Consumer switching costs in the telecommunications industry

This dissertation examines whether consumer switching costs affect price and market share in the telecommunications industry. Consumer switching costs may be defined as the costs consumers incur when switching from one supplier of a good or service to another. There are at least four types of consumer switching costs: investment in physical equipment, informational, transactional and artificial. Some markets may exhibit only one type of consumer switching costs, while others may exhibit several types. The telecommunications industry exhibits several types of switching costs. In particular the Message Toll Service market exhibits informational and artificial switching costs, and the Cellular Telephone market exhibits physical, informational, transactional and artificial switching costs.;In order to test for switching costs in the Cellular Telephone industry, the theoretical model of Beggs and Klemperer (1992) is used. The model consists of a system of three equations: a market share equation for the high market share firm; a market share equation for the low market share firm; and a price difference equation. The data are available for 305 Standard Metropolitan Statistical Areas for 1989, 1991, and 1993. I estimated the system of equations by the method of Three Stage Least Squares with a Hal White correction for heteroskedasticity. Although switching costs are present in the Cellular Telephone industry, the results do not prove conclusively that they affect price and market share. The nonconclusive results obtained from this exercise may be the result of the unavailability of firm marginal cost data and quality issues in the Cellular Telephone industry.;For the Message Toll Service market, I modified the Beggs and Klemperer (1992) theoretical model using the consumer distribution theory of Perloff and Salop (1985) to include more than two firms. For the dissertation I derived a price equation and a market share equation and used these to test the Message Toll Service market. The results for the empirical estimation are inconclusive due to the small data samples. Although I was unable to obtain a large data set to test the theory, I proceeded in doing so in order to test the methodology of my model.