How endogenous risk preferences and sample selection affect analysis of firm survival

Abstract

<p>The roles of selection and endogeneity in measured risk preferences are illustrated using the correlation between risk attitudes and firm survival. Selection bias occurs when risk attitudes are elicited only from current entrepreneurs so that the risk attitudes of unsuccessful entrepreneurs are excluded from the analysis. Risk attitudes measured after agents enter entrepreneurship will endogenously reflect business success. Data on entrepreneurs from the National Longitudinal Survey of Youth 1979 and the Panel Study of Income Dynamics shows that when risk attitudes are measured subject to selection and endogeneity, mixed or even positive correlations between risk acceptance and the probability of firm exit occur. However, firm exits fall monotonically with willingness to accept risk when risk preference measures are not subject to selection or endogeneity related to business success.</p>