Jaden Chen, University of North Carolina-Chapel Hill
Will Cong, Nanyang Technological University
Siguang Li, Hong Kong University of Science & Technology
Abstract: We study endogenous information provision and source authentication when secondary senders can copy primary senders' signals, providing a microfoundation for correlation neglect. Authentication mitigates this duplication bias but hinders information diffusion, creating ambiguous effects on misinformation and welfare. Crucially, we show that policies aimed at maximizing user welfare can be fundamentally misaligned with the goal of minimizing misinformation. Non-verification can be optimal when diffusion is highly valued or primary senders hold strong bargaining power. While factors like signal quality reduce misinformation under exogenous verification, the effects are uneven when verification is endogenous. We also examine intellectual property protection and self-regulation, consistently highlighting a core trade-off between information accuracy and diffusion in shaping platform policy and welfare outcomes.
Discussant: Snehal Banerjee, University of Michigan
Maryam Farboodi, Massachusetts Institute of Technology
Peter Kondor, London School of Economics and Political Science
Pablo Kurlat, University of Southern California
Abstract: We develop a parsimonious model of credit market competition where ex-ante identical lenders endogenously choose screening technologies and interest rates. Two key spillovers shape equilibrium outcomes: rejected borrowers reapply at other lenders, affecting application pool quality, and screening signals across lenders are correlated rather than conditionally independent. The equilibrium features a hockey stick interest rate schedule - a segmented market structure with varying degrees of fragmentation across borrower opacity levels. Lenders in different segments come to resemble traditional banks, fintech firms, or credit card issuers. We apply our framework to study how technological progress spills over across market segments. We show that whilst AI adoption increases financial inclusion, mandatory data sharing regulation counter-intuitively may not benefit underserved populations and can increase inequality in financial access.
Discussant: Pavel Zryumov, University of Rochester
Abstract: I analyze a dynamic model of concurrent bargaining in which multiple prospective buyers compete to trade with an informed seller. When the seller maintains confidentiality over buyers’ past offers, buyers may engage in competitive “price experimentation”: buyers risk early losses to subsequently acquire informational advantages over competitors and expect to earn future information rents. Due to price experimentation, the seller may benefit from maintaining confidentiality over past offers and restricting buyer entry. The model has implications for the strategic choice between auctions and negotiations, and for the common use of “pre-qualification” in asset sales
Discussant: Brett Green, Washington University-St. Louis
