited to wool sorters or trappers, employees who prepared the wool for use by other employees in subsequent manufacturing operations. The Board's sole rationale for finding the requested unit appropriate was that the union had succeeded in organizing the employees in the requested unit and had failed to organize the rest of the company's workforce. The Board opined, "[w]herever possible, it is obviously desirable that, in a determination of the appropriate unit, we render collective bargaining of the Company's employees an immediate possibility." ⁷³ It found the unit appropriate "even if, under other circumstances, the wool sorters or trappers would not constitute the most effective bargaining unit . . . ." ⁷⁴

 [*27] 

Similarly, in Garden State Hosiery Co., ⁷⁵ the Board approved a unit limited to employees of the company's knitting department. The record revealed that knitting department employees began the production process and turned the product over to employees in the auxiliary department who completed the process, stamped, folded and packaged the product. The union had tried unsuccessfully to organize plant-wide, but enjoyed majority support among employees in the knitting department. The Board reasoned:



[I]t is often desirable . . . to render collective bargaining for the employees involved a reasonably early possibility, lest prolonged delay expose the organized employees to the temptation of striking to obtain recognition, and permit unorganized employees, engaged in other work-tasks, to thwart collective bargaining by those who evinced an interest in selecting a representative. ⁷⁶



In the Taft-Hartley Act of 1947, ⁷⁷ Congress expressed its disapproval of this NLRB practice in defining bargaining units. The Taft-Hartley amendments, however, did not change the statutory mandate that the NLRB define bargaining units to assure to employees the fullest freedom in exercising their rights. Rather, Congress added a new section providing that "the extent to which the employees have organized shall not be controlling." ⁷⁸ The potential contradiction between the unamended original Wagner Act language and the new Taft-Hartley Act language is resolved easily through legislative history of the Taft-Hartley amendments.



The prohibition on extent of organization controlling bargaining unit determinations arose in the House of Representatives. The version of Taft-Hartley which passed the Senate contained no such provision. The Conference Committee adopted the House's version. ⁷⁹ Opponents of the provision argued that it would make it extremely difficult, and perhaps impossible, for unions to organize such industries as insurance and public utilities which had highly integrated employers with widely dispersed small operational units. ⁸⁰ Proponents responded, however, that the NLRB was  [*28]  not to abandon application of its community of interest criteria in ways that assured employees their fullest freedom under the Act. Rather, proponents argued, the "extent of organization shall not be controlling" provision was intended to prevent the NLRB from granting unions requested bargaining units that were not supported by any valid criteria. ⁸¹



Taking into account considerations of community reveals how the prohibition on extent of organization controlling bargaining unit definition actually furthers the NLRA's mandate that units be defined so as to ensure employees their fullest freedom under the Act. Effective collective bargaining cannot take place in an artificial community. Employees whose principal commonality is their membership in a union will not be able to bargain effectively with their employer. Employees who are not interdependent and do not share common needs or goals are not likely to maintain solidarity during collective bargaining. Employees who routinely interact with, and are interdependent with, employees excluded from the unit are not likely to develop the psychological ties to the unit that enable its members to stick together during bargaining. Thus, the NLRB's belief prior to Taft-Hartley that basing bargaining units on the extent of organization made collective bargaining an immediate possibility was credible, but was not defended with appropriate reference to the role of community in facilitating effective bargaining. The prohibition on giving controlling weight to the extent of organization is a requirement that bargaining units not only be capable of being organized, but that they also be constructed so that collective bargaining is a real possibility. The inquiry thus should be whether the small unit that the union has requested has sufficient indicia of community so that collective bargaining within such a unit is a realistic possibility.



The NLRB generally has recognized that this is the appropriate inquiry. The Board has held consistently that the issue before it is whether the unit requested by the union is an appropriate unit, not whether a larger unit may be more appropriate. ⁸² However, at times the Board has analyzed the  [*29]  appropriateness of a requested unit not by focusing on whether the unit has sufficient indicia of community, but on balancing the union's interest in ease of organization against the employer's interest in efficient administration. This approach is readily apparent in the NLRB's handling of cases requiring it to choose between single location and multiple location bargaining units.



