Journal ofOccupational Psychology, 1986,59,217-230.

Printed in Great Britain 1986 The British Psychological Society

Union growth and decline: The impact of employer and union tactics
JOHN J. LAWLER* Institute of Labor and Industrial Relations, University of Illinois at Urbana-Champaign

The past decade has generally not been a time of prosperity for organized labour, especially in the United States. It is now estimated that only about 19 per cent ofthe American workforce is unionized (Adams, 1985), compared with an all-time high of approximately 30 per cent. It is widely believed that this results from changes in organizing strategies used by both unions and, in particular, employers. The resistance of American employers to unionization has clearly increased over the past quarter of a century, spawning a multimillion dollar industry (Bureau of National Affairs, 1985). At the forefront of the ' union-free environment' movement is a formidable array of management consultants and attorneys, dubbed' union busters' by a disdainful labour movement. This paper examines the nature of union organizing and employer counterorganizing efforts and the role of the new breed of union resistance specialists. Although written from a North American perspective, the activities and research findings discussed here should be of relevance in Europe and elsewhere. Indeed, many American multinationals have implemented union resistance programmes in foreign subsidiaries and aspects of these programmes have been adopted by some non-American firms.

THE INSTITUTIONAL SETTING

Unlike European countries, both Canada and the US have labour relations laws which include a mechanism by which a union may be certified as the exclusive bargaining agent for defined groups of workers (the bargaining unit). In the US, the certification of a bargaining agent in the private sector occurs under the provisions ofthe National Labor Relations Act (NLRA). This is normally accomplished through an election conducted by the National Labor Relations Board (NLRB), the federal agency which enforces the NLRA, after a petition for an election is filed (usually by the union). While both the employer and union may participate in the campaign that precedes the election, the NLRB has established standards of conduct to regulate campaign activity. These standards are intended to guarantee that elections take place under ' laboratory conditions' to assure the free and informed choice ofempioyees with regard to union representation. Unions are normally recognized at firm or subfirm level, with ample opportunity for the employer to challenge the right of a union to represent its employees. Even after a union is certified, it must continue to be supported by a majority of those in the designated unit or it may be decertified (again, by means of an election). In contrast, the absence of the certification procedure in Europe, coupled with the prevalence of multi-employer bargaining, means there is not the same focal point for individual employer opposition to unionism Requests for reprints should be addressed to John J. Lawler, Institute of Labor and Industrial Relations, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA. 217

218

JOHN J. LAWLER
Attorney/consultant involvement Employer campaign strategy
Contextual control tactics

Contextual influences; Labour market Product market Legal system Political system Demographics

Monitoring and buffering tactics

II
JCampaign outcome I

Employes sentiments and preferences

Monitoring and buffering tactics

Contextuai control tactics

Union campaign strategy

AFL-CIO/consultant involvement

Fig. 1. The organizing process.

as in North America, an important factor in explaining the relatively greater prevalence and effectiveness of union avoidance activities on the North American side of the Atlantic (Kassalow, 1984).
CONCEPTUAL FRAMEWORK

American institutional arrangements generate intense union-management competition over union representation rights. Formal campaigns occur in which unions and employers evolve strategies: general strategies are implemented through specific tactics intended to affect campaign outcomes. This paper is concerned primarily with describing and assessing the impact of such tactics. A model of the organizing campaign process is proposed (Fig. 1) which identifies the basic elements of that process and suggests possible causal linkages among these elements. It also incorporates a typology of employer and union campaign tactics. There are three major phases in the natural history of a union-management relationship in which union organizing and/or employer counter-organizing activities are dominant: (i) pre-unionization phase: no union present, with little or no active effort to organize employees; (ii) recognition phase: internal employee committee and/or external union presence with active organizing effort underway; (iii) de-unionization phase: efforts (either by employer or employee group) to return to non-union status. Campaigns that may occur in each phase are associated with particular outcomes. The pre-unionization phase results either in continued non-union status or the initiation of recognition phase activity. The recognition phase results either in a union's certification as an exclusive bargaining agent or defeat for the union. Finally, the de-unionization phase results either in continued union recognition or a return to non-union status. Campaign outcomes are postulated to be directly determined by employee sentiments and preferences, employer campaign strategies and union campaign strategies. Employee sentiments and preferences are conditioned by contextual influences and by employer and union campaign strategies. The strategies of employers and unions will, in turn, reflect

