Card check recognition: new house rules for union organizing?

I. INTRODUCTION

A significant policy debate has been occurring regarding union organizing methods in the United States. This debate focuses on the appropriateness of granting union recognition based on majority support as demonstrated by union authorization card signatures, also known as "card checking." (1) Critics describe the practice as anathema to basic democratic principles and accuse unions of wanting to deal from the bottom of the deck to secure undeserved representation of employees. (2) Proponents of card check recognition argue that reliance on National Labor Relations Board ("NLRB") organizing procedures fails to protect employees' rights to organize, and forces unions to compete against a stacked deck that unfairly favors employers. (3) Indeed, the labor movement in the United States has long been dissatisfied with the legal framework under which unions operate. (4) This frustration was illustrated by American Federation of Labor and Congress of Industrial Organizations ("AFL-CIO") President Lane Kirkland's statement in the early 1980s suggesting that the National Labor Relations Act ("NLRA") (5) be repealed, thereby allowing unions and employers to operate within the "law of the jungle." (6)

Perhaps owing to sustained union membership losses, unions have recently fought hard for legislation that will facilitate the organizing process. (7) Specifically, unions are supporting amendments to the NLRA, such as the Employee Free Choice Act of 2007 ("EFCA"), which requires employers to recognize a union when the employer is presented with evidence of majority support for union recognition via card check. (8) The EFCA requires the NLRB to develop model authorization language and procedures for establishing the validity of signed authorization cards. (9) The EFCA also provides stronger penalties for employers' violations occurring while employees are attempting to form a union or attain a first contract. (10) These amendments would represent a significant departure from the NLRA, which currently allows for card check organizing based only on voluntary acquiescence of the employer, an unlikely event given the strident opposition to unions by U.S. employers. (11)

Despite their importance, little has been written in the academic literature about the experiences of unions and employers with the use of the card check organizing process. (12) This Article seeks to fill this gap. We collected data on every organizing event the AFLCIO publicly reported between 1998 and 2005. (13) The list of events includes government-conducted elections, card checks, and other events. (14) While the data are not exhaustive of all union organizing events, the data provide a more comprehensive view of the recent experience with card checks than other currently available sources. The data also allow us to evaluate some of the justifications that have been advanced in support of the proposed changes to existing law, as well as to explore the possible consequences of those changes.

We examine the claim frequently made by supporters of the proposed legislation that, for all practical purposes, unions have abandoned the election route to representation in favor of the card check process. (15) Citing data from the AFL-CIO, supporters of proposed reforms claim that less than twenty percent of all union organizing is conducted through Board certified elections and that most new members are now joining the labor movement via card check recognition or some alternative procedure. (16) Our data do not support this claim. In particular, we find that during the period from 1998 to 2005, unions reported organizing the majority of their workers through some type of government-sponsored election. While non-trivial, the number of non-election events (e.g. card checks, mergers, and affiliations) and the number of employees organized through these events do not account for the majority of union organizing activity.

More importantly, we find that the current rate of card check use appears to be limited to a few unions in specific industries. Employers in these industries tend to be those who otherwise might have a reduced incentive to oppose unions. In other words, employers who have abandoned the current system might not be representative of employers in general. In fact, available evidence suggests that most employers oppose card check recognition or, at least, prefer the use of Board election procedures.

This latter finding is important when assessing proposed amendments to the NLRA that would facilitate the card check process. If, as our data suggest, most employers remain reluctant to agree to card check recognition, passing legislation along the lines of the Employee Free Choice Act of 2007 can force employers to acquiesce to card checks, but such legislation will not remove employers' incentives to oppose unions. We suggest that employers will likely respond to such legislation by shifting their union avoidance activities to earlier stages in the organizing process, perhaps even before any organizing effort starts. Employers, for example, could simply engage in more intense monitoring systems to detect union organizing campaigns earlier--that is, they will not have the luxury of waiting for the union to request recognition before launching into a vigorous opposition campaign. This type of reaction by employers could be very effective, and is one which the proposed legislation does not address. We surmise that while the proposed legislation could be a good first step in expanding the ability of workers to organize collectively, it is important to realize that employers are likely to react to such changes by also altering their behavior. Any proposed legislation should anticipate and address this possibility.

Part II briefly discusses the mechanics of NLRB elections and card checks and, more importantly, the incentives both unions and employers have for using one or the other. (17) Part III describes our data, (18) and Part IV reports our findings. (19) Part V discusses the implications of our findings for the current policy debate regarding the card check process. (20) Part VI concludes the article.

II. THE DYNAMICS OF NLRB SPONSORED ELECTIONS AND CARD CHECKS

A labor union can become the majority representative of a group of employees either through a NLRB-sponsored election or through voluntary employer recognition. (21) The employer and the union seeking recognition regularly precede the NLRB-sponsored election with vigorous campaigning, but it is employers' behavior that has tended to generate the most concern. During the campaign process, which usually lasts several weeks, employers engage in a variety of activities to counteract the union. Various scholars have documented both the range of tactics that employers use in the course of an organizing campaign, (22) as well as the effect such tactics have on organizing election outcomes. (23) Some of the practices employers commonly use during organizing campaigns include the communication to employees of the employers' views regarding the presence of a union, via letters, pamphlets, and captive audience speeches. (24) Employers also frequently hire outside consultants to manage the campaign, and often encourage the formation of "Vote No" committees comprised of employees that oppose the union. (25)

Although legally permissible in principle, any communications with employees could become illegal if the employer uses them to either threaten or interfere with employees' rights under the NLRA. (26) For example, while the employer can distribute literature and communicate with employees, the employer cannot threaten to fire, or actually fire an employee because of that employee's union activities. (27) Similarly, it is illegal for an employer during the course of an organizing campaign to threaten to close or relocate operations if the union wins the certification election. (28) It is also illegal in the course of any communications with employees for the employer to promise or actually implement changes in the terms and conditions of employment. (29) Yet these illegal activities occur on a fairly frequent basis. Research on employers' campaign conduct shows that since the mid-1970s employers have become much more willing to engage in illegal activities during the course of an organizing campaign. (30)

The card check process, which involves voluntary recognition of the union by the employer based on a showing of majority support via representation cards, (31) provides an alternative to the NLRB-supervised election process. The use of card checks is not entirely new in U.S. labor relations. (32) In fact, various commentators have noted that following the enactment of the NLRA, the Board frequently certified unions on the basis of card check agreements. (33) During these earlier years, Board supervised elections were used primarily in situations where the employer questioned the status of the union as the majority bargaining representative, while the card check was the default recognition method. (34) A few years after the NLRA was enacted, and due to issues of institutional legitimacy, (35) the Board began showing a preference for elections as the primary means of union certification. (36) The Board's shift was further extended when Congress amended the NLRA in 1947. (37) Thus, secret ballot elections became the primary way for the Board to certify a union as the exclusive bargaining representative of a group of employees. (38)

From the perspective of labor unions, card checks have several advantages. By eliminating much of the campaigning that occurs between the union's request for recognition and the scheduled election, there is less opportunity for the employer to engage in anti-union campaigning, (39) and some of the employer abuses that typically occur during union organizing might be avoided. (40) Moreover, card check recognition should minimize delays that often characterize the NLRB organizing process and which research indicates decrease the likelihood of a union victory. (41) In short, the card check process should make it easier for unions to communicate their message to employees and have the employees make a decision without undue pressure by the employer. (42) A shorter organizing process which consumes fewer union resources might also free up both time and money for unions to pursue other organizing targets (43)

Although advantages to unions can easily be identified, why would employers ever agree to card check recognition? This would seem to be a difficult question to answer given long-standing employer opposition to unions that has characterized U.S. labor-management relations. (44) Yet the answer is clear: employers will agree to card checks if it is in their economic interest to do so. (45) In fact, research suggests that in agreeing to card checks, many employers point out the importance of avoiding the costs associated with mounting a vigorous anti-union campaign. (46) For other employers, however, card checks could result in significant benefits. (47) For instance, the presence of a union could make it easier to recruit, train, and retain employees. (48) Other employers might see card checks as a trade-off necessary to obtain the support of unions in pursuing their legislative agendas. (49) In short, for some employers the potential costs of having a union may be outweighed by other business needs. These employers realize that making it easier for the union to organize is a good business decision.

