The Folly of Unconscious Bias Training
Who is David McQueen?
I am an entrepreneur, professional speaker and executive coach.
I am often approached by potential and existing clients asking if I would run unconscious bias training as a part of their culture change programme and raising awareness around race and gender biases.
I do not do unconscious bias training (UBT), and for one simple reason: it does NOTHING for behaviour change.
Billions have been poured into an industry that “creates awareness”. We are already aware. We know what the issues are, and you cannot equip people to tackle bias through such training. There is more than enough evidence to show that UBT does not work. But here we are. This view often puts me at loggerheads with others in the culture development space, but I make no apologies for not pursuing a practice that has no efficacy or evidence of impact.
There are more deserving areas on which to focus my facilitating and coaching. Such as the systemic design of cultures and inclusive leadership and working with leaders and teams to tackle this holistically. As opposed to a one-off, online or day of training. In my experience, at best, UBT ticks a box, and at worst, it perpetuates biases.
Why the heck would I want to train or coach on something that, for me, does not move the needle on actual change?
Walk with me.
A brief history
UBT is a type of training aimed at increasing awareness of unconscious or implicit biases around protected characteristics, including (but not limited to) gender, age, race, and ability. UBT shares its history with anti-bias training (started in the US) and can be delivered in several ways.
The most common form of delivery is the implicit bias test (IAT): an online assessment where participants are assessed on their positive or negative responses to stimuli linked to the protected characteristics mentioned. The IAT is usually followed by a debrief of the results, an explanation of the theory and suggested techniques for reducing bias. Some practitioners and organisations follow this up with a further test to evaluate bias again.
You also have bias mitigation workshops. These workshops can include methods such as perspective-taking, counter stereotype training and stereotype replacement ( Devine, Forscher, Austin, and Cox (2012)).
Some practitioners have recognised that the process for mitigating bias is long, and especially in areas such as with student bodies, longer sessions for introspection and continuous testing have taken place. There is less evidence of this kind of practice of longevity across unconscious bias training in private and public sector organisations.
Why it’s problematic
“You can’t train the bias out of anyone, but you can equip your people with a practice they can apply to increase equity and inclusion, one question and one decision at a time”. This is a marvellous quote from the folk over at Tidal Equity.
I wholly agree that you can’t train bias out, but rather it is about having a system designed to mitigate bias through values, behaviours and rituals.
One of the challenges for me, as a practitioner, is that this singular approach to both unconscious or implicit bias used by many organisations does nothing to tackle conscious or explicit bias. So we are only tackling half of the problem here. If you have an explicit bias towards a particular gender, race or orientation, are you really going to admit it in a workshop of colleagues?
The second challenge is how to measure the outcome of such an intervention. Yes, awareness can bring things to mind but when we look at the evidence of the impact of many forms of unconscious bias training, it’s difficult to see how it would be part of a wider behaviour or culture change effect.
Thirdly, what do organisations have in place if UBT backfires? There have been several high profile cases where UBT has been shown to make biases more explicit.
Fourthly, many people have felt shamed or have a sense of guilt or blame when they see the results of these tests. I am at a loss to find evidence that shame, blame and guilt help to shape healthy culture change. If this type of training results in feeling guilty or negative around effective diversity, equity and inclusion, how do we really get buy in from a place of empathy, understanding and finding a way to move forward?
There are many versions of the implicit association test (IAT), which is mostly used for this training. According to Project Implicit, the IAT has been used by over 17 million users online since 2015. The most popular of these tests to measure prejudice and bias focuses on the black-white IAT, where you are tested on your response to images of Black and White people using i and e on a keyboard to identify whether the image evokes good or bad words. Your bias is measured here by the speed of your response and how many mistakes you make.
In and of itself, reducing racial bias to a binary of just Black and White ignores the nuances of how racial bias and prejudice plays out across various racial and ethnic groups. For the record, there are more than two.
In psychology, the efficacy of tests like this is usually measured using Test-Retest Reliability Coefficients, where tests are taken more than once in a given period. In a nutshell, two sets of data are compared to find a correlation using the scale listed below.
1 : perfect reliability,
≥ 0.9: excellent reliability,
≥ 0.8 < 0.9: good reliability,
≥ 0.7 < 0.8: acceptable reliability,
≥ 0.6 < 0.7: questionable reliability,
≥ 0.5 < 0.6: poor reliability,
< 0.5: unacceptable reliability,
0: no reliability.
The Temporal Stability of Implicit and Explicit Measures: A Longitudinal Analysis (2017) found the race IAT to have a rating of 0.44.
The other reason why many fellow practitioners find this problematic is this assumption of a clear divide between unconscious and conscious or explicit bias and behaviours. If we are looking for cultural behaviour change, surely it needs to address biases as a whole and not just as a part.
