Misled by Sensations? Quasi-experimental Field Evidence of Expectation-based Attribution Bias

Abstract

We test expectation-based attribution bias using field data from financial analysts. Expectation-based attribution bias refers to the tendency of individuals to misattribute sensations of elation or disappointment, stemming from comparisons of outcomes with prior expectations, and arises because people face difficulties in separating experienced utility into gain-loss and reference-free utility. Employing a regression discontinuity design, we find that analysts whose forecasts have barely been met become increasingly optimistic relative to those whose forecasts have barely been missed. This result is consistent with the misattribution of the gain-loss utility and biased updating of analysts’ beliefs. Furthermore, our analyses show that analysts whose forecasts have barely been missed update their beliefs more strongly than analysts whose forecasts have barely been met, which is consistent with asymmetric misencoding of gain and loss utility. We contribute to the literature by providing important field evidence of expectation-based attribution bias in an environment with high stakes and expert decision makers.

Keywords: Attribution Bias; Reference dependence; Information processing; Expectation formation; Regression discontinuity design

JEL Classification: D81, D83, D91, G41