(3,230 words | Estimated reading time: 17 minutes)
In recent years, the Gambling Commission has imposed fines totalling over £52m against bookmakers for social responsibility failures. These social responsibility failures can have a particularly adverse impact on problem gamblers.
Where improper encouragement and/or a failure to protect a problem gambler has caused loss, might their bookmaker be liable for negligence? A negligence claim was brought against a bookmaker by a problem gambler in Calvert v William Hill Credit [2008] EWHC 454 (Ch).
By way of background, the claimant in Calvert betted large amounts of money with the defendant via the telephone over the course of several months, before asking the defendant to give effect to a self-exclusion agreement, preventing him from making further bets. The defendant agreed to do so, but this agreement was not properly implemented, so that when the claimant sought to reopen his account months later, he was permitted to do so. After engaging in a further betting spree the claimant again requested that he be excluded from placing further bets. The claimant was assured that this request would be effected, but it was not. The claimant went on to lose close to £2m during the course of gambling with the defendant.
Although a narrow duty of care was identified as being owed to the claimant on the facts, Briggs J, as he then was, rejected the existence of a broader duty owed to problem gamblers more generally. Simmler J subsequently stated in The Ritz Hotel Casino v El Geabury [2015] EWHC 2294 (QB) that the issue ‘was addressed comprehensively by Briggs J’ in so doing.
Briggs J’s reasons for rejecting the existence of a broad duty of care may no longer be open to the court, if they ever were in 2008, in light of socio-political and regulatory developments. Briggs J’s reasons are provided at Paragraphs 169-174 of his judgment and shall be addressed in turn.
Assessments of vulnerability
At Paragraphs 169-170 of Calvert, Briggs J notes that problem gamblers do not suffer uniformly as a class and that not all will be so vulnerable as to require special treatment. He deems requiring a bookmaker to identify those who are so vulnerable as tantamount to requiring them to conduct a medical diagnosis, and regards the notion of licensees being able to do so as ‘wholly unrealistic’. This is echoed in The Ritz Hotel Casino v Al Daher [2014] EWHC 2847 (QB) at Paragraph 114.
Provision 3.4 of the Gambling Commission’s Licence Conditions and Codes of Practice October 2019 (‘LCCP’), to which all licensed bookmakers in the United Kingdom are subject, provides that:
1) ‘licensees must interact with customers in a way which minimises the risk of customers experiencing harms associated with gambling. This must include:
a) Identifying customers who may be at risk of or experiencing harms associated with gambling.
b) Interacting with customers who may be at risk of or experiencing harms associated with gambling.
c)Understanding the impact of the interaction on the customer, and the effectiveness of the Licensee’s actions and approach’.
2) licensees must take into account the Commission’s guidance on customer interaction.’ (emphasis added).
Paragraph 2.1 of the Commission’s February 2018 guidance note states that ‘if you suspect a customer may be experiencing or at risk of developing problems with their gambling, you are required to interact with them to offer help and provide support’. Per paragraphs 3.12-3.14 of the Commissions July 2019 guidance note, it is further stated that licensees will ‘need to direct some customers to information about safer gambling…will need to interact with some customers a number of times’, and ‘for some customers, and particularly if the behaviour continues to cause concern, you may need to take a more proactive approach’ (emphasis added).
It is clear from the above that the Commission has not only deemed it realistic for bookmakers to engage with customers and identify those that are vulnerable and in need of protection – it mandatorily requires such assessments to be made.
Note that it was the seminal 2007 version of the LCCP being considered by Briggs J in Calvert (this is noted at Paragraph 7 of his judgment). Provision 2.3 of the 2007 LCCP addresses customer interaction and is the predecessor to Provision 3.4 of the 2019 version above. It states that ‘licensees must implement policies and procedures for customer interaction where they have concerns that a customer’s behaviour may indicate problem gambling’.
