Unconscious Human Nature Affecting “Professional Skepticism”

Thank you for joining us today for our inaugural 2007 Forum on Auditing in the Small Business Environment, and our 19th Forum since we began this program in 2005.  We have learned so much during the past two years in which we have been “on the road,” talking to folks such as you.  We hope that this is a helpful use of your time; we know that is useful for us.  To make the next few hours together as productive as possible, I have a few requests.

  • Please participate.  We have built into our schedule plenty of time for you to ask questions, make comments, and offer suggestions.  All of your presenters today want you to interrupt them with timely questions; your participation will not only improve the quality of our discussion, it will make the entire day much more enjoyable for all of us.  Simply raise your hand so that we can get you a microphone or, if you prefer, write your comment or question on one of the comment cards in your materials.  Hand the card at any time to any of our staff, and we will address it at the next available opportunity.
  • Please turn off your cells phones and PDAs, or put them on silent mode. 
  • Please turn in your evaluation forms, before you leave today.  We rely on your evaluations and other comments to continually improve these forums.  In fact, our 2007 curriculum was crafted primarily based on input we received from prior Forum participants.  I thank you in advance for taking the time to help us.
  • We will mail you your CPE certificates, so as you sign out today please take a moment to make sure that your addresses are correct.

Before we begin, let’s take a rough survey of the audience. How many of you have been to prior PCAOB Forums?  [Response: approximately 40%.]  How many of you currently have 50 or more SEC “issuer”[1] clients?  [Response: approximately 10%.]  How many of you have over 20, but less than 50 such clients?  [Response: approximately 30%.]  How many of you have 20 or fewer issuer clients?  [Response: approximately 60%.]  How many of you are in the process of, or have completed, a PCAOB-inspection?  [Response: approximately 75%.]  Of those of you with your hands up, how many have been through, or are currently in, a 12-month remediation cycle?  [Response: approximately 20% of entire audience.]

As I hope that you can appreciate from this rough survey, you are a diverse audience.  You represent a wide spectrum of the auditing profession.  We have tailored today’s program to meet the needs of most of you.  Some of you may find some of our points too rudimentary; others may find them more challenging.  I ask that you all have patience with us, and understand that this is, indeed, a varied audience.

Before I turn the podium over to the real stars of the day, I’d like to share a few observations that I have accumulated from my 4-plus years on the PCAOB Board.  Let me start by explaining to you a little bit about my background because – as I will discuss shortly – my background colors my perspective.  First, I am an attorney and, like a majority of my colleagues on the Board, not a CPA.  (Insert your favorite attorney joke here!)  Second, I have spent most of my career advising and advocating for investors.  Prior to joining the Board, I was General Counsel for the California Public Employees’ Retirement System (also known as CalPERS), where I spent a total of 16 years.  Lastly, one of my roles at CalPERS involved what we called “corporate governance.”  As some of you may know, CalPERS is known as an “activist” investor – an investor that uses its rights as corporate owner to push for improved corporate performance.  While at CalPERS, we focused our governance attention on companies that we believed had demonstrated a lack of attention toward promoting long-term and sustainable shareholder value.  To put it another way most germane to today’s discussion, I was trained to be a little bit cynical about whether others are truly behaving in a way that most protects investor interests.

When I was first appointed by the Securities and Exchange Commission to sit on the PCAOB Board, this background contributed to several initial presumptions.  I thought that:

  1. Everything that I needed to be concerned about would happen within or surrounding the largest few accounting firms.
  2. The audit profession would “circle the wagons” to challenge the PCAOB at every opportunity.
  3. There were probably serious ethical issues within the profession that we at the PCAOB would need to “root out.”

We all know that initial presumptions are more likely than not to be wrong, and mine certainly were.  Before I related to you some of the conclusions that I have reached as a result of my tenure on the Board, I must remind you that the views that I express today are my own, and do not reflect the opinions of the PCAOB, my colleagues on the Board, or our staff.  I say this not as a disclaimer of any sort.  To the contrary, I own my words.  Please hold me accountable for them.  I simply ask that you not hold the PCAOB accountable for my views.

With experience on the Board, I have found that:

  1. The PCAOB can provide real, meaningful input to impact the quality of work performed by all audit firms – not just the Big 4.  In fact, it is essential that the Board do this, as investor interests can only be met if the market for audit services is freely and openly competitive.  This is one of the reasons why the PCAOB has put significant resources behind these small audit firm forums.
  2. Most of the audit profession not only graciously welcomed the PCAOB, but openly embraced us as we have jointly set about to restore public confidence in the profession.
  3. I have seen no evidence of serious ethical issues that could threaten the auditor credibility that we have all been working to repair.  However, clearly there are issues that the profession needs to address.  During the rest of the day we will be discussing important technical issues that affect auditor credibility.  But during my brief time with you, I want to talk to you about a different category of issues – issues affecting behavior.

