As organizations attempt to become both more diverse and fairer in the way they manage and attract talent, bias suppression initiatives have spread in a bid to help people override their personal views and behave in a fairer manner.  Research from the University of Notre Dame illustrates how difficult it is to maintain these well-intentioned efforts.

This was especially so in workplaces with high levels of accountability for managers to uphold certain values.  This accountability can be crucially important, but the researchers highlight that it can also backfire.

“Across multiple studies, we found that bias suppression with high accountability induces counterfactual thinking,” the researchers explain. “In other words, the decision-maker questions what would, could or should have transpired had they chosen differently. Then they regret the decision they made and ultimately—with subsequent low accountability—reverse their action.”

Held to account

Interestingly, however, this didn’t appear to occur when the accountability in an organization was low initially, before then rising over time.

“We found that when accountability changes from high to low, bias suppressed decisions tend to be reversed,” the researchers explain. “When accountability changes from low to high, biased decisions are sustained. For example, a manager with high accountability may avoid showing favoritism to a subordinate who is also a friend. If the manager no longer feels such pressure in the future, they are more likely to favor that friend over other subordinates. However, a manager who initially has little or no accountability may show favoritism to their friend and continue favoring that friend over subordinates even when they are highly accountable.”

The authors argue that when key decision makers believe they will be held to account and therefore suppress any biases they may have when making their decisions, their emotional and cognitive processing seems to play a crucial role in undermining any subsequent attempts to suppress their biases.

“Even if their bias suppression efforts are initially successful, their counterfactual thinking and feelings of regret for not following their personal instincts or preferences are so strong that people tend to reverse their unbiased decisions,” they explain. “This is surprising because people tend to strive toward consistency in their decision-making. And it means that bias suppression is not self-reinforcing.”

Consistent application

After examining the behavior of CEOs at S&P 500 firms, the researchers were able to establish a negative relationship between attempts to suppress biases and the consistent application of those attempts.

The results suggest that bias suppression is less effective when people initially tend to indulge their biases, with the researchers suggesting that our pride may be playing a major role.

“People often opt to indulge their biases and continue doing so, despite high accountability, because they view the biased decision as the right decision,” they say. “Indeed, that is likely why they feel proud. Although people are unlikely to endorse making a biased decision, we found that they will admit to making a decision based on their self-interest or strong views.”

Staying the course

Interestingly, being accountable intermittently may not be the answer to this conundrum, and it would be unwise of organizations to rely on past fair behavior to bestow trust upon a manager in the future.

When measures such as transparency and monitoring are introduced to try and provide accountability, it can be tempting to remove them prematurely after any initial successes.  This can be especially tempting given the expense involved in maintaining these accountability processes.

As such, if such measures are to be introduced, it’s important that they’re done on a consistent and long-term basis to prevent people from falling back into bad habits again.  Ultimately, the best course of action is to work on the culture of the organization so that instinctive behaviors are modified rather than relying on compliance alone.

“Organizations should be more realistic about employees’ ability or lack thereof to consistently suppress bias,” the researchers conclude, “and may find value in automating decision processes typically mired in bias.”