This discussion suggests that unions should prefer to operate under a system in which employers are required to recognize the union if presented with a card check majority. It is also clear that some employers perceive the card check process to make economic sense, and thus voluntarily agree to be bound by it. But, it must also be true that for many other employers the card check process does not make sound business sense. One only needs to look at the many employer organizations that lined up in opposition to card check recognition to see how unpopular it is in some quarters. (50) That group of employers should be expected to continue opposing any attempts at unionization, either by refusing to recognize a union through the card check process, as is currently their option, or if the law denies that option, by engaging in other activities that minimize the ability of unions to organize the workplace.

III. DATA

We collected data on the occurrence of all organizing events reported in the AFL-CIO's weekly publication, Work in Progress ("WIP"). The AFL-CIO began weekly publication of WIP in 1996. (51) In the first issue, the AFL-CIO described the objective of WIP as "an effort to communicate more quickly and effectively with union leaders, supporters and activists across the country." (52) In later issues, WIP encouraged member unions to send information about new organizing victories since "[o]rganizing new workers is the most important goal of the labor movement." (53)

We read through all issues from 1998 to 2005 and coded several pieces of data related to union organizing activities. The description of the reported organizing events regularly included information on the union or unions involved, the number of employees directly involved in the event, the name of the employer, the location (city and state), and the type of organizing event (e.g., government sponsored election, card check recognition, merger, etc.). (54) In addition, from this information we were usually able to assess whether the employees were private or public sector employees, as well as the industry in which they worked. (55)

Several caveats about the data are in order. First, the information provided in WIP does not claim to be exhaustive. That is, WIP does not list every organizing event in which AFL-CIO unions were involved during this period. Thus, our data do not represent the whole universe of organizing activity during this period, but only those events which were reported in WIP. (56)

Second, the AFL-CIO member unions self-report the organizing events reported in WIP. While there is no way of knowing whether there were any systematic biases in the way unions decided to report their organizing activity to WIP, there is also no reason to expect unions to under or over report one type of organizing method relative to others.

A comparison of our data to the extant data on card checks, though, suggests that our sample is consistent with published research. For example, in their study of union organizing under neutrality and card check agreements Professors Adrienne Eaton and Jill Kriesky surveyed fifty-seven national unions "with 10,000 or more members primarily in the private sector" about their experience with these organizing methods. Of the thirty-six respondents, twenty-three reported having at least one card check agreement. (57) Our data include reports by eighty-one unions, representing employees in both the private and public sectors without any restrictions on union size, of which thirty-six reported having experienced a card check organizing event. Eaton and Kriesky report that the majority of the card check agreements in their data occurred in the service sector, with the majority of these taking place in the hospitality, gaming, and telecommunications industries. (58) As reported below, (59) our data also indicate that service sector card checks represent a large majority of all card check events, (60) and most of these occurred in the communications, hospitality, and health care industries. (61)

Thus, the data available through WIP provide a large union--and industry-wide sample of organizing activities occurring between 1998 and 2005 that is generally consistent with data used in prior research. These data, we submit, present a picture of the recent experience of unions with card check agreements and other organizing methods.

IV. ANALYSIS

A. Reports on Total Organizing Activity

We start by looking at the total number of organizing events reported in WIP. We find there were 3847 organizing events. These organizing events resulted in the addition of approximately 1.9 million unionized workers. Of these, 762,955 were employees who work in industries under the NLRB's jurisdiction ("NLRA Employees"), (62) while about 1.1 million were what we refer to as "Non-NLRA Employees." (63) Of this latter group, the very large majority (93%) were public sector employees. In fact, the majority of organizing activity reported in WIP involved public sector employees (56%). This finding is consistent with other surveys which show that a large percentage of union organizing activity over the last several years has occurred in the public sector. (64)

Table 1 provides a breakdown of the type of organizing events reported in WIP. During the relevant period, unions reported a total of 2896 elections, (65) 730 card check events, (66) and 152 mergers or affiliations. (67) Both the numbers of elections and card checks events reported in WIP declined between 1998 and 2005, with a notable decline occurring between 2000 and 2001. The decline in 2001 could be related to disruptions caused by the September 11, 2001 terrorist attacks against the United States. (68) The data also reflects a notable decline between 2004 and 2005. This was likely caused by the departure of four unions from the AFL-CIO to form their own organization, The Change to Win Coalition, in the summer of 2005. (69) As described below, (70) these unions were among the most active AFL-CIO unions in terms of their organizing activities. Even if we disregard the last year of the data, it is interesting to note that neither the number of elections, nor the number of card checks have bounced back to pre-2001 levels.

Table 1 also provides information about the relative occurrence of different types of organizing events. The 2896 election events constituted about seventy-five percent of the total number of events; card checks represented nineteen percent of the total, and mergers and affiliations constituted about four percent of the total. Thus, the WIP data suggest that although elections continue to be the most common type of organizing event, card checks are fairly common, with mergers and affiliations lagging far behind.

In terms of the total number of employees brought into the AFL-CIO by each of the various types of organizing events, Table 2 shows that about half (52%) did so by means of government sponsored elections. Card checks, on the other hand, resulted in the addition of 245,823 members (about 13%), while new members brought via mergers and affiliations amounted to over 600,000, or about a third of the total. Thus, contrary to recently reported figures, (71) the data collected from the WIP reports indicate that the two most traditional types of organizing events (elections and mergers/affiliations) served as the main avenue for bringing new employees into the labor movement. Although not insignificant, the card-check process ranks third in terms of the number of newly organized employees.

Table 2 also allows us to explore the types of employees organized under each of the various methods. About twice as many "Non-NLRA Employees" (these include public sector and other non-NLRA employees) joined AFL-CIO unions through government-sponsored elections (65.3%) as compared to "NLRA Employees" (34.7%). (72) Exactly the opposite is true for card checks. About sixty-five percent of employees joining via card checks were "NLRA Employees" while a little bit over a third (35%) were "Non-NLRA Employees." (73) These figures suggest that while the card check process is used less frequently than other more traditional organizing methods, card checks are used more frequently to target employees that fall under the jurisdiction of the NLRA. Elections, however, are the most commonly used organizing method for all types of employees.

The data in Tables 1 and 2 raise several important issues regarding the use of card checks under the existing legal framework. The WIP data confirm the proposition that card checks have become a common organizing method, particularly in the private sector where unions have been more likely to use the card check process, thus allowing them to avoid what they perceive to be an unfair NLRA election process. Perhaps because employees not covered under the NLRA operate under legal frameworks which unions perceive as more fair, there is less need for unions to avoid the election process in favor of alternate organizing methods. (74) Recall that a large proportion of what we refer to as "Non-NLRA Employees" is public sector employees. Public employers are widely believed to be less likely to mount vigorous campaigns against organizing drives, and thus are less likely to engage in illegal and dilatory behavior during the course of the organizing campaign. (75)

The WIP data also raise some doubts about the claim made by advocates of the proposed legislation that card checks have become the primary means of organizing new workers. For example, Professor James Brudney argues that less than twenty percent of all union organizing is conducted through Board-certified elections and that most new members are now joining the labor movement via card check recognition or some other alternative procedure. (76) Professor Brudney arrives at this figure by comparing the total number of employees organized via NLRB elections from 1998 to 2003 to the total number of employees the AFL-CIO reports to have organized in these same years. According to Brudney, only about 550,000 employees out of nearly three million (about 18%) were organized through Board-supervised elections. (77) Unfortunately, there are some problems with the conclusions drawn from these data.

First, as Brudney recognizes, the total number of employees he reports as having joined the AFL-CIO during this period includes both private and public employees. (78) Public sector employees, however, are not covered under the NLRA, and thus are not capable of being organized via an NLRB election. Thus, including newly organized public sector employees in the total number of employees organized by AFL-CIO unions to evaluate the relevance of the NLRA election process is inappropriate. Doing so grossly understates the percentage of employees being organized via NLRB-conducted elections. In fact, the WIP data indicate that between 1998 and 2005, more public sector than private sector employees joined the AFL-CIO.

Second, by arguing that card checks have displaced NLRB elections as the dominant method of union organizing, Brudney implies that the majority of the eighty percent of employees not organized through NLRB-conducted elections were organized through card checks. (79) As he acknowledges, however, this figure includes employees who joined via mergers with AFL-CIO affiliated unions or via new affiliations with the AFL-CIO, as well as through other organizing methods. (80) Our data show more employees were brought into the AFL-CIO through mergers and affiliations than through card checks. Moreover, employees "joining" the AFL-CIO through mergers and affiliations hardly constitute newly organized employees into the labor movement and, therefore, do not seem germane to an evaluation of the relevance of the NLRA election process. Again, including them in the total number of employees organized by AFL-CIO unions grossly understates the percentage of employees being organized via NLRB-conducted elections.