In the early years after the Taft-Hartley amendments, the Board faced representation petitions seeking bargaining units limited to a few of the employer's numerous retail stores within a particular geographic area. Under the Taft-Hartley prohibition, the Board properly rejected such bargaining units as inappropriate because their sole distinguishing feature appeared to be the extent to which the petitioning union had succeeded in organizing the employer's employees. ⁸³



In Safeway Stores, Inc., ⁸⁴ the employer operated twenty-two stores in its Waco, Texas district: six in Waco, four in Austin, two in Corsicana, and one each in ten smaller towns. The Waco district covered an area approximating a rectangle two hundred by forty-five miles. The towns in which the stores were located were on average thirty miles apart and were an average of sixty-one miles from Waco. Local store managers had authority to hire and discharge, requisition food stock requirements from the company's central warehouse, purchase local produce, adjust prices to meet local competition, and exercise other aspects of managerial control. Transfers between stores were very infrequent. Local managers kept payrolls and paid them out of local receipts. Employees of the meat department in the Austin stores already bargained separately. Nevertheless, the Board rejected separate units for each city in which the employer operated stores. ⁸⁵ It interpreted its prior decisions, which properly had rejected units based solely on the extent of organization, as standing for the general proposition that "absent unusual circumstances, the appropriate  [*30]  collective bargaining unit in the retail grocery trade should embrace all employees . . . who perform their work within the Employer's administrative division or area." ⁸⁶ Thereafter, until 1962, the NLRB apparently gave controlling weight to the employer's administrative structure in evaluating bargaining units of companies with multiple locations. ⁸⁷



In 1962, the Board changed its policy. It reasoned that its prior policy basing bargaining units on the employer's administrative geographic divisions "impede[d] the exercise by employees in retail chain operations of their rights to self- organization." ⁸⁸ Six years later, the Board declared that single location units in the retail industry were presumptively appropriate. ⁸⁹



The presumption may be overcome, however. ⁹⁰ The analysis looks to whether the employer's administrative structure deserves greater weight than the employees' interests in ease of organization. For example, in Friendly Ice Cream Corp. v. NLRB, ⁹¹ the United States Court of Appeals for the First Circuit described the evaluation of bargaining units as a process of balancing the "employer's interest in bargaining with the most convenient possible unit," ⁹² with the "employees' interest in being  [*31]  represented by a representative of their own choosing." ⁹³ The court opined that, in view of the statutory mandate that bargaining units be defined so as to afford employees the fullest freedom in exercising their rights, "this factor of employee freedom can legitimately tip the balance in determining which of two equally appropriate units should be preferred." ⁹⁴ Consequently, although the court characterized the employer's structure as "a casebook study in centralized control," ⁹⁵ it found that the local manager had sufficient authority in personnel matters such that the employer had failed to rebut the single location presumption. ⁹⁶



The balancing of employer interests in efficient administration against employee interests in organizing and achieving representation is inappropriate. The NLRA requires that the NLRB determine bargaining units to ensure employees their fullest freedom under the Act. ⁹⁷ The prohibition on making extent of organization controlling was not an open invitation to give independent weight to the employer's administrative structure. Rather, prohibiting the Board from making extent of organization a controlling factor further ensures employees their fullest freedom under the Act by ensuring that the bargaining unit will have sufficient indicia of community so that effective collective bargaining is a realistic possibility. The employer's administrative structure is relevant only to the extent that it bears on the employees' interdependence, common needs, and other indicia of community. If the employer's administrative structure is to be given any significance independent of its effect on the employees' community, it must result from other policy reasons specific to the employer or industry. For example, the NLRB has refused to apply the single location presumption to public utilities because the industry "is  [*32]  characterized by a high degree of interdependence of its various segments and . . . the public has an immediate and direct interest in the uninterrupted maintenance of the essential services that this industry alone can adequately provide." ⁹⁸



Electronic workplaces take various forms. Most commonly today, electronic means of communication reduce but do not eliminate employees' attachment to specific physical work sites. They do this in two ways. First, electronic communications free up employees from having to report physically to the same location each day. Telecommuting or flexiplace programs enable employees to work from home or other alternate locations one or more days per week.



