My research lies at the intersection of industrial organization, digital economics, and applied microeconomic theory. I study how data, algorithms, and AI shape competition and consumer welfare, focusing on digital platforms, privacy and data integration, online advertising, and algorithmic pricing.
Working Papers
The Cost of Exercising Rights: Consent Design in Data-Driven Platforms
Under review.
Abstract. Digital platforms can grant users formal privacy rights while making those rights costly to exercise. This paper studies consent design in advertising-funded platforms that choose opt-out frictions before users choose platforms and consent. Higher refusal costs expand monetizable data and may improve data-driven quality, but burden users who refuse. With captive users, platform-chosen frictions are excessive when the marginal refusal burden exceeds the quality value of induced consent, and insufficient when the reverse holds. Under competition, privacy-sensitive users sort across platforms, so refusal costs can repel or attract demand. The welfare effects of friction caps also depend on business stealing.
Implications of LLM Integration in Advertising
with Eduardo M. Azevedo and Pinar Yildirim
Abstract. We examine how richer query information changes the revenue economics of search advertising. We model the move from keyword search to conversational search as a refinement of the platform’s information about query-level click-through rates and ask how this refinement affects expected revenue under different selling rules. We find that, under a fixed weighted second-price auction, finer information weakly improves matching but has an ambiguous effect on revenue. This is because more precise queries can thin out competition within realized auctions. When the platform can adapt the mechanism to the information it observes, finer information weakly raises expected revenue. With constant demand elasticities, the revenue-maximizing auction is a weighted second-price auction with tuned weights and reserves, an auction format major platforms use. Conversational search also raises questions regarding the optimal timing of ads in a conversation. We argue that a platform may prefer to wait before showing an ad, because later turns can reveal more about what the user wants, although the user may leave before then.
Research in Progress
Further research projects are underway.
Policy Reports
Generative AI Outlook Report
European Commission Joint Research Centre, 2025
Underwater Archaeological Tourism: Demand Characteristics, Supply Structure and Growth Levers
SRM – Centro Ricerche e Studi, 2025