Research Interests

Economics of Risk and Uncertainty; Insurance; Behavioral Finance; Behavioral Decision Theory; Experimental Economics

Job Market Paper

[1] Too Risky to Hedge: Theory and Experiment on Narrow Framing (PDF)
Abstract: Here, we study insurance decisions when the policyholder evaluates insurance with narrow framing. We show that due to aversion to risk on the net insurance payoff, narrow framing reduces insurance demand in the form of both coinsurance and deductible insurance. We also show that the optimal insurance contract involves a deductible and the coinsurance of losses above the deductible when transaction costs depend on the actuarial value of the policy. In an incentivized lab experiment, we document substantial effects of narrow framing on hedging. By estimating a structural model, we find that people give a weight of 43% to the utility from evaluating insurance in isolation and 57% to the hedging value of the contract. We also find that individuals with lower cognitive abilities and lower demand from insurance place a significantly higher weight on the evaluation of insurance in isolation.

Working Papers

[2] Insurance in the Household: Individual Risk, Household Risk and Intra-household Inequality
with Hélène Couprie and Astrid Hopfensitz
Abstract: Risk is rarely an individual phenomenon and often shared in groups and households. In the case of informal insurance situations, individual risk preferences have to face the risk preferences of the group. We study for the case of spouses, how individual versus household risk preferences interact in an experimental paradigm. 202 cohabiting spouses (101 couples) participated in a controlled experimental risk taking task. We focus on a household risk task, in which spouses face the choice between an option in which risk is correlated (high household risk, low inequality among spouses) and an option allowing for hedging (low household risk, high inequality among spouses). We show that spouses are mainly influenced by household risk and do not react to inequality as long as payoffs are symmetric. However when payoffs become asymmetric, because one of the spouses risk is reduced, we observe a change in preferences. Specifically our results suggest that households put a higher weight on men’s individual risks. We further observe that married couples put a higher weight on expected utilities from joint household payoffs.

[3] Willingness to Pay for Reductions in Morbidity Risks under Anticipated Regret (An interview by NBS Forum on Risk Management and Insurance can be found here:
Abstract: Prevention decisions that reduce individuals’ morbidity risks are important, generally irreversible, and particularly difficult since they imply a trade-off between two important attributes: the safety and its cost. All those features make regret more likely to be anticipated. In this paper, we study the willingness to pay for reductions in morbidity risks within a framework of anticipated regret. As results, we find that with other things being equal, an individual who is disproportionately averse to large regrets has a higher willingness to pay than a standard expected utility individual. This notion of regret aversion has been shown to be relevant for regret theory to reconcile with many decision patterns which are not in line with standard expected utility theory. Moreover, the effect induced by this notion of regret aversion can be interpreted as if the regret averse individual overweighs risk reductions due to prevention, i.e., probability weighting effect. We further discuss how the resolution of uncertainty may affect the regret averse individual’s willingness to pay, and how the attribution effect in belief formation may bias its estimates.

[4] Belief-based Regret Theory
Abstract: Existing regret theories consider situations where the realized state of nature is observed or at least can be inferred with certainty. In reality, people may only observe the outcomes from their actions and lack of knowledge about how these outcomes were generated. But nothing prevents people from feeling regret when they counterfactually think of what could have happened had they behaved differently given their posterior beliefs about the realized state of nature. In this paper, we provide a decision criterion under anticipated regret, assuming that people feel regret through comparing the obtained consumption utility with the highest expected consumption utility that could have been obtained. This assumption holds if adding (or removing) a state-wise dominated action from the choice set does not affect the optimal choice. We further apply the decision criterion to a self-protection problem and find that regret tends to meditate risk prevention behavior, i.e., a regret sensitive individual invests more effort in self-protection when an expected utility individual invests too low, and vice versa.

Research in Progress (Selected)

[5] Regret and Restricted Choice: An Experimental Study 
with Felix Fattinger and Moritz Loewenfeld
Abstract: In real-world investment decisions, investors face varying choice sets due to a multiplicity of reasons such as institutional or liquidity constraints. Existing studies have shown that individuals might be better off with less choice due to regret, lacking self-control, or choice overload. However, little is known about how variations in choice sets affect investment choices among the available options. We use regret theory to derive predictions about how choice sets will impact behavior in a simple but realistic investment task and test them with an incentivized lab experiment. Moreover, manipulating choice sets provides an exogenous variation in the space of potential regret, which allows us to test regret theory in an incentive compatible way. We thus contribute to the literature in two ways: (1) testing regret theory using an incentive-compatible approach, and (2) examining the impact of choice set restrictions on investment behavior. We find that choice set effects are in the order of one full standard deviation of investment decisions. Moreover, these effects are in line with predictions from regret theory.

[6] Behavioral Determinants on Individual Preferences for Reducing Health risks: A Contingent Valuation Study
with Henrik Andersson and James Hammitt
Abstract: Recent theoretical works in behavioral and health economics suggest that regret and loss aversion can influence individuals’ willingness to pay (WTP) and willingness to accept (WTA) for changes in health risks. Since the estimated WTP and WTA are important for policy evaluation, a better understanding of how individuals form their preferences is of high policy relevance. In this project we empirically test the hypotheses that individuals who are: (1) more sensitive to regret have higher WTP and WTA, and (2) more loss averse have a higher WTA-WTP ratio. A online-survey will be conducted on a random sample of respondents from the French population combined with tests for regret and loss aversion already used in the literature. We expect our findings to raise questions about how public health policies should be designed and implemented. Our study will therefore contribute to the debate on public-health policies and behavioral economics.