Competition and vertical/agglomeration effects in media mergers : bagging bundle benefits

Existing frameworks (such as used by the New Zealand Commerce Commission in its recent evaluation of the proposed merger between Sky Television and Vodafone) require, as a first step, the definition of the relevant markets affected by the merger or vertical integration activity. Historic precedents in the telecommunications sector have tended towards finding that vertical agglomeration effects when network operators integrate downstream into the provision of applications and services to end-consumers are harmful to competition. Such Structure-Conduct-Performance methods of evaluating mergers and other aspects of market performance are problematic when the firm(s) concerned supply many different products, both together in various different bundle forms and separately as individual components. Defining the markets for (merger) analysis on the basis of only one of the components in a possible bundle that the (merged) firm may supply risks overlooking the complex interactions that occur on the demand side when consumers make their purchase decisions. This is especially likely to be an issue in the supply of internet applications and content bundled with broadband internet access. Consumers have heterogeneous preferences for different applications and content (hereafter "content"), and will purchase (or access) many different content types. Even though ownership of rights to distribute one content may confer a degree of market power in for the owner-provider over those consumers with very strong preferences for this content over all others, it is not axiomatic that the firm will be able to exert this power over consumers whose preferences are more evenly distributed. The more variety there is in the content bundles available, and the more heterogeneous are consumers' preferences across the various content types, the greater is the number of possible markets in which interaction is likely to occur and the more problematic it becomes to identify the relevant markets for analysis of mergers and antitrust cases. We propose that classic merger and antitrust analysis based on econometric cost-benefit analysis can be augmented by using simulation and numerical analysis of a range of bundle offers expected to be relevant in decision-making. We develop a simple model and use it to demonstrate how this approach could have informed the recent New Zealand Commerce Commission decision about the proposed Sky-Vodafone merger by offering some quantitative estimates of total and consumer welfare and provider profits under the proposed factual (with bundling) and counterfactual (individual component sales) cases. The approach may also inform other analyses, such as the assessment of the effects of two-sided markets and firm pricing decisions.