Proposed Customer Interaction Guidance: The problems with common sense by decree
The Gambling Commission (the “Commission”) is currently consulting (the “Consultation”) on the proposed Customer Interaction – Guidance for remote operators (the “Guidance”). While this exercise has not yet attracted the same attention as its 2020 predecessor (on remote customer interaction and affordability checks) it is potentially every bit as significant for licensees and consumers. Once implemented, it is likely that its strictures will in time bind non-remote gambling licensees and consumers too. Regulus Partners and Harris Hagan have written a series of co-authored articles to assist stakeholders responding to the Consultation. In the second of these articles, we described how the Guidance threatens to alter the meaning in law of “vulnerability” through its definition of “vulnerable persons”, by rendering universal what had formerly been considered exceptional. In this, our third article, we examine why this matters, in terms of both near-term regulatory compliance and longer term attempts to stigmatise participation in gambling.
The most striking feature of the Guidance is the extent to which it attempts to mandate common sense through a series of ill-defined rules – proposing a vast bureaucracy to facilitate and evaluate customer interactions. Having established a wide range of criteria for classifying customers as either “vulnerable” to, or “at risk” of, gambling harm; two concepts that appear to be definitionally similar but which the Commission, at different times, treats as both distinct and the same, the Guidance then requires licensees to configure a range of actions in response. The Guidance is, however, silent on what specific actions should be aligned to particular criteria (or clusters of criteria).
A little less conversation, a lot more interaction
Such is the breadth of the Commission’s conceptualisation of “vulnerability to harm” (there are literally millions of possible combinations of the “factors of vulnerability” and “indicators of harm” contained within the Guidance), it is inevitable that different licensees will take different actions in response to the same risk factors. More significantly and putting aside the inconsistent use of the differing terms “may be experiencing harm” and “may be at risk of harm” in the Guidance, licensees’ interpretation of what actions should be applied to any particular constellation of “indicators” is likely to differ from the Commission’s (and indeed, within the Commission, between one official and another).
In the absence of clearer direction, licensees will be left to learn from painful experience what “vulnerability” looks like through the eyes of Commission officials as well as what action or response should be considered appropriate (knowing that the Commission’s view on any given day is no guide to future interpretation). More cautious licensees may find themselves conducting customer interactions as a response to the vaguest suggestions of vulnerability; for example, being above or below certain, as yet unspecified, age thresholds. In this way, a degree of standardisation may eventually be achieved; albeit with thresholds set at successively lower levels.
Licensees who adhere to the Guidance, are exposed through a requirement that their staff “need to be trained on the skills and techniques they need to help them carry out customer interactions, including what to do if a customer becomes distressed or there is a risk of suicide.” There is no question of anyone in the gambling industry not wishing to prevent suicide. However, this wording implies that it is the responsibility of licensees or their employees to identify risk of suicide, and act upon it. Suicide risk is always the result of a complex array of factors, that may or may not, include an individual’s gambling. It is the responsibility of qualified professionals to identify that risk, not licensees, and it is dangerous on multiple levels, including in relation to the wellbeing of licensees’ employees, to suggest otherwise in formal Guidance.
The consequences of all this may be guessed at: enforcement action ramped up as the Commission weaves an ever more intricate web of compliance tripwires; licensees absorbing (or passing onto customers) substantial increases in the cost of doing business; and customers facing greater levels of intrusion and inconvenience based upon lifestyle preferences but also a range of factors beyond their control, such as age or disability.
Under the Guidance, licensees will be expected to conduct customer interactions according to a quota system based on DSM-IV and PGSI “problem gambling” prevalence rates from the Health Survey for England 2018 (the “Health Survey”). Online sportsbooks must therefore interact with at least 3.7% of their customers each year; while the minimum quota for online slots, casino and bingo operators is set at 8.5%. Licensees will be required to carry out monthly checks to ensure that they are on course to hit these targets.
The use of quotas raises a number of questions not addressed within the Consultation, which does not even canvass views on the wisdom of such a scheme. First, it reveals a discontinuity in the Commission’s logic – with “problem gambling” belatedly introduced as a proxy for “vulnerability to harm”. It should not, however, be assumed that someone with a DSM-IV or PGSI classification of “problem gambling” will meet the definitional criteria set out by the Commission for “vulnerability” or “harm”. This invites the question of just what problem is to be addressed and who, in particular, licensees are expected to protect: those whose gambling might be considered problematic according to recognised psychiatric criteria, or those who may meet the regulator’s often more nebulous definition of “harm”. The requirement that licensees use “specifically the problem gambling rates for the individual activities” exposes a basic misunderstanding of prevalence surveys. The Health Survey does not in fact provide “problem gambling” rates by discrete activity. Instead, it shows prevalence rates for people who participate in certain activities in combination with others, which is a rather different thing.
