The PoliTics of evaluaTing cannabis RegulaTion in uRuguay1

Evaluation of the impacts of recent cannabis regulation policies includes important political dimensions and should not be reduced either to binary verdicts of success versus failure, or to a unidimensional assessment criteria. Based on the Uruguayan case, the article discusses the complexity of evaluating the results of cannabis regulation, assessing, among other aspects, the political dynamics surrounding implementation. In the case presented, the predominance of concerns about crime and public security was a decisive factor.


IntroductIon
In recent years, the consensus around the international drug control regime has faced major challenges, with a growing debate over alternatives to the so-called prohibitionist paradigm, which "modulated contemporary understandings of psychoactive substances when it established arbitrary limits for legal/positive and illegal/negative drug use" (Fiore, 2012a, p. 9). Institutionally, the most significant changes have occurred in the Americas, related to the regulation of the production and distribution of cannabis. 2 432393/2016-1, and by the Plano de Incentivo à Pesquisa (Research Incentive Plan, PIPEq) of the Pontífica Universidade Católica de São Paulo (PUC-SP), process no. 6529/2018-1.
[2] Since the 1970s, some countries have adopted alternatives to prohibitionism, although these cases have been rare, and usually less extensive. Notable examples include the de facto legalization that took place in the Netherlands, a process that enabled the emergence of the coffee shop model, as well as Spain's flexible legislation, which grants personal cultivation and the opening of cannabis clubs.
[3] For the purposes of this article, the designation "social use" marks a distinction in relation to the medical use of cannabis. Other terms are also widely used for the same purpose, such as "adult use" and "recreational use" (the best known), though we believe these to be less precise.
These regulatory initiatives run counter to the international conventions of the United Nations (un), which set the global agenda on drug control. According to these multilateral fora, the production, distribution and use of cannabis and other drugs -the use of which is considered intrinsically harmful to the individual and to society -should be prohibited, except for medical or scientific purposes (un, 1961, article 4).
Since the 1970s, using a range of legal mechanisms and regulatory norms, some countries have withdrawn possession of cannabis for personal use from the criminal sphere, or even, as in the case of the Netherlands, loosened sanctions for small-scale trade (Eastwood;Fox;Rosmarin, 2016). However, it was in 2012 that the Uruguayan government took the unprecedented step of presenting Act 19.172, which proposed the regulation of importation, production, acquisition, storage, marketing and distribution of cannabis and its derivatives for social use. 3 The bill was approved by the legislature the following year. In the United States, also in 2012, the states of Colorado and Washington established similar legislation by means of state-wide referenda. Since then, thirteen more states and the federal capital have legalized the social use of cannabis in their jurisdictions (medicinal use is permitted, under different rules, in 36 states and 4 territories). In 2018, Canada passed a national law, becoming the first G20 country (the group of the most developed countries in the world) to fully regulate the cannabis market. Via different processes and at varying paces, other countries are moving in the same direction. A good example of this changing state of affairs is the decision by the Mexican Supreme Court, in 2018, which declared the prohibition of cannabis unconstitutional and the approval of national regulatory legislation by Mexico' s Senate in November of 2020.
These cases make it seem increasingly clear that cannabis will gradually lose its illegal status in further countries. From a historical perspective, this represents a profound shift in the way states relate to the phenomenon of drug use and an unprecedented rupture with the prohibitionist paradigm that for a century had made extreme prohibition hegemonic on a global scale. Taking this context into account, this article focuses on a specific aspect of these changes: the political debates surrounding the evaluation of new drug policies. Such debates come about at the ambiguous intersection between, on the one hand, the production and dissemination of data and scientific analysis and, on the other, disputes over social values and state priorities in this area.
The Uruguayan case is taken as a subject for reflection because of its originality, duration -the seven years the legislation has been in place constitute the longest ever period that a comprehensive policy of [4] "Trade-off" is a term used mainly in economics to refer to a decision-making process in which one seeks the balance between two desired situations. In the case of drug policy, such situations tend to gain prominence and importance. cannabis regulation has been active -, and the development of control and supervision mechanisms. The article seeks to demonstrate that the political justification for passing the new law, based on themes of security and the fight against crime, tends to restrict debates on the criteria for evaluating this policy. Considering Uruguay' s geographical and historical proximity to Brazil, and notwithstanding its particularities, the article assumes that the case can provide the national debate on drug policies with important insights.
Our argument is that any drug policy should seek to expand evaluation criteria and metrics precisely because they involve political dilemmas and trade-offs. 4 As such, the topic of drugs is part of a political debate where values and priorities are decisive. The density and rationality of the debate depends on scientific evidence, but in no way ends there.
