Most of what academicians know about comes from large-scale case studies that tell us why large organizations made big mistakes: why, for example, the American military was unable to take advantage of intelligence information warning of an attack on Pearl Harbor; why missiles were not withdrawn from Turkey in accordance with a presidential order; or even why individual members of a bureaucratic organization were denied official recognition for their achievements.' The details of these events, their frequency, and their general significance-how organizations negotiate with each other over turf, how individuals protect their perquisites, which games bureaucrats play successfully and which are filled with risk, and just when substantive issues have to give way to procedural concerns-are of less interest to students of administration than are their consequences. But these familiar consequences are probably exaggerated as general descriptions of how bureaucrats behave. Readers of case studies of bureaucratic politics may be forgiven for a tendency to apply Murphy's Law to all the work of large organizations, since they have no basis for appraising their importance in the context of other problems of government. This article takes the opposite tack, for the moment ignoring consequences in order to examine in detail the kinds of activities which managers perform when they engage in bureaucratic politics. It offers no judgment as to their costs in terms of national goals.2 The activities on which this analysis is based were recorded during the summer of 1984, when nine countries in Southern Africa (Angola, Botswana, Lesotho, Malawi, Mozambique, Swaziland, Tanzania, Zambia, and Zimbabwe) cooperated as a group known as the Southern Africa Development Coordinating Conference or SADCC in authorizing a large-scale study of management in order to identify training needs and other means of improving performance. The managers who collaborated in the research included members not only of the public and private sectors but also of intermediate hybrids such as parastatal organizations, public enterprises, or statutory corporations.3 As one means of appraising administrative performance in the region, the study team gathered reports of management events that displayed either effective or ineffective behavior on the part of these administrators. Each such incident, of which there were more than 1,800, was later coded to identify training needs and performance characteristics that might lead to improved management.4 One such characteristic was bureaucratic politics.5
Read full abstract