Exclusive Interview with Yale’s Daylian Cain: Just a Sugar Pill? Disclosure’s “Ah-Ha!” Moment
On September 24th, 2010, The Committee for the Fiduciary Standard held the Fiduciary Forum. Knut A. Rostad, Chairman, The Committee for the Fiduciary Standard said “[Securities and Exchange Commission] Chairman Schapiro has stressed the investor protection mission of the SEC in her speeches and actions; the Fiduciary Forum 2010 focused attention on the very dramatic limitations of investors and consumers, raising huge questions as to the effectiveness of disclosures as a core investor protection tool. It was disclosures’ ‘ah ha’ moment.”
The star of the Forum may well have been Daylian M. Cain, Assistant Professor of Organizational Behavior at the Yale School of Management. Cain’s groundbreaking research reveals disclosure may not help – and may actually hurt – investors, including ERISA plan sponsors. While he likes the idea of transparency, he feels disclosure can only become more effective when all parties understand its shortcoming. Fiduciary News was fortunate enough to obtain an exclusive interview with Cain.
FN: What first inspired you to look into the possibility disclosure might actually do more harm than good?
Cain: Back in the late 90’s, I was sitting in on a Bioethics class, taught by Chris MacDonald (St. Mary’s U.) in Nova Scotia, and we were discussing conflicts of interest. A classmate brought up disclosure as a remedy. It occurred to me that a disclosure by a doctor (e.g., he had some sort of financial ties, say, to the pharmaceutical industry) would be difficult for a patient to process; after all, should the doctor be trusted less because of a conflict of interest, or trusted more because of strong relations to research? And, I worried that, even if a patient should be inclined to get a second opinion, doing so may involve taking another day off of work, and (in those remote Maritime towns) perhaps another 4-hour drive to see a second specialist. So, I wondered how alerting a disclosure would have to be before it inspired the patient to go through the hardship of getting a second opinion. The day I landed at Carnegie Mellon, where I did my PhD, I began researching these ideas with Don Moore (now at Berkeley) and George Loewenstein (CMU), who were working on conflicts of interest. We knew from Kahneman & Tversky’s work on “Anchoring” (the process by which mental starting points “stick” to affect subsequent judgment) that even what the audience knows to be randomly generated suggestions can affect their judgment; if a random suggestion can affect judgment, then I worried that a mildly discredited suggestion would also likely affect judgment. So, in the beginning, I was merely worried that disclosure would fail to be of sufficient help. It turns out to be potentially worse than that… We have found that disclosure can have perverse effects – making advice worse, creating pressure to follow the advice (the “panhandler effect”), etc… Although in general I think disclosure is probably a good thing, just not a great thing, I still worry that the worst effect of disclosure is that it causes regulators to feel the problems of conflicts of interest are solved.
FN: Elaborate on the “pandhandler effect” you refer to – where might one see it and what how can one avoid it?
Cain: This was inspired by work I do with Jason Dana (U. Penn) on charitable giving, and was researched by Sunita Sah (Duke), Loewenstein and myself. The idea is that people often feel pressured to give to someone who asks, like when a panhandler asks for money. When an advisor leans forward and informs the client he has a financial stake in the client’s decision, it puts pressure on the client to comply with that advice because it is now common knowledge that the advisor is helped by compliance. So, clients may trust the advice less but will feel more pressure to comply. We find this effect to be strongest in one-on-one social situations, where it is clear that the advisor will see how the client responds to the advice. We are currently researching the boundary conditions for when this effect is strong enough to sway behavior and when it is not.
FN: Your paper “The Dirt on Coming Clean: The Perverse Effects of Disclosing Conflicts of Interest” (with G. Loewenstein and D. A. Moore, Journal of Legal Studies, January, 1-25, 2005) suggests disclosure may work with sophisticated estimators, but not unsophisticated estimators. Has later research changed this?
Cain: We have tried several rounds of feedback, and disclosure still failed. Koch and Schmidt (U. Mannheim) found that MANY rounds of feedback do help clients; Church and Kuang (Georgia Tech) look at how sanctions for giving biased advice can help disclosure work better. This later research suggests hope for disclosure, especially for savvy, repeat players, like institutional investors. But, I think the jury is still out. First, while experience may help clients, it can also help advisors who are trying to manipulate them. And, anchoring (i.e., a biased suggestion impacting subsequent judgment) has been shown to be extremely robust to experience. That said, I am hopeful that we can learn how to make disclosure work, and experience with it is surely needed. Perhaps once people understand how big a problem conflicts of interest are – and that they are a problem even for well-meaning professionals who think they are being objective but who are being swayed by their own interests (something echoed in Bazerman, Loewenstein and Moore’s, “Why Good Accountants Do Bad Audits”) – then disclosure of such conflicts might be more properly alarming. We have found that disclosures are not alarming enough, even after being burned by a conflict once or twice. And, after all, some of our biggest decisions are made but a few times in our lives. I am not against disclosure, and I am anxious to learn more about how to make it more effective. But right now, disclosure is often a boiler-plate after thought, printed in fine-print legalese, not the sort of alarm-bell regulators assume it to be.