Let me be clear: I am not talking about behaviors that manifest themselves in conscious collusion or corruption, or those intentional bad decisions and actions that rightfully fall within the “few bad apples” barrel.  Instead, I want to talk about human nature, and about those behaviors and attitudes that affect all of us –behaviors and attitudes that those of us in positions of trust must particularly and diligently be aware of and fight against.

Because we are human, we are not perfect.  Because we are not perfect, we will make errors.  But, because the cost of auditor error is so significant, we must try (within the limits of reasonable human capacity) to minimize the frequency, size and severity of errors.  To put an even finer point to it: I am speaking of those characteristics of human behavior that affect “professional skepticism”[2] – not because we want them to, but because we are usually not even aware of them.

Over two decades of psychological research has revealed at least one truism: even the most well-meaning person unintentionally allows unconscious thoughts and feelings to influence his or her seemingly objective decisions.  We all have what Yale Professor David Armor calls “the illusion of objectivity.”[3]  While we obviously do not have time today to even list, in a comprehensive manner, all of the reasons for this, let me describe briefly a few that I have noted that seem most relevant to the work of an auditor in today’s world.  (Please also note that I am not a psychologist; I am merely a student of human behavior.)

Implicit Prejudice

We have all heard of studies indicating that our society is still, unfortunately, more likely to (for example) equate violent behavior to African American men than to any other group.  This is a prejudice that we know exists, and hopefully all fight against.  But consider other prejudices that may not be so clearly abhorrent.  Consider, for example, another characteristic that affects everyone in this room: age.  When you first meet the CFO of a new client, and that CFO is gray-haired, does your mind reach some immediate presumptions?  Do you believe that person is likely to have deep and relevant experience?  Is he or she more likely to be mature or wise?  Do you suspect that the person is slower, either physically or mentally, than a younger person?  Is he or she old-fashioned?  Now replace that gray-haired CFO with an under-30 CFO.  What presumptions do you have now?  Is the person bright and energetic, or inexperienced and naïve?

Implicit prejudice arises from our ordinary and unconscious tendency to make associations.  Even though these associations are not motivated by ill-will, they may cause us to make costly mistakes.  For example, we may rely on someone else (the gray-haired or under-30 CFO) more or less than objective data would otherwise indicate is wise.[4]

In-Group Favoritism

This refers to the tendency that we all have to favor those people who fit within our comfort zones – those people that are “like us.”[5]  The idea of favoring those with whom we are most comfortable can affect not only who we choose for friends and professional colleagues, but also who we believe, rely upon, and trust.  Applying this concept to your daily lives – as well as to the daily lives of those human beings that embody the PCAOB – we must ask ourselves whether we tend to believe, rely upon or trust someone because (for example) we graduated from the same college, used to work at the same accounting firm, attend the same place of worship, or frequent the same social circles.

Bias Favoring Those Who Can Benefit You

A person may benefit you in a variety of ways, the most obvious of which is through your financial interests.  The SEC’s independence rules, as we’ll discuss later this morning, are intended to manage this type of conflict of interest.  But inherent biases of this type can result simply from human interaction, particularly when that interaction takes place in a commercial context.  These are biases that unintentionally skew decision-making.[6]

Research has shown that when we are motivated to reach a particular conclusion, we usually do.  Without consciously knowing it, we tend to critically scrutinize and then discount facts that contradict the conclusions that we want to (or think we should) reach.  We also tend to uncritically embrace evidence that supports our positions or presumptions.[7]  This has nothing to do with whether one is a good or bad auditor; it is simply the result of being human.

As reported in the November 2002 volume of the Harvard Business Review[8] Max Bazerman and his colleagues cite three structural aspects of auditing that create substantial opportunities for bias to influence professional judgment.  The authors of the article, entitled “Why Good Accountants Do Bad Audits,” also identify three aspects of human nature that can amplify these unconscious biases.  I’ll summarize each of these.

