The first session of the second day of the conference, titled “Auditors’ Biases,” covered all manner of conscious and unconscious bias as well as how they are addressed through professional standards. The panelists were: Harry Cohen, Partner, KPMG LLP, and member of the Auditing Standards Board (ASB); Bob Dohrer, Chief Operating Officer, RSM International; and Audrey Gramling, Professor, Oklahoma State University, and member of the ASB. The panel was moderated by Professor Doug Carmichael, Baruch College, City University of New York.
Research into Bias
Audrey Gramling kicked off the discussion with some background on what cognitive biases are: “In the early ’90s, I was starting my academic career, and it focused on auditor bias research. At that time, I think I would be safe to say that the profession did that have much appetite to talk about auditor bias. There was the belief that these are professional people, they have training, they’re not going to be biased in their decisions—and I think over 20 years’ worth of research suggests that may not be true. And so the idea that we’re here today and we’ve got that term in the auditing standards is pretty exciting to me.”
“When I say ‘auditor bias,’” Gramling continued, “it’s just shorthand, if you will, for cognitive biases in decision making.” She noted that the term was introduced by economists to describe the mental approximations people use to understand an uncertain world and make judgments about it. Gramling provided examples of how such biases can be adaptive as well as misleading. She also explained how anchoring bias can serve to influence the estimates one might make in quantifying items based upon an initial starting point or frame of reference. Gramling then described this framing effect in the context of an audit program, which often asks whether management’s estimates are reasonable assumptions, thus establishing a positive frame.
“Why does this happen?” Gramling continued. “There are limits on our mind’s attention, there’s lots of information out there, lots of things we have to focus on, and we have limits on our attention. It’s difficult for us to ignore information that we know.” She described the line of research about motivated reasoning: “If I know my supervisor wants me to arrive at a certain decision, or I know the client wants me to arrive at a certain decision, unconsciously that is going to affect what I do in my decision-making process.”
Carmichael asked about Statement of Auditing Standards (SAS) 142, Audit Evidence, and its treatment of audit evidence and cognitive bias. “It wasn’t really until recently,” Bob Dohrer replied, “that there has been some movement towards understanding auditor bias plays a significant role.”
“The issue was that you know that it’s terribly difficult to write a specific stand-alone standard on the exercise of professional skepticism and how that relates to forming professional judgments,” said Dohrer, describing his experience on the International Auditing and Assurance Standards Board’s (IAASB) professional skepticism working group. “The fact is that unconscious bias is an element of the exercise of professional skepticism, and of course that impacts professional judgments made throughout the audit, anywhere from designing procedures, all the way through to evaluating audit evidence. And that unconscious bias on the part of the auditor may actually impair the exercise of professional skepticism.”
“If I know my supervisor wants me to arrive at a certain decision, or I know the client wants me to arrive at a certain decision, unconsciously that is going to affect what I do in my decision-making process.”
“It started with this concept of the auditor,” Dohrer said, turning to the AICPA’s process for bringing the importance of professional skepticism into SAS 142. “You can’t turn your head on the fact that the auditor needs to be skeptical when thinking about whether [he has] accumulated sufficient appropriate audit evidence and when thinking about the reliability of audit evidence. Skepticism really came into play in the context of making sure that auditors think through the unconscious act of biases like confirmation bias or anchoring bias—that they’re not dismissing audit evidence that may actually be contradictory.”
“I really look forward to the standards addressing unconscious bias and how you deal with that,” Dohrer added. “And I’m looking forward to hearing a practical, real-life perspective [of] how auditors and firms are dealing with these issues.”
From Awareness to Implementation
“Awareness is a key concept,” Harry Cohen observed. “No matter how long you practice in this profession, we are reminding ourselves continually about those potential biases. It is a necessity in order to overcome them.”
Cohen elaborated on the importance of confronting biases. “At the end of the day, we are in a service industry, and sometimes it’s hard to remember, as professionals, that the actual customer isn’t who we are actually servicing,” he said. “We’re servicing the capital markets at large, but we’re working with members of management on a daily basis—and it’s those daily interactions that could influence people’s conclusions. When you really get down to it, what we’re really doing as practitioners is making a judgment about people and issues. In our delivery of an audit, we need to maintain a questioning mind.”
