← writing

Anti-racism law, the Nowak case, and Reform

This is a longer, opinionated piece rather than a news report. I've tried to be fair rather than neutral: to set out what the law actually says, where I think its application has drifted, and where I'm giving a view as opposed to a fact. The trigger for writing it is a single, awful case — the murder of Henry Nowak — but the argument underneath is older and bigger than any one death.

The laws we actually have

“Anti-racism law” in Britain isn't one statute; it's a layer cake built up over sixty years. It helps to know what each layer does, because most of the public argument blurs them together.

Read together, this is a genuinely impressive body of law, and it grew for good reasons. The Britain of colour bars, “no blacks, no Irish” signs and an unaccountable police response to Stephen Lawrence's killing was a worse country than today's. Almost nobody serious wants to return to it, and it is worth saying that plainly before criticising anything.

Where the line moved

Here's my actual claim. The core anti-discrimination statutes above — the ban on refusing someone a job or a home because of their race — are sound and worth keeping. What has drifted is the layer of duty, guidance and training stacked on top of them: the public-sector equality duty that tells every public body to actively advance equality, the perception-based recording that flowed from Macpherson, and the wider diversity-and-inclusion apparatus that grew up around both. That drift isn't purely extra-legal — some of it is written into the law, and the rest takes its cue from the law — which is why the tidy line between “good statute” and “bad culture” only takes you so far.

Once a racist incident is defined by perception rather than evidence, the recording threshold collapses to whatever a complainant asserts. That was a sensible corrective in 1999, when the danger was police ignoring racism. The side-effect, decades on, is a system trained to treat an allegation of racism as a fact to be acted on first and tested later — including non-crime “hate incidents” logged against people never charged with anything. The aggravated-offence regime adds a second pull in the same direction: because what someone says can convert an ordinary assault into a more serious crime, the words at the scene can end up carrying more immediate weight than the wounds. And around all of it sits a decade of diversity-and-inclusion training — unconscious-bias courses, “lived experience” treated as evidence, the habit of reading any disparity as proof of discrimination — that pushes the same way: believe the accusation, doubt the accused.

None of that is a problem when allegations are true. It becomes a catastrophe when an allegation is a lie — because the machine is tuned to believe it. Which brings us to Southampton.

Southampton

On the night of 3 December 2025, Henry Nowak — eighteen, a first-year accounting student at the University of Southampton — was stabbed five times in Portswood by Vickrum Singh Digwa, then twenty-three. Nowak died at the scene. At trial the prosecution's case was that Digwa was the aggressor; the jury convicted him of murder on 28 May 2026, and on 1 June Judge William Mousley KC jailed him for life with a minimum of 21 years. Digwa's mother was convicted of helping to hide the weapon.

What turned a horrific killing into a national argument was what happened before the ambulance. Digwa told the first officers that Nowak had racially abused him and knocked off his turban. The officers appear to have believed him. Body-worn footage later released shows the mortally wounded teenager handcuffed on the ground, saying he had been stabbed and could not breathe, and an officer responding to the effect of “don't think you have, mate.” The dying victim was treated, briefly, as the racist suspect. The sentencing judge later found that Nowak had said nothing racist at all; Digwa's account was a fabrication.

The Independent Office for Police Conduct is investigating the officers' conduct, and the National Police Chiefs' Council has said it will review its anti-racism guidance. Both are the right responses. My reading — and here I am giving an opinion — is that no statute literally told those officers to disbelieve a bleeding boy; the Equality Act has no clause that does that. But it would be too convenient to stop there and blame “culture” alone, as if the culture had fallen from the sky. The reflex — treat an on-the-spot accusation of racism as the most urgent fact in the scene — was trained over years by the duties, the definitions and the training that the law and the wider anti-racism apparatus put in place. The statute didn't command the failure, but it helped build the instinct behind it.

The “two-tier policing” problem

Almost immediately the case was absorbed into an existing slogan: two-tier policing — the claim that British forces go softer on some groups and harder on others, usually framed as leniency towards minority or left-coded causes and severity towards the white working class. The phrase had gained traction during the disorder that followed the Southport murders in the summer of 2024.

It is worth being careful here, because this is where heat overwhelms light. When the 2024 riot claims were examined — by the Commons Home Affairs Committee and by His Majesty's Inspectorate of Constabulary — the bodies concluded there was no deliberate, codified policy of treating white or right-coded suspects more harshly. I think that narrow finding is probably right: there is no secret instruction and no written two-tier rule, and the conspiratorial version of the charge falls apart on inspection. But that is a much smaller claim than it sounds. “No written rule” is not the same as “no pattern,” and there is evidence — in the recording guidance, the training, the disparity in who gets the benefit of the doubt, and cases like this one — of an instinct that leans one way. The slogan overreaches; the thing underneath it is real.

The Nowak footage is exactly what that looks like in the wild, and that is the political danger for the government. You can tell people the inquiries found no codified two-tier rule; it is much harder to tell them what they have just watched on their phones did not happen. After the verdict, Nigel Farage doubled down on the two-tier charge; Sir Keir Starmer, Baroness Lawrence and Sir Sadiq Khan pushed back hard, and the row drew comment even from Washington. The argument is no longer really about a statistic. It is about whose story the public believes the state is built to hear first.

