Gambling

Written evidence submitted by Professor J Orford (GA 42)

Summary

· I am an Emeritus Professor of psychology with a long-standing interest in gambling and problem gambling. I was adviser for the three British Gambling Prevalence Surveys (BGPSs) and author of a recent book on the subject. My main concern is with the public health aspects of gambling (paras 1 & 2).

· The latest BGPS (2009/10) shows the prevalence of problem gambling amongst adults, particularly young adults, to be sufficiently high that it constitutes a significant public health problem which needs to be addressed. Even larger numbers are adversely affected by the problem gambling of a family member (para 3).

· Since the previous survey (2006/07), problem prevalence appears to have increased by 40 or 50%. That is most likely to have been due to the effects of the 2005 Act which increased access to gambling opportunities. Such an increase was predicted by major gambling policy reviews including that of the Budd report which preceded the Act, and is consistent with public health theory. It constituted a main hypothesis behind the series of BGPS studies (paras 4 -11).

· Certain groups appear from the results of the surveys to be particularly vulnerable. They include those who are unemployed, those on lower incomes, those living in relatively deprived areas and those from Black and ethnic minority groups (para 12).

· Under 16s were not included in the adult surveys but 12 to 15 year olds are known to be especially vulnerable if exposed to gambling. The access of children and adolescents to category D machines should be reviewed as was promised (para 13).

· Results of the 2009/10 survey, including secondary analysis (report available shortly), are throwing further light on the question of whether certain forms of gambling are more dangerous than others. Some forms of gambling, fixed odds betting machines (B2 machines) among them, are associated with much higher rates of problem gambling prevalence and much higher proportions of gambling occasions and of total spend attributable to problem gamblers, than other forms (paras 14 – 16).

· In conclusion, the 2005 Act has been responsible for exposing the British people to a greater danger of problem gambling than previously was the case. The changes which the Act brought about are not popular with the public. The Government should now adopt a more precautionary approach in the interests of public health and should be less influenced by supplier interests (para 17).

1. I am Emeritus Professor of Clinical and Community Psychology in the School of Psychology at the University of Birmingham and an Honorary Fellow of the British Psychological Society. I am a Clinical Psychologist by training but most of my career has been spent jointly in University and NHS settings (in London, Exeter and Birmingham). I have also had continuous career involvement in the charity/voluntary sector as a Trustee/committee member. My main area of interest throughout my career has been the study and treatment of problems relating to the use of substances and activities, like alcohol, gambling or certain drugs, which have the potential for addiction and misuse. Gambling has increasingly been the focus of my work in recent years. I have been one of the academic advisors to the three British Gambling Prevalence Surveys of 1999/2000, 2006/07 and 2009/10, carried out by the National Centre for Social Research (NatCen), the last two for the Gambling Commission. My most recent written work on the subject was the book, An Unsafe Bet? The Dangerous Rise of Gambling and the Debate We Should be Having, published by Wiley-Blackwell in January 2011.

2. My main interest is therefore in the public health aspect of gambling. My submission is mostly about the impact that the Gambling Act 2005 has had on levels of problem gambling, drawing particularly on the results of the 2009/10 survey. I also address other issues such as the regulation of gaming machines and the effectiveness of the Act in protecting children and vulnerable people from the adverse effects of gambling.

3. The British Gambling Prevalence Survey carried out in 2009/10 (BGPS10) used two measures of problem gambling, according to which the prevalence of problem gambling was estimated to be 0.9% or 0.7% of the population aged 16 years and over. Those percentages translate into substantial numbers of problem gamblers: in the region of 450,000 or 350,000 people. These are very significant numbers in public health terms which put problem gambling on a par with many other health conditions which are taken far more seriously. For example, these figures are comparable to estimates of the prevalence of the misuse of Class A drugs in Britain. Because large numbers of Britons do not gamble at all, prevalence figures are substantially higher if expressed as a percentage of those who report any gambling in the past year (1.3% or 1.0%). The proportion of adults with family members they believe to have gambling problems is higher (3.8%). To the figures for problem gamblers can be added an estimated approximately 900,000 adults who are gambling in a way which puts them at ‘moderate risk’ of problem gambling. These figures suggest that as a nation we do have a very significant gambling problem which should be at the forefront of our minds when considering national policy.

