The authors thank Vincent Mercier for his valuable contribution to data analysis.
1Cannabis legalization in Canada was destined to produce a large implementation gap. It is a multifaceted and wicked policy issue (Oldfield et al., 2022, p. 440), whose complexity is only exacerbated by its moral connotation. By virtue of their historically constructed and perceived value-driven character, moral issues such as cannabis are contested and their regulation is a site for the expression of broader ideological fractures (Schwartz & Tatalovich, 2018). Policies that are moralized by political elites are prone to be discussed through symbolic appeals to fundamental conceptions of “right” and “wrong” for political gains (Ferraiolo, 2013; Kreitzer et al., 2019). In doing so, the significance of policy changes on moralized issues is magnified by ideologically charged perceptions of what those changes mean.
2Some features of the legalization process create additional challenges in the transition from formulation to implementation. First, ensuring public safety and providing a safe cannabis supply – the main stated aims of legalization (Task Force on Cannabis Legalization and Regulation, 2016) – can act as competing goals. Public safety is indeed achieved through strict control of the product, while replacing the legacy market with a safe and legal alternative is better achieved through efficient supply that activates consumer integration. Legalization in Canada further involves a wide range of actors at all levels of government (federal, provincial, municipal), as well as non-governmental and corporate actors. Coupled with a short timeline of around a year, these features of legalization led to an intense and complex policy exercise (Wesley & Murray, 2021, pp. 1081-1082), where tensions emerging from competing priorities were inevitable (Beauchesne, 2021). For instance, many provincial governments did not view the previous prohibitionary policy as obsolete and would not have legalized cannabis were it not under federal constraint. As a result, some provinces (Benoit, 2023; Benoit & Lévesque, 2022; Lévesque & Benoit, 2020), as well as many municipalities (Gagnon et al., 2022), used their powers to curb or limit the extent to which legalization took effect on their land. As a shared jurisdiction, cannabis policy therefore produced a multi-layered framework. In this framework, a complex web of agencies with different needs and priorities is tasked with its implementation. More broadly, legalization policies are transformative. Although an illegal market was well-established, legalization created a new formal economy, along with the extensive bureaucratic and industrial apparatus it requires. Policy transformations of that scope usually lead to the reorganization of resources (Zahariadis & Exadaktylos, 2016, pp. 64-65) and the redistribution of power among stakeholders (Galvin & Hacker, 2020).
3In the end, those complicating factors all paved the way for lingering policy ambiguities in the transition from formulation to implementation. But why exactly do these ambiguities appear? How do they shape the behavior of policy stakeholders? In this paper, we argue that, in order to manage the unique challenges of cannabis legalization, governments and central agencies rely on pre-emptive blame avoidance strategies. Ambiguities ensuing from the use of those strategies in turn enhance the responsibilities of implementers, while also limiting their capacity. Overall, this paper contributes to a better understanding of the role of blame avoidance in shaping implementation gaps. We proceed as follows. First, we review the literature on blame avoidance and blame avoidance strategies. Second, we overview our methods, cases and data. Then we present results in three sections, respectively tackling how the use of blame avoidance shapes the explicit objectives of legalization, its implicit necessities, and state capacity. We conclude by underlining the contributions of this paper, as well as suggesting avenues for future research.
4Although Weaver’s seminal “The politics of blame avoidance” (1986) has been cited more than 2,500 times, the concept of blame avoidance remains undertheorized (Hood, 2011a; König & Wenzelburger, 2014, p. 406). According to Hood (2011a, pp. 5-6), blame encompasses two fundamental dimensions: “some element of perceived and avoidable harm or loss – something is seen as being worse for some person or group than it could have been if matters had been handled differently,” and “some attribution of agency – that harm was avoidable because it was caused by acts of omission or commission by some identifiable individual or organization.” Blame avoidance occurs when decisionmakers attempt to avoid the harms that stem from agency attribution by reducing their exposure to blame in the first place (Zahariadis et al., 2020, p. 160).
5Scholarly accounts of blame avoidance are usually guided by economic voting theory, in which the behavior of politicians revolves around seeking reelection. In this approach, blame is evaluated through the prism of rewards and punishments (Pierson, 1994, p. 17), and blame avoidance strategies are used by blame takers to avoid being punished by the electorate (König & Wenzelburger, 2014). Recent research mirrors this game-theoretic focus, as the success of blame avoidance strategies is assessed through the government’s approval ratings (Vis, 2016), attitudes towards policy reforms (Wenzelburger & Hörisch, 2016), reactions to allegations of political misconduct (Schönhage & Geys, 2024), and public acceptance of external sanctions against their country (Schlipphak et al., 2023). In this view, blame attribution requires blame makers, including politicians from opposition parties, public advocacy groups, and the media, to hold those in power accountable in the public sphere (Hood, 2011a, p. 7; Klijn et al., 2022, p. 286). Given the important discursive dimension of blame games, then, extant research focuses on political communication contexts, such as debates over the formulation of policy, to understand blame attribution and avoidance (Hansson, 2024; Leong et al., 2023).
6In this paper, we move beyond policy formulation and explore how blame avoidance shapes policy implementation. Previous research convincingly shows that public officials have developed such a strong aversion to blame that avoidance has become a driving force of “the organization and operation of modern executive government, producing its own curious logic of administrative architecture and policy operation” (Hood, 2011a, pp. 5-6). Blame avoidance includes a preemptive aspect when the design of policies preventively eschews any perceived responsibility. This anticipatory form of blame avoidance differs from his reactive counterpart, which is driven by the actors’ perception of the risk of punishment rather than the policy’s actual risk (Hinterleitner, 2017, p. 243; Hinterleitner & Sager, 2017, p. 589; Vis, 2016, p. 123). Because anticipatory strategies are aimed at preventing blame attribution, policymakers cannot know for certain whether their use of strategies is necessary (Vis, 2016). To avoid drawing attention from blame makers, they can rely on subtle strategies that accentuate policy ambiguities by blurring the policy’s objectives, deflecting the accountability chain and deferring blame on to the more discrete implementation phase.
