The study of union certification elections is well advanced in the academic literature. There is, however, little research on determinants of union certification when more than one union is on the ballot. The work on multiunion elections that does exist is primarily descriptive and conducted at the macro level. This paper presents a micro-level study of a multiunion election. Three sets of variables derived from expectancy theory, the investments model, and the theory of reasoned action, were posited to influence employee choice in a multiunion election. Data were collected via a mail survey of 270 employees who recently participated in a multiunion election. The data were analyzed using multiple analysis of variance and discriminant analysis. The results provide strong evidence that the employee vote in a multiunion election is correlated with (a) an evaluation of the efficacy of a raiding union, (b) employee behavioural investments, and (c) the attitudes of others in the immediate work group. Implications for labour practitioners centre on increasing employee involvement, identifying the union with important others, and delivering desired results. Future researchers may wish to extend this framework to other settings.
The strength of the labour movement impacts directly on societal and firm level operations. At the macro level, a healthy, vibrant labour movement promotes both pluralistic democratic governance and a kinder, gentler society that is responsive to the needs of the less fortunate. At the micro level, unions shift the locus of control and authority from a cadre of managers to a partnership of labour and management, wedded by mutual dependence and charged with joint decision making. The power of organized labour, understandably, has generated considerable attention from scholars interested in workers' propensity to support union organizing efforts. Their inquiries are varied in both approaches and findings. Studies have utilized many explanatory variables and competing theoretical models, and have obtained estimates at both the macro and micro levels of analysis. We shift the focus to an area that has received little attention: a micro-level study of a multiunion election. We begin with a review of the literature. We then outline our theoretical framework and describe the research design and methodology. Finally, we present the results of the analysis and discuss their significance to conclude the paper.
While much has been learned about the willingness of workers to support unionization (Block & Premack, 1983; Fiorito, Gallagher, & Greer 1986; Heneman & Sandver, 1983; Wheeler & McClendon, 1991), little has been written on multiunion elections. What has been written is primarily descriptive in nature and conducted at the macro level. Studies from around the time of the AFL-CIO merger focused on the impact of union raiding (Krislov, 1954, 1959a, 1959b; Troy, 1960). In summary of this work, Chaison concluded that "most of the election data analyses dealing with raids have concerned themselves with the identities of the contestants, the general level of raiding over time and the success rates of individual unions or union types when raiding or being raided" (1973, p. 435).
Chaison extended this work by analyzing National Labor Relations Board (NLRB) multiunion election data from 1970 to 1971. He found that employees are unlikely to reject all unions. The no-union option prevailed in fewer than 5% of the elections, while the incumbent won in 38% and the raiding union won in 57%. The strong overall support for union representation may be the result of (a) strong bargaining-unit preference for unionization, which attracts extra-union intervention, or (b) raiding unions providing an alternative to the no-union choice when employees are dissatisfied with the incumbent. In a later article Chaison (1976) extended his analysis of multiunion elections to include nonraid multiunion elections over the period 1964-1973. …