For an Auditor, Intuition Might Matter!

Thinking Fast versus Thinking Slow: The Effect on Auditor Skepticism

• Wolfe, Christensen and Vandervelde
• A version of the paper can be found here.
• Want a summary of academic papers with alpha? Check out our Academic Research Recap Category.

Abstract:

Based on psychology theory, we propose that intuition can be a key element stimulating auditor skepticism, whereas overreliance on analytical processing can overwhelm auditors’ intuition thereby reducing skepticism. We test our expectations with an experiment containing responses from 85 senior auditors. Our results support our theory. We find that auditors are more likely to judge an asset as potentially impaired if they use their intuition as opposed to analytical processing. When we categorize auditors on their innate use of intuition, our results become more pronounced. We find that auditors using analytical processing, who rarely use their intuition, seldom judge the asset as potentially impaired. Our research suggests that intuition can be of use to auditors, and when ignored, auditors can become less skeptical. These findings should help inform regulators, standard setters, and audit firms as they seek to enhance professional skepticism.

Alpha Highlight:

We continually beat the drum of “computers good; humans bad” when it comes to decision-making. This drum beat is based on the piles of research papers that justify this position. However, not all research suggests that human intuition is ALWAYS worth eliminating from a decision-making process. This paper is an example of that literature and serves as an interesting counterpoint to the argument that decision-making should be purely objective and not involve an element of “human gut-instinct.”

Daniel Kahneman, the winner of 2002 Nobel prize in Economics, has been studying human biases for more than 40 years! His best-selling book, Thinking, Fast and Slow, summarizes much of his research and provides tons of thought-provoking lessons. Prof. Kahneman reminds us how easily we can get trapped by our own behavioral problems. In the book, he describes two modes of decision making processes:

• Thinking fast (Intuitive processing): instinctive, emotional, and even biased gut feeling;
• Thinking slow (Systematic Processing): careful, robust and evidence-based analysis.

This paper, written by three accounting professors, studies these two modes of thinking on a group of “Professional Auditors:” 85 Big Four Senior Auditors.

Hypothesis Development:

Auditors are  trained to have an alert mind when assessing evidence. In the audit industry, such skepticism is called “Professional skepticism.” To appropriately apply professional skepticism in practice, auditors should make reliable impairment judgements fairly, which means they should approach their judgement based on the evidence.

In general, using an objective systematic approach in the decision-making process is beneficial because it strips out the potential “bias” associated with human decision-making. In effect, purely objective approaches strip out “intuition” from the decision making problem.

Removing intuition from most decisions is typically a good idea, primarily because it prevents overconfidence. For example, in the context of investing, allowing a human being to continually pour over SEC filings will not improve their information set, but it will enhance their confidence in their original thesis, thus leading to overconfidence problems.

However, in the context of auditing, where the natural “intuition” is to bias decisions to be less confident, eliminating intuition, which eliminates skepticism at the margin, might actually make an auditor less effective.

In this context, a purely objective audit-decision making process may not capture the various schemes a company might play on an auditor to try to “trick their model.” Only human intuition, which is biased towards busting people in this scenario, can capture this potential element. The paper essentially shows this to be the case — an auditor equipped with natural skepticism and a model, tend to do better than a purely objective model.

Score one for the human — we needed to put some points on the scoreboard!