Horsfield Menzies : Blog vs Behaviour – DANIEL RUBIN : THE TRAINING WORKED. THAT’S THE FUCKING PROBLEM

By /

Daniel Rubin once asked a sensible question: does unconscious bias training actually work? The Government had decided to scrap the training for civil servants, citing weak evidence that it changed behaviour. Rubin was not particularly impressed by the implication that this made the underlying problem disappear. He acknowledged the evidence problem but argued that attitudes buried deep in the subconscious were hardly going to evaporate because somebody watched a PowerPoint, clicked NEXT seventeen times and collected a certificate before lunch.

More interestingly, Rubin went further. He thought the reaction dismissing unconscious-bias training as “political correctness” or “woke” suggested that “something is very wrong” either with the training itself or with the culture of the workplace receiving it.

Thank you, Daniel.

Because that gives us a considerably nastier test than merely asking whether somebody attended the fucking course. It asks what happened once they understood the lesson.

The Ignorance Defence Has Left The Building

Horsfield Menzies cannot plausibly present itself as an organisation unfamiliar with discrimination theory. Its own website has become a small library of the stuff. Daniel Rubin has written about unconscious bias in recruitment, warning about subjective “chemistry” and the tendency to favour “people like us”. Marcia Sumbe has written about bias, comparators, evidence and consistent treatment. Elsewhere, the firm has published material about mental health, disability, equality, investigations, harassment, sickness absence and sensitive information.

These are not concepts Horsfield Menzies occasionally bumps into. They are content, expertise and part of what the firm sells. Its public presentation goes beyond WE KNOW EMPLOYMENT LAW and offers something more flattering: these people understand workplaces, fairness, human behaviour and the subtler ways organisations can fuck people over without necessarily recognising what they are doing.

So I do not need anybody’s unconscious-bias training certificate.

The website is the certificate.

More importantly, they knew the vocabulary.

Knowing Better Is The Interesting Part

That changes the nature of Blog vs Behaviour. If somebody has never seriously considered unconscious bias, an ugly piece of behaviour might at least be explained partly by ignorance. There is a blind spot and the person does not recognise that it exists.

Rubin removes that possibility himself. He understands that bias can operate beneath conscious intention, and that people may sincerely believe they are acting neutrally while assumptions influence the result. Crucially, his article also recognises that awareness means very little unless something eventually changes in the culture and decision-making around it.

This is where the usual corporate diversity routine becomes dangerous. Picture the training room: coffee in cardboard cups, somebody from HR explaining affinity bias, a slide containing several smiling stock-photo employees of improbable enthusiasm and a quiz at the end. Everyone agrees diversity is important. Seventy-five per cent gets you the little green tick.

Congratulations. You are now aware of unconscious bias.

Monday morning, however, still arrives. The difficult person remains difficult. The unfamiliar person still feels unfamiliar. A distressed person continues communicating in ways the professionals around them dislike. Although the PowerPoint has completed its task, the important question remains whether the culture has done anything with it.

That is Rubin’s own fucking test.

The “People Like Us” Test Gets Real

My Cepac litigation supplied a useful practical version. I was not much like the professionals surrounding the case: an unemployed disabled litigant in person dealing with serious mental-health difficulties, representing myself against corporate respondents with professional solicitors. I could be angry, distrustful and profane. Under sustained pressure, my correspondence could become excessive and, as my confidence in the process deteriorated, I pushed harder rather than becoming quieter.

Fuck, I even swore.

None of that makes every reaction admirable. Some were not. Nor does it make the underlying issue disappear.

Rubin already knew the intellectual problem because he had written about “people like us” in recruitment. Human beings tend to respond more comfortably to people who resemble norms they already understand. Familiar communication feels reasonable, familiar behaviour feels safer and difference requires more interpretative effort. Consequently, unfamiliar behaviour can acquire negative meaning before anybody consciously decides they dislike the person because of a protected characteristic.

Move that insight out of the interview room and into litigation. When the disabled claimant does not behave like the lawyers, does distress still look like distress or does it begin to look like misconduct? Does persistence resemble somebody desperately trying to get an issue heard, or does it become vexatiousness? When unfamiliar communication appears, does it trigger curiosity about context, or does everyone start building a fucking bundle?

Those questions matter precisely because Rubin knows why assumptions matter.

A Simple Claim And An Enormous Claimant

The underlying Cepac dispute was relatively compact. Page Outsourcing telephoned me on 21 August 2024. I say an interview was offered for after the bank holiday, we discussed the training salary and Page asked about the gap in my employment history. During that conversation, I disclosed that the gap related to my mental health. A contemporaneous WhatsApp recorded “Interview next week” and “27k training wage up towards 29”.

