Implementing the policy first in jurisdictions where tobacco consumption is dropping and the public supports tobacco control measures could increase the likelihood of success. Some government alcohol stores already sell tobacco products; this would reduce costs and complexities of implementation.Alcohol control systems have encountered opposition in recent years, and some control jurisdictions have relaxed their policies . In the US, no states have established control systems since the immediate post-Prohibition period; thus, expanding such systems might meet opposition from those who would like to eliminate them, including the alcohol industry. In other countries , government retail monopolies may be better accepted and expansion into tobacco products better received. The addition of marijuana sales to government stores and framing expansions of the system as a means to rationalizing substance regulation could also increase public acceptance. Such expansion might create new public health challenges. With emphasis on free trade and free markets, business managers pressure alcohol monopolies to open new outlets and increase profits. Adding tobacco products as a new profit center could contradict public health goals unless endgame targets are built into changes in the monopoly system. Alcohol control bureaucracies might be unwilling to be purveyors of an addictive and deadly product, or regard moving tobacco into their system as suggesting that alcohol and tobacco are equivalent dangers. Again, vertical grow built-in endgame goals might mitigate these objections. The most obvious adverse effect of availability restrictions is an increase in informal market activities .

The tobacco industry’s standard argument against all effective tobacco control measures is an alleged increase in illicit trade. However, a government tobacco monopoly would keep a legal supply available and informal market activities could be limited by effective enforcement. As with ‘plain packaging’ , ingredients disclosure, and product labeling, the tobacco industry might use international law and treaties to oppose government tobacco monopolies. International law and trade agreements constrain domestic regulations, but allow government measures to protect human and environmental health. Restricting the supply and marketing of alcohol and tobacco have been defended successfully against challenge as both necessary and proportionate to achieving government health goals. Tobacco control advocates might disagree about which products should be restricted to government stores. Some would argue that such a policy should apply equally across product types. Others would argue for a ‘harm reduction’ approach in which only the most dangerous products would be transitioned to government stores, leaving smokeless tobacco, e-cigarettes and various nicotine products thought to be less harmful still available at private retail stores. Some tobacco companies claim to support regulation based on the relative harm of tobacco products and could seize the opportunity to capture a larger market for ‘less harmful’ products. It would parallel decisions by some alcohol control regimes about product categories .Predicting the disadvantages of a government tobacco monopoly is complicated by the absence of creation of control state regimes in the last 70+ years and by the differences between alcohol and tobacco products. One objection might be that limited store locations present ‘equity’ issues – some communities might have less access to tobacco products than others.

Disadvantaged communities, however, regard the current situation as inequitable because tobacco products are more available and heavily promoted to them than to others, and more available than other, healthy products. Ensuring that community representatives are involved in planning and allocating resources for the transition would be critical. Making tobacco less convenient might also spur interest in online sales. In the US, online sales are regulated through the PACT Act that is intended to ensure state taxes are collected and proper identification is required. Internet vendors continue to promote tax free cigarette sales and the Act is ineffectively enforced. A high level of credit card fraud associated with online tobacco purchase attempts may limit such sales. Health harms of tobacco are aggravated by simultaneous alcohol use. If both substances are sold at the same locations, dual use could increase. Other potential disadvantages might arise from strengthening government involvement in tobacco sales. Jurisdictions that have eliminated government monopolies on alcohol have lost alcohol revenues; thus, placing tobacco under a control regime might increase tobacco revenues . This could create increased dependency on tobacco revenues, reducing policymakers’ appetite to implement those changes. Earmarking tobacco revenues for tobacco control efforts, including research, prevention, and policy development and implementation, could help resolve these problems. Need for tobacco revenues beyond baseline levels would decrease as use decreased. Establishing an endgame goal of eliminating tobacco sales by a predetermined date from the outset could mitigate such dependence.Integration of tobacco and alcohol control bureaucracies could be problematic. Alcohol control systems were not designed to be and have never been part of a strategy to end alcohol use. Without specific policies designed to reduce sales, a similar philosophy could permeate tobacco control systems, resulting in an institutionalization of government tobacco sales, rather than an endgame.

