Technical ReportPDF Available

EU Presidency Policy Brief: Treaty Compliance Options for Cannabis Regulations in the EU (Models of decriminalisation and legal regulation compliant with International Law and EU Acquis)

  • Independent researcher


This policy brief presents critical challenges to international and European legal compliance raised for national authorities while exploring decriminalisation and legal regulations of non-medical cannabis. The 1961 Single Convention has become ipso facto part of Union acquis. Far from being a barrier, this legal panorama enables 11 options EU Member States could adopt to reconcile domestic reforms with EU law. Prague: Racionální politiky závislostí & CzecHemp. [published on 2 December 2022; released on 14 February 2023] Video of the launch: (slides:
Models of decriminalisation and legal regulation
compliant with International Law and EU acquis
This policy brief presents critical challenges to
international and European legal compliance
raised for national authorities while exploring
decriminalisation and legal regulations of
non-medical cannabis.
The 1961 Single Convention has become
ipso facto part of Union acquis.
Far from being a barrier, this legal panorama
enables 11 options EU Member States could adopt
to reconcile domestic reforms with EU law.
2 Dec. 2022
1.1. Lexicon 5
1.2. Decriminalisation: rationale, context, precedents, and legality 5
1.2.1. Decriminalisation and international law 7
1.2.2. Decriminalisation and EU law 9
1.3. Legalisation: a more complex panorama 10
2.1. International law: drug control 11
2.1.1. 1971 and 1988 conventions 11
2.1.2. Goals of the 1961 Single Convention 12
2.1.3. “Purposes” in the Single Convention 13
2.1.4. Cannabis cultivation 14
2.1.5. Personal activities 15
2.2. International law: non drug-related 15
2.3. EU acquis 15
2.3.1. Primary legislation 16
2.3.2. Secondary legislation 16
2.3.3. Interpretation, non-legislative acts 18
2.3.4. Case law 19
2.3.5. Treaty-making competencies of the EU 20
3.1. Options A: Regulation after denunciation of the Single Convention 24
3.1.1. Option A1: remaining outside 24
3.1.2. Option A2: re-accessing 24
3.2. Options B: Regulation after internal amendment of the Convention 25
3.2.1. Option B1: amending articles 25
3.2.2. Option B2: amending schedules 27
3.3. Options C: Regulation after external amendment of the Convention 28
3.3.1. Option C1: new convention 28
3.3.2. Option C2: inter se amendment 28
3.4. Option D: Direct regulation justified by the superiority of human rights law 30
3.5. Options E: Direct regulation of a “non-medical industry” 32
3.5.1. Option E1: article 2 (9) and 28 (1), unilaterally 33
3.5.2. Option E2: article 2 (9) and 28 (2), unilaterally 34
3.5.3. Option E3: article 2 (9) and 28 (2), multilaterally 36
3.6. Option F: Direct regulation as an experiment (scientific purposes) 37
4.1. Multi-tiered trade relationships 40
4.2. Cannabis-specific EU trade law 41
Acronyms Case law Bibliography
National cannabis law reforms such as decriminalisation and legal regulation trigger complex questions of
compliance with international law and European acquis. The legal grounds for regulations of cannabis
departing from prohibition, within and possibly between Member States, need ascertainment.
The widely-accepted view that the 1961 Single Convention on narcotic drugs excludes all non-medical
uses of cannabis can and should be questioned. Whether there is a violation “depends on the concrete
design of the legal regulations” (Deutscher Bundestag, 2022a), particularly because the interpretations of
the Single Convention “are as diverse as the putative effects of [cannabis] itself” (Dawkins, 1973, p.
Cannabis policy is primarily a competence of EU Member States, but their legislative capacity is clearly
limited under international and European law, albeit differently. In previous years, numerous political and
strategic options have been outlined internationally (e.g. OAS, 2013). Following Germany’s announced plans
for domestic cannabis regulations, several analyses and opinions have been published, often by legal experts
new to the field of drug policy. Unfortunately, most only scratch the surface, and rely on assumptions rather
than a methodological legal approach, while key provisions are also often overlooked.
Apolicy void has been created by arguably unreliable advice amidst the absence of proper assessments of
EU Member States’ legal obligations. Recent developments in human-centred, evidence-based, and
health-oriented drug policies, as well as key events (UNGASS 2016, sustainable development agenda,
change in international scheduling, ruling of the Court of Justice of the EU, an increasing number of
domestic cannabis policy reforms, including within the EU) call for a re-analysis of the legal context for
cannabis reforms.
Massive violations of international law are perpetrated at the Eastern border of the European Union. While
countries in the EU engage in cannabis policy reforms, the excuse of supposed breach of the drug control
treaties has led to untenable accusations. Notably, European countries would “claim the right to judge other
Member States’ implementation of international law [but] have themselves too often disregarded those
principles in their drug-related engagement” (Permanent mission Russia, 2022). Besides the profound
hypocrisy of this statement, it highlights the importance and political relevance of efforts to reconcile treaty
obligations (or lack thereof) with domestic cannabis regulations.
The EU is founded on the values of the rule of law, and committed to upholding and promoting the strict
observance and the development of international law (articles 2 and 3 (5) of the Treaty on European Union).
It should seek pathways to alleviate rule tensions and the risks of norm decay or non-compliance cascade.
This policy brief summarises the feasible and viable policy options that allow EU Member States to
regulate non-medical cannabis while complying with UN and EU legal frameworks.
Scope: this brief does not discuss options in flagrant violation of jus cogens or international law (including
human rights law), options of claimed non-compliance (e.g. Canada; Boister and Jelsma, 2018, p. 459),
unsubstantiated proposals (e.g. selective denunciation based on rebus sic stantibus, an historical error or a
“fundamental change in circumstances” Dawkins, 1973, pp. 363–364) or other unsound proposals (e.g.
relying on the regime of “special stocks” in the Conventions). Likewise, therapeutic, medicinal, wellness and
other hemp-related uses of cannabis are not discussed in this brief (see UNCTAD, 2022).
This brief introduces key terms and concepts (1) before reviewing current international law and European
acquis (2). Section (3) presents six compliant options (A: denunciation of the Single Convention, B: internal
amendment, C: external amendment, D: hierarchical superiority of human rights law, E: direct regulation of
a non-medical industry, and F: direct regulation of an experiment for scientific purposes). Questions relating
to the internal market are discussed in section (4) before the final recommendation (5).
Step 1 (immediate)
Decriminalise personal cultivation, possession and use.
Step 2 (short-term)
Regulate non-medical cannabis industries under articles 2 (9) and 28 (2) of the 1961
Single Convention on narcotic drugs.
Publicise the legal rationale in non-binding Statements between like-minded States.
Address carefully cross-border issues, including anticipating a future progressive
development of trade.
Step 3 (following step 2)
Engage in dialogues with WHO and INCB for guidance and compliance on the harm
reduction obligation (article 2 (9) (a)) and on statistical reporting (article 2 (9) (b)).
Engage in an open-ended dialogue with the European Commission.
Step 4 (long-term)
Explore and study an amendment of the Single Convention under article 47.
1.1. Lexicon
The definitions for “decriminalisation,” “depenalisation,” “legalisation,” “regulation” vary and, at times,
overlap.1In this brief, within the limited context of cannabis, the following definitions are used:
Decriminalisation (= depenalisation): altering domestic laws on the use and possession, and related
preparatory activities to the extent that they are strictly necessary to personal use and possession.
Legal regulations (= legalisation): altering domestic laws on use and possession and other activities
(cultivation, transformation, transport, trade, import/export, sale…) in part or in whole.
If not stated otherwise, the term cannabis refers indistinctly to the cannabis plant, plant parts, resin, or any
derivatives (except for pure THC, not included under “cannabis” in this brief).
TEU is the Treaty on European Union and TFEU the Treaty on the Functioning of the European Union
(both in their consolidated version of March 2020). Schengen Agreement refers to the 1985 Agreement of
14 June 1985 between the Benelux, Germany and France on the gradual abolition of checks at their common
borders, and Implementing Convention to the 1990 Convention implementing the Schengen Agreement
of 14 June 1985 (Council, 2000).
The term “Conventions” alone refers together to these 3 international legal agreements:
Single Convention: Single Convention on Narcotic Drugs, 1961, in its version amended by the 1972
Protocol amending the Single Convention on Narcotic Drugs (UN, 1975),
1971 Convention: Convention on psychotropic substances (UN, 1976a),
1988 Convention: UN Convention against Illicit Traffic in Narcotic Drugs and Psychotropic
Substances (UN, 1990).
Commentary refers to the official UN Commentary of the Single Convention (UN, 1973). Council
refers to the Council of the European Union. When used alone, Commission refers to the European
Commission, while the UN Commission on Narcotic Drugs is referred to as CND.” High Representative
is the High Representative of the EU for Foreign Affairs and Security Policy.
A list of acronyms is present at the end.
1.2. Decriminalisation: rationale, context, precedents, and legality
The non-criminalisation of minor, personal and non-violent drug-related activities is core to any evidence-
and human rights-based approach to drug policies. While the March 2018 Council Conclusions recalled that
“alternative to coercive sanctions can be applied for a range of offences” (Council, 2018), there is a
shared consensus that decriminalisation refers primarily to personal use and possession. However, since use
and possession are only possible after substance acquisition, the activities of buying and cultivating for
personal purposes (in particular, so-called home-growing in the case of cannabis) also fall under the
umbrella of decriminalisation as a consequence. Collective activities in the private sphere are also generally
de facto encompassed (as seen in section 1.2.2 below).
An increasing number of governments or parliaments have taken initiatives to carve personal conduct out of
their law criminalising drug-related activities (Eastwood et al., 2016; INCB, 2019) to avoid interfering with
1For detailed definitions, see EMCDDA (2016, p. 2)
fundamental rights. In the absence of political initiative, Constitutional or Supreme Courts have also often
invalidated provisions criminalising personal conduct based on constitutionally-recognised human
rights (EMCDDA, 2019), for example:
On the grounds of personal autonomy and privacy,” the Supreme Court of Hawaii (1972), the
Supreme Court of Alaska (Brandeis, 2015), and Argentinas Corte Suprema de Justicia (Sistema
Argentino de Información Jurídica, 2009) ruled in favour of decriminalisation.
The Corte Constitucional of Colombia reached similar conclusions based on the right to
autonomous development (del Río Forero et al., 2022), and Mexicos Corte Suprema de Justicia de
la Nación based on the right to the free development of one’s personality (Ibarra, 2022).
