BETA Podcast: When conflict of interest disclosures backfire

Dave:

Hello listeners. Thanks for joining us for another BETA Podcast. I'm your host, Dave. This episode, we're really lucky to welcome one of the world's experts in conflict of interest. Her name is Professor Sunita Sah and she's from Cornell University. She has an incredible body of work. She looks at how disclosing conflicts of interest affects people's decisions, which is particularly relevant in Australia at the moment. Sunita is joined by Janine Bialecki. She heads up BETA's behaviour advice team. Sit back and enjoy.

Professors Sunita Sah, Janine Bialecki, welcome to BETA's podcast. We'll be talking today a little bit about conflict of interest and disclosure. Sunita, first it might be good to just get a brief intro of who you are, where you've come from, what kind of work you do?.

Sunita:

Sure. I'm delighted to be here. As you said I'm Sunita Sah. I'm from Cornell University in the United States. I'm an assistant professor of management and organisations, and most of my research has looked... examines advice, influence, and ethics. Much of it has looked at conflicts of interest in professional advisors like financial advisors or physicians, and policies used to manage conflicts of interests like second opinions and particularly disclosure.

Dave:

Cool. And BETA's very own Janine Bialecki, are you able to give us a quick introduction as well?

Janine:

Hi Dave. Thanks. Yeah, my name's Janine and I work in BETA's behaviour advice team. My background is in economics. I worked at the Treasury Department for a long time before I went and did my master's in behavioural economics. Realised that was a pretty fun and interesting thing to do and now here I am at BETA.

Dave:

Perfect. Well, I can launch straight in. Sunita, essentially we seek advice all the time, but where things get a little bit complicated with conflicts of interests when people have a financial interest in giving you a particular recommendation, people get a little bit worried and often disclosures about that conflict of interest is seen as the way forward. Are you able to give us just a quick definition of how these things work in society, and what the concerns are that people have?

Sunita:

Conflict of interest is when there is a potential clash between a professional responsibility. For example, putting your client first or if you're a physician, putting your patient first, and a personal self interest. So if you have any incentive, usually financial but not necessarily so to give advice that might not put the patient first.

Dave:

And so, where do disclosures come into the picture there. They have become incredibly common, really popular way for people to try to counteract that I suppose. Where do they come into it?

Sunita:

Disclosure is one of the most commonly proposed and implemented solutions to dealing with conflicts of interest. This is across many different professions and many different industries. So way back many, many years ago, 2005, the American Medical Association stated that physicians who receive referral fees for referring patients to clinical trials should disclose that fee to the patient. Disclosure is really informing the advisee about the advisor's conflict of interest.

Dave:

From what I understand now, hearing a little bit more about your research, it's not always that simple, is it? What are some of the behavioural impacts that come into play? How do people actually respond in the real world to disclosures?

Sunita:

Well, disclosure is really popular because at least in theory it provides us potentially useful information. It's decreasing that information gap between the advisor and the advisees so that at least in theory the advisee can make a more informed choice.

However, in reality there's a number of unintended consequences that can occur on both the advisee who's receiving the disclosure and also on advisors who have to make the disclosure.

Janine:

Can I say quite at this point, it's really interesting the way that conflicts play out in a very biological way. I don't know whether you know this paper and they asked people to look at two pieces of art and people were told that they were going to be paid by a gallery. And the logo of the gallery was on one of the pieces of art, and people said that they liked that painting better. When they put them through an MRI machine, they found that looking at the painting of the gallery that was going to pay you was linked with, close to the pleasure centre of the brain. Is that right?

Sunita:

That's right. That's a really fascinating paper and it shows us how these conflicts of interests can influence us often in ways that we're not aware of or that are completely unintentional. So we can end up really believing that the advice that we're giving is the completely best advice for the patient and the client once we're incentivized to give particular recommendations.

Janine:

Yeah, I think the interesting conclusion from that paper is the participants were vehement in saying that they weren't influenced by who was paying them, but the researchers could see it in the brain that they actually were?

Sunita:

Exactly. And most of the time when I look at... When I ask the professional advisers that take part in some of my experiments, they're also the same. They're offended that you could possibly think that they might be giving biased advice because they really do believe that they are doing what's best for their client.

Dave:

So, it's happening on a biological level, which I guess is a little bit scary when people don't even recognise when they're being biassed. Sunita, what are some of the variables that come into this? I know that if it's an expert versus a non-expert giving this advice that it can have a different effect. What are some of the kind of variables that come into this that your research has shown?

Sunita:

Sure. I can talk a little first of all about the effect of conflict of interest disclosure on the behaviour of advisors, particularly since you mentioned the experts versus non-experts. So, much of the research that has been previously done on conflict of interest disclosure on advisors has looked at what I call non-experts. So research participants just playing the role of advisor. What we see in that research is that often advisors tend to give even more biased advice with disclosure than without, and of course this is really quite worrying.