Second, electronic mail and similar technologies enable employees to have more frequent and more prolonged communications with employees outside their physical work site. Electronic means of communication thus facilitate the integration of employees in different physical locations in the same production process. Nevertheless, in most cases, electronic communications are not likely to so eradicate employees' attachment to their physical work sites as to eliminate the physical work location as a basis for community. Employees who physically report to a particular location, even if not on a daily basis, are likely to continue to identify with that location and share common needs and concerns with their coworkers at that location. ⁹⁹ Consequently, in most cases, the single location presumption should apply to electronic workplaces. Nevertheless, the more that electronic communications result in employees from different locations working together frequently, the easier it will be for employers to rebut the single location presumption.

 [*33] 

Far more problematic are cases like Technology Services Solutions, ¹⁰⁰ where employees have no fixed work location and all reporting is conducted electronically. The common needs and concerns of such community members are likely to be derived from the specialized nature of what they have in common-i.e., that they work for the same company. They are not likely to be tied together by other common needs or experiences. They will not support the same sports teams, shop in the same stores, vote in the same elections, fight the same traffic, or have face-to-face dealings with the same bosses. They will not meet by the water cooler or in the cafeteria. Indeed, what characterized the employees in Technology Services Solutions, and what is likely to characterize employees in most such virtual workplaces, is the independence with which they perform their jobs. The common electronic boss is far less likely to be a tie that binds workers together in virtual workplaces than the common on-site supervisor who binds workers together in traditional workplaces. ¹⁰¹



Similarly, membership in an electronic workplace community is likely to be more transitory than membership in a traditional workplace community. The costs of withdrawal from a traditional workplace community are both economic and social. The social costs result from leaving behind coworkers to whom an employee has become attached. The common refrain, "I hate my job but love the people I work with," is far less likely to be heard in a virtual workplace than in a traditional one. The costs of withdrawing from a virtual workplace community are likely to be predominantly economic, i.e., the loss of compensation improvements, such as vested pensions and additional vacation days, that result from increased seniority with the same employer. Here again, the communal tie is employer-wide.



On what basis does a sufficient community of interest exist such that effective collective bargaining can be a realistic possibility in a virtual workplace? The Railway Labor Act (RLA), ¹⁰² with its bargaining  [*34]  organized by system-wide craft or class, provides some helpful analogies. The RLA relies less than the National Labor Relations Act on physical proximity to define bargaining units. From the earliest days after its creation in 1934, the National Mediation Board (NMB) defined bargaining units only in terms of system-wide crafts or classes. ¹⁰³ Craft organization of the railroad industry prevails even today.



The first stirrings of collective bargaining in the railroad industry occurred in the 1870s with locomotive engineers and firemen. After an abortive attempt by Eugene Debs to organize the railroad industry on an industry-wide basis, culminating in the violent strike at the Pullman Company south of Chicago in 1896, organization of the railroad industry existed primarily in operating crafts-conductors, trainmen, and switchmen. ¹⁰⁴ Then, in the 1920s, the shop crafts organized, although still along craft lines. As automation and communication became more important, the signalmen joined telegraphers as other craft groups. The NMB, which enjoys virtually unreviewable authority to define bargaining units, ¹⁰⁵ follows a policy of recognizing system-wide units of "crafts or classes." ¹⁰⁶

 [*35] 

Organization of the airline industry, which did not accelerate until the 1950s, proceeded along craft lines as well. In the late 1970s through the mid 1990s, airline employee representatives attempted to deviate from the system-wide craft or class norm. The Board rebuffed these attempts, holding steadfastly to its system- wide craft or class rule for defining bargaining units. ¹⁰⁷ Of course, system-wide representation rights in a labor organization do not necessarily mean that collective bargaining takes place on a system-wide basis for all issues. ¹⁰⁸ In a 1974 railroad merger case, the Board declined an invitation by the employer to state a policy that would enlarge representation units to conform to the scope of a merged railroad operation. ¹⁰⁹ In the airline industry, however, the Board has adhered more strongly to the principle that system- wide representation rights must be expanded as the airline system expands through mergers and other corporate transactions:



The pattern of representation which has resulted in the railroad industry has, in the Board's judgment led to uneven representation, duplication of effort and confusion; and has significantly reduced the ability of railroads to integrate operations and manage a single rail system. In the absence of compelling facts,  [*36]  judged in each instance on a case-by-case consideration of the situation presented, the Board does not intend to foster a similar pattern of representation in the airline industry.