The Constitutional Court of South Africa relied on the right to privacy to decriminalise use,
possession and cultivation in private (Zondo, 2018; Mhlantla, 2022). Going further, Chiles Corte
Suprema de Justicia expanded decriminalisation to the cultivation of cannabis collectively in a
private circle, with reasoning echoing in part the legal “doctrine of shared consumption” behind
Spain’s jurisprudence on cannabis clubs (Fernández Bautista, 2021; Marks, 2019).
In Europe, the Constitutional Court of Georgia “held it to be unconstitutional to criminally prosecute
people for consuming cannabis” in 2015 and 2018 (Žalimas et al., 2018, p. 194). The Corte Suprema
di Cassazione (2019; 2020) in Italy ruled that the cultivation of cannabis for personal use, regardless
of the THC content, should not be considered a punishable offence, provided that it is cultivation
based on simple methods and with a limited number of plants and yield.
What stems from all these Supreme or Constitutional Courts’ rulings is the primacy of human rights (in
particular, the rights to privacy and personal autonomy) over measures to criminalise personal conduct,
particularly in the private sphere.
Indeed, human rights law provides a normative framework, against which criminalization and
penalties are to be assessed (Heilmann, 2011, p. 280). This primacy is even clearer insofar as there is no
supra-national mandate or obligation to criminalise these activities. In 2016, INCB President declared:
“The evolution of drug control policy and practice, in many countries over recent years, from an approach
relying primarily on criminal justice and incarceration to a more health-oriented approach is welcomed. This is
entirely consistent with the conventions” (Sipp, 2016).
In its “Study on the impact of the world drug problem on the enjoyment of human rights,” the UN High
Commissioner for Human Rights (UN Human Rights Council, 2015) called for decriminalisation, among
other things, to address the disparate impact on ethnic minorities, women and the youth. As similarly
stated by the Pompidou Group (2021), there are countless other benefits to a public policy which
de-prioritises the repression of personal conduct and minor, non-violent offences:
“decriminalisation can facilitate the search for treatment for those dependent on drugs, […] improve public
health outcomes, and protect many people from the devastating impact of a criminal conviction.”
One of them is the possibility of reducing public expenditure (particularly on police, justice and other
non-health related costs; Gonçalves et al., 2015) and reallocating it to the fight against more serious crimes.
In the context of developing regional and international crises, decriminalisation may prove helpful in
mitigating the impact of austerity measures on drug-related health and social services (Pompidou Group,
In practice, decriminalisation can take a myriad of legal shapes. A difference is sometimes made between
depenalisation, which may not require a change in the legal framework (de facto) and decriminalisation
which does (de jure). However, there is still no consensus on terminology. Below, an indicative table
presents the spectrum of modalities (adapted from IDPC, s.d.) from the most repressive on the left to the one
offering more legal certainty with regard to the non-criminalisation of personal activities on the right:
Criminalisation in law
(de jure)
Decriminalisation in law (de jure)
of crimi-
Decriminalisation in practice
(de facto or “depenalisation”)
criminalisation still possible
from the
justice system
Absence of
Discretion in
the criminal
justice system
discretion /
Absence of
offence for
use and
No mention of
“use” and
in the law
Several states
and counties
No known
In a briefing, the UN Commission on Narcotic Drugs (CND) clarified that:
“the Conventions do not absolutely require the punishment of possession, purchase or cultivation for personal
use. That is why, under the Conventions, de-penalisation of possession, purchase or cultivation of controlled
drugs for personal use is possible, under specific circumstances” (CND, 2014, p. 6).
1.2.1. Decriminalisation and international law
In international law, “decriminalisation” is the waiving (or non-application) of criminal or administrative
sanctions to those uses and possession of drugs interpreted as illicit under the drug control conventions
(under an interpretation of Articles 33 and 36 (1)). Despite persisting beliefs, the three international drug
control conventions offer evident flexibility for governments:
“the use of alternatives to conviction and punishment, as provided for by the conventions, can form an
integral part of a balanced and human-rights based approach to drug policy” (INCB, 2022b).
While the Conventions require some State answer to the illicit possession of narcotic drugs, this can be a
warning, a reprimand, or even the provision of health information or (non-compulsory) diversion towards the
healthcare system. As recalled by former INCB president Werneer Sipp (2016), where applicable,
constitutionally-protected rights (such as the right to privacy) allow to waive State responses to
personal conduct involving drugs fully:
“Indeed, the international drug control conventions require proportionate responses to drug-related offences.
While ‘serious offences shall be liable to adequate punishment particularly by imprisonment’, there is no treaty
obligation to incarcerate drug users having committed minor offences. The conventions provide for flexibility
by States in determining appropriate sanctions, including non-punitive or non-custodial measures, for instance
for possession of drugs for personal use.”
The key provisions codifying decriminalisation are the following (Eastwood et al., 2016):
Singe Convention:
Article 33, possession of cannabis is allowed only “under legal authority,”
Article 36 (1) (a) on “penal provisions” requires Parties to punish all activities (except
“use”). These activities must be “contrary to the provisions of this Convention” and the
article is “subject to [the] constitutional limitations” of each Party,
Under Article 36 (1) (b), if those who commit an offence are in particular “abusers of drugs”
a specific health-centred alternative to conviction/punishment can be applied (“treatment,
education, after-care, rehabilitation and social reintegration”).
1988 Convention:
Article 3 (1) (a) (ii) requires Parties to establish as criminal offences under its domestic law
the cultivation of cannabis plants contrary to the provisions of the Single Convention. This
requirement does not apply to cultivation aligned with the Single Convention.
Article 3 (2) calls Parties to adopt measures “to establish as a criminal offence […]
possession, purchase or cultivation of narcotic drugs […] for personal consumption.” These
must be “contrary to the provisions” of the Single Convention. The article is “subject to the
constitutional principles and the basic concepts of the legal system” of each Party. According
to Amber Marks (2019), this article “relieves State parties from the Article's obligation to
criminalise drug possession and cultivation for ‘personal consumption’ when doing so would
conflict with their constitution or basic concepts of their legal system. Spain relied on Article
3 (2) in its decision not to criminalise conduct involving personal consumption.”
Article 3 (4) (c) states: “in appropriate cases of a minor nature, the Parties may provide, as
alternatives to conviction or punishment, measures such as education, rehabilitation or social
reintegration, as well as, when the offender is a drug abuser, treatment and aftercare.”
Article 3 (4) (d): “Parties may provide, either as an alternative to conviction or punishment,
or in addition […], measures for the treatment, education, aftercare, rehabilitation or social
reintegration of the offender.”
It has been known since 1961 that the Single Convention does not mandate the penalisation of personal
conduct (Lande, 1976, p. 597); an impressive scholarship has developed in support of this fact. The
Convention’s Commentary explains that countries can “choose not to impose penalties on the
unauthorised possession for personal use” (UN, 1973, p. 402). Still, for some analysts, the international
legal compliance of decriminalisation schemes may depend on the actual design of the law (Ambos, 2022).
Decriminalisation is not only supported by international law but also by global consensus. The UNGASS
2016 outcome document, the 2019 CND Ministerial Declaration, and at the EU level, the 2018 Council
Conclusions, reaffirmed that the Conventions encompass the principle of proportionality in sanctions and
allow for the development of alternative measures to conviction or punishment (UN, 2016; CND, 2019;
Council, 2018).
Decriminalisation is recommended by the UN Office on Drugs and Crime (UNODC), the UN High
Commissioners for Human Rights and for Refugees, the International Labour Organisation (ILO),
International Office on Migration, UNAIDS, UNDP, UNESCO, UNICEF, UN Populations Fund, UN
Women, World Food Programme, and WHO, among other UN entities (ILO et al., 2017; UN, 2018;
UNAIDS, 2019; 2021; UNDP, 2018; UNDP et al., 2019; UNODC and WHO, 2018).
The UN Secretary-General and the heads of 30 UN agencies have together promoted “the
decriminalisation of drug possession for personal use” and “the principle of proportionality” (UN, 2018).
Recently anew the UN High Commissioner for Human Rights encouraged countries to decriminalise drug
use (Bachelet, 2022), drawing on calls made by virtually almost all UN human rights experts bodies (e.g.,
UN Committee on Economic, Social and Cultural Rights [CESCR] on various occasions), Special
Rapporteurs (e.g., on the Right to Health: UN, 2010) and other mechanisms (e.g., UN Working Group on
Arbitrary Detention), among a more extended list (UNDP et al., 2019). On 14 October 2022, the UN
CESCR, in its Concluding observations on the sixth periodic report of Italy, expressed his concerns
“about the punitive approach to drug use and the insufficient availability of harm reduction programmes
[recommending] that the State party review its drug policy and legislation to bring them into line with
international human rights norms and best practices” (CESCR in: Forum Droghe, 2022, p. 9).
At the peak of the COVID-19 pandemic, 16 UN human rights experts and special rapporteurs also called for
“moratoria […] on enforcement of laws criminalising drug use and possession” (Pūras et al., 2020). Also
because of positive impacts on health, the Global Fund (2022) lists decriminalisation among its “Program
Essentials” under the section “human rights.”
More recently, decriminalisation has been assessed against the backbone of international and European
human rights law (Bone, 2021; Lines, 2017; van Kempen and Fedorova, 2019b). These studies, in particular,
confirm the findings of UN bodies and national Supreme or Constitutional Courts about the primacy of
constitutional human rights to privacy, autonomy, health, or personal development, over measures
criminalising personal use and consumption. They also outline the extraordinary character of the
criminalisation of personal activities, suggesting that the burden of proof is on the State to justify
criminalisation” (Barrett, 2011).
Besides all this, the INCB (2019) notes that:
“the discretion to adopt criminal justice policies that include alternatives to conviction or punishment for minor
crimes that is provided for under the conventions remains underutilized
Highlight: Decriminalisation and international law.
Decriminalisation is supported by international drug control law, international human rights law, and case law.
Because of the amount of evidence suggesting the harmful effects of criminalization of personal conduct, particularly
on health, States willing to continue criminalising must justify the validity of their policies and its adequation to
international legal norms.
1.2.2. Decriminalisation and EU law
The central piece of EU law related to decriminalisation is the Council Framework Decision 2004/757/JHA
of 25 October 2004 (the “Framework Decision”) laying down provisions in the field of illicit drug trafficking
(Deutscher Bundestag, 2022b, p. 5; Jelsma, 2022, p. 95; van Kempen and Fedorova, 2019a, p. 115).
Article 2 of the Framework Decision is formulated to exclude from its scope all drug-related activities2when
committed “exclusively for […] own personal consumption as defined by national law.” Consequently,
personal activities related to drugs are not regulated under the Framework Decision.
