Tuesday 21 July 2026 · The Home of Racing

SportsPlayUK

Loving Every Furlong
The Press Room

One Fair Refusal, Five Years Of Something Else

The Gambling Commission's evidence problem

Racing asked to see the evidence supporting affordability checks, but was refused. The refusal was fair. Then there is the survey the Gambling Commission buried, the black market it waved away, and the harm figures its own statistics regulator queried.

When I wrote about prize money earlier this month, I held back on two of the culprits. Not because the case against them was thin, but because it was too big to bolt onto another argument. Now I can go into more depth about the two bodies that hold more sway over racing’s future than any owner, trainer or racecourse: the Government and the Gambling Commission. This year, both have shown, publicly and on record, how little they understand the sport they are shaping.

There is a deadline attached to this one. The Culture, Media and Sport Committee has given the Commission until 24 July to answer five questions. What follows is why those questions exist, and why racing could not get answers to them on its own.

I will set out what has happened first, plainly, so you can weigh it yourself. My own view I have kept for the end.

The levy review

In March, before the Easter recess, the Government closed its review of the Horserace Betting Levy with a written ministerial statement. The rate stays at ten per cent. There will be no reform.

Two years earlier, the picture had looked different. In May 2024, racing and betting reached a deal that would have lifted the levy to 11.5 per cent, with one per cent ringfenced to market and promote the sport. The early general election killed it before it could be enacted. What racing thought it had won, it then had to argue for all over again. This time it lost.

The levy raised £108 million in 2025. Brant Dunshea called it disappointing that it took almost three years to decide on no change at all. He also pointed to something the Government has never explained. Before the Budget, the DCMS told the Treasury that racing would not benefit from a betting tax carve-out unless the levy rose in line with it. Months later, the same department decided the levy did not need to move. Both cannot be true.

The tax carve-out

There was one piece of good news, and it deserves to be stated clearly. In the November 2025 Budget, the Government introduced a new 25 per cent duty on remote betting from April 2027, while keeping bets on British racing at 15 per cent. Other sports pay the higher rate. Racing does not.

That exemption was not a gift. Racing won it with a one-day shutdown of fixtures in September under the #AxeTheRacingTax banner and months of lobbying behind it. But the Treasury’s own framing takes the shine off. HMRC argues that betting on racing already sits at an effective 25 per cent once the ten per cent levy is added to the 15 per cent duty, which is precisely why it left the duty alone. So the carve-out racing fought for is, in the Government’s eyes, no concession at all. It is the status quo, described generously.

The minister

The gambling minister is Baroness Twycross. On 29 April, in her Parliament Street office, she met Brant Dunshea and Greg Swift of the BHA alongside Grainne Hurst of the Betting and Gaming Council. It was racing’s last real chance to change her mind.

What makes that meeting worth recording is what happened the Friday before it. As Lee Mottershead reported in the Racing Post, the same arguments had been put to DCMS officials, who accepted that fair points had been made. By the time racing sat down with the minister, those points had either not reached her or had not detained her. Mottershead described her as reliant on the paper in her hands and answering with lines lifted from the Gambling Commission’s own script.

Racing also reminded her of a promise. Labour had pledged to adopt the previous Conservative administration’s approach to affordability checks, and the Conservative red line was that the checks must be fully frictionless. We now know they cannot be. The pledge went with it, and a policy that ought to have been tested through secondary legislation was instead left to an unelected quango.

Her department did not respond to the cross-party group of MPs who asked for the checks to be paused. When the Commission confirmed them on 7 July, Twycross welcomed the decision and said the assessments would be brought in “in a careful, phased way”. She welcomed a policy her own department had been asked more than once to pause, weeks after her own former adviser had resigned over it.

The secretary of state

Above Twycross sits Lisa Nandy, and Nandy has been absent from this story altogether. Mottershead reports she has never sat down with the BHA. Her Conservative predecessor, Lucy Frazer, was no expert on racing either, to the point that Julie Harrington drew her a chart explaining how the sport is funded, but Frazer at least met the sport’s leaders. Racing is Britain’s second-most-attended spectator sport and supports 85,000 jobs, directly and indirectly. UK basketball has had the meeting that racing is still waiting for.

