Unconscious Bias Training Doesn’t Work, But It’s Making Billions

By now, we’ve all heard of “Unconscious Bias Training” programs, designed to make us aware of our explicit and implicit biases and to promote diversity in the workplace.

By Freya India4 min read
coca cola be less white Unconscious Bias Training Doesn’t Work, But It’s Making Billions

But, despite their popularity, there’s no compelling evidence that these training sessions actually work. In fact, they can even be counterproductive. 

What Is Unconscious Bias Training?

Unconscious bias training is based on a key assumption within Critical Social Justice theory: that we each possess “implicit biases.” Implicit biases are “attitudes or stereotypes that affect our understanding, actions, and decisions in an unconscious manner,” such as how we perceive individuals of a different race, gender, or sexuality. In aggregate, these individual biases are thought to amount to oppressive structures in society such as “systemic racism,” essentially a “system of people’s unconscious thoughts.” 

In the U.S. alone, companies spend over $8 billion a year on unconscious bias training.

Unconscious bias training aims to help individuals “unlearn” these biases and change their behavior. These programs can take various forms, but they often follow three basic steps: 

  1. Participants take a pretest to assess baseline implicit bias levels (Implicit Association Test)

  2. They complete the implicit bias training task

  3. They take a posttest to re-evaluate bias levels after training

For example, an implicit bias training task may involve “imagining powerful women, hearing their stories, and writing essays about them” or pairing “counterstereotypical traits such as ‘successful' with images of black individuals.”

How Popular Is It? 

In the wake of George Floyd’s death, there was a huge increase in demand for these kinds of programs across the West. As you can see, Google searches for “unconscious bias” spiked dramatically:

Image Source: 

In the U.S. alone, companies now spend over $8 billion a year on unconscious bias training. In 2018, for example, Starbucks closed more than 8,000 stores across the U.S. to complete these programs. Google also introduced 60-90 minute unconscious bias training sessions to its employees, while the United States Department of Justice imposed implicit bias training on more than 33,000 of its federal agents and prosecutors. According to a recent report, 81% of companies now run unconscious bias training in the UK.  

But, Does Unconscious Bias Training Actually Work?

And yet, hundreds of studies dating back to the 1930s suggest that unconscious bias training doesn’t actually reduce bias or change behavior. A 2017 meta-analysis looking at 494 studies (currently under peer review) found that reducing implicit bias didn’t have any significant effect on behavior. 

Hundreds of studies suggest unconscious bias training doesn’t reduce bias or change behavior. 

“I was pretty shocked that the meta-analysis found so little evidence of a change in behavior that corresponded with a change in implicit bias,” psychology professor and co-author of the meta-analysis Patrick Forscher commented. “I currently believe that many (but not all) psychologists, in their desire to help solve social problems, have been way too overconfident in their interpretation of the evidence that they gather...The impulse is understandable, but in the end, it can do some harm by contributing to wasteful, and maybe even harmful, policy.” 

Does It Do More Harm Than Good? 

One Harvard University study found that unconscious bias training can actually make individuals more hostile toward other groups. For example, Kawakami, Dovidio, and van Kamp challenged the efficacy of “counterstereotype training” after they found an increase in participants’ gender bias following training sessions. Using a sample of 2,000 people, Kulik et al. also found that implicit bias training increased prejudice against older participants. 

In one study, white individuals were made to read a brochure critiquing prejudice toward black people. When participants felt pressured to agree with it, their bias against people of color actually strengthened.

Many participants actually report more animosity toward other groups after compulsory courses.

Writing for the Harvard Business Review, sociologists Frank Dobbin and Alexandra Kalev reveal that “people often respond to compulsory courses with anger and resistance — and many participants actually report more animosity toward other groups afterward.”

“Five years after instituting required training for managers, companies saw no improvement in the proportion of white women, black men, and Hispanics in management,” they go on to explain, “and the share of black women actually decreased by 9%, on average, while the ranks of Asian-American men and women shrank by 4% to 5%.” 

