On the Economic Management of Large Organizations: A Laboratory Study

Abstract

A LARGE fraction of most nations' resources is employed by very large organizations. Moreover, the management of these government agencies and giant private corporations has often been decentralized to such a degree that the numerous component subunits have incompatible objectives. These modern tendencies raise special problems of interest to managers, economists, and other social scientists. The magnitude of the resources used by some of these enterprises-and especially by the military departments of the federal government-is so large that the economist can hardly ignore them; the Air Force budget alone is approximately $17 billion a year, which makes it comparable to the entire automobileproducing industry.' A natural question is whether the economic efficiency of these very large organizations can be increased. Economic management of very large enterprises inevitably encounters problems of intra-firm organization. Of special import are information flows, lines of authority, and decision points within an organization. In very large organizations there will be many subunits that often have conflicting interests, may compete for resources, and prefer autonomy. Even within a single military service, subsidiary Commands can thwart, disregard, and occasionally disobey the directives of higher authority. Such tendencies to internal anarchy mean that it is not enough for management to devise optimum economical policies. These policies must also be implemented, and this requires the development of detailed procedures. More important, taking organizational conflicts into account, these procedures must be such that employees, in their daily work, will be able and willing to follow them. In many respects this is still an unexplored area. However, a number of scientists have recently been devising preferred economical policies for the procurement and distribution of aircraft spare parts by the Air Force. In addition, in an attempt to see whether they are practical, these policies and their attendant procedures were enacted for several months in a laboratory. This experiment simulated, with people and computers, a very considerable part of the Air Force supply system. The present paper seeks to describe the economic problems of supporting aircraft with spares, the policies developed, and the laboratory simulation that assessed them. This subject has some interest in its own right. But another purpose is to present a case study in the use of laboratory representation as a I Some economists are likely to despair of analyzing the economic efficiency of government agencies. It is true that, not knowing the worth of extra activity, the desirable scale of remains uncertain. However, tractable suboptimizations remain, such as producing a given output with minimum cost.