As discussed in section 1.2 above, activities which would otherwise be punishable as illicit drug trafficking
are excluded from the Framework Decision when committed for personal consumption.
The Framework Decision further stipulates that its scope is limited to activities “committed without
right.” A right to personal use and possession defined by domestic law would further support the exemption
of these activities from the realm of the penal provisions established by the Framework Decision (discussed
2Production, manufacture, extraction, preparation, offering, offering for sale, distribution, sale, delivery on any terms
whatsoever, brokerage, dispatch, dispatch in transit, transport, importation or exportation of cannabis without right, or
the possession or purchase of cannabis without right with a view to conducting one of the previous activities.
in section 2.3.2). Finally, article 4 establishes criteria for proportionality of penalties: “large quantities” and
“drugs which cause the most harm to health” are to be treated more severely.
The 2018 Council Conclusions further promoted the use of alternatives to coercive sanctions for drug-related
Numerous Case law from the European Court of Human Rights (ECtHR) and Court of Justice of the
European Union (CJEU) have strengthened and specified the number of human rights, in particular the right
to privacy, enshrined in article 8 of the European Convention on Human Rights and article 7 of the EU
Charter of Fundamental Rights.
In Pretty v. the United Kingdom, the ECtHR (2002) explains that personal autonomy is part of privacy:
“[…] the notion of personal autonomy is an important principle underlying the interpretation of [the]
guarantees [of article 8 of the European Convention on Human Rights]. The ability to conduct one’s life in a
manner of one’s own choosing may also include the opportunity to pursue activities perceived to be of a
physically or morally harmful or dangerous nature for the individual concerned and even where the
conduct poses a danger to health or, arguably, life, the Case law of the Convention institutions has regarded the
State’s imposition of compulsory or criminal measures as impinging on private life.”
Another example is Niemietz v Germany, where the ECtHR (1992, §29) confirmed that “respect for private
life must also comprise to a certain degree the right to establish and develop relationships with other human
beings.” Codified together with the freedom of association, this right to develop private relationships with
other persons constituted the basis for the emergence of so-called cannabis social clubs. These groups
sharing cultivation and consumption in a private, non-profit and closed-loop model were given echo by the
supreme Courts of Spain and Chile (Marks, 2019; see also section 1.2).
The waiver (or expungement) of criminal and administrative records for people convicted of offences
which decriminalisation has repealed, is often seen as the natural next step, and a legal and ethical
consequence of decriminalisation. It is shown to positively impact on vulnerable populations (Kilmer et al.,
2021), justifying recent announcements by President Biden in the United States (Politico, 2022; White
House, 2022).
1.3. Legalisation: a more complex panorama
If the modalities of decriminalisation can be illustrated in a simplified table as in section 1.2, domestic
cannabis regulations represent a much more
“complex and multifaceted challenge with hundreds of policy decision points, each presenting opportunities to
either narrow or widen disparities. And social equity goals must be balanced with many other policy objectives
concerning legalisation, such as raising tax revenue, protecting youth, minimising increases in problematic use,
and attending to indirect effects on consumption of tobacco, alcohol, and other substances” (Kilmer et al.,
2021, p. 1040).
Reforming cannabis laws requires recourse to interconnected and multidisciplinary fields and policy areas at
the domestic level (FAAAT, 2021; Rahwanji, 2019; TDPF, 2014). This brief focuses on the options at the
supra-national level.
This section presents international and European laws affecting domestic cannabis regulations. With no
exception, international and European law criminalises cannabis-related activities that are “illicit,” “illegal,”
“without right” or “contrary to the provisions” of particular laws. All levels of law point at the somehow
ambiguous Single Convention as the benchmark for the legality of cannabis-related conduct.
2.1. International law: drug control
The three international drug control conventions (see section 1.2.1), ratified by all EU Member States,
establish legal requirements in relation to cannabis. Furthermore, the EU accessed article 12 of the 1988
Convention, but it is entirely unrelated to cannabis.
“As the [INCB] has reiterated on several occasions, the fundamental principles underpinning the three
international drug control treaties […] are the principle of a balanced approach, the principle of proportionality
and respect for human rights” (INCB, 2019)
Dual in nature and not self-executing, the Conventions pursue harmonisation and coordination in both health
policies on drug abuse and penal approaches to illicit trafficking. “A close look at the [Conventions] reveals
that, although formally binding, the penal provisions prove remarkably soft” (Colson, 2019, p. 81).
“Cannabis flower and resin are scheduled under the [Single] Convention, its principal psychoactive compound
THC is scheduled under the 1971 Convention, and the 1988 Convention mentions cannabis specifically in
articles about cultivation and eradication” (Jelsma, 2022, p. 92).
2.1.1. 1971 and 1988 conventions
The 1971 Convention lists THC (active ingredient of cannabis) in its schedules, but only applies to pure
molecules separated from the plant materials or of synthesis. Article 2 of the 1971 Convention limits its
scope to “substance[s] not yet under international control,” excluding cannabis which was already controlled
under the Single Convention. The 1971 Convention is therefore relevant only to pure isolated THC, but
not to cannabis even though it contains THC (Bayer, 1989, p. 23; Chatterjee, 1981, p. 458; INCB, 2014, p.
68; Schappe, 2001; Tupper and Labate, 2012; UN, 1976b, pp. 3, 25, 385–387).
The 1988 Convention lists controlled precursors, but none relates to cannabis. The 1988 Convention also
strengthens and supplements the 1961 and 1971 conventions (CJEU, 2010, §37), including via measures
related to cannabis. Nevertheless, it does not codify new obligations for cannabis control.
Article 14 paragraph (1) states that the measures to eradicate illicit cultivation and to eliminate illicit demand
must not be less stringent than the measures taken under the Single Conventions. Paragraph (2) obliges
Parties to prevent the illicit cultivation of cannabis and eradicate cannabis plants cultivated illicitly in its
territory. All provisions concern illicit cannabis, as defined under the Single Convention. Fundamentally a
penal instrument reinforcing already-existing treaties, the 1988 Conventions only adds-up penal
provisions to activities already illicit under the 1961 and 1971 conventions:
“In spite of its harsh wording and the specific reference to cannabis, the 1988 Convention clearly builds on the
two previous treaties. […] The key wording in this regard is ‘contrary to the provisions’ of the 1961 and 1971
conventions; and ‘illicit’ cultivation of cannabis plants” (Jelsma, 2022, p. 93).
In the 1988 Convention, illicit or illegal activities are the same as those of the preceding conventions
(Boister, 2001, pp. 97–103), although “to make sense, it should read conduct contrary to national legislation
as, in strict technical terms, the Parties criminalise the conduct, not international law” (Boister, 2001, p. 76).
EU accession to the 1988 Convention: Article 27 (2) of the 1988 Convention explains that “regional
economic integration organisations shall declare the extent of their competence with respect to the matters
governed by this Convention.” Under this provision, the Council (1990) declared the EU to be only
“competent for questions of commercial policy relating to [precursors], questions which are dealt with in
Article 12 of the Convention.” Article 12 is completely unrelated to cannabis. In the same declaration, the
Council “reserve[d] the right to make further declarations in accordance with Article 27 (2)” to expand its
competence. Because it never declared any other competence, the EU has only ratified article 12 of the 1988
Convention, and is therefore not bound as such by any of the other provisions of the 1988 Convention,
contrary to what has too-often been claimed (e.g. in Deutscher Bundestag, 2022b).
Highlight: 1988 Convention and the EU.
In recent months, several documents have circulated, wrongfully assuming that the EU was in and of itself a Party to
the 1988 Convention, in addition to each State being a Party in their own right. This is not entirely correct: the EU
has only ratified article 12 of the 1988 Convention, but not the cannabis-related parts of the Convention.
The 1988 Convention is clear in explaining that regional organisations such as the EU need to declare precisely the
extent of their competence under the Convention. Accordingly, the Council declared EU’s competence only under
article 12 about commercial precursors policy. The competence of the EU can not exceed this declaration without
further ratification by the Council. The EU has therefore not ratified the 1988 Convention, it has only ratified its
article 12.
The 1988 Convention should not be at the centre of the discussions: the EU is not fully a party to it, and in any
case, it does not create any additional specific obligation for States other than those established by the 1961 Single
Convention. Discussions should focus on the 1961 Single Convention which is the only relevant legal basis.
Hence, besides pure THC and illicit conduct, cannabis is not controlled under the 1971 or 1988
conventions. The international legal framework for licit cannabis is therefore determined under the
Single Convention. Accordingly, analyses of the nexus between global drug control and domestic cannabis
regulations tend to focus on the Single Convention. According to INCB (2014, p. 68):
“Under the 1961 Convention […], plants that are the sources of narcotic drugs, such as cannabis plant, […] are
subject to specific control measures. In contrast, […] no plants are currently controlled under [the 1971]
Convention or under the 1988 Convention. Preparations (e.g. decoctions for oral use) made from plants
containing [active ingredients listed in the 1971 Convention] are also not under international control.”
2.1.2. Goals of the 1961 Single Convention
Usual interpretations of the Single Convention have argued that its goal and object is the prohibition of all
kinds of non-medical and non-scientific use of drugs (Boister, 2001; Bruun et al., 1975). These positions are
difficult to reconcile with the letter and spirit of the Single Convention, where the few explicit mentions of
the option to prohibit are discretionary upon the parties (Marks, 2019, p. 213). Prohibition is effectively
contemplated in the Single Convention as a legal and legitimate option but not as an obligation:
“The decision in this respect lies with governments, since they are in a position to evaluate the prevailing
conditions in their countries or territories, taking into account the world narcotics situation” (UN
Secretary-General, 1966, p. 22)
All mentions of prohibition3are conditioned to “the protection of the public health or welfare.” Besides this,
the majority of the convention describes measures of control and for “international cooperation with respect
to governing the proper or improper fluxes of narcotics” (Zitt, 2016, p. 526).
Prohibition does not constitute the Convention’s goal either (Riboulet-Zemouli, 2022, pp. 91–96). The goal
of the Single Convention is to protect the health and welfare of humankind. Criminal laws and
prohibition are merely tools or means to achieve these public health objectives. The objective of the
convention is “humanitarian and social” in nature for Paul Reuter (1968); for the INCB (2022a, p. 15), it is:
“to safeguard the health and welfare of humankind through two overarching goals, namely: (a) ensuring the
availability of controlled narcotic drugs and psychotropic substances for medical and scientific purposes and
ensuring the availability of precursor chemicals for legitimate industrial use; and (b) preventing the diversion
of controlled substances into illicit channels.”
