December 16, 2020 - 3:23pm

I had a phone call from a government minister, not long ago, which is not something that happens to me very often. Kemi Badenoch, the women and equalities minister, wanted to talk to me about unconscious bias training, because she’d read the two articles (1, 2) I’d written for UnHerd on the topic.

The short version of those articles is that while unconscious bias is probably a real thing, the training that is given to reduce it does not seem to work, by any reasonable definition of the word “work”. There is no standard of what counts as UBT, and what training exists doesn’t seem to make people behave in less prejudiced ways; what’s more, it has “potential for back-firing effects”. (Not my words, Lynn. The words of the Equality and Human Rights Commission.)

The UK Civil Service used to undergo unconscious bias training; now, following a review of the evidence commissioned by Badenoch, that training has been scrapped. I’d love to claim that my powerful writing has swayed government policy, but all I really did was put the minister in touch with some scientists who’ve researched the topic, and point to a few studies like the EHRC review above. There’s a good roundup of the evidence in this post by the scientist Patrick Forscher.

What makes me nervous is that this will fall into culture-war territory. It’s very hard to argue against measures to combat racism, even if those measures are ineffective. But I think this is a good thing. UBT is both an expensive waste of money and a cheap sticking plaster: you can spend millions on it without it having a real effect on prejudice; but it’s a damn sight less costly than the concrete steps which might actually improve minority representation in the workplace. It reminds me of what Helen Lewis (among others) calls “woke capitalism”.

Companies (or political parties, or civil service departments) can hold training sessions, fire employees for bad opinions on social media, light up their offices in rainbow flags. These are all cheap things that get good PR and help the bottom line. But it doesn’t achieve the stated goals. For example, “the only question I want to ask big companies who claim to be ‘empowering the female leaders of the future’,” says Lewis, “is this one: Do you have on-site child care? You can have all the summits and power breakfasts you want, but unless you address the real problems holding working parents back, then it’s all window dressing.”

That’s how I feel about UBT. It makes for good PR for companies (or political parties, or civil service departments). But it doesn’t improve outcomes for the people it’s supposed to help. The things that would help might be on-site creches and generous shared parental leave policies, or apprenticeships and training for underprivileged people, or preferential hiring for minority groups.

Stopping UBT at the civil service is good, because UBT doesn’t work. The next step, if I was in charge, would be finding out what measures actually do work, and introducing them. That is going to be more difficult and more expensive than ending UBT, but you’ve got to start somewhere.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers