Cannabis Policy, Implementation and Outcomes

01 June 2003

This report examines what is known about the effects of policies regarding the possession and use of cannabis. Such policies continue to be subject to debate in most if not all European countries. Different governments have made different policy decisions, varying from explicit toleration (but not full legalisation) to strict prohibition. Policymaking would be served by insight in the relationship between different cannabis policies and their outcomes, such as prevalence of cannabis use and social consequences for cannabis users and for society as a whole.

application-pdfDownload the report (PDF)

As the impact of policy is greatly dependent upon its implementation, it is worthwhile to study not just formal policy but also cannabis policy as implemented in practice. Interest in such a study has come from a joint initiative of the Health Ministers of Belgium, Germany, France, the Netherlands and Switzerland. The study reported here was made possible by a research grant from the Dutch Ministry of Health, Welfare and Sports. The current report describes the result of this study.

Discussion and recommendations

In this concluding chapter, we first summarise what we have learned and then formulate recommendations for going forward towards an ultimate goal of cannabis policy that is feasible and has the likelihood of achieving the goals set forth by policy makers.

7.1 What have we learned?

In our review, we first looked at what is known about the relationship between formal cannabis policy and cannabis policy as implemented, and then at the question whether implemented policy affects the prevalence of cannabis use and what the consequences for both individual cannabis users and society are.

Formal cannabis policy and cannabis policy as implemented. Chapter 4 shows that there are real variations in formal policy amongst nations, and amongst regions within nations that are structured along federal lines (e.g., the US, Australia, Germany). Moreover, any notion of coherent implementation of policy at the local level is optimistic; not only is there a sizable gap between formal policy and policy as implemented in a number of countries/regions, but there is no evidence that any country or region has achieved uniform implementation in its several jurisdictions.

The studies examined show that there are various factors that contribute to the sizable gap between formal cannabis policy and cannabis policy as implemented. One factor is that national policy regimes allocate responsibility for policy enforcement. They can officially assign discretionary power to, for example, regional police authorities, enforcement officials, prosecutorial officials, and judicial officials. These officials may opt for a more punitive or more permissive approach, depending on their own or their organisation’s agenda. Another factor is limitations in financial or human resources, which might impede implementation of the formal policy.

The discrepancy between formal cannabis policy and cannabis policy as implemented has consequences for our ability to understand the relationship between policy, prevalence of cannabis use and consequences. Interpretation of the consequences of formal policy is impossible without knowing how that policy is implemented. Unless the implementation practices are known, analysis is tricky at best.

Cannabis policy and prevalence of cannabis use. Given the discussion immediately above, it is not surprising that there can be no definitive statement made regarding policy and the prevalence of cannabis use. Moreover, the range of policies as implemented is presumably more narrow than the continuum of formal policy from "prohibition" to "decriminalisation” would suggest, thus limiting our possibilities to assess the relative effectiveness of implemented policies. Within this restriction, our overall conclusion is that the evidence does not support the notion that policy and prevalence of cannabis use are strongly connected. Neither does the evidence support a definitive lack of connection, but the weight of the evidence, leans towards a lack of connection.

Part of the problem in establishing the connection between policy and prevalence is the variety of ways in which prevalence is measured. In many studies prevalence is measured in three ways: ever used, used last year and used last month. However, longitudinal and internationally comparable data on these three measures are lacking for many European countries, impeding a sound analysis of the connection between policy and prevalence. Moreover, each of these measures has a different meaning for understanding policy. Therefore, any comparisons across policy regimes must be interpreted with caution.

The majority of studies focus on the relation between cannabis policy and prevalence, and do not study the frequency of cannabis use by individual users. However, when studying the consequences of cannabis use, one should be able to distinguish between heavy users and casual users.

Even if there were systematic relationships between implemented policy and cannabis prevalence, these would not be conclusive evidence for a causal link. As we discussed in Chapter 1, policy is influenced by the cultural context of a nation, and the prevalence of cannabis use may also be related to that context. Therefore, it might be impossible to entirely disentangle the relationship between cultural context, policy and prevalence.

Cannabis and social consequences. In Chapter 6, we examined what is known about the social consequences of policies regarding cannabis use, from the perspective of the users and the communities in which users live. Here, there is some conventional evidence available, for example on the number of arrestees. In other areas where evidence would be desirable, for example on the fates of arrestees, it is largely missing.

It might be expected that the more nominally punitive a policy, the more users will face criminal sanctions. But this is not certain; per capita arrests, fines and other sanctions are not highly correlated with formal policy, and when the vagaries of definitions and data collection are taken into account, the relationship is even more tenuous.

Experiencing the criminal justice system has negative consequences for cannabis users beyond the correction of drug taking behaviour. Criminal records and other sanctions reach beyond the actual penalties themselves into almost every aspect of the user's life (for example, finding a job and finding housing), typically in a negative way. Whether these consequences to the user are beyond what would be considered proportional to the offence of cannabis use is a matter for public debate; evidence that the weight of that debate is on the "unfair" side is in the frequency across countries of systematic attempts to dissociate cannabis use from the criminal justice system. Additionally, to the extent that sanctions for cannabis use are viewed as excessive, the perceived legitimacy of the entire criminal justice system may suffer.

The negative consequences to the community of cannabis regulation have not been extensively examined. What evidence exists does not support an hypothesis of extensive harms to a community from cannabis use per se. The effect of regulation is another story. However, the evidence is so scant that no firm conclusions may be drawn.

The costs of enforcement of cannabis policy have not been extensively examined. A few jurisdictions have provided some figures; however the cost-effectiveness of various policies cannot be yet determined. This is an important gap in information that requires filling.

