Background: Although exposure to alcohol marketing has been identified as a driver of alcohol consumption, evidence on the real-world effectiveness of strong alcohol marketing bans in reducing population-level consumption remains debated. The current evidence base is undermined by methodological weaknesses, particularly the absence of reliable counterfactual trends. As a result, it has been difficult to separate the impact of marketing restrictions from underlying consumption trends and broader contextual changes affecting alcohol use. Novel statistical methods using synthetic control groups offer a strong design for causal inference in policy evaluation. They can overcome key limitations of earlier studies and generate stronger evidence to inform alcohol control policy.  

Objectives: We aimed to estimate the impact of the 1993 French alcohol marketing ban (the Loi Evin) on population-level alcohol consumption over the six years following its implementation.  

Methods: We used a synthetic control difference-in-difference framework to estimate the average treatment effect of the Loi Evin policy on adult alcohol consumption per capita (APC) during the first six years post-implementation. Synthetic control trends were constructed in the 15 years prior to 1993 using data from a donor pool of 8 European countries with comparable APC trends and no major concurrent alcohol policy change over the study period. Models were adjusted for covariates including GDP per capita. 

Results: Our findings suggest that the introduction of the Loi Evin marketing ban was not associated with an observed reduction in APC in France relative to its synthetic control. Estimated annual effects in all post-intervention years were small and heterogeneous, ranging from −0.05 to 0.08, and not statistically significant. Results remained consistent across sensitivity analyses, including alternative donor pools and shifts in pre-intervention period lengths. 

Conclusions/Implications for practice or policy: Findings suggest that the French Loi Evin did not produce a measurable short-term reduction in population-level alcohol consumption in France.  This limited effect may be partly explained by the continued authorisation of marketing in certain media and the rapid weakening of some restrictions under industry pressure. Nevertheless, the law may have influenced other outcomes, including youth alcohol consumption and longer-term processes of alcohol denormalisation that could not be evaluated in this study. This suggests the need for comprehensive, consistently enforced marketing bans, along with monitoring of other relevant outcomes to accurately assess public health impact.