The issue of cross-subsidization in multiproduct pricing becomes increasingly important as regulators turn to partial deregulation as a means of promoting efficiency. When previously restricted markets are opened to new entrants, mandated cross-subsidization by the regulated firm encourages "cream-skimming" behavior and could result in entry by inefficient producers. Recent regulatory and technological changes in telecommunications suggest that the local telecommunications market may present opportunities for profitable entry in the not-too-distant future. Network bypass activity among businesses combined with the generally higher rates charged to business customers relative to residential customers suggests that business service revenues may be being used to subsidize residential access to the network. Prior to allowing wide-scale entry into local telecommunications, it is important to understand whether or not there is cross-subsidization in local phone service rates.
Currently available tests of cross-subsidy, originally suggested by Faulhaber (1975), are inadequate for this task. These tests require observations on stand-alone costs of firms or plants that produce each service in isolation. Typically, such observations do not exist. This paper presents two new empirical tests of cross-subsidy that do not require observations on the cost of producing each service by itself. Depending on technological conditions, the revised stand-alone condition and the revised incremental cost condition are necessary and/or sufficient for Faulhaber's original cross-subsidy conditions. Adjusted versions of these tests are developed for firms that receive subsidies from outside sources.
The results of applying the new cross-subsidy test conditions to local telephone revenue and cost data supplied by New England Telephone show that business service revenues subsidize residential service provision at over 70% of the central offices in the sample. Nearly 87% of the offices in the lower six rate group categories exhibit a business-to-residential cross-subsidy. The sufficient conditions for a residential-to-business cross-subsidy fail at all of the offices tested.