Structural Elements

  1. Ambiguity.  Bias thrives whenever there is a possibility of interpreting information in different ways.  Is there anything more ambiguous than much of GAAP?
  2. Attachment.  Humans are creatures of habit.  We tend to want to maintain the status quo.  In this context, we all have a natural tendency to want to maintain our business relationships, or client base, not because we are greedy for fees but because we become attached to our clients.
  3. Approval.  Biases become stronger, this article suggests, when one is in the position of endorsing someone else’s judgment (rather than making the independent judgment oneself).  For example, the “attachment” bias to maintain a business relationship becomes easier to satisfy when we are called on to accept a client’s aggressive accounting practices, than it would be if we were responsible for providing the original advice as to the most appropriate accounting policies.

These theories are supported by a study involving 139 auditors from one of the large accounting firms.  One-half of the group was assigned the role of auditor of a fictional company; the other half assumed the role of auditor of one of this company’s customers.  All participants were given the company’s financial statements in which a series of complex financial transactions were disclosed.  The “company’s auditors” was 30% more likely to conclude that these complex financial statements were GAAP-compliant than were the customer’s auditors.[9]  Thus, given ambiguous accounting rules, auditors are more inclined to support the client to which they were attached and, in the case of the “company auditors,” asked simply to endorse this client’s judgment.

Human Nature Amplification

As I mentioned, the biases that are inherently created by the structure of the auditor-client relationship can be amplified by other aspects of human nature.  These authors note three:

  1. Familiarity: People are naturally more willing to act in a way contrary to another’s wishes if that other person is a stranger.  Clearly, this aspect of human nature can impact the professional skepticism exercised by an auditor toward a long-term client.  Personal connections only exacerbate this bias.  While in some circumstances it may be true that “familiarity breeds contempt,” it also creates relationships – and most of us are reluctant to disturb long-standing relationships.
  2. Discounting.  We are also more inclined to be responsive to immediate consequences (such as the loss of a client resulting from insisting on changes to a financial statement) than to delayed and uncertain ones (such as a future restatement, lawsuit, or harm to your firm’s reputation resulting from looking the other way).
  3. Escalation.  It is easy to make small errors or oversights, and even easier when these errors or oversights are the result of the unconscious biases that we’ve been discussing.  But, we all know that some small errors can become big problems when they accumulate over time.  When this occurs, the auditor is faced with the never-welcome responsibility to admit the prior errors, which may also require acknowledging human biases.  This is often a much harder task – at least in the short-term – than is turning unwitting mistakes into conscious concealment. [10]  At that point, just as is often said of the Watergate scandals of the 1970’s, the cover-up becomes worse than the original error.

Bounded Awareness

The final type of unconscious bias that I’ll mention today refers to circumstances in which cognitive blinders prevent a person from seeing, seeking, using or sharing relevant information even though it is easily accessible and readily perceivable.[11]  For example, it’s been reported that before Merck took the anti-pain drug Vioxx off the shelves in 2004, over 100 million prescriptions had been filled, and the drug had been associated with approximately 25,000 heart attacks and strokes.  But, evidence of these risks was apparently readily available to both company personnel and doctors as early as 2000.[12]  What happened?  Well, the reasons why Merck and its personnel did not act more quickly is the subject of serious litigation focusing, I suspect, on whether consciously self-interested decisions were made.  But the prescribing doctors surely had nothing to gain by prescribing an inappropriately high-risk drug to their patients.  Why did so many doctors ignore available information?

The answer lies, perhaps, in this concept of bounded awareness.  Doctors were more aware of – and unconsciously influenced by – the favorable reports they received from their own patients, grateful for pain relief, than they were of rising statistics.  They felt closer to the patient in front of them than to the patients who were represented in the studies.[13]

There are many other examples demonstrating how bounded awareness can prevent us from seeing or seeking information.  Cornell psychologist Ulric Neisser showed a videotape to a group of people, and asked half of the group to count how many times a basketball was passed between players wearing blue T-shirts.  The other half was asked to count how many times the basketball was passed between players wearing red T-shirts.  The assignment was made more difficult because the two teams had actually been filmed playing at different times, but the footage was superimposed onto one video.  At the conclusion of the film, Professor Neisser asked: how many people noticed the woman walking across the court carrying an open umbrella?  The answer was: not many (21%).  The participants were so focused on their task – counting passes of the ball – that their brains did not register seeing the very obvious out-of-place woman.[14]  We do not see what we do not expect to see.

In another study, participants were told that a particular dangerous object would be concealed in screened luggage 50% of the time.  With this information, the participants performed fairly well, with only a 7% error rate.  However, when they were told that the dangerous object would appear only 1% of the time, their error rate jumped to 30%.[15]  This too demonstrates that we tend to see (or perhaps fail to seek) what we do not expect.

