produced and distributed, the individuals and organizations that are allowed to participate in the market, how each is allowed to operate, and how legalization is implemented. For example, product safety and quality are influenced by decisions related to the cultivation of cannabis, the cannabinoid extraction process, and any other chemicals used to produce the final product. Policies around sales and marketing can educate consumers about products; for example, labeling can reduce accidental consumption and provide consumers with information about dosing and prevention of harmful use. Policies on advertising and promotion shape who can see them, where they are allowed, and what the advertisements must contain. Another consideration relates to licenses, the number and types of outlets where cannabis can be sold and how, and the circumstances in which the

product can be sold—for example, whether it can be sold with food, with or without other intoxicants, and whether consumption can occur on premises. Another consideration is the geographic location and density of cannabis retail outlets.

To evaluate cannabis policy in the United States, the committee considered regulatory regimes worldwide. It then evaluated cannabis policies within the United States, describing observed variations in cannabis policies related to public health. The committee chose to evaluate the variation in state policies because systematically collecting local policy data within and across states was infeasible, and information on compliance with state regulations is scarce. Where available, the committee also considered the limited evidence on the implications of local public health regulations. The committee then considered alternative regulation models, such as those for tobacco and alcohol in the United States and for cannabis in Canada and Uruguay.

CANNABIS REGULATORY REGIMES ACROSS THE WORLD

As of 2021, 64 countries had provisions in national law or had developed guidelines allowing medical use of cannabinoid products (UNODC, 2022). Several countries have adopted or tolerated alternatives to the legalization of the entire supply chain, which offer opportunities to grow or sell cannabis for adult use but do not allow commercial cultivation and production of cannabis (Kilmer and Pacula, 2017; UNODC, 2022). In Spain, Belgium, and 11 other countries in Europe, for example, cannabis clubs and nonprofit collectives allow adults to cultivate, produce, and distribute cannabis collaboratively among themselves (Pardal, 2022). In the Netherlands, the cultivation, production, sale, and possession of cannabis are illegal, yet cannabis sales for personal use at coffee shops are tolerated (Government of the Netherlands, n.d.a). Thus, coffee shops must acquire the product from illegal sources. Some municipalities license coffee shops for selling cannabis. Additionally, court decisions in some countries (e.g., Mexico) have created ambiguity around the legality of home growing (Pardal, 2022). Thus, there has been considerable variability internationally in alternatives to prohibition and models of supply.

Pilot experiments in cannabis regulation offer potential insights into the public health effects of different regulatory models. In particular, the Netherlands and Switzerland have passed laws authorizing studies on cannabis regulation. The Government of the Netherlands (2019) is currently running a 4-year study whereby 10 growers are legally allowed to produce cannabis to sell to coffee shops in 10 municipalities; the researchers will then evaluate the impact of this experiment on public health and crime (Government of the Netherlands n.d.b). Furthermore, in Switzerland, through the Ordinance

on Pilot Trials under the Federal Narcotics Act, cantons, municipalities, and organizations (including universities) will be able to conduct trials to investigate the impact of different cannabis distribution channels (pharmacy distribution, cannabis social clubs, and nonprofit retail outlets), and to test cannabis products of different tetrahydrocannabinol (THC) concentration levels (FOPH, 2023).

In contrast, as of April of 2024, Canada, Luxembourg, Malta, Uruguay, South Africa, one state in Australia, and 24 states in the United States have legalized cannabis supply to and possession for adults. However, their approaches differ meaningfully (Figure 2-3). Luxembourg, the Australian Capital Territory, and South Africa¹ have legalized home cultivation and possession for personal use. Malta has gone a step further: it allows home cultivation as well as the operation of nonprofit cannabis clubs, which are allowed to grow and supply cannabis to their members (Pardal, 2022). Uruguay uses a hybrid approach, with legalization for nonprofits and a highly government-regulated form of for-profit legalization. Home cultivation, cannabis clubs, and retail sales in pharmacies are allowed; however, the government controls large-scale cannabis cultivation, and the product and retail operations

___________________

¹ On May 28, 2024, South African lawmakers legalized cannabis for personal use. The sale and creation of a legal market for cannabis are prohibited (Sabaghi, 2024).

are highly regulated (Cerdá and Kilmer, 2017). Canada also takes a hybrid approach to the government’s involvement in the supply of cannabis in the adult-use market, allowing models to vary by province, with some provinces restricting retail sales to government-run stores and others allowing private retail sales or a combination of both. Finally, the United States represents the fully commercial option on the supply regulation spectrum, with the private sector conducting cultivation, production, and retail sales. In addition, most states also allow some form of home production and sharing.

CANNABIS REGULATION IN THE UNITED STATES

Cannabis regulation in the United States is complicated by the lack of federal involvement in the drug’s legalization. In most areas of public health regulation, U.S. states have historically had a federal partner that has assisted in the regulation of broad market factors such as product quality assurance (through the Food and Drug Administration [FDA]), testing (through the FDA and the U.S. Department of Agriculture [USDA]), prevention (through the Centers for Disease Control and Prevention [CDC]), and industry structure (through the Federal Trade Commission). States have the authority to regulate products under the 10th Amendment of the U.S. Constitution, which establishes that the federal government’s powers are restricted to those delegated by the Constitution and that the states have all remaining powers. While state authorities have limits (e.g., through preemption, through federal drug scheduling, in use of federal funds), they nonetheless have significant experience and history in regulating legal commodities and behaviors that impact public health, from alcohol and tobacco to sugary drinks, safe driving practices, and pesticides used in agriculture.

Local governments also have several mechanisms available for regulating cannabis within their jurisdictions, such as zoning restrictions determining where retail outlets can be located, regulations regarding the types of products that can be sold, rules on additives or ingredients that can be contained in products sold (e.g., flavoring bans), restrictions on advertising, and taxation (Caulkins and Kilborn, 2019; Dilley et al., 2017; Payán et al., 2021). At times, local authorities have implemented stricter regulations than those adopted by the state. Thus, defining “cannabis public health regulations” within any state is complicated because state policies alone do not necessarily define the local regulatory environments.

