This folder contains the data of the Risk-Sharing experiment and the Rscripts used to analyze the Broodfonds survey data and the data of the Risk-Sharing experiment to create the tables and figures as reported in 'Vriens, E., & Buskens, V. (2020). Managing Risk Heterogeneity in Risk-Sharing Groups: A Multi-Method Study on Risk Aversion and Solidarity. Working Paper.' The data of the Broodfonds survey is not included in this data package, because it is only available upon request.
Abstract paper:
A recent revival of organizations that provide insurance in risk-sharing groups (e.g., Broodfonds in the Netherlands, Friendsurance in Germany) brings us back to the age-old question of how risk-sharing groups can survive despite their risk heterogeneity. Data tells us that members within these groups vary substantially in their risk of needing support, essentially meaning that low-risk members pay more to support others than they receive support in return. How can heterogeneous risk-sharing groups realize stable participation rates? In a multi-method study, we examine the potential of risk aversion and solidarity as compensators for heterogeneity by comparing the results of a survey conducted among 5192 members of 230 Broodfonds groups to an online experiment with 430 British subjects of the Prolific platform. While we find that risk heterogeneity has a negative effect on participation in Broodfonds groups, our experimental manipulation of heterogeneity has no significant effect. Moreover, risk aversion does not predict participation in the field study nor in the initial decision to join in the experiment. In the first, it only predicts participation for low-risk members and in the latter, it only explains continued participation. Solidarity motives, finally, are a strong predictor in both settings. These results have important implications for theories on sharing risk (in which traditionally risk aversion is a crucial factor) and for understanding what may make risk-sharing groups successful in practice.
