Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
If we combine my result with those of Niskanen and Mackay/Weaver, we have three necessary conditions for interests to be mutual. First, review committees are composed of high demanders of their bureau's outputs; the committee seeks to maximize the interests of the group it represents, subject to majority approval of the full legislature. Second, the ideal point of the committee must not be ‘too extreme’ in terms of mix. Finally, the multiproduct bureau must be the least-cost institution for public supply of the services, in the sense that joint production is efficient, for the review committee interested in two or more outputs. If any of these does not hold, the interests of the bureau and its review committee diverge. The incentives for the bureau to expand across outputs without regard to efficiency means that in general we could expect the bureau to perceive the review committee as an adversary, because for the committee efficiency in production is an important consideration. What are the implications of this result? There must exist an optimal structure for bureaucratic production of services. Further, there exist substantial incentives for discovering and implementing this structure when legislators create bureaus. Diseconomies of scope imply separate bureaus are better; economies of scope the opposite. If committees optimally structure the bureaucracy in accordance with this result, they ensure that interests do not diverge. So we see that the initial design of bureaucratic organization may be more important than was previously recognized, and calls for closer examination. Copyright Martinus Nijhoff Publishers 1984