7Ambiguity characterizes policy problems that can be understood in multiple ways, and which result in contention over policy challenges and their solutions (Fowler, 2019, 2021; Zahariadis & Exadaktylos, 2016). While ambiguities are widely acknowledged as contributing to implementation gaps (Howlett et al., 2015; Matland, 1995; Taylor et al., 2021; Zahariadis & Exadaktylos, 2016), a perspective that also encompasses blame avoidance can shed light on some of the mechanisms that cultivate them. Nurturing, rather than solving, policy ambiguities shifts blame away from policymakers on to implementers. The chaotic context that implementers are forced to grapple with, and which lead to the dashing of “great expectations” from policymakers (Fowler, 2021; Pressman & Wildavsky, 1973), can thereby be considered a consequence of blame-avoidant policy design. Linking policy ambiguities to implementation gaps uncovers negative instances of blame avoidance, where ambiguous features of the policy design generate breaches that can later be exploited by blame-avoidant actors. Despite being highly consequential for policy outcomes, these invisible anticipatory dynamics remain under-researched (Hinterleitner, 2017, p. 251).
8Observing anticipatory blame avoidance is more challenging than its reactive counterpart (Hinterleitner & Sager, 2017, p. 602), since non-blame events are necessarily less visible than are blame events (Vis, 2016, p. 123). A way to observe it is through the perspective of blame takers’ subordinates, i.e., implementers to which policymakers delegate potentially blame-prone responsibilities. Much of the literature on principal-agent models and street-level bureaucracy underlines that bureaucrats seek to expand or protect their discretion, while superiors try to hold on to it. An approach centered on blame-avoiding policy implementation (Hinterleitner & Wittwer, 2023) contributes to explaining why politicians might deliberately surrender discretion, and why bureaucrats might contest an extension of discretionary power (Catney & Henneberry, 2012; Hinterleitner, 2017; Klijn et al., 2022).
9In this paper, we argue that governments and central agencies leverage the ambiguities of cannabis legalization to clear themselves from blame-prone responsibilities. Confronted by these ambiguities, implementers seek to minimize, mitigate and solve them. We distinguish between two sets of actors: on the one hand, federal and provincial governments as well as central agencies who have the power to design policies and the authority to distribute responsibilities and influence the accountability chain at the implementation stage; on the other hand, blame-taking subordinates, i.e. the intermediate and street-level implementation stakeholders who face design ambiguities and engage in coherence-making activities to cope with these (Mavrot, 2023). The latter are a diverse set, encompassing actors from various sectors (public health, police, municipalities, etc.), and private actors (cannabis producers, retailers, consultants, etc.). We make use of an already vast “catalog” of blame avoidance strategies (König & Wenzelburger, 2014, pp. 405-406) to explore when and how policymakers use preemptive blame avoidance. Hood’s (2011a) three-term typology of blame avoidance, which includes presentational, agency and policy strategies, is useful to that end.
10Presentational strategies aim at shaping public perceptions, notably by casting an issue in the best light possible to limit negative feedback (Zahariadis et al., 2020, p. 161). In cannabis policy, this can materialize as “defensively disproportionate regulation” (Hood, 2011a, p. 8) by emphasizing control over consumption or by targeting organized crime. Politicians can also accentuate the righteous safety objectives of legalization, while leaving out some of its implicit and blame-prone necessities such as the development of a legal cannabis industry or state revenues generated from cannabis sales. Conservative and paternalistic governance strategies still stand strong in the post-legalization era (Wheeldon & Heidt, 2022, p. 2). We therefore expect governments to avoid engaging in those implicit necessities rather than run the risk of being accused of trivializing cannabis consumption.
11Policy actors can also rely on agency strategies (Hood, 2011a) by delegating tasks that summon blame through various types of responsibility-sharing partnerships. In the multilevel context of a federal regime, previous studies suggest that politicians tend to shift blame on to other government levels, tolerate fuzzy institutional structures to dilute reputational threats, or develop multilevel blame avoidance strategies (Bache et al., 2015; Béland & Myles, 2012; Rodríguez R., 2022). Governments can also “pass the buck” to less accountable bodies such as independent regulatory agencies (Hinterleitner & Sager, 2017, p. 593), or to non-governmental and private organizations. In Canadian cannabis legalization, delegating ambiguous educational or retail mandates to lower-level public health bodies, or to private retailers, also creates opportunities for anticipatory blame avoidance. Market conditions are commonly used to justify unpopular decisions and poor service performance. The recourse to “don’t blame me, blame the market” rhetoric indeed allows policy actors to shift the accountability chain on to impersonal forces (Hood, 2011a, pp. 70, 83).
