Exploring the impact of drug decriminalization and legalization policies on mental health outcomes: A scoping review.

In recent years, global drug policy has undergone a significant shift, moving away from punitive, prohibitionist approaches toward more reform-oriented models. These “liberal drug policies”—commonly defined as regulatory frameworks that include decriminalization, legalization, and commercialization—aim to reduce criminal penalties, regulate markets, and prioritize public health. To date, over 30 countries have implemented some form of decriminalization to lessen the legal consequences associated with personal drug use [ 1 ]. In 2018, the United Nations guidance has emphasized the importance of aligning drug policy with public health, development, and human rights objectives, noting that countries should “place people, health and human rights at the center” of drug policy responses [ 2 ]. Despite this momentum, concerns persist about potential increases in substance use and associated harms, particularly among vulnerable populations such as youth [ 3 ]. While punitive policies have largely been shown to be ineffective in reducing drug-related harm [ 4 ], the broader public health effects of liberalization remain contested. Drug policies, including decriminalization and legalization, offer different methods of moving away from traditional prohibitionist frameworks. Decriminalization generally involves reducing or eliminating penalties for the possession of small amounts of controlled substances, focusing on personal use while keeping production and sale illegal [ 5 ]. On the other hand, legalization creates a regulated market for certain drugs, allowing legal production, sale, and use. Opponents argue that by increasing the availability of these substances, legalization may lead to higher consumption and, consequently, a rise in negative health outcomes. Advocates, however, primarily emphasize the public health benefits of legalization and decriminalization, arguing that these policies can save lives by preventing overdose deaths, and through supervised consumption and regulated supply, expanding harm-reduction and treatment services, and avoiding incarceration-related trauma by shifting from criminalization to care [ 6 ]. They also note that legalization enhances control over drug quality, disrupts illegal markets, and generates tax revenue that can be reinvested into public health initiatives [ 7 , 8 ]. Mainstream narratives, often shaped by non-expert voices such as politicians, media figures, and ideologically driven commentators, frequently reflect polarized views. The absence of conclusive evidence, combined with limited public awareness, has further fueled misperceptions about the policy impacts [ 9 , 10 ]. Drug-related outcomes—whether beneficial or harmful—are highly dependent on contextual factors such as health system capacity, enforcement practices, public attitudes, and the pre-existing status of drug use, treatment infrastructure, and criminal justice involvement within each country or jurisdiction. For example, Portugal’s decriminalization efforts have reduced the burden on criminal justice systems and improved public health outcomes by increasing access to harm reduction services, although concerns about potential increases in drug use persist [ 5 ]. Similarly, the Netherlands’ depenalization of cannabis through “coffee shops” has lowered criminal justice costs but has also been associated with higher cannabis use, driven more by commercialization than decriminalization itself [ 11 ]. In Uruguay, the state-controlled cannabis market aimed to regulate the illegal drug trade, though its impact on overall use is still debated [ 12 ]. In Canada, cannabis legalization has reduced cannabis-related arrests and provided social justice benefits but has also led to increased emergency department visits and hospitalizations, particularly among youth [ 13 , 14 ]. In contrast, countries such as France and Sweden, which maintain prohibitionist drug policies, have seen persistent or increasing rates of drug use and related harms, suggesting that restrictive laws alone may not effectively curb substance use [ 15 , 16 ]. Japan, on the other hand, also enforces a highly punitive drug policy, but reports low levels of drug use [ 17 ]. These contrasts suggests that the effectiveness of drug policy cannot be evaluated in isolation from the broader social, health, and legal context in which it is implemented. Beyond legal and economic effects, liberal drug policies may have profound implications for public mental health. Substance use and mental health share a complex, bidirectional relationship: each can precipitate or aggravate the other [ 18 ]. Importantly, poor mental health, trauma, and