Issues in Online Advertising Markets and Applied Econometrics
- Author(s): Gibbons, Charles Edouard
- Advisor(s): Scotchmer, Suzanne
- et al.
This thesis begins by examining the incentives present in online advertising markets. We consider the strategies of online advertising providers, firms, and consumers in the context of ad listings assigned by a generalized second price auction. The first part of the chapter develops a model of consumer responses to ad listings and product offerings from the included firms and uses this behavioral model to derive optimal bidding functions for the firms. We show that the relationship between per-sale margins and product-consumer match probabilities ("relevances") must meet certain conditions to rationalize this equilibrium for consumers and firms; in particular, we give the conditions for consumers to rationally search from the top of the listing downward. Next, we turn to the incentives facing the ad server to alter the relevances and margins of the firms and the search costs and valuations of the consumer pool. While these incentives align with the desires of consumers, they may conflict with those for firms. We calculate the optimal number of slots for the ad server to offer, which is less than that desired by firms and consumers. We also show that the ad server has an incentive to subsidize its own competitor in the product market. These results have important implications for competition policy, innovation, and online content provision.
Next, in a chapter coauthored with Juan Carlos Suarez Serrato and Michael B. Urbancic, we turn to a topic in applied econometrics: regressions with fixed effects. Though common in the applied literature, it is known that fixed effects regressions with a constant treatment effect generally do not consistently estimate the sample-weighted treatment effect. This chapter demonstrates the extent of the difference between the fixed effect estimate and the sample-weighted effect by replicating nine influential papers from the American Economic Review. We propose a model with fixed effects interactions to identify the sample-weighted treatment effect and derive a test that discriminates between this estimate and the standard fixed effects estimate. For all 9 papers in our replication, at least one set of fixed effects interactions is jointly significant; in 6 of 9 papers, there is a sample-weighted estimate that is statistically different from the standard fixed effects estimate. In 7 of 9 papers, the differences are economically significant (larger than 10%); the average of the largest difference between the estimators from each paper is over 50% and the median is 19.5%. Our procedure does not markedly increase the variance of the estimators in 7 of 9 papers.
Lastly, in a chapter coauthored with Michael B. Urbancic, we consider the use and interpretation of estimates from instrumental variables regressions. We elucidate a common mistake in the applied literature in comparing results from OLS and IV models. Often, this is comparison serves as a test for exogeneity, but this logic is flawed. We offer closed-form and graphical examples illustrating that equality of these estimates does not imply exogeneity and discuss how the Hausman test applied to the IV setting is misguided. We illustrate our point empirically by comparing estimates of the returns to education using a trio of standard instruments. We conclude by offering guidance for the applied researcher.