Characteristics that influence purchase choice for cannabis products: a systematic review
grid.25055.370000 0000 9130 6822School of Pharmacy, Memorial University of Newfoundland and Labrador, St. John’s, Canada
grid.25055.370000 0000 9130 6822Faculty of Medicine, Memorial University of Newfoundland and Labrador, St. John’s, Canada
Abstract
Introduction
When non-medical cannabis use became legal, government regulators implemented policies to encourage safer consumption through access to a regulated market. While this market is growing, sales still occur through unregulated channels. This systematic review identifies factors influencing cannabis purchasing to help policymakers understand why consumers still purchase illicit market cannabis (registered with PROSPERO CRD42020176079).
Methods
A comprehensive search strategy included databases in health, business, and social science fields (inception to June 2020). Studies were eligible for inclusion if they were conducted with persons who purchase cannabis and examine at least one attribute that would influence purchase choice and were published in the English language. Studies could be of any methodological design. Two independent reviewers completed two levels of screening, and all extraction was verified by a second reviewer. A qualitative synthesis of the findings was completed. The quality of the included studies was assessed using the Mixed Methods Appraisal Tool.
Results
Of the 4839 citations screened, 96 were eligible for full-text review and 35 were included in the final synthesis. Aspects of price were the most common factors (27 studies). Twenty studies measured price elasticity; most studies found that demand was price inelastic. Many other attributes were identified (e.g., product quality, route of administration, product recommendations, packaging), but none were explored in depth. Eleven studies addressed aspects of product quality including demand elasticity based on quality, potency, and aroma. Studies also explored consumer-perceived “quality” but provided no definition; differences in quality appeared to impact consumer choice. Smoking cannabis appeared to be the preferred route of administration but was only examined in three studies. There was insufficient data to understand in the impact of other attributes on choice. There appeared to be preference heterogeneity for different attributes based on the consumer’s experience, reason for use, and gender.
Conclusion
While price influences choices, demand is relatively inelastic. This suggests that consumers may be seeking lowest-cost, unregulated cannabis to avoid reducing consumption. Beyond price, there is a significant gap in our understanding of consumer choices. Perceived quality does appear to impact choice; however, more research is needed due to the lack of a recognized definition for cannabis quality.
Supplementary Information
The online version contains supplementary material available at 10.1186/s42238-022-00117-0.
Introduction
Cannabis is the second most commonly used psychoactive substance world-wide (Peacock et al., 2018; United Nations Office on Drugs and Crime, 2019). The global estimated annual prevalence of cannabis consumers aged 15–64 was 3.8% in 2017 or approximately 188 million people (Peacock et al., 2018). The number of people who use cannabis annually is also estimated to have increased by roughly 30% between 1998 and 2017 (United Nations Office on Drugs and Crime, 2019).
This rise may be accredited to recent changes in cultural-norms and policies in several countries regarding cannabis use (Bahji and Stephenson, 2019; National Academies of Sciences et al., 2017; National Institute of Health, 2015). Currently, cannabis for non-medicinal use is legal in Canada, Georgia, South Africa, Uruguay, the Australian Capital Territory in Australia, and specific regions in the USA (ACT Government, 2020; BBC News, 2018; Guthrie, 2018; United Nations Office on Drugs and Crime, 2019). Within the USA, there are 19 states and the federal District of Columbia which have legalized recreational cannabis (Solutions, 2019). Several countries have also adopted milder forms of punishment in regulating cannabis without actual legalization, through decriminalization or unenforced laws (Areesantichai et al., 2020; Hanford, n.d.; Smith, 2020). Moreover, medicinal use of cannabis has been prevalent and legalized in many countries for some time.
