The Effect of Prices on Fixed and Mobile Telephone Penetration: Using Price Subsidies as Natural Experiments

Abstract

This paper makes use of a natural experiment performed by a low-income price subsidy program, Lifeline Assistance, to estimate consumer substitution between fixed and mobile service. Variability across states and time in the Lifeline discount and the program’s eligibility requirements aids in identification. Using a large, nationwide household dataset that includes both billing details and demographic information, we measure mobile subscription differences between households receiving the subsidy and those that do not participate. Adopting a differences approach, we estimate access cross-price elasticities in the range 0.25 to 0.31, confirming substitution between the two access services and consistency with other published estimates. Simulations using the estimated mobile subscription models point to limited impact of changes in Lifeline participation on mobile penetration. We further rule out that the response to the price subsidy is an income effect by showing that household consumption of other similar products and services are not affected. We also find that high-income households have cross-price effect not significantly different from low-income households suggesting our results may apply to be broader population. Nevertheless, we find evidence that Lifeline participation is endogenous. To correct for potential bias, we exploit the re-sampling that exists in our dataset. A difference-in-differences analysis of the sub-panel shows little change in estimates of Lifeline’s effect on mobile subscription from our cross-sectional analysis. We do find asymmetry in responses to Lifeline participation depending on whether households added or lost Lifeline between re-samplings. These estimates have implications for reform of Universal Service policy as well as for the degree of consumer fixed-mobile substitution.