PITTSBURGH--Professionals, such as doctors, lawyers and financial advisers, face conflicts of interest (COIs) when they have a personal, and often financial, interest in giving biased advice. Therefore, requiring COI disclosure has become a popular way to try and protect consumers from biased advice, but previous research has shown that mandatory disclosures have little impact on advice recipients, and may even lead advisers to give more biased advice. However, virtually all of the prior studies questioned the effectiveness of COI disclosures that advisers were unable to avoid.
New research from Carnegie Mellon University's George Loewenstein and Georgetown University's Sunita Sah examines situations in which advisers have the ability to not have any COIs -- such as doctors who can decide whether to meet with and accept gifts from pharmaceutical companies. Published in Psychological Science, Loewenstein and Sah found that when COIs can be avoided, disclosure successfully deters advisers from accepting COIs so that they have nothing to disclose except the absence of conflicts.
"Prior research has cast doubt as to the effectiveness of disclosure for managing conflicts of interest, particularly when consumers have the burden of interpreting and reacting to the information," said Sah, lead author of the study and assistant professor of strategy, economics, ethics and public policy at Georgetown. "Our findings suggest that disclosure can become a successful intervention to managing some conflicts of interest if it motivates professionals or providers to avoid such conflicts. Stating that you have no conflicts of interest or 'nothing to declare' can be a signal of quality or integrity to regulators and consumers, which could be viewed as a competitive advantage. The ability for disclosure to change the behavior of providers in a positive way is also advantageous as it avoids relying on consumers to make use of disclosure information."
For the study, the researchers conducted three experiments to determine how COIs influence advisers. In the first experiment, 97 adviser-advisee pairs participated in an online game with Amazon.com gift cards at stake. Advisers advised the advisees on the number of filled dots on a grid. Estimators were paid based on their accuracy, but advisers had a conflict; they were paid more if advisees gave an estimate that was higher than the true value. The set up -- with advisees only seeing a small subset of the complete grid -- was designed to simulate a situation in which a consumer receives advice from a better informed, but conflicted professional. The results replicated previous research and showed that disclosure led advisers to give higher (and more biased) recommendations than nondisclosure.
In the second experiment, the researchers again randomly assigned pairs of advisees and conflicted advisers to conditions in which the conflict was either disclosed or not disclosed. There was, however, an important change from the first study: advisers were given a choice of whether to accept or reject the COI. Without disclosure, a majority of advisers (63 percent) chose the incentives that created a COI, but with disclosure a minority (33 percent) accepted the conflict. Advice was higher (and more biased) for those who chose conflicted incentives than for those who did not, and advisers in the disclosure condition gave significantly less biased advice than those in the nondisclosure condition. When advisers could eschew a conflict therefore, disclosure encouraged them to do so.
Finally, in a study with 248 participants, the researchers added a third condition to the second experiment: voluntary disclosure. In this third condition advisers decided both whether to choose incentives that entailed a conflict and whether to disclose whether they were conflicted. Similar to mandatory disclosure, voluntary disclosure led advisers to avoid COIs, and then disclosed their freedom from conflicts to advisees.
"Disclosure doesn't seem to be much good when conflicts are unavoidable, but it does seem to help when advisers have a choice about whether to subject themselves to conflicts," said Loewenstein, the Herbert A. Simon University Professor of Economics and Psychology in the Dietrich College of Humanities and Social Sciences. "A nice feature of disclosure is that it is, in effect, 'self-calibrating.' Doctors, for example, are unlikely to find it worth it to accept small gifts such as pens or calendars if the gifts are going to be disclosed. Although larger gifts would be more tempting, doctors are likely to be deterred from accepting them because disclosure of large gifts would be more damaging to their reputations."