Quality of the evidence. Regarding the three main questions of the review, there is no dearth of opinions, and the range of these opinions is broad. As we have reported in the previous chapters, the evidentiary basis of these opinions is quite narrow; indeed, the breadth of opinions is probably at least in part a result of the narrowness of the evidence. Moreover, because of the nature of the topics, the evidence can never approach the strength of scientific validity of a randomised controlled trial, or even a case control study. Instead, evidence-based studies attempt to understand the relationship between policy and cannabis by longitudinal studies (mostly surveys of different individuals from the same population), "natural experiments" where uncontrolled differences amongst populations may be found, and internal regression analyses of single or multiple populations. Although these types of studies are less than the standards that the Cochrane Collaboration48 or other researchers committed to evidence-based policy might prefer, they are probably about as good as can be obtained.

What is not about as good as can be obtained is the consistency of approaches to collecting evidence. Although attempts to harmonise data collection for purposes of examining policy towards drugs has been for over 20 years in Europe the objective of many organisations (e.g. the Pompidou Group, the EMCDDA, and WHO-Europe), we are still not there, either within nations (with some exceptions) or across national boundaries. Although the technical quality of data collection and analytic sophistication has greatly increased over the years, it is still not as good as it could be. The conventional wisdom is to state that, because of the efforts of the abovementioned organisations, that things will get better, but that conventional wisdom has been with us for too long, without the magnitude of improvement predicted.

7.2 What needs to be done?

Our literature review showed that the scientific evidence regarding the effectiveness of cannabis policy as implemented is limited, and we believe that it is thus fair to assume that policy-making regarding possession and use of cannabis is not based on solid evidence. The summary of the evidence examined does not lead to the ability to make any recommendations regarding the implementation of policy. It must be admitted that we simply do not know enough about how policy affects cannabis prevalence and consequences. Therefore, the recommendations below are aimed at getting to know enough about the relationship between cannabis policy, implementation and outcomes. Perfect understanding is not possible, but fortunately, it is also not required. With some effort, enough understanding can be obtained to guide coherent policy making.

Recommendation 1: Spend more money on basic information acquisition, and spend it wisely. The dominance of the non-European Anglo-Saxon community (e.g. the US, and Australia) in the scientific community studying policy and drugs is not only because they are willing to spend proportionally more money on studies, but also because they have been more systematic in their approach to collecting and analysing data. There is as yet nothing approaching the American National Longitudinal Survey in Europe, and there needs to be.

Recommendation 2: Start understanding the range of how policies are implemented and why different implementation choices are made. We know that implementation varies. We do not know how it varies by jurisdiction, what the full range of variation is, and why variations occur. Full comprehension of this matter on an EU-scale is a vast undertaking. A more modest beginning can be made by looking at a select number of countries and a select number of jurisdictions in these countries to get a systematic flavour of how much variation there is. In the process, a method can be developed that can be expanded, first to other jurisdictions within the first set of countries and then to other countries. Such information can be collected by interviewing professionals involved in the implementation of cannabis policy. Annex B gives an overview of issues that need to be addressed when such an effort is undertaken.

Recommendation 3: Where differences in "real policy" (in the sense of the policy that is implemented at the local level) can be established, conduct comparative studies of cannabis use, using common measures. Again, to do this on an EU-wide basis is not yet feasible, so starting small is the desirable first step. To the extent that the same implementation strategy is found across different nations (and national policies) and different implementation strategies are found within one nation or region, this study can bring evidence to bear on the question of the effect of policy on cannabis prevalence.

Recommendation 4: Where the differences in cannabis policy as implemented can be found, multidisciplinary studies of the costs of different implementation regimes to the society should be examined, again using common data measures across jurisdictions. The societal costs of different implementation regimes should be measured in terms of the monetary costs of enforcement, the attitudes towards enforcement, the crime rate attributed to cannabis and the nuisance to the community of cannabis users. Studies carried out in fulfilment of this  recommendation will complete the structure needed for as evidence-based a level of comparison of policies as is reasonably possible. Again, the full examination of the question will take time, and starting small is the only way to start on the path of success.

Recommendation 5: To gain more insight in the outcomes of cannabis policy, data should be collected on the consequences for cannabis users. Currently not much is known about the outcomes of cannabis policy for cannabis users. Therefore, we believe it is important to collect the following data:
o Probability of being arrested for cannabis use;
o Sanctions in practice for those arrested: criminal record, caution, incarceration, fine, etc.
o Other consequences: obligatory treatment, losing driver’s license, etc.
If done well, such data might eventually lead to the development of a new set of indicators reflecting the stringency of cannabis policy. When better indicators for the stringency of cannabis policy in actual practice are available, it is possible to gain more insight in the connection between informal cannabis policy and prevalence.

Recommendation 6: Create opportunities for policy makers with differing beliefs about what effective policy should and could constitute to come together and discuss their viewpoints in a non-threatening way. Debates amongst advocates for different current policies typically generate more heat than light, as the advocates defend their current positions in light of the problems that they are currently facing. Opportunities for discussing these matters on a level playing field are rare, because the evidence to create that level playing field is not available.

However, if the previous recommendations are implemented and an evidentiary base begins to emerge, this can be used as a vehicle to engage the policy actors in a nondefensive, future-oriented discussion where possibilities of moving from current positions to harmonised ones that still remain faithful to national cultural bases might be possible.

RAND Europe

Re-Asserting Control: Voluntary Return, Restitution and the Right to Land for IDPs and Refugees in Myanmar - cover