Does unconscious bias training work?

Written by
Clare Grist Taylor

Published
09 May 2019

09 May 2019 • by Clare Grist Taylor

When Starbucks hit the news last year by closing 8,000 of its shops to conduct diversity training after the arrest of two black men for ‘trespassing’ at one of its branches in Philadelphia, debates around the effectiveness of its four hour bias training went mainstream.

While some praised the company for taking racism seriously, others condemned the move as anything from an ill-conceived management box-ticking exercise to a publicity stunt. But one thing was clear: this was a very public airing of concerns around the effectiveness of unconscious bias training (UBT) as a tool for tackling workplace exclusion.

The prejudice in the Starbucks case may have been an extreme example with extreme outcomes, but bias, whether explicit (conscious) or implicit (unconscious), can blight organisations of all types and sizes. Unconscious bias, which equates to the inherent attitudes and stereotypes that influence our understanding, actions and decisions, is considered particularly challenging, as we are not always aware of these.

In the workplace, where we have to make daily judgements about working relationships, it’s easy to see how bias might get in the way of diverse and inclusive cultures.

This matters. More than ever, we have evidence that a diverse workforce is not just positive in its own right, but an important driver of organisational reputation and growth. McKinsey’s 2018 Delivering through Diversity report showed that more gender-diverse boards can improve performance by 21%; for ethnically diverse boards, the figure rises to 35%. The imperative to ‘do something’ in this space has become more compelling. For many organisations, tackling unconscious bias with UBT has become a cost-effective, convenient response.

Fundamentally, UBT is designed to do three things: make people aware of their biases, often using a diagnostic such as the Harvard Implicit Association Test (IAT); provide tools and techniques to adjust automatic ways of thinking and, as result, modify behaviour to reduce bias. But does it work?

The jury is still out. Academics have debated its effectiveness for years. A 2018 report from the Equality and Human Rights Commission found a “mixed picture”: UBT might help to raise awareness of implicit bias, but there was only “limited” evidence that it changes behaviour.

The positive effects of UBT can be transitory and easily forgotten. In the worst cases, mandated or badly delivered training can backfire, actively reinforcing stereotypes by making people feel that, because bias is universal, it’s somehow inevitable, or provoking a backlash from people who feel defensive about being seen as sexist or racist. For minority group members, raising awareness of difference can also heighten feelings of alienation. Increasing awareness around bias is clearly not without its challenges.

Another criticism is that UBT is inevitably focused on individual behaviour rather than broader organisational culture. Add to this heated debates about the reliability of tools such as IAT and the fact that UBT is not aimed at tackling explicit bias (the main culprit in the Starbucks debacle) and the case against it grows damning.

But let’s think critically about what we’re asking UBT to do. According to Tracey Groves, CEO of Intelligent Ethics, tackling bias is about “unlearning often centuriesold attitudes, societal norms and behaviour. That’s why UBT can be so uncomfortable and why unconscious bias is so hard to unpick”.

Groves emphasises that bias can never be entirely tackled by UBT, but it can be a good starting point if approached as a beginning of a more complex journey to build an ethical and inclusive culture: “UBT can raise awareness about difficult issues and open up debates organisations need to have, offering a safe environment to ask difficult questions and confront biases on a collective basis, at all levels of the business and across all protected characteristics,” she says.

“The key thing is for organisations to be clear about the format and purpose of the training, how it fits in with broader company strategy on inclusion as a business imperative – and that it’s not used in isolation.”

If it’s unreasonable to expect deep-seated attitudes to change overnight, it’s equally unreasonable to expect UBT to shoulder the burden of diversity and inclusion (D&I) on its own.

Fears have also been expressed that the limits of UBT might be used as an excuse to close down D&I debates and initiatives. Nancy Roberts, CEO of Umbrella Analytics, shares the current disquiet about UBT, but is equally concerned that this should not derail the D&I agenda more widely.

“UBT is, in many ways, the poster child of D&I work, and if its legitimacy is under question, this may have ramifications for D&I as a whole,” she argues.

“Rather than simply discounting UBT, I’d put in a plea for organisations to be really considered and intentional when they deliver it to make sure its effectiveness is maximised. The question we should ask is not UBT or no UBT, but, rather how can well-deployed UBT contribute to D&I in the round?”

So, rather than retreating at the first sign of challenge to UBT, organisations need to develop a more nuanced approach to D&I, improving their use of UBT as just one of the tools at their disposal to create a range of strategies for bias awareness, reduction and mitigation.

In a 2016 Harvard Business Review article, entitled Why Diversity Programmes Fail, sociologists Frank Dobbin and Alexandra Kalev rehearsed a number of arguments against mandated, command-and control-led diversity initiatives, including diversity training. They also looked at practical strategies for inclusion which don’t focus on control, but instead apply three basic principles:

1. Engage managers in solving the problem

When managers actively help boost diversity in their companies, they begin to think of themselves as diversity champions. Unsurprisingly, asking for volunteers to lead positive programmes for change (“help us find a greater variety of promising employees”) works better than blaming and shaming with rules and re-education.

