It’s easy to accept in principle that there’s no magic bullet, harder to stop holding out hope for one.

If the last few decades of psychology have taught us anything, it is that we are not the rational, deliberative decision makers that we like to think we are. This includes the unavoidable insight that we are biased in ways we are not conscious of. To rectify, or rather help people self-manage this bias, unconscious bias training (UBT) has quickly become a popular learning and development intervention. And on 6 July, responding to the Black Lives Matter movement and in the wake of Labour’s antisemitism crisis, Keir Starmer announced he would be leading the Labour Party in undertaking unconscious bias training.

The training typically starts with an unconscious bias test, explains the psychology of bias and gives examples of how bias can work in practice. Its popularity has grown in recent years, in part because the MacGregor-Smith Review for the UK government called for employers to make it mandatory for all employees. But when seen as the go-to-solution, unconscious bias training just won’t cut it.

What are we trying to change?

Prejudiced ways of thinking or behaving that we do without realising have an ingrained nature. They come from our background or education, learnt over many years, or from cultural assumptions we pick up in adulthood. It’s our essentially human traits, like our tendency to notice and perhaps be wary of people who are different from us that make us susceptible to these biases, and . No matter how good it is, no half-day training session will unravel deeply embedded bias.

People may feel less compromised or criticised by the idea of unconscious bias – ‘I wasn’t aware of it, so it’s not really my fault’. But a lot of bias is explicit – that is, conscious or at least semi-conscious enough that people can’t claim ignorance. Consider a manager who is more likely to hire a man than a woman because the woman may go on maternity leave, or the British Vogue editor Edward Enninful being told to ‘use the loading bay’ on arrival at work. It may be convenient and unthreatening to label such prejudice as unconscious, but it’s hardly convincing. At some point, changing deep-rooted bias and overturning structural disadvantages will be inconvenient and uncomfortable. Let’s not suggest otherwise.

But one can argue it’s harder to keep oneself in check if we don’t realise we’re being prejudiced. So won’t drawing attention to it through training help combat it?  Not necessarily, is the unfortunate truth.

What’s the evidence on unconscious bias training?
Unconscious bias is a huge problem, but it doesn’t necessarily follow that explaining the psychology of it will reduce it. The best research evidence – summarised thoroughly in a review for the EHCR – suggests that unconscious bias training is effective for awareness raising and can reduce implicit bias. However, it can’t eliminate implicit bias, it sometimes backfires and increases bias, it doesn’t reduce explicit bias and any benefits are short-lived and unlikely to change behaviour.

A reason unconscious bias training backfires is when it suggests prejudice is immutable. People can walk away with a sense that it’s not worth bothering: ‘we’re all prejudiced – what can you do?’ It can also be explained by what psychologists call ‘moral licensing’: people feel virtuous or confident having done the training and then stop consciously making an effort (This is the same reason some people can become less healthy after taking a course of vitamins; they feel virtuous and stop making an effort in things like exercising and eating healthily that are what really make the difference.) In either case, there’s a risk we stop striving to be open to people’s differences or ‘otherness’ and to keep our prejudices in check.

What are the alternatives?

It seems that other types of training and development are a better bet than unconscious bias training in supporting workforce diversity. While the evidence is thin, ‘perspective taking’ approaches show more promise. This focuses on sharing stories and involve exercises in which one imagines ‘walking in someone else’s shoes’. It needs to include deep reflection on what the experiences are of minority or disadvantaged groups.

But no training should be seen as the only solution. Other methods include regular communications that reinforce messages, active policies that force change (for example, recruitment targets for different demographic groups) and effective channels for holding people to account.

Making evidence-based decisions

Given finite time and resources, leaders have a responsibility to ensure investments actually drive change. Decisions on interventions must be evidence based, guided by whether it is likely to have a sizable impact, not by whatever the industry standard happens to be. And practices should be evaluated and improved where possible. Evidence-based decision making is at our disposal, so there’s not really an excuse not to do it.

Certainly, it’s not enough to plonk in place a popular policy, or hustle people through an off-the-shelf training programme, and say ‘job done’. Defenders of unconscious bias training often say that it is not intended as the only solution but as part of a suite of actions to combat prejudice. What is clear, given the mixed evidence on its effectiveness, is that it should not be pitched as ‘best practice’ or the go-to solution.

If used, unconscious bias training must contain a very clear message that we each have a responsibility to try to overcome our prejudices. There must be no doubt of its sincerity - this is not a tick-box exercise done for compliance or image reasons. But if it were my money, I’d sooner invest time and resources on discussing inequalities than teaching the psychology of bias.

For more, see the CIPD’s evidence-based review, Diversity Management that Works.