A Theory of"Likes"

2024-08-21

AI Summary

This paper examines the role of recommendation systems in mitigating decision frictions arising from consumer preference heterogeneity and in disseminating economic information. Method: We develop the first Bayesian framework to characterize the divergent impacts of recommendation mechanisms on social welfare and platform profit under non-strategic user behavior. Contribution/Results: We theoretically establish that recommendation systems reshape demand structure through two distinct channels—“information aggregation” and “preference screening.” Crucially, we show that welfare-optimal design entails a fundamental trade-off between recommendation accuracy and coverage breadth, rather than maximizing click-through rates—a structural divergence from profit-maximizing mechanisms. Our analysis provides a rigorous welfare-economic benchmark for platform governance and algorithmic regulation, grounding policy-relevant insights in formal microeconomic foundations.

Technology Category

Multiagent Systems: Mechanism DesignGame Theory and Economic Paradigms: Mechanism DesignData Mining & Knowledge Management: Recommender Systems

Application Category

Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systemsUser Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systemsResponsible Web: Consent frameworks and practices on the web

Abstract

This paper investigates the value of recommendations for disseminating economic information, with a focus on frictions resulting from preference heterogeneity. We consider Bayesian expected-payoff maximizers who receive non-strategic recommendations by other consumers. We show by which channels a recommendation system influences consumer demand and welfare. Our analysis reveals how the welfare-maximizing design of a recommendation system may differ from what a profit-maximizing designer would choose to do.

Problem

• Investigates value of recommendations in economic information dissemination
• Analyzes impact of recommendation systems on consumer demand and welfare
• Compares welfare-maximizing vs profit-maximizing recommendation system designs

Innovation

• Bayesian expected-payoff maximizers analyze recommendations
• Recommendation system impacts demand and welfare
• Welfare-maximizing design differs from profit-maximizing design