Job Hunting

Navigating the Bias Trap

An exclusive interview with Don Moore of Carnegie Mellon offers insights into why good accountants sometimes do bad things.
Lisa YoonNovember 12, 2002

For evidence that corporate scandals have cast accountants as nefarious in the general press, flip open an issue of The New Yorker, where chances are pretty good that the page selected will feature a cartoon depicting accountants as villains. Here’s one: Two Puritan women wearing disapproving expressions are talking about a third, who shamefully wears a large Scarlet Letter-type “A” on her dress. The caption reads simply “Accountant.”

But in this month’s edition of the Harvard Business Review, three scholars suggest that it’s not the corrupt accountant with a thirst for fraud that companies need to worry about. Rather, the real danger is unconscious bias that occurs regularly in accountants who have every intention of doing things right.

In the article “Why Good Accountants Do Bad Audits,” Harvard Business School professor Max Bazerman and Carnegie Mellon University’s George Loewenstein and Don Moore, point out that accountants exhibit biases that are in line with their own objectives. In one study, which simulated the sale of a company, students playing the role of the companies’ auditors consistently favored their clients.

Despite being offered a reward for calculating the most accurate valuation of the seller, the auditors representing the seller valued the company 30 percent higher than other study groups. Imagine, the authors say, the bias that occurs when real companies with real money are at stake.

Such unintended bias is a problem for CFOs, who under the new Sarbanes-Oxley law, must certify company financials. So what’s a finance chief to do? Co-author Don Moore, assistant professor of organizational behavior and theory at Carnegie Mellon’s Graduate School for Industrial Administration, recently talked to about the psychology of bias and what CFOs should do about it. How does unconscious bias work in the context of auditing?

Don Moore: Research—ours and lots of other psychological studies on motivated reasoning—shows that when people to want to arrive at a certain conclusion, it inevitably colors the way they acquire information. Whether it’s “The company I am auditing is fairly representing its financial condition,” or “My child is actually the smartest kid in class”—you can’t analyze information objectively when your heart really wants it to come out a certain way.

You’re biased in the way you review arguments; you’re biased in the way you consider why those arguments might be wrong. In fact, you tend not to consider why they might be wrong if it’s inconsistent with your view. But if the argument is consistent with what you want to believe, then you tend not to subject it to more scrutiny.

It’s not that [accountants] can’t help [performing questionable audits]; it’s that they aren’t aware that they’re doing it. So it’s an oversimplification to say, “This fraud is committed by bad people trying to cheat shareholders.” It just may be that their incentives are such that they would really like to come to a positive audit conclusion. As the system stands now, outside auditors have too many incentives that make them want to serve the interests of management. Isn’t that what proponents of tougher auditor-independence rules have been arguing all along — that auditors’ relationships with their clients are tainted by conflict of interest?

Moore: Yes. But [with most accountants], it’s not that they’re thinking of themselves as in cahoots with management; they’re not weighing the dangers of jail if they’re convicted of fraud with the potential benefits of the additional business they could get by issuing positive audit reports. Instead, it’s that they know where their interests lie, and it’s those preferences that drive the way they think about information concerning the company books. In your research summary, you write that unless people are made aware of their vulnerability to unconscious bias, they’re not going to be able to undo its effects. In the HBR article you and your co-authors also say that because of this, stricter accounting rules aren’t the answer to stemming questionable accounting. Yet, don’t stricter accounting rules make accountants aware of the potential for faulty accounting?

Moore: Well, we would have to be more specific about which audit rules in particular we’re talking about. But if the question is making them aware of their own biases, the research is quite pessimistic.

There was a study my co-author George Loewenstein was involved in, in which subjects were assigned a negotiation exercise. Before they started, the participants were given a little sermon about the danger of unconscious bias. They were told that people tend to be biased in their judgments of fairness; that is, they tend to think what favors them is more fair. Therefore, the subjects were told, be careful and avoid this in your negotiation. The participants became very suspicious of others, but they didn’t feel that they themselves were subject to bias. It’s very difficult to undo bias. If the incentives of external auditors make them prone make to biased judgments, are a company’s internal auditors are even more vulnerable to bias? Independent of legislative or regulatory intervention, what can CFOs do to get the most accurate books from their finance departments?

Moore: That’s something we didn’t get to address in the article, but evidence from research shows that when a group on the internal audit committee is motivated to reach a conclusion—”our company is in good shape,” for example—they are motivated to analyze or search for facts in a different way from when you are motivated to search for the truth.

One strategy for getting around those biases in group decision-making is to appoint a devil’s advocate—someone in internal audit whose job is basically to bring “lawsuits.” These employees would ask: If someone wanted to charge fraud, if someone wanted to paint our financial situation more pessimistically than we would like to, what information could they draw on? Where are we cutting corners, pushing the envelope of GAAP? One useful way for the CFO to extract information that is free of company biases is to appoint devil’s advocates, whose job it is to find those loopholes, and to report them, and get rewarded for ferreting out internal inconsistencies, or people who are pushing the envelope too far. So companies should appoint their own whistle-blowers?

Moore: Exactly. Except that whistle-blowers announce these problems to outside authorities. A fully informed CEO wants to have that information from inside the company. They don’t want to read about it first in the press. What are some possible solutions to accounting malfeasance in the long term?

True auditor independence will have to come from reform, much the same way outside auditors are appointed and changing the way shareholders deal with management on these issues. For example, one possibility is to give the power of appointing auditors to shareholders and not the management.

More realistic, but very expensive, is to eliminate those biases, so that auditors don’t really want to come to a conclusion that favors the client firm. Or, much more problematic and probably impossible, is to untie the CFO from the decision that says “our company is in great shape.” Instead, the CFO could have incentives that make him or her report honestly the condition of the firm.