Revisiting James Q. Wilson’s Bureaucracy : Appointee Politics and Outcome Observability

This study introduces a measurement approach to test a core precept in the study of bureaucratic politics and public organizations. First, we quantify James Q. Wilson's famous agency typology by the relative observability of agency outcomes. We code over 20,000 stated objectives from Performance and Accountability Reports (PAR) for United States federal agencies over the years 2000 to 2012 by whether the attainment of a given objective can be attributed directly to the programmatic activities of the agency (i.e., observable) or if its attainment is dependent upon other actors or other socioeconomic factors outside the agency's control (i.e., unobservable). We then test Wilson's previously untested proposition that “production” and “craft” organizations (i.e., those with more observable outcomes) will be better managed by career leadership, whereas “coping” and “procedural” organizations (i.e., those with more unobservable outcomes) are better managed by appointed leadership. We account for contemporary realities of appointee politics, including the president's strategic use of unilateral appointments in lower executive and upper management positions and the pervasiveness of appointee vacancies at the highest levels. Although scholars frequently rely on the Wilson typology in empirical and theoretical explorations of bureaucratic behavior, its dimensions have not been systematically measured at any level of government. We find, consistent with Wilson, that politicization diminishes performance in Craft and Production agencies. On the other hand, vacancies in presidential appointee positions are associated with lower performance only in those agencies that do not map to ideal types in Wilson's schema. In other words, the relative observability of an agency's goal agenda moderates the effectiveness of its appointed leadership—a contingency that has been insufficiently accommodated in previous work