How to interpret studies on the impact of legalizing cannabis

We appreciate the systematic review of cannabis legalization’s impacts in Canada by Hall et al. [1], whose content overlaps that of a previous review [2]. Summaries of the ever-growing legalization evidence base are important for both researchers and policymakers. For example, Germany’s Health Ministry asked us to review the literature to inform that country’s legislative planning. After surveying 164 studies from Canada, Uruguay and the United States [3], our conclusions were similar to those of Hall et al. However, we wish to highlight three points that merit greater consideration in future research. First, before legalization, cannabis use in Canada had already been increasing for years [4]. Cannabis use prevalence and cannabis-related emergency department visits in Ontario were already rising before 2018’s legalization [5, 6], and there were apparent changes in alcohol sales after medical cannabis usage expanded in 2015 [7]. This means that any results from simple before-and-after legalization comparisons should be interpreted with great caution. We therefore recommend that studies of post-legalization changes account for pre-legalization trends and appropriately identified control groups to avoid overstating legalization’s impacts. This could be conducted, for example, by estimating trends prior to legalization, predicting those trends out to the future and then comparing them with actual results [8]. Secondly, after legalization, it took time for Canada’s new legal market to establish itself: store counts grew every year [9], as did consumers’ willingness to disclose usage [10]. This means that researchers’ post-legalization time-frames greatly affect their likelihood of detecting effects and that literature reviews should not give the same weight to studies of, for example, the first year of legalization as to those covering year 3. For example, when we compared studies with fewer than 2 years of post-legalization data to those with more than 2, the latter provided much clearer indications of increased consumption and health outcomes [3]. We therefore recommend that other researchers place more emphasis on longer-term post-treatment study designs. Thirdly, although Canada legalized nation-wide, there were meaningful implementation differences among its 13 provinces and territories regarding retailing (e.g. government-owned versus business), consumption (e.g. public smoking allowed versus banned) and products allowed. These differences have subsequently been seen in their outcomes. For example, hospitalizations for cannabis poisonings in children increased overall after legalization, but the increases differed according to the degree of commercialization and the product types sold [11]. For academics and policymakers, those interjurisdictional differences are at least as interesting as the national averages. We therefore recommend that researchers pay more attention to this heterogeneity in study designs. For all these reasons, it is important for researchers to treat cannabis legalization as a complex process, rather than an instantaneous binary intervention. Where data quality permits, studies should increasingly account for past societal trends, ongoing market maturation and interjurisdictional differences. Researchers should also complement survey-based studies with other methods, such as analysis of work-place toxicological tests [12] or wastewater data [13]. Jakob Manthey: Conceptualization (lead); writing—original draft (lead); writing—review and editing (equal). Michael J. Armstrong: Conceptualization (supporting); writing—review and editing (lead). Tobias Hayer: Conceptualization (supporting); writing—review and editing (supporting). Daniel T. Myran: Conceptualization (supporting); writing—review and editing (supporting). Rosalie Liccardo Pacula: Conceptualization (supporting); writing—review and editing (supporting). Rosario Queirolo: Conceptualization (supporting); writing—review and editing (supporting). Jürgen Rehm: Conceptualization (supporting); writing—review and editing (supporting). Marielle Wirth: Writing—review and editing (support); Frank Zobel: Conceptualization (supporting); writing—review and editing (supporting). None. Open Access funding enabled and organized by Projekt DEAL. J.M. has worked as consultant for and received honoraria from public health agencies. All other authors do not declare any conflicts of interest. Not applicable.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.