UNION GROWTH AND DECLINE

219

union and management perceptions of employee preferences and sentiments. Contextual influences also shape campaign strategies (for example, legal and political forces may limit the range of permissible activities, or market conditions may affect the perceived consequences of unionization for the employer in question). Although employers may develop campaign strategies in-house, they often use attorneys and consultants. By attorneys, we mean labour law firms specializing in union avoidance services to management, while consultants usually rely on behavioural techniques. Many union avoidance specialists provide both services. Some consulting firms in the US specialize in providing services to unions (Bureau of National Affairs, 1985), but their presence is much more limited. The AFL-CIO Organizing Department also provides assistance to affiliated unions, therefore acting in a consultant-like fashion. Union and employer campaign strategies are likely to be mutually dependent, in that each side develops programmes to counter those of the other. Campaign strategies generate specific tactics, and the most commonly used tactics can be grouped in five categories (Table 1). Tactical categories are described here in general terms, with more detailed examples of tactics in practice provided below. Employers and unions may attempt to influence employee sentiments and preferences by means of persuasion, coercion, and/or manipulation. A second category of tactics is to create buffers between employees and contextual infiuences. On the employer side, buffering is exemplified by the establishment of rules that limit the access of union activists or organizers to employees during working time. Thirdly, unions and employers may endeavour to control the contexts affecting employee sentiments and preferences. While infiuence tactics may be directed at altering perceptions of contextual conditions, control tactics create real changes in contextual conditions. Wages and working conditions may be improved or a plant may be located in an area where unions are weak. Fourthly, monitoring tactics are closely linked to infiuence and buffering tactics, and often involve many of the same instruments. However, monitoring tries to discern employee attitudes and contextual conditions related to organizing activity. Employers may use supervisors to obtain information on actual or potential organizing efforts and supervisor training often focuses on the means of discerning such efforts in their early stages. Union organizers are similarly concerned with identifying sources of support among employees and anticipating possible employer actions. Finally, direct action tactics are not intended to alter or control employee sentiments; rather, these are efforts to change the proximate environment within which support for unionization is (or is about to be) expressed. Direct action tactics include the use of NLRB rules and procedures to delay certification elections or nullify the results ofan election. Our model posits employee sentiments and preferences as central to the determination of campaign outcomes. However, the cognitive processes that determine employee motivational states are not specified. Most psychological studies of union attitudes and behaviours (e.g. union membership, voting in representation elections) assume rational information processing and decision making by employees, and expectancy theory concepts are often used. Brief & Rude (1981) adopt the Fishbein & Ajzen (1975) formulation directly in their voting behaviour model, while Zalesny (1985) proposes variants of both the Fishbein & Ajzen and Triandis (1976) models. Brett (1980), Kochan (1980) and Youngblood et al. (1984) all incorporate elements of expectancy theory into their models. Empirical work related to these theoretical formulations generally supports the principal hypotheses. Consistent with expectancy theory, these models hypothesize that employee union attitudes and behaviours depend on job-related deprivation and the expected impact of unionization on working conditions. However, most micro-level models do not consider the role of the employer and union campaign tactics in the attitude-formation process. As we shall see, there is considerable controversy about how contextual infiuences and campaign tactics affect employee belief systems, with some researchers arguing that employees are largely inured to campaign infiuences.

220

JOHN J.

u CD
CD

CJ CD

itte

(D

_ >

c .2
C O O Q.

E E aniz ctivi
o u a c

c S 3 C . O

aniz ctivi'

. E

c o

.EIIIE"

o o> c

E E o

u
s o

c
3 03 O I- *r O) 03 -.a^ C o D eg b03 Q.

E :a I
c := N T3
. 2 fc C O D

B
o

CD

ternal rganiz

iternal rganiz

Ero-gc o _ ^
<0 = ' 5 2 c O C D O ( D COCD.2Ct;il>3 C p - O ; = C0^T3 8 D

B o

CD 03

o o Si S

I "
.O3

c
CD

a> C (I) D

15

2-2.9

tu

5 c E c 2 BE

a. CD </)U

2 o "= o-2 2

- (s ro ^\aS

I c
c c

.1

I
. 1
.a o

I I
c

CO CD 03 Q

2 .o
C

O 0)

CO *-

I
'" S C i" S- 0 3
03 C sy O o "O O O3 3 a .p .= ra IS" O p t Q UJ 0. D 03 C 2 D

> 3 _

o a>

l
0. 0. Q. .

CD O

EE
o > <i>

C "D D

c c

if

II
Q- c .c "0 aC O
C -D O

o
O) C D

c o u
c

CO (D

s
O)

c o

c o

a5

UNION GROWTH AND DECLINE EMPLOYER AND UNION TACTICS IN PRACTICE

221

We have identified three phases of union growth and decline at the organizational level. We shall now consider union and employer tactics in the pre-unionization and recognition phases. Research findings on the use and eflfectiveness of various tactics are reviewed, with special emphasis on psychological and behavioural work in the area. The pre-unionization phase Union tactics. Although most organizing efforts in the US seem to be internally generated (i.e, a group of workers approaches a union and asks for assistance in establishing a bargaining unit), there are situations in which, for ideological or pragmatic reasons, unions will initiate an organizing effort. Choosing a promising organizing site involves the use of various monitoring and contextual control tactics (Gagala, 1983, pp. 95-144). The organizer may spend time in cafes and bars in the proximity of a likely organizing site to discern signs of discontent and employee receptiveness to unionization. The costs and benefits for the union of organizing the site must also be assessed. Once it seems desirable to initiate an organizing effort, the organizer may make discrete contacts with employees far from the workplace. He/she may also obtain employment in the firm for purposes of internal agitation and monitoring. As union support may be scant and emphemeral in the early stages of the organizing effort, influence tactics must be low key lest the organizer's presence be detected by management. Thus, highly visible instruments of communication (e.g. posters, leaflets, rallies and group meetings) are likely to be avoided in favour of individual contacts. The organizer probably first endeavours to establish an internal organizing committee, normally composed of militant and influential employees. The committee and organizer work jointly to persuade other employees to support unionization. Unfortunately, there has been little research recently on union campaign tactics in the pre-unionization stage. Gagala's (1983) work is descriptive and highly normative (as it is intended as an organizer's handbook). Earlier case studies (Karsh et ai, 1953; Strauss, 1953) provided richly detailed descriptions ofthe organizing process during both its initial and more advanced stages. However, unions now appear much less likely to rely on such traditional approaches to organizing, which are often time-consuming and costly. Unions have substantially reduced expenditures on organizing activity since the 1950s (Voos, 1983), which certainly must have affected the manner in which organizers approach possible organizing sites. The dearth of contemporary research on union tactics during the pre-unionization phase leaves these issues unresolved. Employer tactics. Labour relations consultants, who tend to rely heavily (though not exclusively) on behavioural techniques, are very visible in the pre-unionization phase. A substantial number of books and articles have been published in the US on ' preventive labour relations' (e.g. Hughes, 1976; Myers, 1976; DiMaria, 1980; Kilgour, 1981) and most of this material has been written by professional labour relations consultants, who also frequently conduct seminars on preventive labour relations. However, the major instrument by which consultants provide information on preventive labour relations activities is direct involvement in the design of human resource management systems in client firms. As with union tactics, there is little hard research on employer tactics in the preunionization stage. What we know about preventive labour relations systems comes largely from consultant and union publications and press accounts of programmes in specific firms, but a few studies cast some light on the nature and impact of these programmes. Foulkes (1981) reported the results of intensive case studies in 26 large firms which operate exclusively or predominantly on a non-union basis. Most of these are on Fortune magazine's list of the 500 largest American firms. Foulkes' model of the human resource management process in the non-union firm, based on hisfieldwork, is highly qualitative and lacks statistical evidence in its support. Yet, as an exploratory exercise, his framework

