Gambling Commission director urges industry to be “fastidious” in its use of data

Gambling Commission executive director Tim Miller has bemoaned the misuse of official statistics and has suggested that some parties are deliberately misrepresenting data.

Speaking at the annual GambleAware conference, Miller stressed the importance for news outlets and others to be “fastidious” in their respective use of statistic-based evidence. 

The regulator has been ramping up its push against its deemed misuse of statistics, with Miller’s comments following a blog post from Gambling Commission director of research and statistics Ben Haden in which he said the body would take misuse “seriously”.

The Gambling Commission released its most recent edition of the Gambling Survey for Great Britain (GSGB) today, 5 December, with the regulator looking to clamp down on data from the survey being used “selectively”.

While Miller acknowledged that “no one number can be conclusive [and] no single data set will give you the whole picture”, he noted the GSGB was “important but just one part of the evidential jigsaw”.

Miller said: “Challenging the misuse of official statistics is sadly taking up more and more of our time.

“Our recently published data on children and young people was a prime example of that, where much of the coverage fell pretty wide of the mark.

“Sometimes the misuse is an accidental but well-intentioned act. A genuine attempt to write in an accessible way that ends up oversimplifying a necessarily complex and nuanced picture. 

“Sometimes it occurs due to a lack of care. An attention-grabbing headline that does not properly reflect the content of an otherwise well-considered article.” 

The executive director continued, highlighting that there has been several occasions where statistics have been used incorrectly with intent. 

“Other times it would appear to be a more deliberate misuse of data – where perhaps a strongly held belief is being cloaked in the selective use of the evidence base,” Miller added.

“Whatever the reason, it is vital that all of us are fastidious in the way we use evidence. That means using accurate definitions; it means not selectively picking and choosing data that supports a desired outcome; it means being open about where decisions are still needed to be taken despite evidence that may be inconclusive – what we would call the precautionary principle.”

Freedom of Information (FOI) requests published on the Gambling Commission website since September have detailed several instances in which the regulator has reached out to news outlets and individuals to request alterations to use of stats.

On 4 November, the regulator published redacted correspondence following an FOI request regarding the “misuse of official statistics” since 19 September.

The request revealed that the Gambling Commission had been in communication with several publications, including The Telegraph, Times, Daily Express and iNews.

In The Telegraph’s instance, the outlet was informed that it has misused statistics to “gross up” the prevalence of problem gambling in the UK, something the GSGB guidance warns against. 

The GSGB guidance reads: “The GSGB should not be used to gross up the prevalence of problem gambling or the consequences of gambling to whole population (until further work is completed).”

In the case of iNews, the publication had used GSGB data that estimates 2.5% of adults have a PGSI score of above eight and extrapolated it to deduce there are 1.3 million problem gamblers in the UK.

In the email sent to iNews, the Gambling Commission cited the GSGB guidance and asked for the 1.3 million figure to be removed.

The Times was contacted by the Gambling Commission after quoting the regulator as a source for its claim that one million adults and 55,000 children have an ‘out of control’ gambling addiction. The regulator made clear that there are no statistics published by itself on the rate of gambling addiction. 

Similarly, The Daily Express was advised that by comparing data within the GSGB to previous studies, it risked misleading readers and casting doubt on the survey entirely. 

Also included in the 19 September to 4 November FOI tranche was the Coalition to End Gambling Ads, in which the body had also used the 1.3 million problem gambling claim and grossed up stats regarding related harm experienced by people due to someone else’s gambling.

A previous FOI request, submitted on 19 September, had asked for correspondence regarding the misuse of stats in the year-to-date.

The Gambling Commission said it had been in touch with 15 individuals and organisations since 20 June, including The Guardian, Daily Mail online, Gambling with Lives and the University of Bristol.

A further FOI from 23 October requested correspondence between the Gambling Commission and the Betting and Gaming Council over the use of statistics.

The partially redacted emails show the Gambling Commission requesting the trade body to use the “official” GSGB figures moving forwards, instead of data from the NHS Health Survey.

The NHS Health Survey, which the BGC said was an “accredited statistic”, puts the rate of problem gambling in the UK at 0.4%.

The FOI request showed the BGC informed the Gambling Commission it would continue to use the NHS Health Survey stats as opposed to the GSGB dataset.

There was no suggestion from the Gambling Commission the BGC had misused data. The email from the Gambling Commission read: “We’ve noticed that BGC still refers to previous stats, it’s not a misuse of stat issue but we’d be keen for you to start using the official figure moving forwards.”

A spokesperson for the BGC, said: “The methodology used in this survey is different to those previously conducted into betting and gaming.

“As a result, the Gambling Commission has been clear, GSGB should not be used to make direct comparisons with results from prior surveys, as a measure of addiction, or to gross up problem gambling prevalence to the whole population.

“Our members are concerned these findings may be unreliable because there is a significant risk GSGB overstates gambling participation and gambling related harm. Surveys using predominantly online self-completion consistently produce higher estimates of gambling participation and associated harms, compared to established alternative survey methods.

“The BGC and our members are committed to raising standards and we welcome any robust study that accurately gauges betting and gaming participation and problem gambling prevalence.

“This commitment is backed by record funding, with BGC members voluntarily donating over £170m over the past four years to independent Research, Prevention and Treatment services to tackle problem gambling and gambling related harm.

“Each month around 22.5 million people in Britain enjoy a bet and the overwhelming majority do so safely and responsibly. The most recent NHS Health Survey for England estimated that 0.4% of the adult population are problem gamblers.”