Fast and slow dynamic decision making under ambiguity

• Published in: Journal of risk and uncertainty Vol. 70; no. 2; pp. 89 - 104
• Main Authors: Caferra, Rocco, Hey, John, Morone, Andrea
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.04.2025; Springer Nature B.V
• Subjects: Ambiguity; Decision makers; Decision making; Economic Theory/Quantitative Economics/Mathematical Methods; Economics; Economics and Finance; Environmental Economics; Expected utility; Incentives; Induction; Microeconomics; Monetary incentives; Neutrality; Operations Research/Decision Theory; Uncertainty; C24; Myopic; C23; Ambiguity box; Risk; Resolute; Dynamic inconsistency; Sequential choice; Choice under uncertainty; Expected Utility; Dual-decision process; C91; D81; Sophisticated; Decision time; Probability gap; Ambiguity
• ISSN: 0895-5646; 1573-0476
• DOI: 10.1007/s11166-024-09445-3

Different people think in different ways, and their behaviour can be analysed in different ways. In this paper, we analyse the correlation between the type of behaviour and the time taken to reach a decision in a dynamic context and under ambiguity with different monetary incentives, linking the results with fast and slow thinking processes. Four different types of dynamic decision-makers are identified: Resolute, Myopic, Sophisticated, and Expected Utility (EU). The different types use different methods to solve dynamic problems: A Resolute decision-maker (DM) decides right at the beginning his or her strategy, a Myopic DM simplifies the problem by ignoring part of it, a Sophisticated DM works by backward induction, and an EU DM either works by backward induction or by using the Strategy Method. We use data from (Caferra et al., 2023 ) where subjects were asked to solve a two-stage dynamic allocation problem. In that experiment, there were two treatments, incentivised and unincentivised. We found that their type matters: EU subjects take more time to solve the ambiguity, showing a relationship between dynamic consistency and ambiguity-neutrality with a deliberative thinking process. We also found that subjects in the non-incentivised treatment take less time, indicating that monetary incentives matter. The gap between the probabilities at each stage appears to be a good predictor of uncertainty for uncertainty averse subjects: the higher is the gap, the clearer is the most probable event and the lower is the time subjects spend to solve the decision problem.