Tribal sovereignty presents a unique challenge in cannabis policy for the United States. Within states that have legalized cannabis, tribes retain the authority to establish their own decisions and rules related to legalization, potentially creating a situation in which federal prohibition remains in effect on tribal lands after the state has legalized cannabis. This challenge stems from the inherent sovereignty of federally recognized tribes, which

generally exempts them from state laws within reservation boundaries. Similarly, tribes possess the legal authority to license, regulate, and even legalize cannabis activities on their reservations, even if recreational marijuana sales are not legal in the surrounding state. Importantly, some state statutes explicitly exclude Native American tribes from participating in cannabis licensing processes, creating a potential conflict with tribal sovereignty (Mooney, 2022).

Since the first states legalized cannabis for adult use in 2012, there have been efforts to describe the regulatory frameworks that have either already been adopted by individual states or might be considered by states adopting adult-use policies in the future (Barry and Glantz, 2016; Blanchette et al., 2022a; Ghosh et al., 2016; Pacula et al., 2014a). Because 14 of the first 16 adult-use laws were passed by states through ballot measures (Schauer, 2021), legislators and regulatory agencies were assigned responsibilities based on broad notions of how the populace wanted the markets to operate; they were not carefully designed market systems. Moreover, state agencies were given relatively short periods within which to establish these markets. Initial regulations, therefore, focused on setting up licenses and legal supply chains and addressing voters’ objectives in initiatives to eliminate the illicit market, including the involvement of gangs and other actors engaged in the trafficking of illegal drugs. The early regulations also included a few broad public health objectives, such as preventing the distribution of cannabis to underage people, making a safe product available, and preventing impaired driving. As more time passed, regulators in these early-adopting states began to grapple with some of the more challenging public health aspects of cannabis policy—product regulation and testing, marketing restrictions, and warnings. The delay in addressing some of these public health issues has made it challenging for researchers to understand which state policies are the most effective at promoting public health.

Impact of the 2018 Agriculture Improvement Act

The 2018 Agriculture Improvement Act (2018 Farm Bill) has had a profound impact on the cannabis landscape in the United States and confuses any policy analysis at this time. As discussed in Chapter 1, this legislation redefined “hemp,” allowing its legal sale without its being subject to the Controlled Substances Act (Gottron et al., 2019). According to the 2018 Farm Bill, “hemp” is now defined as “the plant Cannabis sativa L. and any part of that plant, including the seeds thereof and all derivatives, extracts, cannabinoids, isomers, acids, salts, and salts of isomers, whether growing or not, with a [delta-9-THC] concentration of not more than 0.3 percent on a dry weight basis” (PL-115-334, § 297A). This definition has led to legal ambiguities, facilitating the production and sale of cannabinoids derived

from hemp and leading to a largely unregulated, multibillion-dollar industry (Skodzinski, 2024) that competes with the regulated state-legal cannabis industry (Johnson, 2023; Johnson and Willner, 2023). The inclusion of the terms “all derivatives, extracts, cannabinoids, isomers, acids, salts, and salts of isomers” has also led to the sale of intoxicating cannabis products, especially in states that have not chosen to legalize cannabis (Demko, 2024).

One cannabinoid that has garnered tremendous consumer interest is cannabidiol (CBD). A purified form of CBD, Epidiolex^®, is approved for oral administration by the FDA for the treatment of infantile refractory epileptic syndromes. Consumers have also demonstrated interest in CBD’s other potential benefits, such as its antianxiety and anti-inflammatory properties. CBD is added to dietary supplements, foods, drinks, and health and beauty products. In 2018 and 2019, CBD sales proliferated despite regulatory uncertainty. The FDA has said that it is “unlawful” under the Federal Food, Drug, and Cosmetic Act (21 USC §§301 et seq.) to market CBD products as or in dietary supplements and formed a working group to determine a legal pathway for CBD (Johnson, 2019). As of January of 2023, the FDA is working with Congress to develop a new regulatory pathway for CBD following years of FDA review related to CBD product regulation (Johnson, 2023). In July 2023, several members of Congress requested information from stakeholders on how to “provide a legal pathway” for marketing CBD products (Johnson, 2023).

The primary concern for public health, though, is the unregulated market of cannabis products that contain delta-9-THC or similar compounds. Among those concerns are an abuse of the “dry-weight” delta-9-THC definition of hemp (Williams, 2021); the sale of cannabis flower containing notable concentrations of naturally occurring tetrahydrocannabinolic acid (THCA), a precursor to delta-9-THC; and the sale of cannabis products that contain synthetic derivatives of CBD, such as delta-8-THC (CANNRA, 2023).

Because federal guidance has been limited, the cannabis industry is shrouded in uncertainty and conflicting interpretations. For example, some cannabis businesses have tried to leverage the dry-weight concentration of delta-9-THC (0.3 percent) specified in the 2018 Farm Bill by applying it to products created with hemp (Williams, 2021). Cannabis edibles in the form of chocolates or gummy candies allow producers to leverage the “dry weight” distinction because chocolates and gummy candy contain little water. A 5-g gummy candy can contain roughly 15 mg of delta-9-THC (5-g candy × 0.3% = 0.015 g = 15 mg delta-9-THC) and remain within the dry-weight definition of hemp. For comparison, a standard cannabis edible in Colorado contains 10 mg delta-9-THC (Johnson et al., 2023). Legal experts advising the cannabis industry have stated that “dry weight” does not apply to final products, although the practice appears to be common (Williams, 2021).

2018). Many aspects of a chosen cannabis supply framework could influence public health outcomes. However, relatively few cannabis policies have been formally and systematically compared across states until recently (see Chapter 4).

The current state of cannabis regulation in the United States is complex because of the lack of federal involvement. Analysis reveals variations in regulations (e.g., advertising restrictions, product types, THC concentration limits), as well as enforcement (e.g., age verification, product safety). The patchwork of federal, state, and local regulations makes it challenging to study the effectiveness of different regulatory approaches within the states. While initial regulations focused on establishing legal markets, the focus has shifted toward addressing public health concerns such as ensuring product safety, limiting exposure to and use by youth, and preventing impaired driving.

