Fairness Preferences Revisited

Abstract

This study revisits Engelmann and Strobel (2004), who test the theoretical models of fairness preferences with simple distribution games. We execute the experiment by recruiting a panel of subjects via the online platform Amazon Mechanical Turk three years prior and directly within the COVID-19 pandemic. We hypothesize that fairness preferences may be different across subject pools and malleable by the external shock of the pandemic. Based on several replication criteria from the literature, we replicate less than half of the preference estimates. Our findings suggest that efficiency concerns have persistent power for rationalizing allocation decisions while maximin preferences do not explain choices in our context. We further see disadvantageous inequality carry a significant weight. Finally, not all of the preferences seem to be stable over time, which underscores the importance of crisis events on fairness preferences in the short run. When discussing differences in findings from our and the original study we conclude that they may be attributed to differences in peers, fairness beliefs, or social image concerns.

Keywords: Fairness Preferences, Replication, Pandemic, Online Experiment, Mturk

JEL Classification: C91, D63, D90