Warm Glow Gambles: How Risk Fundamentally Transforms Prosocial Behavior
Posted: 20 May 2020 Last revised: 3 Aug 2020
Date Written: July 22, 2020
Abstract
A fundamental puzzle in the social and natural sciences is why humans, in contrast to other animals, routinely help strangers at substantial personal cost. Current literature documents three regularities in prosocial behavior: first, prosocial behavior increases if people have more economic resources (e.g., money) or psychological resources (e.g., positive affect or energy); second, prosocial behavior is positively correlated with “warm glow giving” motivations; third, prosocial behavior decays with repetition. However, these three regularities are overwhelmingly based on studies of riskless prosocial behaviors. We use field and lab studies to show that, for risky prosocial behaviors that might succeed or might fail, all three regularities actually reverse – in contrast to previous findings, theories, and lay beliefs. Our results point to fundamental differences between riskless prosocial behaviors (e.g., donating) and risky prosocial behaviors (e.g., volunteering; prosocial investing). For example, we examine a unique, massive 2013-2018 dataset of nearly 3 million time-stamped prosocial behaviors by volunteers; contradicting the first regularity, we show that positive affect shocks (e.g., positive stock returns, happier days) actually cause prosocial behavior to decrease. We argue that, in contrast to riskless prosocial behaviors which reliably create warm glow, risky prosocial behaviors instead create warm glow gambles: if a prosocial act succeeds, the actor feels better and energized, but if it fails, the actor feels worse and demotivated. While warm glow is driven by the psychology of self-signaling, warm glow gambles are instead dominated by the psychology of risk-taking, as exemplified by Prospect Theory (Kahneman & Tversky, 1979).
Keywords: prosocial behavior; warm glow; prospect theory
JEL Classification: D03,H41, L31
Suggested Citation: Suggested Citation
