The Gambling Commission (UKGC) has told NEXT.io that it does “not recognise” claims that it is about to knowingly publish inaccurate results in tomorrow’s Gambling Survey for Great Britain (GSGB).
Claims of sampling bias have been levelled at the GSGB in the past over its methodological approach.
The first edition was published in July 2024 and the release on 16 July will be the third. With that in mind, the commission has stood by the evolution of the research, its findings and how they stack up against alternative studies.
The commission told NEXT.io: “We recognise that all survey methodologies have strengths and limitations, and we have been transparent about those throughout the development of the GSGB.”
Professor Patrick Sturgis
In 2024, Professor Patrick Sturgis of LSE published an independent review of the first GSGB and found that the participants of the survey may have included a higher proportion of gamblers than is representative of the wider British population.
Because it would seem relevant to them, gamblers are perhaps more likely to respond to an explicitly gambling-related survey than non-gamblers, thereby potentially inflating participation and problem gambling figures – as Sturgis noted, this non-response bias introduces a “non-negligible risk that they substantially over-state the true level of gambling and gambling harm in the population.”
Later experiments, reported on in August 2025 by Sturgis and others from LSE and NatCen Social Research, tested this risk and found that mentioning gambling explicitly in the survey invitation did lead to a 4% increase in reported gambling participation.
The amount of people recording a PGSI score above zero was 1.8% more in those part of the ‘gambling-invitation group,’ though Sturgis states that “this difference did not reach statistical significance.”
In August 2025, the UKGC committed to implementing some of Sturgis’ recommendations and a spokesperson has made reference to the academic and his support, saying: “The GSGB is now the Commission’s designated official statistic. It is based on responses from around 20,000 adults each year and has been independently reviewed by Professor Patrick Sturgis, who concluded that moving to this methodology was the correct decision given the changing survey landscape.”
The Commission also took the opportunity to defend the transparency with which the UKGC has conducted itself during the research process, saying: “Throughout the development of the GSGB we have published detailed technical information, been open about the survey’s strengths and limitations, and engaged extensively with stakeholders and independent experts. The survey methodology has evolved through pilot and experimental phases before becoming the Commission’s designated official statistic.”
Problem gambling in 0.7% or 2.7% of the population?
Critics of the UKGC’s research, including the Betting and Gaming Council (BGC), prefer to cite the most recent problem gambling figures from the NHS Health Survey, which indicates that 0.7% of the population suffer from problem gambling.
There is a clear discrepancy between that figure and the findings of the most recent edition of the GSGB, which notes that 2.7% of participants score eight or higher on the problem gambling severity index (PGSI), indicating that they have experienced some adverse effects from gambling.
In direct response to this variance, the UKGC told NEXT.io: “The GSGB and previous NHS-led surveys use different methodologies and should not be treated as directly comparable. We would encourage policymakers to consider the GSGB alongside other evidence, including operator data, treatment data and wider research, to build the fullest possible picture of gambling behaviour and gambling-related harm.”
It remains to be seen what problem gambling figure tomorrow’s publication will report, but the regulator has rejected the idea that the methodology used is inherently wrong.
The spokesperson added: “We publish comprehensive technical reports alongside the statistics and continue to review and improve the methodology as new evidence emerges. It is therefore wrong to suggest that concerns have been ignored or that the data are known to be unreliable. The methodology has been independently scrutinised and provides an important source of evidence for understanding gambling behaviour in Great Britain.”
