Unraveling the Micro-Economic Dynamics of Broadband and Content Provider Alliances

The services that run on telecommunications providers' infrastructure have evolved from essential to the day-to-day functioning to hyper critical to the minute-by-minute functioning of modern society. Telecommunications firms have however been increasingly exposed to competition and diminishing profitability. In order to face this challenge, it is essential to understand the micro-economic realities of the interaction between content providers and connectivity access providers. In this paper, we unravel options and discuss the insights revealed by micro-economic modeling of the interaction between a retailer of online content and a broadband provider. We show that heterogeneity among consumers leads to heterogeneous outcomes for the firms even where average trends are clear. One of the main results is that the absence of commercial cooperation between the content and the broadband provider delivers bad outcomes in almost all cases. The economic theory underlying the analysis is consumer willingness-to-pay and the resulting economic surplus (often, Marshallian surplus) comprising of the consumer surplus and the producer surplus. It is assumed that the content product increases the cost of delivery of the broadband product due to increased network use. The parameters that are variable are (a) the correlation between consumer valuations of the two products and (b) the additional cost of broadband provision for users of the content product. The paper presents the results of extensive simulation models for different distributions of consumer valuations and incremental additional cost for content product users. Our discrete modelling approach differs from the more common approach of assuming a continuum of consumers (i.e. not only infinitely many but uncountably infinitely many!) and solving for an optimal expected outcome. Our first step is to solve the problem for a finite number of customers and a single allocation valuations (drawn from a specified distribution). This allows one to see what might happen in concrete cases. We then do this many times and compute averages which approximates the expected outcome in the continuum case, of course. This approach allows us to identify more detail than an expected outcome does. We would like to remind the reader than the expected outcome from the roll of a die is 31 2 which is never an actual outcome.