A principal objective of competition and regulation policymakers is to analyse product markets and design appropriate policies to address any market failures. A key input into this process is the analysis of consumer demand. Recent research in industrial organisation has developed and applied methods to estimate demand in many settings, with applications including simulating mergers and quantifying the benefits of new goods. Existing research, however, typically takes the products offered by firms as given. This has important econometric and economic implications. If products are chosen by firms, existing econometric methods are likely to yield biased estimates of key demand parameters. Furthermore, the types of products firms sell can influence welfare as much or more than prices. This project develops two potential solutions to this problem and demonstrates their relevance for US pay television services. Econometrically, we propose to develop 'orthogonal instruments': variables that can be used to consistently estimate key demand parameters even if firms choose their products. Economically, we propose to generalise existing empirical models to allow for firms' optimal choices of prices and product qualities. This will permit us to measure 'quality mark-ups' and quantify competition's impact on reducing distortions in quality provision (as well as prices).

All of the data used in this grant was obtained from data collection firms or agencies. The sampling methods varied by dataset. See attached documentation file for details.