adverse childhood experiences (ACEs) are recognized as significant upstream drivers of substance use initiation and escalation [ 19 ]. Individuals exposed to ACEs, untreated psychological conditions, or chronic stress are at substantially higher risk of developing substance use disorders (SUDs) and related harms [ 20 ]. At the same time, problematic drug use can contribute to a range of psychiatric symptoms, including depression, anxiety, and psychosis [ 18 ]. Even occasional or nondependent consumption may provoke mental health conditions [ 21 ], yet these co-occurring complications are particularly prevalent among individuals with SUDs [ 22 ]. The nature and severity of these mental health effects depend on the type of substance, the pattern of use, and individual vulnerability. For example, early and heavy cannabis use in adolescence can disrupt neural maturation and heighten the risk of psychosis, depression, and suicidal ideation [ 23 ]. Problematic use of opioids, whether prescribed or illicit, are linked to depression, anxiety, and suicidality, especially during withdrawal phases [ 24 , 25 ]. Stimulants such as cocaine and amphetamines may trigger acute paranoia and psychotic episodes, with chronic use often resulting in persistent cognitive deficits [ 26 , 27 ]. Harmful use of legally available substances like alcohol can also elevate the risks of depression and suicide [ 28 ]. These overlapping conditions not only complicate diagnosis and treatment but also point to shared biological and psychosocial vulnerabilities. Despite strong associations, it remains uncertain whether escalating substance use worldwide is contributing to mental health conditions in the context of substance use or whether worsening psychological well-being is driving more individuals to self-medicate with substances. In this context, evaluating how drug policy, be it legalization, decriminalization, or harm-reduction strategies, modulates these intertwined outcomes is an essential focus for future research. While the issue is pressing, studies exploring the mental health impact of changing drug policies remain sparse. A recent scoping review by Fortier et al. investigated the mental‑health impacts of cannabis legalization [ 29 ], identifying twenty‑eight original studies and reports from January 1, 2012, to April 30, 2023. This paper has outlined the limited research available and highlighted the largely inconclusive findings, however, its focus was specifically on cannabis policies. The present review expands the scope to include a broader range of drug policies, such as the decriminalization, legalization or commercialization of psychoactive substances other than cannabis, alcohol, tobacco, or prescribed medications, and extends the analysis period back to January 2001. We also incorporate review papers alongside original studies to map the entire body of existing literature. This approach offers a more comprehensive, up‑to‑date overview of current knowledge. By addressing a wider spectrum of policies and substances, this review aims to bridge existing gaps, establish a stronger foundation for future research, and better inform evidence‑based policy decisions in this evolving and critical field.

In total, 5,607 studies were identified through both searches (4,805 in November 2023 and 802 in December 2024). The breakdown of sources is as follows: Medline (n = 1,191), Embase (n = 2,956), PsycINFO (n = 576), CINAHL (n = 273), Web of Science (n = 605), and manual sources (n = 6). After removing duplicates, 3,373 studies remained for title and abstract screening. Of these, 3,215 ineligible studies were excluded, leaving 158 full texts for review. At the full-text review stage, 103 studies were further excluded. This resulted in a total of 55 studies eligible for data extraction and charting. The most frequent reason for exclusion at this stage was the lack of reported mental health measures as an outcome. Fig 1 presents the PRISMA flow diagram outlining the multiple steps of this scoping review. The PRISMA-ScR checklist is provided in the Supporting Information ( S3 Appendix ). Of the 55 included studies, 16 were review papers and 39 were original research papers. No policy reports or working papers were identified that addressed mental health outcomes. The characteristics of the review and original research papers are described in greater detail in the following sections. Nearly all studies originated from North America, with 29 conducted in the United States and 18 in Canada. One study was conducted in the United Kingdom, and seven review papers took a global perspective, although their investigators were predominantly based in North America. Fig 2 illustrates the geographical distribution of the assessed