Commentary on Zendle and Newall: The need for direct evaluation of the UK's gambling affordability checks policy

Zendle & Newall's [1] article provides an important contribution to the gambling literature which should be celebrated for at least two reasons. First, the collection of open banking data represents a promising new method for the field. This approach circumvents issues with the inaccuracy of self-reported gambling expenditure [2] and provides a comprehensive overview of spending. Secondly, the authors should be commended for their openness and larger contribution to the field through data sharing. *

Their article provides strong evidence for understanding the relationship between gambling expenditure and harm. Unlike almost all existing gambling research, expenditure data were not limited to a single type of gambling, venue/operator or bank account. This offers a more complete picture of how spending relates to risk.

However, the first conclusion in their Abstract reads: ‘The UK government's proposed affordability checks for gamblers should rarely affect people who are not experiencing gambling-related harm’. I contend that this conclusion is not well-supported by their analysis, and more direct evaluation is required.

To summarize the approach used to reach their conclusion, the authors computed a total net-spend value for their participants based on their outgoing gambling expenditure minus withdrawals into their accounts over 12 months. They then divided this value by 12 to produce a monthly average. This showed that no-risk gamblers had a value of £16.41—far below the proposed £125 threshold which would trigger an initial affordability check. This value was £79.6 for low–moderate risk gamblers [Problem Gambling Severity Score (PGSI) of 1–4] and £208.91 for high-risk gamblers (PGSI > 4).

One update the authors could not have foreseen is that on 1 May 2024, the UK Gambling Commission published a revised plan for introducing the now-named ‘financial vulnerability checks’. The proposed updates actually lend credence to the authors’ findings, as the measure of affordability is set to change from ‘losses’—a difficult-to-calculate outcome that can include re-gambled winnings—to ‘deposits minus withdrawals’ (net-deposits); that is, the same metric used by Zendle & Newall. Customers will receive a check only if this ‘net-deposit’ value reaches £150 or more in 30 days, commencing 28 February 2025 [3]. †

A concern, however, is that the net-spend values computed by the authors included all gambling expenditure—on-line and land-based. The proposed checks still appear to relate only to on-line gambling and may, at least initially, ‡ need to be performed at an individual site level. Their outcomes, therefore, do not fully elucidate the relationship between spending with a single operator and harm or spending on-line more broadly and harm.

The authors’ conclusion that affordability checks will not impact no-/low-risk gamblers was based on the average monthly net-spend value for each risk group. This is problematic for two reasons. First, these values tell us little about whether someone crossed the original threshold value of £125 on any given month in the year. § Gambling behaviour may vary substantially over time, with high expenditure occurring during restricted periods followed by periods of decreased spend [4]. Second, these values do not appear to account for those who started gambling within the 12 months. If, for example, one individual had only been gambling for 3 months but with a consistent monthly net-spend of £300, they would have an average net-spend of only £75 using their methods.

A reanalysis of Zendle & Newall's data could more directly simulate the impact of affordability checks, should the policy have been implemented during the window covered by their banking data. To do this, the authors would need to calculate the proportion of eligible months (i.e. all those including and after the first month a person started gambling) in which each individual would have surpassed the threshold. These rates could then be compared between risk groups to determine how the proposed threshold relates to harm. The ability to differentiate at-risk and not-at-risk groups could also be compared across different thresholds, including the old (£125) and new (£150) values.

Zendle & Newall's conclusions will probably be of significant interest to United Kingdom and international policymakers who may adduce them to support gambling affordability checks. While I support the introduction of these checks, the authors have not directly evaluated their potential impact. They have, however, collected the requisite data to undertake a more direct evaluation that can inform the optimal implementation of affordability checks in the United Kingdom.

Robert M. Heirene: Conceptualization; writing—original draft; writing—review and editing.

No funding was received for this article.

R.H. has worked on a project funded by Responsible Wagering Australia (a representative body of Australian on-line wagering operators; University of Sydney, 2019–21) and as an independent, subcontracted statistical consultant for PRET Solutions Inc. on a commissioned project (funded by the Australian Casino operator Crown; 2023). In 2023, R.H. was a co-investigator on a successful grant from the International Centre for Responsible Gaming, which now partially supports his Research Fellow role at the University of Sydney.