Lessons from tobacco and alcohol control can provide frameworks for thinking about cannabis policy, and each has been well studied. Studies have compared different regulatory models of the tobacco industry’s influence on youth access to cigarettes, and on the early initiation and popularity of smoking (CDC, 2012; Chaloupka, 1999; DeCicca et al., 2022; Higgins et al., 2019). Similar literature exists for alcohol and alcohol-related harms (Cook, 2007; Nelson et al., 2013, 2015; Office of the Surgeon General et al., 2007; Toomey and Wagenaar, 1999). Evidence about cannabis regulations is limited because the variation in state regulatory approaches to adult use has been well documented only recently (APIS, 2023a; Blanchette et al., 2022b; Schauer, 2021). While analogies can and have been made to tobacco and alcohol (Barry and Glantz, 2018; Hall, 2017; Orenstein and Glantz, 2018; Pacula et al., 2014b; Steinberg et al., 2020), there are important differences among cannabis products. The cannabis plant is more than a cannabinoid, with plant hybrids having unique chemical profiles (Procaccia et al., 2022). The cannabinoid mixtures within different plant hybrids can have different health effects, and there is therapeutic value in consuming particular cannabinoids when trying to manage some medical symptoms and conditions (Lynch and Campbell, 2011; Wang et al., 2021). Tobacco and alcohol do not have health benefits and thus are vastly different. Thus, it is difficult to know in advance the extent to which specific regulatory strategies targeting alcohol or nicotine and tobacco would be similarly effective for cannabis.

In 2019, researchers from Boston University and RAND organized a group of public health experts and asked them to nominate and rank state-level cannabis regulatory policies they believed (based on their knowledge of the scientific literature studying cannabis, alcohol, tobacco, and opioids) were likely to be the most effective at achieving three public health objectives associated cannabis legalization: (1) minimizing excess use of cannabis by the general population, (2) limiting youth use, and (3) reducing

cannabis-impaired driving (Blanchette et al., 2022a). Through a modified Delphi process, the group of public health experts identified state regulations likely to achieve the public health aims (see Table 2-1): the adoption of a state monopoly, restrictions on physical retail availability, tax strategies, retail price and operating restrictions, and product design restrictions and requirements. Policies on youth access and advertising restrictions were also deemed likely to be highly effective in reducing youth access. For

have imposed more restrictions on the number of outlets per capita. Seven states (Arizona, Connecticut, Illinois, Nevada, Rhode Island, Virginia, and Washington) impose a cap on the total number of retail outlets allowed.

States impose laws or rules regarding retail storefronts and place limitations on where these businesses can operate. These laws are often created to strike a balance among accommodating the burgeoning cannabis industry; preventing oversaturation of the market in certain areas; and addressing concerns about public safety, youth exposure, and “community aesthetics.” Laws are common that require retail cannabis businesses to remain a specific distance from public areas and child-centered institutions—typically 500–1000 feet from schools, childcare centers, and community centers. Certain states also have regulations that prohibit retail stores from locating within a specified distance of religious institutions or places of worship, such as churches and synagogues. States also generally allow localities to increase setback requirements, thus placing further limitations on store access.

Admin. Code § 17:30-3.4). Local jurisdictions can also benefit from excise and wholesale cannabis taxes, which can be used for such purposes as education, public safety, and criminal justice reform. However, most states do not mandate such allocations at the state level, allowing municipalities to decide where to disperse the funds (Schauer, 2021; Tax Policy Center, n.d.).

Retail Price Restrictions

States can modulate cannabis prices by taxing cannabis products (discussed above) and by setting minimum pricing standards that establish a floor price below which a cannabis product cannot be sold. Alcohol policy provides important insights about the potential impact of minimum THC unit pricing policies (Humphreys, 2017). The introduction of a minimum alcohol unit price in Scotland in 2018 was associated with reduced alcohol purchases, particularly among the top fifth of households that purchased the greatest amount of alcohol (O’Donnell et al., 2019). Increases in alcohol prices in England were also associated with reduced alcohol use and reduced alcohol-related emergency room visits, injuries, and deaths (Purshouse et al., 2010). A recent World Health Organization (2022) report summarizes the empirical and simulation evidence evaluating the impacts of minimum unit pricing for alcohol in various high-income countries, including provinces of Canada and Australia. WHO concluded that these policies, when set at a price that is passed on to the consumer, do lead to reductions in alcohol consumption, alcohol-related traffic collisions, sexually transmitted diseases, and declines in violence and crime.

According to data from the Alcohol Policy Information System, as of January 1, 2023, 12 of the 21 states where adult use of cannabis was legal had imposed pricing controls. Pricing controls are regulations on pricing that limit the ability of retail stores within those jurisdictions to offer cannabis at a discounted price or at a loss to attract customers (APIS, 2023b). States also have other regulatory levers available to keep stores from offering cannabis at a discounted price, such as bans on happy hours or giveaways.

Retail Operating Restrictions

A range of restrictions can be placed on retail outlets that will influence their operations. In addition to retail price restrictions, these include rules regarding hours or days of operation, minimum legal purchase ages, maximum sales limits, rules on home delivery, mandatory employee training, cash-only purchases, and more. The retail operating restrictions imposed are influenced by where cannabis is sold. For example, the ability to use loss leaders is much reduced when cannabis is sold mostly in cannabis-only stores, as opposed to grocery or convenience stores. As many of these rules

are applied at the local rather than state level, information on the degree to which such rules are applied and their effectiveness at addressing public health concerns is limited.

State-level policy regarding customer age restrictions is consistent across all jurisdictions. In no state is a person under age 21 allowed to purchase cannabis, with certain states, such as Colorado, offering exceptions for those over age 18 with a medical cannabis card (Schauer, 2021). Furthermore, people under age 21 may not be employed by a retail cannabis store, and many states, such as Washington, specifically mandate that all employees be trained on these rules to ensure that they are implemented regularly (Washington State Legislature, Initiative 502). States also universally prohibit employees from openly consuming cannabis products on the premises of the retail outlet.