12A final avenue for avoiding blame is found in policy or operational strategies, which involves selecting policies that expose elected officials to as little risk of blame as possible (Hood, 2011a, pp. 20, 103). This strategy notably encompasses the shaping of administrative capacity, or the extent to which public organizations are enabled to fulfil their mandates (Skocpol & Finegold, 1982; Soifer, 2008). Administrative capacity is a central force in policy implementation (Fowler, 2021; Hill & Hupe, 2022). It includes cognitive abilities, such as the translation of ambiguous concepts into actions within complex implications, the reduction of uncertainty about policy tools, or the interpretation of programs consistent with broader policy goals (Fowler, 2021; Taylor et al., 2021). Without these capacities, ground-level implementers face multiple challenges that can lead to policy failure. A policy strategy for blame avoidance can be enacted through an abstinence-based approach, when “choosing not to operate or provide facilities or services (including record-keeping) that attract blame or have the potential to do so” (Hood, 2011a, p. 102). Not investing in state capacity can lower the visibility of a policy’s controversial aspects. This is an effective policy strategy since lowering visibility decreases the likelihood that the issue becomes politicized, and thereby blameworthy (Vis, 2016).
13This paper addresses the relationship between blame avoidance and policy ambiguities in the case of cannabis legalization in Canada. Legalization is an interesting policy for addressing this theoretical issue because, as hinted above, it involves multiple dimensions of state intervention at once. In the Canadian case, this is further magnified by the fact that legalization involves federal and provincial jurisdictional responsibilities and delegated regional and municipal responsibilities, while a substantial portion of implementation rests with private entities. This makes for a regulatory landscape that is multi-layered and broad in scope. Given this, we limit our analysis to the subcases of Quebec and Ontario. These two provinces were selected as they represent a contrast, the former having a decidedly state-oriented policy model and the latter being more market-oriented. Moreover, these subcases are the two largest provinces in the Canadian federation and are thus a significant focus for cannabis producers and federal regulators. In Quebec, successive governments from the Liberal and Coalition Avenir Quebec parties developed a model that sought to restrict consumption and limit cannabis’s for-profit opportunities. This was justified on public health grounds, most forcefully through framings focused on the safety of children and young adults (Lévesque, 2022). In Ontario, a government change from the Liberal to the Progressive-Conservative party, a few months prior to legalization, led to the development of a for-profit model in the name of fiscal responsibility. The Progressive-Conservative Ontario government nonetheless emphasized public safety as its primary objective (Ibid.).
14We make use of the multi-level setting of cannabis policy implementation in Canada to gain a better understanding of the tensions arising from policy ambiguities during implementation. We do so through internal comparison, where we rely on subcases – key issues, challenges and outcomes, some of which are extreme examples and others more representative of the general state of legalization in the federation – to produce evidence about the whole (Lange, 2013, p. 58). Using this evidence, we develop a constitutive argument (Pacewicz, 2022), i.e., an analytical description of the pathways that account for the emergence and persistence of policy ambiguities.
15We rely on two main data sources. First, we conducted two rounds of interviews with cannabis policy stakeholders in the public health, public administration, industry, and law enforcement sectors. In Summer 2020, we conducted a first round of loosely structured interviews with 30 participants across Canada to identify the major themes, concerns, and challenges of policy implementation. From these interviews, we developed a semi-structured guide which we administered from July 2021 to February 2022 to 29 cannabis policy stakeholders, i.e., actors with decision-making power over policy implementation in the governmental, administrative or corporate spheres, from Ontario (n=13), Quebec (n=12) and federal (n=5) instances. Interviews were conducted over Zoom and generally lasted between 60 and 120 minutes. Meetings were recorded and manually transcribed by an external firm. For every interview, a consent form was read and signed by the participant. We coded interviews in NVivo. We developed an initial coding grid and enriched it inductively through flexible coding (Deterding & Waters, 2021). This procedure allows for both a structured approach to data analysis, in which core themes present in the data are systematically picked up, and iterative recognition of sub-themes and exceptions that contribute to a nuanced handling of the data. Second, we conducted an institutional literature review. The review focused on publicly available documents retrieved from the websites of policy stakeholders such as wholesaling agencies, regulators, cannabis business associations, etc. It also included bills, regulations, and legislative debates, as well as publicly available documents (e.g., media pieces, scientific publications, policy briefs, etc.) referred to by interviewees.
16Data from our two rounds of interviews allow us to make claims about the implicit norms under which legalization operates in our cases by interrogating the meaning-making practices of participants (Lamont & Swidler, 2014). This enables for a better understanding of the dynamics of blame attribution and avoidance, as well as the points of contention between different types of stakeholders, and how these stakeholders in turn make sense and act upon these dynamics. As a complement, the use of institutional literature allows us to get a sense of the regulatory environments in which stakeholders operate, as well as to identify policy ambiguities. Finally, we use the institutional literature review to corroborate some of the statements by interviewees.
17Unlike U.S. states where commercialization and tax revenues have dominated decision-making (Rehm & Fischer, 2015, p. 541), cannabis legalization in Canada explicitly pursues a public health and safety agenda. Nonetheless, these objectives can be conflicting as they can lead either to restriction or promotion of access. Immediate ambiguities therefore emerge along the existing trade-offs between integrating illegal consumers into a new legal market and limiting or shaping, rather than bolstering, consumer demand for harmful products (Wesley & Murray, 2021). These ambiguities are further magnified by a federal regime that enables provinces to draft their own consumption and retail regulations based on these shared overarching goals. The moral dimension of cannabis policy, the ambiguities embedded in federal legalization objectives, and the jurisdiction of provinces over important aspects of policy altogether create opportunities for blame avoidance to emerge. Governments of Quebec and Ontario adapted differently to the challenging task of selling controversial products while shaping consumer demand in a public health perspective. In this section, we argue that they both preemptively avoided blame. Ontario did so by using agency strategies to delegate cannabis retail to external organizations, while Quebec relied on presentational strategies to emphasize public control over consumption.
