Events at CERGE-EI

Tomas Jagelka, Ph.D.

University of Bonn, Germany

Authors: Tomas Jagelka, Christian Belzil

Abstract: Preferences, skills, and other latent personal attributes (PSAs) are key drivers of inequalities in life outcomes. We propose a novel framework for quantifying, and accounting for, individuals’ effort and cognitive noise which confound estimates of PSAs based on observed behavior. We establish the ability of our framework to quantify the noise content of a given experimental design and to de-bias estimates of PSAs, in an application to a large-scale experimental dataset measuring risk preferences. While the two elicitation designs we study were used interchangeably in the past, we estimate that a change from the more complex design to the more intuitive one results in a 30% decrease in (rational) inattention. On the one hand, failure to properly account for decision errors results in estimates of risk preferences biased by 50% for the median individual. On the other hand, accounting for endogenous effort allows us to empirically reconcile competing models of discrete choice. Furthermore, the estimated individual effort propensities have external validity. We show that they capture low-stakes motivation which generalizes to other settings and predicts, inter alia, an individual’s performance on the highly influential PISA achievement test.