Personality, Preferences, and Reference-Dependence

Introduction

In recent years, the study of personality (noncognitive skills) has significantly increased in economics. There has been an explosion of empirical work that shows the promise and relevance of incorporating such factors into the field.

However, we lack a definition of relevant personality traits in terms of primitive preferences, the way we do with risk and time preferences, for instance. Most studies rely on self-reported surveys, but this is only part of the issue: without a sharp, agreed-upon definition, it is difficult to build models, make predictions, or elicit these measures in a way that unambiguously separates the primitives from the outcomes. My aim is to provide these notions.

Part 1: Attitudes Towards Success and Failure

Part 1 focuses on attitudes towards success and failure, and the thresholds that separate one from the other. Such attitudes are not only present in the way we discuss personality (e.g., perseverance and tenacity), but they also appear in seemingly distant strands of economics, notably through the widely used notion of gain-loss sensitivity, notions of aspirations, etc.

Within a standard decision-theoretic setting, we define such attitudes as features of the primitive preferences over lotteries. We characterize the corresponding utility representations and provide indices of the intensity of each attitude, in a manner analogous to the Arrow-Pratt coefficients for risk aversion.

Implications

These notions have crucial implications on two main levels:

1. They generate testable predictions for choice under risk and novel insights about central notions within the established economics toolkit.
2. They expand the domain of economics methodology to the study of personality by providing choice-based measures grounded on preferences, thereby bridging the gap between insightful notions from psychology and the desiderata of rigorous economic analysis.

Parts 2 and 3: Experimental Approach

These two aspects are the focus of Parts 2 and 3, respectively, which tackle the problem through a series of large-scale experiments.