To exempt the Flight Attendants from the Board's finding that USAir and the Shuttle are a single transportation system would lead to an inconsistent and disruptive pattern of representation the Board has sought to avoid. Therefore, the Board applies its determination [that Shuttle flight attendants should not be kept in a different bargaining unit and subject to different rates of pay, hours and working conditions] to all crafts or classes on the combined system. ¹¹⁰



The effect of the system-wide craft or class bargaining unit rule of the RLA is to strengthen occupational affiliation forces and to weaken those based on location. Organizing collective bargaining along craft or class lines in the railroad and airline industries reflects the realities of community in those industries. Airline pilots, railroad engineers, airline mechanics, and railroad police share expertise, have common experiences, and face labor market pressures that transcend geographic boundaries. They work across considerable distances, serve customers who are themselves in transit, and frequently find themselves away from home. For them, the relevant workplace community is not a fixed facility, but a mobile flight deck, locomotive cab, airplane, or squad car. This is reflected in the NMB's emphasis on the work-related community of interest the employees share.



For example, in Independent Ass'n of Continental Pilots, ¹¹¹ the National Mediation Board restated the community of interest factors it uses in defining a craft or class for RLA representation purposes: actual duties of employees, nature and set-up of operations, work environment, qualifications of employees, job retention requirement, interaction of employees, and role of major equipment. ¹¹² According to the Board, the factor of "work-related community of interest" is particularly important. ¹¹³ The motivation for this factor is to "ensure a mutuality of interest in the  [*37]  objective of collective bargaining." ¹¹⁴ In the case before it, the Board concluded that ground school instructors were not part of the same craft or class as flight instructors because the ground instructors, unlike the flight instructors, did not need experience as Continental pilots. The ground school instructors had significantly different teaching responsibilities, worked in separate organizational units, and received markedly different benefits and compensation. ¹¹⁵ This analysis does not emphasize social interaction, but interaction on matters of employment.



Similarly, communities of interest in virtual workplaces are more likely to organize around occupations and job responsibilities rather than physical locations. The service technicians in Technology Services Solutions ¹¹⁶ in all probability identified with each other as service technicians, not as employees working out of any one particular location. The tie that bound them together was their occupation. Indeed, a bargaining unit limited to service technicians in one state, as requested by the union, might have been easier to organize, but would not have had a sufficient sense of community and solidarity to make successful collective bargaining realistic. The odds that a union victory among service technicians in one state would prove to be illusory are high. It is far more likely that collective bargaining can be successful in virtual workplaces when organized along occupational lines rather than geographic location.



The structure of communities in electronic workplaces is likely to make organizing such workers more difficult. The electronic communica-tion system itself is a critical component of the employees' community in an electronic workplace. Access to such electronic communication systems is also critical for unions seeking to organize employees in electronic workplaces. Accordingly, Part V explores the access issues.



V. Access to Employees in Electronic Workplaces



The legal regime governing access to employees has developed entirely within the confines of traditional physical workplaces. The initial inquiry  [*38]  focuses on the employer's actions in exercising its right to exclude individuals physically from the property. When the employer has properly exercised that privilege, it yields to employee section 7 rights ¹¹⁷ only when the employees are so isolated physically as to render alternative means of communication with the employees infeasible. When employees already are lawfully on the property, the employer's privilege to forbid solicitation is much more circumscribed. ¹¹⁸ In electronic workplaces, however, the distinction between physical exclusion from and invitation onto the property is far less meaningful. An ongoing case in California illustrates how this is so.



In late 1998, a California state trial court in Intel Corp. v. Hamidi enjoined Ken Hamidi from sending e-mail to employees of Intel Corporation. ¹¹⁹ Hamidi, a former Intel employee, distributed e-mails en masse to current employees criticizing Intel's employment practices. The court regarded Hamidi's actions as a trespass and, therefore, enjoinable. ¹²⁰ Hamidi was not a current employee of Intel and was not seeking to organize current employees or otherwise communicate with them for their mutual aid and protection. Consequently, Hamidi does not raise issues under the NLRA. It does, however, highlight the conceptual difficulties of applying legal concepts that initially developed in physically defined spaces to electronic communications.



