BERL, an economic consultancy firm tasked by the New Zealand Ministry of Justice to model the impacts of the CLCB, estimated that the legal cannabis sector facilitated by the CLCB would create 5000 full time jobs, representing wages and salaries of $210 million per year, and contribute $440 million to GDP . One rurally based Māori medicinal cannabis company has been established with these development and employment goals in mind . Conversely, there was strong opposition to the CLCB amongst right leaning voters in New Zealand , again consistent with findings from the U.S . The opposition of National Party voters is understandable given the party’s traditional right-wing conservative base. The opposition of ACT voters is, on the face of it, less easy to understand given ACT describes itself as a “classical liberalism” political party that promotes “individual choice” and “small government”.However, economic conservatives in Western democracies often adopt conservative views on not only economic issues but also social issues to more closely align themselves with their socially conservative allies . We also found an association between considering frequent cannabis use to be a high health risk and not supporting the CLCB. The health risks of cannabis use are also cited in the U.S. as a leading reason to oppose legalisation . It is interesting to note that perceptions of the health risk of frequent cannabis use , as opposed to merely “trying” or “using cannabis weekly or less often”, is the dominant predictor. This suggests there is a somewhat nuanced understanding of the health risks of cannabis, consistent with the findings from New Zealand longitudinal research . Those who had tried illegal drugs other than cannabis in their lifetimes were more likely to oppose the CLCB once we controlled for moral views of cannabis use. Recent use of other drug types has also been found to be associated with opposition to marijuana grow system legalisation amongst young adults in the U.S. . This may represent specific concerns or views about cannabis, as opposed to other drug types.

Lifetime experience of other illegal drug use may include those who have experienced negative experiences from drug use in the past, and this may translate into opposition to drug liberalisation in the present. In the U.S., a lack of support for cannabis legalisation in some counties has been explained by high levels of illegal cannabis cultivation in these areas and the desire to maintain black market income streams . Finally, we found that reading summaries, parts of, or the whole CLCB was a significant predictor of support for the bill. It appears that knowledge of the regulatory controls of the legal cannabis market proposed in the CLCB increased the likelihood respondents would support legalisation. However, the causality of this association can be questioned. One interpretation is people were convinced to support the CLCB once they actually read the Bill’s content. An alternative explanation is that those already positively inclined to support legalisation were more likely to spend time reading the CLCB, and thus the details merely served to reinforce their pre-existing voting intentions.On March 11th, 2020 the World Health Organization  declared the severe acute respiratory syndrome coronavirus  a pandemic , and by March 13th, the United States  president declared it a National Emergency . Within weeks of these declarations, states and localities across the US began to institute stay-at-home orders, closure of non-essential businesses, and many Americans began the transition to remote work . These disruptions have led to rising unemployment , market volatility , housing and food insecurity , and social isolation . As a result, concerns about a ‘second pandemic’, constituting an increase in psychiatric and substance use disorders, began to emerge . Since the early 2000′s, the US has seen a shift in state cannabis policies, which have been found to be associated with increased prevalence of cannabis use and cannabis use disorders among sections of the population , with 29 states operating medical cannabis dispensaries and an additional 8 states operating both recreational and medical dispensaries as of March 2020 . Studies of the effects of cannabis policies on adult use have shown increased cannabis use and use disorders in states with medical cannabis policies . In addition, epidemiological surveys of substance use in the US have shown increases in cannabis use since 2002, with significant increases among certain sociodemographic groups including men, Black people, young adults, low-income groups, and those never-married .

With regards to accessibility of cannabis products, most states with operating dispensaries declared them essential businesses during the COVID- 19 pandemic , along with liquor and tobacco retailers, which should have provided continued access to cannabis products for purchase. However, it is unclear how accessibility to cannabis products was impacted in states with no legal cannabis options. Moreover, other factors beyond cannabis accessibility may have impacted cannabis use behavior during the COVID-19 pandemic  and studies investigating trends in cannabis use over the course of the pandemic are needed. Emerging literature on the effects of the pandemic on mental health and substance use have shown symptoms of depression , anxiety , loneliness , and alcohol consumption  increasing in some segments of the population early on in the pandemic. Elevated mental distress in the population due to COVID-related stressors may increase the use of cannabis as a coping strategy or self-medicating behavior. A cross-sectional study of emerging adults in Canada found that self-isolation and motives to use cannabis for coping with depression were associated with cannabis use during the pandemic . A study of individuals who used cannabis in the Netherlands showed that 41% of respondents reported increased cannabis use since lockdown measures were instituted, with stress and mental health significantly associated with reported increases . However, studies on cannabis use in the US during the COVID-19 pandemic remain sparse and further investigation of long-term outcomes of the pandemic, including potential changes in substance use behaviors, is warranted. As the literature grows in monitoring the effects of the pandemic on mental health and alcohol use in the US, it is also important to examine potential changes in cannabis use. The aims of this study were to  describe changes in days of past-week cannabis use from March 10th through November 11th, 2020 among US adults who reported cannabis use in a nationally representative panel and  characterize differences in trends of use within sociodemographic subgroups and by state cannabis policy status.Participants were sampled from the Understanding America Study , a probability-based, nationally representative Internet panel of adults .