B. Reports of Organizing Activity by Union

The data collected from WIP also allow us to identify the unions involved in the various reported organizing events. Table 3 provides a breakdown of organizing events by union. (81) Column 2 of Table 3 reports the total number of organizing events. Column 3 reports the number of election events. Column 4 provides the number of card check events. Column 5 reports the number of employees organized through all the reported organizing events. Columns 6 and 7 break down the number of employees organized according to whether they were "NLRA" or "Non-NLRA" employees. The numbers in parentheses represent the percentages of those employees who were organized through card checks.

The Table 3 figures indicate that the Service Employee International Union ("SEIU") was the most active union both in terms of organizing events and in terms of the total number of employees organized, followed by the American Federation of State, County and Municipal Employees ("AFSCME"). In terms of organizing events, AFSCME was followed by United Food and Commercial Workers ("UFCW"), the International Brotherhood of Teamsters ("IBT"), and the Communications Workers of America ("CWA"). While these three unions were very active in terms of the total number of organizing events, their contribution to the reported total number of newly organized employees ranked toward the middle of those unions listed in Table 3. (82) Rounding out the top five in terms of the total number of employees organized are the International Association of Machinists ("IAM"), the American Federation of Teachers ("AFT"), and the United American Nurses ("UAN"). (83)

The data in Table 3 also reveal that unions differed widely in their use of card check recognition. Six unions--AFSCME, CWA, Hotel Employees and Restaurant Employees Union ("HERE"), SEIU, UFCW, and Union of Needletrades, Industrial and Textile Employees Union ("UNITE")--account for more than three quarters (76%) of all the reported card check events. (84) AFSCME reported seventy successful card checks or fourteen percent of its total number of organizing events. The CWA reported 103 successful card checks (41% of its reported organizing events), while HERE reported 144 (about 62% of its reported organizing events). The SEIU reported 138 successful card checks, or about eighteen percent of all their organizing activity, while UFCW reported fifty-five, or about fifteen percent of its total number of reported events. Finally, UNITE reported forty-five successful card checks, or about thirty-five percent of its total number of events. For all but three of the remaining unions listed in Table 3, card checks represented no more than about fifteen percent of the total number of reported organizing events. (85)

These figures suggest that the use of card checks during the period under study was concentrated in a few unions. For some unions the card check was especially important in organizing "NLRA Employees." Columns 6 and 7 in Table 3 show the number of "NLRA" and "Non-NLRA" Employees organized by each union. For instance, eighty-two percent of the 45,791 "NLRA Employees" organized by HERE were organized through the card check process. (86) For the CWA, the corresponding figure was 58.7%. For "Non-NLRA Employees," only the International Union of Operating Engineers ("OPEIU") organized more than fifty percent of new employees via card checks. On the other hand, most of the other unions included in Table 3 organized no more than about twenty-five percent of employees through card check recognition.

C. Reports of Organizing Activity by Industry

Although in recent years unions have been increasingly willing to cross traditional industry jurisdictional boundaries to organize outside their core industries, (87) most unions are still strongly identified with one or a few main industries. Consequently, one would expect that since card checks are used primarily by a few unions, there will be a corresponding concentration in terms of the distribution of industries in which card checks have become most common. Table 4 provides support for that proposition. (88) Table 4 lists the total number of reported organizing events (column 2), the number of reported elections (column 3), the number of reported card checks (column 4), and the total number of employees organized (column 5), by industry classification. (89)

The data in Table 4 suggest that the use of card checks is concentrated in four industries. Of the 730 card check events reported in WIP, 427 (58%) involved the following four industry classifications: Accommodations and Foodservices (117 reported card checks), Health Care and Social Assistance (111 reported card checks), Information (73 reported card checks), and Public Administration (126 reported card checks). Furthermore, card checks accounted for the majority of new union members in five industries, including the aforementioned "Information" and "Accommodation and Food Services" industries, as well as "Apparel Manufacturing," "Amusement, Gambling and Recreation," and "Food and Beverages."

Not surprisingly these industries have traditionally been the core organizing grounds for the unions that we identified above as being the most active users of the card check process. For example, the SEIU has traditionally organized workers in the health care and public service sectors. (90) UNITE and HERE have enjoyed a strong presence in the apparel, hospitality, and food services industries, (91) while the CWA has organized workers in broadcasting and telecommunications. (92)

The data in Tables 3 and 4 show that under the present legal landscape, the use of card checks is fairly concentrated among a few unions and within those unions' traditional industry organizing targets. Although a systematic analysis of the reasons behind this finding is beyond the scope of this Article, we submit that economic factors related to the industries in which card checks are most common may help explain this finding. If this is correct, and assuming that there is no change in the existing legal framework regulating the organizing process or in industry conditions, one would expect the use of card checks to continue to be largely limited to those industries where card checks are currently used.

If, on the other hand, the NLRA is amended along the lines of the proposed Employee Free Choice Act of 2007, and unions are given the right to demand card check recognition, unions will likely rely on card checks more heavily in order to avoid the more cumbersome and time consuming election process. (93) At the same time, to the extent employer opposition to unions is based on economic or industry driven factors, one should not expect employer acquiescence to union organizing. Instead employer opposition to unions would likely appear at other points in the organizing process. In other words, employers would adapt to the new legal paradigm regulating union organizing in ways that protect their interests.

To see this, consider the reasons employers might have for opposing an organizing drive. Employer opposition to unions has been modeled primarily as an economic decision. (94) According to this model, when facing a union organizing drive employers make a fairly straightforward cost-benefit calculation and act accordingly. In particular, employers evaluate the likely effect that engaging in opposition tactics will have on the likelihood of a union victory, the costs associated with mounting a campaign against the union, and the extent to which the presence of a union will affect the firm's profitability. Employers are more likely to oppose the organizing drive in situations where such opposition is likely to impact the outcome of the election, (95) where the costs associated with mounting an anti-union campaign are small, (96) and the impact that a union could have on the firm's profitability is large. (97)

Recent industry-level studies of the industries that our data suggest have been the primary targets of card check activity provide evidence that in each of these industries the factors mentioned above were such that one would expect less employer opposition to union organizing. For example, consider the health care industry. Because of an increased demand for health care services, fueled in part by the aging U.S. population and the increased demand for medical care, there has been a corresponding increase in the demand for nurses. (98) This heightened competition for nurses resulted in an industry-wide push for higher wages. (99) As a result, the union/ nonunion wage differential among all nurses has been relatively small as compared to all other occupations. (100) The relatively small union/non-union wage differential among nurses suggests that, other things held constant, the union's effect on the profitability of the firm is likely to be relatively small. In turn, employers in the industry should be less likely, as compared to employers at large, to oppose a union organizing campaign. (101) The willingness of employers to voluntarily agree to recognize the union by means of a card check process is evidence of this.

A similar incentive has been identified in the hotel and hospitality industry. In particular, the upper-segment establishments in the hotel industry (102) have adopted a high-price/high-quality business strategy, under which competition is based on the delivery of high-quality service to their customers. (103) This business strategy requires the existence of a stable and well-trained labor force. There is some evidence that employers in the hotel industry viewed unions as a means to reduce employee turnover and facilitate the accumulation of human capital via training and a longer term employment relationship. (104) To the extent that employers believed that having a union could lead to lower turnover and training costs, one would expect less opposition to union organizing activities. Again, this lower level of opposition is reflected in the willingness of employers to enter into card check agreements.

The telecommunications industry experienced some similar dynamics. In the 1990s, employers in the information and telecommunications industry realized that an amicable co-existence with the union was a pre-requisite to gain the union's support in obtaining necessary regulatory relief and in acquiescing to necessary corporate restructuring. (105) The major employers in the telecommunications industry (which were for the most part the regional components of the Bell System) sought changes in the rate structures for local telecommunications markets. (106) As those rates were regulated at the state level by state public services commissions, (107) an opportunity opened for the use of union political leverage. The unions, however, were reluctant to cooperate unless the major employers agreed to provide unions with some guarantee of institutional security. These guarantees came in the form of promises to avoid opposition campaigns to union organizing, relying instead on card checks and neutrality agreements. (108)

Finally, the high level of reported card checks in the public sector is not surprising. Public sector employers have traditionally been less hostile towards unions than their private sector counterparts. (109) Survey data indicate that public sector employees perceive their employers as generally more likely to welcome a union than private employers, and also more willing to share authority over workplace issues. (110) Research also indicates that public sector employers are less likely to engage in both legal and illegal antiunion tactics in the course of an organizing campaign. (111) Thus, to the extent that public sector employers have a higher level of tolerance for unions than private sector employers, it is not surprising that they have been frequent users of the card check process.

