Photo by Markus Spiske on Unsplash

Adapted: Originally posted on Medium September 13, 2020

Targeting institutional power by building frameworks to prevent prejudice in decision-making is more effective.

In recent years organizations that long sat in the comfort of complacency have been forced to take a stance on racial justice. Most have realized the only viewpoint that guarantees continuity in a multicultural, multi-ethnic and globally connected world embraces equity, diversity and inclusion. This recognition, however, comes with the responsibility of comprehensive and continuous action. Unfortunately, most organizations don’t know where to start.

Many organizations have stumbled through attempts at an anti-racist transformation. In part because of a lack of diversity and expertise in organizational leadership — providing sound guidance and direction. In part because the organization’s transformation likely started at a time of profound economic disparity — due to the COVID-19 pandemic.

As absurd as it is to conflate the two — organizations have struggled with simultaneously establishing values and balancing books. As a result, many are keen on finding quick “solutions” for addressing organizational racism. Recent trends show that unconscious bias training may be one of the “solutions” being considered by the masses — with organizations ranging from prestigious colleges to tech giants investing millions in training staff.

There’s just one problem — if the goal is ending organizational racism, unconscious bias training doesn’t work — in fact, it is woefully ineffective.

What is unconscious bias training?

Unconscious bias training, as the name suggests, is a mechanism that assists individuals with identifying the biases they unknowingly hold towards a distinct group of people. Unconscious bias can be about any distinguishing factor — race, age, gender, sexual orientation, etc.

There is little debate about the existence of unconscious biases. However, training people to be aware of their unconscious biases only became popular in the ’90s following a seemingly groundbreaking psychological experiment at Yale — the Implicit Association Test.

The Implicit Association Test, or the IAT, measures unconscious bias in various areas. The experiment consists of a series of cues and responses. First, the subject is presented with images and words and must sort them into groups as quickly as possible. For the race IAT, the photos are faces of Black people and faces of white people; the words are categorized as ‘good’ or ‘bad.’ Your level of bias is measured by tracking the speed at which you organize the ‘good’ or ‘bad’ words after seeing images of Black or white individuals.

The first rounds of the IAT experiment at Yale were unsettling for many of the subjects — whose results challenged their egalitarian self-beliefs but exciting for researchers — who seemingly had developed a method to measure unconscious ideology. Even more enticing for entrepreneurs was the timing of the IAT experiment — coinciding with the internet boom — which created the perfect conditions for the rise of a multi-million dollar industry.

Over the past 20+ years, workplace unconscious bias training programs have grown exponentially, most rooted principally in the findings and application of the IAT. Now, you can find unconscious bias training online and hire facilitators to conduct unconscious bias training at your organization. If you work in human resources, education or customer service, you are likely mandated to participate in unconscious bias training before you start your job. It’s tried and tested — but is it true?

It turns out the answer isn’t a simple yes or no. There is continuous and current academic debate about the effectiveness of unconscious bias training on individuals. But, there is one thing many researchers agree on. Unconscious bias training will do little, if anything, to improve diversity, equity, and inclusion in organizations — here’s why.

It is impossible to measure the effectiveness of unconscious bias training.

Implicit bias exists, but it is impossible to measure with accuracy. Unfortunately, this is an often overlooked and misunderstood point when considering the effectiveness of unconscious bias training.

In psychology, a test-retest model is used to measure the reliability of an experiment — the maximum score is 1. For an experiment to be considered strong, it requires a minimum score of 0.7 — meaning if a subject participates in the experiment multiple times, under the same conditions each time, they are likely to receive the same, or a similar, score 70% of the time. An analysis of the race IAT scores it’s reliability a 0.44 or 44% — meaning the subject’s results vary widely each time they take the race IAT. From a scientific standpoint, this is considered unacceptable. Yet, the IAT is still widely used as the foundation for unconscious bias training and as a standalone assessment.

What does this mean?

This doesn’t mean implicit bias doesn’t exist or that the IAT is a lousy test. It just means the IAT is not an accurate enough assessment to base concrete conclusions about implicit bias. The problem is that a significant portion of unconscious bias trainers use the IAT, or offshoots of it, in their facilitation.

The typical structure of unconscious bias training often begins with the facilitation of an abbreviated IAT experiment. The intention is to lay the foundation, with scientific reliability, that implicit bias exists. This approach is compelling, as it immediately puts the trainees in a position of self-reflection, making them more susceptible to engaging with the training content.

While the net result of this approach may be positive, it can also be damaging. With a skewed baseline assessment of unconscious bias, it is impossible to accurately identify changes in belief or behaviour, making it impossible for organizations to assess the effectiveness of unconscious training on staff. Without data demonstrating effectiveness, organizations investing in unconscious bias training is more performative than productive.

Awareness does not correlate to action.

Even if the IAT met the threshold for scientific reliability, unconscious bias training is still unlikely to make a tangible impact on organizational racism — because there is no correlation between awareness of implicit bias and prevention of explicit action. That was the finding of a 2019 Meta-Analysis of the IAT conducted by researchers from the University of Arkansas, Washington University, Duke University, and the University of Virginia.

What does this mean?

Unconscious bias training is built on the assumption that when people are aware of their biases, they will consciously correct them. This study proves that that is not guaranteed; awareness does not correlate to action. Even if an individual is made aware that they have a strong unconscious bias towards Black people and actively participate in unconscious bias training, how they treat their Black co-workers is unlikely to change in the medium term. Their absence of behaviour change is not because they are intentionally callous. Instead, it reflects the power subconscious thought has over our actions, a power often greater than conscious effort.

Despite an individual’s will — sustained changes in behaviour are not instantaneous. Instead, they require consistent and prolonged personal effort and adjustments to both environmental and social influences.

If unconscious bias training doesn’t work, how do you tackle organizational racism?

The most elementary understanding of racism is prejudice plus power. Prejudices are personal beliefs about individuals or groups of people; these may be implicit or explicit. By definition, prejudice is synonymous with bias. Power is the authority and influence we have over individuals or groups. When power intersects with prejudice, it creates the conditions for racism and other forms of discrimination.

People in power at organizations — hiring managers, supervisors, executives — will, at some point in their tenure, allow their prejudice, implicit or explicit, to influence their decision-making. In the context of race, this is likely to result in discriminatory hiring practices, inequitable treatment of racialized employees, and limited diversity in senior leadership.

Unconscious bias training ineffectively targets individual prejudice to address organizational racism. When in fact, it is more effective to target institutional power by building frameworks to prevent prejudice in decision-making.

Training recruiters about unconscious bias will not prevent them from screening out qualified Black candidates before the interview stage; anonymizing the hiring process by removing names and pictures from resumes will.

Training managers about unconscious bias will not prevent them from producing negatively skewed performance evaluations for Black employees; a standardized, metric-based evaluation process will.

Training executives about unconscious bias will not prevent them from dismissing or discouraging contributions from Black team members; Black representation at the executive level will.

Unconscious bias training is about changing the perspectives and behaviours of individuals. While this is a worthy endeavour for individuals, building anti-racist organizations is about strategic adjustments to institutional power. Organizations committed to addressing racism must realize there are no quick “solutions.” Instead, change requires comprehensive and continuous action.