18Ontario’s cannabis regulatory model was introduced by the Progressive-Conservative (OPC) government with laissez-faire and fiscal responsibility in mind. The OPC stated a clear preference for low state involvement and correlate administrative costs, as well as for fostering an environment where consumers were the final judges over products. As then Government House Leader Todd Smith summarized:
We concluded that a system of government-run stores was incapable of seriously competing with the illegal market and, in turn, would leave our communities more vulnerable and susceptible to the underground market. […] A private retail store model would allow Ontario businesses to participate in the retail system. Increased competition would offer consumers more access and better combat the illegal market (Legislative Assembly of Ontario, 2018).
19This symbolic stance regarding market competition and consumer choice hardly represents Ontario’s model, as the state is involved in almost every aspect of supply. The Alcohol and Gaming Commission of Ontario (AGCO) oversees retail stores, from licensing to regulatory compliance, and the Ontario Cannabis Store (OCS) wholesales cannabis from producers to retailers and acts as the sole online retailer. Despite this, since the main component of cannabis supply is private brick-and-mortar retail stores, consumers are more likely to interact with private than with state organizations as they navigate this system. The interplay between symbolic appeals to a market-based approach and relatively invisible areas of the state’s crucial role opens opportunities for blame avoidance for policymakers and regulators. Blame avoidance was performed mainly through an agency strategy, through which the provincial government reallocated decision-making to other actors.
20A key area of implementation where Ontario’s regulatory model allows for blame avoidance at the provincial level is geographical accessibility, both upstream and downstream. The sole restriction that Bill 36 imposed on store location is a 150-meter distance from schools. Otherwise, at face value, the legislation delivers on its promise of laissez-faire. However, Bill 36 included an opt-out clause for municipalities, leaving them the possibility of prohibiting cannabis sales on their land. Per the AGCO, around 18% of municipalities used this prohibition clause (AGCO, 2019). This delegation of power in restricting access to cannabis, which may have curtailed resistance to private supply from municipal associations (Lévesque, 2023), also resulted in a delegation of blame to municipalities for unequal geographical accessibility.
21This upstream shift of blame from the provincial government is coupled with downstream blame shifting from the state to private stakeholders. On the one hand, the accessibility of a legal supply is important for the success of legalization since it is likely to influence whether or not a consumer transitions to the legal market (Armstrong, 2021). In a private retail model, any lingering inequalities in access can be attributed to the market, i.e., they can be explained by the industry’s unwillingness to provide access to cannabis in remote communities. Five years after legalization, the Ontario market still presents major geographical inequalities. Available data suggests that more than half of cannabis stores are located in the 10 largest cities, with about one in four in Toronto (OCS, 2022). This leaves 706, out of 1,460 stores, for the remaining 434 municipalities of Ontario. Although there has been a notable decrease in regional discrepancies, the Northern region of Ontario is still disadvantaged in the current geographical distribution of stores: it has less than 1% of stores while having around 90% of total land mass and around 7% of total population. Market forces do have a responsibility for the current situation. This, however, obscures the fact that the provincial government could correct these inequalities through policy, and that – given proper means – the AGCO could alleviate the unequal geographical distribution of stores (e.g., by prioritizing licensing in communities with less access).
22On the other hand, the market is blamed for the over-accessibility of cannabis in certain areas. As participants from the public and private sectors highlighted, over-accessibility might eventually contribute to adverse health outcomes among users. Toronto was mentioned by many as an exceptionally problematic case in this regard. Again, containing the number of retailers can be undertaken by provincial governments. For example, in British Columbia, a limit of eight stores per license holder has been in effect since legalization (Government of British Columbia, 2018). Not only are there no such restrictions in Ontario, but the AGCO is hardly a hurdle for prospective licensees. Public notice periods on new store licenses last only 15 days, after which the agency decides whether or not to grant an authorization (AGCO, 2022). Participants working around the agency, in both public and private sectors, indicated that refusal upon negative public comments is very uncommon. One interviewee from the retail space suggested that, overall, the eligibility criteria for allowing licenses was not a strong barrier to retailers.
23The problem of over-accessibility should thus be considered as an instance of non-decision-making on the part of the Ontarian government and state agencies, just as much as it is the result of the unfolding of market forces. In effect, it hints at an oft-neglected aspect of cannabis policy implementation: due to its moral component, the political costs to policymakers of revising policies far outweigh the potential benefits, even when there is consensus on a problem among implementers. Overall, the development of a private cannabis supply system opened an opportunity for the Ontario government to pre-emptively shift blame to private retailers. Given that the Ontario government successfully performed a laissez-faire cannabis policy at the formulation stage, the costs of its inaction during implementation have been marginal.
24Quebec political authorities have been reluctant to cannabis legalization (Benoit & Lévesque, 2022; Lévesque & Benoit, 2020), and successive Liberal and Coalition Avenir Québec (CAQ) governments adopted bills that reflect this reluctance in important ways. It is showcased in comparatively strict regulations over consumption: home cultivation is prohibited, the legal age is 21 (vs. 18 for alcohol), and use is prohibited in all public spaces. It is also reflected in the retail strategy of the state-owned monopoly Société Québécoise du Cannabis (SQDC), which offers a relatively narrow range of products and has limited the rollout of storefronts since legalization.
25These choices reflect a “defensively disproportionate regulation strategy.” (Hood, 2011b, p. 8). By emphasizing control over cannabis consumption and by targeting organized crime, public authorities use symbolic cues to reassure the public, to show that elected officials are taking the issue seriously, and to sanction cannabis use as unacceptable behavior. This framing reproduces stigma surrounding cannabis and complicates implementation. For instance, restricting consumption sites may be a way for governments to signal that they are taking cannabis consumption seriously, but can in turn lead to important policy gaps: by “disallowing areas to use legally, the government paradoxically necessitates that people using a legal substance do so in a way that is considered illegal” (Wheeldon & Heidt, 2022, p. 10).