4. BGPS10 showed a substantial percentage increase in the prevalence of problem gambling since the previous survey was carried out in 2006/07. The same two measures were used in both surveys. The increase in prevalence was 50% in the case of one measure, and the increase was statistically significant; the increase was 40% in the other case, and of borderline statistical significance. There has been much discussion about how the apparent rise in the prevalence of problem gambling in the three years between 2006/07 and 2009/10 should be interpreted. I have heard the view expressed that this should not be attributed to the effects of the 2005 Act, either because a) it is not a real effect, or b) the rise can be explained in other ways. That view seems to me to be dangerously complacent and/or biased towards serving producer interests. In my view the results constitute strong evidence for a real increase in the prevalence of problem gambling as a consequence of the changes introduced under the Act. Let me explain.

5. Gambling carries some danger because it is potentially addictive. That has long been known and for a number of years now problem gambling has been included in the major taxonomies of mental health problems (the WHO International Classification of Diseases and the American Psychiatric Association’s Diagnostic and Statistical Manual). In many respects it is similar to alcohol which is also legally available, indulged in without problems by most, but which carries dangers and gives rise to very significant public health problems. Almost certainly some forms of gambling carry greater dangers than others (see paras 14-16 below).

6. As it was meant to, the 2005 Act approved the removal of many restrictions on the provision of gambling which were considered to be outdated. The result is, as intended, that gambling opportunities are now more widely advertised, more accessible, and more diverse in Britain than they were. The leading relevant public health model, sometimes known as the total population consumption model, predicts that such a change in accessibility would lead to a general increase in the engagement of the population in gambling, an increase in the numbers gambling in a risky way, and thence to an increase in the prevalence of problem gambling. The same would be true, for example, for an increase in the availability of alcoholic beverages or fatty foods, leading to a general increase in the population’s drinking or body mass index, and thence to an increase in the prevalence of problem drinking or obesity. The rise we are seeing in the prevalence of problem gambling is therefore predictable on theoretical grounds.

7. An increase in prevalence would also have been predicted according to all the major reviews of a nation’s gambling and gambling policy, including highly authoritative reviews by the Australian Productivity Commission and by the US National Research Council, both in 1999, and in Britain the report of the Gambling Review Body in 2001 (the Budd Report), the recommendations of which formed the basis for the 2005 Act. It is particularly telling that the Budd Report, which was always inclined to free British gambling from what were seen as unnecessary constraints, should have recognised the likely effects of its recommendations. Paragraph 17.7 of that report states, ‘a central question for us has been whether increasing the availability of gambling will lead to an increase in the prevalence of problem gambling. The weight of evidence suggests that it will do so.’

8. An increase in the prevalence of problem gambling was therefore widely expected. This expectation constituted one of the main reasons for, and perhaps the principal rationale of, the series of BGPS studies. Knowing that there was a danger that problem gambling prevalence might rise as a consequence of the Act, the surveys were considered a very important check on that possibility. A major hypothesis was that prevalence would rise, or, to put in more formal scientific language, the null hypothesis was that prevalence would not rise. The findings now enable us to fairly confidently reject the null hypothesis. A survey of the general population is always going to have difficulty in detecting changes in the prevalence of a health or social problem which affects a minority of the population, unless the sample size is very large. Sample sizes for the BGPS have been between 7,000 and 9,000 although a number of us argued that sample sizes should be twice that number in order to have much chance of detecting a statistically significant change in prevalence. It therefore came as a considerable surprise that the percentage increase in prevalence was so great that the results were reaching (in the case of one measure) and almost reaching (in the other case) statistical significance.