The CJEU, concerning cannabis and “in accordance with settled Case law,” confirmed:
the Single Convention is based, inter alia, on an objective of protecting the health and welfare of
mankind. It is therefore appropriate to take that objective into account when interpreting that
convention’s provisions. Such an approach is all the more compelling since a reading of the commentary on
the Single Convention published by the United Nations relating to the definition of ‘cannabis’ for the purposes
of that convention leads to the conclusion that, having regard to the purpose and general spirit of that
convention, that definition is intrinsically linked to the state of scientific knowledge in terms of the
harmfulness of cannabis-derived products to human health” (CJEU, 2020, §§73, 74).
Asking the question, “Do the drug Conventions commit governments to implement a prohibitionist regime?”
the CND (2014, p. 1) answers:
“Although the Conventions restrict the use of controlled drugs to medical and scientific purposes, this
limitation should not be considered as the justification for a repressive “prohibitionist” regime, but as the
foundation of a drug control system protecting the health of people from the inappropriate use of narcotic
drugs and psychotropic substances.”
Moreover, this restriction or limitation to medical and scientific use is not absolute. From this limitation to
medical/scientific uses, a series of exceptions and exemptions can be carved out for other purposes.
Highlight: Prohibition is not an obligation.
The idea that prohibition is at the core of the drug control conventions needs to be overcome. Prohibition is only one
option available for States; not mandatory despite persisting beliefs.
2.1.3. “Purposes” in the Single Convention
Article 4 (c) expresses the general obligations of States:
“Subject to the provisions of this Convention, [the parties shall take such legislative and administrative
measures as may be necessary] to limit exclusively to medical and scientific purposes the production,
manufacture, export, import, distribution of, trade in, use and possession of drugs”
3Prohibition is possible for schedule IV drugs, article 2 (5) (b) and for cannabis cultivation, article 22. Article 39 allows
countries to apply “more strict or severe” measures than those of the convention (UN, 1975, pp. 109, 120, 129).
Although this article “could be understood as a requirement for use to be prohibited in national legal systems,
such a requirement is not spelled out” (de Ruyver et al., 2002, p. 20). The CND (2014, p. 3) refers to:
“the popular discourse of equating drug control with a purely “repressive regime” and the limitation to medical
and scientific use with “prohibition”. This is not the vision of the Conventions, which aim at protecting public
health, providing a legislative and normative framework that addresses the use of controlled narcotics drugs
and psychotropic substances within qualified clinical interventions.”
If the conventions establish a framework for medicines, it could be argued that it is a sophism to take from it
that the conventions rule out any non-medicine aspect of drugs. A close reading of article 4 (c) and its
Commentary confirms that the “general provisions” of the Single Convention are two-fold:
it lays down the general legal framework, limited to medical and scientific purposes,
by contrast, it also contemplates “exceptions expressly permitted by the Single Convention” (UN,
1973, pp. 110, 402).
The words “subject to” in article 4 (c) offer a balance between the limitation to medical/scientific purposes
and those “cases in which non-medical consumption or industrial use is exceptionally permitted by the
Single Convention” (UN, 1973, p. 113). This translates by distinct legal regimes:
Drugs for medical or scientific purposes, including traditional medicine: the Convention
extensively codifies medical/scientific uses of narcotic drugs (Bone, 2021, p. 105; Bruun et al., 1974,
p. 46; Rexed et al., 1984, pp. 33–50) for what has sometimes been described as a framework
convention for the medico-pharmaceutical sector (Riboulet-Zemouli, 2022).
Drugs for purposes other than medical and scientific:
Drugs for traditional non-medical purposes used to be subject to special provisions under
article 49 for countries that so desired. The applicability of article 49 lapsed decades ago.
Drugs for industrial non-medical purposes may be exempt under article 2 (9) under
specific conditions (denaturing or preventing abuse and harms, and a reporting to INCB).
Other provisions not related to cannabis (e.g. coca leaf exemption).
Drugs for “special purposes” (“special stocks”) correspond to sensitive governmental stocks used
at their entire discretion for situations which may be large-scale emergencies, military purposes, etc.
Within these regimes, prohibition ultimately always remains a compliant option.
2.1.4. Cannabis cultivation
For the cultivation of cannabis plants, the Single Convention codifies two regimes:
Article 28 (1) requires the same obligations for the cultivation of cannabis as those for the cultivation of
opium poppy (i.e. article 23 of the Single Convention). Only “cannabis and cannabis resin” are mentioned.
Yet, article 28 (1) is generally applied as the standard regime of cultivation of any cannabis for medical
or scientific purposes, including “extracts and tinctures of cannabis” and “cannabis leaves.”
Article 28 (2) exempts some cannabis cultivation from all obligations under the Convention. Although only
“fibre and seed” are mentioned, this regime is generally also applied to “cannabis and cannabis resin” when
they are not used in medicine (for instance, to herbal CBD products and resin; UNCTAD, 2022). Standard
regime for the cultivation of hemp,two possible interpretations of article 28 (2) can be made: either it
exempts only the cultivation of “fibre and seeds” (thus excluding CBD), or it exempts any cultivation for
industrial/horticultural purposes.4
2.1.5. Personal activities
In the Conventions, “decriminalisation” refers to personal activities which are unauthorised or illicit but
upon which no sanctions or penalties are applied. States can opt for “decriminalising” personal activities
under articles 33 and 36 (1) of the Single Convention (see section 1.2.1 above).
However, some personal activities can be authorised or licit (e.g., personal use or possession for medical and
scientific purposes). Legal regulation would vest personal activities for non-medical use with a legal status
equivalent to medical/scientific purposes rather than a status of decriminalisation under articles 33 and 36
(1). States can opt for “legalising” personal activities under one of the options presented in section 3.
For Marks (2019, p. 222), personal activities could be defined as “everything that pertains to the
constitutionally protected realm of the private.
2.2. International law: non drug-related
In designing domestic regulations for non-medical cannabis, attention should be given to other
legally-binding instruments, which can include obligations and/or guidance for certain licit activities
involving cannabis.
In addition to international and European human rights law, when legislating over a plant, accounts should be
taken of binding instruments in other fields than drug control, such as the Framework Convention on
Climate Change, the Convention on Biological Diversity and its Nagoya Protocol, and the International
Treaty on Plant Genetic Resources for Food and Agriculture (all ratified by all EU Member States and the
EU itself), conventions of the International Labour Organisation (ILO), the “Geneva Act” Appellations
treaty, among others relevant international instruments (FAAAT, 2021; Riboulet-Zemouli and Krawitz,
2021; Wyse and Luria, 2021).
2.3. EU acquis
Since the 7-point action plan of French President Mitterrand (1989), drug policies have become increasingly
discussed at the EU level. Pertaining to all three pillars before 2009, the Treaty of Lisbon has only confirmed
the transversality of drug-related policies across EU legal areas (Boekhout, 2004, p. 53–55).
Today, “[t]here is no harmonised EU law on cannabis use” (EMCDDA, 2018, p. 9) which is reflected in a
broad diversity of domestic approaches (Boekhout, 2004; EMCDDA, 2016; Mravčík, 2015; Sánchez-Avilés,
2020). Nevertheless, provisions affecting domestic cannabis regulations are present in several EU laws and
4The first of these two interpretations (limiting the scope of this article to fibre and seeds) is quite unorthodox in
international law (McNair, 1961, pp. 399–410) and opposes the rules of botany: which legal regime would apply to the
cultivation for the leaves, or for the non-fibre parts of the stem (hurd/shives) which not mentioned in any of these two
provisions on cultivation? It also seems to go against the intent of the drafters of the Single Convention
(Riboulet-Zemouli, 2022, pp. 54–59; UN, 1973, pp. 312, 314).
One of the critical findings in this brief, about EU drug control law, is that the international drug
control Conventions have become ipso facto part of the Union acquis” (Erlbacher, 2019, p. 1648) due to
their mention in many Union acts and CJEU Case law.5Two types of mentions exist:
-Some EU acquis are directly applicable to the substances that are listed in the Conventions or rely
upon and complement legal provisions from the Conventions (e.g., the Schengen acquis, or the 2004
Framework Decision, both analysed below).
-Other pieces of EU legislation exclude from their scope the substances listed in the drug
Conventions (for instance, Regulation No. 178/2002 laying down the general principles and
requirements of food law in the EU, which excludes from the definitions of “food” any “narcotic or
psychotropic substances within the meaning of” the drug control Conventions). Because they
exclude cannabis from their scope, these parts of EU law are not relevant (see also section 3.2.2.(ii)).
2.3.1. Primary legislation
In primary EU law, article 83 (1) of the Treaty on the Functioning of the European Union (TFEU) mandates
the European Parliament and the Council on the “areas of crime” of “illicit drug trafficking.” Article 168
gives the EU the competence to “complement the Member States’ action in reducing drugs-related health
damage” while respecting “the responsibilities of the Member States for the definition of their health policy.”
Integrated into the framework of the EU, the Schengen acquis commit Parties to coordinate their action
against “illicit drug trafficking” (articles 8 and 9 Schengen Agreement, articles 71–76 Implementing
Convention; analysed in: van Kempen and Fedorova, 2019a, pp. 103–112). Article 71 (1) calls for “measures
to prevent and punish the illicit trafficking” to be adopted in accordance with the international drug control
Conventions. The Schengen acquis also contain provisions for Parties:
To seek to “harmonise laws and regulations” on narcotic drugs (article 19 Schengen Agreement);
To “notify each other of the measures taken in order to monitor the legal trade” in drugs (article 76
(3) Implementing Convention);
“Not to jeopardise” stricter control measures in the territory of other Parties (art. 76 (1),
Implementing Convention), notably if a country “departs from the principle referred to in Article
71(2) in connection with its national policy on the prevention and treatment of addiction to narcotic
drugs and psychotropic substances” (Final Act, Implementing Convention).
Other pieces of primary legislation also focus on illicit activities such as the 2000 Convention on Mutual
Assistance in Criminal Matters between the EU countries (EUCMA).
2.3.2. Secondary legislation
Three pieces of secondary EU legislation support this treaty core:
Council Decision 90/611/EEC of 22 October 1990 accessing the 1988 Convention on behalf of the
EU, although with a competence limited to trade in precursors (see section 2.1.1).
5The implicit requirement for States to access the drug control Conventions before joining the EU (under Chapter 24)
seems to corroborate that ipso facto incorporation into EU acquis.
Regulation (EU) 1307/2013 of 17 December 2013, establishing rules for direct payments to hemp
farmers under the common agricultural policy in its articles 32 (6) and 35 (3).6Hemp corresponds to
certain varieties of cannabis plants, as listed in a catalogue.
Regulation (EU) 1308/2013 setting up common markets in hemp products and conditions of import.