That the validity and reliability of using such a method to tackle behaviour change in the workplace have been called into question, makes me call into question this approach. So how do we change workplace behaviour to reduce or mitigate such biases?
The Case for Culture Change
There needs to be an integrated approach to how leaders and organisations explore blind spots in their approach to attracting, engaging and retaining staff. Talent management is no easy task, and it is also expensive when it goes wrong.
The cost of staff turnover can be a huge drain for most companies. In 2014, a report called The Cost of Brain Drain was written by Oxford Economics in partnership with Unum. In it, they explored the logistical cost of staff turnover, including advertising, agency fees, hiring temporary workers, HR processes and interviewing (as shown in the chart below).
According to the report, the average cost of turnover per employee is £30,641 – based on the baseline of an employee earning £25,000 or more per annum. When you consider this cost likely multiplies the larger an organisation you have, it makes sense to provide adequate investment for retaining, developing and engaging good staff.
In addition to this cost of employee turnover, there are out-of-court settlements to consider. There have been many instances of companies paying employees who have been discriminated against because of their protected characteristics. While the amounts paid out has decreased and companies are making stronger strides toward more robust policies, the fact remains, organisations are still resorting to out-of-court settlements (McGregor Smith Review and Race Disparity Audit). In some cases, for people who decide to leave the organisation, these settlements even come with arbitrary NDAs.
For leaders who have the insight to think wider than this, and see inclusion as something more than a compliance issue, there are solutions that better serve the organisation than an online test. I include three basic approaches that practitioners, including my team, have taken to address this void.
An audit of the existing culture and processes can be used to both appraise and mitigate bias. Such an audit can include, amongst others, exploring talent management.
How do you attract, hire, retain and promote staff?
What does your interview process look like?
Are they structured?
Do you use interview panels?
If so, have those staff been trained on bias mitigation?
How do you collate the feedback?
The same can be said for leadership development.
What is the process for developing leadership?
Does your succession plan have a broad base of experiences and candidates?
Is there an inclusive sponsorship and mentoring programme from emerging leaders upwards?
And, again, have existing senior leaders had coaching or training on how to lead across cultures and characteristics?
You can extend these even further to choices made around suppliers, product design, customer experience, employee engagement, and so many other processes where a series of questions can unearth how inclusive your organisation is.
After this appraisal, policies can be developed, allowing the onus to be removed from the individual and making way for an organisational or cultural approach to the issue.This doesn’t mean that individuals cannot make choices, but rather, we look at the system through which these choices and decisions are made.
Development of a training programme that has inclusion at its heart. This includes training for cross-cultural communication and, in addition to popular courses like public speaking, training and coaching on listening and conflict management. This training should be clearly aligned with the values and behaviours of the organisation.
Very often, organisations and cultures would single out underrepresented groups or characteristics as part of this training. What we suggest is that architects and curators of these programmes make sure the learning is designed to be more aware and inclusive of the context of how all groups, majority or underrepresented, perceive information. Whether through focus groups, pulse surveys or other assessment tools and feedback, to see where there are possible gaps and leaning on that feedback to shape the programmes.
A big part of how we tackle biases is learning how we communicate, how others communicate and having a sense of understanding where we meet in the middle. Without blame, shame or guilt.
Whether individual coaching of senior leaders and managers or group and team coaching, it’s a powerful and effective approach. I will break these down accordingly.
Individual or one-to-one coaching around things like leadership, performance and self-awareness is often offered to more senior staff. Given the responsibility they have for others, a coach would get them to explore how they communicate and how they lead. Exploring things like developing strategic narratives, handling conflict, building resilience, being agile, etc.
Group goals are focused on leveraging the outcomes of individuals who may or may not work together but are looking to develop competency around a specific theme. For example, you could coach a group of HR business partners around how wide they can recruit or how they can explore different interview techniques to reduce bias.
Team coaching is targeted at professionals who work together. Often focused on leveraging resources, thinking and approaches that can facilitate them to be high performing.
These three approaches make up an effective form of learning design which leaders have used (including some of our clients) to explore how individuals, teams and their organisations can approach situations and processes that might throw up biases. Using this asset-based approach, as opposed to the deficit approach of unconscious bias training, I strongly believe you can address Awareness, Behaviour change and Commitments to learning more sustainably and collaboratively.
Given the pushback against unconscious bias training as a method of behavioural change in organisations, I think it folly to pursue this as a continuous path. Even if the alternatives I have listed do not sit well with you, at least pursue something that can be measured and is more sustainable than an online test and discussion that starts from a place of negativity.
Judgement Under Uncertainty
Equality Human Rights: Assessment of Unconscious Bias Training
Implicit Bias Trainings are Flawed