But for experts such as real practising physicians, and financial advisors, what we found was something a little bit different. It was that when these expert advisors have to disclose their conflicts of interest, they actually gave better quality advice. And what seems to be triggering this is that the disclosure itself serves as a reminder for the norm of the context in which the advice has been given, the norm to either put clients first or to put self interest first. Professional advisors whether they're physicians or financial advisors, both actually in my experiments put the advisee first. And so, they gave better quality advice with disclosure than without.

What we think was going on with non-experts is that we were imagining being a financial advisor or a real estate agent and our descriptive stereotypes of these types of advisors made us feel that it was appropriate to put self interest first. So disclosure really serves as a reminder of the professional norm.

Dave:

Interesting. So, it's kind of what we expect of a particular sector versus what actual professionals expect of each other and themselves?

Sunita:

Exactly.

Dave:

It can be quite beneficial for advisors then, which is quite interesting.

Sunita:

It could be beneficial. It very much depends on the norms of the context, but it could be beneficial if the norms are that we are supposed to put advisees first and not self-interest.

Dave:

Interesting. What about the impacts on advisees on consumers? Where does that come in?

Sunita:

There's a couple of unintended consequences on advisees that I have documented. One is what we call the burden of disclosure effect, and this means that even if disclosure leads to less trust in the advice, which arguably could be the intended purpose of disclosure, is to alert someone to some uncertainty as to the quality of advice, it could also lead to increased pressure to comply with that advice.

This could be due to two psychological mechanisms that we've written about. One is insinuation anxiety, which is that even though we feel less trust in the advice, we don't want to signal that distrust to the advisor. And so, we go along with the advice to avoid having to insinuate that you don't trust the advice and you think your advisor is biased.

The second one is what we call the panhandler effect, which is even though disclosure is intended to communicate the idea that I will receive more if you take option X rather than Y, what the consumer might hear is, please take X because it will benefit me. Disclosure becomes almost like an implicit favour request. These mechanisms increase our pressure to comply with advice that comes with disclosure.

Dave:

It's almost playing on people's social awkwardness, which I would definitely [crosstalk 00:09:21].

Sunita:

Exactly, very much so.

Janine:

It's also interesting the environment in which this happens because typically if a consumer is asking for advice, it's a situation where they don't feel comfortable or they're simply not able to make a judgement themselves. And so, they're in this situation where they've asked for advice, there's been a disclosure, but they're often not in a position to do anything else. I guess in some circumstances you can get a second opinion. But if I've taken my car to the mechanic say, and they've told me they'll get money from a certain oil company, it's quite a bit of hassle for me to go take it to another mechanic. And actually I don't really know what I'm doing with my car, so I'm probably just going to do it anyway.

Sunita:

A lot of the decisions that we make are really because we want to maintain harmonious relationships. And if you hear a disclosure, even though you might like the advice less than if you hadn't heard it, you still feel incredibly uncomfortable to turn down that advice. In some cases even more so with disclosure than without, so you could end up complying more.

Janine:

Yeah. And often you get that disclosure at the point where if it's a doctor they're giving you the medical advice then and there, and you have to make a decision. Or you've taken your car to the mechanic, do I get this thing done to my car? You have to make the decision. As you say, it's kind of awkward to say, "I don't trust you. I'm not sure whether you're doing the right thing by me. I might go over there and seek other advice."

Sunita:

Exactly. And that's a really good point. When you do have to make the decision in front of your advisor, you're more likely to comply. But one way to reduce the pressure would be if you could make your decision in private, or at least get a cooling off period. You could go away, think about it. And if you really had felt pressured, you're probably going to change your mind, then hopefully you don't need to go back to the same advisor to implement it. Because if you do, that's going to diminish that protective effect a little bit.

Janine:

And you've done some work, have you... You've studied the cooling off period, and how that plays out, and you did find different results. Is that right?

Sunita:

Yeah. What we found with the cooling off period is that people were more likely to comply when they made the decision in front of the advisor. Once the advisor had gone and they had this opportunity to change their mind, a lot of them did change their mind, which really shows that their private decision that they want to make is very different from the public decision that they were displaying earlier.

Janine:

I think that's really interesting because in other contexts, I think behavioural economics has found that calling off periods maybe don't do that much. Say you've bought a product and there's... When I bought my mattress there was sort of a 90 day return policy, but I think they probably don't get many people returning the mattresses. And so, it's a great advertising angle, and they look like they're doing something great. But once the consumer buys the mattress they feel a sense of ownership and it's hard to return. It's interesting that cooling off periods work quite differently in an advice context

Sunita:

I think the difference there is that you haven't been given the products already, right? So, it has not already been implemented because once it has been implemented then you're going against the default, which we know how powerful defaults are. And so, in this situation, if it's easy for someone to change their mind without a cost and without repercussion. Without the hassle of having to return anything, then we saw that they did.

Janine:

Great.

Dave:

And so, obviously this kind of delayed decision can be really effective. What other ways can we help protect consumers and make disclosures a bit more effective for people? Are there any other things that we can look at doing?