26Quebec policy stakeholders underline ambiguities resulting from the legislation’s performative aspect. For example, the promotion of abstinence among youth to ensure their safety has led to a ban on vaping products and strong restrictions on edibles because these are perceived as too attractive. Annual Quebec Cannabis Surveys, conducted since 2018, show that interest has indeed grown in these modes of consumption, especially among younger generations (ISQ, 2022). For a participant from the cannabis industry, however, strong restrictions over vaping and edibles are inconsistent with the objectives of substituting illegal supply with a safe and legal alternative:
Quebec regulations prohibit anything that is easily sold and that is in demand by removing anything that is gummies, chocolate, etc. – anything that is attractive, anything that may appear attractive to people under 21. All people who want to consume, ingest, or drink fashionable products must necessarily fall back on the black market. (Participant from cannabis industry, August 2021).
27Meanwhile, participants from health authorities understand the problem in an entirely opposite way:
Edibles were marketed quickly, I think, and drinks in particular. The objective is really to compete with the illegal market, but I have never seen an illegal retailer selling cannabis drinks. For me, this is a demand from industry. What they managed to gain was acceptance that these products present a lower risk behavior than is smoking pot. I agree with this, except that it is not true that pot smokers turn to drinkables. These are new consumers (August 2021).
28Product diversification is perceived by health authorities as an industry strategy to attract new consumers. Another participant from a health authority highlights the ambiguity that emerges from these contrasting views: “the principle of public health requires that we can capture the illicit market with a legal offer, I would say consistently but responsibly, and in a way that protects young people. Finding this balance remains a big challenge.” (August 2021).
29The abstinence approach to youth consumption has shown its limits. A conference of the National Institute of Public Health of Quebec on cannabis vaping in 2023 emphasized its harmful effects. The cannabinoids concentration in these products is generally higher than in smoked cannabis. Moreover, as few norms regulate the manufacturing of e-cigarettes, several devices are cheap and poorly cap the operating temperature, thereby increasing toxicological risks (Malboeuf, 2023). In a survey of adolescents aged 12-17, 66% had used cannabis vapes compared with only 20% in 2019 (Ibid.) Because they are underage, we can infer that illegal providers have both successfully captured the vaping market and reached the youth market. As vaping becomes increasingly popular among adult consumers as well, a considerable public health issue emerges, and the current prohibitionist approach would appear ill-suited to addressing it.
30For several stakeholders, Quebec’s regulations are so strict that they can only be weakened over time. The issues of legal age and product diversification are often raised to illustrate this point:
Participant
I do not disagree with progressive regulation with strict measures that will ease over time.
Interviewer
Progressive in the sense that you think adjustments are to come?
Participant
Yes, it is certain that there will be adjustments. For one, the chaos that some predicted with legalization did not happen. Two, many public health people say that, if you don’t go to the SQDC, you will go elsewhere, and it is much more harmful to take the uncontrolled product than a controlled product. Health rules, production rules are so strict that we have no choice but to have a product that is free of any bacteria, of any problem. (Participant from cannabis industry, September 2021).
31Participants from public health also agree that some characteristics of the Quebec model should be adjusted to give a better reflection of a public health approach. This is the case with the prohibition on home cultivation, since “with a public health objective, if you want to remove cannabis from a lucrative and commercial logic, for me, it is wise to allow consumers who want [to grow] to have four plants at home.” (Participant from public health, August 2021) This incrementalist approach to legalization is not shared by the government. Indeed, the home cultivation case rather reveals the government’s commitment to a strict approach. Since the Quebec ban contradicts federal law, this jurisdictional conflict was brought before the Supreme Court which ultimately ruled in favor of the province. This judicialization is itself a form of blame avoidance, where both federal and provincial governments preferred to settle this dispute through means that were not overtly political rather than by engaging in contentious negotiation. The provincial government’s position further exemplifies a defensively disproportionate regulation in performing a public health and safety approach to cannabis nevertheless inspired by a prohibitionist paradigm. Overall, this has led to ambiguities during implementation.
32Wesley and Murray (2021, p. 1098) note that governments in Canada designed policies that contributed to demarketing cannabis for public health purposes. This is most evident in federal restrictions on commercialization, which imposes neutral labelling and prohibits almost any form of advertisement. In this strict federal commercialization regime, state and private cannabis retailers alike are viewed as somewhat responsible for public health outcomes, and are frequently framed as providers rather than sellers (Wesley & Murray, 2021, p. 1099).
33These attempts to decommodify cannabis supply evacuate implicit necessities that stem from legalization under a private seed-to-sale regime. Whether retail is run by a private or state corporation, cannabis production, transformation and distribution remains entirely private in Canada. This generates an implicit necessity for the industry to be successful, and fosters an expectation from corporations that public agencies will account for industry’s barriers to development (Train & Snow, 2019, pp. 567-568). It also ties market performance to the explicit objective of providing a safe cannabis supply to consumers. A second necessity that is evacuated by these attempts at decommodification is that of state revenues from legalization. In this section, we argue that these two implicit necessities have been shaped by agency blame avoidance strategies, and that they deepen ambiguities between what is expected from public agencies and from private stakeholders.
- 1 Interestingly, whether cannabis production would be private and for-profit was subject to very litt (...)
34Some stakeholders view the cannabis industry as a necessary evil, while others welcome it as a potential driver of economic development. Regardless, the establishment of a private cannabis supply system across the federation means that legalization policies must pander to what numerous participants categorized as the “health” of the cannabis industry.1 Without conditions for it to thrive, suboptimal market outcomes could harm the objective of providing a safe cannabis supply and tarnish some of the economic advantages of legalizing cannabis in the first place: industry centralization, rising production prices and/or diminishing quality, bankruptcies, etc. These potential outcomes were feared by participants from both private and public sectors.
