European observations on cannabis legalization



The picture of diverse cannabis supply laws may be broadened by including recent draft laws from European Union countries, where objectives are more to protect public health than to raise revenue. Challenges for evaluation can be reduced by a well-designed policy and clear objectives—and data sources—from the start.

Kilmer & Pacula outline the diversification of supply laws around the world, and researchers’ ability to evaluate the consequences of the changes [1]. This picture of diversification may be broadened by including recent draft laws. In the European Union no national government has shown any interest in officially legalizing cannabis supply for recreational purposes, but this has not stopped politicians elected to the national parliament from proposing legislative models. While these are not in force—indeed, most have already been rejected—they give an indication of the range of laws that might need to be evaluated and the challenges this brings.

Recent drafts submitted by European politicians to their parliaments include proposals by a single parliamentarian in Ireland [2], the Palikot Movement party in Poland [3], the ecologist group in France [4], the Left Bloc in Portugal [5], a group including the Green Party in Germany [6], an opposition party in the Netherlands [7] and a cross-party group in Italy [8]. In Ireland, France, Germany, the Netherlands and Italy the model proposed was sale via licensed outlets (taxed in Ireland and Germany), with Italy and Ireland also providing for cannabis social clubs (collective growing without sale); in Portugal, only personal cultivation and cannabis social clubs would be permitted; in Poland, only clubs. In Spain, three Supreme Court decisions of 2015 declared organized cultivation by clubs open to new members as a trafficking crime [9]. However, a legal change in the same year defined the offence of personal cultivation, similar to that of use, to be committed only if the growing was in public view [10], perhaps implying that personal cultivation and use in private is now not an offence of any kind, comparable to Washington, DC.

Kilmer & Pacula consider four factors that affect supply and evaluation; namely, purpose, producer, purchaser and products. However, these may, in turn, depend upon the basic objectives of the laws, which may determine an important standard against which they should be evaluated. In the objectives of the four US retail systems, stated at the start of the laws, the central aims are the re-prioritization of law enforcement resources to focus upon violent and property crime, and to raise tax revenue. By comparison, the stated objectives of most of the European proposals are to protect young people’s health through access and quality controls. However, the implementation evaluations that have been conducted in Colorado and Washington State are not always able to address the stated objectives of the policies [11, 12]. The Colorado report notes that several topics for evaluation requested by Senate Bill 13–283 actually had no baseline data set in practice. There is relatively little coverage of evaluation of the officially stated objectives of the change put to the voters—re-prioritization of resources to fight violent and property crime—but the debate, as with Kilmer & Pacula’s paper, focuses upon the increased threat to public health, examining statistics of drug use and problematic use. Tax revenues, seductive in the post-recession recovery, are rarely acknowledged in public debate as being less than 1% of the total state budget [13].

Proposals from European political parties consider using evaluations at different points in the policy cycle. Cost benefit analyses, performed ex ante, can be a strong empirical support to decision making processes. Some European proposals consider a comprehensive model of public health and law enforcement costs and benefits, rather than focusing upon revenue and public expenditures. Process evaluations consider if a policy is being implemented as planned or frustrated by unexpected limitations (e.g. smoking bans contributing to unexpected consumption of edibles in Colorado). Ex-post evaluations that aim to show positive or negative impact are difficult to undertake due to data limitations, but these difficulties can be reduced by a well-designed policy and clear objectives—and data sources—from the start. Kilmer & Pacula’s paper illustrates clearly how cannabis policies are evolving globally. There is no doubt that the need for better data to evaluate these policies—including data on consumption of other substances—will become a key part of the discourse [14].


Green light for medicinal cannabis but AMA says proceed with caution

Australian Medicine
Volume 29 Issue 3 (07 Mar 2017)

Abstract: The Federal Government has given the green light to the legal sale of medicinal cannabis, with Health Minister Greg Hunt announcing that companies will be allowed to apply to distribute cannabis oils and other medical marijuana products.

Institutional users Login to access article

To cite this article: Johnson, Chris. Green light for medicinal cannabis but AMA says proceed with caution [online]. Australian Medicine, Vol. 29, No. 3, 07 Mar 2017: 9. Availability: <;dn=703616704877745;res=IELAPA> ISSN: 2202-1701. [cited 26 May 17].

Personal Author: Johnson, Chris; Source: Australian Medicine, Vol. 29, No. 3, 07 Mar 2017: 9 DOI: Document Type: Journal Article ISSN: 2202-1701 Subject: Cannabis–Therapeutic use; Cannabis–Law and legislation; Cannabis–Social aspects; Australian Medical Association;


As of January 2016, 23 U.S. states and the District of Columbia have legalized medical or nonmedical cannabis, with more likely to follow. This dynamic policy context represents a substantial challenge for policy evaluation. Part I of this article provides a summary of state-level cannabis policy components across states and federal action regarding state-level policies. Part II presents a detailed history of cannabis policies in Washington State from 1998 to 2015 and analyzes the potential impacts of the policy changes on cannabis supply and demand. As an early adopter of both medical and nonmedical cannabis policies, Washington State provides an excellent example of the key elements to be considered in evaluating the relationship between policy changes and cannabis use. We highlight the importance of the interplay of federal enforcement priorities and previously adopted state-level cannabis regimes in interpreting the potential impacts of new cannabis policies.

