The Cost of Free Spectrum

TL;DR

This paper analyzes how adding unlicensed spectrum affects social welfare and consumer surplus, revealing that it can sometimes decrease overall welfare due to congestion and competition effects.

Contribution

It provides a theoretical model quantifying the impact of unlicensed spectrum on social welfare and consumer surplus, highlighting potential negative effects.

Findings

Adding unlicensed spectrum can decrease social welfare in certain conditions.

Consumer welfare may decline with unlicensed spectrum when user sensitivities vary.

Welfare impacts depend on the amount of unlicensed spectrum and user congestion sensitivities.

Abstract

There has been growing interest in increasing the amount of radio spectrum available for unlicensed broad-band wireless access. That includes "prime" spectrum at lower frequencies, which is also suitable for wide area coverage by licensed cellular providers. While additional unlicensed spectrum would allow for market expansion, it could influence competition among providers and increase congestion (interference) among consumers of wireless services. We study the value (social welfare and consumer surplus) obtained by adding unlicensed spectrum to an existing allocation of licensed spectrum among incumbent service providers. We assume a population of customers who choose a provider based on the minimum delivered price, given by the weighted sum of the price of the service and a congestion cost, which depends on the number of subscribers in a band. We consider models in which this weighting is uniform across the customer population and where the weighting is either high or low, reflecting different sensitivities to latency. For the models considered, we find that the social welfare depends on the amount of additional unlicensed spectrum, and can actually decrease over a significant range of unlicensed bandwidths. Furthermore, with nonuniform weighting, introducing unlicensed spectrum can also reduce consumer welfare.