studies. Data on countries’ policies were collected from Eastwood et al. [ 1 ]. The figure is generated using Python (Base layer source: Natural Earth, 1:110m Cultural Vectors (Admin 0 – Countries), available at https://www.naturalearthdata.com/downloads/110m-cultural-vectors/ ). We could not locate any studies examining the impact of policies decriminalizing or legalizing ‘non-cannabis psychoactive drugs’, including substances other than cannabis, alcohol and tobacco, on mental health outcomes (n = 0). Consequently, all included studies were focused on cannabis policies (n = 55). In terms of mental health outcomes, self-harm (including suicidal ideation, suicide deaths, and intentional self-harm) was the most frequently assessed outcome (n = 22), followed by psychosis (n = 21), depression (n = 11), anxiety (n = 9), and psychiatric ED visit/hospitalization (n = 11), as shown in Fig 3 . Tables 1 and 2 present the characteristics of the original studies and review papers and their findings. ML: Medical Legalization, RL: Recreational Legalization, COM: Commercialization, DCR: Decriminalization, ED: Emergency Department, SMI: Serious Mental Illness, PTSD: Post-Traumatic Stress Disorder, N/A: Not available a Overall policy impact on mental health outcomes. A positive impact indicates improvement in mental health symptoms or a reduction in adverse outcomes. A negative impact indicates worsening symptoms or an increase in adverse outcomes. b No significant impact of the policy on mental health outcomes (i.e., null or inconclusive findings). c Mixed results were observed across different subgroups or outcome measures. ML: Medical Legalization, RL: Recreational Legalization, COM: Commercialization, DCR: Decriminalization, ED: Emergency Department, SMI: Serious Mental Illness, PTSD: Post-Traumatic Stress Disorder N/A: Not available a Original studies included in this scoping review that were also reviewed in the included review papers. b Overall policy impact on mental health outcomes. A positive impact indicates improvement in mental health symptoms or a reduction in adverse outcomes. A negative impact indicates worsening symptoms or an increase in adverse outcomes. c Mixed results were observed across different subgroups or outcome measures. d No significant impact of the policy on mental health outcomes (i.e., null or inconclusive findings). Our search covered a 24-year period from January 2001 through the end of 2024. However, as shown in Fig 4 , all included studies were published within the past 10 years, with none identified prior to 2014. This highlights a notable gap in the literature, especially considering that cannabis decriminalization in the United States began as early as the 1970s, followed by the legalization of medical cannabis in California in 1996 [ 92 ] and the introduction of medical cannabis legislation in Canada in 2001 [ 93 ]. Among the 55 reviewed papers, 29 evaluated the effects of recreational legalization, with or without including the impact of opening commercial dispensaries; 7 focused only on medical legalization; and 18 analyzed the impacts of both medical and recreational laws. Thirteen studies considered commercialization and opening of the cannabis dispensaries in their analyses. One review paper only examined the literature on commercialization and access to cannabis retailers [ 79 ]. Notably, none of the original studies or review papers directly examined decriminalization as a distinct drug policy framework. A majority of original studies took place in the United States (n = 24) [ 38 – 40 , 42 , 44 , 47 – 51 , 53 – 55 , 59 – 61 , 63 , 64 , 68 – 71 , 73 , 76 ] and fourteen studies were conducted in Canada [ 41 , 45 , 46 , 52 , 56 – 58 , 62 , 65 – 67 , 72 , 74 , 75 ]. We identified one study from the United Kingdom, where a liberal drug policy is not broadly implemented and only allows restricted cannabis prescriptions for medical purposes [ 43 ]. For the US studies, analyzed data was obtained from the following states: Colorado (n = 6) [ 47 , 48 , 54 , 60 , 71 , 73 ], California (n = 4) [ 40 , 42 , 51 , 60 ], Washington (n = 3) [ 47 , 48 , 60 ], Alaska (n = 2) [ 47 , 60 ], Oregon (n = 2) [ 47 , 60 ], Massachusetts (n = 1) [ 60 ] and Nevada (n = 1) [ 60 ]. The remaining studies in the US incorporated data from nearly all states. In Canada, Ontario was the province predominantly studied (n = 10) [ 41 , 45 , 46 , 56 , 57 , 62 , 65 – 67 , 74 ], followed by Alberta (n = 4) [ 45 , 46 , 66 , 75 ], Quebec (n = 3) [ 58 , 66 , 72 ], and British Columbia (n = 2) [ 52 , 66 ]. Most studies employed a quasi-experimental(n = 14) [ 38 , 39 , 42 , 45 – 48 , 53 , 57 , 60 , 63 , 70 , 75 , 76 ] or a cohort design (n = 12) [ 40 , 41 , 49 – 52 , 58 , 61 , 65 , 71 – 73 ]. Nine were cross-sectional