Council Framework Decision 2004/757/JHA (the “Framework Decision”, see section 1.2.2), the
central piece of secondary legislation to consider in the context of domestic cannabis regulations (see
section 1.2.2 ; Council, 2004).
The Framework Decision 2004/757/JHA especially has been extensively analysed (Deutscher Bundestag,
2022b, p. 5; Jelsma, 2022; van Kempen and Fedorova, 2019a, p. 115).
Indicative as to its goals, the explanatory memorandum refers to the Framework Decision as an instrument
targeting “the transnational dimension of the offence” of illicit drug trafficking and as “a further development
of the law enforcement side of the Schengen acquis” (European Commission, 2000).
Its Article 1 (1) mentions the drug control Conventions, relying on their definitions of “drugs” (although it
departs from it in article 1a, adding new psychoactive substances to the definition of “drugs” in EU law). In
article 4, the Framework Decision features another difference with the standard of the conventions: a
proportionality of penalties grounded on quantitative and qualitative distinctions (“large quantities” and
“those drugs which cause the most harm to health”).
As discussed in section 1.2.2 above, activities which would otherwise be punishable as illicit drug trafficking
are excluded from the Framework Decision when committed for personal consumption. Article 2 (1) further
stipulates that the scope of the Framework Decision is limited to activities committed “without right.”
Activities which are based on a right are not specified. It can be assumed that they include the following:
Rights contained in EU law, for instance:
hemp cultivation under Regulation (EU) 1307/2013 and use of hemp products under
Regulation (EU) 1308/2013 (and others on cosmetics, foods…),
activities related to cannabis medicines under Regulation (EC) 726/2004 and Directives
2001/83/EC and 2004/24/EC,
activities of police or customs officers (e.g. pursuant to relevant Schengen dispositions, to
the EUCMA…);
Rights contained in domestic legislation, such as:
activities undertaken by authorities, for example, in handling seizures of drugs,
activities related to personal consumption, under article 2 (2) of the Framework Decision,
activities of coffee-shops in the Netherlands (a critical point).
There has long been a debate as to the extent to which a right covers coffee-shops in that sense, and how that
might extend to legally-regulated contexts. There is a profound ambiguity as to the exact meaning of
“without right,” although an indication can be found in the explanatory memorandum of the first draft
proposed by the European Commission (2000): “Essential criteria […] are the notions of acting ‘for profit’
and ‘without authorisation’.” The terms “for profit” have today been replaced by the concept of personal
purposes, and “without authorisation” by “without right.”
This debate extends to the structural question of non-medical cannabis regulations and to the conditions
under which domestic regulations would constitute a right in the meaning of the Framework Decision.7
7Jelsma (2022, p. 95), and van Kempen and Fedorova (2019a, p. 134) discuss this. Two EU Directives –not related to
cannabis– state precise definitions of without right: “not permitted under national law” (Directive 2013/40/EU),
6Contrary to impressions, hemp crops or derived products which do not meet the criteria of this Regulation are not
automatically illicit, but simply the rules set out in the Regulation do not apply (as confirmed in CJEU, 2020, §58, §96).
Although considered central to cannabis policy, the Framework Decision deals exclusively with illicit
activities and is focused on “develop[ing] the Schengen acquis from the perspective of criminal law” (van
Kempen and Fedorova, 2019a, p. 116). It provides little guidance regarding licit activities (the Regulations
on hemp and medicines not being mentioned, for example). It may have limited relevance to
lawfully-regulated activities in a Member State.
2.3.3. Interpretation, non-legislative acts
As part of the acquis, the Conventions must, therefore “be taken into account for the interpretation of
(secondary) Union rules” (Erlbacher, 2019, p. 1648). Numerous illicit activities in EU law remain defined
by direct or indirect reference to the international drug control Conventions (a situation similar to the
1988 Convention which criminalises activities which are “illicit” according to, or “contrary to the
provisions” of the 1961 and 1971 conventions).
Joint Action 96/750/JH of 17 December 1996 on the “approximation of the laws and practices […] to combat
drug addiction and to prevent and combat illegal drug trafficking” should be considered for matters of
interpretation (van Kempen and Fedorova, 2019a, 112–115). It was repealed in 2016, and is therefore not
binding. But the Framework Decision which replaced the Joint Action “is entirely in keeping with [its]
spirit” (European Commission, 2000). That spirit, again, was the coordination and cooperation of authorities
in the “fight against drug addiction” and illicit trafficking. The Joint Action also stated:
“Member States shall ensure that their obligations under the United Nations Conventions on narcotic drugs and
psychotropic substances of 1961, 1971, and 1988 are applied strictly and effectively” (Council, 1996).
In line with the Schengen acquis, the Joint Action called the Member States to “endeavour to approximate
their laws to make them mutually compatible to the extent necessary to prevent and combat illegal drug
In this regard, article 19 of the Implementing Convention (placing drug laws among the fields of future
harmonisation) is an essential to interpreting EU cannabis-related acquis throughout the Schengen area.
Finally, there are non-legislative acts of the Union to take into consideration:
Council Resolution of 29 November 1996 on “measures to address the drug tourism problem,” still
in force today, lays down a series of recommendations for administrative cooperation, noting that the
phenomenon differs in every country. “Drug tourism” is also mentioned in the repealed Joint Action.
Council Resolution of 16 December 1996 on “measures to combat and dismantle the illicit
cultivation and production of drugs” is mainly focused on preventing “the use of cannabis seeds for
illicit cultivation.” Still in force today, it invites Member States “to ensure the banning of the
cultivation of cannabis under glass, under polythene tunnels, and indoors, with the exception of
such cultivation, for example for the purposes of scientific research, where a special licence has been
obtained […]”8
Council Resolution 97/C 10/02 of 20 December 1996 on “sentencing for serious illicit drug
trafficking” is no longer in force since it became obsolete when the Framework Decision was
adopted, but it is referenced in some pieces of law still in force today.
Despite its straightforward title, Council Resolution on Cannabis of 26 July 2004 is not a general
framework for cannabis. It mostly takes note of developments in the illicit cannabis market and
8This provision may have a place in the discussion on the agricultural models to adopt for legal cannabis crops.
“medical, scientific or similar purpose […] activities carried out under domestic legal powers” (Directive 2011/93/EU).
However, these definitions are always limited in their scope to the specific context of the Directive containing them.
requests various research and analyses to Europol, the EMCDDA and the European Commission. It
calls on Member States to increase youth prevention, and capacity-building among parents, teachers
and healthcare professionals. Notably, it is the only piece of text which does not use the term illicit,
calling on Member States to take measures against cultivation and trafficking of cannabis within the
Although non-legislative acts, these documents have influenced numerous CJEU decisions, which form part
of Union law.
2.3.4. Case law
In Josemans, the CJEU upheld specific restrictions to the fundamental freedoms about cannabis, finding
“that a Dutch municipal authority could legitimately require ‘coffee-shop’ owners selling cannabis to deny
entry to non-residents” (Tomkin, 2019, p. 727). For Craig and de Búrca (2011, p. 798), this is surprising “in
view of the consistent previous case law” related to other activities for which Member States can justify
special restrictions (or not) on grounds of public policy (ordre public), ethics or morality, as is the case for
drugs (CJEU, 1999a; 2010; de Witte, 2013).
There are two notable differences in CJEU Case law between cannabis and other sorts of activities where
ethical or moral approaches vary between the Member States:
- Several EU acts call for States to take measures against drug tourism, which is not the case for other
activities restricted on the grounds of public policy, morality or ethics, that the CJEU analysed,
- The fait accompli is that currently, all Member States prohibit cannabis, which is also different
from other activities. The Court stated that “[w]hile Member States had diverse approaches to
regulating prostitution, there was a common prohibition on the trade of narcotics” (Tomkin, 2019, p.
727). Similarly, the CJEU noted that lotteries, while restricted in some Member States, are not
“activities whose harmful nature causes them to be prohibited in all the Member States” (Craig and
de Búrca, 2011, p. 798).
In addition, it should be noted that the Court does not seem to consider prohibition as an international or EU
legal obligation, but rather as a “legal position [which] complies with” the Conventions (CJEU, 2010,
§37; 2020, §60). It stems from Court rulings that there is a prohibition in all the Member States on marketing
cannabis “under international law and European Union law” (CJEU, 2010, §77), which should be construed
as meaning in compliance with, not under mandate of. For the Court, it is on the initiative of Member
States that prohibition is implemented, on the grounds of “harmfulness” (CJEU, 1982, §8; 1988, §17;
2010, §36; 2020, §59; see also section 2.1.2. above). The CJEU (2010, §41, §78; 2020, §61) understands that
drugs are prohibited “because of their very nature,” not because of treaty obligations:
“since the harmfulness of narcotic drugs, including those derived from hemp, such as cannabis, is generally
recognised, there is a prohibition in all the Member States on marketing them, with the exception of strictly
controlled trade for use for medical and scientific purposes” (CJEU, 2010, §36).
There are some inconsistencies in the address of cannabis by the CJEU. In most Case law related to criminal
proceedings under domestic drug legislations, the Court only mentions options of licit trade for cannabis in
the context of “medical and scientific” or “pharmaceutical and medical purposes” (CJEU, 1982, §8; 1988,
§17; 2010, §36) but failed to mention hemp as another strictly controlled trade under EU law. However, in
the same judgements, sometimes in the same sentence, the Court referred to cannabis as hemp-based
drugs” (CJEU, 1988) and “narcotic drugs […] derived from hemp” (CJEU 2010, §36; 2020, §59).
This is slightly startling, taking into account the parallel Case law where, for example, the CJEU prohibited
quantitative restrictions and measures having equivalent effect on hemp (CJEU, 2003), including on CBD
extracted from hemp (CJEU, 2010), insisting that CBD “extracted from the Cannabis sativa plant in its
entirety [including] its flowering or fruiting tops […] is not a drug within the meaning of the Single
“Community law must be interpreted as precluding national legislation which has the effect of prohibiting the
cultivation and possession of industrial hemp” (CJEU, 2003)
“Articles 34 and 36 TFEU must be interpreted as precluding national legislation which prohibits the marketing
of cannabidiol (CBD) lawfully produced in another Member State when it is extracted from the Cannabis
sativa plant in its entirety and not solely from its fibre and seeds, unless that legislation is appropriate for
securing the attainment of the objective of protecting public health and does not go beyond what is necessary
for that purpose” (CJEU, 2020).
Zigzagging rulings, differences with settled Case law on similar activities, and a unique situation (the
generalised prohibition among Member States), in addition to evolving doctrines at the CJEU (de Witte,
2013, pp. 1558–1576), suggest that the Court’s opinion is likely to evolve as domestic laws do. Increased
reliance on procedural proportionality would balance and mediate the progressive normalisation of cannabis
in part of the Union.
