Human Capital of Misconduct in the U.S. Securities Industry

Abstract

When misconduct is committed by an organization or by its members in the course of working for the organization, there is often a debate about whether it resulted from the actions of a few rogue members (i.e., bad apples) or from the characteristics of the organization as a whole (i.e., bad barrels). In this respect, I seek to examine the human capital of misconduct by estimating the relative importance of individual versus organizational characteristics in explaining the likelihood of misconduct. To do so, I examine the mobility and matching patterns within the U.S. securities industry. Specifically, I exploit the licensing database of the U.S. securities industry’s self-regulatory authority to build a unique dataset of the careers of 10,000 U.S. stockbrokers, including information on their employers as well as instances of organizational misconduct. I apply two-way fixed effects models to estimate the percentage of variation in misconduct that can be attributed to fixed effects of individuals versus fixed effects of firms. My analyses across two different random samples of stockbrokers suggest that the variation in organizational misconduct is largely explained by individual differences rather than organizational differences – i.e., misconduct by the stockbrokers in the context of brokerage firms is more a product of “bad apples” rather than “bad barrels.” I also find support for negative, rather than positive, sorting in the market based on ethics. Specifically, I find that, on average, bad apples are sorted into employment at less misconduct-facilitating firms – and vice versa. I employ several robustness checks to mitigate a potential bias in correlation between stockbroker and firm fixed effects by focusing on sub-samples with higher observed stockbroker mobility – these checks support my findings. I discuss my contributions to scholarship as well as the implications of my research for practice and policy.