Incentivizing unobservable effort in risky environments, such as in insurance, credit, and labor markets, is vital as moral hazard may otherwise cause significant welfare losses including the outright failure of markets. Ensuring incentive-compatibility through state-contingent contracts between principal and agent, however, is undesirable for risk-averse agents. We provide a theoretical intuition on how pro-social preferences between agents in a joint liability group con-tract can ensure incentive-compatibility. Two independent large-scale behavioral experiments framed in an insurance context support the hypotheses derived from our theory. In particular, effort decreases when making agents’ payoff less state-dependent, but this effect is mitigated with joint liability in a group scheme where agents are additionally motivated by pro-social concerns. Activating strategic motives slightly increases effort further; particularly in non-anonymous groups with high network strength. The results support existing evidence on joint liability groups and further suggest that even if peer pressure to ensure effort provision is absent, such group policies can improve efficiency when agents are pro-social.
