A Goal Programming/Constrained Regression Review of the Bell System Breakup

Abstract

The recently implemented court decision to break up Bell (=American Telephone & Telegraph Co.) to accord with U.S. anti-trust laws represents a highly significant policy decision which is proving to be influential in other countries as well as the U.S. The telecommunication industry is of such size and importance that even relatively small economies that might be lost with Bell's breakup as a “natural monopoly” could involve substantial welfare losses to consumers and producers. Studies commissioned by the U.S. Justice Department that approached this topic by econometric methods reported that the evidence failed to support the contention that Bell was a natural monopoly. Here a goal programming/constrained regression, as developed in the Management Science literature, uses the same functional form and the same data but nevertheless reverses the main findings of the econometric studies in every one of the 20 years covered. This kind of difference in results obtained by two different methods of analysis points up a need for drawing on persons from different disciplines who are capable of checking each other's methodologies when important policy decisions may be influenced by results that depend on the methodologies that these disciplines customarily use. Advantages of doing this are further illustrated by data deficiencies that escape detection by the econometric methods employed in the Justice Department-commissioned studies.