# The effect of television advertising on gambling behaviour: a quasi-experimental study during the 2022 Qatar FIFA World Cup

## TL;DR

This study shows that TV gambling ads during live football in England increase men's betting, suggesting stricter ad rules could help reduce gambling harm.

## Contribution

A novel pseudo-randomized quasi-experiment to assess the impact of TV gambling advertising on real-time betting behavior.

## Key findings

- Betting frequency was 16% to 24% higher during games with gambling ads compared to those without.
- Participants were 22% to 33% more likely to bet during games with gambling advertising.
- Current ad restrictions may not be sufficient to curb gambling behavior during live sports.

## Abstract

•This study indicates that television gambling advertising increases betting during live football for men in England.•These increases occur despite current industry-led restrictions on advertising during live sporting events.•Results were highlighted using a novel, pseudo-randomised quasi-experiment.•Restricting advertising during live sport may be a useful policy tool to reduce gambling behaviour and potential harm.

This study indicates that television gambling advertising increases betting during live football for men in England.

These increases occur despite current industry-led restrictions on advertising during live sporting events.

Results were highlighted using a novel, pseudo-randomised quasi-experiment.

Restricting advertising during live sport may be a useful policy tool to reduce gambling behaviour and potential harm.

Gambling is a recognised public health problem, and population-level advertising restrictions may be an effective way to reduce gambling-related harm. This study fills an important evidence gap by employing a quasi-experiment to estimate the impact of television gambling advertising on gambling behaviour during the 2022 Qatar World Cup. It exploits the variation in gambling advertising between two broadcasters to evaluate its impact on gambling behaviour in a real-world setting.

Poisson and Logistic panel data regression models using the broadcaster in place of gambling advertising are employed with individual fixed effects and match-level controls. Betting data is from a purposive sample of 365 men aged 18 to 45 who bet on football in England.

Frequency of betting on football was 16 % to 24 % higher during games televised on a channel with gambling advertising compared to one without [IRR: 1.16 – 1.24,p < 0.01]. Furthermore, participants were 22 % to 33 % more likely to place a football bet during games that contained television gambling advertising [OR: 1.22 – 1.33, p < 0.01].

Using a pseudo-randomised quasi-experiment, this study found that television advertising significantly increased both the likelihood and frequency of betting during live football games in the sample studied. This raises concerns about the adequacy of current advertising restrictions in the UK and suggests that a policy which restricts television gambling advertising around live football might be an effective part of a wider public health strategy to tackle gambling-related harms. Future studies should replicate this design, using larger, more representative samples, to inform policy.