The Social and Economic Consequences of Cannabis Legalization: A Randomized Controlled Trial

Eligible and consenting participants will be randomly allocated to one of two experimental groups and a control group to measure the causal effects of legal access to cannabis and a new self-regulation program (“Harm Reduction Plus, HR+):

• Control group “NoAccess”: Participants in the control group will not have access to legal cannabis for the entire duration of the RCT and will therefore continue to rely on the illegal market to obtain cannabis.

• Experimental group “Access”: Participants in the "Access" group are given legal access to cannabis products, which they can obtain in person from stores in the canton of Zurich run by the "Swiss Cannabis Research" association (“Verein Cannabis Research”, VCR) for up to four years. Participants in the “Access” group gain access to a product range that mirrors products available on the illegal market. This includes cannabis flowers, cannabis resin (hashish), cannabis for consumption with vaporizers or vape pens, edibles, and oils .

• Experimental group “Access & HR+”: Participants in the "Access & HR+" group are also given legal access to cannabis products, which they can obtain in person from the stores in the canton of Zurich run by VCR, just as participants in the "Access" group for up to four years. Participants in “Access & HR+” additionally receive an invitation for a voluntary self-regulation program to prevent excessive cannabis consumption called Harm Reduction Plus (HR+).

2024-05-02

2029-05-02

The primary outcomes are:
1. Cannabis consumption
2. Criminal behavior
3. Employment outcomes
4. Educational attainment

We analyze these primary outcomes within the scope of four separate work packages / projects, which focus on different research topics (see corresponding pre-analysis plans) .

In addition, we will analyze the following secondary outcomes:

1. Consumption of other drugs (i.e., alcohol and illicit drugs)
2. Health and wellbeing
3. Cognitive capacity


We will conduct further exploratory analyses using the following other outcomes:

4. Nicotine consumption
5. Social relationships
6. Self-evaluation
7. Perceived stigmatization of cannabis use
8. Perceived health risk of cannabis consumption
9. Customer satisfaction
10. Sources of illegal cannabis

For descriptive purposes, we will also collect cannabis samples from the illegal market from a random subset of our participants to get laboratory measurements of cannabis quality.

Finally, we assess take-up of the three elements of the self-regulation program in treatment Access & HR+ and capture how often participants reflect about their cannabis consumption as a manipulation check.

The 2021 amendment to the Federal Narcotics Act and the associated ordinance on pilot trials (BetmPV) sets the boundaries for scientific pilot trials involving cannabis for recreational purposes. Trials are limited to a duration of five years and may not exceed 5,000 participants gaining access to legal cannabis. Participant may not be given access to more than 10 grams of pure THC per month.

Recruitment and eligibility:

We aim to recruit and enroll up to 7,500 participants (of which a maximum of 5,000 will receive access to legal cannabis) in several cohorts. Each cohort participates in the trial for up to four years. All participants will be recruited according to the inclusion and exclusion criteria for this RCT, which are defined in article 14 of the BetmPV. Individuals can join pilot trials if they (i) can demonstrate their current consumption of narcotics of the cannabis type, (ii) reside in a community where the pilot trial is taking place, (iii) agree to the conditions of the scientific study, and (iv) provide written consent. Participation is not open to minors, pregnant or breastfeeding individuals, those who are incapable of judgment, or those with a medically diagnosed illness for which cannabis use is contraindicated. To be eligible, individuals must have consumed cannabis at least once in the six months leading up to onboarding.

Randomization and treatment assignment:

Eligible participants who provided informed consent and completed the baseline survey will be randomly assigned into one of the three experimental conditions with equal probability (see more details below).


Data Collection:

The analysis is primarily based on administrative register data and administrative data from VCR to measure the impact on the participants’ labor market outcomes, criminal behavior, educational attainment, and consumption of legal cannabis. In addition, we will conduct regular online surveys administered through Qualtrics to measure the consumption of illegal cannabis and other substances to get a better understanding of the underlying mechanisms. We also collect data from the study’s web application for information regarding the take-up of the self-regulation program HR+ and general usage of the web application. Additionally, we are collecting descriptive data on the development of illegal cannabis quality (assessed through laboratory analysis) over the course of the RCTs.

Not available

Randomization done by statistical software

To minimize the risk of spillover effects we implement a clustered randomization. We define clusters as a group of neighboring and/or acquainted participants. To create these clusters, VCR collects participant addresses and asks participants during the onboarding process to name any acquaintances that are also part of the pilot study. VCR then assigns people that know each other (one-sided links) or that are neighbors (i.e., living within a radius of 50 meters) to the same cluster. Due to the rolling admission process, we do not know the number or size of all individual clusters beforehand. To minimize the effects of such different cluster sizes, we will perform a size-stratified cluster randomization . Using this method, for each cohort we randomly assign clusters to one of the three experimental groups so that the distribution of cluster sizes in each group is approximately equal.

Yes

Based on the recruitment of the first cohort with a sample size of 7500 participants we would expect roughly 6300 clusters.

To comply with the legal upper bound of 5,000 participants gaining access to legal cannabis, we recruit a maximum of 7,500 participants, randomly allocating applicants from each cohort in equal parts to one of the two experimental groups (2,500 participants in each treatment group; 5,000 in total) and the control group (2,500 participants). The actual sample size will probably turn out to be less than 7,500 as it is determined by the number of eligible applicants who provide informed consent and complete the baseline survey for each recruitment cohort. If enrollment progresses slower than expected, we may seek permission from the Federal Office of Public Health (FOPH) to recruit subjects from cantons other than Zurich to reach the maximum sample size of 7,500 participants. Moreover, we are planning a second pilot trial in the canton of St. Gallen with a maximum of 5,000 participants. While this pilot trial has a different focus it will, among others, include treatments NoAccess and Access. Depending on a timely approval of the St. Gallen pilot trial by the FOPH we will pool observations from all cantons for treatments NoAccess and Access to further increase statistical power.

The maximum number of participants per treatment arms are:

• Control group “NoAccess”: 2,500 participants
• Experimental group “Access”: 2,500 participants
• Experimental group “Access & HR+”: 2,500 participants

Note that conditional on being able to pool the data with our second pilot trial in St. Gallen the maximum sample size for Access and NoAccess might exceed 2,500 participants.