We consider the optimal policy problem of a benevolent planner, who is uncertain about an individual's true preferences because of inconsistencies in revealed preferences across behavioral frames. We adapt theories of expected utility maximization and ambiguity aversion to characterize the planner's objective, which results in welfarist criteria similar to social welfare functions, with intrapersonal frames replacing interpersonal types. Under paternalistic risk aversion or ambiguity aversion, a policy is less desirable to the planner, holding all else fixed, when it leads to more disagreement about welfare from revealed preferences. We map some examples of behavioral models into our framework and describe how this notion of robustness plays out in applied settings.
Authors
- Acknowledgements & Disclosure
- For valuable discussions and comments, we thank Scott Elliott, Jacob Goldin, Louis Kaplow, Ben Lockwood, Yusufcan Masatlioglu, Emel Filiz-Ozbay, Alex Rees-Jones, Joel Slemrod, Luminita Stevens, and Dmitry Taubinsky. Canishk thankfully acknowledges financial support from the Economic and Social Research Council DTP [Grant No: ES/P000622/1]. The authors have no relevant material financial or other interests that relate to this research paper. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.
- DOI
- https://doi.org/10.3386/w32813
- Pages
- 57
- Published in
- United States of America
Table of Contents
- Introduction 3
- General Model 6
- Core Assumptions on Individual Choices and Welfare 6
- Discussion of Setup and Relationship to Prior Literature 7
- Behavioral Optimal Policy Problems 9
- Known Normative Frame 9
- Unknown Normative Frame 10
- Probabilistic Uncertainty versus Ambiguity 11
- Known Normative Weights/Probabilistic Uncertainty 12
- Ambiguity 15
- Approaches to Comparability 18
- Money Metric Welfare 18
- Relationship to Interpersonal Comprability 20
- Perturbations 21
- Unstructured Welfare Weights 21
- Applying the Envelope Theorem with Fully Specified Objectives 22
- Under Risk Neutrality 23
- Under Paternalistic Risk Aversion 24
- Under Paternalistic Ambiguity Aversion 24
- Examples 25
- Example 1: Biases Versus Strange Preferences 25
- Example 2: Present Focus 28
- Example 3: Is a Feature of the Environment a Frame? 30
- Discussion: Identifying Normative Weights. 31
- Robust Optimality 33
- Robustness and Perturbations 33
- Robustness in General 34
- Robustness in Applications 34
- Optimal Defaults 35
- (Unconstrained) Optimal Reference Points 37
- Corrective Taxation 39
- Conclusion 40
- Convex Sets of Frames and Interpretation of Normative Weights 46
- Continuous frames and convex hulls 46
- Basics 46
- Linearity and Equivalence to Previous Objectives 46
- Counterfactual Normative Consumers and Identification of Normative Weights 47
- Proofs for Section 2 (General Model ) 46
- Proofs for Section 3 (Approaches to Comparability ) 47
- Proofs for Section 5 (Examples ) 53
- Proofs for Section 6 (Robust Optimality ) 54