Crime and the Legalization of Recreational Marijuana

18 PagesPosted: 5 Feb 2017

Davide Dragone University of Bologna – Department of Economics

Giovanni Prarolo University of Bologna; Fondazione Eni Enrico Mattei (FEEM)

Paolo Vanin University of Bologna – Department of Economics

Giulio Zanella University of Bologna


We provide first-pass evidence that the legalization of the cannabis market across US states may be inducing a crime drop. Exploiting the recent staggered legalization enacted by the adjacent states of Washington (end of 2012) and Oregon (end of 2014) we find, combining county-level difference-in-differences and spatial regression discontinuity designs, that the legalization of recreational marijuana caused a significant reduction of rapes and thefts on the Washington side of the border in 2013-2014 relative to the Oregon side and relative to the pre-legalization years 2010-2012. We also find evidence that the legalization increased consumption of marijuana and reduced consumption of other drugs and both ordinary and binge alcohol.

Keywords: cannabis, recreational marijuana, crime

JEL Classification: K23, K42

Dragone, Davide and Prarolo, Giovanni and Vanin, Paolo and Zanella, Giulio, Crime and the Legalization of Recreational Marijuana. IZA Discussion Paper No. 10522. Available at SSRN:

Drug law enforcement (DLE) research has been poorly funded relative to drug treatment research. The literature is slight in volume and not yet very insightful. Taking the lack of funding to represent a chronic lack of public interest in the effects of DLE, the article offers a set of suggestions for how to create of a stronger DLE research community.

Research on drug treatment is highly sensitive to the nature of the intervention involved. A typical study will be of a specific type of treatment, for example contingency behaviour therapy, administered to a specific patient population with well-described outcomes (e.g. Copeland, Swift, Roffman, & Stephens, 2001; Kakko, Svanborg, Kreek, & Heilig, 2003). A caveats section will identify the numerous qualifications to internal and external validity. Each study is a contribution to a literature that is large enough that it has already its own literature of meta-analyses (e.g., Minozzi et al., 2006; Prendergast, Podus, Chang, & Urada, 2002).

Contrast that with research on drug law enforcement. The studies are few, the interventions are usually described in very broad terms, there is little effort to specify the settings (population, environment, history) in analytically relevant terms. Caveat sections are brief. The literature is not cumulative and meta-analyses are few and impoverished, in terms of the studies that fit the criteria for systematic reviews (Mazerolle, Soole, & Rambouts, 2006; Werb et al., 2011).

Why these differences, all of which point to the greater strength of treatment research? One answer is surely money. In the USA, the principal source of research funding, the budget for NIDA is over $1 billion, of which perhaps $600 million goes to treatment related research. Other agencies such as the Department of Veteran Affairs, also provide substantial funding. For enforcement research there is primarily the National Institute of Justice, with a paltry $50 million for all law enforcement research, of which a small fraction goes to drug law enforcement studies. Foundations have shown little interest in drug enforcement research; for example, the Robert Wood Johnson Foundation’s large Substance Abuse Policy Research Initiative in the 1990s and 2000s included few projects on law enforcement. There is at least an order of magnitude difference in the funds available for drug treatment and research and those available to study the effects of enforcement.

The problem though is not just money. There are also specific challenges to work in this field that are not encountered in an individually oriented and experimental field such as treatment. For example, enforcement experiments are possible but so far only of very narrow interventions in atypical settings, such as policing of crack houses in the poorest neighbourhoods of declining American cities (Sherman et al., 1995, Weisburd and Green, 1995). These experiments are expensive to mount and require lengthy negotiations with suspicious enforcement agencies. There are many outcome measures that are fundamental to the field but only available at very aggregate levels whereas learning will come from analysis at much more local levels. For example, federal nations might have variation in the intensity of enforcement across provinces/states but almost no country reports drug prices below the national level, so that the price effects of differences in enforcement intensity cannot be studied.

A further complication is that enforcement goals are multiple and sometimes contradictory. In addition to reducing the use of drugs by shifting the supply curve through increasing the risks faced by sellers, enforcement also aims to make users and communities safer and to reduce the power of drug selling organizations in transshipment and producing countries (Caulkins & Reuter, 2010). A useful research literature will take all these outcomes into account but only rarely are these multiple outcomes available at the same level of aggregation. Thus individual studies usually emphasize just one outcome because of data availability (e.g. systematic quantitative data on prices but little on user victimization).


ELSEVIER  Uruguay’s middle-ground approach to cannabis legalization

Assessing the concordance between illicit drug laws on the books and drug law enforcement: Comparison of three states on the continuum from “decriminalised” to “punitive”

Volume 41, March 2017, Pages 148–157



Variations in drug laws, as well as variations in enforcement practice, exist across jurisdictions. This study explored the feasibility of categorising drug laws “on the books” in terms of their punitiveness, and the extent of their concordance with “laws in practice” in a cross-national comparison.


“Law on the books”, classified with respect to both cannabis and other drug offences in the Czech Republic, NSW (AU) and Florida (USA) were analysed in order to establish an ordinal relationship between the three states. Indicators to assess the “laws in practice” covered both police (arrests) and court (sentencing) activity between 2002 and 2013. Parametric and non-parametric tests of equality of means, tests of stationarity and correlation analysis were used to examine the concordance between the ordinal categorisation of “laws on the books” and “laws in practice”, as well as trends over time.


The Czech Republic had the most lenient drug laws; Florida had the most punitive and NSW was in-between. Examining the indicators of “laws in practice”, we found that the population adjusted number of individuals sentenced to prison ranked across the three states was concordant with categorisation of “laws on the books”, but the average sentence length and percentage of court cases sentenced to prison were not. Also, the de jure decriminalisation of drug possession in the Czech Republic yielded a far greater share of administrative offenses than the de facto decriminalisation of cannabis use / possession in NSW. Finally, the mean value of most “laws in practice” indicators changed significantly over time although the “laws on the books” didn’t change.


While some indicators of “laws in practice” were concordant with the ordinal categorisation of drug laws, several indicators of “laws in practice” appeared to operate independently from the drug laws as stated. This has significant implications for drug policy analysis and means that research should not assume they are interchangeable and should consider each separately when designing research.