Intertemporal Aspects of Risk Management and their Behavioral Implications
Final Report Abstract
This research project investigated how intertemporal considerations and related behavioral biases influence insurance demand and risk management decisions. The project was motivated by the observation that insurance contracts are inherently intertemporal – premiums are paid before risks materialize – yet traditional models often overlook this temporal structure. Three sub-projects combined theoretical modeling with incentive-compatible experiments to explore psychological mechanisms that affect insurance choices. Sub-Project 1: Risk Management and the Autonomy of Labor. A framed field experiment on Amazon MTurk examined whether individuals integrate preventative risk management with intertemporal scheduling decisions. Workers in the high-autonomy group were 9.7 percentage points more likely to engage in preventative effort, supporting the broad bracketing hypothesis. These results inform the design of incentive systems and autonomy structures, especially in the gig economy. Sub-Project 2: Insurance and the Preference for Early Resolution of Uncertainty (PERU). Building on Kreps-Porteus-Selden and Hansen-Sargent models, this sub-project tested whether individuals who prefer early resolution of uncertainty are more willing to pay for insurance that prevents late resolution of uncertainty. A laboratory experiment with 204 participants showed a directional effect: those in the “late resolution” condition had an 8.4% higher willingness to pay, but the difference was not statistically significant. Structural estimates confirmed weak PERU tendencies. Sub-Project 3: Insurance as Commitment? This part explored whether people use insurance as a commitment device to counter presentbiased behavior. A theoretical model with β-δ preferences predicted a preference for commitmentbased insurance contracts. In a four-week real-effort experiment with 250+ participants, three contract types (flexible, personalized commitment, and imposed commitment) were compared. No statistically significant differences in willingness to pay were observed. Structural estimates reve- ˆ aled that participants did not perceive themselves as present-biased (β = 1.05), aligning with the null result. Conclusion. While the theoretical models identified plausible behavioral effects, empirical support – particularly in sub-projects 2 and 3 – was limited. Nonetheless, the findings enhance our understanding of how intertemporal behavior might (or might not) influence insurance behavior. The interested public can use the findings to better understand the temporal dimension of insurance contracts, how it influences human behavior, and potential implications for contract design and public policy.
Publications
-
Risk Management and the Autonomy of Labor. National Bureau of Economic Research.
Dambaur, Andreas; Ericson, Keith Marzilli; Jaspersen, Johannes & Zoller, Sandra
-
Insurance and the Preference for Early Resolution of Uncertainty. SSRN Electronic Journal.
Jaspersen, Johannes Gerd; Peter, Richard & Ragin, Marc A.