Hours of operation are often not imposed at the state level; numerous states leave this mandate to municipalities and local governments. There are exceptions, however. A few states set statewide hours during which it is illegal for a retail cannabis store to have its doors open. In New York state, for example, cannabis retail stores cannot be open between 2 a.m. and 8 a.m. (New York Cannabis Law 9 § 116.7). Other states require retail cannabis businesses to be open for a minimum period; otherwise, they risk forfeiture of their retail license. In Washington state, if a cannabis retailer is not open at least 3 days a week for at least 5 hours a day between the hours of 8 a.m. and 12 a.m., its license will be taken under the pretense of the business not being “fully operational” (Washington State Legislature Initiative 502).

As of January 1, 2023, 6 states with legal cannabis prohibit home delivery statewide, while another 10 impose restrictions or limits on home delivery services (APIS, 2023b). In the states that allow it, home delivery represents a growing percentage of online sales for large online stores and local brick-and-mortar outlets. Delivery services may promote more at-home than in-community use and target a higher-income, more tech-savvy consumer. However, underage cannabis purchases may increase with home delivery because age verification is not done in the store. The public health impacts of home delivery have yet to be well studied, primarily because of the lack of data on home delivery transactions within a geographic area (Matthay et al., 2023).

Limits (caps) on the amount of cannabis a retail operator can sell to a consumer in a single transaction are standard practice across the states. The limits vary from 1 oz to 2 oz of dried flower and 3.5 g to 15 g of concentrate (Pacula et al., 2021). Research has shown that imposing weight-based limits rather than limits on total THC purchased has important implications. Assuming that individuals purchased average-concentration products in the marketplace in 2019, Pacula and colleagues (2021) show that consumers in all states with legal cannabis could purchase more than 500 10-mg doses of

THC in a single transaction within state-specific sales limits (Pacula et al., 2021). In six states, the amount that could be purchased, assuming average concentration, was greater than 1,000 10-mg doses of THC and in two states was greater than 1,500 10-mg doses (Pacula et al., 2021).

As with alcohol, some states require retail employees to receive server training as part of regular operating requirements and licensure. The training can vary in content and orientation. Some training focuses on administrative rules and penalties related to the law and what procedures retailers must follow (e.g., in Oregon, no underage sales or exportation out of state). Other programs (e.g., in New Mexico) provide more content that brings awareness of the risks to those who use cannabis in the event of excess consumption or use with other substances. Understanding of the extent to which such training programs and their specific content influence server behavior is just beginning to emerge and represents an important area for further work (Buller et al., 2019, 2021).

All states that have legalized cannabis thus far have set up retail license systems so that only licensed cannabis stores can sell cannabis products. Restaurants, convenience stores, and grocery stores are not currently allowed to sell cannabis products, thereby restricting its general availability. Since passage of the 2018 Farm Bill, however, cannabis products are found everywhere unless the states have instituted additional policies to restrict their sale and are being sold even in states that have not legalized cannabis, confusing many consumers.

Product Design

Unlike provinces in Canada or Uruguay (see below), most adult-use states impose few restrictions on the types of cannabis products that can be sold and purchased. States allow a wide array of smokable, vaporable, edible, infused, and concentrated products to be sold in legal adult-use cannabis retail stores (Schauer, 2021). Requirements regarding shelf-life stability and perishability, meant to minimize food safety risks, are mandated in at least three states: California, Michigan, and Washington. As of January 2021, all states have serving-size limits on the amount of THC permitted in edibles and other consumable cannabis products that can be contained within a single package. These limits differ considerably from limits imposed on the total amount sold in a single transaction. Four states (Alaska, Oregon, Massachusetts, and Vermont) have a limit of 5 mg of THC per serving and up to 50 mg per package. Most other states have limits of 10 mg of THC per serving and up to 100 mg per package. Nonedible THC-infused cannabis products are not regulated in most states, leaving many highly concentrated THC products available for purchase. Vermont alone has placed a cap on the THC concentration in products other than edibles,

limiting cannabis flower to no more than 30 percent THC and cannabis oils to 60 percent THC, and prohibiting all oils and concentrates other than cartridges for vape pens (Schauer, 2021).

As a result of the recent uptick in health issues associated with cannabis products, many states have instituted policies that limit certain ingredients in retail cannabis products (Schauer, 2021; see also Chapter 3). These include excipients (media for delivering a drug), diluents, terpene flavoring blends, and other compounds added to vape cartridges. Many states have banned or tested for vitamin E acetate, often found in cannabis vape cartridges, since it causes e-cigarette, or vaping, product use–associated lung injury, commonly referred to as EVALI. Certain states, such as Colorado and Oregon, have placed restrictions on specific ingredients proven to be unsafe for aerosolization, such as polyethylene glycol; squalane; propylene glycol; and triglycerides, including medium-chain triglycerides. Only a few states have placed limitations on flavors allowed in cannabis products, such as nonnatural artificial flavoring, or prohibiting their use altogether.

Policies Limiting Youth Access and Exposure to Promotions

A central premise of state legalization was that adopting these policies would make it easier to keep the products away from youth. All states impose a minimum legal purchase age of 21. However, states vary in the extent to which these rules are enforced through unannounced compliance checks of retailers. While most states conduct random retailer inspections, they do so under the auspices of checking as to whether the retail premises are ready to open or (if they are newly opened) generally following state rules. Random retailer inspections involve an employee of the supervising agency visiting without prior notice and serving as a source of information and assistance for retailers, answering retailers’ questions, offering training, discussing issues, and ensuring proper signage or use of the seed-to-sale system in the store. Compliance checks, on the other hand, tend to be more punitive. These typically involve a minor who appears to be 21 attempting to purchase products illegally, with law enforcement witnessing a sale to a minor and penalizing the retailer (through fines) for violating the law. All states with adult-use laws have established retail compliance inspection programs, but most have not (as yet) established a mechanism for conducting compliance checks. Early-adopting states (e.g., Washington, Colorado, Oregon) have done both. The specifics of these programs differ across states in terms of frequency and severity of penalties imposed and under what conditions, and the process for determining which outlets to check.