The selection of sources and data for the research arose from the authors' participation in the congress Cinco Años de Regulación del Cannabis (Five Years of Cannabis Regulation), which took place in December 2018, at the Universidad de la República, in Montevideo. The event brought together Uruguayan researchers dedicated to this topic, as well as international guests. Accessing the papers presented and their bibliographies, interacting with the researchers (with whom we spoke informally), attending the debates, all this helped us gather information on different aspects of Uruguayan drug policy. The approach that initially sparked our interest was that of Rogeberg and colleagues (2018), based on the multi-criteria decision method. In the course of the research, however, other studies focusing on the evaluation of drug policy were incorporated into and came to shape our approach. Some sources were particularly relevant for data collection, such as reports from the Instituto de Regulación y Control del Cannabis (Institute of Cannabis Regulation and Control, ircca) and from the Ministry of the Interior of Uruguay, along with evaluations from the Monitor Cannabis, a research institute linked to the Universidad de La República. We also drew on Uruguayan newspapers to understand local political dynamics, and reports by non-governmental organizations dedicated to evaluating the regulation of cannabis in the country.
The article has six sections, including this introduction. In the second section, we present some peculiarities of the drug policy debate. In the third section, we summarize the process of cannabis regulation in Uruguay, and then, in the fourth, we analyze the centrality that the issue of security had in this process. In the fifth section, we discuss the importance of expanding the criteria and metrics for evaluating drug policies. Finally, in the sixth section, we conclude the article, highlighting the political nature of the drug regulation debate.
[5] This does not mean that debates around drug policy only began in the last quarter of the twentieth century. However, there they clearly intensified during this period, evidenced by a huge growth in the specialized literature.
[6] With regard to this point, Caulkins, Kilmer and Kleiman (2016) propose that removing formal barriers to the consumption of cannabis, in contexts of regulated use, would be important for strengthening informal social norms capable of protecting those who are potentially more vulnerable to its excessive use. Fiore (2012b) has also argued that criminalization is a significant obstacle to social controls on drug use.

drug polIcy AssumptIons
The monitoring and evaluation of public policies has acquired such significance in contemporary states that it has come to constitute a specific sub-field of the social sciences. Different means of evaluation are developed using broad methodological toolkits, ranging from econometric analysis to ethnographic investigation.
Evaluation of drug policy, as in the other areas, must take into account the diversity in both nature and scale of the data and criteria analyzed. This makes it possible to overcome the kind of one-dimensional interrogation that tends to guide the debate, usually conveyed by the press in questions such as, "Does prohibition work?", "Which drug policy can solve the problem of violence?", "Is legalization the best policy to prevent young people from using drugs?".
This one-dimensional conception is related to the way the prohibitionist paradigm situates the state vis-à-vis the issue of drug use, elevating it as the sovereign arbiter of whether certain psychoactive substances are permitted while ignoring several other possible regulatory models. These approaches continue to impede the development of more complex understandings of what a drug policy means, even after forty years of intense political and academic debate. 5 In fact, it must be recognized that drug policies encompass both the global regimes that underpin national laws on a set of psychoactive substances and state prevention, care and treatment systems across different levels of government. While acknowledging the breadth of issues falling under the definition of drug policy, this article focuses on the regimes that organize national -or sub-national, in cases like the usa -legislation on cannabis.
To begin with, two considerations on drug policy must be taken into account. The first is to emphasize once again that the main theme of the drug policy debate is the role that the state plays in response to the issue of drug use. Evidently, the meanings ascribed to drugs and use thereof go far beyond and even exist independently of the state, since historically they far precede modern states. Drug policies are related to, but do not encompass the multiple informal norms and controls -beyond the scope of states action -that regulates and gives symbolic meaning to substances and their uses. For example, religiously motivated aversion to the consumption of a psychoactive substance among part of the population is not necessarily expressed in the form of state legislation, norms and practices, but nonetheless strongly influences them (Fiore, 2012a). 6 The second consideration is the peculiarity of the drug market in relation to other economic activities. This involves recognizing, in various different ways, that drugs are not, as stated by Babor and colleagues (2010), "ordinary commodities". Substances that, when consumed, alter perceptions and sensations are part of complex systems of meanings, moral valuations and economic materializations, hence requiring, when taken as commodities, specific state policies corresponding to the extent and intensity of their impacts, whether associated with their consumption or production and circulation.

the new uruguAyAn polIcy for regulAtIng cAnnAbIs
In 2013, Uruguay became the first signatory state to the un's international drug control regime to legalize the production, distribution and use of cannabis for non-medical and non-scientific purposes. In addition, unlike the governments of Colorado and Washington, in the United States, which in the previous year had established regulation through popular referenda, the Uruguayan law was proposed by the executive and approved by votes in both legislative houses.