9. The increase in prevalence was therefore predictable and in line with a major hypothesis. It is also consistent with a number of other findings of BGPS10. As predicted by the total population consumption model, there are indications of the kinds of general population changes which it is thought mediate between greater accessibility and increased problem prevalence. Notably, for the first time the proportion of the population reporting gambling in the past year on any form of gambling other than the National Lottery has exceeded 50% (56% compared to 48% in 2006/07 and 46% in 1999/2000). Engagement in certain forms of gambling has markedly increased, notably betting on sports events other than horse or dog races, playing fixed odds betting machines (FOBMs or B2 machines), and online gambling. A standard set of questions about attitudes towards gambling was also included in 2006/07 and 2009/10. Attitudes in all social groups remain mostly negative towards gambling: large majorities believe, for example, that gambling is not good for society, that there are too many opportunities for gambling, and that gambling should be discouraged; although the majority are against prohibition, disagreeing that it would be better if gambling was banned altogether. But attitudes had moved slightly, and significantly, in a more positive direction by 2009/10. These changes in gambling behaviour and attitudes are all indications that in just three years since the 2005 Act came into operation it is possible to detect the way in which gambling has become more normalised in British society. This is evidence of the mediating link between increased accessibility and increased problem prevalence which is a key component of the total population consumption model.

10. It is important to add that NatCen, who were responsible for carrying out BGPS10 and writing the report of the survey, are very cautious in their interpretation of any results which they produce. They have therefore looked hard at the possibility that the apparently increased problem prevalence might be explained as an artefact of, for example, the changing socio-demographic makeup of the general British population or any possible changes in the way the survey was administered. They have not found any such factor to be responsible and therefore conclude that it is mostly likely that BGPS10 has picked up a real increase in prevalence.

11. I have dealt at some length with this issue of the evidence for increased problem gambling prevalence and the interpretation of this finding because I believe it is so important. Because the increase seems to be a real one, because such an increase was expected and was a major reason for carrying out the series of BGPS studies, and because there is other evidence about how gambling behaviour and attitudes have changed which fit with the major public health model in the field, I believe it is correct to conclude that the most likely interpretation is that the 2005 Act has been responsible for an increase of as much as 40-50% in the prevalence of problem gambling amongst British adults (equivalent to around 100,000 to 150,000 more people with such problems). Research can never pin down any one interpretation with absolute certainty. But I believe it is hard to avoid the interpretation I have given.

12. Problem gambling is unequally distributed across different social groups. Young adults, young men in particular, appear to be most vulnerable. Worryingly, problem gambling prevalence has increased most in those aged 16-44 and according to BGPS10 prevalence amongst all 16-24 year olds is now in excess of 2% (over 3% for 16-24 year old past year gamblers, and 4% for 16-24 year old male past year gamblers). Those who are unemployed appear to be a particularly vulnerable group, having a higher rate of engagement than others in all forms of gambling other than the National Lottery and other lotteries, being over-represented amongst those who spend a relatively large amount of time gambling, reporting gambling for enhancement and coping reasons which are associated with problem gambling, and having a significantly higher prevalence of problem gambling. Black and ethnic minority groups showed a pattern of results suggesting that they are more likely than others to abstain from gambling, but of those who do gamble a higher proportion experience problems with their gambling than is the case for the majority white population.

13. The BGPS does not include under-16s but there is a great deal of other research evidence from Britain and elsewhere that younger adolescents (12-15 year olds) are more vulnerable to problem gambling than adults. It is therefore a continual source of surprise that Britain allows children and adolescents to play category D gaming machines. The Budd report expressed unease about this, and Government, aware of widespread concern, promised a review of this aspect of regulations. This anomaly – no other country allows this – should be put right.