If Hamidi had physically come onto Intel's premises and distributed leaflets to Intel employees, he clearly would have trespassed. If, instead, he had stationed himself on a public sidewalk outside Intel's offices and distributed the same leaflets to employees heading to work, he clearly would not have trespassed even though many of those leaflets would have been brought by Intel employees onto Intel's property. In the actual case, Hamidi sent e-mail to Intel employees over the Internet. It is not intuitively obvious whether such actions are more analogous to distributing leaflets on  [*39]  Intel's property or to distributing them on the public sidewalk and leaving it to Intel employees to bring them onto the property. ¹²¹



Electronic communications present extremely powerful tools for organizing employees for unions generally and around specific issues. ¹²² For example, a union organizer could engage in a conversation with prospective members via cyberspace in a non-union workplace:



I'm trying now to organize [name deleted] company. The employees all have computer terminals on their own desk. There's nothing in the law that prevents me from sending e-mail from my office to theirs, unless their employer puts a firewall up. It's legal! So if I'm going to reach them, I'm going to flood them. As for getting e-mail addresses, that's easy: You get someone inside to provide them. It's usually their first initial and last name-at "com." ¹²³



Consequently, the stakes will be high when the NLRB and the courts consider issues concerning electronic access to employees. In considering electronic access issues, they will attempt to build on a legal regime that developed in workplaces completely bounded by physical space.



Any analysis of union access rules must go back more than half a century to the Supreme Court's decision in Republic Aviation Corp. v. NLRB. ¹²⁴ Republic terminated an employee for soliciting coworkers to join  [*40]  a union during his lunch break in violation of the company's blanket ban on solicitation on company property. The Court upheld the Board's approach that such no-solicitation rules were presumptively illegal when applied to employees soliciting their coworkers during nonworking time. ¹²⁵ In such instances, they violated section 8(a)(1) by interfering with, restraining, and coercing employees' exercise of their rights under section 7 to engage in concerted activities for mutual aid and protection. ¹²⁶ The employer could overcome the presumption by showing that special circumstances warranted restricting solicitation to protect a valid business purpose.



The Board applied Republic Aviation to nonemployee union organizers as well as employee solicitors. The Supreme Court, however, disagreed. In NLRB v. Babcock & Wilcox Co., ¹²⁷ the Court found the distinction between employee and nonemployee solicitation to be crucial.



The Court characterized the distinction between employee and nonemployee solicitors as "one of substance." ¹²⁸ Nonemployee solicitors were trespassers and, the Court stated, "an employer may validly post his property against non-employee distribution of union literature if reasonable efforts by the union through other available channels of communication will enable it to reach the employees . . . ." ¹²⁹ The Court made clear that it was protecting the employer's property rights, specifically the "right to exclude from property." ¹³⁰ Such rights would yield to section 7 rights only "when the inaccessibility of employees makes ineffective the reasonable attempts by nonemployees to communicate with them through the usual channels . . . ." ¹³¹



Twenty years after Babcock, the Court seemed to invite the NLRB to exercise its expertise in developing an approach to reconciling employee section 7 rights with the employer's strong property right to exclude nonemployees from its property. In Hudgens v. NLRB, ¹³² the Court held  [*41]  that the Board improperly relied on the First Amendment in holding that a striking union had the right to picket an employer at the employer's retail store in a large shopping mall. ¹³³ It remanded the case to the Board to consider "[t]he locus of that accommodation" between section 7 rights and employer property rights. ¹³⁴



Taking the Court up on its apparent invitation, the NLRB established a general approach to all access cases in Jean Country. ¹³⁵ The Board indicated that it would balance the impairment of section 7 rights that would be posed by a denial of access against the impairment of private property rights that would be posed by compelling access. In the balancing process, the availability of reasonable alternative means of communication would be entitled to particular weight. ¹³⁶



In Lechmere, Inc. v. NLRB, ¹³⁷ the Supreme Court held that the Board's Jean Country approach conflicted with Babcock. The Court regarded Jean Country as impermissibly eroding Babcock's holding that an employer may post its property against solicitation except where reasonable alternatives are not available to the union, ¹³⁸ and as impermissibly sanctioning "reasonable trespass." ¹³⁹ The Board was not authorized to require access merely because "nontrespassory access to employees may be cumbersome or less-than-ideally effective . . . ." ¹⁴⁰



