*Ian Walker, is an economics professor at Lancaster University and recently co-authored **new research on problem gambling**. Here, he shares why we need to reconsider what we know, or don’t, about problem gambling and its impacts on wellbeing.*

The received wisdom – if that’s the right expression for near ignorance – is that problem gambling is relatively uncommon and the damage to wellbeing is limited to the individual and perhaps their immediate family.

**Around 300,000 problem gamblers in UK**

What we know is that psychologists have devised a way of spotting problem gamblers by asking 10 questions and counting if they get three or more ‘wrong’.

In large and (more or less) random sample surveys it turns out that 0.7% of UK adults get classified as problem gamblers – so a tiny problem? Well 0.7% is a third of a million adults; about the same number as the city of Sheffield has.

So, not so small scale, but still not a worry if it’s limited in effect. But if the impact is bigger than we suspected for each individual problem gambler, it would be a *really* big problem for all of us.

**Aggregate cost of £30bn each year for all problem gambers
**What can psychologists tell us about how big a problem this is, per problem gambler. Well, it seems that nobody bothered to quantify this! The first to ask, and answer, this question is new research: How much of a problem is problem gambling.

The question is hard to answer, as are so many unanswered questions, but our estimate of the answer is that the cost aggregated across all problem gamblers is probably around £30 billion each year. Suddenly what we thought of as a small problem, is actually a huge one in economic terms.

**A different approach: asking problem gamblers how they feel**

So how did we go about estimating this? We adopted the methodology pioneered by Richard Layard and his colleagues in the Centre for Economic Performance at the London School of Economics. Despite economists preferring to infer how people feel from how they behave, Layard and colleagues have been busy promoting the idea that it’s worth asking people how they feel, on a numerical wellbeing scale.

The idea has gained a lot of traction in the policy arena. So much so that even hard-nosed economists, who are naturally inclined to be wellbeing sceptics, are beginning to take notice. Such wellbeing questions now find their way into many social surveys. Thanks to the Gambling Commission and David Forrest, a specialist in the economics of gambling at Liverpool University, the question was tacked onto the end of the 2010 British Gambling Prevalence Survey. This survey also asked the 10 magic problem gambling questions to about 10,000 randomly selected households, as well as asking them about their spending on gambling products.

**Problem gamblers in lowest 10% for self-reported wellbeing scores**

The methodology first asks what happens to self-reported wellbeing for players who get no questions ‘wrong’ for those 10 problem gambler identification questions.

The graph below shows what we found. People who give more and more incorrect answers (the so-called DSM score), correspondingly report their wellbeing to be lower and lower.

On average, those with no wrong answers said that their wellbeing was about a reasonably contented eight. About 30% score eight, and about 50% score more than eight.

But the average wellbeing for those diagnosed with a gambling problem, by the DSM score, was just over six. Less than 10% of the general population sampled were this unhappy. So this establishes how much lower is the wellbeing of problem gamblers relative to non-problem gamblers.

*The relationship between wellbeing and DSM score*

**Problem gambling has similar social costs to alcohol misuse**

Next, we ask what the relationship between wellbeing and income looks like. It turns out that problem gamblers and non-problem gamblers are different in other ways that affect wellbeing. The difference in wellbeing between problem and non-problem gambers is 1.5 on a log scale. What does this mean? Instead of explaining log scales, let me put it in money terms: a wellbeing difference of 1.5 is about equivalent to a difference of around £90k each year. A huge difference.

So this finding shows the difference in wellbeing terms between problem and non-problem gamblers is close to three times as large as the effect of doubling income.

If we take that £90k average and apply it to the 750,000 adults identified as problem gamblers, the aggregate loss in wellbeing is equivalent to around £30 billion. This is much bigger than we expected it to be. In effect, it’s the same order of magnitude as the social costs of alcohol abuse, and it doesn’t get much bigger than that.

When we discovered this, we had many reservations about this huge number. We worried that the rather crude way in which the survey measures problem gambling would imply that our estimate was too small. We worried that the crude way in which income is measured would mean that it was too large. We examined the robustness of this number to the reservations that we had as best we could with the data. But we could not convince ourselves that it was unrealistic estimate of the costs of being a problem gambler.

**The data doesn’t match what we know from the gambling industry**

So, the research tells us that problem gambling is a huge problem. But it doesn’t tell us what it is that makes problem gamblers so miserable. Nor does it tell us about how to deal with it? We would like to know what problem gamblers spend on gambling, by type of gambling and how much they lose. But, guess what? People lie about their gambling spending. The data, in most cases, doesn’t match what we know about gambling spending and losses from the industry. Only in the case of National Lottery products do we get information from the survey that matches what we know from the operator.

Our survey provides self-reported data on spending. We expect that problem gamblers underestimate their play (and losses) by (probably much) more than do recreational gamblers.

But the differences in our data are already really scary – even though they are probably underestimated to a large degree. Problem gamblers report average monthly gambling spending of £300 – fifteen times greater than non-problem gamblers.

Problem gamblers spend around 10 times as much on scratchcards as do non-problem gamblers, and around 30 times as much on casino (online and offline) games, and around 50 times more on machines found in bookmakers that electronically mimic fruit machines and casino tables: so-called FOBTs (fixed odds betting terminals) where lots of money can be lost very quickly. In contrast, problem gamblers spend only 10 times as much on scratchcards and less than twice as much on lottery draw games as do non-problem gambers.

In one or two Scandinavian countries in the world, there is automatic collection of gambling spending by individuals in real time and better research will need such data. For the moment, in the UK all we can say is that we are confident that lotto is not the crack cocaine of this industry; and scratchcards probably play only a small role in the misery.

**Researchers and regulators haven’t been aware**

Such is the size of the problem gambling problem, and the size of our ignorance about how this occurs, we are left wondering how we have got into this state of affairs. It seems that researchers and regulators have not been aware of the size of the problem.

This is not surprising since the resources spent on problem gambling treatment and research is just a tiny fraction of that spent on alcohol abuse and tobacco use. And right now, we don’t have great quantitative evidence on what works for problem gambling – and some treatments, like Cognitive behavior therapy, has limited geographical availability.

Profiling people who are developing a problem will probably help a little – but apart from the problem being worse in young males the defining characteristic is excessive levels of expenditure. Self-control strategies are available but can be worked around too easily.

Schools might play a role since young people are vulnerable to risky products. Most importantly, it seems like a bad idea to have the resources available for training, treatment and research to be driven by contributions from the industry – this is not something that we adopt for smoking and drinking, it makes no sense to adopt it for gambling.