Survival, Attrition and Biased Decision-Making

Dmitriy Krichevskiy, Elizabethtown College


This paper considers the persistence of earnings differentials between wage earners and the self-employed through modeling overconfidence. The model examines challenges of limited information and its impact in the decision-making process. Using a Bayesian setting and concentrating specifically on the entrepreneur’s entry decision, the model shows how limited information can lead to substantial overestimation of potential profits and other outcomes even if the decision-maker has unbiased prior beliefs. The driving force behind the conclusion is the assumption that decision-makers, such as entrepreneurs, are not aware that available information is limited. For entrepreneurship, the resulting bias leads to overconfidence, excessive entry and the persistence of lower returns to entrepreneurship when compared to returns in the wage sector. This situation may arise wherever attrition is present, from military to business schools to medicine. Going the extra mile to gather and incorporate all relevant information into decision-making pays off in better quality decisions.