Illegal cannabis sales are still largely prevalent in Canada and beyond, with only 48% of Canadian cannabis consumers making their last purchase from a legal source and illegal retailers in California outnumbering legal retailers three to one (Wadsworth et al., 2021). People who use cannabis attribute the persistence of the illegal market to numerous issues that may decrease the appeal of legal cannabis. According to the media, cannabis consumers reported issues such as high cost (Deschamps, 2020; Esfandiari, 2019; Fahmy, 2019; Johnson, Glen et al., 2019; McCabe, 2019; Shackford, 2019; The Canadian Press, 2020; Tunney, 2019a), poor cannabis quality (Ahearn, 2018; Turvill, 2020), product moisture (Israel, 2019; Turvill, 2020), limited supply (CBC News, 2019; Cecco, 2019; Esfandiari, 2019; Geraghty, 2019; Johnson, Glen et al., 2019; Mazur, 2019; Tunney, 2019a; Williams, 2019), distance to licensed stores (Esfandiari, 2019; Johnson, Glen et al., 2019; Tunney, 2019b), and inconvenient packaging (Lamers, 2019). Through crowdsourced cannabis prices, Statistics Canada confirmed that the price of legal cannabis is more expensive compared to illegal cannabis (Statistics Canada, 2020). From 2018 to 2019, the average price of legal cannabis in Canada increased from $9.69 per gram to $10.30, while the average price of illegal cannabis dropped from $6.44 per gram to $5.73 (Statistics Canada, 2020). This mirrors the experience in the USA where illegal cannabis prices dropped substantially in states where it became legalized (Smart et al., 2017).
The multi-attribute utility theory (Torrance et al., 1982) states that when individuals make decisions, their choices are based on their preferences towards certain attributes of that choice. Likewise, there are many attributes or factors that people consider when making the choice between legal or illegal cannabis. A better understanding of the degree to which these specific factors influence decisions can help inform cannabis policy. Research to date has predominantly examined the effect of cannabis price on consumer demand by measuring price elasticity of demand. Price elasticity of demand (Gilroy et al., 2020) represents the degree to which demand for cannabis changes as price fluctuates. A common method to examine price elasticity has been the marijuana-purchase task (MPT) (Aston and Meshesha, 2020). The MPT is a simulated purchase scenario which evaluates consumers’ demand for cannabis in relation to a change in price (e.g., from free to $10 over 20 increments). This method is also used to look at demand elasticity in relation to characteristics other than price, such as product quality.
There are considerations or attributes beyond price that are important to consumers when they purchase cannabis. Some research has been done to focus on factors like quality, aroma, potency, packaging, and warning labels. However, because cannabis legalization is a new in many countries, there is a lack of research that attempts to bring existing research evidence on cannabis choice behavior together. Understanding the role that all of attributes of choice play in decision making may be informative for refining cannabis policies to better support public health and safety as well as meet consumer needs. This can also offer insight for countries looking to legalize cannabis for either medicinal or non-medicinal use. The purpose of this systematic review is to identify what factors influence cannabis purchasing behavior to inform the design of a cannabis choice modeling study. The secondary objective was to identify gaps and limitations in the existing evidence base.
Methods and analysis
Study design
This study was designed in accordance with the PRISMA statement on systematic reviews (Moher et al., 2009) and is registered (CRD42020176079) with PROSPERO (International Prospective Register of Systematic Reviews). The Covidence online systematic review software was used to assist in screening, selection, and data extraction.
Eligibility criteria
The research team used the SPIDER search strategy tool (Cooke et al., 2012) to define the key elements of the review question. This tool is designed specifically for research questions that lend themselves better to qualitative or mixed methods approaches. The inclusion and exclusion criteria along with the SPIDER search protocol are described in Table 1.Sample: individuals who have consumed cannabis for either medical or non-medical purposes Phenomenon of interest: consumer choice for cannabis products (could be either legal or illegal sources; and for either medicinal or non-medicinal purposes) Design: any study design, including but not limited to focus groups, interviews, case studies, observational studies or surveys were included. Non-English articles, systematic or literature reviews were excluded. Studies where only abstracts were available were also excluded. Evaluation: at least one situational attribute of choice (e.g., product characteristics, retailer characteristic) Research type: qualitative, quantitative or mixed methods
Data sources and search strategy
A comprehensive search strategy which aimed to find both published and unpublished studies was developed in conjunction with an experienced librarian (MS) and peer reviewed by a second librarian. The search included databases in health (Medline, EMBASE, PsycINFO), business (ABI/INFORM, Business Source Complete), and social science (ASSIA, IBSS, SocINDEX, Sociological Abstracts) fields. A broad index search in Scopus was also performed. The complete search strategy is included in the appendix. The strategy was first created in Ovid MEDLINE and was modified to fit other databases’ search criteria. Reference lists of key articles were also screened (JD). The search was conducted from inception to June 2020 to each database.
Screening and selection process
Two reviewers (JD, OS) independently screened articles in Covidence utilizing a two-stage screening process based on the eligibility criteria. In the first stage, articles were screened based on the title and abstract. Articles which did not meet the inclusion criteria were excluded. Disagreements were subsequently resolved via discussion until consensus was achieved. In the second stage, full-text screening of the included articles was independently performed by both reviewers to determine eligibility. Reasons for exclusion during this stage were documented. Disagreements were again resolved through discussion. A third reviewer (LB) was consulted in select cases when meeting the inclusion criteria was unclear.
Data extraction
Extracted information included study characteristics, participant characteristics, and attribute characteristics. Study characteristics included year of publication, methods used, country, time period of data collection, and sample size. Participant characteristics included gender/sex and non-medicinal vs. medicinal vs. dual use (both non-medicinal and medicinal) use of cannabis. Attribute characteristics deemed to be relevant to consumer choice, as well as a narrative summary of these characteristics, were also extracted.
Data synthesis
Because these attributes were investigated using a diverse range of methodologies and recruited a variety of study populations, there was no attempt to combine studies statistically. Only qualitative synthesis was completed. Where possible, data exploring differences in preferences among sub-groups of the population or between the legal and illegal market were highlighted.
Quality assessment
The Mixed Methods Appraisal Tool (MMAT) (Hong et al., 2018) was used to assess the quality of each of the included studies. The MMAT tool provides five critical appraisal questions for each of the five possible study design categories. Only two categories were required for appraisal in this systematic review. These were “qualitative” and “quantitative descriptive.” For qualitative studies, questions cover appropriate method, findings adequately derived from the data, interpretation substantiated by data, and coherence between qualitative sources. For quantitative descriptive studies, questions cover: the appropriateness of sampling, representativeness of sample, appropriate measure, overall risk of bias being low, and appropriateness of statistical analysis. Each question has three possible responses: yes, no, and cannot tell. Quality assessment was completed by two reviewers (JD, OS), and disagreements were resolved through discussion. Studies were not excluded based on not meeting a quality threshold, but rather quality assessment was considered in the interpretation of the findings.
Results
A total of 4839 titles and abstracts were screened after duplicates were removed. Ninety-six articles were eligible for full-text review; of these, 61 were excluded due to (1) no attributes of choice (n = 26), (2) not a research study (n = 15), (3) abstract only (n = 10), (4) duplicate study (n = 1), (5) duplication of data (n = 4), (6) unable to find text (n = 2), and other (n = 3). A total of 35 publications were included (Fig. 1, Table 2). Most were conducted within the USA (n = 25); five were carried out in Canada, six in other international locations, and one was of unknown location (three studies were conducted in more than one country). The most frequently examined attribute was price, with twenty-seven studies looking at some measure of the impact of price on choice. Most studies were conducted in a population where cannabis was not legalized for non-medical use (n = 19), some were conducted in legalized environments (n = 9), while other had unknown or mixed legalization status (n = 8). Only fourteen studies (Aston et al., 2019; Boehnke et al., 2019; Capler et al., 2017; Chait and Burke, 1994; Cole et al., 2008; Gilbert and DiVerdi, 2018; Goodman et al., 2019; Goudie et al., 2007; Halcoussis et al., 2017; Reinarman, 2009; Riley et al., 2020; Shi et al., 2019; Shukla, 2003; Wadsworth et al., 2019; Williams, 2004) explored non-price attributes, further categorized into the following themes: quality, route of administration, packaging, and product recommendations (Fig. 2).Author, year Method (hypothetical/revealed)* Attributes Population
Non-medical legalization statusSex (% female)
Mean age (SD)Time period Location Amlung, 2019 MPT (hypothetical) Price elasticitySubstitutability 289—adult cannabis users;Immediately before cannabis legalization 40.1%31.7 (9.9) November 2017 to February 2018 Hamilton, ON, Canada Amlung, 2019 MPT (hypothetical) Price elasticitySubstitutability 724—adult cannabis users (> 21 years);States where cannabis is legalized 52%34.13 (10.02) September to December 2017 Alaska, California, Colorado, Massachusetts, Maine, Nevada, Oregon, Washington, USA Aston, 2015 MPT (hypothetical) Price elasticity 99—frequent cannabis users (18–44 years); cannabis not legalized 37.4%21.4 (4.4) 2010–2011 Rhode Island, USA Aston, 2016 MPT (hypothetical) Price elasticity 83—frequent cannabis users (18–44 years); cannabis not legalized 34.9%21.6 (4.7) 2010–2011 USA Cole, 2008 MPT (hypothetical) Quality elasticity 80—polydrug users 18–44 years; cannabis not legalized Polydrug users: 36.3%21.0 (1.2)Cannabis users: 36.3%21.0 (1.2) 2006 England Collins, 2014 MPT (hypothetical) Price elasticity 59—young participants (18–25 years) who regularly used cannabis; cannabis not legalized 46%21.64 (1.98) Unknown Buffalo, USA Goudie, 2007 MPT (hypothetical) Income elasticityQuality 40—polysubstance users. 38—reported at least a single lifetime use of cannabis; cannabis not legalized Polysubstance users: 27.5%23.8 (4.9Cannabis users: unknown 2005 England Hindocha, 2017 MPT (hypothetical) Price elasticity 24—non-dependent cannabis and tobacco smokers (18–60 years); cannabis not legalized 50%24.46 (3.96) Unknown London, UK Nisbet, 1972 MPT (hypothetical) Price elasticity, expenditure elasticity 926 UCLA students(437 users of cannabis); cannabis not legalized Unknown Unknown Los Angeles, California, USA Patel, 2019 MPT (hypothetical) Price elasticity 749—adults with cannabis use in the previous 6 months; states where cannabis is legalized 53.2%37.7 (10.2) September to December 2017 USA Peters, 2017 MPT (hypothetical) Price elasticity 82—frequent cannabis users, 18 years or older (use 20+ days in past month); unknown legalization status 54.9%32.4 (8.8) Unknown USA Strickland, 2017 MPT (hypothetical) Price elasticity 78—frequent cannabis users, who consumed at least once in the previous 2 weeks, and at least 50 lifetime uses; unknown legalization status 51.6%30.2 (7.3) Unknown USA Strickland, 2019 MPT (hypothetical) Price elasticity 83—non-medical prescription opioid users > 18 years of age; unknown legalization status 63.9%34.0 (8.0)Cannabis users: unknown May to September 2018 USA Teeters, 2019 MPT (hypothetical) Price elasticity 132—college students who reported use on 4 or more days in past month; unknown legalization status 46.2%19.94 (3.23) 2014–2016 USA Vincent, 2017 MPT (hypothetical) Price elasticity 683—young adult (18–25 years) non-medicinal cannabis users; unknown legalization status 16%21.2 (2.2) Unknown USA Aston, 2019 Qualitative interviews (revealed) Route of administration 25—medical cannabis registration card holders; cannabis not legalized 60%47 (12) 2016 Rhode Island, USA Ben Lakhdar, 2016 Secondary analysis of data from the French National Identification System for Drugs and Other Substances (SINTES) and TREND system (both surveys) (Revealed) Price elasticity, quantity discount 268 cannabis users (249 for the elasticity calculation); cannabis not legalized 23.1%27 (7.234) 2005 Mainland France (Lyon, Marseille, Metz, Paris, Rennes, Toulouse) Caulkins, 2006 Secondary analysis of survey data from the National Household Survey on Drug Abuse (revealed) Price discounts National probability sample of the civilian noninstitutionalized population (> 12 years of age).8339 respondents reported using cannabis in the past 12 months; cannabis not legalized Unavailable 2001 USA Davis, 2016 Secondary analysisCrowdsource platform PriceOfWeed.com (revealed) Price elasticity 23,000 actual cannabis transactions where price, quantity, and quality are reported; mixed legalization status N/A Between September 2, 2010, and August 29, 2011 USA (excluding Alaska and Hawaii) Desimone, 2003 Secondary analysis of data from the National Household Surveys on Drug Abuse (revealed) Price elasticity 92,784 individuals aged 18 to 3943,147 individuals aged 12–17; cannabis not legalized Individuals aged 18–39: 50.9%29.05 (6.28)Individuals aged 12–17: 48.8%14.49 (1.68) 1990–1997 USA Halcoussis, 2017 Secondary analysis of crowdsourced data on prices (PriceOfWeed.com) (revealed) Price elasticity, quality 29,461 transactions; mixed legalization status N/A September 2010 through March 2012 USA Hansen, 2017 Administrative dataWashington State Liquor and Cannabis Board (revealed) Tax reform (i.e., price) Cannabis legalized N/A Washington, USA Reed, 2020 Qualitative interviews (revealed) Price, source 60 individuals who consumed cannabis; cannabis legalized 64%21.8 (2.53) 2014–2016 California, USA Reinarman, 2009 Survey and Interviews inAmsterdam andSan Francisco (revealed) Source, price, potency, accessibility Experienced users (at least 25 occasions in their life); cannabis not legalized (California), Decriminalized (Amsterdam) Amsterdam: 41%34.2 (standard deviation not reported)San Francisco: 47%37.1 (standard deviation not reported) Amsterdam 1995/1996San Francisco 1997 Amsterdam (216), San Francisco (266) Riley, 2020 Cross-sectional survey (revealed) Price elasticity, quantity discount, quality 1961 cannabis consumers; cannabis not legalized 27.2%Mean age not reported August and September 2017 South Africa Smart, 2017 Secondary analysis of administrative data from Washington State’s cannabis traceability system Price elasticity, potency on price, quantity discount A total of 44,482,176 million cannabis purchases, including 31,052,123 cannabis flower purchases; cannabis legalized N/A July 2014–September 2016 USA (Washington) Wadsworth, 2019 Survey - International Cannabis Policy Study (ICPS) (revealed) Quantity discount, source 1227 Canadians aged 16–65 years who reported purchasing dried cannabis in the past 12 months; immediately before cannabis legalization Unweighted sample: 48.8%Mean age not reportedWeighted sample: 39.8%Mean age not reported August–October 2018 Canada Boehnke, 2019 Survey (revealed) Cannabinoid content, cannabis strain, potency, administration routes, dispensary/friend recommendation, smell, visual properties, described effects, name Medical cannabis patients (≥ 18 years) with chronic pain (1321)715—medicinal only606—used both non-medicinally and medicinally (dual); in states with cannabis legalization 781, 59%49.8 (13.9) January and August 2018 USA and Canada Capler, 2017 Secondary analysis of survey dataCannabis Access for Medical Purposes Study(revealed) Quality, safety, availability, efficiency, cost, feeling respected, product type 445 Adult medical cannabis users (215 accessed from dispensaries; 230 from other sources); cannabis legalized for medical purposes 33% (dispensary 32%; non-dispensary 34%)39.3 (dispensary 45.5Non-dispensary 36.3) 2011–2012 Canada Chait, 1994 Choice blinded trials (two independent choice trial—with low (0.63% THC) and high (1.95% THC) potency cannabis) (revealed) Potency 12 volunteers judged to be healthy with no history of substance use disorder; cannabis not legalized 25%23 (standard deviation not reported) Unknown Chicago, USA Gilbert, 2018 Rating of 13 samples using standard sensory evaluation techniques with untrained consumers (revealed) Aroma 61 adults 21 years of age or older (current, former and non-users)Female: 46% (29)Age: 28.2 (8.4); cannabis is legalized 45.9%28.2 (8.4) Unknown Colorado, USA Goodman, 2019 An experimental choice taskPart of the online International Cannabis Policy Study (hypothetical) Packaging, warnings Participants aged 16–65 from Canada (n = 9987) and US states with “legal” (n = 7376) and “illegal” (n = 9682); Canada pre-cannabis legalization, USA a mix of legal and illegal states 61.5%44.5 (15.5) August–October 2018 Canada and USA Shi, 2019 Online cross-sectional survey with discrete choice experimentSurvey performed(hypothetical) THC concentration, CBD concentration, warning message, price (WTP) 2345 adults aged 21 and over (1186 past-year nonusers and 1159 past-year users); cannabis legalized in included states Past-year nonusers: 67.96%Mean age and standard deviation not reportedPast-year users: 68.51%Mean age and standard deviation not reported October 2017 USA (California, Colorado, Massachusetts, Nevada, Oregon, and Washington) Shukla, 2003 Key informant interviews(revealed) Illegal market considerations, availability, cost 51—purposeful sampling of individuals who made choice about cannabis use; cannabis not legalized 47%31.52 (standard deviation not reported) Unknown USA Williams, 2004 Australian National Drug Strategy’s Household Surveys (NDSHS)(revealed) Price, price elasticityLegal status 5468 observations from non-institutionalized population aged 14 years and older; cannabis not legalized (one region with decriminalized status) 54%Mean age and standard deviation not reported 1988–1998 Australia
Quality appraisal
Generally, studies were of appropriate quality to address the relevant study question. The design elements that were most difficult to assess were “representativeness of the sample” and “low risk of non-response.” Studies which collected data through crowdsourcing (e.g., Amazon Mechanical Turk) or participant self-identification, compared to administrative data or national surveys, are more likely to be subject to selection bias, non-response bias, and recall bias. In studies where sampling was appropriate and representative of the population, it is important to note that several studies had very narrow inclusion criteria, and therefore, the samples are only representative of that particular subset of the population. All but three studies used a quantitative methodology and were assessed using the “quantitative descriptive” set of questions. There was a wide range of methodologies used, some with more complex analyses; however, measures and statistics were generally appropriate for the specific methods used. A complete table outlining the results of the quality appraisal can be found in the online appendix.
Discussion
This systematic review sought to examine attributes that influenced cannabis consumers’ purchasing decisions. While price was the most researched attribute, other attributes like characteristics of quality, packaging, route of administration, and product recommendations also influenced purchase decisions. Media reports often claim that attributes such as high price, poor quality, limited supply, distance to stores, and inconvenient packaging of legal cannabis products are reasons why consumers continue to purchase from illegal sources (CBC News, 2019; Cecco, 2019; Esfandiari, 2019; Geraghty, 2019; Johnson, Glen et al., 2019; Lamers, 2019; Mazur, 2019; Tunney, 2019a; Turvill, 2020; Williams, 2019). However, as cannabis legalization is relatively recent, the number of studies which have explored these attributes is limited. In general, there is a dearth of evidence to support understanding of the role that any attributes play on cannabis consumer choice, outside of price.
The attribute of price constituted a majority of the reviewed literature. These studies were conducted mostly in populations where cannabis had not been legalized for non-medical use, while others were in populations where cannabis status was legalized, unknown, or mixed. Legalization creates a shift in the demand curve and therefore, it is important to consider this aspect when interpreting demand functions. It is also important to recognize that there is considerable heterogeneity in how the values for elasticity are derived across MPT studies. Some studies examine revealed choices by looking at transaction records, while others with examine stated choices using hypothetical scenarios. The prices and units of cannabis included varied from puffs to whole joints or grams of cannabis. Instructional vignettes and choice parameters used to describe the purchase decision varies greatly across studies which can impact demand (Aston and Meshesha, 2020). Finally, included studies used a variety of different demand equations; therefore, comparison of demand predictors cannot be compared directly to one another. The purpose of this study was not to examine the impact of price in detail. For a more in-depth interpretation of price and price elasticity measured through using MPT design, Aston et al. (2019) provides a comprehensive review.
When consumers were faced with a choice of different sources of cannabis offering the same product at different prices, people chose the product at the lower cost as shown in the study by Shi et al. (2019). In Canada, cannabis purchased from Health Canada licensed producers was associated with the greatest price per gram out of the examined sources, being over double the price per gram of illicit sources (Wadsworth et al., 2019). Quantity discount might also explain the large difference in price per gram between legal and illegal cannabis. It appears that the effect of quantity discount and a general lower price per gram of illegal cannabis may offer an explanation as to why the illicit market continues to thrive despite cannabis legalization.
Packaging also appeared to influence product selection, however studies that explored packaging did so through hypothetical questionnaires and focused solely on branding and warning messages (Shi et al., 2019; Wadsworth et al., 2019) and did not investigate legal cannabis’s oversized and wasteful packaging as described in media reports (Lamers, 2019). Given public pressure to be more environmentally conscious, excess packaging may have the potential to influence where consumers purchase cannabis. No studies have looked at the impact of packaging on real purchase decisions.
Media reports claim that legal cannabis is of lower quality than illegal cannabis (Turvill, 2020). However, there is insufficient evidence to support this claim as quality was either insufficiently defined or not examined in the studies reviewed here. Quality could be interpreted as any combination of label accuracy, potency, presence of contaminants or pesticides, curing process, ability to give desired effect, size, visual properties, and aroma (“How to buy good weed,” 2020). More research is needed to explore cannabis quality and how that is defined by consumers. One aspect of quality that perhaps does provide some insight into the strong illegal market is the higher potency of cannabis available on the illicit market (Mahamad et al., 2020). Generally speaking, medicinal users preferred high CBD content, while dual and non-medicinal users preferred high THC content (Shi et al., 2019).
Exploring gender differences that influence purchase decisions is an important consideration given that that cannabis use was more prevalent among males than females (CCSA, 2019). There also appeared to be sex-based physiological, behavioral, and neurobiological differences in cannabinoid effects, which may play a role in product selection (Fattore and Fratta, 2010). Although many of the studies examined included male and female participants, there were only a few areas where gender preferences were highlighted. Men tended to choose products with higher potency and preferred smoking or vaping, as compared to women who chose lower potency products and preferred topicals or tinctures (Boehnke et al., 2019). Men also tended to consider strains and the smell of the product when selecting cannabis to purchase (Boehnke et al., 2019). However, it is unknown if there are gender differences when considering other attributes in product selection, such as price, quality, packaging, and product recommendations. A recent scoping review reported on how gender norms influence patterns of cannabis use (Hemsing and Greaves, 2020). Further research on gender differences when choosing cannabis products is needed.
The goal of cannabis legalization in many jurisdictions is to protect public health through safety and quality regulations (ACT Government, 2020; Spithoff et al., 2015). However, cannabis is still purchased from the illegal market (Canada, 2020; George-Cosh, 2019), so a better understanding of the attributes that people consider when purchasing products will help inform the reasons for choosing between the illegal and legal markets. This study has provided a better understanding of these attributes; however, it also highlights that there are significant gaps in our knowledge in this area. A more thorough knowledge of cannabis consumer purchasing preferences can help policy makers refine the existing policies that will protect public health and safety while meeting the needs of consumers.
Limitations
There were several limitations in this systematic review. Given cannabis legalization is relatively recent and restricted to a few countries, the literature regarding this topic is limited and the number of studies exploring each attribute is scarce, especially in a post-legalization context. Although price was the most common attributed examined, many of the studies that examined aspects of price did so using distinct methodologies and data sources that should not be considered together. In studies that captured data on purchase history through a survey, prices were often exaggerated or recalled incorrectly, whereas administrative data captured actual purchase behavior. Studies using the MPT design captured data in an experimental setting, and the effect of price on purchases were considered more objective despite being a hypothetical measure.
