The Impact of Excise Tax Structures for Retail Marijuana on Marijuana Consumption

The investigators will follow probability-based cohorts of adult non-medical marijuana users (N=1,500; ages 21+) and susceptible adolescents and young adults (AYA) users (N=1,000; ages 15-20) over four years. The investigators will use a prospective survey to collect marijuana use information and will conduct volumetric choice experiments (VCEs) where participants choose quantities to purchase among three legal product prototypes: flower, edible, and concentrate; and one illegal flower product, under simulated retail marijuana tax structures. The presentation of edibles and concentrates will change over time to capture the range of product varieties in the market and to accommodate new products and characteristics as they evolve. Choice experiments including VCEs use a series of manipulations as a measurement device to elicit preferences for products and consumption. Therefore, the experimental design is within-subjects with repeated measures, which allow for causal identifications of the impact of design attributes (e.g., tax bases and rates) on preference and consumption outcomes. The attributes and levels are selected to reflect THC levels in the market, existing (weight, price, tiered price by forms) and novel (potency) tax bases, existing (10%-40% of retail price or equivalent in other bases) and potential (high 60% and 80% of retail prices or equivalent) tax rates, and price distribution for each product. Price distribution levels (5 to 95 percentiles) will come from our price data collection and will be updated over time and vary by participants' state and locality of residence.

Before the experiment, study personnel will instruct participants 1) how to answer the choice experiment question, including considering their weekly expenditures when answering VCE questions; 2) that there is no access to other products, no stockpiling, no selling or giving away products to others, etc., to ensure that participants choose products that reflect their usual purchase patterns for self-use; 3) that there are no right or wrong responses. Study personnel will instruct participants to answer all questions honestly, thoughtfully, and to the best of their understanding as if they were actually in this situation, to reduce hypothetical bias; 4) that there may be potential costs or issues associated with purchasing illegal products, including risks of penalties, undefined potency, possible contamination, and inconsistent product quality; and 5) that participants should keep their weekly budget in mind when making purchase choices of the products. In VCEs, the investigators will manipulate THC levels, tax rates and bases, and price levels. Participants will be asked to answer the following: "Imagine these are the marijuana products that you (participants) can purchase for non-medical use, how many of each product would you (participants) most likely buy next time?" Participants can choose more than one product, and as they choose quantities, the total balance will be displayed. To make the tasks easier, study personnel will 1) set a warning to inform participants if their reported budget exceeds their weekly budget for spending; 2) provide a dropdown list for quantities or, if they would like more units than the listed, they can fill in a number; and 3) all products are described as high-quality to allow for a comparison of prices. Half of the participants will be randomized to see itemized pre-tax tag price, excise taxes, and sales taxes, as requested in some states (e.g., Illinois); the other half will be randomized to see only prices that are inclusive of excise taxes. This will allow us to whether excise taxes are more effective when applied at checkout compared to hidden in the tag prices.