2.3.5. Treaty-making competencies of the EU
Treaty-making9in the field of drugs is the most recent field of expansion of the EU’s competencies. Article
218 TFEU provides the ordinary treaty-making procedure, essentially in the hands of the Council. These
dispositions are also valid for agreements that have not been directly accessed by the Union but by Member
States, like the Single Convention (CJEU, 2014; Erlbacher, 2019, p. 1661; Schutze, 2018, pp. 207–212).
In 2017, a precedent confirmed that the EU’s treaty-making capacity applies to the drug control
Conventions. Article 218 (9) and 83 (1) TFEU were interpreted as subjecting the actions of EU Member
States during CND scheduling votes to an imperative mandate (Council, 2017a; Sánchez-Avilés, 2020). This
fact was clarified last year when the European Commission (2021) activated an infringement procedure
against Hungary for breaching these rules.10
Modalities of the treaty-making competencies of the Union depend on the area in which the alteration
of an international agreement relates. The general procedure is as follows: The European Commission
submits recommendations to the Council ahead of the opening of negotiations (for common foreign and
security policy, the High Representative). The Council decides by qualified majority (except if the
amendment covers a field for which unanimity is required for the adoption of a Union act) to nominate the
negotiating team, to authorise the opening of negotiations, to adopt negotiating directives, and to designate a
special committee to follow negotiations. It also authorises the signing of the amendment and its conclusion.
The European Parliament needs to be fully informed throughout the procedure, and, in some cases, its
consent is required before concluding the agreement.
10 The EC opened the procedure on “Hungary for failure to follow the Union’s position when voting on the WHO
recommendations on cannabis and cannabis-related substances at the UN Commission on Narcotic Drugs in December
2020. […] The Union position –adopted by the Council in November 2020– is binding on EU Member States, who have
to vote accordingly in the Commission on Narcotic Drugs, in line with Article 218 (9) TFEU.”
9Capacity to negotiate, sign, conclude, accede, amend, suspend, terminate international agreements. competencies of
Member States, the Union, or both, are defined in TEU (Title V; Final Act) & TFEU (articles 2 (1), 3, 207 (3), 216,
Notably, by derogation to article 218 TFEU, articles 3 (1) and 207 mandate a special procedure for
treaty-making on trade-related agreements giving more leeway to the European Commission during the
“[T]he EU is exclusively competent in a few policy areas (most prominently external commercial relations,
broadly defined […]) and shares treaty-making competencies with its Member States in various other policy
areas, or supplements those Member State powers” (Klabbers, 2015, p. 54).
By contrast, in Germany v Council, the CJEU (2014) ruled that:
“it is clear […] that the situations in which the Union has an external competence, in accordance with [Article
216 (1) TFEU], are not limited to the various scenarios set out in [Article 3 (2) TFEU].”
Highlight: Policy options and EU competencies.
The ascertainment of competencies is determinative regarding the policy options chosen in case of lex ferenda
domestic cannabis regulations (i.e. requiring a modification of the conventions or a change in the ratification
status), which would alter the rules and framework of a part of Union acquis). Crucially, the legal basis for
treaty-making competencies can be challenged by Member States and, in some cases, by EU institutions.
The Union has a role to play in the event of a lex ferenda option (modifications of the rules and framework of the
Union). This is not necessarily the case for lex lata approaches that do not alter the formal framework, and rely
on the exclusive State competence of treaty interpretation.
In Commission v Council (Appellations Agreement), the CJEU (2017) ruled that a failure to respect the
specific procedure for negotiating trade-related agreements lead to annulment of the Council Decision that
had opened the negotiations, devoiding the Decision of legal effect. The negotiation of any international
agreement concluded having a “direct and immediate [effect] on trade” (CJEU, 2017, §74) falls within the
exclusive competence of the EU. In Commission v Council (Rotterdam Convention), the CJEU (2006)
“The choice of the legal basis for a Community measure, including one adopted with a view to conclusion of
an international agreement, must be based on objective factors which are amenable to judicial review and
include in particular the aim and content of the measure. If examination of a Community measure reveals that
it pursues a twofold purpose or that it has a twofold component and if one of those is identifiable as the main or
predominant purpose or component, whereas the other is merely incidental, the act must be based on a single
legal basis, namely that required by the main or predominant purpose or component.”
11 Options are presented here, encompassed within six general paths (A–F). This selection offers the
opportunity to align the design and implementation of domestic cannabis regulations in compliance with
international and EU obligations. Because of the interconnectedness of EU acquis with the international drug
control conventions (section 2.1), the two legal frameworks should be considered jointly.
Among the options, some require textual changes in pieces of international law, or changes in the status of
ratification, prior to implementing domestic cannabis regulations (lex ferenda; options A–C). Others
consider that compliant regulations can be achieved without the need to formally modify the drug control
Conventions (lex lata; options D–F).
Political pathways can also be categorised as follows (Deitelhoff & Zimmermann, 2013; Wunderlich, 2014):
Options A, C and to a lesser extent D, convey a rebuttal, critique or challenge of the mere validity of
all or part of the Conventions. These options might be at odds with Title V TEU, and repeated calls
by European leaders to protect and reinforce multilateralism (Macron et al., 2021);
Options B, E, and F are characterised by a proactive attitude of norm entrepreneurship,
constructively challenging some enforcement aspects of the Single Convention without questioning
its validity as a whole: relying on its provisions and claiming compliance.
With these two sets of criteria, the following matrix of options A–F can be drawn (indicative only):
In the matrix above, in yellow, lex lata reform proposals (laws can be adopted within the current Single
Convention) are on the left side. Lex ferenda options (changes in international law or ratification status are
required prior to regulation) are on the right side, in blue.
This visual presentation is similar in the flowchart summarising options A–F on the next page:
3.1. Options A: Regulation after denunciation of the Single Convention
There are two approaches to a withdrawal from (= denunciation of) the Single Convention.
3.1.1. Option A1: remaining outside
Withdrawing from the Single Convention (art. 46) and remaining outside, thus departing from all
commitments (including questions unrelated to cannabis). States are relieved from international obligations
six months after notifying their withdrawal.
This option presents a high likelihood of negative impacts on the licit trade in medical drugs and on the
cooperation in the field of law enforcement and crime prevention (Don, 2014, pp. 234–235; Jeanroy et al.,
2022, p. 22). From a technical standpoint, this option will not resolve all issues. It might be left partly
without effect insofar as the 1988 Convention refers to illicit activities as those defined under the Single
Convention. A withdrawal from the 1988 Convention would, therefore, also likely be needed.
(i) Precedent
There is no precedent. Beyond the question of drugs, withdrawing from international treaties has been
fueling several crises of multilateralism in recent years.
(ii) EU legislation
Because the Single Convention is ipso facto part of Union acquis, the mere legality of a withdrawal is
questionable. Regardless, as in the case of the 1988 Convention, the practical effects of a withdrawal risk
being superseded by the continued presence of references to the conventions throughout EU drug legislation.
(iii) Intermediary recommendation
Withdrawing from a legal agreement ratified by 186 countries would also likely incur high political costs and
be perceived as undermining the rule of law as a whole: “the potential ramifications of withdraw from the
Single Convention outweigh any benefits” (Don, 2014, p. 235). Option A1 is not recommended.
3.1.2. Option A2: re-accessing
Withdrawing from the Convention (art. 46) and re-accessing with a reservation on cannabis (art. 50).
Reservations to the Single Convention under article 50 (3) can only be made by a “State which desires to
become a Party.” The Convention “expressly prohibits the statement of reservations following initial
accession to the treaty” (Don, 2014, p. 235). Hence, the Convention needs to be denounced before a
reservation for cannabis could be placed, upon re-accession. The reservation and re-accession becomes
effective if it is not “objected to by one third of the States” within a year.
(i) Precedent
Bolivia left the Single Convention on 29 June 2011 (with effect 1 January 2012) under article 46 (2). It then
re-accessed on 29 December 2011 with a reservation for the cultivation, trade, possession and use of coca
leaves “for cultural and medicinal purposes,” but its accession took effect on 10 February 2013 (UN
Secretary-General, 2013), effectively remaining fully outside of the Single Convention for 405 days.
15 countries objected to Bolivia’s reservation (including six EU countries) but “fell far short of the
one-third of parties (sixty-two) required to block it from becoming effective” (Crook, 2013, p. 462).
(ii) EU legislation
The same hurdles as those outlined in Option A1 would arise. An EU Member State withdrawing from the
Conventions risks infringement of European law for a prolonged period.
Reservations made upon recession raise the questions of the EU’s treaty-making competencies (section
2.3.5): a Member State may be at risk of not being the sole decision-maker of the modalities of its
reservation. A minimum, information to the Commission and other EU Member States, and possibly
consultations within the Council, would be needed (see also below Option B1).
Previous objections made by EU Member States to Bolivia’s reservation in 2011 are an additional negative
precedent, and “the unique circumstances requiring Bolivians to protect coca as cultural patrimony may not
be transplantable to emerging cannabis policy” (Bone, 2021, p. 90).
(iii) Intermediary recommendation
This Option has often been presented as giving Member States back their full sovereignty. However, in the
European context, it seems that any reservation upon re-accession would be subject to consultations within
the Council, likely triggering greater involvement of the Union, and possibly conditioning the terms of the
reservation. It also carries similar political costs as Option A1 (and other validity contestation approaches)
and would leave the Member State a year or more in infringement of EU acquis.
Therefore, Option A2 is not recommended.
3.2. Options B: Regulation after internal amendment of the Convention
The Single Convention contains two modalities of amendment, “the two least controversial options […]
which use uncontested and established treaty procedures” because they “show respect for treaty procedures
and willingness to discuss emerging dilemmas with treaty parties” (Jelsma, 2022, p. 97, 98). Article 47 of the
Convention provides for amendments to the main body of text; article 3 for amendments to the schedules.
3.2.1. Option B1: amending articles
Amendment under article 47 is a priori the most logical option. By adopting an amendment to enable
domestic cannabis regulations, “those countries who wish to pursue such legislation would be permitted to
do so, and those countries who remain in opposition would be able to remain parties to the original treaty”
(Don, 2014, p. 237). Two pathways exist:
Silent procedure: a draft amendment is circulated to all state parties. If no objection is raised within
18 months, the amendment enters into force for all parties. This procedure is unlikely to happen for
Diplomatic conference: where the draft amendment would be negotiated, with provisions adopted
with a two-thirds majority. States are free to attend and stay or not at the conference. “Once an
amendment has been proposed and adopted, parties are free to decide if they will become a party to
the amendment” (Don, 2014, p. 236).
