The significance of an adequate administrative machinery and its importance as a bureaucratic power base for planning and implementation have frequently been noted by scientists.' Three main purposes of such machinery can be distinguished. The first is the translation of policies as outlined by governmental authorities and Parliament into overall strategies (mediumor long-term plans), the formation of strategies into sectorial, content, or annual programs, and the further translation of these programs into specific projects (intersectorial, regional, or local, annual or longer periods, and so forth).2 The second purpose is the actual implementation of the plan, and the third includes the development and utilization of control, evaluation, and feedback processes.3 The specific course of action to be followed in accomplishing each of these purposes is clearly influenced by, and indeed often directly dependent on, administrative rationality as conceived by the administrators themselves. Administrative rationality can be defined as a multivariable process in which conceived costs and outputs are put together into the most satisfactory state within a certain time interval. The costs may be economic, social, ecological, or psychological, such as overcoming inner resistance to change of present procedures or personal adjustments. In the same way, outputs might include personal successes, prestige, achievement of goals, or smooth-running machinery. The following case study is intended to facilitate discussion and analysis of the impact of administrative rationale during a sensitive educational change: that is, during a comprehensive desegregation process. The consequence of such rationality may, but does not necessarily, overlap formal policy objectives. However, since the nature of the data throws light on comprehensive educational reform, they may also be of value in themselves.