How debunked stats and a selective approach to publishing research has misled the gambling reform debate

• Read part two of our special report here 

Last summer Gambling Commission chief executive Andrew Rhodes wrote an open letter to all sides of the gambling divide to warn them about the misuse of statistics in the often fierce debate on the subject.

"Much as everyone is entitled to present their arguments, what is wholly unacceptable is the misuse of statistics to support that argument," Rhodes said in the letter.

He concluded: "I therefore ask anyone commenting on this area to take a greater degree of care to ensure they are using evidence and statistics correctly, accurately and in the proper context and with any necessary caveats applied."

There are, however, observers of the industry who are critical of the way the Gambling Commission itself handles evidence and statistics, and this will become an issue again this week as the regulator prepares to publish the first annual report of the Gambling Survey for Great Britain (GSGB).

The GSGB will become the new source of official statistics for gambling, covering important issues such as problem gambling rates, and has the potential to open a new front in the debate over gambling with potentially negative consequences for punters and British racing.

There are concerns too about the reliability of the figures it will contain due to the methodology behind it.

One recent example that raised questions about the commission and its approach to crucial evidence is the publication after three years of the results of a short survey carried out by the regulator on attitudes towards affordability checks.

The information was revealed, in the Racing Post, only following a freedom of information (FOI) request – despite repeated calls for it to be published.

Gambling Commission chief Andrew Rhodes said misuse of gambling statistics were "wholly unacceptable", but his organisation's own use of evidence has been called into question

Two years ago Rhodes was told by Julian Knight MP, the then chair of the Culture, Media and Sport select committee, that it was "very strange" the survey's information was not made publicly available. Yet even after that hearing it remained under wraps.

Last year the Gambling Commission turned down a previous FOI request made by industry analysts Regulus Partners for disclosure of the survey results, claiming there was no public interest in releasing the information at that point and it would be published in due course.

Then, in February, three years after the survey, Rhodes said it would be published alongside the response to last year's consultation on financial risk checks. Instead, it emerged as just a summary which dealt with responses in broad terms, saying they contained "a wide range of views" but not detailing the weight of opinion behind those views.

When the results were finally revealed following a further FOI request they showed widespread opposition to affordability checks with, for example, more than 77 per cent of the 12,000 respondents saying they did not agree with the concept.

This was important as many felt the survey results should have been used to inform the government's gambling white paper and the financial risk checks consultation that followed.

It was not the only information omitted from that process by the Gambling Commission.

A survey of customer attitudes conducted by market research firm 2CV for the Gambling Commission in 2019 was also missing.

That piece of work had also found opposition towards hard interventions by operators, with more than 70 per cent of respondents rejecting controls.

In its response to the 2023 consultation, Arena Racing Company said of the 2CV research: "For reasons that are unclear, this evidence has not been included in this consultation and was not included in the call for evidence on affordability checks."

It would not take a hardened cynic to wonder why the evidence was not included and come to the conclusion that it was inconvenient for the Gambling Commission. At the very least, its approach was selective.

The regulator did this year publish the results of consumer research carried out by polling firm Yonder to gauge attitudes towards the financial risk checks included in the last government's gambling white paper.

In total contrast to the previous surveys, Yonder found, to use the Gambling Commission's words, "high levels of overall support for the proposals", as well as "a recognition that the proposals are necessary to protect online gamblers from harm".

Why should that be?

The commission did say that Yonder was "concerned that the volume of specialist media coverage that the Financial Risk Assessment consultation had commanded might bias response from some types of respondents".

It added: "Accordingly, Yonder developed research materials for each of the proposals."

In addition to that, the commission created what it described as a series of "short, informational videos, providing core information".

It said the videos were "intended to ensure respondents’ answers were well informed".

Respondents were shown articles about affordability checks from the Racing Post – but they were set against emotive articles from the mainstream press, often mentioning affordability checks only briefly.

The GSGB will also produce estimates of a wide range of harms, including suicide attempts, one of the most sensitive yet most commonly cited statistics used in the debate about gambling.

Only this month, an article in the New Statesman magazine claimed gambling addiction led to 400 suicides a year. It was the latest repetition of a claim made numerous times in recent years.

It was a figure used by Labour MP Paul Blomfield in the parliamentary debate over financial risk checks in February, when he said: "According to Public Health England, over 400 people take their lives each year as a result of gambling."

That brought an intervention from former Conservative MP Sir Philip Davies, who said the figure had "been discredited many times" and was one "with which the Gambling Commission certainly would not align itself".

