Séminaires LEMMA

"Consumers' Privacy Sensitivity in Digital Markets: Evidence from the Mobile App Industry"

Hugo ALLOUARD (ESSEC Business School)

Tuesday, 25 November 2025, 11h-12h

Lemma - 4 rue Blaise Desgoffe, 75006 Paris. Salle Maurice Desplas

Hugo is post-doctoral researcher in economics at ESSEC Business School. His research focuses on industrial organization, digital, public, and labor economics, as well as structural econometrics.

Abstract : Prior research consistently shows that individuals are sensitive to privacy, yet it often fails to separate the benefits from the costs of sharing personal information. This paper addresses that gap by estimating a structural demand model using real-market data from the mobile gaming industry between 2015 and 2021 in Canada, France, Germany, and the United States. The results reveal a clear trade-off for both users and firms. Although sharing personal data generates disutility due to privacy concerns, consumers also benefit when developers use that data to deliver targeted features, advertisements, and other data-dependent enhancements. The analysis shows that ignoring these quality improvements leads to systematic underestimation of both willingness to pay for privacy and privacy sensitivity. Once these benefits are incorporated, estimates of willingness to pay for privacy converge toward those found in prior experimental studies, helping to explain the gap between observational and experimental results in the literature.

"Linking Mechanisms with Few Messages"

Maël LE TREUST (CNRS), en collaboration avec Tristan Tomala

Tuesday, 18 November 2025, 11h-12h

Lemma - 4 rue Blaise Desgoffe, 75006 Paris. Salle Maurice Desplas

Abstract : We investigate the mechanism design problem formulated by Jackson and Sonnenschein in 2007, where we consider that agents do not have enough messages to reveal their type. This problem is deeply related to the mismatched distortion-rate problem formulated by Lapidoth in 1997, in the Information Theory literature. The characterization of the set of single-letter incentive compatible distributions is an open problem. We provide inner and an outer bounds and we show these bounds match in several special cases.