FN: You just said “while experience may help clients, it can also help advisors what are trying to manipulate them.” Give us an example of how this might help advisors manipulate clients. How can clients avoid this manipulation?
Cain: If clients learn to discount information more, advisors can learn to give even more manipulative advice. Now, one might think that biased advisors are already trying to be as manipulative as possible, but we don’t find this to be the case. We find that even biased advisors feel a moral obligation to keep an eye on the truth. But, the less truth the client expects, the greater moral license the advisor feels to give increasingly biased advice. That said, I don’t think experience is a bad thing. I think that once people learn (perhaps through experience, or perhaps through exposure to the research like this) that conflicts of interest are problems even for well-meaning advisors, then maybe disclosures will become sufficient warning devices. In fact, I am all for transparency, disclosure, experience, and such things. I just doubt that these things alone will solve the problem. Years of research from social science can tell us that once we let biased advice out of the bag, like an untrue but scandalous rumor, it is difficult to undo its effects.
FN: In the “Dirt” paper, it appears the conclusion is that the estimator’s (i.e., the client) interests are best served in the case where the adviser has no conflict-of-interest. In practical terms, this would appear to suggest, rather than relying on disclosure, it may be more effective to simply remove the broker exclusion from the Investment Adviser Act. Is there anything in the subsequent papers that leads to a different conclusion?
Cain: To date, I still believe that the best solution to conflicts of interest is to avoid them, at least with more vigor than we currently apply. Christopher Robertson (Harvard) has also been writing on this: We need to encourage the production of unbiased advice. In a forthcoming paper with Loewenstein and Moore (Journal of Consumer Research), we show that the mere availability of unbiased advice is not enough, since clients often are not sufficiently motivated to seek it out when disclosures are made about the advice the clients are currently receiving. So, we may need to put unbiased advice alongside conflicted advice. Of course, unbiased advice may be hard to come by, and some conflicts of interest may be intractable, but I insist that we take a harder look at these conflicts.
FN: You mention practical problems with a “cooling off” period (giving the client some time after signing the paperwork before the service is actually delivered) in a recent review paper with Sah and Loewenstein. For example, this “cooling off” period might cause the client to miss an investment opportunity. Do you still have those reservations?
Cain: I do not have research-based problems with a cooling off period. I just wonder about the full ramifications of such a hindrance to a fast transaction. There are costs and benefits that I am not in a position to weigh. Our research is too preliminary on this. Again, regulators will have to weigh the pros and cons of any change in legislation; I just want them to go in with a more skeptical eye when it comes to disclosure.
FN: What have you seen in the financial industry that makes you believe they understand and appreciate your research on disclosure?
Cain: Certainly the regulators are more sensitive to the idea that disclosure may not be a panacea, and there is lots of discussion around this very issue, especially since the last financial crisis. The regulated (doctors, big pharma, wall street, etc.) may prefer disclosure, rather than sever the ties that line their pockets; but most people are coming around to the idea that someone need not be intentionally corrupt to have difficulty in objectively navigating a conflict of interest.
FN: What have you seen in among regulators that would lead you to believe they understand and appreciate your research on disclosure?
Cain: More and more, regulators are asking, “So, what should we do, then?” Well, for the last 5 years, my colleagues and I have been saying that the current treatment (disclosure) is not working. And, if all I do is inspire a sincere look at the problem, I will consider my job well done. But, of course, I appreciate the “what then?” question; it is what everyone, myself included, is working on. What I can say is that, in the past, we thought disclosure worked better than it did, we thought the clients could assimilate disclosure better than they could, and we assumed that conflicts of interest were only problems for the intentionally corrupt. Now, regulators are revisiting these issues with a more skeptical lens. We need to be more aggressive in eliminating conflicts of interest, more wary of those who have conflicts of interest, more energetic about getting second opinions, and more on the look-out for unbiased advice.
FN: What is the best possible solution – in your mind – that is consistent with your research?
Cain: I think of disclosure as half medicine, half sugar pill. If we throw a sugar pill at the cancer that is conflicts of interest, it may not cause direct harm (although, in some cases, it may), but we could be doing harm if that “treatment” replaces more effective measures. Sure, clients have the right to know what their advisors are up to. But even if disclosure does not make matters worse, disclosure often fails to correct from the biasing forces of conflicts of interest. So, what then? We should aim to eliminate conflicts of interest wherever feasible. Easier said than done. But, I say to regulators: Treat the world as if conflicts of interest are more dangerous than you thought, harder for even well-meaning professionals to objectively navigate, and treat disclosure as if it does not work as well as you’d like; then go from there. With Sah and Loewenstein, we are examining how to make disclosure work: cooling off periods, different types of disclosures, providing unbiased advice, etc. The research is still at the early stages. Certainly, we need to start giving hard consideration of measures to encourage getting unbiased advice out there. As said in “The Talking Cure”, (Talk of the Town Column, James Surowiecki, The New Yorker, December 6, 2002), “Transparency is well and good, but accuracy and objectivity are even better. Wall Street doesn’t have to keep confessing its sins. It just has to stop committing them.”
FN: Thank you, Daylian. It’s been a delight to talk to you and learn from you. I’m sure our readers will agree.