CONSULTANTS, UNIONS, AND NLRB ELECTIONS (LOGIT MODELS)

This dissertation is about a significant and growing force in labor-management relations in the United States. This force is the outside consultant hired by management either to prevent unions from gaining a foothold in their company or, in the case where the union is already present, to oust the incumbent union. This dissertation focuses on several questions. First, what is the role played by consultants in National Labor Relations Board (NLRB) representational elections? Second, what is the impact of these consultants on the outcome of NLRB elections?; A profile of the characteristics of NLRB elections involving consultants is drawn from survey data provided by the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO). The following election characteristics exhibit a statistically significant relationship with consultant activity: geographical location, size of the unit, company is part of larger corporation, number of years at the election site. The extent of the pervasive consultant involvement in NLRB election campaigns is suggested by the findings of this study that over 70 percent of the election campaigns were run by outside consultants. Two specific campaign tactics were found often in consultant-led campaigns. Consultants were found to be involved in a significant number of elections where the union did not obtain the requested bargaining unit. In elections where the union did not obtain the desired unit, it lost a disproportionate number of those elections. Another trait of consultant-led campaigns is the use of front-line supervisors. Here again, the union lost a significant number of elections when the consultant used front-line supervisors in the campaign. In order to assess the impact of these consultants on the outcome of NLRB elections, a logit regression model was developed using data from the U.S. Department of Labor and the AFL-CIO. The results of the logit regression model indicate that consultants are responsible for approximately a twelve percentage point "swing" in votes for the employer. A simulation was performed to gauge the impact of consultants on NLRB elections held in 1979, 1980, and 1981. The results of the simulation suggest that if consultants had not been employed in any of the elections over the three-year period, union victories would have increased by 2,241. This would have meant an addition of almost a quarter of a million members.