Afterwards, the expected confirmation disappeared and Cepac denied that an interview had been offered. Employment Judge T. R. Smith subsequently described the proceedings as a “simple claim”, adding that I was essentially upset because I had not got a job. Complete fucking bollocks. By then, the Tribunal file stood at approximately 1,941 pages.

The recruitment call had not become 1,941 pages complicated. Instead, the claimant had become extraordinarily interesting.

Emails accumulated alongside complaints. Medical evidence became contested territory, previous litigation arrived, TCAP entered the bundles and posts on X followed. Language, reactions and behaviour increasingly became evidence from which a portrait of the claimant could be constructed. Eventually, that conduct case became powerful enough to dispose of the discrimination claim before the original recruitment dispute received its completed merits determination.

Now return to Rubin. This is the man who had already written about the danger of subjective judgment attaching significance to people who do not comfortably resemble “people like us”. You might therefore expect that insight to produce a degree of intellectual caution when the awkward outsider becomes increasingly central to the case.

I saw very little.

Awareness Training Is Not An Exorcism

Rubin’s unconscious-bias article is useful precisely because he does not pretend awareness magically cures anything. These attitudes, he says, can sit deep in the subconscious, and changing them may require time and repetition. In other words, knowing the phrase unconscious bias does not transform somebody into an unbiased human being.

The same applies to institutions. A law firm can write about equality without becoming equal, publish mental-health articles without responding particularly well to distressed people and advise clients about investigations without investigating itself impressively. Likewise, endless discussion of bias does not prevent an organisation becoming remarkably comfortable with its own assumptions.

In fact, the vocabulary can make the performance more convincing. Once everybody knows the correct language, institutional behaviour acquires a flattering narration. Nobody is disregarding disability; they are managing procedure. Nobody is reacting negatively to difference; they are addressing conduct. Nobody is overwhelming a litigant; they are protecting the client’s position.

Every sharp edge gets a professional label.

The terminology changes while the pressure remains.

Daniel’s Own Culture Test

This is where Rubin’s short article becomes considerably more damaging than he probably intended. His diagnostic is straightforward: if people reject the whole subject as political correctness or woke nonsense, something may be wrong with either the training or the workplace culture.

Horsfield Menzies passes the awareness part magnificently because nobody there appears to be rejecting the vocabulary. Quite the fucking opposite. They publish it, market it and advise other employers about it.

This is not a firm where somebody needs to explain for the first time that unconscious bias exists. Horsfield Menzies has enough material about equality and good employment practice to run the training itself, invoice somebody for it and publish a follow-up blog before the biscuits have gone stale.

So suppose those principles do not survive contact with an inconvenient disabled opponent. What excuse remains? Missing training is hardly convincing. Neither is the suggestion that the concepts were too difficult or nobody understood the risks.

Daniel fucking Rubin did.

His own article therefore leaves another possibility sitting on the table.

Culture.

Values Integral To The Business

That word becomes particularly entertaining because Horsfield Menzies uses it about itself. The firm says its values are integral to the business because they help define its culture and distinguish Horsfield Menzies from other firms.

Wonderful.

Now we have something measurable.

Culture is not the wording on a values page. If it were, every business with INTEGRITY written in reception would qualify automatically and corporate ethics could be outsourced to whoever supplies the vinyl lettering. Instead, culture reveals itself when an organisation faces something inconvenient.

How does it react to an awkward complainant? What happens when somebody challenges one of its own people? Does curiosity survive hostility? Can awareness of disability alter the interpretation of difficult behaviour, or does protecting the client eventually consume every competing instinct?

Those are culture questions.

Another fucking webinar will not answer them.

The Training Meets The Complaints Department

Rubin eventually received an even cleaner practical test. On 23 March 2025, I sent a detailed private complaint about Senior Associate Sam Butler directly to Angus Menzies, Simon Horsfield and Daniel Rubin. The allegations concerned disability-related medical information, contact involving my medical centre, disputed litigation statements, aggressive costs pressure and correspondence I considered obstructive.

That presented an opportunity to demonstrate actual culture rather than website culture. A person the firm already regarded as difficult had complained about one of its own solicitors. The relationship was hostile, the litigation remained live and every incentive encouraged Horsfield Menzies to interpret another message from me as yet more irritating correspondence from an irritating opponent.

Which is exactly when assumptions become interesting.