The crucial distinction between Babcock-Lechmere and Republic Aviation is the difference between nonemployees, who are strangers to the property and may be physically excluded by the property owner, and employees who are licensed by the property owner to be on the property. The Court itself has cited the primacy of this distinction in other access cases. ¹⁴¹ Professor Cynthia Estlund has debunked this distinction  [*42]  effectively. She has demonstrated that the common law of property at the time of Republic Aviation and Babcock provided that a licensee who exceeded the scope of the license, such as an employee who violated a no-solicitation rule, became a trespasser. ¹⁴²



Nevertheless, the licensee-trespasser distinction has been given paramount weight in access cases. For example, in California, shopping center owners do not have an absolute right to exclude solicitors. ¹⁴³ The NLRB has held that Lechmere does not apply to California shopping centers because union solicitors are privileged under state law to be on the property. ¹⁴⁴ More generally, the Board has refused to apply Lechmere whenever the employer's property interests do not include control over physical access to the property. ¹⁴⁵



The rationale behind the licensee-trespasser distinction becomes more comprehensible when we realize that what is at stake is the property owner's naked right to exclude individuals from the propertyregardless of reason or justification. As respected commentators have observed, for example, the only apparent reason for Lechmere's exclusion of the union organizers was to prevent their message from being communicated. ¹⁴⁶ There was no showing that the solicitation disrupted the employer's business in any material way.



In determining the boundaries of this naked property right, the licensee- trespasser distinction defers, in the first instance, to the property  [*43]  owner/employer's determination as to who shall have access to the property. It takes that determination as a given and applies settled modes of analysis to the property owner's solicitation restrictions. Thus, if the property owner has decided to exclude certain individuals physically, including nonemployee union organizers, the distinction applies the very property-protective Babcock-Lechmere analysis and upholds the solicitation restrictions in most instances. On the other hand, if the property owner has decided to invite certain individuals, such as its employees, onto the property, it has opened itself up to the more section 7-protective analysis of Republic Aviation. In other words, the law does not impose access requirements on the employer/property owner; the employer/property owner imposes them upon itself by deciding whom to invite onto the property.



This approach of taking the employer's basic exercise of property or entrepreneurial rights at face value and applying established modes of analysis to them is common in labor law. For example, the law respects a successor employer's basic entrepreneurial right to hire whomever it wants and to structure its workforce any way it wants. The law does not compel the successor to hire the predecessor's employees. ¹⁴⁷ However, once the successor has decided to hire the predecessor's employees, an established mode of analysis allows the NLRB to compel the successor to recognize and bargain with the predecessor's union. ¹⁴⁸ Similarly, the law accepts the workforce structure as developed by the employer and applies established modes of analysis to that structure to determine who, if anyone, will have collective bargaining rights. ¹⁴⁹



Thus, once an employer has decided to allow solicitation by strangers on its property, it may not pick and choose which solicitation it will  [*44]  countenance; at the least, it may not pick some solicitors that it will allow and choose not to allow union solicitation. An exception to an employer's right lawfully to post its property against union solicitation when the employer discriminates in the exercise of that right was recognized by the Court in Babcock ¹⁵⁰ and Lechmere. ¹⁵¹ Recently, the circuits have divided over whether an employer who allows charitable and other community solicitation on its property may lawfully prohibit union solicitation. ¹⁵² Those courts that allow an employer to discriminate in favor of charitable and community solicitors and against union solicitors lose sight of the basis behind the discrimination exception to the naked right to exclude. An employer who decides to open up the property for some solicitors has, by its own actions, chosen to allow solicitation on the property and invites scrutiny when it then seeks to bar union solicitors. ¹⁵³



A similar approach underlies the NLRB's analysis of employer restrictions on the use of company bulletin boards. Since at least 1941, ¹⁵⁴ the Board has held that an employer lawfully may prohibit employees from using company bulletin boards and similar areas for the posting of notices; but if the employer allows employees to post some notices, it may not  [*45]  prohibit them from posting union notices. ¹⁵⁵ The Board regards such prohibitions as discrimination against union notices. Two circuit courts of appeals have approved the Board's approach. ¹⁵⁶