35In tension with this implicit necessity of a thriving cannabis industry are some of the overt choices made in favor of decommodifying cannabis supply. Some participants underlined a tension between the difficulties of doing business in a low-margin industry and a federal regime that contributes to this reality by imposing stringent and costly regulations. Fixed capital costs in the cannabis business are very high and appear early in the process, as cannabis license applicants must already have acquired a production facility before being approved. Federal fees for doing business in the cannabis space are also uniquely high compared with other agricultural products (see Justice Canada, 2022). Participants further noted that regulatory compliance (i.e., security, surveillance) is an overwhelming part of their overall production cost, reaching 50% in some cases. One unanimously despised element of current regulations among industry participants is the 10% excise tax imposed on cannabis production, viewed as “prohibitive […] from a consumer point of view” (Participant from industry, September 2021). Some further argued that the excise tax is impeding the goal of competing with the illegal market:
So why are we doing that? It is intentionally designed to hurt the industry, an industry that we are at the same time relying upon to succeed in order to displace the illicit market. So it is counterproductive to be slapping all sorts of duties and taxes in a way that is not going to allow the industry to effectively compete with the market that you are saying you want to displace because it is associated with, you know, contaminated products, organized crime, violence, all of these other bad things. So it is, again, we went into this from the perspective of, like, “We do not like it and we do not want to promote it, but we are going to kind of hold our nose and allow a legal source to exist.” But we have got to get past that now. We have got to get past that and have a paradigm shift to – it is in everybody’s interests for this industry to succeed. (Participant from a consulting firm, December 2021).
36While governmental review reports celebrate the federal framework’s ability to provide a competitive environment for industry to thrive in (Government of Canada, 2022), participants from industry were overwhelmingly frustrated by existing regulatory costs.
37In Quebec, industry participants also criticized the continued stigmatization of cannabis, in part fueled by the government. As a professional who participated in provincial policy formulation in 2017-2018 mentioned, “we had agreed to consider cannabis as a tobacco product and to apply the guidelines of the [World Health Organization] regarding tobacco control, which means that the government should not in any way aid or favor the industry or enter into communication with them.” (September 2021) The objective was to prevent corporate lobbying from influencing the content of the law. The industry is concerned that this watchword spilled over on to implementation, with two main consequences. On the one hand, there are no communication channels between corporate and government stakeholders. On the other hand, the cannabis industry is formally excluded from programs offered to the otherwise highly subsidized agricultural sector. For one cannabis producer, whose partners are more traditional agricultural producers, this exclusion was “nonsense for real producers.” (October 2021) This appears even more unjustified because support programs exist for the alcohol industry. Overall, agency strategies of blame avoidance are used by both federal and provincial decision-making bodies to create a context in which they can claim merit for the successful commercialization of legal cannabis but completely eschew responsibility for the industry’s hardship. This institutionalized sidelining of an industry that already faces a challenging economic context minimizes its ability to thrive.
38One of the inevitable by-products of legalization was the development of new services that are naturally tied to new state expenses. Depending on each province’s framework, organizations had to be built, infrastructures had to be developed, and public servants had to be hired. Provincial governments across the federation were keen to emphasize that legalization should be a low-cost operation. This was done in various ways, from full disengagement from the cannabis supply chain in Saskatchewan to the “lean” managerial approach of the state-run SQDC in Quebec. Governments also mobilized presentational strategies by avoiding framing cannabis legalization as an opportunity for state revenue, which – according to some participants – would have led to public backlash and criticism (Participant from the public sector, July 2020). Meanwhile, governments expect revenues from legal cannabis to cover some of the costs of policy implementation. For participants, this expectation of revenues led to short-term thinking about the economic benefits of legalization. By the same token, public agencies fell into a revenue-seeking logic:
I think the government has an objective of raising funds from the cannabis industry, through direct taxes, sales taxes, corporate taxes, income taxes, etc. However, by focusing so much on direct taxes, we are perhaps not always considering the net benefit of everything else, the income tax paid by cannabis employees, of the economic benefits. (Participant from industry, January 2021).
39How do public agencies negotiate financial viability and revenue-seeking? In Ontario, the taboo around state revenues has led the OCS to act in a fashion that some term predatory. In addition to being the sole provincial online retailer, the OCS wholesales cannabis in the province, thus acting as the middleman between licensed producers and retail stores. From March 2020, the OCS relied on a value-based mark-up approach, by which their own perception of market value determined product mark-ups. The OCS claimed that this approach was based on best business practices and was more flexible (OCS, 2021, p. 23). According to many participants, however, this strategy was used to maximize the agency’s revenue at the expense of licensed producers, especially smaller ones who can ill afford to trim their margins. This contributed to bitter sentiments towards the OCS, with some hypothesizing that this behavior was caused by the necessity of generating revenue:
Through their own legislation, the Ontario Cannabis Store is set up to be a direct competitor to the retailers [...] they also have their own interests with regard to their own sales because they sell directly to Ontarians through the mail. They have that in mind whenever they are talking policy with us because it is very important for them to be able to hit their revenue targets and do their projections appropriately so that they can tell the government how much money they are likely to bring in as a Crown corporation. Like any Crown corporation, there is pressure on you to not just meet your financial targets from last year but to exceed them. (Participant from the public sector, August 2021).