A final concern would be placing the imprimatur of government on tobacco sales. Tobacco control advocates believe that the widespread availability of tobacco products signals that they are ‘normal’ and thus less dangerous than they really are. Moving products to a government store places them in a ‘special’ category. The intended message is that they are too dangerous to be widely available. It might instead signal that they are ‘officially approved’. Tobacco users might reasonably assume that the government would not sell products known to kill their users. It could also create the perception that the government was in partnership with tobacco companies, a direct conflict with its public health objectives.The incongruity of warning the public about the dangers of tobacco products, while continuing to allow their ubiquitous sale, could be addressed with a transition of sales to existing government operated stores. Tobacco in government stores would improve policy coherence, integration, and proportionality, thus reducing current regulatory disparities among harmful substances. Such a move would allow continued sales of tobacco products, but allow governments increased options for regulating them. It could also serve as a step toward an endgame goal of eliminating sales of commercial tobacco. While transitioning tobacco sales to government-run alcohol stores may runcounter to trends in privatization of public services, the policy dynamics around health care costs, marijuana legalization, and dropping smoking rates suggest that it may be timely to examine this option.Cannabis use disorder is a substance use disorder defined by the same symptoms as other substance use disorders, including problematic patterns of use, tolerance, withdrawal, craving, neglect of other activities, clinically significant distress or impairment, and associated psychosocial and health-related problems. While many individuals can use cannabis without harm, 33% of those who use cannabis regularly develop cannabis use disorder ; among U.S. adults age 18 and older, the prevalence of CUD is 5·9%, or 15 million adults. The legal status of cannabis varies across countries and time periods. In the United Kingdom, cannabis use is illegal, with stricter legal penalties under discussion. In contrast, Canada legalized cannabis nationally for medical and recreational use in 2001 and 2018, respectively, with legislation that emphased public health concerns. In the United States, legalization has occurred on a state-by state basis: since 1996, 39 U.S. states enacted medical cannabis laws , and since 2012, 21 states enacted recreational cannabis laws . Much of the U.S. legislation has been industry-friendly rather than public health focused. Over the last two decades, the proportion of Americans favoring cannabis legalization has increased to 68%, perceived risk of cannabis use has decreased, rollling grow table and rates of U.S. adult cannabis use and CUD have increased. Adjusted for nationally increasing trends, MCL and RCL have been shown to play a role in the increasing prevalence of cannabis use. However, fewer similarly-adjusted national studies examined MCL and RCL effects on trends in the prevalence of CUD, two using the same survey data from U.S. general population and one using data from Veterans Health Administration patients. 

These studies indicate that MCL and RCL played a role in the increasing prevalence of CUD, but that their effects were modest. The modest-sized effects could occur if MCL and RCL effects differ across subgroups, obscuring important differences when groups are combined. Specifically, studies have not investigated whether state MCL or RCL effects differ between individuals with and without chronic pain. Chronic pain, is an increasing U.S. health problem. Veterans have high rates of chronic pain. Pain and CUD are associated. The prevalence of CUD diagnoses has increased disproportionately in VHA patients with chronic pain, particularly older patients, possibly due to increasing use of cannabis to manage pain, expanding the pool of individuals using cannabis who are vulnerable to CUD. Most U.S. adults, particularly in MCL and RCL states, believe that cannabis is effective for treating pain despite inconclusive evidence, and pain is the most common qualifying condition for medical cannabis. Given that cannabis legalization plays a role in increases in CUD prevalence, that rates of chronic pain are high in VHA patients, that pain is associated with CUD, and that beliefs that cannabis is useful for pain are more prevalent in MCL and RCL states, we hypothesized that MCL and/or RCL would play a greater role in the national increases in rates of CUD in VHA patients with chronic pain than in other patients. If so, the information would be important to convey to healthcare policymakers tasked with formulating laws and regulations regarding cannabis use, and to clinicians treating patients with chronic pain in different jurisdictions. The VHA is the largest integrated U.S. healthcare system. We leveraged VHA electronic health record data from 2005-2019 to investigate whether the role of MCL and RCL enactment in the increases in rates of CUD diagnoses differed between patients with and without chronic pain. Because MCL and RCL have stronger effects in older adults and disproportionate increases in CUD are greatest in older patients with pain, we also examined results by age group. Additionally, because the impact of MCL or RCL enactment could take time to emerge, we explored a 1-year time lag after enactment. Because enactment does not necessarily change cannabis availability, we also explored the dates that dispensaries were legally permitted. The ICD codes are available elsewhere. We excluded cancer-related and acute pain, e.g., fractures. We required diagnoses for specific conditions from ≥2 outpatient visits or ≥1 inpatient hospitalization within each study year. A binary variable was created indicating any chronic pain condition each year from 2005-2019. Primary exposures were state cannabis laws: state-year variables indicating state MCL and/or RCL enactment. Patient state-of-residence was indicated by last healthcare encounter for each year. States were categorized each year as No-CL, MCL-only, and MCL/RCL. For sensitivity analysis involving state legal dispensaries, which can occur subsequent to MCL or RCL enactment, we used RAND-USC OPTIC marijuana policy data30 to create state-year variables indicating the years that legally-protected dispensaries were operational. Age-group definitions followed previous studies: 18-34, 35-64, and 65-75 years. Patients ≥76 years were excluded because of their very low CUD prevalence. Control covariates included variables previously shown to be related to chronic pain and to CUD4 : age, sex, race and ethnicity . Time-varying yearly state control covariates included those previously shown to be related to state rates of various health conditions: state/year rates from American Community Survey data: % male; non-Hispanic white; non-Hispanic Black; Hispanic; ≥18 years of age; unemployed; below poverty threshold; and yearly median household income. One-year and 5-year estimates were used for 2005-2008 and 2009-2019, respectively, using R tidy census.We conducted the analyses in two stages. We began by describing overall time trends in yearly CUD prevalence by pain status among patients living in states that had enacted MCL-only or MCL/RCL or neither by the end of the study period, 2019. The purpose was to provide a general summary of the magnitude of differences and trend patterns across these three groups of states regardless of when laws were enacted during the period, while adjusting for background differences in demographics. To do this, we grouped diagnosed CUD prevalence across years 2005-2019 by chronic pain and by state cannabis law status at the end of the study period : No-CL; MCL-only; and MCL/RCL. Adjusted yearly prevalence estimates were obtained from a linear binomial regression model interacting state law status in 2019 by year, controlling for age, sex, race, ethnicity, and time-varying state covariates.