Cold Hand Decision-making

Daniel Chen (ETH Zurich/Center of Law and Economics)
Tobias Moskowitz (University of Chicago/Booth School of Business)
Kelly Shue (University of Chicago/Booth School of Business)

Abstract: We document negative autocorrelation in judgment and decision-making in three field con- texts: refugee judges, baseball umpires, and loan officers. Previous research on the hot hand fallacy suggests that agents often think of sequential streaks of 0’s or 1’s as evidence of bias even though such streaks are likely to occur through flips of a fair coin. We hypothesize that the hot hand fallacy and active response to fairness perceptions lead agents to engage in cold hand, i.e. negatively autocorrelated, decision-making. A judge worried about becoming or appearing too lenient or tough if she issues a streak of affirmative or negative decisions may actively adjust her decisions in the opposite direction. We find evidence of negative autocorrelation, particularly among moderate decision-makers, and stronger effects after recent streaks of decisions and in situations when agents face weak incentives for accuracy. We show that our findings are un- likely to be driven by sequential contrast effects and we build a model suggesting the degree of coarse-thinking agents must engage in to explain our results.


Download the paper