Part of what gives the grievance its charge is a memory of contrast. When George Floyd died in 2020 after a Minneapolis police officer knelt on his neck for more than nine minutes — the officer, Derek Chauvin, was later convicted of murder — Britain — a different country, an ocean away — saw some of its largest demonstrations in years: a statue went into Bristol harbour, footballers took the knee for two seasons, and boardrooms and broadcasters reorganised themselves around the moment. Plenty of people who thought that response proportionate have looked at the reaction to Henry Nowak — a British victim, mistreated by British police, on British video — and concluded that the official capacity for outrage is selective. It is fair to say the comparison is imperfect: the cases differ in almost every particular, Floyd was killed by a serving officer, and the two countries carry very different histories. But a felt double standard does not have to be statistically exact to be politically real, and naming it is something Reform has done plainly while the larger parties have tended to look away.

What it does to Reform

This is where the case matters beyond its own tragedy — and it needs handling with some care. It would be easy, and wrong, to treat a teenager's death as a political gift. It would be just as wrong to wave away the anger it released as manufactured or merely opportunistic. The grievance underneath is legitimate and, on the facts here, reasonable: people watched the state mistreat a dying victim on the strength of an accusation that turned out to be a lie, and they were right to be appalled. Reform did not invent that feeling. What it has is the timing, the standing and the willingness to give it a voice that the larger parties have struggled to.

Start with the polling. Through the first half of 2026 Reform has been the largest party in voting-intention surveys, sitting around the high twenties — a YouGov poll at the turn of June put it on 27%, with Labour and the Conservatives tied near 18% and the rest of the field fragmented between the Greens and Liberal Democrats. In a five-way split, a party in the high twenties with a disciplined message can win a very large number of seats on a minority of the vote. Reform doesn't need a majority of Britons; it needs a plurality in the right places and a divided opposition, and at the moment it has both.

And this is no longer just polling, which is the part that should worry the other parties most. Reform has now banked the votes twice over. In May 2025 it won some 677 council seats and took outright control of ten councils — including Essex, which the Conservatives had run for a quarter of a century — its first time controlling any local authority at all. Then in May 2026, only weeks before the Nowak verdict, it went further: on the night it added around 1,450 councillors and seized control of fourteen more councils, while Labour lost control of dozens. Its national vote share actually edged down a little that year — a reminder that under first-past-the-post a concentrated, efficiently-spread lead converts into far more seats than a broader but flatter one, and that the headline win was in councils rather than raw votes. But seats are what take councils, and constituencies, and Parliament. A party can lead the polls and still be a protest balloon; a party that is actually winning the count, in real wards, two years running, is something else. The Nowak case did not create Reform's momentum — it landed on top of a machine that was already moving.

Now add the issue. Reform's standing pledge is to repeal the Equality Act 2010 — to scrap what it calls the “divisive notion of protected characteristics” and the diversity-and-inclusion apparatus around it — and replace it with something vaguer (a “Workplace Fairness Act” has been floated, though no draft exists). In the abstract that is a hard sell: most voters quite like the underlying protections and don't think about the Act at all. A case like Nowak's changes the frame. It lets Reform argue not against fairness but against a system that, in one televised instance, handcuffed the victim and believed the killer. That is a far easier story to tell on a doorstep, and it converts a wonkish, easily-caricatured policy into a visceral one.

For Labour the bind is acute, and it is structurally the same bind it has on immigration and on grooming-gang inquiries. Defend the police and the existing framework and you look as though you are explaining away footage the country has already judged; concede too much and you ratify Reform's premise and demoralise your own coalition. The Conservatives, who built much of this legal architecture and governed over the 2024 disorder, have even less room to claim the issue as their own. The result is that Reform gets to own a grievance that a large slice of the public feels is real, while its two larger rivals are left arguing about definitions.

“For better or worse” is the right hedge, and I mean both halves of it. For worse: a genuine reckoning — which is needed — risks being flattened into a slogan, and a real, partly legal problem gets answered with one blunt instrument, repealing the lot, when the honest fix is more surgical — the duties, the recording rules and the training, not the ban on discrimination itself. For better: a complacency that had hardened into orthodoxy — the assumption that the perception-first, accusation-first instinct was simply settled progress — is finally being forced into open argument, and that argument was overdue.

My own conclusion is unglamorous. The core ban on discrimination is worth defending; the duties, definitions and training stacked on top of it have overcorrected and need honest, surgical reform; and the political system is, at the moment, structurally incapable of doing that careful second thing without handing the whole subject to a party whose answer is the blunt first one — abolish the lot — instead. That gap is the opening Reform is walking through, and nothing in the current alignment of polling and events is working to close it.

Sources & notes

A note on how this page was made: I researched it with the help of AI tools and then checked the specifics against the reporting below. It is a personal essay — the framing and the opinions are my own, not a settled account — and any mistakes are mine. If you spot one, please let me know.