222

JOHN J. LAWLER

provides several useful insights. He maintains that a basic element in effective preventive labour relations programmes is the articulation of a corporate philosophy that stresses the importance and welfare ofthe individual employee in the overall objectives ofthe firm. This may include an explicit commitment to operate only on a non-union basis (though Foulkes says this is not an essential ingredient). The management-oriented literature also stresses this point, often suggesting that firms should explicitly communicate their intention to remain non-union to current and potential employees. More generally, this approach can be seen as part of organizational culture buildinga major theme in the contemporary US organizational behaviour and human resource management literature. The building of a corporate culture serves to foster employee commitment and organizational indentification is argued to be an important part ofthe human resource management function (Lawrence, 1985; Mills &Balbaky, 1985). Foulkes identified substantive policies designed to implement the organizational philosophy and promote a culture unfavourable to unionization. Contextual control tactics underlie several of these policies. Successful non-union firms seek to manage environmental conditions by situating plants in areas where unions are especially weak (e.g. the southern and south-western regions ofthe country, as well as rural areas). The choice of production technology, plant size, and even of product, may involve considerations ofthe potential for unionization. Relatively small plants with a concentration of white-collar and professional employees are less likely to be unionized and seem to be characteristic of non-union firms. A key element seems to be the establishment of elaborate personnel programmes that serve many of the functions of a union, albeit under management control. Well-structured internal labour markets and attractive compensation systems are integral aspects of such programmes. Monitoring and buffering are accomplished through the careful selection and development of managers. Maintaining a high manager-employee ratio allows management to keep a watch on what is happening on the shop fioor. Many firms develop elaborate rules that restrict the access of union organizers to employees and thereby prohibit prounion employees from soliciting support. Managers, particularly first-line supervisors, are trained in the enforcement of these rules (so as to avoid violating provisions on the NLRA) and in how to recognize embryonic organizing efforts. Finally, influence tactics are manifested in various ways. Employee orientation programmes lay the foundation for culturebuilding efforts. Periodic attitude surveys, managerial counselling of employees, periodic meetings between employees and high-level managers, employee grievance plans and other similar communications devices all serve as influence tactics. Since communication channels provide feedback to management, they also act as monitoring devices. One study, Kochan (1980, pp. 183-191), focusing on general styles of employer resistance, noted differences between direct union suppression and indirect union substitution. The latter involves techniques similar to those described by Foulkes, while the former involves aggressive and confrontational tactics not uncharacteristic of the historic approach of American management (e.g. intimidation and coercion of employees, discharge of union supporters, surveillance of employees, use of security guards and/or local police to discourage picketing or leafieting). Of course, employers may well use a carrot-and-stick approach by combining elements ofboth stategies. Kochan was primarily concerned with ascertaining the determinants of such activity, and noted an association between organizational characteristics and resistance styles. Firms operating in high profit and growing industries which employ relatively skilled workers tended to use union substitution strategies (e.g. Kodak, IBM, Dupont), while firms operating in competitive, declining and/or low profit industries which employ less skilled workers were more apt to use union suppression strategies (e.g. the American textile industry). This difTerence is partly due to the generally higher costs of implementing a union substitution programme, but companies in the high growth sectors (especially high technology) seem generally more willing to experiment with less authoritarian styles of management.