studies [ 44 , 54 – 56 , 59 , 62 , 64 , 66 , 67 ], including those with repeated cross-sectional design. One study used an ecological cohort approach [ 69 ], and another referred to their methodology as an event study [ 68 ]. There were two qualitative studies in which data were collected through semi-structured interviews: one with healthcare providers in a psychiatric hospital setting [ 74 ], and the other with individuals who had a prescription for medical cannabis and their carers [ 43 ]. Most papers utilized data from national or regional administrative databases, such as vital statistics, emergency department presentations, hospital admission/discharge records, or insurance claims. A few studies relied on previously collected surveys or existing longitudinal cohort data [ 40 , 49 , 63 , 70 , 76 ]. The majority of studies (n = 24) compared pre- and post-legalization trends, primarily using quasi-experimental (e.g., interrupted time series) and longitudinal designs. Fifteen studies compared states or regions that adopted the policy with those that did not, either alongside a pre-post approach or as a standalone comparison. Two studies constructed synthetic control groups to represent counterfactual scenarios in which the policy had never been implemented [ 42 , 48 ]. Additionally, two studies used conditions or emergency department visits unrelated to cannabis as comparison outcomes [ 45 , 54 ]. Suicide and self-harm (including suicide deaths [ 38 , 42 , 47 , 48 , 53 , 55 , 59 , 60 , 70 , 71 ], intentional self-harm [ 46 , 48 , 61 , 67 ], and suicidal ideation [ 69 ]), along with psychosis [ 41 , 45 , 50 , 57 , 58 , 64 – 66 , 72 , 73 , 75 , 76 ] and depression [ 40 , 41 , 51 , 52 , 56 , 63 , 69 , 75 ] were the most frequently evaluated non-SUD mental health outcomes. These were followed by anxiety [ 40 , 41 , 52 , 56 , 62 , 75 ], overall mental health [ 39 , 40 , 43 , 70 ], non-specified psychiatric admissions/visits [ 54 , 57 , 68 ], serious mental illness (SMI) [ 49 , 69 ], personality disorder [ 73 , 75 ] and somatization [ 40 ], as shown in Fig 3 . One study analyzed prescription fill rates for psychiatric medications as an indicator of trends in psychiatric symptomatology [ 44 ]. It is worthwhile noting that the measures used to assess outcomes varied across studies. Some used more reliable diagnostic methods, such as ICD codes and DSM-based interviews, while others relied on self-report instruments [ 39 , 40 , 52 ]. One study used prescription fill data as a proxy indicator, which may be influenced by factors unrelated to policy changes and should therefore be interpreted with caution when evaluating mental health impacts. Table 2 presents a summary of the included review papers. Sixteen review papers were included, comprising nine narrative reviews [ 78 , 80 , 81 , 84 – 87 , 89 , 91 ], six systematic reviews [ 77 , 79 , 82 , 83 , 88 , 90 ] and one scoping review [ 29 ]. Half of the reviews looked over papers published globally [ 29 , 77 , 79 , 81 , 82 , 88 , 90 ], one third specifically focused on the United States [ 80 , 84 , 85 , 87 , 91 ], and the remaining reviewed the policy impact in Canada [ 78 , 83 , 86 , 89 ]. Notably, 20 of the original studies included in our scoping review were also examined in one or more of the included review papers. This overlap reflects the limited pool of available evidence in this area and has been specified in Table 2 to acknowledge repetition and ensure transparency in the interpretation of findings. Most of the review papers examined a broader range of outcomes beyond mental health conditions. Drug use patterns, overdose rates, SUD prevalence, cannabis-related traffic incidents and physical health issues were among the outcomes analyzed, alongside non-SUD mental health outcomes. Among these, psychosis was the most commonly explored non-SUD mental health outcome [ 29 , 79 , 82 , 83 , 86 – 90 ], followed by, suicide or self-harm [ 29 , 79 , 84 – 87 , 90 , 91 ], psychiatric ED visit or hospitalization [ 77 , 78 , 80 – 83 , 85 , 89 ], depression [ 29 , 86 , 87 ], anxiety [ 29 , 82 , 86 ], and SMI [ 84 , 94 ] (See Fig 3 ). Given the substantial variation in policy frameworks, study designs, and reported outcomes, this section does not attempt to quantify the strength of evidence or draw conclusions tied to specific legal contexts. Instead, we offer a high-level, descriptive summary of the overall direction of findings related to mental health outcomes. The effects of drug policy reforms can differ significantly across jurisdictions, shaped by local enforcement practices, health system capacity, and supporting regulations. Therefore, the findings should be interpreted with caution and understood as indicative of general trends rather than direct comparisons