In the case of tobacco, early studies showed no changes in adolescent tobacco use following the introduction of the federal minimum legal purchase age (MLPA) until Congress enacted the Synar Amendment in 1992,

requiring states to enforce MLPA laws by conducting random inspections of tobacco retailers. Early evidence showed that states adopting comprehensive and aggressive tobacco retailer inspection programs experienced reductions in adolescent smoking relative to states that did not (Chaloupka and Pacula, 1998; DiFranza and Dussault, 2005; Sloan, 2000; Stead and Lancaster, 2000). In the case of alcohol, similar studies have examined the impact of inspections of retailers’ checks of adolescent IDs on underage drinking. This research has shown that these inspections, too, have been effective at reducing access to alcohol, heavy drinking, and alcohol-related traffic fatalities among adolescents (George et al., 2021; Grube et al., 2018; Holmila et al., 2010; Schelleman-Offermans et al., 2012; Schweitzer et al., 2017; Scribner and Cohen, 2001).

While advertising cannabis is legal in most states with legalized adult use, states vary as to the amount, type, and location of legal advertisements, which can influence how frequently youth encounter them. Delaware and Montana are the only two adult-use states that prohibit advertising cannabis and cannabis products entirely (Allard et al., 2023). All states but New Jersey and Virginia restrict targeting or appealing to those under age 21 in advertising. The same states, excluding Arizona, also prohibit advertising false or misleading claims (Allard et al., 2023). Many states have adopted limitations on advertising based on the age of the viewership audience (most states mandate that more than 71.6 percent of the intended audience be age 21 and over in cannabis-related advertising, using the standard set for the alcohol industry) instead of completely restricting advertising to individuals over age 21. Yet millions of children could still be exposed to cannabis advertising if offered via a popular media channel, such as online. Since cannabis advertising can be placed on billboards, buildings, or storefronts, exposure of adolescents to proindustry messaging can occur readily in neighborhoods and areas with retail outlets (Allard et al., 2023; Firth et al., 2022; Shi and Pacula, 2021; Swinburne, 2022).

More than half of states with legal adult use place restrictions on the physical location of advertising. In California, for example, cannabis advertisements cannot be placed within 1,000 feet of a school, daycare center, or youth center where children are present. However, these restrictions will not eliminate children’s exposure to cannabis advertising because they can still be exposed where they live. Other policies found commonly across states include restrictions on retail building signage, guidelines specific to internet-based advertising, prohibitions on depicting product consumption, and requirements to include warning statements (Allard et al., 2023; Schauer, 2021).

In compliance with the standards of the U.S. Poison Prevention and Packaging Act, all states with legal adult use require cannabis and cannabis-infused products to be dispensed in child-resistant packaging (Schauer,

2021). Furthermore, most states mandate that package visuals cannot appeal to or directly target underage people with cartoons, toys, shapes, or designs. Some jurisdictions explicitly state that packaging cannot resemble any product that does not contain cannabis or cannabis concentrates, namely items typically marketed to children, such as food. Several states prohibit the use of the word “candy” or “candies” on labeling, and others prohibit using specific fonts that may appeal to underage people. Additionally, at least three states (Maine, Massachusetts, and Washington) require a visual symbol on the product package indicating that the product is not safe for children (Schauer, 2021).

Policies for Reducing Cannabis-Impaired Driving

Every state, whether it has adopted an adult-use law or not, has rules regarding cannabis-impaired driving, but state standards vary considerably. Three states with adult-use laws (Arizona, Michigan, and Rhode Island) have adopted zero-tolerance laws, prohibiting drivers from having any amount of THC or its metabolites in the body when driving (APIS, 2023b). Three adult-use states (Illinois, Montana, and Washington) have adopted specific “per se laws,” which prohibit drivers from driving with a detectable amount of THC—from 2 nanograms per milliliter (ng/mL) to 5 ng/mL—in their blood, regardless of evidence indicating whether that amount would impair the average driver. Colorado similarly specifies that drivers with more than 5 ng/mL of THC in their blood can reasonably be presumed to be impaired. Still, that state allows a defendant an affirmative defense, meaning that even if the defendant tests above 5 ng/mL, they can provide other evidence to demonstrate they were not impaired. Most adult-use states, however, specify that it is necessary to determine whether the driver was under the influence (i.e., impaired) by THC identified in the body, even if the amount exceeds a specified threshold. The public health value of these different approaches is difficult to ascertain.

The variability in state laws for blood THC limits stems at least partially from the lack of correlation between the level of delta-9-THC in the blood and the degree of impairment. This lack of correlation is due to the ability to develop tolerance; individual differences; and other factors, such as mode of use. With more frequent use, such as daily, and use at multiple times during the day, individuals have higher levels of blood THC reflecting accumulation that do not correspond with recent use or impairment. In a meta-analysis of 28 studies, McCartney and colleagues (2022) found a weak association between THC biomarkers and impairment for individuals who use cannabis occasionally and no association with impairment among individuals who use cannabis regularly. They therefore concluded that blood THC level is a poor indicator of impairment. Other studies using

driving simulators have similarly found that blood THC levels are not well correlated with impairment or recent use. Given the ability to develop tolerance to some of the cognitive and psychomotor effects of cannabis, it is unclear whether recent use necessarily leads to driving impairment among those who use it regularly (Colizzi and Bhattacharyya, 2018). In a study of individuals using cannabis for medical reasons, for example, recent use did not significantly affect cognition or performance on neuropsychological assessments or simulated driving, despite measurable levels of THC in the blood (Arkell et al., 2023; Manning et al., 2024), highlighting the challenges of using blood THC–based assessments to determine impairment (Arkell et al., 2021; Brooks-Russell et al., 2021; Marcotte et al., 2022).

Furthermore, blood THC levels differ importantly based on the route of administration. With edibles and other forms of oral ingestion, blood levels achieved are substantially lower than they are with inhaled cannabis because THC goes through first-pass metabolism and is converted to an active metabolite (see Figure 1-6 in Chapter 1). Here too, then, considering THC level alone will not indicate recent use or impairment (Newmeyer et al., 2016; Spindle et al., 2021).

