Estimating and comparing the effects of price- and potency-based taxes on cannabis purchase patterns in an experimental cannabis marketplace.

Background and aims: Recreational cannabis legalization has made high-potency cannabis products more readily available, raising public health concerns. While price-based taxes are common, potency-based taxes have been adopted as a potentially more effective approach to mitigate the harms associated with high-potency cannabis consumption. This study aimed to estimate and compare the effects of price- and potency-based taxes on cannabis purchase patterns.

Design: In May 2024, we conducted an online Experimental Cannabis Marketplace experiment, where cannabis products varied in product categories, delta-9-tetrahydrocannabinol (THC) levels, and prices. Participants completed hypothetical purchase sessions with varying tax rates and tax types. Potency-based taxes were designed to be higher for high-THC products and lower for low-THC products than price-based taxes. The Poisson random effects model was used to estimate the association between tax rates and each outcome for the two tax types. The Wald tests assessed differences in coefficients between the two tax types.

Setting: U.S. states with recreational cannabis legalization.

Participants: A total of 1250 adult cannabis users.

Measurements: Cannabis purchase patterns, including quantity demanded, THC demanded, and expenses.

Findings: For both price- and potency-based taxes, a higher tax rate was associated with a lower quantity demanded (total, by product category, and by THC level), lower total THC demanded, increased total expenses, and increased tax expenses (all ps < 0.01). The price elasticity of quantity demanded was estimated to be -0.46 and the price elasticity of THC demanded was estimated to be -0.48 to -0.52. Compared with price-based taxes, potency-based taxes were associated with a lower quantity demanded for high-THC products (elasticity for potency-based taxes = -0.59 and for price-based taxes = -0.49; p for elasticity difference = 0.046) and a lower proportion of total expenses on high-THC products (elasticity for potency-based taxes = -0.11 and for price-based taxes = -0.05; p for elasticity difference = 0.028). No differences were observed between the two tax types in quantity demanded for all products, total THC demanded, or total expenses.

Conclusions: Potency-based taxes on cannabis products appear to be more effective at reducing demand for high-potency products compared with price-based taxes.