Affordability checks confirmed

On 7 July, the Gambling Commission confirmed that it will introduce financial risk assessments. The BHA called the decision one of “self-harm on an immense scale”, and Dunshea called the Government’s refusal to take the decision itself a clear abdication of duty.

The thresholds matter, and they have moved. Implementation begins in phases: assessments first apply to customers over 25 at the largest operators who deposit more than £5,000 net in a rolling 24 hours. At full implementation, the trigger drops to £1,000 over 24 hours, or £3,000 over 90 days. During the early stages, the Commission will not take enforcement action against operators who fail to act on the findings of an assessment.

The cost estimates are not the BHA’s invention. Independent consultants Regulus Partners modelled the hit to racing at £250 million over five years. The Betting and Gaming Council estimates 120,000 racing punters could face checks, that around 96,000 would refuse to hand over documents, and that the levy would lose roughly £13 million a year as a result.

The Commission’s case is that the checks are proportionate. Its pilot, it says, showed 97 per cent of flagged accounts could be handled without friction, and that only around half a per cent of customers would reach the threshold. It also rejects the term everyone else uses. This is not an affordability check, the Commission says, because it makes no assessment of income or of what a customer can afford, and there are no proposals or plans to introduce affordability checks.

Both figures come from the Commission. Neither has been independently tested, because the independent test has not been published.

The adviser resigns

In May, Dr James Noyes resigned from the Gambling Act Review Evaluation Advisory Group, the panel advising the Government and the Commission on the white paper. Noyes was not opposed to affordability checks. He was one of the first to argue for them. He resigned anyway, calling it “clearly unacceptable” to roll the policy out before any meaningful, independent evaluation.

His specific objections deserve noting. The data used to set the spending thresholds, he said, was outdated and eroded by inflation. Credit reference agencies were returning different results for the same customer, which meant the checks could not be the frictionless process punters had been promised. The measures would, in his words, be extremely detrimental to racing.

His departure exposed something else. Most people in racing did not know the advisory group existed. The BHA, which had engaged with the DCMS for years, said the group had never been mentioned to it, not even in answers to parliamentary questions. Its full membership will not be published until after its findings are out; a clear example of throwing the baby out with the bath water. When the man who helped invent the policy walks away, calling the process unacceptable, that is worth more than a footnote.

The missing evidence

The independent evaluation of the pilot, run by NatCen, has not been published. There has been no public update on it since May 2025. The Commission is holding its full consultation response until the autumn, after the checks have already been signed off. Members of its own implementation groups will not see the evidence either. Asked at a press briefing why it had not published, the Commission’s director of major policy projects and evaluation, Helen Rhodes, said a great deal of the pilot information was already out, and more would follow in the autumn, though some of it would be summarised because of commercial sensitivities involving the credit reference agencies and operators.

There is a structural problem underneath this. NatCen was evaluating the white paper for the Government at the same time as it was working as an evaluation partner on the pilot. And the advisory group, the body meant to provide external scrutiny, was told it could not advise on financial risk assessments because those had not yet been implemented. So the single most contested measure in the entire review sat outside the remit of the group set up to scrutinise it.

The FOI request, and why it failed

By the end of April, the BHA had run out of patience. It took the extraordinary step of using the Freedom of Information Act against a fellow regulator, requesting access to correspondence between the Commission, the DCMS and NatCen.

The Commission refused it as vexatious on 1 June, and racing reacted with fury. Having read the letter rather than the coverage of it, I think racing is wrong about this one, and the reason matters.

Look at what was actually asked for. The request covers all correspondence, documents, and communications relating to affordability checks, expressly including emails, letters, briefing papers, WhatsApp, SMS, Teams, and Slack messages, agendas, minutes, notes, readouts, submissions, and internal analysis, running from 1 January 2020 to 23 April 2026. Six years. Every mailbox. Every messaging platform.

Under section 14(1), a public body can refuse a request where the burden of answering it would be grossly oppressive, and that is exactly the ground the Commission used. It said it would have to review large volumes of material across several systems, that potentially exempt information was scattered throughout and could not be easily isolated, and that diverting staff would interfere with its other work. It confirmed that the information exists. It noted that it was under no legal obligation to explain its reasoning and chose to do so anyway.

Racing’s counterargument, put forward by Greg Swift, is that the Commission cannot characterise the request as vexatious while also characterising the subject as a live matter of public interest. It is a good line, and it does not work. Vexatious under section 14(1) is a term of art about burden, not about impertinence, and the ICO’s own guidance requires public interest to be weighed against burden rather than to defeat it. Both statements can sit in the same letter.