The applicability of this option to cannabis in the existing geopolitical context has long been questioned
(Bruun et al., 1975, p. 52; Jelsma et al., 2018). Different proposals on what to amend have been made:
Modifying article 47 to facilitate reservations without the need for prior denunciation (Riedel, 2017,
pp. 742–746; Zitt, 2016),
Modifying article 49 to extend the temporary reservation (Riedel, 2017, pp. 740–742),
Modifying article 2 (9) (Riboulet-Zemouli, 2022, p. 120),
Altering a series of provisions throughout the Convention (Room and MacKay, 2012).
(i) Two precedents
In 1972 the Single Convention was successfully amended, following a procedure initiated in 1971 by the
United States (Bruun et al., 1975, p. 18; Room and MacKay, 2012, p. 9). It entered into force in 1975.
In 2009, Bolivia proposed amending the Convention to delete its provisions barring coca leaf. This proposal
did not gain sufficient support and failed” (Crook, 2013, p. 461). Two years later, Bolivia opted for
denunciation under Option A2 (Riedel, 2017; Room and MacKay, 2012, p. 10; Zitt, 2016).
(ii) EU legislation
Given precedents on voting procedures at the CND (see Option B2 below), an amendment of the Single
Convention is likely to “affect common [EU] rules and alter their scope” (Article 3 (2) TFEU), hence falling
under exclusive competence of the Union (see section 2.3.5 above). In this context, the amendment would
bind all Member States (article 216 TFEU). The amendment may also be formulated so that the EU
could access it.
If it is framed outside of the Union’s exclusive competence, an amendment process may not be possible in
closed-loop. It would likely be subject a minima to “prior consultations within the Council” (article 28 (3)
TEU) and in any event to consultation, cooperation, and an approach of solidarity with other EU Member
States (art. 32 TEU). Incorporating the amendment into Union rules risks being problematic if not all
EU Member States join it. Such an approach of shared competencies may be attempted, although it is also
likely to weaken the overall negotiating position if the Member States are divided (Erlbacher, 2019, p. 1651)
as is the case on this topic.
The steps to take by an EU Member States to initiate an amendment process would depend on the area of
competence within which it is framed.
The process under article 218 TFEU (see details in section 2.3.5. and appendix 2) is informed by prior
recommendation of the European Commission or High Representative. The Council nominates the
negotiating team and issues negotiating directives. Decisions are taken by the Council, usually via qualified
majority (unanimity is required in areas where common acts are adopted by unanimity). The European
Parliament must be kept informed, and its prior consent is sometimes required before concluding. If the
amendment was framed as a trade agreement, special provisions under article 207 TFEU would apply,
giving more prominence to the Commission in the negotiating process (it nominates the negotiator
instead of the Council). A qualified majority is also required, except in some cases involving trade in
services, intellectual property, and direct foreign investment, where unanimity may be required.
The legal basis for entering the negotiation of an amendment could be challenged by a Member State. In
addition, under article 218 (11) TFEU, “the European Parliament, the Council or the Commission may obtain
the opinion of the Court of Justice” on the compatibility of the amendment with EU treaties.
(iii) Intermediary recommendation
Option B1 is not recommended in the short term but could be recommended in the longer-run once
significant support is obtained (at least equivalent to a qualified majority vote).
3.2.2. Option B2: amending schedules
Amendment of the schedules under article 3. The modification of drugs’ listing in the schedules is a
process of amendment (articles 1 (1) (u) and 3; McNair, 1961, p. 748; UN, 1973, p. 29). Lowering cannabis
within the schedules would alter the modalities of its control for medical/scientific purposes. Entirely
withdrawing cannabis from the schedules would impact domestic (non-medical) cannabis regulations much
This option has an important legal hurdle: even in the case of a full descheduling, cannabis might remain
subject to the obligations of schedule I because it is mentioned explicitly in article 2 (6) that cannabis is
subject “to the measures of control applicable to all drugs in Schedule I” (Riboulet-Zemouli and
Krawitz, 2022, pp. 11–12). This sentence would require an amendment (under article 47, see Option B1
above) to release cannabis effectively from the obligations attached to Schedule I.
Member States can request the WHO to re-assess cannabis but cannot influence the outcome. The criteria
WHO uses to issue its recommendations require that any cannabis-like substance be listed in the schedules,
making it difficult to recommend a withdrawal (see Jelsma, 2022).
Finally, this option is the only proposal whose outcome would be binding upon all State Parties to the
Convention, even those disagreeing with a withdrawal. According to the Commentary:
“The Schedules may be amended in a different way than the other parts of the Single Convention. A special
procedure, that of article 3, is provided for their revision. Amendments of the Schedules, but not that of other
sections of the Single Convention, can become binding on Parties to that treaty without their express or implied
consent” (UN, 1973, p. 29).
(i) Precedent
In 2020, schedule IV was amended by a CND vote, and “cannabis and cannabis resin” was deleted from it
(remaining in schedule I). The scientific review process by WHO leading to the vote had been initiated four
years earlier (Riboulet-Zemouli and Krawitz, 2022) and yielded a total of 9 recommendations. Removal from
schedule IV was the only recommendation accepted: one did not call for a vote, three were rejected and four
were not voted upon.
(ii) EU legislation
Amendments to the schedules “have direct repercussions on the scope of application of Union law in the area
of drug control” and are “directly incorporated into common Union rules” (Council, 2017b):
“The CND’s scheduling decisions also have legal effects in the EU legal order by virtue of Union law, namely
Framework Decision 2004/757/JHA. Changes to the schedules […] have direct repercussions for the scope of
application of this EU legal instrument” (Council, 2017a).
The decision-making procedure is now well-established through the Horizontal Working Party on Drugs.
There is an unintended consequence of this Option in EU law: if cannabis was entirely withdrawn from the
schedules, it would no longer be “a narcotic drug” in the meaning of the Single Convention. In consequence,
the pieces of EU legislation which were not applicable, because they exclude narcotic drugs from their scope,
could start becoming relevant to unscheduled cannabis. This is for instance the case for the general principles
and requirements of food law in the EU (laid down in Regulation No. 178/2002). See also the introduction
of section 2.3.
(iii) Intermediary recommendation
If the WHO recommended a withdrawal, the procedure would be similar to the one followed for
scheduling decisions at CND every year. It is unlikely that there would be a supporting majority at CND.
Still, in the event of a majority in favour of descheduling, a procedure under article 47 would have to be
made, likely under exclusive EU competence (see Option B1).
The technical difficulties linked to treaty provisions, and the polarised discussions around previous cannabis
recommendations in 2018–2020, suggest that further efforts to deschedule may not be achieved easily or
quickly, and are “unrealistic” (Jelsma, 2022, p. 98). Finally, because of its unintended consequences at EU
level and since it is the only option which would be binding upon all Parties, Option B2 is not
3.3. Options C: Regulation after external amendment of the Convention
Another option consists in the conclusion of a new legal agreement intended to amend, complement, or
replace the Single Convention. Here also, there are two approaches.
3.3.1. Option C1: new convention
There have been proposals for new conventions intended to supersede the existing ones for countries
adopting the new convention(s)” (Room and MacKay, 2012, p. 68).
(i) Precedent,
Like other regional instruments, the Shanghai Cooperation Organisation’s 2004 Agreement on combating
illicit trafficking could be considered as a precedent in this regard.
(ii) EU legislation
The arguably strenuous idea of an EU-specific “drugs convention” may resolve many issues outlined in
section 2. In practice, the process outlined in sections 2.3.5 and 3.2.1.(ii) is likely required.
(iii) Intermediary recommendation
Because it remains vague and has not been subject to any research, Option C1 is not recommended.
3.3.2. Option C2: inter se amendment
often called inter se amendment,” it consists in amending the Single Convention without relying on the
procedure of article 47, but instead, on the 1969 Vienna Convention on the Law of Treaties (VCLT). The
inter se option for treaty modification” proposes that:
“a group of two or more like-minded states could conclude agreements among themselves that permit the
production, trade, and consumption of cannabis for non-medical and non-scientific purposes” (Jelsma et al.,
2018, p. 7)
This option “require[s] a clear commitment to the original treaty aim to promote the health and welfare of
humankind and to the original treaty obligations vis-à-vis states not party to the agreement” (Boister and
Jelsma, 2018). There are advantages to the inter se amendment generally (Groenendijk, 2011): besides the
improvement of the textual framework (like in Option B1), the mechanism could allow overcoming the
political divide and altogether avoid the need to discuss the text of an amendment with pro-status quo States,
finding “a balance between the stability of treaty regimes and the necessity of change in absence of
consensus” (Boister and Jelsma, 2018, p. 493).
Technically, some legal questions remain open. If the amendment mechanism of article 47 of the Single
Convention were binding upon all Parties, recourse to an external inter se amendment process would be
legitimate. Nevertheless, reliance on external mechanisms seems unnecessary because article 47 already
allows for amendments to the Single Convention between a group of States, only binding upon them.11
Finally, the VCLT (ratified by 25 member States but not by France and Romania) took effect in 1980: some
argue that it “cannot, in any case, apply to the Single Convention because of its non-retroactivity clause
(Dawkins, 1973, p. 364; see also McDade, 1986). Therefore, the process should rely upon customary
international law instead of the VCLT, making it more delicate.
(i) Precedent
There is no precedent.
(ii) EU legislation
This method shares many challenges with Option B1 (section 3.2.1) with added specificities. Acquis and
practice support the conclusion of inter se agreements as a matter of principle; ordinary EU
treaty-making provisions (article 218 TFEU) do not apply to inter se agreements concluded between Member
States of the Union (Erlbacher, 2019, p. 1661):
“The process of European integration is full of examples that started as inter se cooperation before being
embedded in the EU legal order. […] Yet, the conclusion of inter se agreements in the past was almost
exclusively pursued in areas lying outside the scope of EU law, or at least in areas where the Union had not
taken any action at all” (Dimopoulos, 2015, pp. 288–289).
In contrast, two substantive alterations of the Union’s legal framework contributed to “rendering the use of
inter se unnecessary” (Dimopoulos, 2015, p. 290), calling for caution in addressing inter se in present
The Treaty of Amsterdam introduced the enhanced cooperation mechanism (now Article 20 TEU
and 326–334 TFEU). Regarded as a measure of last resort, enhanced cooperation can apply between
at least 9 countries, outside of the areas of common commercial and common foreign and security
The Lisbon Treaty deleted previously-existing provisions facilitating inter se between EU Member
States (an exception for Benelux countries has remained, in article 350 TFEU).
