27 Nov The Alternative Facts of Life and Death
Saying that one cares about something and actually caring about something are two very different things. Right now, a number of high profile politicians, journalists and public health lobbyists profess to care very deeply about the issue of gambling-related suicide. Some of these people are in earnest; others perhaps less so. In this piece we argue that those exercising curiosity and circumspection are more likely to represent the voice of concern than those who simply declaim.
It is currently in vogue in Britain’s Parliament and press to bandy around the “fact”that around 700 people a year take their own lives as a result of their gambling problems. It is a shocking statistic. In fact, it is so shocking that it warrants examination; but curiously this is something that few have bothered to do.
So far as we can tell, the ‘700 suicides’ figure started out with the concern group, Gambling with Lives. Its website states that the number of gambling-related suicides in the United Kingdom may be within the range of 250 to 650 a year. In other words, the upper end of the range is below the number now being regularly quoted as a “fact”.
This follows a familiar pattern in policy debates. First a range of costs is established; the bottom end of the range is then lopped off to form a central “fact”; and then hocus pocus, that “central fact” becomes the low-ball number. This happened in 2016 when the Institute of Public Policy Research estimated that the cost to the British state from problem gambling was somewhere between £260m and £1.2bn. It wasn’t long before commentators (some of them self-proclaimed “experts”) had successfully detached the £260m figure so that £1.2bn became the central “fact”. Soon after, we were told that this estimate was conservative. Thus, through a succession of small steps, the upper estimate had become the lower.
Of course, whether 250 people or 650 people take their lives in relation to gambling, these are still shocking numbers. So where do they come from? Gambling with Lives has been fastidious in citing its sources, which is why it is puzzling so few people have bothered to investigate. After all, curiosity – not outrage – is the mark of genuine concern.
The 250 figure is derived from a study (carried out by a team from the University of Manchester and published last year) of suicides amongst young people. It found that within a 20-24 year-old age cohort, 20% of suicide victims had “recent financial problems”and that 4% involved gambling problems. Thus the 250 suicides estimate assumes that 4% of all recorded suicides in Britain (5,965 in 2016 according to the Office of National Statistics) involve problem gambling.
The calculation is beautifully transparent – but is the conclusion right? First of all, there is the matter of sample size. The University of Manchester team studied 106 cases in the 20-24 year-old cohort; and found evidence of problem gambling in four of them (a mercifully – but unhelpfully – small sample size). The study makes no mention of gambling involvement in the under-20s age cohort so it is possible that the rate for the entire sample (391 cases) was closer to 1%.
Aside from issues of sample size, there is also the question of representativeness. We know that people in their twenties and early thirties are at elevated risk of problem gambling. Assuming that the rate of problem gambling in young adult suicides is the same for all suicide victims is an example of what statisticians call base rate neglect. Tellingly, the University of Manchester team did not suggest that their research ought to be used to gauge population-wide trends.
So much for the lower end of the range. What about the upper estimate of 650? This derives from a different study carried out by a team from the University of Hong Kong and published in 2010. The research project examined 150 cases of suicide in Hong Kong and found that 17 of the victims (or 11.3%) were “probable pathological gamblers”. Applying this rate to suicides in Great Britain provides a figure of 674 deaths.
It doesn’t take a genius to spot the flaw in this calculation. It is based on data from Hong Kong – a culturally and politically distinct jurisdiction situated almost 6,000 miles from Great Britain that is home to one of the largest bookmakers on the planet (the Hong Kong Jockey Club) and just a 55-minute Turbojet ride across the straits from Macau – a licensed gambling market four or five times the size of the Las Vegas Strip.
In 2005, the problem gambling rate in Hong Kong was recorded at 5.5% compared with a rate in Great Britain at roughly the same time of 0.8% (the mean of the rates from the British Gambling Prevalence Surveys in 2007 and 2010). In other words, problem gambling in Hong Kong at the time of the study may have been as much as seven times higher than in Britain; another obvious example of base rate neglect.
Hong Kong is quite simply not a very good benchmark for understanding gambling-related harm in Britain. For example, the University of Hong Kong team found that of the 17 gambling-related suicides in their study, five had received menaces of harm from loan sharks in the days leading up to their deaths.
This finding hints at another important point. The decision to take one’s own life is unlikely to be due exclusively to one single cause. Indeed, to dumb down the matter (as certain commentators have done) may in itself result in harm. The respected political journalist, Isabel Hardman (who knows a thing or two about mental health) makes this point in her recent book, ‘Why We Get the Wrong Politicians’. She writes (in relation to claims that benefits cuts have led to suicides): “It is irresponsible to suggest that suicide has one clear cause, not least because it suggests to others affected…that ending one’s own life is the only solution.”
At present, the official statistic for Great Britain is that between one and two suicides a year are linked to problem gambling. This at least is the figure cited by the Office for National Statistics which reports that between 2001 and 2016, 21 cases were linked to problem gambling. Of course, this is not a reliable figure as the ONS disclosure makes clear: “these figures represent how many records mentioned gambling on the death certificate but are likely to be an undercount.”
Our own search of internet sources (mainly press reports) found 20 cases of suicide apparently related to problem gambling in the seven years between 2012 and 2018 (with seven of them occurring in 2017). This is also likely to be an approximation of the tip rather than the iceberg entire.
So what is the true scale of gambling-related suicides in Great Britain? The simple answer is that we don’t know – a fact that Gambling With Lives itself acknowledges, stating on its website: “methods of recording, mean that it is not possibleto identify the number of deaths in the UK related to gambling”.
This absence of evidence prompted Dr Heather Wardle of the London School of Tropical Medicine and Hygiene (and our most experienced social researcher on gambling matters) to recommend this year that efforts should be made to ensure coroners consider (and record where appropriate) the involvement of gambling in suicide verdicts.
In time we will obtain more reliable official data. Until then, some may feel justified in inventing figures in order to draw attention to the subject. It is easy to sympathise with this point of view – and to understand the frustrations of the bereaved who are struggling to get gambling-related suicide taken seriously as a societal issue.
However, in another sense, there is something a little disturbing about these ‘alternative facts’. The danger in citing large (and unverified) numbers is that we reduce human tragedy to statistical abstract (and end up debating numbers rather than lives). The very fact that people (often but not exclusively young men) have taken their own lives in response to gambling problems is enough to suggest that not enough is being done – by the industry, by the state and by society – regardless of whether their number is two or 700 a year.
In the interests of decency, all sides must learn to eschew spin and be prepared to explore this most poignant of subjects in a spirit of unblinking and honest inquiry.