UNION GROWTH AND DECLINE

223

Kochan also reported the results from a survey of several hundred firms conducted by the Conference Board (a management-oriented research organization). Top level managers in these firms were asked to identify organizational priorities in labour-management relations, and the level of importance attached to avoiding unionization was correlated with several organizational characteristics. A strong negative correlation was found between the importance of avoiding unionization and the proportion of current employees organized (r= 0 43). However, most other organizational characteristics studied (e.g. wage level of employees, profit rates, industrial concentration ratios, relative labour costs) were found to be either weakly or insignificantly correlated with the priority assigned to avoiding unionization. Kochan therefore concluded that the major determinant of a firm's propensity to resist unionization was its current level of unionization: highly unionized firms were less likely to oppose new organization while weakly organized firms were more likely to do so. Yet there is a certain circularity to this argument, for while an association may exist between these two variables we have no way of knowing the causal order involved. The recognition phase Although preventive steps taken during the pre-unionization phase are apparently widespread, much greater attention has been focused on union avoidance techniques used by employers during election campaigns. This reflects a concern that employer campaigns, often developed and orchestrated by consultants and/or attorneys, undermine the intent (if not the letter) of American labour relations law by compromising workers' free choice. A number of fairly sophisticated studies of the nature and impact of employer tactics have been published, several of which deal with their psychological impact, and there has also been some research on union tactics during the recognition phase. Typical employer and union tactics are described first below, followed by a consideration of research findings. Union tactics. Unlike the pre-unionization phase, the recognition phase involves active and aggressive union intervention in a firm. The union must first seek a ' showing of interest' among employees by obtaining signed' authorization' cards in which employees designate the union as their bargaining representative. If the employer will not voluntarily recognize the union based on card signatures, then an election must take place. Union tactics during card drives and election campaigns are wide-ranging, encompassing all ofthe general types identifled above. There appear to be two general strategies shaping union conduct in the recognition phase. Traditionally, unions have followed a unit-focused approach in organizing new members: unit-focused campaigns are designed to obtain support for unionization among employees in a potential bargaining unit and to secure union victory in a representation election. By contrast, at least some unions now seem to rely increasingly on environment-focused campaigns, which are intended to put direct pressure on firms by affecting their public image and obtain broader social support for unionization. Virtually all efforts to obtain recognition involve some type of unit-focused campaign, whilst environment-focused campaigns usually come in the wake of a weak or unsuccessful unit-focused campaign. Initially, the unit-focused campaign is a continuation of the process begun in the pre-unionization phase. The union may wish to maintain secrecy at the onset of the cardsigning drive in order to catch management off guard with a demand for recognition. Hence, influence tactics are apt to remain low key with the organizer and committee members involved in considerable one-to-one persuasion efforts. Once sufficient support is obtained to request recognition or an election, secrecy is no longer imperative. Meetings and social gatherings will then often be used as a means of solidifying union support. While the internal committee is central to most successful campaigns, the importance of the organizer and his/her support staff should not be underemphasized, especially during the

224

JOHN J. LAWLER

latter part of the campaign. The impact of various organizing styles and the nature of the interaction that takes place between the employee committee and the organizer are discussed by Karsh (1982). Union organizers and employee committee members may use monitoring and buifering tactics in an attempt to anticipate and deflect employer campaign activities. Indeed, predicting employer moves and taking defensive measures is important to a successful union campaign. However, unions do not have the same ability to construct elaborate monitoring and buffering devices as do employers (e.g. rules restricting organizing activity, supervisory training). Contextual control is also more diflicult for the union to achieve after the organizing campaign is underway, though context may be controlled by the manner in which the bargaining unit is configured. Disputes over which employees are to be included in the unit are resolved by the NLRB. Consequently, the union will seek to include in the bargaining unit individuals with a high probability of voting in favour ofunion recognition. Although the NLRB is obliged to apply certain standards in determining bargaining units (i.e. employees within bargaining units should share a significant' community of interests'), it has considerable leeway over the interpretation of these standards. Thus, union (as well as employer) representatives have the opportunity and the incentive to argue for the inclusion or exclusion of certain employees. Direct action tactics are at the core of the environment-focused campaign (Craft & Extejt, 1983). Most involve some effort to influence third parties to apply pressure on the employer to take a more reasonable stance with respect to the union. Some direct action tactics are not especially new, and consumer boycotts are a venerable tradition in American labour relations. Yet massive national product boycotts undertaken by several unions in recent years have involved much greater use of the media and have achieved higher levels of public awareness than earlier boycotts (which often involved only union members). Prominent examples of the more recent approach include boycotts of non-union lettuce and grapes, Gallo wine, Coors beer and Farah slacks. The lettuce and grape boycotts, undertaken by the largely Hispanic United Farm Workers and its charismatic leader Cesar Chavez, were probably the most successful of these. Closely followed by the press and given considerable aid by the AFL-CIO, these boycotts enjoyed wide popular support which was undoubtedly enhanced by public concern over discrimination and the plight of disadvantaged workers. Unions have also introduced a number of innovative direct action tactics in environment-focused campaigns. The' corporate campaign' involves a coordinated, multifaceted assault on recalcitrant employers. A public relations effort is mounted to generate awareness of the employer's intransigence and a consumer boycott is often encouraged as part of the public relations campaign. The union may acquire shares in the company and attempt to use annual shareholders' meetings to attack the company's anti-union policies, thus embarrassing management. Generating financial pressure on firms that have extensive dealings with the target company appears to be a very effective tactic. For example, in its ultimately successful effort to gain recognition in J. P. Stevens, a major textile manufacturer, the Amalgamated Clothing and Textile Workers' Union (ACTWU) threatened large union withdrawals from a major bank which had a Stevens official on its board of directors (Craft & Extejt, 1983). More recently, the Food and Allied Service Trades Department (FAST) of the AFL-CIO, a consortium of unions which organize in the rapidly growing service sector, have adopted a similar organizing technique, termed the ' comprehensive campaign'. This principally involves an extensive financial analysis of a target firm, with the intention of identifying activities or relationships that are potentially embarrassing to the firm. As the Reagan Administration appointees to the NLRB are generally seen as having weakened the position of unions in Board proceedings, FAST unions try to avoid representation elections, going instead for direct recognition by the employer, and using financial information obtained in the initial phases of the campaign.