The singular feature of the Uruguayan law is the strict control over the cannabis market by the state. The state is responsible for controlling the production carried out by licensed companies and for authorizing pharmacies interested in retailing the drug. Only five strains are officially produced with standardized levels of potency -measured by proportion of Tetrahydrocannabinol (thc), the main active ingredient present in cannabis. To buy them at authorized pharmacies, users must register with the regulatory body specially created for the implementation and operation of the policy, the ircca. In addition to access through pharmacies, cannabis users can individually grow up to six female plants (the gender that produces thc) at homes or join a cannabis club. These clubs can grow crops of up to 99 plants, depending on the number of members, and are prohibited from making a profit. Whether production is individual or collective, registration with the ircca is mandatory and only Uruguayan citizens or permanent residents are eligible. It is prohibited to drive vehicles under the influence of cannabis, carry out advertising of any kind, consume cannabis in public places or sell any products derived from cannabis, even over the internet (Pardo, 2014). In short, according to Walsh and Ramsey (2015), the Uruguayan measures are quite "state-centric" and aim to prevent commercial exploitation of the drug in any of its production stages, as well as limiting its use.
The state body responsible for evaluating the regulatory policy is the Unidad de Monitoreo y Evaluación (Monitoring and Evaluation Unit), linked to the División de Sustancias Controladas (Division of Controlled Substances) of the Ministry of Public Health, which started operating in 2015. Other non-governmental entities, including transnational ones, have also conducted evaluations of cannabis regu-lation policy in the country, including academic groups (like Monitor Cannabis), foreign foundations (the Friedrich-Ebert-Stiftung and Open Society Foundations) and advocacy and research organizations (Advocacy for Human Rights in the Americas).
Over its six years of existence, there have been a number of challenges to the regulation: the population' s initial rejection to the change, which reached almost 70% in 2013; the disagreement on the part of un agencies and some of their member states; domestic criticisms of the mandatory government registration of consumers and cannabis clubs; restrictions by the international banking system on the inflow of financial values from the cannabis trade; and, more recently, an increase in homicides and robberies, whose relationship with cannabis or other drug trafficking is controversial and will be addressed below.
There are still criticisms about the inability of government production to meet demand and the slowness in producing data on the policy. In any case, the regulatory policy remains in full operation, having even survived the defeat, in 2019, of the political force that implemented it, the Frente Amplio (Broad Front) coalition, ending a 15-year period of leftist government in the country.

securIty And polItIcs: the uruguAyAn cAnnAbIs InItIAtIve
The regulation of cannabis in Uruguay enacted by Law 19.172, its background and subsequent evolution show the political centrality of the theme of security in contrast to other aspects of drug policy. It is no accident that this very issue has most heavily shaped the international development of drug control throughout the 20th century.
Social and political youth organizations advocating for the right to cannabis use, and others concerned with public health issues linked to harm reduction, 7 have always had a strong presence in the debate on regulation in Uruguay, which dates back to the 1980s and gained greater prominence during the 2000s. The influence of users' organizations is evident in the spread of "World Cannabis Marches", which played an important role in challenging the prejudice and stereotypes associated with users, as well as regularly placing the issue on the public agenda. Likewise, it must be recognized that the first proposals to reformulate cannabis control in the Uruguayan political system, in early 2012, were the result of an articulation of social movements through two law proposals related to the guarantee of rights: the release of growers from prison and the legalization of personal cultivation and cannabis clubs (Collazo;Robaina, 2015, p. 4).
This scenario changed dramatically during the presidency of Pepe Mujica, as the security office of the federal executive assumed greater influence, launching its Estrategia por la vida y la convivencia (Strat- [7] The concept of "harm reduction" is the subject of political and academic controversy; a possible definition is the argument that care and attention should be provided to those who do not want to or cannot stop using one or more drugs. Historically, the harm reduction approach has been consolidated since the 1980s, in the context of controlling the spread of HIV/AIDS and viral hepatitis. egy for life and coexistence) in 2012. According to Repetto (2014, p. 126), this shift implied the reassessment of the cannabis issue largely as one of security, distancing itself from perspectives related to human rights and public health. The Estrategia por la vida y la convivencia was a government proposal whose stated aim was to respond to social violence, identified "as a priority problem for the population" (Uruguay, 2012). 8 In fact, even though it ranks among the safest countries in Latin America, with low crime and homicide rates, perceptions of insecurity were widespread in Uruguay. The increase in crack consumption, theft, robberies and homicides, which reached a record level in 2012, helps to explain such perceptions (Walsh; Ramsey, 2015). The Estrategia por la vida y la convivencia was launched in the wake of three tragic and violent events that occurred in quick succession in Montevideo and were quickly linked to drugs, fostering strong popular support for "mano dura" (hardline) response by the state against crime (Repetto, 2014, p. 125).