“Now more than ever, information in audit evidence is more obtainable or attainable through a variety of sources. A couple of decades ago, there wasn’t the Internet, there wasn’t Google,” Cohen observed. “Now the relevance and reliability of that information is certainly a challenge, and we as practitioners have an ability to develop audit procedures that are intended to confirm what the company or management has presented to us. But we have the ability to obtain information to develop our own expectations about what an estimate should be, and that is an awfully good way of overcoming bias.”
“There’s an awful lot of things that, as professionals, we’re expected to focus in on. But judgments and estimates and the subjectivity around those certainly require more senior-level involvement. I try to describe my observations—my expectations about what we’re doing, how we’re doing it, when we’re doing it—to my team and coach them throughout the exercise,” Cohen continued. “From a consultation perspective, no matter how much experience that we have, reaching out for consultation is very powerful—because any sort of bias we might have that we’re unaware of, a consultation from outside of the engagement team with people who are further removed from the facts and circumstances and the interactions, makes it more likely we are to flush out questions or concerns or biases that we as practitioners are falling into a trap, without recognizing it.”
“No matter how long you practice in the profession, we are reminding ourselves continually about those potential biases. It is a necessity in order to overcome them.”
Carmichael asked whether auditors are sometimes biased in favor of their clients being honest. “In the standards today, we recognize that professional skepticism, and therefore the formation of professional judgments, is pervasive throughout an audit engagement,” Dohrer said. “I think that extends back into client acceptance.” He added that the AICPA quality management project recognizes that “evidence needs to be accumulated to support any decision made, from the very inception of client acceptance, all the way through issuance of deliverables and reports.” Dohrer suggested that nonauthoritative guidance and academic research would be helpful in building upon the groundwork laid in the new standards.
“I think that awareness is already built into the standards, for example, through fraud brainstorming,” Gramling added. “Now, could the application guidance or the standards be improved with respect to how best to conduct brainstorming? Maybe. I think there is academic literature out there that would be helpful.”
“As is often the case with standards setting,” Dohrer said, “it’s a fine line that you walk because while everybody can agree about the desirability of, I’ll call it, a robust consideration of contradictory evidence that might exist, once you set a standard that would require an auditor to go looking for contradictory evidence, this opens up that full other scale.” Dohrer continued on the topic of how standards should address the documentation of contradictory and confirmatory evidence: “I think the way we’ve tried to address that in standards is to elevate and bring emphasis to this awareness of bias that might exist, and how that should drive auditor behavior.” Gramling added that academic research has uncovered a “likability bias,” which indicates that fraud assessments can be influenced by whether clients are warm and welcoming, or tense and rude.
“The way we’ve tried to address [evidence] in standards is to elevate and bring emphasis to this awareness of bias that might exist, and how that should drive auditor behavior.”
“Over the last several years, there’s been an increased focus on how we execute the audit and how that influences how we overcome professional skepticism, or how we become more professionally skeptical and overcome the potential for those biases,” Cohen responded to a question regarding documentation requirements for the judgments made by audit teams. “We are not only focusing on the nature and extent of those procedures, but the timing of those procedures, the documentation of those procedures, and the review of those procedures. Not only how and what we’re using to get our jobs done has changed, but also when we’re doing those activities has certainly changed—it’s accelerated earlier in the audit process.”
“Over the last several years, there’s been an increased focus on how we execute the audit and how that influences how we overcome professional skepticism.”
“I think, to Audrey’s point, we use fraud specialists to help bring to life things that the core engagement team might not be as sensitive to,” Cohen continued. “All of these tools become more meaningful when we’re executing them earlier in the audit process. That certainly mitigates the timing elements; it facilitates a change to our initially designed procedures and our ability to coach the clients, management, and company stakeholders, and explain why we’re doing something different and why these incremental requests are coming. All of those reactions might be an undue influence—which is historically why people didn’t do things that we really want them to do as part of the audit process.”