This short overview of these four industry groups suggests that there exist various industry-specific factors which help explain the higher levels of card check numbers in these industries. In particular, it appears that employers operating in these industries had some clear economic reasons for voluntarily agreeing to card checks. While card checks occurred in several other industrial sectors, (112) the fact that these four sectors accounted for a vast majority of the total number of card check events suggests that industry specific economic characteristics might be playing an important role. To be sure, the unions that are primarily identified with these industries (the CWA, the SEIU, and UNITE-HERE) have been traditionally characterized as creative and entrepreneurial in their organizing efforts, and they have successfully used card checks in various other industries. (113) As such, it is possible that over time they themselves, or in association with other unions, could expand the use of card checks to other industrial sectors. On the other hand, the data reported in WIP appear to indicate that the success of card checks has been limited to some fairly specific industries.

V. IMPLICATIONS

Our data provide a new and somewhat more nuanced picture of the use of card checks in relation to the various forms of organizing activities unions frequently utilize. In particular, our data show that while important, card checks do not appear to be the most frequently used organizing tactic. According to the data collected from WIP, about 18% of all organizing activities and about 20% of all newly organized employees were organized via card checks. We find, however, that card checks have become an important tool in some specific contexts. Over two thirds of the reported card checks (64.6%) involved private sector employees covered under the NLRA, which suggests that card checks are primarily used among employees who otherwise would have to be organized through the NLRA process. Thus, although the claim that "[b]oard elections have ceased to be the dominant mechanism for determining whether employees want union representation" (114) appears to be an overstatement, card checks are an important component of organizing private sector workers covered by the NLRA. In particular, our data show that for some unions card checks have become an extremely important organizing tool. For at least one union (UNITE-HERE) card checks represent the majority of their organizing activities. Similarly, for several other unions, (e.g. CWA, UFCW), sizeable proportions of their newly organized private sector employees have been organized via card checks.

Our data also show that a substantial proportion of card check activity has occurred in a small number of industries. We argue that the concentration of card checks in these industries is not surprising because of conditions that make it more economically feasible for employers to deal with unions. These conditions reduce employers' incentives to aggressively fight union organizing efforts.

By the same token, the lower level of card check activity in other industries represented in our data suggests that employers in those industries may face economic incentives to oppose organizing efforts, or at least to refuse to voluntarily recognize the union on the basis of a card check showing of interest. Amending the NLRA to allow the Board to certify unions on the basis of card checks will not change the economic incentives employers in these industries face for avoiding unions and, consequently, they are likely to adopt various tactics designed to prevent union organizing. A change in the law may simply cause a shift in the timing at which these employers will mount their anti-union campaigns.

In particular, we suggest that employers who presently have no incentives to agree to card check agreements could adopt preemptive tactics designed to reduce the likelihood of a union organizing campaign. As is true of tactics which employers currently use during an election campaign, some of these tactics might prove to be legal and some illegal. For example, employers could respond by adopting positive/proactive measures intended to eliminate the employees' need for union representation. These measures could include: establishing an adequate and equitable compensation system; developing positive supervisory-employee relations; establishing open channels of communication; and adopting some form of alternative dispute resolution system to deal with employees' concerns. (115) These tactics are intended to eliminate the types of issues that a union could use to generate support for an organizing campaign. (116) Employees who are satisfied with their current working conditions are generally less likely to participate in organizing campaigns and less likely to ultimately vote for union representation. (117)

Employers, however, could also adopt negative/proactive tactics. For example, employers could aggressively screen out pro-union job candidates during the hiring process. (118) This screening could be accomplished in a variety of ways. Employers could question job applicants about their union sentiments. While asking such questions directly is illegal, (119) various observers have noted that companies could use indirect methods to achieve the same objective. (120) For example, employers in an industry with traditionally high unionization rates might seek to hire employees with no prior work experience in the industry, on the assumption those employees are less likely to have belonged to a union. (121)

Although the NLRA protects employees against some of these practices, research shows that the NLRA is virtually unknown to a large number of employees. Section 7 of the NLRA explicitly protects all employees who engage in "concerted activities for the purpose of ... other mutual aid or protection." (122) This protection applies not just to employees represented by labor unions, but to nearly all private sector employees in the United States, and the United States Supreme Court has directly upheld the NLRA's broad-scale protection of non-unionized workers in this regard. (123) At the same time, the application of the NLRA in non-union settings is, as Professor William R. Corbett has noted, "one of the best-kept secrets" of employment law. (124) It is quite possible that many employees will simply be unaware that employers' tactics designed to screen out those with pro-union tendencies are illegal. (125)

Thus, the proposed amendments to the NLRA, intended to facilitate card check recognition, raise the possibility of increased union avoidance behavior by employers at pre-organizing stages in the employment relationship. (126) The shift in employers' tactics could prove to be particularly effective given the lack of employees' awareness of their rights under the NLRA. It could be argued that in anticipation of this change in employers' tactics, proposed amendments to provide card check recognition should include language addressing the potential illegality of the type of pre-organizing union avoidance campaign tactics we identify above.

In fact, the Employee Free Choice Act of 2007 includes a section establishing stronger penalties for employers' actions violating sections 8(a)(1) and 8(a)(3) while employees "were seeking representation by a labor organization or during the period after a labor organization was recognized as a representative" of the employees. (127) This language appears to expand currently available remedies once the campaign has started and later when the union is certified, but the proposed legislation does nothing to prevent the type of behavior we suggest might ensue on the part of employers.

VI. CONCLUSION

Card check recognition has been around for a long time and our data confirm the perception that card checks have become an important organizing tool for unions. Our data also support the assertion that card checks have become a particularly useful organizing tool for unions operating under the NLRA. We also find that while important, card checks have not, as of yet, become the primary approach to organizing new members. Thus, in addition to seeking legislation protecting the card check process, unions and their supporters should also continue to push for other ways of strengthening organizing protections provided under existing law.

For example, during President Clinton's administration the NLRB expanded the use of section 10(j) injunctions with the intent of further deterring section 8(a)(3) violations. (128) The Employee Free Choice Act requires the Board to use injunctions in cases where there is a reasonable cause to believe the employer has discharged or discriminated, or threatens to discharge or discriminate, against employees, or engaged in conduct that significantly interferes with employee rights during an organizing or first contract drive. (129) Regardless of what happens to the card check provisions, unions will benefit from the more assertive use of the 10(j) injunction. Similarly, prior attempts to reform the NLRA, such as the Labor Law Reform Act of 1977-78, focused on issues of equal access by employees to the workplace. (130) Improving the ability of unions to communicate with employees would also be of value to union organizing efforts. Of course, whether any of these proposals would likely pass in today's political climate is highly doubtful, particularly given the fact that the AFL-CIO and labor unions generally are in a much weaker economic and political position today. (131)

Our finding that card check recognition is largely confined to certain industries suggests that employers who presently refuse unions' requests for card check recognition are likely to find ways of opposing organizing drives, even under the proposed legislation. These new tactics could be extremely effective in undermining the ability of employees to form unions. Existing legislative proposals are silent in this regard. Our research suggests that the possibility of a shift in employers' tactics should be recognized and adequately addressed.

APPENDIX
OFFICIAL FULL NAME OF UNIONS INCLUDED IN TABLE 3

Union     Name
Acronym

AFGE      American Federation of Government Employees
AFSCME    American Federation of State, County and Municipal
          Employees
AFT       American Federation of Teachers
CWA       Communications Workers of America
HERE      Hotel Employees & Restaurant Employees Union
IAM       International Association of Machinists
IBT       International Brotherhood of Teamsters
IFPTE     International Federation of Professional & Technical
          Engineers
OPEIU     International Union of Operating Engineers
SEIU      Service Employees International Union
UAN       United American Nurses
UAW       United Automobile, Aerospace & Agricultural Implement
          Workers of America
UFCW      United Food and Commercial Workers International Union
UNITE     Union of Needletrades, Industrial and Textile Employees
USWA      United Steelworkers of America

(1.) See Steven Greenhouse, Labor Presses for Measure to Ease Unionizing, N.Y. TIMES, Dec. 16, 2006, at A28 (describing the ongoing effort by the labor movement to enact legislation regarding the use of card checks).