between studies conducted in diverse settings. Fig 5 illustrates the distribution of the “overall impact” of implemented drug policies on mental health outcomes in both original studies and review articles. A notable majority of original studies found no significant evidence linking policy enactment to changes in mental health outcomes (n = 14) [ 41 , 45 – 47 , 50 , 53 , 56 , 58 , 60 , 61 , 69 , 71 , 76 ], a conclusion also supported by four review papers [ 86 , 87 , 89 ]. One-quarter of original studies [ 49 , 51 , 54 , 63 , 64 , 66 ] and over one-third of review papers [ 80 , 81 , 85 , 88 , 91 , 94 ] reported increases in the prevalence, incidence, or trend of adverse mental health outcomes, such as psychosis and depression, suggesting a potentially negative impact. Conversely, six original studies (15%) reported a positive impact, noting reductions in adverse mental health outcomes [ 38 , 39 , 42 , 43 , 68 , 75 ]; however, no review papers corroborated this finding. Lastly, over 20 percent of original studies [ 48 , 52 , 65 , 70 , 72 , 73 ] and six of sixteen reviews [ 78 , 82 , 83 , 90 ] presented mixed findings, pointing to inconclusive nature of the existing evidence. Studies from both the United States and Canada predominantly found that cannabis policies had no significant impact on mental health outcomes, with 34.5% of U.S. studies and 38.9% of Canadian studies reporting no overall effect. The second most common finding in U.S. studies was a negative overall impact (31%), whereas in Canadian studies the second most frequent was mixed finding (33.3%). A comparison of overall policy impacts between studies from the U.S. and Canada is presented in Fig 6 .

This review underscores a significant evidence gap in understanding the mental health impacts of liberal drug policies—particularly those targeting substances beyond cannabis. While research has increasingly addressed outcomes such as substance use disorders and overdose, non-SUD mental health conditions remain underexplored and inconsistently reported. The limited and context-dependent nature of existing findings highlights the need for more rigorous, long-term studies that account for policy design, implementation setting, and population differences. Addressing these gaps is essential to support evidence-informed, equitable drug policy reforms that consider both intended and unintended mental health outcomes.

We conducted a scoping review to provide an overview of the global literature on the mental health outcomes of drug decriminalization and legalization policies. We adhered to the Joanna Briggs Institute (JBI) guidelines for scoping reviews (Chapter 10) [ 30 ] and used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR) checklist to guide our reporting [ 31 ]. A review protocol was developed that outlined our search strategy, eligibility criteria, and the key study features to be extracted. We executed the initial search on November 30 th , 2023 and then ran a second search on December 21 st , 2024 to update our results and capture the most up-to-date evidence. We searched through MEDLINE (Ovid), Embase (Ovid), CINAHL, PsycInfo (EBSCO), Science Citation Index and Social Sciences Citation Index (Web of Science). Government and national institution websites such as EMCDDA and UNODC (United Nations Office on Drugs and Crime) were explored manually for working papers and reports. A search strategy was developed and tested with the help of a professional librarian. We used search terms that include any relevant clusters of terms related to recreational or illicit drug categories, such as “cannabis”, “opioid”, “stimulant” or “psychedelic”, in combination with keywords pertaining to policy regulations, including “decriminalization”, “legalization”, “liberalization” or “commercialization”. We also included MeSH terms for some specific concurrent disorders in the context of substance use. A copy of the full draft of the Ovid EMBASE search strategy can be found in the Supporting information ( S1 Appendix ). Eligible studies included 1) original research employing qualitative, quantitative, or mixed-methods approaches, 2) review papers such as systematic reviews, meta-analyses, narrative reviews, and scoping reviews, and 3) policy reports and working or white papers that had not undergone formal peer review were also considered. However, we did not include commentaries, letters to editors, books, dissertations, theses, protocols, formative research, and conference papers. Studies published before January 1 st , 2001 and those not written in English were also excluded. The review included studies involving adults aged 18 years and older, as drug use or possession typically remains illegal for youth and adolescents, even under liberal