Essays in Empirical Industrial Organization.

Summary This thesis consists of three chapters. The first chapter studies the effects of a subsidy in the case where firms can change the price as well as the characteristics of the product they offer. The second chapter (joint with Christian Bontemps and Cristina Gualdani) constructs and estimates a two-stage model of airline network competition. The third chapter (joint with Charles Pébereau) studies the adoption of real-time electricity pricing. In the first chapter, I study the economic effects of subsidies on the market for electric cars, which are subsidized around the world because they are seen as a key element in the de-carbonization of the transportation sector. In response to these subsidies, electric car producers have the ability to adjust not only the price of the cars but also their range. My analysis shows that the way in which these subsidies are implemented has an important impact on these price and range choices. In my study, I find that a subsidy directly indexed to range leads firms to sell electric cars that are more expensive and have more range. In contrast, a fixed subsidy leads firms to sell cheaper electric cars with less range. These effects have important implications for policy makers who have two objectives: While the quantity of electric cars sold is maximized with the fixed subsidy, minimizing CO2 emissions requires a trade-off between maximizing sales of electric cars and substituting high-polluting conventional cars. This trade-off is resolved with an intermediate subsidy. If a policymaker cannot achieve both goals with the same subsidy, he or she may well maximize electric car sales and the surplus of poorer consumers with the same subsidy. In the second chapter, written in collaboration with Christian Bontemps and Cristina Gualdani, we construct and estimate a two-stage model of the airline industry. In this model, airlines first choose their direct flight network before competing with its rivals. This two-stage game allows us to take into account the interdependence of the routes chosen by the firms. In addition, the model allows us to perform counterfactual analyses capable of providing robust predictions with respect to the price level, but also to the change in the firms' network. We show that hub-and-spoke networks lower marginal cost and increase fixed cost. We evaluate a merger between American Airlines and US Airways and compare it to the scenario of an American Airlines bankruptcy. We also evaluate countermeasures imposed on the merging firms and find that these measures were successful in limiting the loss of consumer surplus. In the third chapter, Charles Pébereau and I study the introduction of real-time electricity pricing in New Zealand and offer explanations for the low uptake of this pricing. Real-time pricing exposes consumers to the current price of electricity, which changes every half hour. We find that consumers who have adopted this technology most recently are very sensitive to current prices. Adoption rates fall sharply when current prices are high. During a crisis in current electricity prices, the share of consumers abandoning real-time pricing declines with their experience. These results suggest that, over time, consumers are less responsive to immediate price changes. Our results are useful in the debate about how to encourage consumers to adopt real-time pricing, such as opt-in or opt-out programs.