As Harvard Psychology Professor Dan Gilbert has said:

            The information that life serves us is not necessarily the information that one would order from the menu.  But, like polite dinner guests … people seem to accept what is offered rather than banging their flatware and demanding carrots.[16]

An even more sobering example can be found in the space shuttle Challenger disaster.  Post-accident investigations found that, on the day before the fateful launch, NASA personnel were quite aware of and concerned with the impact that the cold weather could have on O-ring performance.  To determine the severity of the risk, scientists examined seven previous launches in which O-ring damage had occurred, and concluded that there was no link between these past problems and cold temperature.  However, if the scientists had broadened their perspective and examined all 24 shuttle launches (including the 17 in which no O-ring damage occurred), the data set would have “unambiguously pointed to the need to delay Challenger.  Later analyses suggest that, given the low temperature, the probability of disaster exceeded 99%.”[17]  Clearly, the failure to seek all relevant information can be disastrous.  Moreover, as we’ve discussed earlier, the impact of our biases (for example, to reach a favorable outcome) can significantly influence what we consider to be “relevant” to the question at hand.

Bounded awareness can also occur because of a failure to share information.  Awareness of circumstances in which people are reluctant to share information – unwittingly and with no bad motives – is particularly important for auditors since so much of your work depends on shared information.  For example, research has shown that when team members are all given the same information and brought into a meeting to discuss it, there is likely to be good interchange.  However, when some members of a team are given unique pieces of information and brought into the same team meeting, those with the unique information tend to be reluctant to share it.  It seems that people, when offering information to colleagues, are comforted when there is supportive and corroborating participation by other members of the team.  However, when corroborating support does not exist (because the offered information is unique), team members may become reluctant to “go out on a limb” or to be perceived as outliers.[18]  


What can we do to minimize risks of unconscious attitudes impairing audit quality?  Clearly we do not have time today for an exhaustive examination of that question.  But, I can leave you with a few thoughts, and ask that you further consider these issues as you go back to your offices.  Studies show that when we are mindful of relevant risks, we will be more conscientious about managing them.[19]  One way to be mindful of unconscious biases that may affect the way in which we interact with audit clients is to collect data aimed at revealing how these same biases may exist in your own offices or professional relationships.  Look at the internal management of your firm: Who do you hire?  How do you reward good performance?  How do you define “good performance”?  On whom do you tend to place greater trust?  With this introspection, then ask yourselves:  Does any of this data indicate an unconscious bias that may also impact the way in which you relate to your client and plan and perform audits?

Thank you for your attention.  Unless there are questions, I’ll turn the podium over to our first panel to discuss the audit process, and avoiding the potholes that may occur during each phase of this process.  As we proceed with the rest of the day, please be aware of the fact that, although much of today’s information is derived from Board statements and documents (and thus can and should be relied upon), our speakers may also be called upon to give their own opinions and views on various topics.  This is particularly likely to occur in response to your questions.  We will do our best to respond to all questions, and will attempt to be clear about when an answer is clearly derived from a Board rule or policy and when it represents our own opinions; in the latter situation, please understand that our opinions are our own and should not be relied upon as PCAOB positions.

[1] The term "issuer" means an issuer (as defined in Section 3 of the Securities Exchange Act of 1934), the securities of which are registered under Section 12 of that Act, or that is required to file reports under Section 15(d) of that Act, or that files or has filed a registration statement that has not yet become effective under the Securities Act of 1933, and that it has not withdrawn.  See Rules of the PCAOB, Rule 1001(i)(iii) (available at http://www.pcaobus.org).

[2] See AU sec. 230.07-.09, adopted as a PCAOB interim standard on April 23, 2003.  See Rules of the PCAOB, Rule 3200T (available at http://www.pcaobus.org).

[3] Mahzarin R. Banaji, Max H. Bazerman, Dolly Chugh, How (Un)ethical Are You?, Harv. Bus. Rev., December 2003, at 56.

[4] Ibid.

[5] Ibid.

[6] Max. R. Bazerman, George Loewenstein, Don A. Moore, Why Good Accountants Do Bad Audits, Harv. Bus. Rev., November 2002, at 96.

[7] Ibid. 

[8] Ibid.

[9] Ibid.

[10] Ibid.

[11] Max R. Bazerman, Dolly Chugh, Decisions Without Blinders, Harv. Bus. Rev., January 2006, at 88.

[12] Ibid.

[13] Ibid.

[14] Ibid.

[15] Ibid.

[16] Ibid.

[17] Ibid.

[18] Ibid.

[19] Ibid.

Related Information