Bank data can provide objective measures of gambling expenditure and associated outcomes that is longitudinal (7 years) and mass-scale (6.5 million people)

New research in @NatureHumBehav w/ Muggleton @PParpart Leake @johngathergood & Stewart

https://www.nature.com/articles/s41562-020-01045-w

Prevalence surveys are the main tool governments use to understand gambling in the population. They typically involve small sample sizes (e.g., 10,000) and require people to accurately and truthfully report their gambling.

Whereas bank data…

Records online gambling expenditure and many offline expenditures

43% of people made at least one gambling transaction (similar to prevalence levels)

The average gambler spent £125 in a year, and the top 1% of gamblers spent £22,000

But bank data can tell a lot more…

It can correlate gambling expenditure with financial outcomes (savings, missed credit card payments) but also social and health outcomes (social activities, prescription spend), and even mortality

31 outcomes in total were recorded

The x-axis in each graph ranks gamblers by the proportion of their monthly spend that is on gambling, so that 100% = the gambler with the highest proportion of spending on gambling

1%, 4%, 15%, and 72% of spending goes on gambling at 50%, 75%, 90%, and 99% on each x-axis

Each y-axis shows a different outcome

This graph shows mortality as an outcome. Gambling is associated with increased rates of mortality, such that someone gambling 30% of spending has about a 37% increased risk of mortality

Gambling persistence can also be investigated this way. Of a sample gambling 10% or more of spending in 2015, over 25% were not gambling at all in early 2012, whereas less than half that proportion had stopped gambling by the end of 2018. Gambling can be a behavioral addiction.

These are correlational (noncausal) results. This research highlights the value of longitudinal and objective data for the gambling field.

Research supported by @ESRC @WarwickBSchool @warwickuni @CQUni @uniofoxford @DSPI_Oxford @summerfieldlab

My email is in my bio if anyone who cannot access a pdf would like one