UNION GROWTH AND DECLINE

225

Employer tactics. Union avoidance campaigns during the recognition phase differ in many ways from preventive labour relations programmes implemented in the pre-unionization phase. The union avoidance campaign is intended as a stop-gap measure undertaken when unionization is imminent. As the employer is often caught off guard by an organizing campaign, it is not surprising that the services of consultants and/or attorneys specializing in union avoidance activity are so frequently sought. The AFL-CIO estimates that union avoidance consultants and attorneys are used in roughly 80 per cent of all representation elections (Bureau of National Affairs, 1985). Lawler & West (1985) found that consultants were used in about 20 per cent of all NLRB election campaigns, although attorneys were present in most campaigns. The apparently low propensity of employers to use consultants perhaps reflects the fact that attorneys often provide many of the same services as consultants; and some of the larger firms employ industrial psychologists and others with behavioural training. LTnion avoidance campaigns typically mix influence and direct action tactics against a background of buffering, monitoring and contextual control tactics. Direct action tactics include efforts to affect the composition of the bargaining unit in a favourable manner, challenges to the validity of authorization cards submitted in support of an election, procedural challenges to the union's request for recognition (e.g. the employer might claim to be outside the jurisdiction ofthe NLRB for several different reasons), and efforts to delay the election as long as possible. Direct action tactics normally require an attorney's involvement as they are usually effected through the quasi-judicial hearing procedures of the NLRB (or relevant agency). Direct action tactics, intended to avoid the need to confront the union in a representation election, are not primarily psychological in character, but election delay is an exception. Union avoidance specialists maintain that delaying the election as long as possible is an extremely potent weapon (e.g. DiMaria, 1980, pp. 47-48; Kilgour, 1981, p. 259). Most organizing campaigns are spurred by one or more critical incidents which consolidate what may be largely latent support for unionization (e.g. the discharge of an employee, the impjementation ofan unpopular rule, wage reductions or layoffs). As this support is apt to dissipate quickly, the union has a limited ' window of opportunity' to exploit. While the NLRA and similar laws are intended to promote rapid elections, employers normally have little difficulty using Board procedures to delay elections for up to six or seven months, thus benefiting from the erosion of union support. The delay also gives employers the opportunity to take remedial action; and employee attrition and selective hiring and job assignment decisions may also contribute to an improved chance of victory, A number of influence tactics are commonly employed. These include ' captive audience' speeches to large groups of employees (so called since employee attendance may be mandatory), small group meetings, and one-to-one sessions (frequently with a supervisor or a consultant). Letters and other written communications are also used. A particularly effective device seems to be films depicting the negative aspects of unionization. While many of these techniques may be used in preventive labour relations programmes, the content of the messages transmitted is somewhat different in the union avoidance campaign. Rather than promoting organizational socialization and culture building, persuasive communications in union avoidance campaigns generally stress the costs of unionization (e.g, payment of dues, possible strikes), the potential difficulty of the union in changing wages and working conditions, the disruptive influences of ' outsiders' in employee-management relations, and management's intention to deal aggressively with any union at the bargaining table (Fulmer, 1981). Contextual control, monitoring and buffering tactics are much the same as in preventive programmes, though perhaps less prominent than influence and direct action tactics. In general, the effects of contextual control and buffering tactics probably take longer to develop, and a viable organizing drive suggests, in fact, a failure of tactics ofthis type. One exception to this rule is illegal job discrimination against union activists and supporters, and

