The impact of alcohol taxation on alcohol-attributable health inequalities in the US: a microsimulation study

Background: This study aims to use a novel microsimulation model to investigate how increasing alcohol taxation could impact education-based inequalities in the alcohol-attributable health burden in the United States (US).

Methods: An individual-level dynamic microsimulation model of alcohol use for the years 2000 to 2019 was developed that is representative of the US national population for subgroups defined by age (18-69 years), sex, race and ethnicity, and educational attainment as indicator of socioeconomic status (SES). Risk functions for ten alcohol-attributable causes of death were used to model each simulated individuals’ risk to die given their alcohol use. Model parameters were calibrated using a Bayesian approach. Counterfactual scenarios were generated to explore the impact of a one-time increase in the federal alcohol excise taxes on years of life lost due to alcohol use by sex and education group in the year of the policy (2015) and subsequent years. We tested scenarios of a 20%, 50%, and 100% tax increase, accounting for price effects on drinking participation and alcohol consumption levels.

Results: Simulated alcohol-attributable years of life lost were lower in scenarios with higher alcohol taxation increases. These savings in the years of life lost varied by education groups, with steeper decreases in years of life lost in lower education groups. Some variations in these education-based differences were observed across causes of deaths and sex. The observed effect of the one-time intervention diminished in the years after the tax increase.

Conclusions: Higher federal alcohol excise taxes are expected to lower the alcohol-attributable health burden in the US adult population in general and among those with low socioeconomic status in particular. This policy can therefore help to narrow recently increasing socioeconomic inequalities in life expectancy in the US.