Why Do Donors Give to Less Effective Charities? (Study 4)

This study is a follow-up to our project “Why Do Donors Give to Less Effective Charities?” In our initial experiments with 1,320 U.S. donors recruited via Prolific, we examined how donors respond to information about charity effectiveness across four causes—education, climate, health, and clean water. We found that a substantial share of participants were splitters who divided their donations equally between two charities, even when one was demonstrably more effective. This pattern was largely driven by fairness concerns toward the charities themselves rather than toward recipients, suggesting that many donors view equal allocation as a fair and appropriate strategy, reflecting broader egalitarian and non-meritocratic fairness preferences. In this original study, differences in cost-effectiveness were embedded within specific cause domains (e.g., health or education), and the fundraising appeal included an industry-typical mission statement describing the concrete interventions of each charity. This leaves open the possibility that the results were influenced by intervention-specific preferences or perceptions of the interventions themselves rather than by fairness concerns toward the charities as institutions. To address this, we replicate the experimental design while presenting charities as abstract institutions that differ only in their cost-effectiveness, without specifying any details of their interventions. This design allows us to test whether donors’ tendency to split persists when the decision context is fully neutralized with respect to intervention-related preferences. In addition, we further investigate the concept of charity fairness by including a vignette module that explores splitters’ fairness preferences in greater depth. In this module, we manipulate the perceived source of inequality in charity effectiveness—specifically, whether differences arise from effort, luck, talent, or the maturity of the organization—and measure how these attributions affect donors’ willingness to split their donations. Together, these extensions enable us to test the robustness and generality of our previous results and to better understand how donors interpret and justify inequality between charities.