Past theories of representative bureaucracy have four weaknesses: they assume that traditional controls are ineffective without empirical evidence, rely on secondary variables, omit the effects of lifetime socialization, and do not consider the role of individual bureaus. Because of these weaknesses, a representative bureaucracy need not be a responsive bureaucracy. Although restricted by secondary analysis, this paper seeks to eliminate these failings and empirically demonstrate the unrepresentative nature of the United States federal bureaucracy. The representativeness of various grade classifications, special services, and bureaus is also measured; and the United States upper civil service is compared to that of five other nations. After an attempt to measure the values of bureaucrats, the future concerns of the theory of representative bureaucracy are outlined.