The comfortable response was dismissal. By contrast, the culturally difficult response was curiosity.

Maybe Butler had done nothing wrong. Perhaps I had misunderstood everything. A meaningful investigation could have established precisely that and explained why.

What came back was not a meaningful or constructive investigation outcome but, if memory serves, little more than a couple of lines from Rubin amounting to a flat denial. No engagement with the substance. No apparent attempt to investigate what I had actually raised. Just the door closing.

Daniel had written about culture.

Then culture knocked.

The Reputation Man

Rubin’s professional profile adds another useful layer. Horsfield Menzies presents him as a pragmatic employment lawyer acting mainly for large corporate clients, with expertise spanning sensitive disputes, board-level matters, reputation management and investor confidence.

That tells us what Daniel is professionally good at.

When something threatens a corporate organisation, he thinks strategically about exposure, risk and outcome. Clients pay experienced employment lawyers to do exactly that, and there is nothing inherently improper about it.

However, reputation management and unconscious-bias awareness make an interesting combination. One asks what assumptions might be distorting our perception of a person; the other asks what a problem means for the client. Sometimes those questions coexist happily. At other times, I suspect they pull in rather different directions.

In the Cepac case, the reputational gravity eventually shifted dramatically away from the disputed recruitment decision and towards the claimant’s behaviour. The claim ended through strike-out, followed by a £20,000 costs order, without the recruitment merits receiving their completed determination.

I cannot tell you Rubin designed every step towards that outcome.

I can tell you which side the reputation man represented when the reputational problem changed fucking owners.

The Danger Of Becoming Very Well Trained

There is a darker possibility hiding inside corporate ethics training: knowing the language can become mistaken for possessing the virtue. An organisation teaches unconscious bias and therefore regards itself as progressive. It has a mental-health policy and consequently sees itself as compassionate. Managers attend equality training, so discriminatory thinking becomes something less enlightened organisations do.

Awareness becomes an alibi.

Rubin’s own article comes close to recognising the danger because he understands that awareness alone does not guarantee behavioural change. Accordingly, the useful measurement is not whether somebody can define bias after training. What matters is what happens afterwards.

That is why Horsfield Menzies makes such a useful specimen. The firm’s public output suggests considerable awareness and its lawyers are plainly intelligent people who understand discrimination professionally. More than that, they repeatedly publish about the behavioural concepts underneath it.

If the behavioural record still looks ugly, more vocabulary is not the answer.

You cannot train somebody out of ignorance they no longer possess. Eventually, choices become more interesting than awareness. Incentives start mattering more than terminology and, as Daniel himself suggested, attention turns towards the fucking culture.

This Was Never About A PowerPoint

The nastiest reading of Rubin’s article is therefore not that unconscious-bias training failed at Horsfield Menzies. I have no idea what formal training any particular individual there completed, and it hardly matters.

The intellectual training plainly landed.

Rubin understands unconscious bias well enough to publish about it. He understands recruitment bias well enough to identify “chemistry” and “people like us” as danger zones. Sumbe understands comparators and evidential consistency. Elsewhere, the firm demonstrates enough theoretical knowledge about disability, investigations, mental health and workplace culture to sell expertise on all of them.

Nobody needs another definition.

The gap is between understanding and behaviour.

That is considerably harder to fix because it does not live inside a training module. There is no RETAKE QUIZ button for professional instinct, commercial loyalty or the way an organisation behaves when somebody it dislikes becomes inconvenient.

At that point, the PowerPoint is innocent.

Does Unconscious Bias Training Work?

So let us answer Daniel’s question.

The evidence may be mixed. Rubin knew that when he wrote the article, believed awareness required repetition and worried that dismissing the subject entirely could reveal something unhealthy about workplace culture.

Fair enough.

Horsfield Menzies has certainly had the repetition. Bias, equality, diversity, mental health, investigations, fairness, evidence and complaints recur throughout its public output. Again and again, people inside the firm have published the correct concepts in the correct order.

Then came a disabled litigant who was angry, unfamiliar, difficult, unrepresented and increasingly easy to regard as the problem.

The firm did not lack awareness.

That is what makes the record sting.

Daniel Rubin asked whether unconscious-bias training works. In the case of Horsfield Menzies, I think the more uncomfortable possibility is that it worked perfectly well.

They learned the words.

They understood the theory.

And if the behaviour still failed the test, Daniel already told us where else to look.

The fucking culture.

Lee Thompson – Founder, The Cummins Accountability Project

Sources

Scroll to Top