Gambling in the UK is a very widespread activity, with 47 per cent of adults in the 2019 Gambling Commission Survey reporting that they had gambled at some point in the last four weeks. Although gambling is by no means a new phenomenon, the digital age has changed the ways that people can gamble, and the UK Government has launched a review of the Gambling Act 2005 to ensure it reflects these transformations.

Against this backdrop, the Policy Institute at King’s College London was commissioned by Action Against Gambling Harms (AGH) to understand better the available evidence around gambling harms. Our objective was to ascertain where there were gaps in the research base and to make recommendations about which avenues of research would be both feasible and valuable. As requested by AGH, we focused our attention on four key topics:

Within each of these areas, we focused our efforts on a) harms, b) characteristics of those experiencing harms, c) behaviours associated with harm and d) interventions to mitigate harm. To achieve this in the limited timeframe allowed by the announcement of the Government’s Review, we conducted a rapid evidence assessment to identify where the research gaps were; a fuller description of our method can be found in the introduction below.

Our headline finding is that there is a real lack of empirical research on gambling harms in the UK. In particular, we found significant evidence gaps around the harms that gambling in general, and problem gambling as a specific subcategory of overall gambling, might cause. We also note that, if this gap is to be filled with robust empirical research, then more funding needs to be channelled to gambling related research through independent sources.

Our review identified a series of specific research gaps that cut across the four areas on which our study focused. First, the UK appears to lag behind other countries in empirical research into gambling harms. In comparison with other countries such as Australia, Canada and the US, we found relatively little research on gambling harms in the UK context. This was true across the four areas that we explored and appears to be a characteristic of the wider evidence base. This is important as there are limits to the extent to which overseas findings are translatable to the UK context, given the potential differences in socio-cultural factors and the gambling ecology (for example, legal gambling age, availability of gambling products and venues etc.)

Second, we found a lack of conceptual literature on what constitutes a ‘gambling related harm’ and a related lack of scholarship on how to measure such a harm. Although not a targeted part of the review per se, we were struck by the limited body of work defining harms and how to measure them.

Third, although we found evidence to suggest a strong correlation between poverty and gambling, we found no research which established the direction of causation between the two, nor the conditions under which poverty causes gambling and gambling causes poverty.

Fourth, we found a very significant gap on the costs of gambling to the public purse in the UK context. At best we found some estimates of the extra fiscal cost of problem gamblers in a few very specific settings, but we found no study in the UK context that systematically mapped out the costs of gambling on public services, despite the fact that studies of this kind have been conducted in other countries (eg Australia and New Zealand). Innovative methodologies will be needed to establish these costs, including public health and ‘burden of disease’ style approaches.

Fifth, we found a relative lack of research that charted the effects of gambling over a long timeframe to fully understand the whole ‘journey’ of gamblers, and the wider effects that gambling causes over a longer duration.

Sixth, we identified some methodological limitations associated with the existing evidence base. In particular, there is a heavy reliance on self-report data, which is at risk of a range of biases, rather than the use of objective measures of gambling related harms. Second, much of the data used is of a cross-sectional nature, precluding conclusions about the direction of the relationship between gambling behaviours and other outcomes. The analysis of data from prospective longitudinal studies, which follow the same participants over time, would help to address this gap in the evidence, though we note that these studies can be costly and difficult to establish. A more practical and feasible approach may be to ensure that existing longitudinal or cohort studies include questions on gambling.

Seventh, the evidence we identified in this review points to differences in motivations and characteristics between subgroups of problem gamblers, as well as the harms that they experience. Much more needs to be understood about these differences and how to effectively tailor public health initiatives and harm reduction interventions to the meet the specific needs of these populations.

Finally, we note that although our review was targeted at specific research areas, within these, we found little mention on the effects that Covid-19 might have on gambling behaviours and the harms of gambling, both in the short-term and in the longer-term.