In this context, the proposal to regulate cannabis began to be justified by the argument that it would remove part of the funding from criminal groups by narrowing the illegal drugs market, of which cannabis is one of the main products. It was expected that violence would decrease as a result, in particular homicides related to the drug trade (Baudean, 2017a, p. 10). In the 2012 draft, the justification sent to Congress highlighted the bankruptcy of the war on drugs; the cost of repressive measures on the public purse, including the overcrowding of prisons and workload created for the judicial system; and, worst of all, the monopoly of the lucrative cannabis market by criminal organizations, which undermined democratic values, encouraged money laundering and fuelled various forms of violence (Garat, 2013, p. 11). President Mujica, on several occasions, reaffirmed this position in his proposals for regulating cannabis, pointing out that "it is pursuing the battle against drug trafficking" and "it is not a law that promotes addiction. It is a way to combat the black market economy" (Bowater, 2014). This justification resonated with public opinion. A survey conducted by the Latin American Public Opinion Project, in 2014, found that the objective of the law most cited by Uruguayans was to prevent crime and combat drug trafficking, accounting for 37% of responses (Walsh; Ramsey, 2015, p. 15).
The Uruguayan government, for its part, also presumed that the proposal would win more support if it was linked to public security rather than promoting public health or the rights of users. According to Repetto (2014, p. 133), in proposing the security narrative as a justification for the regulation of cannabis, the government sought to guarantee the support of the public and the neutrality of the most right-wing opposition parties. The argument would also guarantee some international legitimacy for a measure whose stated aim was to strengthen the state' s sovereignty vis-à-vis criminal groups.
Internationally, this line of argument had already been adopted in the United Nations Office on Drugs and Crime (unodc) 2009 report, "Organized Crime and its Threat to Security", which stated: Global drug control efforts have had a dramatic unintended consequence: a criminal market of surprising proportions. Criminal organizations have the power to destabilize society and governments. The illicit drug business is worth billions of dollars a year, part of which is used to corrupt government officials and poison economies. (Apud Rolles et al., 2016) 9 This explains unodc' s positive assessment of Uruguay' s regulation on this specific theme, as expressed in its "Statement on the bill regulating the production, sale and consumption of marijuana in Uruguay" (unodc, 2013).
The proposal to regulate the drug market as a way to tackle crime has strengthened over the past few years, and has been the most commonly cited argument for abandoning the prohibitionist paradigm, at least in the case of cannabis. A recent report by the Global Drug Policy Commission, for example, states that "regaining control of this market and taking power away from organized crime remains a central motivation for regulation" (Global Commission, 2018).
In the Uruguay' s pioneering agenda, this argument was crystallized in article 4 of Law 19.172, which defines its objective to Protect the country's inhabitants from the risks arising from the illegal sale and trafficking of drugs, seeking, through State intervention, to tackle the devastating social, economic and health consequences of the problematic use of psychoactive substances, as well as reducing the incidence of drug trafficking and organized crime. (Uruguay, 2013, emphasis added) Thus, its main objective is to sever the link between cannabis users and drug trafficking, reducing the revenue and power of criminal groups in Uruguayan society. Interestingly, "the security arguments previously used to ban cannabis have been used, since its reformulation, to promote regulation" (Repetto, 2014, p. 133).
Electing for public security as the main argument for implementing cannabis regulation had consequences for the evaluation of its results: some of the expectations created seem to have been unreasonable, especially in the short term (Baudean, 2017a). Critics of the model, whether in Uruguay or in other countries, have, for example, used the increase in homicide rates in the country to point out its ineffec- [9] According to the authors, this text appeared on the UNODC web page in 2009, but was taken down in late 2015.
tiveness in reducing violence linked to drug trafficking. When he was still as a pre-candidate, current president Luís Lacalle Pou affirmed that "the regulation of cannabis has totally failed in fighting drug trafficking", since "today we have mafias killing in plain sight while the government is tolerant" (Radio Uruguay, 2019). Still, his government has not imposed further restrictions; on the contrary, it has considered expanding the regulations, such as permitting retail trade for tourists (El Observador, 2020).
The evidence cited to support such criticisms is the increase in the homicide rate, which in 2018 was 11.8 per 100,000 inhabitants -the fourth highest in South America, behind Brazil, Venezuela and Colombia (Melgar, 2019). Two years before the cannabis regulation was introduced, in 2011, that rate was 5.8. In other words, homicides rose by 103% over seven years. If we compare it to the year the regulation was passed, 2013, when the homicide rate stood at 7.6, the increase was 55.2%. When the data are disaggregated according to the supposed motivation of homicides, the current highest percentage results from the conflict between criminals, which doubled between 2012 and 2018 (from 29%, in 2012, to 58%, in 2018). This is significantly higher than other causes, such as domestic violence and robbery (Uruguay, 2018).
