Projects
Recommendation systems
Bye-box: An Analysis of Non-Promotion on the Amazon Marketplace by Matthias Hunold , Ulrich Laitenberger , and Guillaume Thébaudin.
Abstract: We study seller and product recommendations of the hybrid e-commerce platform Amazon. Using web-scraped data, we find that Amazon makes the visibility of offers of third-party suppliers in the "buybox" dependent on prices on competing marketplaces like Walmart and eBay. Amazon’s own offers are visible regardless of their competitiveness. We find that the absence of seller recommendations makes recommendations to related products more effective and Amazon tends to steer consumers in these situations more often to products it sells itself. We discuss that this behavior is difficult to reconcile with the hypothesis of an independent marketplace operator.
Presented at: German-French Workshop on E-Commerce; Dyson School of Management Strategy Seminar; Mannheim ICT Conference 2021; EARIE Conference 2021; Centre for Competition Policy Seminar; Economics Seminar, University of Siegen; German IO Committee 2021; University of Würzburg; Goethe University Marketin Semianr; KU Leuven MSI Seminar; Düsseldorf Institute for Competition Economics; Economics of Platforms Seminar (Video) École de Mines, CERNA Seminar; CESIfo Conference on the Economics of Digitization 2022; University of Gießen; University of Duisburg; Workshop on Information Systems and Economics, Copenhagen.
Choosing between recommender systems – Experimental Evidence by Felix Schleef
Abstract: I study how subjects choose from lists when they are ordered by algorithms, and how subjects evaluate and value recommender algorithms.
As a growing body of literature documents biased intermediation by large online platforms, policy makers have started to design regulations for recommender systems. In this project, I use a laboratory experiment to study what happens when consumers are given the choice of multiple different recommender algorithms.
Platform liability
(Hybrid) Platform Screening by Marc Bourreau and Guillaume Thébaudin
Abstract: Online platforms can invest in screening tools to detect harmful products before they are listed on their marketplace. Using a model of monopolistic competition with free entry of sellers and information asymmetry regarding seller types, we study the incentives for platforms to do so. We find that screening policy ambiguously affect sellers’ entry decision, and a necessary condition for platforms to engage in screening is that higher screening intensity raises seller entry. Second, we find that hybrid platforms, which sell their own products on their marketplace, have lower incentives to engage in screening than pure ones, for a given commission rate. Hybrid platforms indeed benefit from the presence of harmful third-party products because it makes their own products more attractive to consumers. We finally investigate the robustness of this result with endogenous commission fees. Our results have important policy implications in the context of the implementation of the Digital Service Act in Europe.
Data sharing
Abstract: Considering a monopolistic hybrid platform, we investigate the effect of a mandated data sharing policy on market outcomes across different data functionalities (price discrimination and cost reduction) and different market structures (perfect and imperfect competition). We find that mandated data sharing has no effects on welfare if data can be used to price discriminate consumers who buy homogeneous goods, while mandated sharing of cost-reducing data improves welfare by lowering the average price in the markets. When goods are horizontally differentiated, data sharing with price discrimination (cost reduction) may instead negatively (positively) affect consumers. We argue that in markets where competition is softer, mandatory data sharing may damage the very agents it is intended to protect, namely consumers and (efficient) sellers.
Information provision in hybrid platforms by Marco Magnani and Federico Navarra
Abstract: We study the incentives of a monopolistic hybrid platform in sharing its superior market information with the third-party seller hosted on its marketplace. After observing platform information-sharing policy, the seller competes in prices with the platform over a horizontally differentiated good. Despite platform duality, an equilibrium in which the platform shares information with the seller occurs. We highlight how the platform has incentives to share information either for relaxing price-competition or for increasing the volume of transactions. Platform incentives to share information are strongest for intermediate degrees of product differentiation. Information provision results in consumer surplus extraction such that the total welfare is reduced. Although entering as a seller and providing market information is profitable, when analysing platform entry as the acquisition of one of the sellers we may observe equilibria in which the platform either sticks to agency or does not provide information since this would increase the entry cost.