So, Why Doesn’t Unconscious Bias Training Work?

Of course, racism, sexism, and other forms of prejudice exist. But the question is: Do they influence our every decision, and can they be unwired from our subconscious simply by a corporate training program? Can we quantify these biases in any meaningful and reliable way to measure exactly how they affect behavior?

The results suggest that we can’t. Repeatedly, the Implicit Association Test inherent to unconscious bias training produces different scores when participants take the same test numerous times. In fact, the “test-retest reliability is so poor that it falls short of scientific standards.” 

There are over 200 cognitive biases that affect our everyday decision-making and behavior.

This is likely because the human mind is extremely complex. There are thought to be over 200 positive and negative cognitive biases that affect our everyday decision-making and behavior. For instance, our actions can be distorted by:

  • Perception bias: when we believe something is typical of a particular group of people based on cultural stereotypes or assumptions.

  • Affinity bias: when we feel as though we have a natural connection with people who are similar to us.

  • Halo effect: when we project positive qualities onto people without actually knowing them.

  • Confirmation bias: when we try and confirm our own opinions and pre-existing ideas about a particular group of people. 

Unconscious bias training tends to focus on only one type of bias (stereotype bias), while largely ignoring other distortions. To identify a specific, unconscious belief among a miscellany of intersecting and complex schemas, and unlearn it, is an impossible feat.

Attitudes aren’t really strong predictors of our behavior, either. Studies show that there’s rarely more than 16% overlap (correlation of r = 0.4) between attitudes and behavior,” and the figure is even lower when it comes to prejudice and discrimination. Therefore, you simply can’t predict whether someone will act in a racist or sexist way from their attitudes alone. What’s more, “the overlap between unconscious attitudes and behavior is even smaller (merely 4%.)” This means that, even if unconscious bias training did change people’s attitudes, it’s unlikely to change how they behave.

Are There Alternatives?

Despite a body of evidence suggesting that these training sessions don’t produce reliable results, and may actually be harmful, unconscious bias programs are compulsory in many corporations, schools, and other institutions.  

Fortunately, co-author of the Grievance Studies Hoax Helen Pluckrose recently announced Counterweight, a new organization providing letter templates for employees to use if they disagree with Critical Social Justice initiatives in their workplace. The letters ask employers important questions such as:

“How will management calculate the baseline of unconscious bias in the company to assess the efficacy of the proposed training, in light of the poor evidence base for implicit association testing?”

Companies could instead promote “moral courage” and encourage employees to speak the truth.

Counterweight also offers alternative suggestions to combat racism and foster equality in the workplace. For instance, companies could instead promote “moral courage”  — that is, encouraging their employees to speak the truth and potentially engage in disagreements, ensuring viewpoint diversity. 

They also recommend the OpenMind Platform, a curriculum that uses psychological principles to help individuals learn about the morals and values of different cultures, replacing divisive/us-vs-them rhetoric. 

These alternatives provide a form of diversity training without the shame and guilt attached. For instance, Coca-Cola recently promoted an online training course entitled  “Confronting Racism” in association with the author of White Fragility, Robin DiAngelo (who charges anything from $30,000-$40,000 for a few hours of training). The course featured a slide entitled “Try to be less white.” Is this really the best way we can promote equality?

Closing Thoughts

It’s not that those who criticize unconscious bias training think that biases, cognitive distortions, and heuristics don’t exist — there’s just not enough convincing evidence that unconscious bias training does anything to reduce them, or that they even affect our behavior. 

Looking at the research, it seems that the popularity of these programs has more to do with companies ticking a PR box than actually striving for diversity and inclusion. 

Ultimately, we have to remember that these training programs, no matter how well-intentioned, often cause more harm than good. For those of us who want to bring society closer together, rather than worsening the divide, it’s therefore crucial that we resist the uncritical adoption of these initiatives in the workplace.