Experimental Design
Context:
In May 2021, the Swiss government amended the Federal Narcotics Act (NarcA), enabling pilot trials for recreational cannabis under the BetmPV ordinance (BetmPV, 2021). Our study is part of a pilot trial in the Canton of St. Gallen and is conducted in collaboration with our field partner, the non-profit Cannabis Research Association (Verein Cannabis Research, VCR). VCR is responsible for the operational and legal aspects of the pilot trial, including establishing distribution channels and serving as the primary point of contact for study participants. The research team is responsible for the study design and independent scientific evaluation.
The pilot trial comprises two parts that draw on a common subject pool but address independent research questions.
Part 1 (this preregistration) precedes Part 2 and studies the WTP for legal cannabis, as detailed in this preregistration
Part 2 (to be preregistered separately) is a randomized controlled trial (RCT) that measures the effects of legal access to cannabis on consumption, crime, employment, and education, with a particular focus on distribution models (i.e., physical retail outlets versus home delivery).
Recruitment and Eligibility:
Part 1 and 2 of the pilot trial draw from a common participant pool. Screening and eligibility assessment are conducted jointly for both parts and follow the inclusion and exclusion criteria set out in the federal regulation (Article 14 BetmPV, 2021).
The study uses a cohort-based rollout. Recruitment and eligibility screening occur continuously, but participation begins at fixed points in time. We plan to recruit multiple cohorts with start dates spaced about two to three months apart. All participants who are considered eligible by a given cutoff date enter the study together as a cohort.
Within each cohort, participants are randomly assigned to either Part 1 or directly to Part 2. Assignment to Part 1 will continue until either (i) the overall target sample size of 2,500 participants who have completed both Survey 1 and Survey 2 is reached or (ii) two years after the start of the first cohort. The remaining participants proceed directly to Part 2 without participating in Part 1. Part 1 has a duration of approximately ten weeks. Only those participants who complete all required surveys in Part 1 are eligible to continue to Part 2.
Part 1 includes a four week buffer period before the start of Part 2. Participants who are not assigned to Part 1 enter Part 2 after a waiting period of approximately ten weeks. The share of participants randomly selected for Part 1 within each cohort depends on the number of eligible applicants. For Part 1 we aim for a total of 2,500 participants who have completed Survey 1 and 2.. We additionally recruit up to 10 participants for an initial test cohort. This cohort will start before the official launch and is intended to test all study procedures of Parts 1 and 2. Data from the test cohort will not be included in the main analysis.
Randomization and Treatment Assignment:
Eligible participants who are selected for participation in Part 1 will be randomly assigned (further details on randomization below), with equal probability, to one of the four experimental conditions.
Data Collection:
As part of the enrollment procedure all participants from Parts 1 and 2 complete a baseline survey. Participants will be invited (on a voluntary basis) to submit 2 grams of their own illegal-market cannabis for laboratory analysis. Upon submitting a sample, they are also asked to complete a very short questionnaire about their illegal-market sample (the “illegal-market sample survey”).
Participants who are randomly selected for Part 1 will be invited to complete two online surveys (Survey 1 and 2) that are administered four weeks apart. These surveys serve as our main data source for Part 1 and are used to elicit participants’ WTP using an incentive-compatible BMD method. In the BDM, participants may use part of their participation fee of CHF 80 to purchase 4 grams of a legal cannabis product.
In each survey round, we elicit participants’ WTP for four legal cannabis products: (i) cannabis flowers with THC concentrations of 5% THC, (ii) 12% THC, (iii) 20% THC, and (iv) hashish with 20% THC. After completion of Surveys 1 and 2, the computer randomly selects one BDM decision and a corresponding price for each participant and implements the outcome accordingly.
The surveys also elicit participants’ qualitative and quantitative prior and posterior beliefs about the prevalence of potential health risks in illegal-market cannabis. We also measure beliefs about psychoactive potency, consumption experience, and purchasing convenience of legal-market cannabis. Subjects further complete a hypothetical BDM to measure their WTP for illegal-market cannabis. All survey instruments and the full BDM instructions are included as supporting documents with this preregistration.