(2.) See Steven Greenhouse, Employers Sharply Criticize Shift in Unionizing Method to Cards From Elections, N.Y. TIMES, Mar. 11, 2006, at A9 (describing employers' opposition to the card check process). See generally Carl F. Horowitz, Why Union Card Checks Are Coercive, NAT'L LEGAL & POL'Y CENTER, http://www.nlpc. org/view.asp?action=viewArticle&aid=1638 (last visited Jan. 02, 2008); James Sherk, How Union Card Checks Block Workers' Free Choice, HERITAGE FOUND., Feb. 21, 2007, http://www.heritage.org/Research/Labor/wm1366.cfm.

(3.) See ADRIENNE EATON & JILL KRIESKY, AMERICAN RIGHTS AT WORK, FACT OVER FICTION: OPPOSITION TO CARD CHECK DOESN'T ADD UP (Mar. 2006), available at http://www.americanrightsatwork.org/dmdocuments/ARAWReports/IBFact OverFictFinal.pdf (examining the validity of claims by anti-union groups that card check campaigns leave employees more vulnerable to union pressure than during National Labor Relations Board elections); see also GORDON LAFER, AMERICAN RIGHTS AT WORK, FREE AND FAIR?: HOW LABOR LAW FAILS U.S. DEMOCRATIC ELECTION STANDARDS (June 2005), available at http://www.amerieanrightsatwork. org/dmdocuments/ARAWReports/FreeandFair%20FINAL.pdf (assessing the extent to which National Labor Relation Board elections embody democratic principles).

(4.) See Paul F. Clark et al., Private-Sector Collective Bargaining: Is This the End or a New Beginning?, in COLLECTIVE BARGAINING IN THE PRIVATE SECTOR 1, 8-9 (Paul F. Clark et al., eds. 2002) (discussing the complaints unions have voiced about the current legal framework regulating the collective bargaining process); see also THOMAS GEOGHEHAN, WHICH SIDE ARE YOU ON? 252-56 (1991) (discussing various unions' concerns regarding the existing labor laws).

(5.) Pub. L. No. 74-198, 49 Stat. 449 (1935) (codified as amended at 29 U.S.C. [section][section] 151-169 (2000)).

(6.) See Cathy Trost & Leonard M. Apcar, AFL-CIO Chief Calls Labor Laws a 'Dead Letter'--Kirkland Says the Federation Would 'Seriously' Study Repeal of All But the Basic, WALL ST. J., Aug. 16, 1984, at 8 (noting Kirkland's frustration with President Reagan's administration of the NLRA).

(7.) See generally Linda Chavez-Thompson, Labor and the American Dream, DENVER POST, Sept. 3, 2007 available at http://www.denverpost.com/search/ci_6787565 (discussing the recent failed attempt by Congress to enact legislation authorizing the use of card checks).

(8.) S. 1041 [section] 2(a), 110th Cong. [section] 1 (2007).

(9.) Id.

(10.) Id. [section] 4(b). The proposed amendments provide for civil fines up to $20,000 per violation against employers found to have willfully or repeatedly violated employees' rights during an organizing campaign or first contract drive. The EFCA also increases the amount an employer is required to pay when an employee is discharged or discriminated against during an organizing campaign or first contract drive to three times back pay. Finally, the EFCA requires the Board to seek a federal court injunction against an employer whenever there is reasonable cause to believe the employer has discharged or discriminated against employees, threatened to discharge or discriminate against employees, or engaged in conduct that significantly interferes with employee rights during an organizing or first contract drive. Id.

(11.) See Thomas A. Kochan et al., The Effects of Corporate Strategy and Workplace Innovations on Union Representation, 39 INDUS. & LAB. REL. REV. 487 (1986) (noting that a significant percentage of employers considered being nonunion their major labor relations goal).

(12.) But cf. James J. Brudney, Neutrality Agreements and Card Check Recognition: Prospects for Changing Paradigms, 90 IOWA L. REV. 819, 828-30 (2005); Adrienne E. Eaton & Jill Kriesky, Union Organizing Under Neutrality and Card Check Agreements, 55 INDUS. & LAB. RAZE. REV. 42 (2001) [hereinafter Eaton & Kriesky, Union Organizing]; Jennifer Dillard & Joel Dillard, Fetishizing the Electoral Process: The NLRB's Problematic Embrace of Electoral Formalism (Working Paper Series, Aug. 24, 2007), available at http://ssrn.com/abstract=1009636.

(13.) The data were collected from the AFL-CIO's weekly publication Work in Progress. Part III of this Article discusses the origins of the publication as well as the advantages and disadvantages of relying on this source.

(14.) For example, the reported events include reported mergers and affiliation agreements between members of the AFL-CIO and independent unions. See infra notes 65-67 and accompanying text.

(15.) See, e.g., Brudney, supra note 12, at 828 (discussing the increasing use of card checks, and corresponding decline in government sponsored elections); Rick Valliere, Unions Turning Away from NLRB Elections as Primary Way of Organizing, Raynor Says, DAILY LAB. REP. (BNA), Jan. 12, 2006, at C-3 (reporting on comments by Bruce S. Raynor, president of UNITE-HERE, to the extent that only ten percent of the employees joining UNITE-HERE during 2005 joined via NLRB conducted elections).

(16.) See Brudney, supra note 12, at 828-29.

(17.) See infra notes 21-50 and accompanying text.

(18.) See infra notes 51-61 and accompanying text.

(19.) See infra notes 62-113 and accompanying text.

(20.) See infra notes 114-27 and accompanying text.

(21.) For a description of the organizing process, see ROBERT A. GORMAN & MATTHEW W. FINKIN, BASIC TEXT ON LABOR LAW: UNIONIZATION AND COLLECTIVE BARGAINING 53-67 (2d ed. 2004).

(22.) See Kate Bronfenbrenner, The Role of Union Strategies in NLRB Certification Elections, 50 INDUS. & LAB. REL. REV. 195, 201-05 (1997) (listing several tactics commonly used by employers in the course of organizing campaigns); William T. Dickens, The Effect of Company Campaigns on Certification Elections: Law and Reality Once Again, 36 INDUS. & LAB. REL. REV. 560, 563-67 (1983) (analyzing the effect of employer tactics in organizing election outcomes); John J. Lawler, The Influence of Management Consultants on the Outcome of Union Certification Elections, 38 INDUS. & LAB. REL. REV. 38, 38-39 (1984) (describing the effects of the use of management consultants on union organizing elections).

(23.) See RICHARD B. FREEMAN & JAMES L. MEDOFF, WHAT DO UNIONS Do? 23436 (1984) (discussing the findings of several studies regarding the effect of employers' campaigns in election outcomes); see also Kate L. Bronfenbrenner, Employer Behavior in Certification Elections and First Contract Campaigns: Implications for Labor Law Reform, in RESTORING THE PROMISE OF AMERICAN LABOR LAW 75, 80-82 (Sheldon Friedman et al. eds., 1994) (discussing the impact of a variety of employers' tactics both in election outcomes and in successful first contract negotiations).

(24.) Captive audience speeches refer to mandatory meetings held at the workplace, during work time, in which employers present their case for opposing the union. See TERRY L. LEAP, COLLECTIVE BARGAINING & LABOR RELATIONS 146-48 (2d ed. 1995).

(25.) See Lawler, supra note 22, at 39.

(26.) See NLRB v. Gissel Packing Co., 395 U.S. 575 (1969).

(27.) See Central Valley Meat Co., 346 N.L.R.B. No. 94 (Apr. 28, 2006) (holding that employer violated the NLRA by threatening to fire and firing various employees during the course of an organizing campaign).

(28.) See Reeves Bros., 320 N.L.R.B. 1082 (1996) (finding that the employer had violated section 8(a)(1) of the NLRA by reading to employees during a captive audience meeting two letters from two of the company's clients indicating that the clients would reduce or eliminate their business if the union won the organizing election).

