News

New insights into gambling behaviour

Article dated: 25 June 2017

Gambling behaviour has been a topic of research for decades due to its complex implications. However, most of the evidence has been gathered from cross-sectional studies, which are only able to draw conclusions at a certain point in time and only provide an overview of how gambling behaviour has changed in the population, rather than with the individuals. 

The UK Data Service is pleased to announce the release of the Gaming and Betting Study: Survey of Loyalty Card Customers, Waves 1 and 2, 2014-2016 data which examines individuals' gambling behaviour across 2 waves.

The Gaming and Betting Study: Survey of Loyalty Card Customers was commissioned by the Responsible Gambling Trust (now GambleAware), as part of a programme of research looking at users of machines in bookmakers. The study addresses the lack of data in this sector and allows a more in-depth investigation of individuals’ gambling behaviours. The baseline survey, wave one data, was designed to examine whether industry data generated by machines in bookmakers could be used to distinguish between harmful and non-harmful patterns of play; the follow-up survey, wave two data, was targeted to explore changing patterns of gambling behaviour over time, to examine changes in problem gambling behaviour, and to identify who is more likely to change their 'problem gambling status'. The final version of the study contains data for all 1552 individuals who gave a full interview at both waves.

Their 'problem gambling status' was measured by the Problem Gambling Severity Index (PGSI), based on the answers to nine questions and categorises people as non-problem gamblers, low risk gamblers, moderate risk gamblers or problem gamblers.

Almost half of the participants in the study have seen their PGSI status change between the baseline and follow-up study: 6% of participants became problem gamblers at follow-up and 9% of participants moved from problem gambler to at risk or non-problem gambler status; 29% of the non-problem gamblers at the baseline had increased their PGSI scores, becoming at least low risk gamblers at the time of the follow up, and 1% becoming problem gamblers. However 41% of the problem gamblers had decreased their PGSI score, no longer classified as problem gamblers at the follow-up, with 7% becoming non-problem gamblers. 

From a socio-demographic perspective, those aged 44-65, those from Black/Black British ethnic groups, and women, were more likely to be classified as problem gamblers at follow-up compared to baseline; moreover, those from minority ethnic groups and with lower levels of personal income were more likely to have increased their PGSI scores than others.

When addressing the participation in gambling activities, researchers have discovered that those who were unemployed or who lived in the most deprived areas of England, Scotland and Wales were more likely to start gambling on machines in bookmakers than others; and 60% of participants who started gambling on bookmakers machines in the past four weeks tended to increase their engagement in other gambling activities.

Another interesting finding is the fluidity of patterns of gambling participation, as 28% of respondents participated more often in gambling activities, and the same percentage participated less often, when comparing baseline activities to follow-up activities.

Furthermore, by analysing the data, researchers have identified the key predictive socio-economic factors associated with becoming a problem gambler are age, ethnicity and income.

Further information about the survey and the results is available in the NatCen Baseline and Follow-up reports. You can download the data from the UK Data Service.

Discover UK Data Service