Even if we accept the perspective -in our view mistaken -that evaluations of drug policy should be restricted to public security impacts, it is important to take a more careful approach to what occurred in Uruguay after the regulation policy was implemented. The pressure for immediate results encourages hasty and politically biased interpretations that lack proper methodological foundations. This is what tends to occur when data is presented through a "before and after" framing, as was also observed in the cases of Colorado and Washington, the first two states to regulate cannabis in the us (see Makin et al., 2018).
To respond to this debate centered on the issue of security, which received regular media attention (Búsqueda, 2018), the government defined four groups of indicators related to security to evaluate the national regulation of cannabis: the illegal market, criminal violence, human security and coexistence, and perceptions of insecurity. According to Baudean (2017a), the elaboration of such indicators was based on five assumptions: 1. If the regulation is successful, the illegal cannabis market will shrink in terms of the number of consumers and of the drug traded volume. 2. If the illegal cannabis market declines, the presence and activity of drug trafficking will be weakened because cannabis is an important part of that business.
3. If the presence and activity of drug trafficking decreases in the long run, systemic homicides linked to that market will also fall. 4. In the short term, it is possible that changes in the market due to regulation (a reorganization of actors in response to shrinking businesses opportunities and struggles for dominance over territories) will produce an intensification of conflicts between trafficking groups. Thus, an increase in systemic homicides (among members of organized gangs) is to be expected. 5. Other aspects of security and social coexistence will be affected by the way the regulation impacts on the market for illegal drugs. In the short term, an increase in perceptions of insecurity is to be expected if the criminal conflict does, in fact, intensify.
In the long run, an improvement in perceptions of security are expected as a result of an eventual reduction in violence associated with drug trafficking.
This planning anchored in a logical chain demands a consistent production of data and objective analysis, without which any statement tends to rest on a high degree of speculation. The problem, however, is that, according to Baudean (2017b), there is no single organization that centralizes the collection and systematization of such data, a role that would originally have been the responsibility of the Unidad Especializada en Evaluación y Monitoreo. University researchers and independent institutes dedicated to the topic often complain about the reluctance of the ircca and the National Drug Board to share the little data produced by the government, which has raised concerns that authorities are trying to exercise control over information in order to monopolize evaluation of the regulation policy (see Hudak;Walsh, 2018).
Other important Uruguayan security bodies, such as the Ministry of the Interior and the judiciary, do not consistently produce data to support policy evaluation. This omission has a significant impact on analysis of security impacts, especially with regard to homicides. There is no precise data on the relationship between homicides and drug use or trafficking, which means researchers are limited to working with secondary data produced by civil society organizations, raising questions about their reliability. In the specific case of homicides, Monitor Cannabis evaluators have opted to present the evolution of homicides of individuals who had a criminal record, particularly those that are drug-related. They also analyzed the evolution of the data in the category "homicide due to criminal conflict", the classification used by the Ministry of the Interior, and data on the regional distribution of the crimes since 2012. Even though it may indicate important trends, this combination is insufficient to establish the relationship between cannabis regulation and the evolution of different types of homicides, since data is not disaggregated according to the type of drug within each category of law violation (Baudean, 2017b).
A second aspect of the dilemmas surrounding evaluation of the relationship between cannabis regulation and public security concerns the knock-on and the reciprocal impacts of the different drug markets, among which cannabis is only one and has very particular characteristics. According to Caulkins, Kilmer and Kleiman (2016, p. 124), illegal cannabis markets tend to be much less violent than those of illegal synthetic drugs, such as cocaine/crack and methamphetamines, or of opiates, such as heroin. This is due, among other factors, to their economic dynamics, as the production, distribution and consumption of cannabis does not normally involve sufficiently high-value transactions to support violent armed groups and, therefore, drive high homicide rates. Furthermore, part of the production is local and structured around a smaller and less complex production chain. In this sense, it is a comparatively cheap drug, which, in order to become economically viable, demands a high volume of sales and competition with non-professional, small-scale dealers (Werse; Bernard, 2016). It is true that such assumptions need to be questioned, since they are largely based on the functioning of the cannabis market in rich countries, especially in Western Europe, the United States and Canada. Studies of the Latin American context may reveal different dynamics, and may even vary significantly within the region.