Outside of price, it is difficult to draw clear conclusions on the influence of other attributes on purchase decisions. Many studies focused on only one, or a small number of factors, and therefore, very little is known about the relative importance of each. There was insufficient data on many of the attributes, including aroma/taste, described effects, product recommendations, provenance, product strain, stigma, product safety, and personal safety, thus limiting the ability to make any estimate on the degree to which these attributes influence choice. Heterogeneity among the study methods and samples also makes it difficult to generalize many of the findings. Several studies had very narrow inclusion criteria and were therefore only representative of that subset of the population, or the sample size was very small. For several of the studies that met our inclusion criteria, choice attributes were often not the primary outcome examined and therefore lacked detail regarding those choice attributes.
Preferences for cannabis products likely differ across consumers based on frequency of use. There were no studies that broke down findings based on consumer use frequency. Qualitative methods were used in only three studies, which limits the depth of understanding especially around non-price attributes. There was also a lack of youth perspectives when making purchase decisions and only a few studies identified gender influences on product choice.
Finally, the literature to date has mainly focused on choices for dried flower cannabis. However, attributes of purchase choices likely differ across product types. For example, visual appeal may be less important for a cannabis beverage purchase compared to dried flower. There is currently no research evidence that helps us to appreciate heterogeneity in choice behavior by product type or route of administration.
While this study is a thorough review of the available literature on consumer preferences for cannabis products, the limitations noted above prevent us from drawing any specific conclusions based on the data.
Future research
Future research is needed to develop a more thorough understanding of the non-price related attributes that people use when choosing cannabis products as well as the relationship between these various attributes. Factors cited by the media, such as distance to licensed stores, cannabis supply, product moisture, and bulky and wasteful packaging, lack evidence and remains to be studied. Quality of the product is poorly studied and has varied meanings, so research is needed to determine what quality means to people and how it influences purchase decisions. With the increasing use of cannabis use among youth (Canada, 2019), it will be important to explore the factors for their choices. Finally, an appreciation for potential heterogeneity among choices based on consumer characteristics (e.g., gender, reason for cannabis use, frequency of use) as well as type of cannabis product (e.g., dried flower, oil, edible) is needed.
Conclusions
This systematic review presents a summary of findings from current literature regarding attributes of choice when consumers purchase cannabis products. The majority of studies focused on price-related attributes whereas three studies contributed a large proportion of findings for non-price attributes. Demand is generally inelastic with respect to price, but the degree of elasticity varies by age, gender, and experience with cannabis. Preferences were greater for products with higher potency of either THC or CBD, but this also changed based on reason for use and gender. There is insufficient evidence to understand the true impact of other attributes on the choices of cannabis consumers and the relationship between attributes. Going forward, additional research will support a more thorough understanding of these attributes, which can offer a better explanation of consumers’ thoughts and opinions. This information will be useful for helping policy makers refine the existing policies to better support public health and safety and meet consumer needs. It can also offer insight for countries looking to legalize cannabis for either medicinal or non-medicinal use.
Supplementary Information
Untitled section
Abbreviations
- CBD
- Cannabidiol
- DACU
- Driving after cannabis use
- MMAT
- Mixed Methods Appraisal Tool
- MPT
- Marijuana Purchase Task
- THC
- Tetrahydrocannabinol
Acknowledgements
The authors would like to acknowledge Alison Farrell for her peer review of the search strategy used in this systematic review.
Funding
This study has been supported through a grant jointly given by the Canadian Centre for Substance use and Addiction and the Canadian Institutes for Health Research (Grant Number: 201910CPV). Neither organization played a role in conducting this research.
Availability of data and materials
Complete search strategy and detailed study tables are included in the supplementary appendix.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.