The difference between this version and the 2019 version is of tantamount importance. The current requirements to identify, interact with, direct, and support, all in a proactive fashion, together establish a duty to take specific actions which protect problem gamblers as a class. The earlier versions more reservedly require only that policies be put in place which might lead to this outcome, without specifying in such terms what said policies might entail. This left it open to Briggs J to make his determination that it would be unrealistic to expect bookmakers to make assessments of vulnerability.
Deference to the views and authority of the Commission is notable in Briggs J’s judgment, most notably at Paragraph 171 when he allows himself to be guided by what is and is not permitted by the Commission. It is therefore reasonable to suppose that he would, with view of the LCCP as currently drafted, have concluded differently.
In respect of prospective claims in relation to losses suffered prior to October 2019 or February 2018, it is open to courts to determine that the updated LCCP and guidance issued by the Commission on these dates respectively confirmed the existence of a duty that already existed at common law. The Commission’s recent developments can be regarded as an expansion on what was previously said, rather than a rewriting of the rulebook. It confirms that compliance with the policies which bookmakers have needed to put into effect since 2007 did and does necessitate a degree of proactivity and concern that is incompatible with the High Court’s decision in Calvert. This not being previously made explicit, Briggs J’s decision can be overruled without being subject to criticism.
How might vulnerable online gamblers be identified?
In light of the above, many bookmakers may question what steps they should be taking to satisfy their duty to identify problem gamblers.
Many problem gamblers gamble online, and many bookmakers collect data on the time played and amounts betted by online customers. Outside of instances where individuals hold themselves out as possible problem gamblers, abnormally high amounts of time played and amounts betted are, it would seem, the only other metrics by which a bookmaker might identify that an individual is potentially a problem gambler.
This being so, and given the Commission’s requirement that bookmakers be proactive, such that never doing more than waiting for customers to hold themselves as a problem gambler is unlikely to satisfy Provision 3.4 of the LCCP, it would appear that the Commission envisages bookmakers making use of this data in the identificatory process that they are required to undertake.
It seems that this data is already considered by some bookmakers in their identification of customers eligible for what are commonly referred to in the industry as VIP package deals, and bookmakers may be well-advised to exercise caution in their approach to these offerings.
The link between VIP accounts and problem gambling is considered in the All-Parliamentary Party Group’s Interim Report on gambling related harm of November 2019, which notes that:
‘operators told us that a much higher proportion of VIP accounts are closed (one operator said around 30% were closed last year) which strongly indicates that the issue of problem and at-risk gamblers being permitted these accounts is widespread’.
The Report also links the treatment of VIP accounts with the abovementioned fines imposed by the Gambling Commission in respect of social responsibility failures:
‘VIP accounts should be restricted: We recommend that far greater assessments are taken by operators to assess a person’s suitability to have a VIP account and that the Gambling Commission adopts more vigilance towards these accounts. We are concerned that fines to companies for offering inducements have very little impact on this well-resourced and well-funded industry’. (emphasis in original text)
Concerns with VIP packages are more explicitly referenced in a “secret report” that the Guardian claims (in their article dated 2 January 2020) to have obtained – a report which they say shows that the Gambling Commission is considering the banning of VIP schemes altogether.
Especially if bookmakers already have processes in place to identify suitable candidates for VIP membership, it may not be considered unrealistic or overly onerous to expect that they incorporate further steps to identify, in light of this information, whether these individuals might be problem gamblers in need of assistance in satisfaction of a broad duty of care.
Exclusion of liability
At Paragraph 171 of Calvert, Briggs J observes that ‘the Gambling Commission’s own policy has, thus far, been to permit liability exclusions as compliant with the licencing regime’, which he regards as leaving it open to bookmakers to choose not to enter into a self-exclusion agreement with a customer requesting to be excluded. He regards this as compliant with a more general principle:
‘in principle, where A asks B to do something upon which A may rely…B is entitled to choose between three alternatives. He may decline altogether…he may do that which is requested with a disclaimer of responsibility for the consequences, or he may simply do it without qualification…in the present case… [the defendant] adopted the second of those three alternatives’.