So the BHA drafted a request so wide that the Commission barely had to think about refusing it. Ask for six years of everyone’s WhatsApp, and you hand the other side an easy out. The Commission took it. That is not a regulator stonewalling. That is a regulator being handed a free pass by an opponent who reached for everything instead of the one thing that mattered.

The resubmission and the committee

Then two things happened that changed the picture.

The BHA went back and resubmitted a narrower request. This is the move that counts. Strip out the six-year sweep and the Slack messages, ask instead for the NatCen evaluation of the pilot and the analysis that went to the board, and the burden defence evaporates. A tight request for a single finished document cannot be grossly oppressive to answer. It is one report. The Commission either has it or does not, and it has already confirmed it does.

That is why the resubmission puts the Commission on the back foot rather than the BHA. The first request let it argue about volume. The second one forces it to argue about substance, and substance is the argument it has spent three years avoiding.

Then, on 10 July, the Culture, Media and Sport Committee wrote to acting chief executive Sarah Gardner, posing five questions and setting a deadline of 24 July.

The first asks whether the Commission will publish the full dataset, the evidence base, and the methodology behind its decision and its thresholds. The second asks it to estimate whether more or fewer recreational bettors will be asked for documents than under the arrangements running today. The third asks it to set out its engagement in detail, following stakeholders’ telling the committee that engagement had been insufficient.

Read those against everything above. They are racing’s questions, almost word-for-word, asked by a cross-party committee of MPs rather than by an interested party. There is no burden exemption against a select committee. The Commission has confirmed it will respond by the deadline.

The 2021 consultation

None of this is new. In 2021, the Commission ran a survey on affordability checks that drew around 12,000 responses. I know because I was one of them. More than 77 per cent disagreed with the idea. Two-thirds were uncomfortable with a company accessing their financial information. Those results did not appear in the 2023 white paper. The Commission refused a freedom of information request for them in 2023, saying there was no public interest in releasing them at that point. Andrew Rhodes, then chief executive, promised publication the following year. What appeared was a summary of a wide range of views, without indicating where the weight of opinion lay. The full figures only surfaced after a further FOI request, by which point the white paper was already written. A separate piece of customer research, carried out in 2019 and also showing opposition to hard interventions, also went missing from the process.

Note the difference from the BHA’s request. Nobody could call a survey the Commission itself commissioned a grossly oppressive burden to publish. It had the results. It sat on them until they could no longer influence anything.

One more, because it bears directly on the checks. In 2023, the Commission’s then chief executive told MPs that the threat from the black market was overstated and that its size was very small. In 2025, the Government handed the Commission £26 million over three years to tackle that same black market. A risk cannot be both trivial and urgent at the same time.

The statistics

Then there is the evidence the Commission does put out. Its Gambling Survey for Great Britain, launched in 2023 and treated as official statistics since early 2024, puts the rate of problem gambling at 2.7 per cent of adults, about 1.4 million people. The NHS health surveys have long put it closer to 0.4 per cent. The Office for Statistics Regulation reviewed the survey, made nine recommendations, and noted the Commission’s own admission that the figures may overstate the true level of harm. Independent analysis by Regulus Partners, comparing the survey against real operator records, found it overstated participation in some activities by as much as 694 per cent.

In October 2025, the Commission removed its earlier advice against scaling those figures up to the whole population. A More in Common poll in January found that just one in three adults trusts the Commission to regulate betting properly. On this record, one in three feels high.

My Take

Here is my verdict, and the facts above have earned it.

Let me start by giving the Commission what it deserves. The vexatious refusal was fair. The BHA asked for six years of every email, every Slack message and every WhatsApp across an entire organisation, and any public body in the country would have refused it on grounds of burden. Racing lost that round because racing swung wildly. I say that because the rest of what follows should be read by someone who knows I will concede a point when the evidence says so.

Now the rest.

