Working Paper: NBER ID: w27134
Authors: Laura Alfaro; Ester Faia; Nora Lamersdorf; Farzad Saidi
Abstract: Social interactions and social preferences play a central role in public health domains. In the face of a pandemic, individuals adjust their behavior, whereas in SIR models infection rates are typically exogenous. City-level data across countries suggest that mobility falls in response to fear, proxied by Google searches. Incorporating experimentally validated measures of social preferences at the regional level, we find that stringency measures matter less when individuals are more patient, altruistic, or exhibit less negative reciprocity. To account for these findings, we extend homogeneous and networked SIR models by endogenizing agents' social-activity intensity and incorporating social preferences. Our quantitative predictions markedly differ from those of the naïve SIR network model. We derive the planner's problem, and show that neglecting agents' endogenous response leads to misguided policy decisions of various non-pharmaceutical interventions. Any further mobility restrictions, beyond agents' restraint, result from aggregate externalities which are curtailed by social preferences.
Keywords: Social interactions; Pandemics; Fear; Altruism; Reciprocity
JEL Codes: D62; D64; D85; D91; I10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| fear (proxied by Google searches for coronavirus) (D83) | reduced mobility (J62) |
| higher levels of altruism and patience (D64) | less mobility response to government stringency measures (O24) |
| fear (Y60) | mobility response beyond government responses (J61) |
| altruism (D64) | lower likelihood of increasing mobility post-lockdown (J62) |
| trust (G21) | dampened mobility responses to government interventions (O24) |