However, these points partially contradict the idea that there is a highly structured articulation between these illegal markets, since one of the main explanations for the increase in homicides rests on the claim of "strangling the illegal drug market". According to this argument, offered by both research institutes and the government, the increase in homicides is due to criminal groups adjusting to a shrinking market (Búsqueda, 2018). In an interview with El País, researcher Marcos Baudean stated that the cannabis consumer market in Uruguay is estimated at 40 million dollars, of which 10 million has already shifted into the legal sector (Martinez, 2018). If this is true, analysts should still expect a huge increase in criminal violence in the country, since the success of the policy depends on the further expansion of the legal cannabis market in the coming years. Up to July 2019, it was estimated that the legal market represented only 20% of the total market, that is, only one in five citizens over eighteen years of age who declared having used cannabis at least once in the last twelve months claimed to have obtained it via a regulated, non-medical route. The illegal market therefore still serves 80% of drug users in the country (ircca, 2019, p. 3).
The reorganization of the illegal market and of the groups operating in it were also accompanied by a geographical reorganization of criminal violence in the capital, Montevideo. The city has the highest homicide rates in the country (11 per 100,000 inhabitants, compared to a national average of 8.1), concentrated mainly in the city' s poorest and most peripheral neighborhoods (Martinez, 2018). According to Baudean (2017b), it is possible to speculate that the regulation of cannabis benefited prosperous areas of the capital, at the same time as it increased the concentration of sale points for illicit drugs in the peripheries of the metropolitan area. Such urban spaces have, in the last twenty years, experienced increasing residential segregation and socioeconomic isolation, another point normally ignored in evaluations of Uruguayan drug policy.
Evaluation of cannabis regulation also demands attention be paid to temporal variables and other intervening factors. If it is true that the violence increased after the creation of the legal market, it must also be noted that this trend had already occurred over prior years. Figure 1 shows that homicides in the country have risen, albeit inconsistently, since at least 2006, and peaked in 2012, a year before cannabis regulation was approved.
Furthermore, the steady and more pronounced increase after 2013 must be weighed against the different stages of policy implementation. While non-commercial forms of access -self-cultivation and cannabis associations -were already permitted in the early years of the law, purchase from pharmacies did not actually begin until July  2017 10 (ircca, 2019, p. 5). In other words, how can a constant evolution in homicide rates, with a peak in 2012, be explained if regulation started in 2013 and widespread access to the legal market only occurred four years later?
Finally, evaluation of the security issue must take into account the fact that criminal logics have not disappeared in the context of cannabis regulation. According to Baudean (2017b), some analysts have defended the idea that Uruguay has introduced an exception policy for cannabis within a wider system of prohibition, which can be qualified as a third way. This is because the state continues to criminalize the entire production and distribution chain that does not exactly adhere to the established rules. This position is stated in anti-crime legislation passed in 2018. Here, illegal trafficking of cannabis is a crime that carries penalties equivalent to that of other illegal drugs. In July 2020, in the context of this general orientation towards security concerns, among which drug trafficking has a central place, the country' s Senate approved the Ley de Urgente Consideración ("Emergency Law"), proposed by the current conservative government of Lacalle Pou. This package of legislative reforms covering diverse areas is also a response by the current government to the 2014 reform of the Criminal Procedural Code, which it is considered to be weak and responsible for increasing impunity and, therefore, criminal activity. According to the President, it is a "popular, just and necessary" package of laws (El País, 2020). Two measures stand out with regard to the discussion on drugs: it institutes a mandatory prison sentence of at least 12 months for activities regarded as drug trafficking; and it restricts temporary release from prison for those who have served at least two thirds of sentences for drug trafficking sentences (as is the case for rape, murder and kidnapping) (Uruguay, 2020).
Therefore, even if we restrict the evaluation of the impacts of drug regulation to security issues, it is essential to consider a wider scope of arguments and possible responses to the problem. A report by the Global Drug Policy Commission, for example, recommends articulating the fight against organized crime and the strengthening of law enforcement with reforms in the police and an improved accountability of police abuse and corruption (Global Commission, 2018). In addition, planning for the legalization of cannabis and other drugs elsewhere must pay special attention to the types of sanctions that will be imposed on those who do not follow the implemented rules, especially those that normally fall on the most vulnerable parts of the population, whether based on socioeconomic or ethnic and racial factors. This is especially sensitive in cases like the Uruguayan model, where the state is at the center of cannabis production and distribution.

A broAder evAluAtIon perspectIve
Security has been very prominent in the debate over cannabis regulation in Uruguay, but that does not mean that other issues have been completely overlooked. The discussion takes place not only in the political and media spheres, but also among experts and activists. Regarding the latter pair, we would like to highlight the diversity, both thematic and disciplinary, that characterized the aforementioned congress Cinco Años de Regulación del Cannabis, held in Montevideo. The event brought together a large number of Uruguayan researchers specialized on the topic, as well as international guests. On the agenda, many papers discussed issues that went beyond the topic of security. They addressed the prevalence and patterns of cannabis use, the origin and types of cannabis available, how the system of accessing the drug operates, and how users access health care, to name just a few.
