Despite the significant costs of unconscious bias in the workplace—here defined as ingrained negative attitudes toward people based on race, ethnicity, gender, or other sociodemographic factors—efforts to reduce it have proved largely ineffective.
In 2020, American companies lost an estimated $280 billion from absenteeism, decreased productivity, and turnover caused by employees’ perceptions that they had been victims of workplace bias, according to a study by the Society for Human Resource Management.
Combating workplace bias is costly. As of 2017, companies were spending more than $8 billion per year on bias-focused employee training, according to McKinsey. Some organizations are starting to experiment with new virtual-reality training tools that target unconscious bias.
Those investments have delivered few returns to date. A 2019 study discussed in Harvard Business Review found that training meant to address unconscious bias leads to little behavioral change among men and white employees, and in some cases may even exacerbate biases.
To understand why most bias training fails and what leaders can do about it, Workflow spoke with Francesca Gino, a behavioral scientist and professor of business administration at Harvard Business School.
The interview below has been edited and condensed for clarity.
We’ve found that people often don’t take training seriously, especially when it’s mandatory. You find people saying, “This seems to be human nature. There’s nothing I can do about it.” Or the examples used in the training are such extreme cases of abuse or harassment that they say, “I’m not that type of person. I’d never do such a thing,” so they fail to listen.
[Read also: The impact of hybrid work on gender equality]
Beyond that, many organizations look at unconscious bias training as a quick fix. The attitude is very much “check the box.” But the reality is that it requires a commitment to much larger changes. It’s an opportunity to get the best out of diversity within the organization. To do that requires a willingness to embrace the messiness of it.
Leaders need to take a close look at all the processes that are part of the human resources cycle—from hiring, promotion, and compensation to performance management. They need to see whether there are opportunities to make changes that fundamentally help with the broader goal of making the workplace more inclusive.
For example, many companies include self-evaluation as part of their performance review. Research by my colleagues at Harvard Business School shows that women and people of color tend to be less likely to talk in these assessments about their accomplishments, compared to their colleagues. Many managers look at self-evaluations before writing their own reviews, which can impact raises and promotions.
As long as there is bias in self-assessment, a performance review from the manager is going to be biased as well.
Empirical data suggests that unconscious bias tends to be most acute in fields that are male-dominated, as technology has been. In my experience working with leaders and employees in these organizations, their bias training and even their more well-developed initiatives are good efforts but don’t go far enough, particularly in terms of representation.
Lack of representation can limit employees’ feeling of inclusivity. People want to see others like them in leadership positions. It goes back to the idea of making a broader commitment to policies that foster a more inclusive and diverse environment.
In the next five years, we’re going to see much more experimentation with VR and AR in training. For example, a male manager in financial services preparing for a performance review for a female direct report might practice their conversation with an avatar to ensure that their feedback isn’t biased [such as real-time responses from the avatar within the training session or post-session analysis of body language and tone of voice].
Technology can also help track inclusivity through metrics like data on hiring, promotions, raises, and team leadership roles. Another tool might be an app that sends a brief survey every week to ask anonymously, “To what extent are you feeling included in your workplace?”
Technology like that can help us become much more efficient at tracking the employee experience of inclusivity than an annual survey.
Effective training tends to use examples that highlight common and relatable forms of bias, rather than extreme examples. They help someone see, “Yes, in fact, in a meeting, I might be interrupting my female colleagues more than my male colleagues.”
Seeing yourself in the scenarios that you’re being trained on is quite important in changing your own behavior, and ultimately the culture of the organization.