Academic journal article Economic Inquiry

Product Substitutability and Competition in Long-Distance Telecommunications

Academic journal article Economic Inquiry

Product Substitutability and Competition in Long-Distance Telecommunications

Article excerpt


The growing availability of detailed firm-level data and advances in econometric techniques have led to the estimation of industry structural parameters in order to evaluate proposed policy changes (Hausman, Leonard, and Zona [1994]). In particular, estimates of own and cross-elasticities of demand are used to draw inferences about competition under various assumptions about producer behavior (e.g., Hausman [1994] and Stern [1994]). Generating estimates of structural parameters requires not only price and demand data but often requires firm-level factor price data to be used as instrumental variables. Even when firm-level factor price data are available, they may not fulfill the independence requirements of residual demand estimators because the competitors purchase from the same input market. Because the data requirements for the estimation of Marshallian demand elasticities are less severe, they may yield unbiased estimates. Residual demand elasticities can then be constructed under various assumptions about the extent of pricing coordination among producers.

This article carries out this process for the U.S. long-distance telecommunications industry. A unique and detailed data set that includes firm-level output, price, and factor price information is employed. The analysis focuses on the so-called Basket 1 services, which are ordinary interstate and international message toll service, operator assistance, and calling-card service. These services are generally regarded as the ones in which AT&T is likely to possess the most market power. Despite the detailed nature of the available data, some of the factor prices are shown to lead to biased estimates when used as instrumental variables. Results from specifications that omit these instruments indicate that AT&T faces a Marshallian demand elasticity of about -10. Residual demand elasticities are estimated for various assumptions regarding the degree of pricing coordination in the industry. For plausible assumptions regarding the degree of price coordination among firms, AT&T's residual demand elasticity is likely below-5, implying an industry dead-weight loss of less than 1.5% revenues.

Since the courts and the Federal Communications Commission (FCC) opened U.S. long-distance telecommunications to competitive entry, the number and size of entrants has grown to the point that AT&T, once the legal monopolist, now supplies less than half of the total market. The FCC's incremental approach to deregulating AT&T decreased regulatory oversight in specific product markets as competition became more vigorous and culminated in its 1995 decision to declare AT&T "nondominant," essentially deregulating AT&T. The results of this analysis tend to support this decision. Nevertheless, debate over the extent of competition continues, primarily to evaluate the claim that allowing the Bell Operating Companies (BOCs) into these markets will yield consumer benefits from still more vigorous competition. This analysis suggests that these benefits are likely to be small. Hence, as Brennan [1987] points out, the case for permitting BOC entry into long-distance should focus on whether they are likely to provide the service at lower cost or whether they have scope for discriminatory actions, rather than on whether they will make the long-distance market substantially more competitive.


The extent of competition in the U.S. long-distance industry and the effects of price regulation have been the subject of policy interest and the focus of a few empirical studies. MacAvoy's [1995] analyses lead him to conclude that the industry is increasingly collusive. His event study finds that a portfolio of MCI and Sprint stock realizes positive abnormal returns when an AT&T price increase is reported. While this indicates that the industry is imperfectly competitive, it could result from either unilateral or coordinated pricing behavior. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.