ECONOMIES OF SCALE IN COMPUTER USE; INITIAL TEST AND IMPLICATION A FOR THE COMPUTER UTILITY

This study is concerned with the existence of economies of scale in the production of data processing and other computing services, and the possible regulatory and public policy implications of such economies. The rapid development of the technology of computation since the Second World War has raised many questions as to the supervision of public authorities of the use and progress of this technology. A study was initiated by the Federal Communications Commission in 1966 in an effort to consider that Commission''s role in the production and distribution of computing services where the use of communications facilities, supplied by regulated carriers, forms a integral part of the computing system. The present investigation is concerned with the production of computing services per se; the direction that public policy takes will be greatly dependent upon the nature of the production of computing services, and perhaps secondarily upon the interdependence between computer systems and the communications suppliers. The relative economies of the use of large computing systems have been known for some time, in terms of the relationship between some measure of the quantity of output of a machine and its cost. Indeed, it is demonstrated here that, when one considers, in addition to the cost of the computer hardware itself, the various categories of operating expenses associated with a computer installation, the relative advantages of large facilities become even more significant. Yet the evidence would seem to indicate that, despite these apparent efficiencies of large systems, the overwhelming majority of installed computers were generally fairly small operations. In an attempt to determine whether actual experience of users was that, all things considered, there were no true economies of large size, an analysis was made of data on nearly 10,000 computers installed at firms in manufacturing industries, using the survival technique, which uses market experience as a basis for studying levels of optimum plant size. The results of this analysis suggested that users did operate computers as if there were significant economies of scale in their use. None of the evidence, in fact, suggested that even the largest size system available today is the most efficient possible use of "plant" hence, the key implication for the formulation of regulatory policy toward the computer is that such policy should encourage, to the greatest possible extent, the shared use of large systems by those who require computing services. Those barriers that do exist which tend to mitigate such shared use should be reduced or eliminated. Public utility status would be indicated only if the costs associated with shared computer use - distribution , software development, system overhead and administration - are less than the potential direct savings resulting from use of large systems. This is at least as much a technological problem as it is regulatory; the future of the computer utility c