Finally, the detection of cannabis-impaired driving is further hindered by the limitations of current roadside detection technology and approaches. For example, studies with placebo-controlled designs have found that field sobriety testing, which was initially developed to detect alcohol impairment, has relatively low sensitivity and specificity for detecting cannabis impairment (Bosker et al., 2012; Downey et al., 2012; Marcotte et al., 2023).

Additional Aspects of Regulations Relevant to Impacting Supply, Product Sold, and Possession/Use

Restrictions on Cultivation

“Home cultivation,” also referred to as home growing or self-cultivation, refers to growing cannabis plants at home. States have different quantity limits, prerequisites for plant maturity, and licensing requirements. Currently, two states—Illinois and Washington—fully prohibit home cultivation (Wadsworth et al., 2022b). All other states allow home cultivation to some degree, although the number of plants allowed, particularly plants in a flowering state, varies. Most states limit cultivation to 6 plants, with up to 3 in a flowering state. However, some permit only 2–4 plants (Oregon, Maryland, Virginia, Washington, and the District of Columbia), and two states permit the cultivation of more than 6 plants (Minnesota and Maine—up to 8 and 18 plants, respectively—if 50 percent are flowering) (Wadsworth et al.,

2022b). Home growing is not subject to mandatory testing requirements, so there are no controls on possible contaminants in homegrown cannabis.

Commercial cultivation of cannabis is defined as growing cannabis from seedlings or immature plants to maturity with the intent to distribute or sell, or any cultivation larger than the state limits for home cultivation (Schauer, 2021). States regulate commercial cultivation of cannabis predominantly through preapproved cultivation licenses. Distinctions among state regulations arise in the type and range of licenses offered. Licenses are often categorized into tiers based on permissible canopy size and location (e.g., indoor, outdoor, mixed). Vertical integration (the ability of the same party or entity to grow, process, and sell cannabis) is legal in all states except Washington. In Washington, neither the cannabis producer nor the processor may have a vested financial interest in any cannabis retailer.

Restrictions and Requirements on Packaging and Labeling

Food and beverage products regulated by the FDA are subject to clear guidelines on labeling and packaging, as these are critical to consumer safety. Similarly, uniform standards for packaging and labeling of cannabis products allow consumers to make decisions about product safety and risks associated with use. While packaging and labeling restrictions are mandated by every state that has legalized adult cannabis use, the regulations vary widely across states (Schauer, 2021). Only three states require plain packaging, and they all define it differently, from requiring that the package be only one color with no information but the required labeling to requiring that it be plain without bright colors; nine states require opaque packaging for cannabis products (Swinburne, 2022).

At least eight states require a “universal symbol” providing a visible notation that the product contains cannabis to help prevent accidental ingestion of products that may look like noncannabis products. The universal symbol used varies from state to state (except for Massachusetts and Maine, which share the same symbol) in terms of color, design, and the content of the warning (Schauer, 2021). At least four states (Colorado, Massachusetts, Maine, and Nevada) require that their “universal symbol” be printed onto each serving of multiple-serving edible cannabis products (Schauer, 2021). States almost universally require that THC and CBD content be listed on the label, although little uniformity exists regarding requirements on how this content is to be presented. More than 80 percent of states require inclusion of the batch number, product tracking, and manufacturer contact information. Many states also require that packaging include a warning label. However, the specifics of this warning again differ among states,

and the information is presented as a long legal disclaimer in small font, limiting its effectiveness in communicating potential risks. Some states require the inclusion of warnings regarding pregnancy or breastfeeding, delayed intoxication, driving, and operation of heavy machinery, while few mandate the inclusion of general health and dependence risks. Various items, ranging from usage instructions, nutritional panels, potency statements, and food allergens, are required by only a few states (Kruger et al., 2022).

Restrictions on Possession and Use

Limits on the amounts of cannabis possessed typically coincide with the quantity of cannabis that can be sold in a single transaction. While there is a moderate amount of variability across states in the specific amounts that consumers can possess or carry without threat of penalty, the penalties for possession of amounts above these limits can be significant. Many states also impose restrictions on where people can possess or use cannabis. Eleven states prohibit the use of cannabis in public, and another eight states have restrictions on use in public. Only Missouri, Rhode Island, and Virginia have no explicit public-use rules. Consumption on any federally regulated land is prohibited at the state level. Only a few states, such as California, allow individual municipalities to decide on appropriate public uses of cannabis with no state intervention. A few states authorize social uses of cannabis, such as in cannabis consumption sites (APIS, 2023b).

Cannabis Regulation in the United States: Findings

The lack of data documenting the systematic enforcement of the public health–oriented regulations, and flagrant evidence of violations of them, leads to tremendous skepticism and uncertainty as to the real public health benefit of the existing state laws. In many states there are clear violations of laws on sales of youth-oriented products (Luc et al., 2020) and on promotion of cannabis products to youth (Cui et al., 2024; Krauss et al., 2017), as well as violations on marketing rules, including posting health claims (Berg et al., 2023; Shi and Pacula, 2021). Retailer trainings are not targeting safety of products to consumers, as evidenced by research showing sales to pregnant women (Barbosa-Leiker et al., 2022; Dickson et al., 2018). This patchwork of federal, state, and local regulations—coupled with a lack of information on actual enforcement of existing regulations—makes it challenging to study the effectiveness of different approaches within the states. Additional data are needed before careful evaluations can be conducted.

allowed to operate online stores where those above the legal purchase age could order products to be delivered by mail. Today, only seven provinces allow private retailers to make online sales; such sales in the other provinces are run exclusively by the government (CCSA, 2024).

The provinces’ retail markets vary considerably. Only government-owned retail stores are allowed to sell cannabis in four provinces (New Brunswick, Nova Scotia, Prince Edwards Island, and Quebec), while five provinces (Alberta, Manitoba, Nunavut, Saskatchewan, and Yukon) have strictly private retail licensees operating retail stores (CCSA, 2024). The remaining provinces have a mix of government-run and private outlets. There is important variation among the provinces in the number of retail outlets, with some provinces, such as Alberta and Saskatchewan, having more than 15 outlets per 100,000 people, and others, such as Quebec and Prince Edward Island, having fewer than 2.5 per 100,000 (Rosenberg et al., 2023). All provinces but Quebec and Manitoba allow for home production (CCSA, 2024), but variation exists in the number of plants allowed (Pardal and Wadsworth, 2023).

