Understanding the (perverse) effects of disclosing conflicts of interest: A direct replication study

Abstract

Advisors are often subject to conflicts of interest—a potential clash between their professional responsibilities and personal interests. Such conflicts can increase bias in advice. Although disclosure is frequently proposed to manage conflicts of interest, it can have unintended consequences on both advisees and advisors. In seminal work, Cain, Loewenstein, and Moore (2005, 2011) demonstrated that advisors give more biased advice with disclosure than without, to the detriment of advisees’ financial payoffs. However, recent experiments by Sah (2018), have revealed that in certain contexts, advisors give less biased advice with disclosure relative to nondisclosure. To further understand the contradictory findings and increase confidence in the original results, I conducted a direct replication of the main study from Cain et al. (2011) on advisors (with slight changes for advisees). I replicated the original finding that advisors give more biased advice with conflict-of-interest disclosure (vs. without) when giving recommendations on the sale price of houses to advisees who have less information. Like Cain et al., I found little or no support for effects on advisees, such as advisees giving higher estimates, being less accurate, or discounting more, with conflict-of-interest disclosure relative to nondisclosure. However, a meta-analysis revealed that across the original study and this replication, advisees were financially worse off with disclosure than without. I discuss when conflict-of-interest disclosure can lead advisors to give more or less biased advice.