Template-Type: ReDIF-Paper 1.0
Title: Inflated Recommendations
Author-Name: Martin Peitz
Author-Email:Martin.Peitz@gmail.com
Author-Name: Anton Sobolev
Author-Email:anton.sobolev@uni-mannheim.de

Classification-JEL: L12, L15, D21, D42, M37
Keywords: recommendation bias, recommender system, asymmetric information, experience
good, intermediation

Abstract: Biased recommendations arise naturally in markets with heterogeneous consumers.
We study a model in which a monopolist offers an experience good to a population
of consumers with heterogeneous tastes and makes personalized purchase recommendations.
We provide conditions under which a firm makes welfare-reducing purchase
recommendations with positive probability, resulting in inflated recommendations. We
extend this insight to a setting in which an intermediary makes the recommendations,
whereas a seller sets the retail price. Regulatory interventions that forbid inflated recommendations
may lead to higher social welfare or may backfire.

Note: 
Length: 53
Creation-Date: 2024-03
Revision-Date: 
File-URL: https://www.crctr224.de/research/discussion-papers/archive/dp336
File-Format: application/pdf
Handle: RePEc:bon:boncrc:CRCTR224_2022_336v2