In Guardian Industries Corp. v. NLRB, ¹⁵⁷ however, the Court of Appeals for the Seventh Circuit rejected the Board's approach. The court reversed a Board decision which found that the employer had violated the NLRA by prohibiting the posting of union meeting notices on company bulletin boards after allowing the posting of employee notices of various items that were for sale. The court reasoned that employee for sale notices were not comparable to union meeting notices and that, therefore, the employer did not discriminate when it permitted the former but disallowed the latter. Interestingly, the court suggested that the Board might require that the employer give employees access to its bulletin boards to post union meeting notices if it found that the employer's nondiscriminatory refusal to post meeting notices for outside groups interfered with employee section 7 rights in a manner comparable to the no-solicitation rule in Republic Aviation. ¹⁵⁸ Perhaps because the NLRB characterizes its findings in bulletin board cases as findings of discrimination, the Seventh Circuit failed to realize that the Board's rule concerning access to bulletin boards is no different from Republic Aviation's rule concerning solicitation prohibitions.



The starting point of analysis in both instances is recognition that, as a property owner, an employer has a right to exclude anyone from its premises. Once it invites employees onto the premises, however, albeit for the limited purpose of performing their job responsibilities, it may not prohibit or restrict them from soliciting their coworkers to join a union unless the prohibition or lesser restriction is necessary to ensure order and prevent disruption of the business operation. Similarly, an employer has a property right to exclude all employee use of its bulletin boards. Once it  [*46]  invites employee use of the bulletin boards, however, albeit for the limited purpose of posting for sale notices, it may not prohibit or restrict them from posting union meeting notices unless the prohibition or lesser restriction is necessary to prevent disruption or interference with the employer's business purpose. ¹⁵⁹



Thus, an employer who chooses to invite certain types of persons on the property, stranger solicitors in the case of the Babcock-Lechmere discrimination exception, or employees in the case of Republic Aviation, becomes subject to the Republic Aviation mode of analysis. The employer may still prohibit or regulate solicitation when necessary to the operation of the enterprise. Consequently, the employer may limit solicitation to nonworking time and may prohibit solicitation during nonworking time where special circumstances justify it. ¹⁶⁰ The employer may not prohibit literature distribution, but may limit it to nonworking areas and may prohibit it completely upon a showing that distribution even in nonworking areas creates excessive litter or otherwise unduly disrupts operations. ¹⁶¹



The different modes of analysis applied to employee invitees and nonemployee trespassers have led many unions to convert their organizers into employee invitees. The process is known as "salting." Paid union organizers seek to be hired by the employers they have targeted for organizing and thus become employee invitees who enjoy broader rights to solicit under the Republic Aviation mode of analysis. In NLRB v. Town & Country Electric, Inc. ¹⁶² the Supreme Court held that such salts are employees under the NLRA and enjoy the same section 7 rights and section 8(a) protections as other employees. ¹⁶³ Although the Court's rationale in Town & Country was quite narrow-it relied on the plain meaning of the term "employee" as defined in section 2(3) of the NLRA ¹⁶⁴-as Professors Gely and Bierman have aptly demonstrated, the salting issue properly is  [*47]  understood as an access issue. ¹⁶⁵ The privilege of salting an employer tends to level the playing field of access to employees.



A second area where the Board and the courts have attempted to level the uneven playing field of access to employees involves home visits, although this area does not implicate employer property rights. The NLRB has held that employer solicitation of employees at their homes is inherently coercive and per se violative of the NLRA. ¹⁶⁶ However, to offset their lack of access to employees at the workplace, the Board has refused to prohibit unions from soliciting employees at their homes. ¹⁶⁷



The playing field that the Board and the courts have developed is a markedly physical one. In Lechmere, the Court emphasized that physical space is what is at issue:



[T]he exception to Babcock's rule [that an employer may validly post its property against union solicitation by nonemployees] is a narrow one. It does not apply wherever nontrespassory access to employees may be cumbersome or less-than-ideally effective, but only where "the location of a plant and the living quarters of the employees place the employees beyond the reach of reasonable union efforts to communicate with them." Classic examples include logging camps, mining camps, and mountain resort hotels. ¹⁶⁸



