Friends of Victims: Personal Relationships and Social Preferences Note: This Paper Is Being Revised to Incorporate 2 More Studies

This paper studies the impact of personally knowing a victim on social preferences for the welfare of other victims of the same misfortune. We begin with the analysis of a survey of volunteers, which shows that they tend to volunteer for organizations that target misfortunes previously suffered by their friends and relatives. In order to control for informational asymmetries, among other potential confounds, we conducted an experiment in which we randomly assigned participants to befriend a victim (a fellow participant who lost an initial endowment of $10) or a non-victim and found that participants who were paired with a victim gave more generous dictator offers to another anonymous victim. *Small: 700 Huntsman Hall, 3730 Walnut Street, Philadelphia PA19104, email: [email protected], homepage: http://marketing.wharton.upenn.edu/people/faculty/small.cfm. Simonsohn: 562 Huntsman Hall, 3730 Walnut Street, Philadelphia, PA19104, email: [email protected], homepage: http://dp.wharton.upenn.edu We are thankful to Eli Finkel, George Loewenstein, Jason Riis, Maurice Schweitzer, and attendees of seminars at Princeton and Stanford universities for helpful comments and discussions. We also thank Jessica Goldenberg, Laura Hotaling, Kwan-Kit Lui, Nicole Verrochi, Wesley Williams, and Yingyin Zhou for help with data collection, and the volunteer coordinators from Action Aids, Alzheimer’s Association, and the Special Olympics for collaborating with our survey.