Can indirect methods be used as an approximation to estimate alcohol excise taxation revenue?

Introduction: Alcohol excise taxation is one of the World Health Organization’s (WHO) “best buys” for reducing alcohol consumption and attributable harms. Despite its proven effectiveness, alcohol taxation remains globally underused, partly due to Member States’ concerns about potential revenue losses. Robust estimation methods are essential to predict future revenues after policy changes. This study compares two approaches for estimating alcohol excise tax revenue per capita and assesses their validity against reported figures. Methods: Reported alcohol excise tax revenue per capita for 2022 was retrieved from official government sources for 17 countries in the European Union and the United Kingdom. These figures were compared to estimated values generated using two different methods: 1) a tax-and-consumption-based approach using beverage-specific tax and consumption data, and 2) an excise share-based approach using the linear relationship between excise tax share and revenue. Comparisons were made using visual inspection of scatterplots, relative error plots, and Bland-Altman plots, correlation analysis and linear regression models. Results: Both methods showed strong correlation with reported revenue (tax-and-consumption-based: r = 0.959; excise share-based: r = 0.857) and high explained variance of the reported revenue. The tax-and-consumption-based method demonstrated higher precision, narrower limits of agreement, and significantly better model fit. Conclusions: Both methods provide useful alternatives in the absence of reported data, but the tax-and-consumption-based approach offers a more accurate and consistent estimate of alcohol excise tax revenue per capita. These findings support its use in global monitoring and fiscal modelling, particularly in the context of WHO technical support to Member States.