Finally, while the Cannabis Act specifies a minimum age for possessing and purchasing cannabis (18 years), provinces can choose to raise the minimum legal purchase age above this level. Today, only Alberta has a minimum age of 18; the rest of the provinces have set it at 19 except for Quebec, which in January 2020 raised its minimum age from 18 to 21 (CCSA, 2024).

When retail sales began in Canada in October 2018, very few products were allowed to be sold. Initially, only dried flower and some oral oils were allowed. Nationally, these products were taxed at a rate of 10 percent, or $1 Canadian per gram. When vape cartridges, concentrates, and edible products started being sold for nonmedical purposes in early 2020, they were taxed as a function of their delta-9-THC content ($0.01 per mg of THC).

The variation over time within and among provinces in the number and type of retail outlets, minimum legal purchase ages, and types of products sold over time has been used to investigate the role of cannabis policy in various health outcomes. First, the introduction of edibles, vape oils, and other products into the adult-use retail market in January 2020 led to a rise in the prevalence of consumption of these higher-THC-concentration products (Hammond, 2023). The introduction of these products occurred at the same time as a rapid expansion of the commercial availability of cannabis retail outlets in a few territories, particularly Ontario. A series of studies shows that the combination of these two factors led to significant increases in the total number of emergency department (ED) visits involving cannabis (Myran et al., 2022); the number of ED visits attributable to cannabinoid hyperemesis syndrome (Myran et al., 2022), the number of

ED visits for cannabis-involved traffic and motor vehicle injuries (Myran et al., 2023a), and the number of ED visits for cannabis-involved psychosis (Myran et al., 2023b). Notably, these studies did not show similar increases in ED visits following the simple legalization of the retail sale of flower products and home cultivation. Other studies have found additional evidence of public health harms associated with the introduction of higher-THC-concentration products in other provinces or across Canada, including a dramatic rise in the rate of ED visits for cannabis-involved poisoning, particularly among children (Myran et al., 2023c; Varin et al., 2023; Yeung et al., 2021).

Studies examining either self-reported impaired driving, administrative data on traffic crashes, or ED visits associated with motor vehicle crashes in Canada have shown no statistically significant relationship with the opening of adult-use markets in the first phase of legalization, when only dried flower could be sold (Callaghan et al., 2021; Imtiaz et al., 2024; Nazif-Munoz et al., 2023; Walker et al., 2023). These findings provide an interesting contrast to studies finding a rise in crashes when products containing higher THC concentration were allowed (Myran et al., 2023a).

Cannabis legalization in Canada is also associated with a reported rise in easy access to cannabis between 2018 and 2019 (Wadsworth et al., 2022a) and other impacts on health. Callaghan and colleagues (2023) evaluated weekly counts of ED visits associated with cannabis-related disorder and poisoning among underage youth in Ontario (<19 years of age) and Alberta (<18 years of age) from April 2015 through December 2019. Using a time series model, they found a 20 percent increase in underage cannabis-related ED visits associated with legalization. Yeung and colleagues (2020) examined monthly cannabis-related ED visits (2013–2019) and poison center calls (2016–2019) in Calgary and Edmonton, Alberta, using a pre–post time series design. They found a small but statistically significant increase in cannabis-related ED visits and a more considerable increase in poison center calls. Alberta was one of the provinces that experienced an immediate rapid expansion of retail stores with legalization.

A comprehensive review of the public health impacts of legalization in Canada by Hall and colleagues (2023) found that legalization resulted in a substantial decline in cannabis-related arrests, decreases in the legal price of cannabis, and a substantial increase in the THC concentration of cannabis products, but only a modest increase in past-month use by adults and mixed findings on impacts on use by youth. The authors note that the rise in acute ED visits involving cannabis among adults occurred in areas with rapid expansion of retail outlets and that the greatest increases in adolescent poisonings occurred after the introduction of edibles in the adult-use market.

___________________

² Calculated assuming an adult population of 3,423,108.

contact with illegal dealers (Queirolo, 2020). At the same time, constraints such as a production shortfall and lack of retail sites may have prevented the system from being implemented fully (Isorna et al., 2023). Furthermore, concerns exist that the highly regulated approach may have discouraged cannabis purchasers from using the legal market. While registration among those who use cannabis has grown since it was first legalized, it is estimated to reach just 51 percent of those who use cannabis in Uruguay, including 34 percent who obtain cannabis directly from the legal market and 17 percent who obtain it from others who obtained it through the legal market.

Limited data exist on the effects of Uruguay’s cannabis legalization on health. To date, research has shown no impact on cannabis use or perceived risk of use among adolescents (Laqueur et al., 2020). While a transitory increase in risky and frequent cannabis use was observed in 2014 immediately after cannabis legalization among 18- to 21-year-olds enrolled in school this increase was not sustained over time (Rivera-Aguirre et al., 2022). No analyses have been published on the effects of cannabis legalization in Uruguay on cannabis consumption among the general adult population. A study of pregnant persons in a public hospital in Montevideo did find a rise in the use of alcohol and cannabis before and during pregnancy in 2016 compared with 2013 (Pinto et al., 2020). However, the study’s cross-sectional design with no comparison group limits the ability to draw inferences about the extent to which this change was due to cannabis legalization.

Two peer-reviewed studies examine the relationship between legalization and traffic outcomes in Uruguay. Using an interrupted time-series approach with weekly data on fatal automobile crashes from 2012 to 2017, Nazif-Munoz and colleagues (2020) concluded that the enactment of legalization in December 2013 may have been associated with an increase in fatal motor vehicle crashes, particularly among car drivers and in urban settings. Kilmer and colleagues (2022) used department-level variation in registrations for legal cannabis over time to examine the association with traffic crashes involving injuries. While they found no evidence that total registrations were associated with these crashes, they did find a consistent, positive, and statistically significant association between the number of individuals registered as self-cultivators and the number of traffic crashes with injuries. This finding is generally consistent with that of Nazif-Munoz and colleagues (2020), which focused mainly on the period before pharmacies began selling cannabis.

