Disclosures About Disclosures: Can Conflict of Interest Warnings be Made More Effective?

Published date01 June 2015
DOIhttp://doi.org/10.1111/jels.12071
Date01 June 2015
AuthorAhmed E. Taha,John V. Petrocelli
Disclosures About Disclosures: Can
ConïŹ‚ict of Interest Warnings be Made
More Effective?
Ahmed E. Taha and John V. Petrocelli*
People regularly rely on advisors who have conïŹ‚icts of interest. The law often requires
advisors to disclose these conïŹ‚icts. Despite these disclosures, people generally insufïŹciently
discount conïŹ‚icted advice. This might be partly due to people interpreting the very fact that
the advisor is disclosing a conïŹ‚ict of interest as a sign that the advisor is trustworthy,
undermining the purpose and effectiveness of the disclosure. This article presents the results
of an experiment indicating that requiring advisors to also disclose that they are legally
required to disclose their conïŹ‚ict of interest makes people discount their advice more. This
occurs, at least in part, because such advisors are viewed as less trustworthy than advisors who
merely disclose their conïŹ‚ict of interest without also stating that the disclosure is legally
required.
I. Introduction
People regularly rely on advisors who have conïŹ‚icts of interest. The most common policy
approach to this problem is to require the advisors to disclose these conïŹ‚icts. Despite these
disclosures, however, advisees still insufïŹciently discount advice given by advisors with
conïŹ‚icts of interest. Thus it is important to determine if the disclosures can be made more
effective.
Much of the rationale behind these disclosure requirements is that disclosure of the
advisor’s conïŹ‚ict will make advisees more skeptical of the advice. However, the very act of
disclosure might be interpreted by some advisees as an act of honesty. In other words, they
might view the fact that the advisor is disclosing the conïŹ‚ict of interest as evidence that
the advisor is trustworthy. This could undermine the purpose and effectiveness of the
disclosure.
This article presents a controlled experiment investigating whether requiring man-
dated disclosures of conïŹ‚icts of interest to also state that the disclosures are legally required
*Address correspondence to Ahmed Taha, Pepperdine University School of Law, 24255 PaciïŹc Coast Hwy., Malibu,
CA 90263; e-mail: ahmed.taha@pepperdine.edu. Taha is Professor of Law, Pepperdine University; Petrocelli is
Associate Professor of Psychology, Wake Forest University.
For helpful comments and suggestions, the authors thank Robert Anderson, Omri Ben-Shahar, Jonathan Cardi,
Michael Green, Mark Hall, Molly Mercer, Gregory Parks, Matthew Phillips, Mark Scarberry, Sunita Sah, Ronald
Wright, and the anonymous referees. The authors are also grateful to Lindsey Chessum and Alex Rothschild for
excellent research assistance.
bs_bs_banner
Journal of Empirical Legal Studies
Volume 12, Issue 2, 236–251, June 2015
236
would make them more effective. It tests whether this information would make people less
likely to attribute the disclosure to the advisor’s trustworthiness, and thus less trusting of the
advisor.
Participants in the experiment read a version of an email from an independent
insurance agent (the advisor) discussing two identical auto insurance policies, each from a
different insurance company. In the email, the agent recommended one of the policies,
allegedly because that insurance company provides better customer service. Versions of the
agent’s email differed in whether they also disclosed that the insurance agent receives a
higher commission on the recommended policy, and in whether they stated that this
disclosure was required by law. After reading the email, participants stated which policy they
would be more likely to buy, the strength of their preference, and how trustworthy they
consider the agent.
Consistent with prior studies, we ïŹnd evidence that advisees discount advisors’ advice
if the advisors disclose their conïŹ‚icts of interest. In addition, however, we ïŹnd that advisees
discount the advice even more if the disclosure also states that the disclosure is required by
law. This greater discounting appears to occur, at least in part, because advisors who
disclose that the disclosure is legally required are perceived as less trustworthy. This suggests
that requiring conïŹ‚ict of interest disclosures to also state that the disclosures are legally
required will make the disclosures less likely to be perceived as acts of trustworthiness, thus
making the disclosures more effective.
The next two sections of this article present brief background information. Section II
discusses why advisees generally insufïŹciently account for bias created by advisors’ conïŹ‚icts
of interest, even when the conïŹ‚icts are disclosed. Section III explains why requiring advisors
to state also that mandated disclosures of the conïŹ‚icts are legally required might be more
effective. Section IV presents the experiment that we used to test the effect of mandating
that advisors inform advisees that a disclosure of a conïŹ‚ict of interest is legally required.
Section V discusses the experiment’s ïŹndings and their implications for laws requiring
disclosure of conïŹ‚icts of interest.
II. People Insufficiently Discount Advice for Conflicts
of Interest
A conïŹ‚ict of interest exists when a person’s self-interest conïŹ‚icts with his or her professional
obligations. Advisors’ conïŹ‚icts of interest can skew the advice they give, possibly harming
those who naively follow the advice. For example, substantial evidence exists that sell-side
securities analysts who work for companies that also provide investment banking services
give overly optimistic recommendations of particular companies’ stock to help secure
investment banking business from those companies (see Fisch 2007). Similarly, medical
researchers with ïŹnancial relationships with pharmaceutical companies are more likely to
report research results favorable to those companies than are other medical researchers
(Angell 2000; DeAngelis 2011).
The most common policy approach to conïŹ‚icts of interest is to require their disclo-
sure (Cain et al. 2011). Disclosure’s popularity stems in part from its being less intrusive
Disclosures About Disclosures 237

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT