Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


An important question has pervaded the field of labor-management relations for over fifty years: Is political bias present in the decisions rendered by National Labor Relations Board members in unfair labor practice cases? This study develops and tests a model of Board member decisions in selected unfair labor practice cases over the period 1962-1983. By studying 2147 decisions made by seventeen individual members, a relationship was found among the decisions and several variables of political party affiliations, personal background factors of members, and conditions within the procedure by which allegations of unfair labor practice violations are processed for decision. Political congruency between Republican members originally appointed by Republican presidents was found to exert a pro-management bias on decisions. Procedural process variables found significant included employers as the party initiating the charges of ULP violations; support of the administrative law judge of the allegations irrespective of whether the filing party was a union or an employer, and dismissal by ALJs of charges filed by unions. Personal background factors found to be significant included the age of the member at the time of the decision, with older members more likely to decide favoring unions over employers than younger members, and the type of undergraduate degree possessed, with members having been awarded degrees in business/economics less likely to make decisions in favoring unions over employers than members with non-business/economics degrees. Results were analyzed using the logit technique. Logit is a non-linear probability model particularly suited to analyzing dichotomous variables (e.g., pro-union/pro-management decisions, Democratic/Republican member). Logit provides insights (where categorized data are involved) that are not available from more commonly used techniques (e.g., regression analysis and analysis of variance) which are based on assumptions (e.g., that the dependent variable is continuous or at least ordinal) and were therefore not appropriate to this study. Practical implications of the results of this study are described and thoughts are offered for future research concerning the roles of the regional director and the administrative law judges in decisions.