If they ever were, the first and second of the three alternatives offered now seem unavailable to bookmakers in light of the express, mandatory language of the LCCP and accompanying guidance, though the notion that it was ever within the gift of a bookmaker to refuse to help a problem gambler who has requested self-exclusion, under the 2007 version of the LCCP, may be challengeable. Note also that this point addresses whether or not there could have been a breach of a broader duty on the facts of Calvert by virtue of the defendant’s policy in relation to self-exclusion agreements. It is not an argument against recognising the existence of a broader duty more generally, despite being framed as such in Briggs J’s judgment.
Customer autonomy
At Paragraph 172 of Calvert, Briggs J acknowledges that problem gamblers who are not provided support might ‘suffer an aggravation of [their] condition’, but states that:
‘the law should be very slow to…justify a requirement to take protective steps to restrain a gambler from exercising his liberty to gamble on his own responsibility, where his status as a problem gambler may mean no more than that he is experiencing mild and occasional difficulties of control’
And concludes by noting that:
‘such a duty would…be an invasion of his autonomy, in relation to an activity for which he is primarily responsible for the consequences’.
This sentiment is echoed in Paragraph 116 of Daher. Llewelyn J agrees with the above findings, and further comments that although one might regard a gambler’s choice to bet with large sums of money as irrational, ‘the choice of Parliament has been to permit casinos to be licensed, and gamblers to gamble in them, as a matter of their own autonomy’. Similarly, at Paragraph 46 of Aryeh Ehrentreu v IG Index Limited [2018] EWCA (Civ) 79, the root of Briggs J’s reasoning behind not identifying a broad duty in Calvert is deemed to be the notion that a problem gambler’s use of gambling services amounts to ‘deliberately inflicting economic harm on himself’.
Briggs J’s cautious approach in 2008 is understandable. However, it cannot be reconciled with the updated LCCP and accompanying guidance. The above conclusions that the problem gambler is ‘exercising his liberty’, ‘deliberately inflicting economic harm on himself’, and conducting a ‘matter of their own autonomy’ might also be incompatible with the reclassification of problem gambling as an addiction disorder in the DSM-IV, as of 2013. Prior to 2013, and as is acknowledged by Briggs J at Paragraph 117 in Calvert, problem gambling was understood to be an impulse control disorder, which inhibits the ability to resist acting upon one’s desires. The reclassification reflects a developing understanding that the neurological effects of problem gambling are more complex and bear more strongly on the individual than was thought to be the case in 2008, affecting the individual’s ability to exercise meaningful control much in the same way that addiction to drugs can.
Increasing awareness of the needs of problem gamblers, and their relative inability to exercise autonomy, has resulted in a recent surge of support from related third party institutions. In November 2019, HSBC introduced mechanisms which allow customers to voluntarily block usage of their cards on gambling websites, following in the footsteps of Lloyds Bank, Bank of Scotland and MBNA. The Guardian reports that in the first couple of weeks following the launch of these controls, over 15,000 customers had sought to activate them.
Further to this, the Gambling Commission announced on 14 January 2020 that from 14 April 2020, a wholesale ban on the use of credit cards will be imposed in respect of all gambling products except for lotteries in furtherance of a good cause.
The NHS is also taking action. Building upon support currently offered by the NHS’ National Problem Gambling Clinic in London, there are plans for up to 14 more dedicated clinics for problem gamblers to be established across the nation. On 8 January 2020, the first of these new additions opened its doors in Sunderland.
There is therefore regulatory force and, it would appear, rapidly increasing social consensus behind the conclusion that it is inappropriate to regard problem gamblers as fully autonomous, and correct to identify them as vulnerable, such that it would be reasonable to impose a duty on those who provide regulated services which might exacerbate their condition to, at the very least, not improperly encourage or exploit said vulnerabilities.