2 Expose them to people from different groups

More contact between groups can lessen bias. Cross-functional or self-managed teams and cross training all make a difference.

3 Encourage social accountability for change

People generally want to be seen to be doing the right thing. Improving transparency to highlight discrepancies, creating diversity taskforces and having a nominated diversity manager tends to make people consider and amend behaviours more readily.

There is growing evidence that strategies like these are being investigated and deployed.

Louise Byrne, vice president of global talent at IHG, has been leading on the development of conscious inclusion ‘workouts’ designed to make awareness of, and action around, D&I more integral to ‘business as usual’.

As part of these programmes, UBT is seen merely as a resource to build awareness, used alongside tools such as Accenture’s powerful #InclusionStartsWith video. The emphasis of the workouts is on creating empathy, getting to know and understand people as individuals rather than representatives of particular demographics or cultures.

The workouts were first used at a meeting of 75 of IHG’s senior leaders in November 2018. After initial scene setting, participants were encouraged, in smaller break-out groups, to share their own experiences of, and perspectives on, difference and to discuss how they could become more active in spotting and tackling systemic bias.

“The sessions gave our leaders the space and permission to have often quite difficult conversations,” says Byrne. “It introduced them to a vocabulary and sets of tools to help them acknowledge tensions, but still be prepared to face them.”

The company’s leaders left the meeting with personalised action plans, and the workout programmes are now being rolled out across the company. For such an openly customer-facing business as IHG, respecting difference is not just a matter of risk management, it also opens up business opportunities.

“Understanding D&I is not a nice-to-have for us,” says Byrne. “It’s a business priority. The workouts have definitely moved on the debate within the company.

“It’s increasingly clear that empathy-led initiatives such as separate check-in floors for Muslim women travelling alone or reward clubs that offer advice about  LGBTQA-friendly  travel destinations can give organisations the competitive edge.”

Another bias-reduction strategy gaining traction is immersive experience. This gives participants the opportunity to spend stereotype-busting time with particular groups, to understand their needs and the barriers they face in the workplace.

An initiative between Ashridge at Hult International Business School and Barclays, to create and evaluate an immersive CSR activity as an alternative to diversity training shows how powerful it can be. This brought together 39 able-bodied Barclays executives with an equal number of people with disabilities to sail one of the Jubilee Sailing Trust’s accessible tall ships as working crew. Members had no alternative but to work together to reach their common goal of voyaging around the UK.

Measures of implicit and explicit attitudes taken before and after the voyage showed a significant shift in attitudes among the able-bodied executives. In follow-up interviews and focus groups, participants revealed that working with people with disabilities, connecting with them, offering and receiving assistance and being exposed to their abilities and competencies rather than their limitations, resulted in their feeling more comfortable and confident when meeting people with disabilities. Immersion had made them less likely to view their crew mates’ disabilities as their defining characteristic.

In the UK’s Civil Service, bias training is compulsory for anyone with a responsibility for recruitment. But for Jane Nicholson, former HR director for D&I at the Home Office, it was apparent that, while the training created a culture of awareness, it wasn’t leading to a marked change in the department’s make-up.

“The HR team was winning awards for the D&I infrastructure we had in place, but we needed to move beyond D&I being seen as the exclusive preserve and responsibility of HR,” she explains. “We had to find ways to integrate D&I into our business strategy more widely.”

The breakthrough came by focusing on diversity data. Nicholson’s team presented functional heads with detailed breakdowns of their staff by gender, ethnicity and disability, showing the differences by grade level and comparing the results against national and even regional population data.

She reports that “focusing on the micro-level enabled managers to see clearly where there were discrepancies, which allowed us to support teams with programmes and improvement plans to target specific groups”.

In one group, two women from BAME backgrounds were awarded temporary promotions to give them the experience they would need to apply for more permanent roles at that level. In another, the tendency only to appoint team members with an elite education was challenged by new approaches to the pipeline, including apprenticeship schemes.

“By giving managers ownership, and providing the data to underpin smaller-scale and more manageable programmes for improvement, we created a much wider network of inclusion champions,” says Nicholson. “D&I is now seen as a form of talent management in its own right.”

UBT is based on the reasonable assumption that it’s important to recognise and address the attitudes and stereotypes that make up unconscious bias.

And there is evidence that, carefully deployed, it can be useful for awareness raising. But deployed in isolation, its ability to change behaviours – the real drivers of change – is limited.

For organisation-wide change, long-term, top-down cultural change must be driven by everything from tone and language to organisational structures, policies and procedures. For that to happen, businesses need to experiment with and deploy a whole range of strategies, techniques and tools, such as the examples highlighted here.

Ultimately, we need to view bias training as the tool that it is rather than as the silver bullet it’s often perceived to be.