40The official, renewed, mandate of the OCS supports this view. In its 2021-2022 mandate letter to the Chair of the OCS, Rod Phillips, ex-Ontario Minister of Finance, states, “As part of the government of Ontario, I expect all agencies to act in the best interest of Ontario by being efficient, effective, and providing value for money.” (Phillips, 2021). Overall, the mandate’s orientation towards revenue generation is helpful in understanding the agency’s behavior. The OCS’s 2021-2024 business plan also reflects these priorities (OCS, 2021). These revenue-seeking practices were addressed in a December 2021 report of the Auditor General (Office of the Auditor General of Ontario, 2021). This report exposed the agency’s lack of transparency and accountability and called for changes in how it engages with producers. Since then, the OCS has established a fixed mark-up system with publicly available data (Ontario Cannabis Store, 2023).
41Pre-emptive blame avoidance helps in understanding the reluctance of state agencies to address these implicit necessities transparently. Both these implicit necessities require exposing themselves to criticism by the public: in one case, by loosening restrictions on private companies, and in the other by assuming state revenues as an integral part of legalization. As long as these implicit necessities remain unaddressed, they are bound to produce tensions with the explicit objectives of legalization.
42One participant from public health noted that, since Canada is the first large, industrialized country to legalize cannabis for recreational purposes, “we did not have the benefit of learning from everyone else.” (November 2021) The uncertainties of this novel policy called for developing unique and thorough administrative capacities. So far, this call seems to have gone unanswered. The limited capacity of public agencies to command cannabis policy has led to implementation gaps, from which they have disinvested in blame-avoidant fashion by using operational strategies, especially through an abstinence-based approach. This, in turn, has consolidated policy ambiguities. As it pertains to state capacity, abstinent blame avoidance strategies can be used for two specific aspects of legalization: first, by not committing enough resources to support the highly regulated legal industry; second, by not investing in educational capacities.
43Health Canada regulations were thought by industry participants to be incredibly complex and demanding, especially those related to security and surveillance. However, as some suggested, the agency’s capacity to follow through on those regulations is very limited: “Health Canada charges back to the industry for the cost of regulating them but does not meet anywhere near the service standards that they are obligated to do” (Participant from industry, September 2021).
44This limited capacity is evidenced in the license approval process. For example, a producer pointed out that, for some of his licenses, it had taken a year-and-a-half before obtaining all the necessary security clearances. Signs of limited capacity were also shown after having obtained all approvals. A participant working for a licensed producer noted that it had only received about a quarter of the planned visits from Health Canada since it was licensed. Limited capacity was further showcased in the handling of promotion regulations. Very strict marketing rules are apparently only weakly followed through by Health Canada:
Promotion [regulations] says you cannot do anything – you cannot do anything – to promote cannabis or cannabis accessories or services related to cannabis unless it falls within one of the very narrowly construed exceptions. And because it is so strict that it actually leads to some absurd results that it has just been very lax in its enforcement and because it has been lax, somewhat lax, in its enforcement than everybody starts to take advantage of that and starts pushing the boundaries. […] It really makes no sense to the point where it starts to just get ignored because it makes no sense (Participant from consultancy sector, we underline, December 2021).
45In parallel, a recurrent theme in discussions with industry participants was the lack of commitment from Health Canada to their own standards. In other words, it seems that, in many respects, the agency has refused to settle regulatory questions that are unclear or that fall into grey areas:
[I]t has been sometimes very difficult to get any answers out of Health Canada [and they] will say, ‘Get legal advice’. And I will say, ‘I am the legal advisor. And I do not, like I do not know. What do you mean by this’? You know, the industry has been crying out for more detailed guidance, more, you know, prescriptions of forms for documents that we have to submit that, you know, we are not really sure exactly what level of detail needs to be included. It would be nice to have, you know, prescribed forms. It would be nice to have a more detailed guidance document from Health Canada on promotion. ‘What do you mean by reasonable steps to ensure something cannot be accessed by a minor? Like what does that actually mean to you’? (Participant from the consulting industry, we underline, December 2021).
46By avoiding stating and disseminating clear rules regarding politically sensitive issues (for instance, what exactly constitutes a product attractive to young people), Health Canada leaves the industry to fend for itself and face criticism in the event of behavior deemed illegitimate. Disengagement from the state is also evident in research and development. Several participants from industry hope that Health Canada or its provincial counterparts would engage in research on harm reduction practices, or on the potential of cannabis as a substitution drug. One industry participant considers “that there is not only a lack of commitment, but a disengagement of governments to conduct and finance research. The industry cannot begin. If it is only industry that conducts the research, you and I know that it will receive little or no recognition from our authorities.” (September 2021).
47The Cannabis Act explicitly set out to enhance public awareness of the health risks associated with cannabis use. Legalization created an opportunity for a more open dialogue about the effects of cannabis and the ways of alleviating some of its risks without the stigma associated with illegal behavior. Despite being legalized, cannabis remains a moralized issue in Canada and some public health participants expressed deep concern over the potential trivialization of the substance:
By being exposed to the context, people are now used to it. We must ensure that legalization is not an object of trivialization. People are more used to it now, it’s less and less exclusive, less and less surprising. People get used to it. Morals are evolving. We have the challenge of keeping this subject topical, contemporary, and that it does not become trivialized among the population. (Participant from public health, August 2021).
48According to this viewpoint, the education aspect of legalization is limited regarding the dangers of cannabis use. There is still shared apprehension among public health stakeholders that talking about cannabis means promoting its use. In different settings in Canada, this leads to an abstinence blame avoidance strategy, consisting of choosing not to provide education campaigns and services that could attract blame. By limiting dissemination of public-health and fact-based prevention, public authorities avoid the risk of being accused of encouraging consumption and trivializing cannabis.