Similarly, the home visits doctrine also is based in physical space. The employer's physical property is off limits to nonemployee union organizers, but the employee's physical property, i.e., the employee's home, is off limits to the employer. This model is based on traditional industrial workplaces where employer and employee time and territory are segmented with relatively strict boundaries and ongoing border skirmishes fought over such matters as mandatory overtime and the taking of personal phone calls during working time. ¹⁶⁹

 [*48] 

In physical workplaces, the solicitor and the target of the solicitation usually are in the same place at the same time. The solicitation takes place in a discrete transaction which has a beginning and an end. Solicitations in electronic workplaces do not share these characteristics. The solicitor may be anywhere and the target of the solicitation may receive and read the solicitor's message minutes, hours, or even days after it is sent.



The NLRB's forays into access issues in cyberspace have, thus far, been very limited. Its decisions concerning employee use of electronic mail for section 7 purposes have been limited to traditional issues of employer discrimination and the boundaries between protected and unprotected conduct.



In Washington Adventist Hospital, Inc., ¹⁷⁰ the employer discharged an employee for sending a "break message" via e-mail to all individuals on the employer's system sarcastically criticizing the employer's restructuring and reductions in force. The break message was one that automatically appeared on the screens of every computer that was logged on and required intervention by the computer user to delete it. The discharged employee sent the message during the time period that the hospital's computer system was at its peak usage. The NLRB adopted an administrative law judge's (ALJ) decision that recommended dismissing the complaint, holding that the employee's actions were not protected by section 7 of the NLRA. ¹⁷¹ The ALJ reasoned that the employee had taken over the employer's computer system at a time when a great deal of medical information concerning patients was being entered and communicated, and "arrogate[d] to himself the decision [as to] whether the hospital's computer-communication facility should cease being used for hospital purposes and be used for his own purposes: to communicate his dissatisfaction with hospital policy . . . ." ¹⁷² The ALJ further found the employee's actions unprotected because his use of a break message interrupted the work of employees using the computer system "during the rush hour, such work being the care of patients in an acute hospital setting." ¹⁷³

 [*49] 

In Timekeeping Systems, Inc., ¹⁷⁴ the Board adopted an ALJ recommended decision holding that an employee's use of the employer's e-mail system to disseminate a message critical of an employer's policy was protected. The employee was discharged for sending a flippant and grating e-mail message to all persons on the system criticizing an employer memo which announced a new vacation policy. The ALJ gave the employer's reliance on Washington Adventist Hospital short shrift. ¹⁷⁵ The ALJ observed that the case before him did not involve the interruption of transmissions regarding patient care. The employer had conceded that "employees were permitted to post 'simple' e-mails to each other," make personal phone calls, and engage in similar personal matters during work time. ¹⁷⁶ Furthermore, in the ALJ's view, the e-mail message could not have taken more than a few minutes to digest. There being no material disruption of the employer's operation, the ALJ concluded that firing the employee violated the NLRA. ¹⁷⁷



At issue in Washington Adventist Hospital and Timekeeping Systems was the application of traditional NLRA doctrine concerning the boundaries of protected concerted activity. In Washington Adventist Hospital, the nature of a break message was critical. The break message interrupted all transmissions to all terminals. The evidence established that such messages were used only to warn users when the system was going down. ¹⁷⁸ Under any reasonable analysis, the employer had legitimate business reasons for restricting the use of break messages, as opposed to routine e-mail messages. Similarly, in Timekeeping Systems, the primary focus was on the tone of, and language used in, the e-mail message. The message itself was otherwise little different from routine personal e-mail messages that the employer permitted employees to send. The analysis and the result are not unlike those for employee invitees who engage in union solicitation during personal time on the employer's premises.

 [*50] 

Another relatively straightforward application of traditional labor law doctrine occurred in E. I. du Pont de Nemours & Co. ¹⁷⁹ In du Pont, the Board held that an employer violated the Act by denying the union access to the e-mail system to distribute union literature while allowing employees to use the system to distribute a wide variety of other material, a clear case of illegal discrimination.



Eventually, the Board will face a case in which an employer broadly prohibits employees from using the computer system for any nonbusiness purpose and where the special circumstances present in Washingt