(29.) See NLRB v. Exchange Parts Co., 375 U.S. 405,409 (1964) (comparing benefits granted during an organizing campaign to a "fist inside the velvet glove," the Court held that the granting of benefits during an organizing campaign was a violation of the NLRA).

(30.) Charles J. Morris, A Tale of Two Statutes: Discrimination for Union Activity Under the NLRA and RLA, 2 EMP. RTS. & EMP. POL'Y J. 317, 330-31 (1998) (noting that the violations of section 8(a)(3)--the type of unfair labor practice charge most likely to be filed in the course of an organizing campaign--per election held, increased four times between 1968 and 1998); see also JOHN SCHMITT & BEN ZIPPERER, CENTER FOR ECONOMIC AND POLICY RESEARCH, DROPPING THE AX: ILLEGAL FIRINGS DURING UNION ELECTION CAMPAIGNS (Jan. 2007) (finding a steep rise in the 2000s relative to the last half of the 1990s in illegal firings of pro-union workers). Although the upward trend in employers' violations of the NLRA is well documented, the reasons behind the increase are less clear. Some commentators argue the increase could be attributed to the relatively weak penalties provided under the NLRA. See Morris, supra, at 318.

(31.) See Brudney, supra note 12, at 825-26.

(32.) See Dillard & Dillard, supra note 12, at 16.

(33.) See Brudney, supra note 12, at 828-30; Dillard & Dillard, supra note 12, at 11.

(34.) See Dillard & Dillard, supra note 12, at 14.

(35.) Id. at 16 (arguing that the Board's shift towards preference for secret ballot elections was a move taken in response to challenges to the Board's impartiality and institutional validity).

(36.) See In re Cudahy Packing, 13 N.L.R.B. 526, 531-32 (1939) (indicating a preference for elections where two unions claimed majority status).

(37.) Among the amendments to the NLRA Congress adopted in 1947 was section 9(c)(1)(b) providing that "[i]f the Board finds upon the records of such hearing that such a question of representation exists, it shall direct an election by secret ballot and shall certify the results thereof." 29 U.S.C.A. [section] 159 (c)(1)(b) (West 2000).

(38.) See Dillard & Dillard, supra note 12, at 18.

(39.) See Brudney, supra note 12, at 833.

(40.) Id. at 832-35.

(41.) See Bronfenbrenner, supra note 23, at 78-79.

(42.) Id.

(43.) See Paula B. Voos, Union Organizing: Costs and Benefits, 36 INDUS. & LAB. REL. REV. 576, 576-77 (1983) (discussing the costs associated with mounting an organizing campaign).

(44.) See THOMAS A. KOCHAN ET AL., THE TRANSFORMATION OF AMERICAN INDUSTRIAL RELATIONS 37-45 (1986) (describing historical trends in employers' opposition to labor unions).

(45.) See Brudney, supra note 12, at 835-40.

(46.) These costs could include hiring the consultant, running the campaign, lost work time, and legal expenses. For those employers that have an existing bargaining relationship, an additional cost is the potential harm to the labor-management relationship associated with an anti-union campaign. The decision by employers to agree to a card check procedure can be motivated as well by the desire to avoid the negative business consequences associated with a union led corporate campaign. See Eaton & Kriesky, Union Organizing, supra note 12, at 48-51; see also Adrienne E. Eaton & Jill Kriesky, Dancing with the Smoke Monster: Employer Motivations for Negotiating Neutrality and Card Check Agreements, in JUSTICE ON THE JOB" PERSPECTIVES ON THE EROSION OF COLLECTIVE BARGAINING IN THE UNITED STATES 139, 147-50 (Richard N. Block et al. eds., 2006) [hereinafter Eaton & Kriesky, Employer Motivations].

(47.) See Eaton & Kriesky, Employer Motivations, supra note 46, at 144-47.

(48.) Id. at 146.

(49.) See Brudney, supra note 12, at 838.

(50.) See Steven Greenhouse, Clash Nears in the Senate on Legislation Helping Unions Organize, N.Y. TIMES, June 20, 2007, at A16 (pointing out that business groups have mounted a major campaign against the proposed legislation).

(51.) See AFL-CIO, The First Bulletin, WORK IN PROGRESS, (Jan. 23, 1996), http:// www.aficio.org/aboutus/thisistheaflcio/publications/wip/wip01231996.cfm (announcing the first issue of Work in Progress).

(52.) Id.

(53.) See AFL-CIO, Editor's Note, WORK IN PROGRESS, (July 27, 1998), http://www. aflcio.org/aboutus/thisistheaflcio/publications/wip/wip07271998.cfm.

(54.) Our ability to code each case depended on the level of detail provided in WIP. If the information was not provided, or if the information was not clear, we coded it as missing data.

(55.) For example, WIP might identify the employees as "public employees." When WIP did not directly identify the employees as private sector or public sector, we attempted to identify the public or private nature of the employees by relying on the employer's name.

(56.) As far as we know, there is not public data identifying the total number of types of organizing events (i.e., elections, card checks, etc.) occurring in the Unites States. This lack of data is especially problematic with regard to card checks, since card checks are not recorded by the NLRB. See NATIONAL LABOR RELATIONS BOARD, N.L.R.B. ELECTION REPORT: CASES CLOSED AUGUST 2007, (Sept. 2007), http://www.nlrb.gov/nlrb/shared-files/brochures/Election%20Reports/ August2007.pdf (noting that the monthly election reports, on which the NLRB's annual reports are based, list the outcome of secret-ballot voting in NLRB-conducted representation elections); see also SCHMITT & ZIPPERER, supra note 30, at 7-8 (discussing how the lack of NLRB data on card checks affects the empirical analysis of illegal employers' activities during election campaigns).

(57.) See Eaton & Kriesky, Union Organizing, supra note 12, at 45.

(58.) Id.

(59.) See infra notes 87-92 and accompanying text.

(60.) Of the 730 card check events reported in our data, 523 (72%) can be categorized as occurring in the service sector, e.g., "Food & Beverages/General Merchandise Stores," "Information," "Educational Services," "Health Care and Social Assistance," "Performing Arts, Spectator Sports, and Related Industries," "Amusement, Gambling and Recreation Industries," "Accommodation and Food Services," "Personal and Laundry Services," and "Public Administration." See infra p. 275 tbl.4.

(61.) Of the 523 events occurring in the service sector, 384 (or 73%) involved employers in the communications, hospitality and health care industries. See id.

(62.) The NLRB jurisdiction is rather broad, covering any private sector enterprise whose operations affect commerce. See 29 U.S.C. [section] 160.

(63.) This category includes public sector employees (federal, state, and local governments). Public sector employees do not fall under the jurisdiction of the NLRB, and instead are covered by a variety of other federal and state collective bargaining laws. "Non-NLRA Employees" also include railroad and airline employees, who are covered under the Railway Labor Act, 45 U.S.C.A. [section][section] 151-188 (West 2007), and employees organized by AFL-CIO affiliated unions in Canada. Section 2(2) of the NLRA specifically excludes "It]he United States or any wholly owned Government corporation, or any Federal Reserve Bank, or any State or political subdivision thereof, or any person subject to the Railway Labor Act...." See 29 U.S.C. [section] 152(2). Section 2(3) excludes "agricultural laborers" from the NLRA's coverage. See id. [section] 152(3).

(64.) See Henry S. Farber, Union Membership in the United States: The Divergence between the Public and Private Sectors (Princeton Univ. Indus. Rel. Section, Working Paper No. 503, 2005) (showing that while between 1972 and 2004 there has been a steady decline in union membership rates in the private sector, union membership rates in the public sector increased dramatically during the 1970s and have remained fairly steady since then).

(65.) The "election" category includes all government conducted activities. Thus, it includes elections under the NLRA, as well as elections conducted under the various public sector laws (both federal and state). See infra p. 273 tbl.1.

(66.) The "Card Check" category includes all those events reported as a card check in WIP. See id.

(67.) The "Mergers and Affiliations" category includes all those events reported in WIP as such. "Mergers" refer to the merging of two unions. See, e.g., AFL-CIO, Driving and Flying, WORK IN PROGRESS (Jan. 8, 2001), http://www.aflcio.org/ aboutus/thisistheaflcio/publications/wip/wip01082001.cfm (reporting the merger between the unaffiliated Chicago Truck Drivers Union and Teamsters Local 710). "Affiliations" include situations in which a previously independent union became affiliated directly with the AFL-CIO or with a union that was already a member of the AFL-CIO. See, e.g., AFL-CIO, Public Employees Join SEIU, WORK IN PROGRESS, May 26, 1998, http://www.aflcio.org/aboutus/thisistheaflcio/publications/wip/ wipO5261998.cfm (reporting the decision by members of the Public Employees of Riverside County organization to affiliate with the SEIU). In addition to the three categories identified in the text (i.e., elections, card checks and mergers/affiliations) WIP reported other forms of organizing events. These included, for example, an employer signing an area wide existing collective bargaining agreement, the reporting of some new members joining an existing bargaining unit, and an election conducted by a non-governmental organization.

