men exchanging money under the table

Full Disclosure: We're Conflicted

brandon_routman
Brandon Routman
Writer

EDITOR’S NOTE: As you may know, I wrote a book basically about conflicts of interest – “Get Rich Cheating.” I’m fascinated by them. Get someone’s trust and then suggest they buy a product or service you just happen to own? Brilliant! (Full disclosure: By “brilliant” I mean “horrible.”)  There’s much more to conflicts of interest and other behavioral science folk have discussed them in more useful ways, but those aren't as funny.

What is important – funny or not – is what we might do about them. As commercial webs have expanded grown more complex and the incentives behind our actions and recommendations become more hidden, how do we limit the damage of conflicts of interest?  What are some solutions?  Requiring disclosure of conflicts is one idea.  Does it work? Hmmm. Read this article that I just happen to have edited to find out…

 

Professionals often face decisions in which their personal self-interest is pitted against their professional responsibility. If you are a realtor, you might bag a larger commission if you convince a home buyer to purchase a bigger, expensive home rather than a smaller, more appropriate one. If you are a politician, you might be tempted to give preferential treatment to companies or industries that give you generous campaign contributions. If you are an attorney, you might advise your client to go to court when that is not necessarily in his best interest. (Editor’s note: If you’re an editor, you might publish an article based solely on how often a writer says, “No, that doesn’t make you look fat and also you’re funny.” Hint hint.)

One tool policy-makers have to try and discourage these outcomes is to require you to disclose your conflicting interests. The presumption is that this alleviates potential for abuse. You are less likely to give bad advice to a client when all relevant factors are disclosed and your client is better able to properly evaluate that advice, and perhaps ignore it, when he or she knows what might motivate your advice. Transparency is, if not the best disinfectant, certainly a good one.

Disclosing a conflict of interest also respects the freedom consumers have to make their own decisions and, in a broader sense, the operation of the free market. It reduces legal liability when problems arise. The client was warned of potential risks, Your Honor. And to some, it serves as a more politically palatable substitute for regulatory oversight. We don’t say you can’t do something, we just say you have to tell people you’re doing it (and let the free market decide!). Legislation like the Affordable Care Act and Dodd-Frank are chock-full of such requirements.

Consequently, disclosures have proliferated. Car dealerships are required to tell us about problems with their sell. Landlords have to fess up to their property’s code violations. Physicians have to reveal economic or research interests they have in performing medical procedures. Mandating disclosure has become so popular a tool, in fact, that its effects are sometimes ridiculous: If you sell caskets in California, don’t forget to warn potential buyers that “THERE IS NO SCIENTIFIC OR OTHER EVIDENCE THAT ANY CASKET WITH A SEALING DEVICE WILL PRESERVE HUMAN REMAINS.”

Yet surprisingly, the effects of disclosures are not particularly well understood. Cass Sunstein and George Lowenstein write that there is a “paucity of research that seeks to understand when, why and how [disclosure requirements] work.” Indeed, it is possible that the very ubiquity of disclosures undercuts their effectiveness. After learning about the inability of caskets to preserve human remains, you might just tune out disclosures altogether. Furthermore, even if you tried to pay attention to them, you’d have a hard time doing so. In many cases, you have to navigate obscure, legalistic language and wade through a deluge of irrelevant information. I’ll spare you an example.

In short, this is not a fair fight. Regular people can digest only so much information and have only so much patience to try. The result is predictable. We sign along the dotted line or check the box without knowing what we are consenting to. Of course, this subverts the purpose of mandating disclosure in the first place. (Editor’s note: See, related, “Terms of Disservice.”)

 

This is not a fair fight. Regular people can digest only so much information and have only so much patience to try.

Just for the sake of argument, though, let’s strip out these particular concerns and focus on how disclosures could – and perhaps should – work. How do people respond to a disclosed conflict of interest that is written clearly and concisely and in a context in which people pay focused attention? Two separate lab experiments look into this question and find less-than-positive results.

In the first experiment, one person, referred to as the estimator, tries to guess the value of a jar of coins, after looking at it briefly and from a distance. The closer his guess is to the actual value, the more money he earns. He is given a suggestion regarding its value by a second person, the advisor, who has examined the jar more closely. The advisor also makes a private estimate regarding the jar’s value which is not shared with the estimator.

This game plays out in three different conditions.

  1. In the first, there is no conflict of interest between the players. Both are paid according to how close the estimator’s guess is to the real value. In this case, the advisor suggests a value of $16.48, which is $.82 more than his private estimate.
  2. In the second scenario, there is a conflict of interest. The advisor is paid according to how high the estimator’s guess is. Here, the advisor suggests a value of $20.16, which is $3.32 more than his private estimate.
  3. In the third condition, the same conflict of interest exists and it’s publicly disclosed. How does this affect the advisor’s behavior? Does he give a more reasonable estimate since he knows the estimator will be suspicious of a highball number? Or does he give an even more biased estimate since he knows that the estimator will naturally counteract it? The answer seems to be latter. He suggests a value of $24.16, which is $7.10 more than his private guess.

What about the estimator’s response? His guesses are lowest in the first scenario ($14.21), higher in the second ($16.81), and highest in the third ($18.14). The good news is that he discounts the advice he is given when the conflict of interest is disclosed. But the bad news is that he does not discount it enough to compensate for the more inflated figure he receives. In sum, disclosing the conflict of interest lines the pocket of the advisor and depletes that of the estimator.

A second experiment adds more realism to this set-up. Here, an estimator is given a packet of information about a particular home and guesses its sales prices. Just as before, he is paid according to how accurate his guess is and is given a suggested value by an advisor who has extra information, like the tax-assessed value of the home.

  1. In the first condition, interests align. Both players are paid according to the accuracy of the estimator’s guess. Here, the advisor’s suggestion to the estimator is $204,000, which is about $1,000 more than his private estimate.
  2. In the second condition, the advisor is paid according to how high the estimator’s guess is and the advisor’s suggestion is $236,000, which is about $31,000 more than his private estimate.
  3. Finally, in the third condition, the same conflict of interest exists and it’s disclosed, and the advisor’s suggestion is $255,000, which is about $51,000 more than his personal estimate.

As before, the estimator’s guesses – based upon those suggestions – are lowest in the first condition ($202,000), higher in the second ($221,000), and highest in the third ($229,000). He discounts advice when the conflict of interest is disclosed, but not enough to make up for the fact that he receives worse advice because of this disclosure.

One caveat to remember is that these experiments take place in a lab. The stakes are small; players don’t have reputational concerns to uphold; they haven’t pledged to abide by set of professional or ethical guidelines; they aren’t fined or jailed for behaving badly; and they can’t seek second opinions. The real world is more complicated. Still, the studies suggest the worrying possibility that mandating a disclosure of a conflict of interest can do more harm than good – even in ideal conditions in which disclosure is clearly stated and is well understood.

 

Studies suggest the worrying possibility that mandating a disclosure of a conflict of interest can do more harm than good.

That doesn’t necessarily mean we should do away with such requirements. Given how deeply embedded they are in local, state and federal regulatory practice, doing away with them would be tremendously hard to do. Nonetheless, it is probably time for policy-makers to take a more thoughtful and nuanced approach to the subject to better serve their constituents. Disclosures never seemed intended to end conflicts of interest, just to limit them. Now, it seems, they may not even do that.

 

brandon_routman
Brandon Routman
Writer

Subscribe

Get the latest behavioral
science insights.