226

JOHN J. LAWLER

there is clear evidence that conduct of this sort, often involving the discharge of an employee, is on the increase. Weiler (1983) reports that while the number of NLRB representation elections held annually have not even doubled since the mid-1950s, charges of discrimination filed with the board have increased sixfold. Moreover, the proportion of charges found to have merit by the NLRB has also substantially increased. Discriminatory action by employers represents a contextual control tactic insofar as sources of agitation are removed (though discrimination may also serve as an influence tactic by intimidating other employees). Unfortunately, the enforcement machinery of the NLRB is slowcases may take two years or more to resolveand the potential sanctions against the employer are quite limited. Research findings. While published studies of the impact of union and employer tactics during the pre-unionization phase are fairly limited, there has been considerable research in the US on tactics during the organizing phase. Most of this work has centred on employer tactics, though there has been some work on union tactics. The leading study of campaign effects during the recognition phase involved an unusual collaboration between two labour lawyers and an industrial psychologist (Getman et al., 1976). The authors used a panel study design to assess the impact of both employer and union tactics on the voting behaviour of several hundred individuals in 31 NLRB elections. Getman et al. found, contrary to the expectations of most observers, that the formal organizing campaign has little impact on voter behaviour. Employees were not found to be especially familiar with themes developed during the campaign, identifying, on average, less than 10 per cent of the issue raised by either the union or employer. Precampaign voting intentions predicted the employee's actual vote for 81 per cent of the individuals studied. None of the employer tactics studied, either legal (captive audience speeches and meetings, personal contact with employees, and written communications) or illegal (threats of reprisal or promises of gain related to union activity, employment discrimination) were related in a statistically significant manner to voter choices. However, some of the union tactics studied (speeches, written communications, personal contact) were found to have a marginal effect in influencing both undecided and initially pro-company employees to vote in favour ofthe union. These results are clearly inconsistent with the rational choice models of employee union attitudes and behaviours discussed above. The authors explained these findings in terms of behavioural theories of decision making which stress rationalization and selfjustification (Getman et al., 1976, pp. 141-144). Pre-campaign preferences and sentiments are seen to determine both the vote cast in the election and the employee's response to the campaign. Strong predispositions decrease the attentiveness of employees to the campaign (as indicated by the generally low familiarity of employees with campaign themes). Moreover, employees attend selectively to campaign messages, exposing themselves to messages that reinforce pre-campaign attitudes. Finally, employees may cognitively distort campaign messages so as to fit pre-existing beliefs. Jeanne Brett (nee Herman), the psychologist involved in the study, expanded on these themes elsewhere (Brett & Hammer, 1982), observing that even though' the decision to vote for or against unionization is limited to two alternatives... there is no evidence of rational decision making [by employees]'. Arguments drawn from social influence theory (Salancik & Pfeffer, 1978) bounded rationality theory (Simon, 1957), and the 'garbage can' theory ofchoice proposed by Cohen etal. (1972) were used by Getman et al. to model voter behaviour as decision making under uncertainty. While the Getman et al. study broke new ground, several authors have argued that it contains significant conceptual and methodological flaws. Flanagan (1976) expressed concern over the author's failure to specify and test a mulitivariate model (Getman et at. only report bivariate relationships). Adequate control variables were not incorporated into the

UNION GROWTH AND DECLINE

227

analysis, so that the authors report a series of non-significant findings without really explaining voting behaviour. Their reliance on the ordinary least squares method to analyse a dichotomous dependent variable (voting behaviour in a representation election) is also problematic. To make matters worse, a partial reanalysis of their data, in which some of these problems were remedied, seemed to contradict their results. Dickens (1983) used the Getman et al. data in a probit analysis of voter behaviour. Various demographic and organizational characteristics were included in the equation, along with measures of both legal and illegal employer election conduct. While pre-campaign voting intentions were found to be a very strong predictor of actual vote, Dickens also found that both legal and illegal employer campaign activities reduced the probability of an employee voting in favour of collective bargaining. While different specifications yielded somewhat different results, the aggregate average reduction in the probability of an employee voting in favour of collective bargaining was estimated as 15-17 per cent for illegal employer practices, 40-47 per cent for legal practices and 54-65 per cent for total employer campaign. In terms of election outcomes, this means that while unions won 36 per cent of the 31 elections in the Getman et al. study, the victory rate would have been about 66 per cent had employers refrained from campaigning in all elections. Conversely, an intense employer campaign in every case would have reduced the victory rate to around 5 per cent. In responding to Dickens' study, the authors questioned the reliability of his parameter estimates and simulation results (Goldberg et al., 1984). They also noted that Dickens failed to include measures of union tactics in his analysis. Another criticism of the Getman et al. study is that it inappropriately focuses on individual voting behaviour rather than election outcomes (Weiler, 1983). Several recent studies have analysed election outcomes in terms of employer tactics and, contrary to Getman et al., generally found strong effects. Cooke (1983) studied direct action tactics, reporting an approximate 1 per cent decrease in the probability of union victory for each month's delay in the holding of an election after a union petition was filed. Procedural manoeuvring, such as pre-election challenges to the composition of the election unit, was found to decrease union victory chances by around 14 per cent, suggesting this to be a particularly potent weapon. Cooke (1985) also demonstrated that discriminatory discharges of union activists significantly decreased the probability of union victory by 17-23 per cent. However, this effect was much stronger for discrimination which occured before the election campaign, a finding that is consistent with the results of Getman et al. Lawler & West (1985) assessed the effects of several employer campaign tactics using a multivariate model, and reported a strong aggregate effect for the set of activities studied (these included captive audience speeches, meetings, supervisor training, election delays, threats and inducements). Most of the tactics studied reduced the probability of unionization by between 2 and 5 per cent, though some had a particularly strong impact. The use of surveys to monitor employee sentiments decreased victory chances by about 7 per cent; providing the union with an incorrect list of employee names and addresses lowered victory chances by about 13 per cent (the NLRB requires that such a list be provided to the union after an election petition is filed; an error-ridden list would make union campaigning difficult). Murrmann & Porter (1983) found that a wide range of employer influence, buffering and direct action tactics decreased union victory chances in representation elections. In a multivariate analysis, they observed that employer efforts to affect the composition of the election unit had by far the greatest impact, followed by the intensity of the employer's influence efforts (through speeches, letters, etc.), restrictions on union campaigning during work hours, and election delays. In addition to studies of employer and union tactics, there has been some research on the effects of consultants and attorneys on the outcomes of organizing campaigns. In the model presented earlier in Fig. 1 consultant and attorney effects on such outcomes are posited to be indirect, acting through employer tactical choices. However, some researchers