49Instead, authorities strictly enforce advertising restrictions enshrined in Section 17 of the Cannabis Act (Cannabis Act, 2018, p. 18), which prohibits all retailers from marketing the product in any way that “could be appealing to young persons” or that “evokes a positive or negative emotion or image of a way of life such as one that includes glamour, recreation, excitement, vitality, risk, or daring.” In practice, this is used by public authorities to avoid investing in prevention strategies, as a participant from an Ontario regional public health unit exposed:
Participant
We cannot go into high schools with a harm reduction model. Even though we know that the abstinence model is ineffective for sexual health or substance use, we and the school board are mandated to go in with abstinence model and that makes it a bit more challenging.
Interviewer
Are you talking about abstinence in high school?
Participant
Yeah. [Schools] have to because they are not allowed to use a harm reduction approach or suggest ways to decrease your harms because that would go against the policies and regulations around that piece of the puzzle (December 2021).
50It also can be used to avoid providing information which is nevertheless important for responsible use. When asked about the information that a customer wondering how to cook cannabis recipes might receive in-store at an SQDC in Quebec, a public services manager answered:
We tell them that they can find that on Google and that we are not here to give recipes. On the other hand, they are told to pay attention to the dosage. We always try to move things towards health protection. As it happens, there is CBD and THC inside. There are other things the counsellor needs to pay attention to. But he should not dwell too much on the medical aspect [because SQDC provides only non-medical cannabis]. (December 2021).
51Referring to Google and avoiding addressing medical issues hardly hits public-health and fact-based prevention targets. Wesley & Murray (2021, p. 1088) report that an SQDC public servant declined to participate in their cannabis marketing study by responding that “by law, we’re not allowed to do any kind of promotion in any way, shape or form, including discussing about our brand in a way that could be subjective or relate it to marketing efforts.” The advertising ban is in this case instrumentalized to avoid a discussion of the major issues of cannabis use, which could draw blame should it lead to an increase in consumption. This abstinence approach is criticized by the cannabis industry. As a Canadian cannabis consultant put it, people curious about cannabis have “a hard time accessing a trusted source of education about it. Government is not in the business of promoting cannabis use so they are not providing the education, but industry is not allowed to educate, so you have a void.” (September 2021).
52In this paper, we argue that governments and central agencies rely on pre-emptive strategies to avoid blame in the implementation of cannabis legalization in Canada. Ambiguities that ensue from the use of such strategies fall on to lower-level agents. Overall, cannabis policy implementation has been characterized by lingering ambiguities on the objectives and necessities of legalization as well as by limited state capacity to address implementation gaps.
53The findings of this study contribute to scholarship on policy implementation in several ways. Extant literature indicates that controversial decisions based on ambiguous objectives, and which bring together interdependent stakeholders with disparate visions, are particularly conducive to implementation gaps (Hill & Hupe, 2022; Pressman & Wildavsky, 1973). We explored this assumption further by highlighting how attempts by decision-makers to avoid blame shapes not only policy design but also implementation in significant ways. Our analytical focus on blame avoidance strategies provides insights into mechanisms that create and sustain ambiguities in implementation. Notably, through the reference to presentational, agency and policy blame avoidance strategies, we uncovered contradictory behavior from public authorities. On the one hand, authorities performatively enact restrictive components of the legislation and frame legalization as a major public health and safety issue. On the other hand, they invest little in administrative capacities to ensure that legalization unfolds according to its stated objectives or to adapt to emerging challenges. This leads to implementation gaps that remain invisible since policy stakeholders are reluctant to reopen debates on cannabis policy. As one industry participant noted, “After all, it is not a sexy topic. It has become part of our customs – we don’t bother, everything is fine. Why should anyone put this back on the agenda?” (August 2021).
54This paper also contributes to studies on drug policy in two ways. First, in our case study, strict regulations generated pervasive policy effects, especially in instances where they were not reflected in sufficient administrative capacity to surveil and sanction undesirable corporate behavior. Second, previous studies built on strong theoretical assumptions about distinct regulation models (Pardo, 2014; Ritter et al., 2016; Rogeberg et al., 2018). Our study of two contrasting models highlights many similarities in their implementation as a consequence of blame avoidance. Similarities arose, especially in the tense relations between the cannabis industry and provincial authorities, as well as in the reluctance of authorities to invest in education strategies. These common challenges, albeit very different policies, invite scholars to move beyond common assumptions about regulation models, as they might overstate the practical relevance of formal differences in design.
55Despite its contributions, this study presents several limitations and leaves some questions for future research. First, while the theoretical insights about the relationship between blame avoidance and implementation may well hold, the interpretive analysis presented here limits our ability to generalize empirical evidence in terms of other policy or national contexts. Second, the moral connotation of cannabis legalization may restrict applications of the findings to similarly moralized policies such as abortion, gun control, or LGBTQ+ rights. Future studies should thus undertake similar questions with policies that are thought to be demoralized (Kreitzer et al., 2019), or engage in comparison between moralized and demoralized policies to validate findings across these policy contexts. Third, by its multilevel institutional design (federal, provincial, regional, municipal, private) and simultaneous relevance for multiple policy areas (health, economic development, public safety, agriculture, etc.), cannabis legalization presents a variety of opportunities for stakeholders to shift blame either across levels or across areas at the implementation stage. This perhaps makes pre-emptive blame avoidance more likely than in other policy contexts. Future studies ought to investigate how blame avoidance shapes implementation in contexts where opportunities to eschew blame are limited, whether this influences the type of strategies used by blame takers, and whether this facilitates the work of blame makers in highlighting policy failures.