Briggs J’s concern, cited above, that one’s liberty might be restrained when they may be exhibiting no more than ‘mild and occasional difficulties of control’ is a concern about balance. In attempting to assist those in need, the newly cautious bookmaker might cast the net too wide. But processes can be refined over time. And in the meantime, though potential offense might be caused to more occasional gamblers by the undue suggestion that they might be exhibiting signs of problem gambling, this must be weighed against the need to protect actual problem gamblers who are vulnerable and in need of support.
Fairness
At Paragraph 173 of Calvert, Briggs J states that a broad duty would not be fair because ‘it would place a burden on the bookmaker pursuant to which the gambler could freely take home his profits, but look to the bookmaker for the return of his losses, without even seeking the bookmaker’s assistance to help him control his gambling’. Agreement with these comments is provided at Paragraph 141 of Al Geabury.
A duty of care towards individuals identified to be problem gamblers specifically would not allow gamblers more generally to ‘look to the bookmaker for the return of his losses’. If a problem gambler is successfully able to claim back losses, there will have first been a determination that there has been a breach of duty by their service provider. There is nothing unfair about requiring said service provider to bear losses that they have brought on themselves as a result of their breach, regardless of whether the problem gambler wins or loses.
To the extent that this point is not accepted, questions as to the fairness of the recoverability of monies lost are context-specific and might turn on the facts of a given case, but are at any rate separate from and should not impinge upon the court’s determination of whether a broad duty should, generally, be deemed to exist.
And Briggs J’s position on what might fairly be recovered is inconsistent. If it would be unfair for a problem gambler to retain profits and recoup losses arising from breaches of a broader duty, why is it any less unfair that one be entitled to do so in circumstances where they are owed a narrower duty of care, as was determined to be possible by Briggs J on the facts of Calvert? This inconsistency cannot be sensibly reconciled.
“The incremental cross-check” – customer autonomy Part II
At Paragraph 174 of Calvert, Briggs J cross-checks his conclusions regarding the recognition of a broader duty of care against other contexts in which a duty has or might arise. Unlike other companies or governmental institutions which might owe a duty to their dependents, Briggs J concludes that gambling companies assert ‘nothing like the control over [their] customer’s activity [like that] which was exerted over [the claimant in another case], or for that matter a control equivalent to that of a prison authority over a prisoner’, and observes that decisions recognising the existence of a duty of care go out of their way to emphasise the importance of the total incapacity of the… claimant’.
Applied to this context, these comments suppose that problem gamblers have relatively complete control over their actions, and that bookmakers have relatively limited control. It has been noted above that these conclusions may no longer be acceptable or appropriate.
Briggs J also notes at Paragraph 174 that in cases where individuals have lacked capacity by virtue of intoxication, it has been necessary to show that this was ‘known to the defendant…or…probable by reason of the circumstances’ in order for a duty of care to be capable of arising between the defendant and intoxicated person. In respect of online gamblers, bookmakers generally have access to data showing the amount of time played and amount betted by individual gamblers. As noted above, bookmakers have a regulatory duty to proactively identify individuals who might be at risk of problem gambling and it is not clear how they might do so without recourse to an analysis of these metrics. Given the general availability of relevant data and existence of the regulatory duty, it may not be too onerous to impose an assumption that individuals who play for longer than a given period or bet more than a certain amount are, by reason of the circumstances, probably problem gamblers, until sufficient steps are taken to conclude otherwise.
This article does purport to comprehensively address what might constitute high amounts of time played or amount betted. It may be helpful for bookmakers to engage in a public dialogue with each other in this respect.
Conclusion
This article has provided a paragraph-by-paragraph analysis of Briggs J’s findings in Calvert regarding the existence of a broad duty of care owed by bookmakers to problem gamblers, and has shown each of the points made to be susceptible to challenge, and/or not relevant to the question of whether a broad duty of care should be deemed to exist. The issue has not, therefore, been addressed comprehensively as was suggested by Llewelyn J in Al Geabury.
Given the amount of individuals affected by the social responsibility failures identified during the Gambling Commission’s imposition of over £52m in fines, and given the momentum driving changes in our approach to problem gamblers more generally, it is only a matter of time before the court has cause to revisit Briggs J’s decision.