228

JOHN J. LAWLER

have included measures of consultant/attorney participation as proxies for employer campaign tactics in the analysis of election outcomes. Moreover, the effectiveness of employer campaigns may be influenced by consultant and attorney participation. Lawler (^1984) reported that consultant participation in representation elections reduced the pro-union vote in representation elections by aboutt 28 per cent. Yet he also found evidence to suggest that the propensity of employers to resist organizing efforts (other than by use of a consultant) appeared to increase the pro-union vote. This finding is explained in terms of a ' backfire' effect, i.e. employer over-reaction to an organizing effort may actually create an environment favourable to unionization (if, for example, the employer becomes exceptionally abusive). In a subsequent study using a broader sample, Lawler & West (1985) observed a negative consultant impact on the probability of union victory, after controlling for a variety of employer tactics. They also reportedfindingsconsistent with the' backfire' effect: in particular, that consultant and attorney involvement in the same organizing campaign increased the probability ofunion victory. On the other hand, Murrmann & Porter (1983) found that consultant involvement, under certain circumstances, reduced the likelihood of union victory. Research on union campaign effects has been much more limited. Getman et al. (1976) did find that both undecided and initially anti-union employees were affected by union campaign tactics, although their analysis was limited to influence tactics. In the case of undecided voters, the correlation between familiarity with the union campaign and a prounion vote was 0- 34 (significant at the 0-01 level). Employees initially opposed to bargaining were somewhat less likely to be influenced by the union campaign: the correlation between union campaign familiarity and pro-union vote was 019 (also significant at the 001 level). Walker & Lawler (1986), in a panel study of individual voting behaviour of university professors in a representation election, found considerable stability in voting intentions over a protracted period, which is consistent with the Getman et al. (1976) study. Despite considerable organizational change and intense union organizing activity, over 70 per cent of the survey respondents voted in the election just as they had intended to five years previously. Union influence tactics (which varied across the several campuses of the university system) had no appreciable impact on the way the faculty voted in the election (employer campaign tactics could not be examined). Voos (1983) reported that the overall intensity of union organizing efforts, as measured by real expenditures on organizing, strongly impacts on union organizing success. She estimated organizing costs at between $450 and $1500 per employee successfully organized (at 1984 prices). Although these costs were high, they were substantially exceeded by the benefits in most settings (where benefits were measured in terms of increased member wages attributable to membership gain). Voos (1983) therefore concluded that the decline in union expenditures on organizing was a major factor in the explanation of falling union membership. Finally, Craft & Extejt (1983) reached mixed conclusions about a variety of union direct action tactics, noting that the effectiveness of these activities seemed to depend on contextual conditions, especially the willingness of unions to commit adequate resources to them.

DISCUSSION

This paper has explored the role of union and employer tactics in union growth and decline, and a general model of these tactics has been proposed. Examples of the ways in which various tactics are implemented at different phases ofunion growth and decline were presented and research relating to the impact of such tactics was discussed and analysed. While discussion has centred on tactics intended to impact either directly (influence tactics) or indirectly (buffering, monitoring and contextual control tactics) on the cognitive processes of employees, it has been necessary to consider other kinds of tactics (direct action tactics) which are designed to complement or enhance the impact of behavioural tactics.

UNION GROWTH AND DECLINE

229

Particular attention was paid to employer efforts to resist unionization and to the role of consultants and attorneys in shaping management campaigns. Intense management resistance to the establishment and/or continuation of unions is, at least among the industrialized Western nations, a largely American phenomenon. The US has the lowest rate of unionization of any country in this group, a rate which has fallen dramatically in the past quarter of a century. This drop has coincided with the emergence of a broad range of sophisticated union resistance tactics, many of which employ behavioural science techniques. Although American business has historically been quite aggressive in resisting unions, there was an accommodation between labour and management in the years following the Second World War. The declining dominance of American firms in their domestic markets, as a result of increased foreign competition, was a major contributor to the resurgence of employer militancy in the 1970s and 1980s. So too was the changing nature of management training: the omnipresent MBA, with extensive training in organizational behaviour, created a business environment more open to the use of psychological methods in union resistance and, more generally, in personnel management programmes. While American management has become more aggressive and sophisticated in responding to unions, the American labour movement displays some malaise. Dwindling resources and ideological intransigence have severely limited unions' responses to the new management challenge. While employers rely heavily on the expertise of external consultants, most union leaders are highly suspicious of those outside ofthe labour movement. Although it is quite evident that employers are doing more to resist unionization and unions are doing less to promote their own cause, these changes may not be the primary determinants ofunion decline in the US. Large numbers of American workers express little confidence in trade unions, and view them as archaic and/or corrupt. Broad changes in the nature of the American economy pose serious threats to the viability of the traditionally pluralist, adversarial approach of mainstream unions. Thus, changing employer tactics and the increased role of consultants and attorneys may be more of a symptom than a source of the weakened position of organized labour. While the research reviewed in this paper is less than conclusive on this point, there is considerable evidence to suggest that such tactics have made some impact on union growth and decline. However, contextual influences appear in almost all studies to exert strong effects on employee preferences and behaviours independently of tactics and/or the use of ' union busters'. Consequently, new limitations on employer conduct and the restriction of consultant and attorney activity are unlikely to be a panacea for union decline, as many unionists seem to believe.
REFERENCES ADAMS, L. (1985). Changing employment patterns of organized workers. Monthly Labor Review, February, 25-31. BUREAU OF NATIONAL AFFAIRS (1985). Labor Relations Consultants: Issues, Trends, and Controversies. Washington, DC: Bureau of National Affairs. BRETT, J. (1980). Behavioral research on unions and union management systems. In B. M. Straw & L. L. Cummings (eds). Research in Organizational Behavior. Greenwich: JAI Press. BRETT, J. & HAMMER, T . H . (1982). Organizational behavior and industrial relations. In T. Kochan, D. Mitchell & L. Dyer (eds). Industrial Relations Research in the 197O's: Review and Appraisal. Madison, WI: Industrial Relations Research Association. BRIEF, A. & RUDE, D . (1981). Voting in union certification elections: A conceptual analysis. Academy of

Management Journal, 6,26\-lin.

