Research team

Structural analysis of negative interdependent preferences using revealed preference. 01/10/2022 - 30/09/2026

Abstract

Individuals care not only about income per se, but also about how much they earn compared to others. Social comparisons and interdependent preferences have a profound effect on individual welfare. With negative interdependent preferences, better material outcomes for peers reduce the well-being of the individual. Social comparisons also shape the behavior of individuals. To describe, explain, and predict this behavior, the current proposal (STANDING) incorporates social comparisons in state-of-the-art labor supply models. The research has two distinguishing features. First, it shifts attention from identifying peer groups –a notoriously difficult task– to discovering properties of the relative income concerns. STANDING focuses on relative income concerns between given parties (players in an experiment, spouses in a couple) or simply uses reference points stated by the respondents. This will facilitate integration of social comparisons in commonly used economic models. Second, STANDING will develop novel nonparametric, revealed preference, methods to bring the models to the data. These methods impose very few ad-hoc restrictions on the preferences for material outcomes and relative income, and this allows a robust analysis of the latter. STANDING will use the new structural analysis, based on revealed preference methods, to study interdependent preferences in a variety of contexts. First, individuals who dislike their position in the income distribution may be willing to reduce others' –and own– income to improve it. Money burning experiments demonstrated that individuals sometimes destroy others' pay-offs even if this is costly. This raises the question whether choices driven by negative interdependent preferences are still consistent with basic economic theory. WP1 implements a modified version of the money burning experiment that is well-suited for testing choice consistency. Second, a household may judge its position in the income distribution by the visible consumption of others. But visible, conspicuous consumption sends biased signals about the true income distribution. What is the effect of misperceptions on the household's willingness-to-work? WP2 builds a model in which households value leisure and their (perceived) position in the income distribution. The perceived position may differ from the actual one. A counterfactual analysis will demonstrate how misperceptions distort labor supply decisions. Finally, negative interdependent preferences may also exist within (!) the household, in the form of gender identity norms. How will female labor supply respond to a wage raise if there are strong breadwinner norms in favor of the husband? WP3 proposes a structural nonparametric analysis of interdependent preferences in the household. This can be used outside specially tailored laboratory experiments. We will apply the analyses of WP2 and WP3 to data from the Longitudinal Internet Studies for the Social sciences in the Netherlands. This dataset contains detailed information on perceived income distributions and gender norms.

Researcher(s)

Research team(s)

Project type(s)

  • Research Project