Working Paper: NBER ID: w21001
Authors: Andrew Caplin; John Leahy; Filip Matjka
Abstract: Popularity is self reinforcing. The attention garnered by popular options propels further interest in them. Yet rather than blindly follow the crowd, most pay attention to how well these items match their tastes. We model this role of social learning in guiding selective attention and market dynamics. We confirm that attention focuses on options that quickly achieve popularity. Information externalities render the chosen set smaller than socially optimal. This rationalizes antitrust policies that encourage early experimentation. When attention costs are based on Shannon entropy, optimal policies are computable. With rich data, optimal choices can be identified for all consumer types.
Keywords: Social Learning; Selective Attention; Market Dynamics; Antitrust Policies
JEL Codes: D83
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| social learning (C92) | selective attention towards popular options (D91) |
| selective attention towards popular options (D91) | market dynamics (D49) |
| social learning (C92) | market dynamics (D49) |
| attention costs based on Shannon entropy (D80) | choices of agents (L85) |
| increased attention to popular options (D72) | market concentration (L11) |
| antitrust policies aimed at reducing initial market share disparities (L49) | social welfare (I38) |
| social learning (C92) | private learning (H42) |