Look at what the Commission protects and what it burdens, over the years rather than in one letter. In 2021, it ran a survey, got an answer it did not want, and held it back until the white paper was written and the moment for influence had passed. A survey it had already paid for and already had in its hands. No burden argument was available there. It just chose not to publish. Research from 2019 pointing the same way went missing too. In 2023, it told MPs the black market was too small to worry about, and by 2025 the Government was handing it £26 million to fight the thing it had waved away. Its own harm figures sit six times above the NHS estimate, its own statistics regulator flagged the risk of overstatement, and its response in October was to make those figures easier to quote, not harder.

That is not one bad decision. That is a direction of travel, running the same way for five years, and every deviation lands on the same side.

Now set the pattern against the promise. The Commission says its checks are 97 per cent frictionless and touch half a per cent of customers. Both numbers are its own. The independent evaluation that would test them is unpublished and sits with NatCen. The consultation response is scheduled for the autumn, safely after sign-off. Its own implementation groups are working in the dark. You cannot ask a country to accept a number and decline to show the working, and you especially cannot do it with a record like this one behind you.

Judge the promise by the record. Here is mine. Two years ago I had accounts closed within a month by two different bookmakers for refusing to share documents. I had held both for between five and ten years. They wanted payslips, bank statements, and my passport or driving licence, from a customer of that standing. I refused. They offered to keep the accounts open with my stakes capped at £1. I refused again, and I no longer have accounts with either. Both were in profit. I had won more than I had lost. I walked away on principle.

I have been following horse racing since I was nine or ten, standing next to my grandad. That is over forty years of loving this sport with no issue of problem gambling. The point of the story is not me. It is that the checks people were promised would be light-touch and rare had already reached an ordinary punter betting modest sums, two years before the Commission signed them off.

This is not new either. Back in 2022, the Commission’s guidance already expected operators to ask higher-spending customers for three months’ payslips, P60s, tax returns or bank statements. When the Racing Post put that to the Commission, it confirmed that was its position. A gambling lawyer who spent nine years inside the Commission, David Whyte of Harris Hagan, put it plainly at the time: the regulator was treating everyone who occasionally bet above its comfort level as vulnerable, and had no legislative mandate for imposing that on the overwhelming majority who showed no sign of harm. Bookmakers acted on the steer and applied it defensively, out of fear of enforcement. Two years ago, it reached me. Today, the Commission says it never mandated affordability checks and has no plans to introduce them. Both cannot be true, and the paper trail runs against the version it is telling now.

Then ask the man who invented the policy. Noyes did not resign because he opposed affordability checks. He resigned because credit reference agencies were returning different answers for the same person, which is precisely what makes a check unfrictionless, and because the thresholds rested on data that inflation had already overtaken. When your own architect says the building will not stand, the burden is not on the rest of us to prove him right.

Then there is the proportion of the thing. By the Commission’s own reckoning, this regime bites on half a per cent of customers, and by its own survey, generous as that survey is, the overwhelming majority of the millions who bet each week do so within their means and come to no harm. To reach a troubled few, the Commission is asking racing to absorb £250 million and asking ordinary punters to hand over payslips and bank statements to keep doing something perfectly legal. That is not a system designed around risk. It is a system designed around suspicion.

So I will say it plainly. The Gambling Commission has been acting in bad faith since the 2023 white paper. Not in that FOI letter, which was defensible, but in the five years of choices around it, every one of which buried, delayed or inflated the evidence in the same direction. It treats every punter as a problem gambler in waiting and racing as collateral it is willing to lose, in the name of a minority who genuinely need help and would be far better served by targeted support than by blanket suspicion.

The Government’s failure is quieter, and no less real. It walked away from a levy deal. It dressed up a worthless carve-out as a favour. It broke its own pledge to make checks frictionless before imposing them. It ignored a cross-party call to pause. It let its own adviser resign rather than listen to him. Its secretary of state has never met the sport. Then the minister welcomed the result and talked about careful, phased implementation. When your own expert walks over a policy, you are not making an informed decision. You are making one in the dark and calling it consumer protection.

Which is what makes the next four days worth watching. The BHA has learned its lesson and asked again, narrowly, for the one document that matters. The select committee has asked the same question, with Parliament’s authority behind it and a date attached. Neither can be answered with a burden argument. Neither can be answered in a blog post. The Commission can publish the evidence, or it can explain to a cross-party committee of MPs why it will not.

Racing has spent five years being told to trust the process. On 24 July, we will find out whether the process can survive a straight question.

Never miss a lesson

New guides, previews and history every week. Free to your inbox.

Join the mailing list