The UAS has recruited participants using Address Based Sampling  since 2014, in which postal records are used to select a simple random sample from a listing of residential addresses across the US. The recruitment involved invitation by mail, with potential participants without prior internet access were provided with tablets and broadband internet connections to facilitate data collection. Once respondents enrolled in the panel, they were surveyed via computer, mobile device, or tablet. Individuals are considered eligible to join the panel if they are aged 18 years or older and are a member of the contacted household . Additional details regarding the UAS methodology can be found at the UAS website . This study used data from 16 waves of the UAS’s COVID-19 Longitudinal Survey, a high-frequency longitudinal data collection with baseline data collection running from March 10 to March 31, 2020. All existing members of the UAS were invited to participate in the survey. Starting on April 1, respondents were invited to participate in bi-weekly surveys according to a staggered schedule, whereby one fourteenth of the sample was invited every day. Every respondent had 14 days to complete the survey; thus, the waves following baseline overlap in calendar time. In the initial survey, respondents were asked for consent to participate in the bi-weekly surveys. Only respondents who consented were asked to complete a subsequent survey on their assigned day. As not all eligible participants had consented by the start of the second wave, the response rate as a percentage of the complete UAS sample was lower in earlier follow-ups. Overall, there were 8547 eligible panel members invited to participate in the March survey. Among those invited, 6932  completed the survey at baseline, March 10 – March 31, 2020. For purposes of our analyses, we included only those participants who reported at least one day of cannabis use across the survey period. The percentage of participants reporting cannabis vertical farming use at each wave ranged from 9.2% in wave 14 to 11.3% in wave 3 . On average, those who reported ever using cannabis included higher proportions of individuals who were younger, identified as being Black or Hispanic/Latinx, were living at or below the Federal Poverty Level, and came from states with both medical and recreational cannabis policies.

Comparisons between adults who reported using cannabis and those who did not report using cannabis during the survey period are displayed in the online supplement Table S2. Surveys with complete data on cannabis use and all sociodemographic characteristics of interest were included in the analytic sample. In total, 1761 unique participants were included; 39.9% completed 16 surveys, 12.6% completed 15 surveys, 7.6% completed 14 surveys, and the remaining 39.9% completed between one and thirteen surveys . Details for participant inclusion and exclusion at each survey are displayed as a flow diagram in online supplement Figure S1.Analyses were conducted in three parts. First, the associations between each of the sociodemographic characteristics and the frequency of cannabis use were examined across the full survey period. Second, a single model with the splines for days since March 10 as covariates examined trends in cannabis use across time. Third, a sequence of models with interactions between the splines for days since March 10 and each of the identified sociodemographic characteristics determined whether trends in cannabis use over time differed between subgroups. We used mixed-effects linear regression models with a random intercept for participants to accommodate repeated measures. The general specification for these models is provided in the online supplement. Joint Wald tests were used to determine if interactions were significant. The margins and the xbrcspline commands in Stata were used to generate linear predictions of cannabis use and to estimate differences in the frequency of cannabis use on given survey dates compared to March 11, respectively, in the overall analytic sample and stratified by each sociodemographic subgroup . March 11 was used as the reference date rather than March 10 due to a higher number of observations on that survey date . We conducted additional sensitivity analyses using the entire UAS sample, including participants who reported no cannabis use over the entire study period. All analyses incorporated survey weights that accounted for probabilities of sample selection and survey non-response at baseline and are aligned with Current Population Survey benchmarks. Statistical significance was assessed at the p<.05 level. Analyses were conducted using Stata version 16  and R .To our knowledge, this is the first study to examine trends of cannabis use during the COVID-19 pandemic in a general population sample of U.S. adults.

The prevalence of cannabis use in this study population ranged between 9.2 and 11.3% at each wave. This is slightly lower compared to national estimates of past month cannabis use of 11.9% in 2019 among US adults 18 or older . Our analyses show that, within the overall sample of adults who reported any cannabis use during the study period, there were statistically significant increases in the number of past-week cannabis use days at the start of the pandemic  compared to baseline ; thereafter, these increases returned to levels comparable to March for the remainder of the study period . When comparing cannabis use at the first of each month to the start of the study, several of the identified sociodemographic groups also demonstrated increased cannabis use in April, May, and June, including: women, non-Hispanic White people; individuals living with only their partner; those reporting a household income not at the FPL; and those living in states with medical cannabis but not recreational only policies.