Making sense of vague evidence is always tough, but is especially precarious for auditors

It is a bad time to be an auditor in SA, with several of the big four accounting firms heavily involved in a string of corruption and fraud scandals. Why has this happened?

It is widely recognised that root cause of these issues is the conflict of interest that is built into the auditing profession. While there are a number of regulatory changes on the way designed to further curb conflicts of interest, the majority of the proposed changes appear to misunderstand the problem they are designed to address.

Conflicts of interest in auditing stem from the fact auditors are hired by the companies they audit, and that the accounting firms that auditors work for might provide a range of additional — and lucrative — services for that company. This gives auditors an incentive to keep their clients happy.

How do these incentives affect the quality of an auditor’s work? Whenever I ask accounting students this, they tend to assume that conflicts of interest tempt auditors to consciously collude in deceiving stakeholders about the accuracy of a company’s financial statements. When I ask for suggestions about how to combat this problem, they usually favour introducing stricter oversight and tougher penalties to scare the unscrupulous away from this temptation. This is largely how the treasury is thinking about this problem, too.

However, the lesson from behavioural economics is that conflicts of interest affect us much more deeply than we assume. The danger of conflicts of interest is not only that they create temptations for the unscrupulous, but also that they create the preconditions for the self-serving bias to take hold.

The self-serving bias is a bias towards what we want to be true. It is the tendency to see things the way we want them to be, rather than the way they are. Research suggests that, when the evidence at hand is ambiguous and complex, everyone — even the most morally pure of us — is susceptible to this bias.

Imagine that you really want Donald Trump to be guilty of colluding with the Russian state to throw the 2016 US elections. In conditions of ambiguity, where the evidence at your disposal could reasonably be read either way, research implies that you will tend to honestly believe that he did collude with Russia, with a degree of confidence that outstrips the evidence. The same applies to those with desires going in the opposite direction.

This is exactly the point at which conflicts of interest become an issue for auditors. Auditors are forced to work in conditions where the evidence at their disposal is limited. By some estimates it is not unusual for auditors to look at less than 5% of a company’s transactions, given the immensity of a company’s books. And the fact remains that auditors have reason to want to provide an opinion that makes their client happy. So even the most honest and careful auditor will have a tendency to interpret ambiguous evidence the way their client wants them to.

This makes the issue of conflicts of interest much more significant than the accounting students I spoke to assumed. It is not just unscrupulous auditors who will be prone to giving biased opinions: it is everyone.

And since the majority of people are not doing it deliberately, tactics designed to scare liars into reporting honestly will not prove effective.

So how do we solve the issue? Rooting out this bias would require seismic changes to an auditor’s incentive structure, and there is no political will for the enormous — and costly — disruption this would cause. In the absence of radical overhauls, the best auditors can do is limit the effect of the self-serving bias as much possible. This has proven maddeningly difficult to do.

However, researchers have found that getting test subjects to focus on weaknesses in their own opinions, in addition to informing them about the self-serving biases, resulted in a significant lessening of the bias’s effect. This means auditors could use a cognitive strategy called “consider the opposite” — a mental game that requires players to ask themselves the question “what are some reasons that my initial judgment might be wrong?” The evidence suggests that re-evaluating the question from this perspective significantly offsets the bias.

However, the success of such strategies depends, crucially, on buy-in from auditors themselves. This is why the first step is to simultaneously tone down the moralising on the issue and increase education.

Conflicts of interest do not pose a problem simply because auditors are especially greedy or incompetent: the self-serving bias is a bias that affects the morally pure and impure, experts and laypeople.

Auditors having to operate in an environment where there are conflicts of interest is a bit like having to work in an extremely loud, distracting workspace — it is simply not an environment conducive to producing your best work, no matter how qualified and well-intentioned you are.

• Flockemann lectures philosophy and accounting ethics at Rhodes University