policy frameworks. Studies focusing on adult sub-populations, such as people who use drugs, individuals with mental health conditions, and those with chronic conditions, were considered in this review. Our initial aim was to assess the effects of liberal policies targeting non-cannabis psychoactive substances (e.g., opioids, stimulants) at national or jurisdictional level. Here, the “policy framework” encompasses three main regulatory approaches: Decriminalization: removal of criminal penalties for possession or personal use, typically replacing them with administrative fines or diversion to treatment [ 32 ]. Legalization: establishment of a lawful pathway for production, distribution, and consumption under government oversight [ 33 ]. Commercialization: introduction of a regulated market in which private entities can produce, market, and sell substances, often alongside taxation and advertising controls [ 34 ]. Because our targeted search for studies on substances other than cannabis yielded no eligible papers, we broadened our scope to include evaluations of cannabis policies. Consequently, any study examining the impact of decriminalization, legalization, commercialization, or similar regulatory changes was considered eligible, regardless of substance class. The primary outcome of interest was non-SUD mental health conditions such as depression or anxiety disorders among populations with exposure to psychoactive substances, regardless of whether the individual meets criteria for a substance use disorder. These conditions may be pre-existing, co-occurring, precipitated, or influenced by substance use. This includes both acute mental health conditions triggered by substance use (e.g., psychotic episodes during intoxication) and longer-lasting or recurrent disorders that may persist beyond the immediate effects of the substance [ 35 , 36 ]. We included studies reporting population-level measures derived from national statistics, administrative health records, or large cohort studies. Both large, population-based datasets and smaller clinical or community samples were accepted, provided they met our diagnostic or measurement criteria. Studies were included if they employed clinical diagnoses (based on DSM or ICD criteria), validated diagnostic interviews, prescription-fill records for major psychiatric medications, or standardized self-report scales to ascertain mental health outcomes. We excluded any studies that: Reported solely on patterns of drug consumption (e.g., frequency, quantity) or on problematic use indicators (e.g., dependence, abuse) or all-cause drug-related emergency department visits (e.g., overdose rates, drug-related deaths, intoxication, or withdrawal) without measuring psychiatric symptoms; Focused exclusively on substance use disorders (SUDs). While SUDs are recognized substance-related mental health conditions, they were excluded to allow a focused examination of non-SUD mental health outcomes in the context of substance use, such as depression, anxiety, or psychosis; Investigated physical health outcomes, civic and social impacts, motor vehicle accidents, drug-related arrests, law enforcement activity, or any aspect of drug trafficking, manufacturing, or cultivation; Addressed mental health conditions associated exclusively with prescribed medications, alcohol, tobacco, or caffeine. Prescribed cannabis and psychedelics were treated as distinct exposures, as their medical legalization falls within the scope of this review. Here “prescribed medications” refers specifically to psychoactive substances commonly used without policy-specific regulation (e.g., benzodiazepines). The search results were imported into the Covidence platform for screening, full-text review, and data extraction [ 37 ]. Titles and abstracts were screened independently by two reviewers, with conflicts resolved by a third reviewer. All potentially eligible studies underwent a full-text review by two reviewers, and any disagreements during this stage were resolved through discussion. Data extraction and charting focused on key study features and findings, including: (1) study design, (2) country and region of the study, (3) data sources, (4) type of drug-related policy (e.g., decriminalization, legalization, commercialization), (5) relevant outcomes, and (6) reported findings. A copy of the template data charting tool is available in the Supporting Information ( S2 Appendix ). A descriptive summary was synthesized to provide an overview of the characteristics, distribution, and findings of the studies. To avoid overstating results, we charted the characteristics and findings of original studies—including quasi-experimental, cross-sectional and cohort studies—separately from review articles.

(DOCX) (DOCX) (PDF)