COHEN, M . D . , MARCH, J. G. & OLSEN, J. P. (1972). A garbage can model of organizational choice. Administration Science Quarterly, 17,1-25. CooKE, W. (1985). The rising toll of discrimination against union activists. Industrial Relations, liX, 421-442. CRAFT, J. & EXTEJT, M . (1983). New strategies in union organizing. Journal of Labor Research, 4,19-32. DICKENS, W . T . (1983). The effect of company campaigns on union certification elections: Law and reality once again. Industrial and Labor Relations Review, 36,560-575.

230

JOHN J. LAWLER

DIMARIA, A . (1980). How Management Wins Union Organizing Campaigns. New York: Executive Enterprises. FISHBEIN, M . & AJZEN, I. (1975). Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley. FLANAGAN, R . (1976). The behavioral foundations of union election regulation. Standford Law Review. 28, 1195-1205, FouLKES, F, (1981). Large, nonunion employers. In J. Stieber, R. McKensie & D. Q. Mills (eds), US Industrial Relations 1950-1980: A Critical Assessment. Madison, WI: Industrial Relations Research Association. FuLMER, W. (1981), Step by step through a union campaign. Harvard Business Review, July-August, 94-102. GAGALA, K . (1983). Union Organizing and Staying Organized. Reston, VA: Reston Publishing. GETMAN, J., GOLDBERG, S, & HERMAN, J, (1976), Union Representation Elections: Law and Reality. New York: Russell Sage Foundation. GOLDBERG, J,, GETMAN, S. & BRETT, J. (1984). The relationship between free choice and labor board doctrine: Differing empirical approaches. Northwestern University Law Review, 79,721-775, HUGHES, C, (1976), Making Unions Unnecessary. New York: Executive Enterprises Publications, KARSH, B. (1982), Diary ofa Strike, 2nd ed, Urbana, IL: University of Illinois Press. KARSH, B., SEIDMAN, J., & LILIENTHAL, D . (1953). The union organizer and his tactics: Acase study, American Journal of Sociology. 50,113-122, KASSALOW, E, M , (1984), The future of American unionism: A comparative perspective. The Annals. 473, 52-63. KiLGOUR, J. (1981). Preventive Labor Relations. New York: AMACOM. KOCHAN, T. A. (1980). Collective Bargaining and Industrial Relations. Homewood, IL: Richard D, Irwin. LAWLER, J. (1984). The influence of management consultants on the outcome of certification elections. Industrial and Labor Relations Review, 38, 38-51, LAWLER, J. & WEST, R , (1985). Impact of union-avoidance strategy in representation elections. Industrial Relations. 24,406-420. LAWRENCE, P. (1985). The history of human resource management in American industry. In R. Walton & P, Lawrence (eds), HRM: Trends and Challenges. Boston: Harvard Business School Press, MILLS, D , Q , & BALBAKY, M , (1985), Planning for morale and culture. In R, Walton & P. Lawrence (eds), HRM: Trends and Challenges. Boston: Harvard Business School Press, MURRMANN, K . & PORTER, A, (1983), Employer tactics and NLRB election outcomes: Some preliminary evidence. In Proceedings of the Thirty-Fifth Annual Meeting of the Industrial Relations Association. Madison, WI: Industrial Relations Research Association, MYERS, S, (1976), Managing without Unions. Reading, MA: Addison-Wesley, SALANCIK, J. & PFEFFER, J. (1978). A social information processing approach to job attitudes. Administration Science Quarterly, 23,224-253. SIMON, H . (1957). Models of Man. New York: Wiley, STRAUSS, G . (1953). Factors in the unionization ofa utilities company: A case study. Human Organization, 12, Fall, 17-25, TRIANDIS, H . C . (1976). Values, attitudes and interpersonal behavior. In Nebraska Symposium on Motivation. Lincoln, NB: University of Nebraska Press, Voos, P, (1983), Union organizing: Costs and benefits. Industrial and Labor Relations Review, 36,576-592, WALKER, J, & LAWLER, J. (1986). Union campaign activities and voter preferences. Journal of Labor

Research.!, 19-40.
WEILER, P. (1983), Promises to keep: Securing workers rights to self-organization under the NLRA. Harvard Law Review, 96,1769-1827. YouNGBLOOD, S., DENISE, A., MOLLESTON, J. & MoBLEY, W. (1984). The impact of work environment, instrumentality beliefs, perceived labor union image, and subjective norms on union voting behavior. Academy of Management Journal, 27,576-590. ZALESNY, M . (1985), Comparison of economic and non-economic factors in predicting faculty vote preference in a union representation election. Journal of Applied Psychology, 70,243-256,