Sunita:

Well, one aspect is allowing sufficient time to deliberate on the conflict of interest disclosure. Actually this is related to what the second unintended consequences on consumers. The first one was that even though you trust the advice less you feel increased pressure to comply. The second one is that sometimes if you don't deliberate on the disclosure, the disclosure could be used as a cue to the advisor's expertise. And so, you actually end up trusting the advice more and complying with the advice more with disclosure.

We want to allow people adequate time and resources and ability to actually deliberate on the disclosure because this unintended consequence increases when they process the disclosure automatically?

Dave:

It's counter intuitive, isn't it? That actually this could lead to increased trust because they're not keeping anything hidden [crosstalk 00:13:46].

Sunita:

Absolutely.

Dave:

I was really, really fascinated to hear about when you did a study on blogging, and you looked at disclosures up front and the different approach people take to processing that information.

Sunita:

This is also related to deliberation. So, if you put the disclosure at the front, at the top blog, you hopefully have people deliberate on the disclosure a little bit more than at the end of the blog. Some of the results suggested that people do do that, but the results weren't as strong as what we hoped. And so, actually asking people to deliberate on the disclosure had a much larger effect.

We also know that some people are more likely to deliberate than other people. And for those people that are more likely to deliberate disclosure you don't see what we call this expertise cue, this increased trust in the bloggers expertise as much. So that gives us some hope that if we can encourage recipients to deliberate on the disclosure, give them the disclosure separate from other information that might compete for their attention, that they would infer potential bias. And not have this unusual effect of perceiving increased expertise in the blogger.

Janine:

I think it's a really interesting avenue of research because behavioural economics, and a lot of the work we've done at BETA has been about simplifying information for consumers. I think because disclosure has been the remedy to everything, and lawyers have gotten involved. We often find when consumers are trying to engage in markets these long disclosures that are just difficult for consumers to comprehend.

And so, behavioural economics has done a lot in simplifying those disclosures and in helping consumers who want to make a quick to conflict decision to feel confident about that decision. But yours is going down the other avenue of slowing people down to make a better decision.

Sunita:

Slowing people down and giving that information separate to the other information. I'm not talking about having more disclosures here because a lot of time the disclosures are not helpful if people don't have a choice. So if there is a choice that's available to people. For example, most of my work has looked at conflict of interest disclosures. If people can see a list of advisors, and they can tell which advisors have a conflict of interest or not. It's much easier for them to choose and have that disclosure upfront in a very consumer user-friendly fashion such as a grade rating from A to F or a five star rating. That would simplify the disclosure and it would also give the information upfront and for them to choose.

Janine:

Yeah, I think that's a really interesting idea and certainly worth pursuing. It might work in practise from two perspectives because it might work for consumers who want to engage in the market, they can see the rank. But also, there was another study, I think it was in India where they published firms, how much or how little they were polluting. They made that public, and they found without any kind of consumer response, the firms at the bottom of the list suddenly knew they were at the bottom of the list, and it sort of became a little bit of a race to the top. They upped their game.

Sunita:

Yes, yes. We've seen that in many different environments and it's fantastic. It's what I call the self calibrating nature of disclosure, where we've seen it with restaurant hygiene ratings in Los Angeles where as soon as they were introduced, restaurants improve the quality of their service so they could get a higher grade. I think that's a wonderful way for disclosure to affect the behaviour of advisors. So, even the naive consumers can be protected.

Janine:

Yeah, I think it's an area that has a lot of promise.

Sunita:

Yeah. Absolutely.

Dave:  

Yeah. We've seen it with Britain's Ofsted system of school rankings and also hospital ratings more recently that actually kind of... It's a helpful procedure to help people understand where they seat in comparison to their peers.

Where do we go from here? So obviously, we've seen lots of instances where the incentives are wrong and where people giving advice aren't necessarily doing in people's best interests. How can policymakers, and even people, and consumers navigate making these complex decisions? How can we make it easy for people?

Sunita:

Well, ultimately the goal would be to eliminate the conflict of interest. So rather than having a bunch of policies designed to work around the conflict of interest or try to manage the conflict of interest, I think people have gone as far as to say, "Disclosure actually doesn't manage the conflict of interest." And if we really are concerned about it, what are other solutions that could be put into place?

One, of course, if it's possible, would be to realign the incentives to eliminate the conflict of interest and perhaps even eliminate the need for disclosure unless it's just disclosing the absence of a conflict of interest, which has been shown to be very positive. People do like to hear about the absence of conflicts of interest.

Dave:  

Interesting. Great. Well, thanks very much for coming on the podcast guys. Much appreciated.

Janine:

Thanks Dave

Sunita:

Thank you.

Dave:  

That's it for another episode. I hope you enjoyed hearing about Sunita's work. Don't forget to follow us on Twitter and visit our website to read a bit more about what we do. Until next time, thanks for listening.