This broad scope reflects the international research agenda on drug policy. For example, the report published in The Lancet, "Public health and international drug policy", suggests the need to evaluate drug policies on multiple fronts, with an emphasis on health, human rights and development, in dialogue with Sustainable Development Goals (sdgs) formulated by the United Nations in 2015 (Csete et al., 2016). More recently, the European Monitoring Center for Drugs and Drug Addiction (emcdda), a European Union agency, released a technical report assessing changes in cannabis policies in the Americas. Echoing the logic proposed by The Lancet, the document includes themes ranging from prevalence of use and tax policies to impacts on the criminal justice and health systems (emcdda, 2020). According to Collins (2016, p. 9), placing drug policies under the rubric of development is to return to approaches that prevailed before the "War on Drugs" reshaped the debate in the 1970s. This widening of analytical criteria represents an a priori questioning of the prohibitionist paradigm, whose objectives diverge from those associated with the sustainable development goals. At the same time, it places the onus on policymakers adopting alternative models, as in the case of Uruguay' s cannabis regulations, to incorporate broader social and economic criteria into their evaluations, such as impacts on public health and inequality reduction.
As this indicates, there is a growing interest in interdisciplinary monitoring and evaluation strategies. In general, these approaches seek to identify key outcomes by surveying costs and benefits across different areas, such as health, economics, culture and politics (Collins, 2016, p. 13). Among the different methodologies adopted, studies that seek to measure social costs with an economic approach, by making use of econometric tools, stand out in particular. Gonçalves [11] A recent example of use of counterfactuals for evaluating Uruguayan drugs policy is the analysis of its impacts on prevalence of use by adolescents, carried out by Laqueur and colleagues (2020). Here, the Chilean data were treated in order to allow comparison with a "control country" that did not regulate cannabis. The results indicate that, to date, there have been no significant changes. and colleagues (2014), for example, evaluated Portuguese policywhich is based on the decriminalization of drug possession for personal use -and concluded, among other things, that the country was able to cut costs by reducing the need to mobilize policing, legal and carceral apparatuses. Another methodological possibility, known as "Cost Benefit Analysis", was applied by Shanahan and Cyrenne (2019) to econometrically measure the scale of the expected balance of regulatory cannabis policies in Canada.
Also from an econometric perspective, Ritter (2009) has sought to evaluate drug policy based on the harm associated with drug use, dividing this into two types: "individual" and "social", or, more directly, between the damage that consumers are subject to and those that potentially affect others. Using this same approach,  has proposed a comparison between the prohibitionist and strict regulation regimes of cannabis, pointing out that the latter causes less damage overall. Similar conclusions are reached by Rehm and Fischer (2015) in their review of the relevant literature, though in this case, going beyond estimates of economic damage alone to include minimizing damage to individual health and maximizing public health.
One notable challenge to such types of evaluation is the inevitable spirit of experimentation of policies being implemented, which makes unlikely the goal of foreseeing positive and negative consequences (Kleiman;Ziskind, 2014, p. 78). Drug use is an especially dynamic phenomenon and it is therefore imperative that the institutional design of policies is endowed with considerable adaptive and dynamic capacity so that permanent adjustments can be made during their implementation (Caulkins, 2007). To this end, it is essential that monitoring and evaluation mechanisms are provided for in the policy development phase.
The collection and, especially, appropriate production of data are fundamental to the drug policy evaluation process. But a prior challenge is to define exactly what is being evaluated and how the counterfactuals 11 of that evaluation will be tested. In the absence of this, arriving at causal inferences about political changes is even more difficult. Some states that have experienced these changes, such as Canada, have identified the need to collect baseline data at the time the policy is introduced in order to measure their changes over time. In the case of the United States, some monitoring policies have created control groups, which is fundamental to understanding impacts on particular variables, such as prevalence of use (emcdda, 2020, p. 43). However, even if data properly constructed and collected, it is still necessary to understand the broader context in which policies are implemented. This means, for example, that the impact on policing, its costs and how it operates must be assessed differently in jurisdictions that, prior to regulation of the cannabis market, did or did not criminalize drug users (emcdda, 2020, p. 44).
Given the novelty of cannabis regulation policies, such concerns have not yet been reflected in most monitoring and evaluation strategies. Studies like Rogeberg and colleagues' (2018) suggest, however, a type of grounded approach that is likely to gain more prominence in the coming years. The emphasis here is on the challenges posed by the inherent trade-offs involved in drug policy, although there is little consensus about what they are. The general orientation is to make the best decisions on the cost-benefit ratio of choices related to the regulatory model, assuming that in "balancing trade-offs, the 'best' policy is the one that results in the best 'overall' bundle of outcomes" (Rogeberg, 2018, p. 153). To achieve this, they "need to balance the harms of use against negative consequences of restrictive policies and the pleasures and benefits that the majority of users may claim to experience" (Rogeberg et al., 2018, p. 144).
