Can a Vertically Integrated Provider Use QoS to Unreasonably Advantage Itself Over OTT Content Providers?
17 Pages Posted: 30 Mar 2015 Last revised: 15 Dec 2015
Date Written: September 25, 2015
Most broadband providers are moving their circuit-switched video to video-over-IP and are multiplexing this traffic with broadband Internet traffic. These vertically integrated video services are thus competing with over-the-top (OTT) video services for customers and for network capacity.
Broadband providers are deploying quality-of-service (QoS) technologies to improve the performance of video-over-IP by reserving network resources for or prioritizing their video traffic. Broadband providers may sell QoS to OTT video service providers or directly to end users. Alternatively, vertically integrated broadband and video service providers may refuse to provide access to QoS to competing video service providers.
A key net neutrality question is whether a vertically integrated provider can unreasonably advantage itself over competing content providers by selling QoS at unreasonably high prices or by refusing to provide QoS to competing content providers. There is a substantial academic literature comparing various types of neutrality. Most, however, focus on an absolute prohibition of prioritization. In contrast, we focus here on the effects of a prohibition of third party paid prioritization on competition between a vertically integrated provider and an OTT provider.
We develop a mathematical model, grounded jointly in network architecture and economic theory, of competition between one vertically integrated provider and one OTT provider. The two offer horizontally differentiated services that differ by the amount and type of content, based on a Hotelling model. End users are similarly differentiated by their preference for the type of content. End users decide which service to subscribe to (if any) so as to maximize surplus.
The broadband provider decides whether to deploy QoS, which incurs an incremental network cost per user. QoS, if used, increases user utility proportionally. We consider two types of markets: a market in which the broadband provider charges the OTT content provider for QoS, and an alternative market in which the broadband provider charges end users directly for QoS. The vertically integrated provider sets its video service price and the QoS price to maximize profit, defined as the sum of video service revenue and QoS revenue minus the corresponding incremental costs. The OTT provider sets its video service price to maximize profit, define as video service revenue minus the corresponding incremental content cost and the QoS cost.
We first analyze the duopoly competition game under a fixed QoS price. The price, market share, and profit of the ISP and of the OTT provider at the Nash equilibrium are derived. We then consider the broadband provider’s decision to deploy QoS, and when it does, how it will set the QoS price. In both types of markets, we analytically determine when the broadband provider will sell QoS and when the OTT provider or users will purchase QoS.
We find that the price for QoS when sold to the OTT provider may be higher than when sold directly to users. If users are relatively homogeneous in their preferences, then the broadband provider may charge a higher price for QoS than is social welfare maximizing. If the QoS price is set to maximize social welfare, then social welfare may be higher if QoS is sold to users than if it is sold to content providers.
Finally, we present numerical results based on current Internet statistics. In addition to verifying the analytical results, we illustrate when the broadband provider’s decision does not maximize social welfare; the effect of QoS price on content prices; the variation of each content provider’s market share with QoS price; and the variation of QoS price, content prices, and market shares with the benefit of QoS and the amount of content differentiation.
Keywords: competition, market structure, end-user welfare, mathematical modeling, net neutrality
Suggested Citation: Suggested Citation