The Guidance becomes even more muddled in proposing interaction quotas for licensees with online sportsbooks and casinos – suggesting the use of a combination of problem gambling rates weighted for the percentage of revenue derived from each activity. It overlooks a simpler, more targeted and more logical method – using ‘problem gambling’ prevalence rates for customers who only use sportsbooks when gambling online, those who only play online slots, casino or bingo and those who do both. We have provided the figures below from the Combined Health Surveys for 2016: it would be a relatively straight-forward matter for the Commission to update these for 2018.
|Activity||DSM-IV/PGSI ‘problem gambling’ rate|
(% of customers participating)
|Online slots, casino & bingo||7.3%|
|Online betting and online slots, casino & bingo||8.7%|
The Guidance states that these quotas may be recalibrated in the future – with particular reference to how “problem gambling” or “gambling-related harm” are measured. Here it should be recalled that the Commission is part-way through a process to wrest control of the measurement of gambling and “problem gambling” prevalence from the NHS – through the replacement of the Health Surveys with its own vehicle. The Commission’s Update: Pilot of survey questions to understand gambling-related harm published in May 2022 (the “Pilot Survey”) produced a combined PGSI problem gambling’ rate of 1.3% – more than three times higher than the result from the Health Survey. Analysis by Regulus Partners has highlighted previously a number of serious errors with the Pilot Survey (including a failure to carry out cross-checks with regulatory returns data; and overlooking the impact of Covid-19). The Commission has neglected to address these flaws and states, as an article of faith and without supporting evidence, that Health Surveys under-report “problem gambling”.
The Pilot Survey also contained questions about gambling-related harms, although the Commission has been rather selective in releasing these results. Some of these “harms”, as we explained in our previous article, include reduced attendance at the cinema, spending less time with loved ones or “feeling like a failure”. Rather unsurprisingly, they are experienced by a much larger group of gambling consumers than those likely to be classified as “problem gamblers”. Licensees may find therefore that interaction quotas are ramped up significantly in the future as a consequence of surveys controlled entirely by a regulator apparently intent on demonstrating that “problem gambling” or “gambling harms” are widespread in the population.
One obvious difficulty with issuing quota requirements is that they lead inevitably to a tick-box approach to compliance based on quota fulfilment. The Guidance sets out a very catholic definition of what might be considered an “interaction” – from generic safer gambling messages or pop-ups right through to treatment referrals and exclusion. Thus, in order to satisfy the quota, licensees may simply have to make sure that around one-in-five or one-in-ten customers receive a generic safer gambling message at least once a year – a figure that is likely to be well below current levels where responsible operators are concerned. In time therefore, the Commission may either drop the quota system or increase its complexity, with quotas for specific types of interaction (e.g. quotas for self-exclusion).
When combined, all of the above issues will make it incredibly difficult for licensees to act where there are “strong indicators of harm” as there is a lack of clarity in the Guidance as to what the Commission considers to be a “strong indicator of harm”. In the Consultation the Commission acknowledges previous concerns raised about this, and states that it “does not consider it appropriate at this time to set requirements which would remove the discretion or ability on the part of operators to tailor processes to their businesses and customers”. One might take the view that this is precisely what the Guidance does, particularly when it contains formal requirements.
Evaluation and Impact
The greatest area of complexity is likely to reside within the requirement that licensees conduct assessments in order to “understand the impact of individual interactions and actions on a customer’s behaviour”. The Commission is correct to highlight the importance of evaluation, even if it might strike some as hypocritical, given its own aversion to scrutiny, as safer gambling initiatives are often implemented or mandated on the basis of face validity rather than scientific observation. Evaluation is critical therefore if we are to improve and move beyond what should work in theory and understand what works in practice.
However, the implication here is that if the customer’s gambling activity does not change for the better (i.e. stop or reduce), they are suffering, or continue to be at risk of suffering harm. This cannot be correct: a customer may continue to gamble at previous levels or even increase their spend following an interaction for various reasons. There are several problems with the demand that licensees demonstrate the effect of changes to customer behaviour of every single customer interaction. First, it will be vastly bureaucratic and costly to implement, given the number of interactions that licensees will be encouraged to undertake. This might be acceptable if the benefits of such a system outweighed the costs, but this is unlikely to be the case. Gambling behaviour, and particularly disordered behaviour, is complex. To suggest that each individual licensee’s action might be separately assessed for discrete impact goes against the balance of research opinion as well as the Public Health whole systems approach. Neither the Guidance nor the Consultation give any consideration to proportionality, in direct contravention of the Responsible Gambling Strategy Board’s evaluation protocol.
