Endogenous quality selection by multiproduct firms

In vertically differentiated markets, the effects of firm entry differ substantially when firms compete over both price and quality rather than just price. In Chapter 1, I define a measure of dissimilarity between cable programming packages, and use analysis of variance to quantify the effects of satellite entry and horizontal integration on cross-market variety in content offerings. Entry reduces variation across integrated firms but causes more tailoring of contents to local market conditions, resulting in greater overall variance. In Chapter 2, I estimate a structural model of supply and demand for subscription television that takes into account endogenous quality choice. Using counterfactual analysis, I decompose the effect of entry into a price response and an endogenous quality response (measured by changes in programming content). I find that raising both price and quality for the most comprehensive subscription package---competing "head-to-head"---is the rational incumbent response in markets with relatively homogeneous consumer types, and accounts for why observed cable prices increase "in spite of" entry. Elsewhere, incumbents respond less aggressively and target lower-end consumers following entry. Endogenous quality also leads to social inefficiencies relative to pure price competition. In particular, head-to-head competition results in "crowding" of quality choices toward the high end of the market and inefficiently low product differentiation. Chapter 3 deals with endogenous quality choice under more general demand specifications. The ability to set a distinct price for all possible combinations of goods leads to weakly higher profits relative to more restrictive schemes. In practice, firms usually use simpler plans such as component pricing (separate price for each item) and bundle-size pricing (prices based on the number of items bought). Using computational experiments, we assess how closely various simpler plans approximate optimal profits, under a broad range of assumptions about costs and demand. Bundle-size pricing attains nearly optimal profits and outperforms component pricing, barring exceptional circumstances. We also propose and estimate a demand model for a theater company selling bundled season subscriptions. We observe the performances selected by each subscriber, allowing identification of the correlation structure for consumer tastes. Using these estimates, we compute profits under various counterfactual price schemes.