Together with a wider group of specialists, Rogeberg and colleagues (2018) proposed that the evaluation of specific policies for alcohol and cannabis should be based on a wide range of criteria. It offers a set of 27 criteria, divided into eight thematic axes and which are not restricted to the potential damage linked to the use of the substance, also encompassing wider impacts of policies, such as those associated with the potential consequences of the market regulation process. These are shown in Table 1.
The importance of the evaluation criteria proposed by Rogeberg and colleagues (2018) lies much more in broadening the political perspective assumed than in the specific metrics, since measuring each of them presents enormous methodological challenges and can only be conceived in a particular context. For example, how can we objectively evaluate the promotion of family cohesion juxtaposed with impeding the influence of the drug industry on government? Better defined criteria can be found in the field of public health, such as those proposed by Fischer and colleagues (2019), who list ten key measures for monitoring and evaluating cannabis regulatory policies based on the scientific consensus on harm associated with drug use. In any case, what the proposal by Rogeberg and colleagues (2018) helps identifying is that the evaluation of drug policies is strengthened when it is not restricted to a few factors and more comprehensive criteria can be debated, even if their nature is variable and not objectively measurable. The inclusion of topics such as the promotion of human rights and social and economic development offers a more democratic perspective on the debate about regulatory models (Csete et al., 2016). As a result, this more inclusive evaluation format enables the involvement of different social actors with multiple kinds of expertise. In this article, we have discussed some of the dilemmas surrounding the evaluation of emerging cannabis regulation policies in the Americas, which have the potential to spread globally. These initiatives represent a break with the prohibitionist paradigm and demand new criteria and metrics capable of guiding the design, implementation and adaptation of these progressive policies.
While reflecting on the Uruguayan case, we have demonstrated how the association of the policy with security and crime reduction objectives was primarily a political strategy that responded to popular demands. This tended to limit the debate on evaluating of cannabis regulations, restricting the criteria considered for analysis and encouraging misinterpretations about the impacts of the policy on public security. The attempts to legitimize cannabis regulation based on its supposed public security benefits, as those made by different political actors in the Americas, may prove problematic in the medium term and risk the continuity of such policies in case the desired results are not achieved. Once again, the importance of expanding the themes involved in this debate is reinforced, in order to guarantee its political viability.
Uruguayan civil society and academic researchers have already understood this and have been working to broaden the terms of the debate, diversifying criteria, reorienting objectives and challenging government approaches to data production and dissemination from an interdisciplinary perspective. This is a positive aspect of new drug regulation initiatives being maintained over time: they enable social learning about possible ways of updating the policy and its objectives.
In addition, the diversification of evaluation criteria is part of a growing process of legitimizing the field of study in drug policy as a relevant scientific object in different areas of knowledge.
Year after year, the public debate on drugs has gained density and drawn on richer sets of data and evidence, distancing itself from the opinionated and moralistic character that marked these debates throughout the twentieth century.
Following the inexorable trend of evidence-based public policies, the field of drug policy, in its contemporary feature, can no longer afford to ignore scientific data. However, it is epistemologically incorrect and strategically misguided to ignore the political nature of the drug policy debate. Redundancy is common in debates conducted around a prohibitionism versus anti-prohibitionism dichotomy, in which actors emphasize the scientific evidence that best supports their arguments and attribute ideological motivations to those who oppose them (Fiore, 2018).
Such a priori disqualification rests on the mistaken assumption that it is not possible to sustain different political positions in this debate based on the same data. Thus, if it is true that in the absence of evidence arguments are limited to general guiding principles -for example, that "citizenship should include the right to use certain substances" or "drugs take away human autonomy" -it is not possible for drug policies to ignore positions that are guided by collective moral values, such as freedom, health and social cohesion, for example.
Even if we accept the pragmatic argument that drug policy should be evaluated by its results and not by the intentions driving it (Mejia;Restrepo, 2014, p. 32), the differences and inequalities that underpin which results are expected should make us permanently attentive to the political values that drive decisions. Generic and non-questionable objectives in a minimally reasonable debate, such as reducing violence and reducing damage to public health, may take on contradictory and even antagonistic meanings when seen from different social perspectives, for example, between poor and rich, young and elderly, women and men, white and black, or religious and atheist. However well established the data and evidence may be, they do not, by themselves, determine the correct path to be taken in drug policy, but only guide interpretations that enable social groups to conduct a well-grounded political debate. Thus, we agree with Rogeberg (2015, p. 348), for whom "if different groups support different policies because they emphasize different concerns, this is a disagreement over normative questions that the research cannot answer". As such, it seems that the political agendas that lie behind drug policies are decisive to the architecture of the new regulations and also influence the criteria used to evaluate them.