Alternative Regulatory Models: Findings

Canada and Uruguay legalized cannabis federally and have more consistent policies nationwide compared with the United States. Canada has mandated minimum-age limits and national restrictions on advertising, and some provinces have a state monopoly. At least initially, there has been no

increase in impaired driving after legalization, but a rise in ED visits has been associated with higher-THC products and rapid retail expansion. Uruguay has a highly regulated model with limited retail outlets and product types. Studies thus far show a shift from illegal cannabis to legal options but may not have fully captured the market because of limitations on access. Data on public health impacts in Uruguay are limited but suggest no increase in adolescent use other than a possible rise in risky use among young adults shortly after legalization. There is conflicting evidence on traffic crashes.

CONCLUSIONS AND RECOMMENDATIONS

With federal cannabis prohibition in the United States creating a patchwork of state laws, the legal cannabis industry operates in a complex and often contradictory environment. The market for hemp products resulting from the 2018 Farm Bill is a prime example of the confusion resulting from limited federal involvement. The ambiguous definition of hemp in that legislation has led to a largely unregulated market for semisynthetic cannabinoids. These products raise significant concerns about safety, accurate dosing, and potential misuse, especially among young adults who may have easier access.

Conclusion 2-1: The redefinition of the federal meaning of “hemp” in the Agricultural Improvement Act of 2018 (2018 Farm Bill) has created considerable uncertainty and confusion as to what cannabis products are legal and has led to a massive new market in semisynthetic cannabinoids with little regulatory or public health oversight.

Recommendation 2-1: Congress should refine the definition of “hemp” to state clearly that no form of tetrahydrocannabinol or semisynthetic cannabinoid derived from hemp is exempt from the Controlled Substances Act.

State-to-state variations in regulations leave public health guidance unclear and limit efforts to prevent harmful use. The lack of federal oversight has fostered a fragmented industry with inconsistent regulations, oversight, and enforcement standards. In contrast, some countries have adopted a more measured approach with stricter government control over cannabis legalization. Such a stricter regulatory framework may better protect public health.

Conclusion 2-2: The federal government has not provided adequate guidance on public health policies that might minimize the adverse consequences of cannabis legalization. States that have legalized cannabis have created regulatory frameworks that have prioritized commerce

over public health. The significant state-to-state variation in regulations on products, retail sales, and use has resulted in inconsistent applications of public health safeguards. A better understanding of the influence of this variation on public health is needed.

Conclusion 2-3: Other countries have taken a centralized, government-regulated approach to protecting public health by placing stricter controls on the access to, availability of, and safety of cannabis products.

The committee’s evaluation of policies that limit youth exposure to cannabis found significant variation among the states. While all states require those who use cannabis to be age 21 and older, enforcement through random checks is limited. Advertising restrictions are also inconsistent among the states. Most states allow cannabis advertising with some limitations on who sees it (not necessarily age-restricted) and where it is placed (e.g., not near schools), with the result that millions of children are exposed to procannabis messages. Loopholes further weaken these restrictions. Unlike stricter countries, some U.S. states permit advertising with enticements such as coupons, health claims, and even depictions of product use. Additionally, few limitations exist on targeting people outside the state or using public platforms such as billboards. Although product packaging is regulated to prevent child appeal, the lack of enforcement and weak advertising restrictions create a situation in which young people in the United States are still subject to cannabis promotion. It is important to note that if cannabis is legalized for sale in the United States at the federal level, advertising restrictions will become more difficult because of First Amendment protections for the advertising of legal products. Advertising for tobacco and alcohol is restricted because other policies gave states the authority to do so (the 21st Amendment in the case of alcohol and the Master Settlement Agreement for Tobacco) (Lange et al., 2015). Additionally, best practices for limiting advertising to youth need to consider where youth are receiving the information; restrictions on advertising on social media thus are likely more critical than restrictions on traditional media outlets.

State-level cannabis legalization is illegal under federal law unless cannabis, like tobacco or alcohol, is removed from the Controlled Substances Act. Still, given that the federal government has been allowing the states to create commercial markets for cannabis under federalism, federal agencies could assist the states that have chosen to legalize. The Council on State and Territorial Epidemiologists, a nonprofit organization of member states and territories representing public health epidemiologists, which includes the CDC, has guidance and resources on public health surveillance. Similar guidance could be created for other public health functions.

The National Conference of State Legislatures (NCSL) and the National Governors Association (NGA) may be able to provide leadership on how

jurisdictions can protect youth access and exposure to cannabis products. The NCSL provides bipartisan policy research, training resources, and technical assistance to every state legislator and staffer. The NGA is a nonpartisan political organization founded in 1908, representing 55 states, territories, and commonwealth governors. The two organizations often work together to provide examples of legislation.

Recommendation 2-2: In conjunction with other federal agencies, the Centers for Disease Control and Prevention should conduct research on and develop best practices for protecting public health for states that have legalized cannabis, drawing from tobacco and alcohol policies. These best practices should encompass marketing restrictions (e.g., on advertising and packing), age restrictions, physical retail and retail operating restrictions, taxation, price restrictions, product design, and measures to limit youth access. Other strategies for protecting public health that warrant identification of best practices include reducing cannabis-impaired driving, promoting state retail monopoly, and encouraging cultivation practices that limit contamination of both products and the environment. The best practices should be reconsidered and updated periodically as new research emerges.

Recommendation 2-3: The National Governors Association, the National Conference of State Legislatures, and other public health stakeholders should develop model legislation concerning best practices related to marketing restrictions (e.g., on advertising and packaging), age restrictions, physical retail and retail operating restrictions, taxation, price restrictions, product design, and measures to limit youth access, as well as strategies for reducing cannabis-impaired driving, promoting state retail monopoly, and encouraging cultivation practices that limit contamination of both products and the environment. Once the Centers for Disease Control and Prevention’s best practices have been developed, they should be incorporated into the model legislation.

REFERENCES