(68.) See Clark et al., supra note 4, at 5-7 (describing the effects on unions and union related activity of the September 11, 2001 events).

(69.) See Steven Greenhouse, Four Major Unions Plan to Boycott A.F.L.-C.I.O Event, N.Y. TIMES, July 25, 2005, at A1, A16. These unions included the International Brotherhood of Teamsters, the Service Employees International Union, UNITE-HERE, and the United Food and Commercial Workers International Union. These four unions were later joined by three other unions in the Change to Win Coalition (the Laborers International Union of North America, the United Farm Workers, and United Brotherhood of Carpenters and Joiners of America).

(70.) See infra notes 81-82 and accompanying text.

(71.) See Brudney, supra note 12, at 828.

(72.) Table 2 shows that of the 982,926 employees organized through elections, 641,757 (588,206 "Public Employees" and 53,551 "Other Non-NLRA Employees") were not covered by the NLRA. In comparison, only 341,169 of all the employees organized through elections were covered by the NLRA. See infra p. 273 tbl.2.

(73.) Of the 245,823 employees organized via card checks, 159,022 were employees covered by the NLRA, while 86,801 were not covered by the Act. See id.

(74.) For example, Charles Morris argues that employers operating under the Railway Labor Act are significantly less likely to engage in illegal campaign activities than employers operating under the NLRA. See Morris, supra note 30, at 317-18.

(75.) See Richard B. Freeman, Through Public Sector Eyes: Employee Attitudes Toward Public Sector Labor Relations in the U.S., in PUBLIC SECTOR EMPLOYMENT IN A TIME OF TRANSITION 59, 79-80 (Dale Belman et al. eds., 1996).

(76.) See Brudney, supra note 12, at 828-30.

(77.) Id. at 828.

(78.) Id. at 829.

(79.) Id.

(80.) Id.

(81.) See infra p. 274 tbl.3.

(82.) See id.

(83.) Id.

(84.) These six unions reported having successfully participated in 555 out of the 730 card checks included in the data. Id.

(85.) UNITE-HERE, created in 2004 as a merger between UNITE and HERE reported fourteen card check events (about 74% of all its reported events). For the International Federation of Professional and Technical Engineers ("IFPTE") and the International Union of Operating Engineers ("OPEIU"), card checks represented about 21% of all their organizing activities. Michael Bologna, UNITE, HERE Delegates Approve Merger, DAILY LAB. REP. (BNA), July 9, 2004 at A-9.

(86.) The same was true for UNITE-HERE. Eighty-seven percent of the 3782 "NLRA Employees" organized by UNITE-HERE joined via card checks. Id.

(87.) See Jack Rasmus, The AFL-CIO Split: What Next for American Labor?, Z MAG. ONLINE, Sept. 2005, http://zmagsite.zmag.org/Sep2005/ rasmuspr0905.html.

(88.) See infra p. 275 tbl.4.

(89.) We used the North American Industry Code System (NAICS), which is used in the National Labor Relations Board's Election Reports. "The NAICS was developed jointly by the U.S., Canada, and Mexico to provide new comparability in statistics about business activity across North America." U.S. Census Bureau, North American Industry Classification System, http://www.census.gov/epcd/ www/naics.html (last visited Jan. 3, 2008). For NLRB-conducted elections, we used the same entry as reported in the Election Reports. For those events not included in the Election Reports (e.g., card checks, or those involving Non-NLRA Employees), we decided the appropriate industry code based on the information provided in WIP. When necessary we looked at various sources to assess the appropriate industry category. See infra p. 275 tbl.4.

(90.) See Service Employees International Union, SEIU History, http://www.seiu.org/about/seiu_history/index.cfm (last visited Jan. 4, 2008).

(91.) See UNITE-HERE, What is UNITE HERE?, http://www.unitehere.org/ about/ (last visited Jan, 3, 2008).

(92.) See Communications Workers of America, Profile & History, http://www.cwaunion.org/about/profile.html (last visited Jan. 4, 2008).

(93.) See supra note 8 and accompanying text.

(94.) See Richard B. Freeman & Morris M. Kleiner, Employer Behavior in the Face of Union Organizing Drives, 43 INDUS. & LAB. REL. REV. 351, 364 (1990) (developing an empirical model of employer opposition to organizing campaigns).

(95.) Where the employer thinks that the union is very likely to win, or where the union is highly unlikely to win, one would expect the employer to be less inclined to engage in opposition activities, On the other hand, where the outcome of the election is uncertain, employers are likely to actively oppose the organizing drive. Id. at 352-53.

(96.) Conducting a campaign in opposition of the union could be an expensive proposition for the employer. For example, hiring consultants and training supervisors to respond to the organizing efforts is costly both in terms of direct payments, as well as lost productivity time. In addition, employers might choose to engage in marginally or clearly unfair labor practices, which will require the employer to pay for legal advice and potentially legal representation. Id.

(97.) In particular, the larger the wage differential the union is expected to negotiate, the more likely the employer is to oppose the organizing drive. Id.

(98.) See Paul F. Clark, Health Care: A Growing Role for Collective Bargaining, in COLLECTIVE BARGAINING IN THE PRIVATE SECTOR 91, 118 (Paul F. Clark et al. eds., 2002).

(99.) Id. at 118, 122-23.

(100.) The union/non-union wage differential for nurses in 2001 was 13%. The union/ non-union wage differential for all other occupations was 28%. Id. at 122.

(101.) The fact that employers in the health care industry might be more likely, as compared to employers in general, to agree to card checks, does not mean that all health care employers will have exactly the same incentives to acquiesce to such demands by unions. Thus, it is not surprising to see unions organizing this industry using corporate campaigns to put pressure on employers to recognize the union. See, e.g., Jessica Fargen, Union Taking Aim at Hospitals: Kicks Off Ad Campaign to Unionize Workers, BOSTON HERALD, Sept. 15, 2007, at 5.

(102.) "Upper-segment establishments" include "high price upscale hotels, as well as mid-priced full-service properties. See C. Jeffrey Waddoups & Vincent E. Eade, Hotels and Casinos: Collective Bargaining During a Decade of Expansion, in COLLECTIVE BARGAINING IN THE PRIVATE SECTOR 137, 140 (Paul F. Clark et al. eds., 2002).

(103.) Id. at 141.

(104.) Id.

(105.) See Jeffrey Keefe & Rosemary Batt, Telecommunications: Collective Bargaining in an Era of Industry Reconsolidation, in COLLECTIVE BARGAINING IN THE PRIVATE SECTOR 263, 278-79 (Paul F. Clark et al. eds., 2002).

(106.) Id.

(107.) Id.

(108.) Id. at 300.

(109.) See Farber, supra note 64, at 13.

(110.) See Freeman, supra note 75, at 79-80.

(111.) See Richard Freeman, Unionism Comes to the Public Sector, 24 J. or ECON. LITERATURE 41, 49 (Mar. 1986) (discussing unionization trends in the public sector).

(112.) In fact, in some of the other industrial sectors included in our data, card checks represented a fairly large percentage of all their organizing activity. For example, in the Amusement, Gambling, and Recreational Facilities Industry, 77% (10 out of 13) of all organizing events were card checks. In the "Apparel Manufacturing" sector 13 out of 24 total events (54%) involved card check organizing.

(113.) For example, Table 4 shows that 73% of the 5066 employees organized in the "Apparel Manufacturing" industry and 96% of the 11,994 workers in the "Amusement, Gambling, and Recreation" industry were organized via card checks. See infra p. 275 tbl.4. Both of these industries have been traditional UNITE-HERE targets. See Bologna, supra note 85, at A-9 (noting that at the time of their merger, UNITE organized workers in the clothing, textiles, and laundry industries, while HERE organized workers in the hotel, restaurant, and gaming industries).

(114.) See Brudney, supra note 12, at 824.