This section of the Guidance gives rise to possible negative (and so presumably unintended) consequences. It fails to consider the substantial costs that such a system would impose on licensees or what this might mean in terms of customer experience where such costs are passed on through pricing. Most importantly, however is that it may impede efforts to protect consumers. Licensees swamped by assessing the impact of potentially millions of individual interactions may suffer a loss of perspective, impairing their ability to identify and understand what is happening to those at genuine risk of harm. People who fall under the Guidance’s discriminatory gaze (by virtue of being too old, too young, not physically fit enough or too trusting) may find themselves subject to repeated harassment by multiple licensees. Finally, there is the risk that a requirement to carry out evaluation on such a microscopic basis will in fact deter licensees from undertaking interactions over and above the level demanded by the quota system.
The Guidance of course, is not satisfied by requiring evaluations of every single customer interaction: licensees must also demonstrate impact. This stipulation reveals a fundamental flaw in the Commission’s thinking. As we explored in our second article, the basis for interaction in the Guidance is a range of “indicators” denoting “vulnerability” or “harm”; but indicators are not the same as actual vulnerability or harm. In demanding that all interactions must demonstrate impact, the Commission appears to conflate probability with certainty; and in doing so ignores the presence of false positives that are a feature of any diagnostic system. Under the Guidance, licensees will be required to demonstrate that interactions alter customer behaviour, regardless of whether reform is necessary, or to document, frequently, their decisions as to why such alteration was considered unnecessary despite the indicators identified. Worse, licensees will feel obliged to take progressively more heavy-handed approaches with such customers until a change is observed, the regulatory equivalent of factitious disorder imposed on another.
This impact becomes even more absurd when one considers the nature of some of the indicators in the Guidance that may trigger an interaction. To illustrate, we use the same hypothetical customer from our second article – a 24-year-old with dyslexia who bets in-play on football and cricket and typically spends slightly above the average for his age group. Now imagine this benighted individual suffers a bereavement. Displaying seven indicators of vulnerability or harm, some licensees may consider him an appropriate target for interaction – but what precisely should be the outcome? The licensee cannot alter his age, address his dyslexia or alleviate the distress of personal loss; and so must presumably either deter him from betting in-play (an activity branded an “indicator of harm” on the most spurious grounds, as we show in our next article) or encourage him to spend less. Even if a reduction in spending is achieved (via coercion of an individual displaying no actual symptoms of harm) the customer will still be considered vulnerable on four counts and at risk of harm on two – and therefore subject to further interactions in the future (until he ceases to be ‘younger’ or ‘bereaved’ perhaps). The consequences of such a regime are unlikely to be in the best interests of that customer; and while ‘black market’ risk can be overplayed, it seems legitimate to cite it where such blatantly anti-consumer logic is concerned.
These are the near-term implications of what the Commission proposes in its Guidance. It is possible too that the Guidance hints at longer term aims or outcomes: the addition of further regulatory restrictions which stigmatise betting and gaming as pastimes. The codification in the Guidance (which introduces “formal” requirements) of “younger adults” as intrinsically vulnerable, may in time be used to lobby for the legal age for gambling to be raised (with perhaps a new maximum age being introduced to address risk among “older adults”); while certain modes of gambling (most obviously in-play betting) may be curtailed or banned on the same basis.
The Guidance contains some useful insights but ultimately falls down in its attempt to inculcate common sense by diktat. In doing so, it fails to recognise the essence of common sense as something that cannot be circumscribed by rules. Whilst some of the factors set out in the Guidance are sensible and may have an effect on a customer’s powers of self-regulation, it is impractical to seek to enforce this in “formal” Guidance, particularly when that guidance is unclear. Further, there is no suggestion anywhere in either the Guidance or the Consultation that the Commission has taken the time to understand the basis of current licensee practices (a review of enforcement cases is, by definition, no way to assess the market as a whole); or to consider how desired behaviours might be encouraged rather than coerced. In our last two articles in this series, we will examine the evidence presented by the Commission for classifying in-play betting as an ‘indicator of harm’ in its own right; before closing with a summation of the reasons why all gambling licensees and consumers, alongside those concerned about personal freedom in other domains, should take the time to respond to this most worrying of consultations.
With thanks to Dan Waugh from Regulus Partners for his invaluable co-authorship.