The source of the figure came from a report by Public Health England (PHE) which claimed 409 suicides a year were "associated with problem gambling only".

In January last year it was succeeded by a report from the Office for Health Improvement and Disparities (OHID) which offered an estimate of between 117 or 496 suicides "associated with problem gambling".

Both the PHE and OHID estimates were based on a 2018 study of hospital patients in Sweden with a wide range of mental and physical health conditions, which found those who had been diagnosed with gambling disorder were 15 times more likely to commit suicide than the general population.

That ratio was applied to NHS Health Survey estimates of problem gambling to come up with a figure for gambling-related suicides in England.

However, the group in the Swedish study suffered from an array of conditions, placing them at elevated risk of self-harm regardless of the presence of gambling disorder.

The Swedish researchers concluded gambling disorder "did not appear to be a significant risk factor for the increase in suicide and general mortality when controlling for previously known risk factors"; but this finding was ignored by OHID.

Dan Waugh of analysts Regulus Partners has led the way in highlighting problems with the claims.

He said: "While people with gambling disorder are at elevated risk of self-harm, the figures bandied about by politicians and activists are based on junk science.

"They rely, for example, on the assumption there is no association between suicide and any mental health condition aside from gambling disorder."

There have been claims made about gambling self-harm that have been shown to be questionableCredit: Edward Whitaker

To return to Davies's point about the Gambling Commission, the industry regulator has been aware of the problems with suicide estimates for some time.

Documents obtained following an FOI request showed the commission carried out a review of the suicide claims as far back as April 2022, describing them at the time as not being based on "reliable data".

However, according to Waugh, FOI requests have found no evidence of the Gambling Commission challenging the OHID or informing the DCMS.

More recently, Rhodes has challenged the use of suicide statistics.

Last autumn he wrote to Dame Caroline Dinenage, then chair of the Culture, Media and Sport Committee which was holding an inquiry into gambling, to point out there are no "robust figures" for gambling-related suicide. Nevertheless, he still quoted the OHID estimates of 117 to 496 suicides.

There has also been criticism of the Gambling Commission for the way it has allocated some of the more-than-£90 million of regulatory settlement funds, raised from action against gambling operators, to a wide range of organisations.

A number of those organisations produced questionable research about gambling or could be described as being openly anti-gambling.

In one example, the Gambling Commission approved funding for a report by the National Institute for Economic and Social Research (NIESR), which claimed problem gambling was associated with direct fiscal costs to the public purse of £1.4 billion a year, a report Waugh described as “possibly the worst piece of analysis” produced as part of the Gambling Act review.

Waugh said that almost 60 per cent of the NIESR's cost estimate was based on a dataset that contained no information whatsoever about problem gambling.

He added: "NIESR invented mental health classifications out of thin air. It decided, for example, that anyone who had won £500 over a two-year period and was economically inactive due to ill health was an 'at risk' gambler.

"The scientific basis for this study was not simply slanted, it was non-existent."

The commission has approved a number of payments to the activist group Gambling With Lives (GwL), including a £350,350 payment to develop "education materials" for young people – awarded despite the fact the regulator has criticised GwL for misuse of statistics.

Asked for a response to these points, a Gambling Commission spokesperson said: "While it is not our role to police the debate around gambling, where we identify examples of people misusing official statistics we publish, we robustly challenge that, up to and including referring to the Office for Statistics Regulation."

The Gambling Commission has said in the past it wants to be known as the "trusted voice" on gambling.

Some of its actions in recent years have raised doubts over whether it is worthy of that trust.

Data published from the experimental stages of the GSGB has suggested it will produce a problem gambling rate several times greater than the previous figures produced by NHS Health Surveys.

The Gambling Commission has been warned the survey might "substantially" overstate the true level of gambling and gambling harm. However, it is set to publish that figure with that knowledge and in the likelihood it will reignite the debate about gambling – a debate fuelled by conclusions wrought from highly questionable statistics.

Read these next:

Gambling Commission's reluctance to publish affordability survey results should be concerning to all 

Revealed: Gambling Commission survey demonstrated considerable opposition to affordability checks 

'The whole thing hasn't gone away' - wary response to Gambling Commission's six-month pilot for affordability checks 

Subscribe to Racing Post Members' Club Ultimate Monthly and get 50% off your first three months! 

Available to new subscribers purchasing Members' Club Ultimate Monthly using code MEMBERS24. First three payments will be charged at £24.98, subscription renews at full monthly price thereafter. To cancel please contact us at least seven days before subscription is due to renew.