(115.) See LEAP, supra note 24, at 146-48.

(116.) Id. at 147.

(117.) Id.

(118.) Id. at 148.

(119.) See, e.g., Center Construction Company, Inc., Cases 7-CA-46490, 2004 WL 2138582 (NLRB Sept. 21, 2004) (finding questions concerning union sympathies in the context of job application interviews to be inherently coercive); Rochester Cadet Cleaners, Inc., 205 N.L.R.B. 773 (1973) (finding that the employer violated the NLRA by asking a job applicant whether her former employers were unionized and whether she had belonged to a union).

(120.) See Gregory M. Saltzman, Job Applicant Screening by a Japanese Transplant: A Union-Avoidance Tactic, 49 INDUS. & LAB. REL. REV. 88, 91 (1995) (discussing some of the tactics used by Japanese automobile plants operating in the United States to screen out union sympathizers).

(121.) Id.

(122.) 29 U.S.C. [section] 157.

(123.) See NLRB v. Wash. Aluminum Co., 370 U.S. 9 (1962); Morris, supra note 30, at 345-46.

(124.) See William Corbett, Waiting for the Labor Law of the Twenty-First Century: Everything Old Is New Again, 23 BERKELEY J. EMP. & LAB. L. 259, 267 (2002).

(125.) See generally Pauline T. Kim, Bargaining with Imperfect Information: A Study of Worker Perceptions of Legal Protection in an At-Will World, 83 CORNELL L. REV. 105, 106 (1997).

(126.) These proactive union avoidance practices are different than those practices adopted by employers in response to an organizing drive. In addition to screening employees with pro-union sentiments, employers can also establish a culture not conducive to union representation. See LEAP, supra note 24, at 146-48.

(127.) The EFCA allows for civil fines, increases back pay awards, and requires the NLRB to seek an injunction against employers in cases involving threats or terminations that interfere with employee rights during an organizing or first contract drive. S. 1041, 110th Cong. [section] 1 (2007).

(128.) See Morris, supra note 30, at 345-46.

(129.) S. 1041, 110th Cong. [section] 1.

(130.) See Comment, Labor Law Reform: The Regulation of Free Speech and Equal Access in NLRB Representation Elections, 127 U. PA. L. REV. 755 (1979). Despite the existence of a pro-union Democratic President supporting NLRA reform legislation, and pro-union Democratic majorities in both the United States House of Representatives and the United States Senate, the reform proposal ended up being successfully filibustered in the United States Senate. Id.

(131.) See Kris Maher, Signs of a Possible Power Shift in Congress Have Unions Going All Out to Reach Voters, WALL ST. J., Aug. 28, 2006, at A2.

Rafael Gely * and Timothy D. Chandler **

* Judge Joseph P. Kinneary Professor of Law, University of Cincinnati.

** Robert H. and Patricia Hines Professor of Management, William and Catherine M. Rucks Department of Management, E.J. Ourso College of Business, Louisiana State University.

TABLE 1
AFL-CIO REPORTS OF ORGANIZING METHODS,
1998-2005 *
                           Card       Mergers &
  Year      Elections     Checks     Affiliations   Other   Total

  1998         415          69            19          7      510
  1999         431         141            24          9      605
  2000         478         122             8         21      629
  2001         358          83            24         12      477
  2002         370          81            26          5      482
  2003         366          84            28          7      485
  2004         304          91             9          8      412
  2005         174          59            14          0      247
All Years     2896         730           152         69     3847

* There were ten organizing incidents for which we were unable to
determine the organizing method.

TABLE 2
AFL-CIO REPORTS OF EMPLOYEES JOINING
THE LABOR MOVEMENT THROUGH
VARIOUS ORGANIZING
METHODS, 1998-2005

NUMBER OF REPORTED INCIDENTS *

                                              Mergers &
     Total         Elections   Card Checks   Affiliations   Other

 All Employees      982,926      245,823       624,054      38,193
                    (2886)        (721)         (152)        (69)
 NLRA Employees     341,169      159,022       246,310      15,998
                    (2048)        (562)          (73)        (51)
Public Employees    588,206      82,356        368,020      13,069
                     (737)        (136)          (73)        (12)
 Other Non-NLRA     53,551        4,445         9,724       9,126
   Employees         (101)        (23)           (6)         (6)

* Missing data on the numbers of employees organized, whether or
not employees were covered by the NLRA, and/or the organizing
method reduced the usable observations for this table.

TABLE 3
AFL-CIO REPORTS OF EMPLOYEES JOINING
THE LABOR MOVEMENT BY UNION,
1998-2005 *

(Percent from Card Checks)

                                   Reported
           Reported    Reported      Card       Total
Union **   Incidents   Elections    Checks    Employees

AFGE           34          30          3       31,556
                                                (3.5)
AFSCME        493         401         70       177,629
                                               (20.0)
AFT           191         143         23       147,233
                                                (7.6)
CWA           249         126        103       63,306
                                               (49.1)
HERE          233          82        144       48,254
                                               (80.0)
IAM           110          92         12       154,895
                                                (1.9)
IBT           319         285         26       47,340
                                               (17.1)
IFPTE          19          11          4       56,645
                                                (1.6)
OPEIU          23          13          5       33,576
                                                (5.0)
SEIU          775         592        138       507,402
                                                (9.3)
UAN             8           6          0       105,500
                                                (0.0)
UAW           207         176         29       89,989
                                               (20.6)
UFCW          365         303         55       74,727
                                               (24.5)
UNITE         130          80         45       33,826
                                               (28.0)
UNITE-         19           5         14        4,053
HERE                                           (82.6)

                         Non-
             NLRA        NLRA
Union **   Employees   Employees

AFGE          200       31,356
             (0.0)       (3.5)
AFSCME      23,516      154,113
            (11.0)      (21.4)
AFT         15,185      132,048
             (4.1)       (8.0)
CWA         46,463      16,843
            (58.7)      (22.4)
HERE        45,791       2,463
            (82.1)      (40.8)
IAM         13,620      141,275
            (21.1)       (5.0)
IBT         27,860      19,480
            (11.2)      (25.5)
IFPTE       54,110       2,535
             (1.4)       (4.8)
OPEIU       31,815       1,761
             (1.7)      (64.7)
SEIU        162,230     344,030
            (16.7)       (5.8)
UAN         100,500      5,000
             (0.0)       (0.0)
UAW         54,782      35,207
            (20.0)      (21.6)
UFCW        64,785       9,942
            (23.8)      (29.2)
UNITE       31,456       2,370
            (30.1)       (0.0)
UNITE-       3,782        271
HERE        (87.9)       (7.7)

* Used 30,000 new employees as selection cut-off, except for
UNITE-HERE.

** See Appendix for a list of the full names of the unions
included in the table.

Notes: Table 3 excludes the California State Employee Association
(CSEA). The 175,000 CSEA members affiliated with the AFL-CIO in 2001.

TABLE 4
AFL-CIO REPORTS OF EMPLOYEES JOINING
THE LABOR MOVEMENT BY
INDUSTRY, 1998-2005 *

(Percent from Card Checks)

                        Reported    Reported     Reported       Total
      Industry          Incidents   Elections   Card-Checks   Employees

Food Manufacturing          80          67           12        20,058
                                                               (21.9)
Apparel                     24          11           13         5,066
Manufacturing                                                  (73.3)

Transportation             112          85           22        89,562
Equipment                                                      (11.9)
Manufacturing

Food & Beverages/           58          34           23         9,037
General Merchandise                                            (53.6)
Stores

Information                168          83           73        42,743
(Broadcasting &                                                (58.5)
Telecommunications;
Information Services
& Data Processing)

Educational Services       120         103           13        23,884
                                                                (3.2)

Health Care and            865         720          111        317,034
Social Assistance                                               (8.6)

Performing Arts,            37          23           12         9,434
Spectator Sports, and                                          (37.0)
Related Industries

Amusement,                  13           3           10        11,994
Gambling, and                                                  (96.1)
Recreation Industries

Accommodation and          219         97           117        42,527
Food Services                                                  (78.6)

Personal and               113         71           38         24,139
Laundry Services                                               (31.0)

Public                     857         655          126        912,738
Administration                                                  (8.6)

* Report only includes industries having at least ten reported card
check organizing successes. There were 193 organizing events for
which we were not able to identify the industry. These 193 events
included 131 elections and 35 card checks, and resulted in the
addition of 165,659 employees.