14. I alluded earlier to the likelihood that some forms of gambling are more dangerous than others. It has always been assumed that forms of gambling that allow rapid, repeated play are likely to be more dangerous (casino games and machine gambling, for example, as opposed to playing a lottery or the football pools). But it has been difficult to confirm that from BGPS findings because many people gamble on more than one activity. For example, BGPS10 found playing poker in a pub or club to be the form of gambling associated with the highest prevalence of problem gambling (12.8%) but people engaging in that form of gambling also reported engaging in an average of six to seven other types of gambling. The high prevalence rates found among poker players cannot be attributed to poker necessarily. I have been undertaking some secondary analysis of BGPS10 data which may throw further light on this by using data on the reported frequency of engagement in different forms of gambling, and the amount typically spent in a month, by problem and non-problem gamblers. I have used these data to calculate, for each of fifteen different forms of gambling, the percentage of all days play which can be attributed to problem gamblers, and the percentage of total spend which can be attributed to problem gamblers. The results are currently being checked and the report of the secondary analysis will be available within the next two months. But I can describe the preliminary findings.

15. It is clear that these attributable fractions of days play and spend vary greatly between the different forms of gambling, and that in some cases they are very large. Three forms of gambling – playing casino games in a casino, playing FOBMs (B2 machines), and betting on dog races – stand out as having 20-30% of all days play attributable to problem gamblers (in other words 20-30% of all attendances at these forms of gambling are by problem gamblers), and two of these – betting on dog races and playing FOBMs have percentages of total spend attributable to gambling of between 20-30%. Other forms of gambling which have 10-20% day’s play and spend attributable to problem gamblers are poker playing in a pub or club, playing other types of gaming machine, online machine style games/instant wins, other sports betting, and betting on non-sports events. Betting on horse races, private betting, football pools, scratchcards and bingo have lower figures for attributable days play and spend, and the National Lottery draw and other lotteries have the lowest figures. These figures begin to give a fuller indication of how prominent problem gamblers are as contributors to different forms of gambling, both in terms of the proportion of problem gamblers who attend and the contribution they make to spend.

16. To obtain a complete picture of the contribution which problem gamblers make to British gambling, it is necessary to take into account two further pieces of information: the relative popularity of different form of gambling in the population as a whole and the different average amounts of money which BGPS respondents reported spending per month. Betting on dog races then ceases to be so prominent because it is less popular and average spend is lower than it is for betting on, say, horse races. Playing casino games in a casino continues to be one of the most prominent forms of gambling, particularly because it is the form of gambling with the highest average monthly spend. The form of gambling which remains in the first position, however, is FOBMs which are associated with a relatively high problem gambling prevalence, percentage days play and spend attributable to problem gamblers between 20% and 30%, and a reported average monthly spend more than twice as great as that reported for other forms of machine gambling (in-person or online). This does not constitute incontrovertible proof that FOBMs are themselves more dangerously addictive than other forms of gambling. It could be the case that they are particularly popular amongst problem gamblers, and that a relatively high proportion of FOBM spend is attributable to problem gamblers, but that the origin of these players’ problem gambling lies in their playing of other forms. Although that is possible, it seems unlikely, particularly since FOBMs combine many of the features which are believed to make certain forms of gambling dangerous and because agencies providing treatment and advice for problem gamblers in Britain are now reporting that their clients very often describe difficulties controlling their FOBM play. Therefore, although the case is not proven beyond all dispute, there is sufficient evidence to suggest that B2 machines, introduced only a few years ago without much consultation and without a proper impact assessment, constitute a particularly dangerous form of gambling.

17. In conclusion, results of the latest British Gambling Prevalence Survey provide convincing evidence that the prevalence of problem gambling among British adults constitutes a significant public health problem which should be taken more seriously, that groups such as unemployed people and those living in more deprived areas are most at risk, and that problem prevalence has risen substantially since the 2005 Act came into operation. Some forms of gambling are more dangerous than others, notably category B2 machines. The 2005 Act has been responsible for exposing the British people to greater danger than was previously the case. Public attitudes are predominantly opposed to the greater accessibility to gambling which the Act has brought about. Policy should now move towards a more precautionary approach and should be less influenced by provider interests.

June 2011

Prepared 1st August 2011