Research & consultations

British Gambling Prevalence Survey 2007 secondary analysis: gambling participation and problem gambling

This analysis was undertaken by Professor Leighton Vaughan Williams, Director of the Betting Research Unit at Nottingham Business School, Nottingham Trent University, in conjunction with Dr Lionel Page, Dr Jonathan Parke and Jane Rigbye.

The team applied sophisticated statistical techniques to the dataset of the British Gambling Prevalence Survey 2007 to gain further insights about patterns of gambling participation and problem gambling.  The report makes a number of recommendations for improving the methodology of future prevalence surveys and is based on priorities for analysis identified by the Commission in conjunction with the Department for Culture, Media and Sport.

The full report can be found at:

British gambling prevalence survey: secondary analysis - October 2008

Key findings

Although many of the findings mirrored those of the BGPS 2007, this analysis offered further insights into problem gambling behaviour.

The authors reported that when considering the association between gambling activity and problem gambling it is important to consider the frequency with which gamblers participate in that activity rather than simply past year and past week participation. This finding should be interpreted as inviting caution about the order of activities associated with problem gambling in the BGPS 2007 rather than as suggesting an alternative ranking. It was not possible through this analysis to learn what causes problem gambling.

The BGPS 2007 report tended to suggest, in the absence of considering frequency, that a higher level of problem gambling was linked with spread betting, fixed odds betting terminals (FOBTs) and betting exchanges. This secondary analysis found that when frequency of play was taken into account the ranking of the activities most linked with problem gambling (ie horse racing, slot machines and scratchcards) was not the same as the BGPS 2007. This must be interpreted very carefully, however as the estimation technique did not allow us to say if these gambling activities are significantly more closely linked with problem gambling than other activities.

Individuals who were male, single, had a lower level of qualifications, lower incomes and whose parents had gambled were more likely to report gambling problems. However, more generally, any individual regardless of gender, marital status, education and family background was deemed to be at risk.

Most gamblers tend to participate in core activities (such as the National Lottery, bingo and scratchcards) though there was a distinct subset of gamblers who participate in more procedurally sophisticated or technical types of games (such as spread betting, betting exchanges and online gambling).

Participation in gambling activities varied across a range of socio-demographic variables. For example, younger players tended to prefer to gamble on slot machines or through private betting, males were more likely to gamble online, on spread betting, on betting exchanges and on FOBTs.

The study also made recommendations for improving the methodology of future prevalence surveys, including collecting data on additional variables that may provide better insights into the determinants of problem gambling and gambling behaviour. 

 

Page last reviewed: April 2011