EU Member States can still conclude inter se agreements under strict conditions, except in the cases of
exclusive competence of the Union (Dimopoulos, 2015). However, the principle of loyalty (article 4 (3)
TEU) “prohibits undermining or circumventing the mechanisms for cooperation provided by the Treaties,
11 Consequently, it may be legitimate to consider such external mechanism for an amendment of the schedules inter se.
such as enhanced cooperation” (Klamert,2019, pp. 53–54). Except for Benelux countries, enhanced
cooperation is likely to be seen as the preferential mechanism over the conclusion of inter se
agreements. Requests to initiate enhanced cooperation are made to the European Commission, which may
or may not submit it to the Council. The latter grants authorisations to proceed, after consent of the European
Parliament (article 329 TFEU).
So far, proposed inter se amendments have been framed in terms of legal cross-border trade in non-medical
cannabis (Boisteer and Jelsma, 2018; Jelsma et al., 2018; Jelsma, 2022, pp. 99). If involving non-EU
countries, this would account in effect to a multilateral trade agreement, an area of exclusive EU
treaty-making competence, giving a greater role to the European Commission in the negotiations (see
section 2.3.5 and appendix 2; Erlbacher, 2019, pp. 1646–1665) in a process which usually takes several
years and “involves over 30 stages” (European Commission, 2018). The principles of pre-emption and
conferral should be borne in mind (see Dimopoulos, 2015; Schütze, 2006).
(iii) Intermediary recommendation
Although sharing many challenges with the regular amendment procedure (Option B1), this approach has
some comparative benefits politically. In purely legal terms, however, Option B1 is preferable. In the event
of a failed attempt with Option B1, the inter se option may take more relevance (Jelsma, 2022, p. 98) than if
it was pursued as a first-choice option.
Suppose the inter se was to engage in trade relationships with non-EU third countries, the area of common
commercial policy risks being triggered. Avoiding it would require the conclusion of a non-trade-related
agreement. If engaging only EU States, the amendment could, for instance, address the internal market of
cannabis on the grounds that the Union has not yet done so. The enhanced cooperation would appear as the
optimal basis to proceed; it also presents better safeguards and offers minimum rights to non-participating
Member States (Dimopoulos, 2015, p. 302).
Unless limited to Benelux, unless Option B1 has previously failed, or unless there are nine EU Member
States to start an enhanced cooperation, and because of critical legal uncertainties added to complexities at
the EU level, Option C2 is not recommended.
3.4. Option D: Direct regulation justified by the superiority of human
rights law
States have obligations under international and European human rights law to protect the right to
health and other human rights. It is possible to claim (and substantiate) that this obligation is better met
through domestic cannabis regulations than prohibition. At the same time, States claim that their obligations
under international drug control law prevent them from regulating as well as from respecting their human
rights obligation.
This option is based on the idea that the superiority of the body of human rights law over drug law
allows for prioritising the former over the latter, providing legal grounds for domestic regulations. For
Heilmann (2011, p. 284):
“human rights are—despite not being spelled out in the drug conventions—a tacit component of all legitimate
drug control measures. Human rights law is applicable at all times when Member States enforce domestic
control measures based on the conventions. The obligations under the international drug control regime cannot
be invoked for a measure that unjustifiably runs counter to human rights.”
In recent years, substantial scholarly contributions have further deepened the analysis of these questions
(Bone, 2021; Lines, 2017). Van Kempen and Fedorova (2019b) showed that “[t]he regulation of […]
recreational cannabis, because of the interests of individual and public health, security and crime control, can
find its basis in positive human rights obligations.” They postulate that five conditions need to be met in
order for States to prioritise their human rights obligations over the drug control ones:
1. relevant human rights-based interest embedded into the law;
2. a claim of greater effectiveness of human rights protection under a regulated environment;
3. large, democratic support for reforms in the country;
4. the creation of a system which does not negatively affect other States;
5. policies discouraging the use, in particular via increases in public awareness of cannabis harms.
Besides these valuable criteria, there are legitimate concerns about the direct applicability of this
method: “the application of human rights law to the recreational use of drugs remains controversial, and
there is still much to be resolved” (Marks, 2019, p. 222). Doubts remain as to whether such a dynamic
human rights-based interpretation can be leveraged effectively by a State:
“Positive human rights obligations can help to justify cannabis regulation if that can more effectively protect
the right to health, the right to life or the right to privacy, and there are good arguments to prioritise human
rights […]. Still, pointing out those contradictions does not automatically resolve the resulting non-compliance
with the UN drug conventions” (Jelsma, 2022, p. 97).
In practice, in case of conflicts between human rights and other legal instruments, a solution of harmonious
compromise should be sought. The invalidation or overruling of particular provisions based on the
superiority of other provisions should be a last resort in any circumstance. This option should therefore only
be pursued if there is no other, more harmonious manner to reconcile the conflicting provisions. The various
options described in this brief suggest they are more reliable ways forwards, making Option D probably a
better fit as a complement to other Options.
(i) Precedent
There is no precedent. Uruguay and Canada used some arguments related to human rights, but never
formally endorsed this approach. Enacting such an option might require the adoption of a CND resolution to
support it (an unsuccessful attempt was made by Uruguay more than a decade ago; CND, 2008). Optionally,
the International Court of Justice (ICJ) may be called to resolve the issue (article 48, Single Convention).
(ii) EU legislation
This method would require further identifying conflicting provisions at the level of EU acquis and European
human rights law (done in large part by van Kempen and Fedorova, 2019b) and justifying the hierarchical
superiority in this field at the European level. The degree of acceptability of this method among EU
institutions, in particular the CJEU, as well as Council and the Commission, remains unknown.
(iii) Intermediary recommendation
Rather than an option to follow as a self-sufficient way to justify domestic cannabis regulations, the analyses
of interrelations between human rights and international drug law, and conflicting obligations of States,
might be better used in support of whichever other Option is chosen. Notably, the argument of positive
human rights obligations complements Options B, C and E.
In any event, the five conditions listed above should be used as guidance in designing domestic cannabis
regulations that protect, promote and enhance human rights. Option D is recommended only as a
secondary and reinforcing argument.
3.5. Options E: Direct regulation of a “non-medical industry”
This Option does not require changes in treaty texts or in ratification statuses. It consists of the design of
domestic regulations under the rules of articles 2 (9) and 28 of the Single Convention. Based on the
Commentary (UN, 1973, pp. 110, 402), this option interprets the limitation to medical and scientific
purposes in article 4 (c) as modulated by exemptions for non-medical and non-scientific purposes. Key
to this option is article 2 (9), which provides that:
“Parties are not required to apply the provisions of the Convention to drugs used in industry for non-medical
and non-scientific purposes. Article 4(c) is expressed to be ‘[s]ubject to the provisions of this Convention’.
Exceptions to the general obligation of Parties in Article 4(c) to limit drugs to medical and scientific purposes
are provided in Article 2(9) […] and Article 49” (Room and MacKay, 2012, p. 86)
Article 49, not applicable anymore, provided the option to allow traditional non-medical use temporarily.
Article 2 (9) is still applicable today, and gives the option to allow industrial non-medical use.
The wording of article 2 (9) is vague, at times appearing inconsistent. It is better understood with the doctrine
of effet utile in international law, according to which the maximum effectiveness needs to be given to
every word of a treaty provision (a principle embedded into the concept of pacta sunt servanda).12
This doctrine triggers an inter-temporal interpretation (Elias, 1980; Kolb, 2016, p. 158), meaning that article
2 (9) can be interpreted within the language and context of our times (as long as this remains aligned with the
object and purpose of the treaty), thereby giving full effect to the article. Under these general rules of
interpretation, a correspondence can be established between the Convention’s industrial, non-medical
and non-scientific purposes in article 2 (9) and what is today commonly known as the cannabis industry
(Riboulet-Zemouli, 2022, pp. 86–90), giving effect to the provision, as long as such cannabis industry
contributes to the health and welfare of humankind.
Room and McKay (2012, p. 85) believe that under article 2 (9), “drugs cannot be recovered or restored from
their industrial use for consumption.” However, there is no mention of this in the Single Convention, and no
other form of ban on the consumption of cannabis that is exempt under article 2 (9).
The exemption under article 2 (9) is not absolute, and it includes two measures of control:
Upstream: harm reduction. Obligation to prove that the cannabis industry contributes to the health
and welfare of humankind. Subparagraph (a) requires States to take any appropriate measures to
reduce abuse and harmfulness from non-medical cannabis. For this, they can either “denature” the
drug, or they can apply “other means” of reducing harm and abuse. This provision articulates with
other health-centred measures on abuse elsewhere in the Convention (e.g. article 38; see appendix 1).
12 International Law Commission (ILC): “Principle of effectiveness (ut res magis valeat quam pereat): Treaties are to
be interpreted with reference to their declared or apparent objects and purposes; and particular provisions are to be
interpreted so as to give them their fullest weight and effect consistent with the normal sense of the words and with
other parts of the text, and in such a way that a reason and a meaning can be attributed to every part of the text
(ILC, 1965, pp. 53–62) see also Kolb (2016, pp. 154–155).
Downstream: data reporting. Subparagraph (b) requires the basic statistical returns to INCB (via
its Form C)13 of annual amounts of cannabis circulating in the industry (Chatterjee, 1981, p. 420).
UN Secretary-General’s (1966, p. 5) Administrative guide for the Single convention confirms that:
“Parties are not required to apply the provisions of the Convention to drugs, i.e. substances covered by
Schedules I and II which are commonly used in industry for other than medical and scientific purposes, subject
to the provisions contained in Article 2, paras. 9(a) and (b).”
Some commentators argue that these provisions only apply to hemp (i.e. CBD products). However, the
Single Convention mentions neither THC nor the concept of threshold or limits. There are precedents of
non-denatured drugs being exempt under this provision (UN, 1973, pp. 72; Riboulet-Zemouli, 2022, p. 49).
All this corroborates the applicability of this article to all cannabis, placing industrial non-medical cannabis
in a legal regime akin to that of industrial hemp.
In such a situation, the non-medical cannabis industry established under article 2 (9) would not be
contrary to the provisions of the Single Convention (as in the wordings of the 1988 Convention and EU
legislation) as long as the obligations mentioned above are met (Jeanroy et al., 2022). Similarly, the concept
of possession “under legal authority” (article 33) also includes other than medical and scientific purposes
(Room and MacKay, 2012, pp. 126–127; UN, 1973, p. 113). Possession of cannabis obtained via the licit
non-medical cannabis industry would be allowed under the Convention.
Generally, compliance with article 2 (9) places the State in a level of compliance similar to that of
medical/scientific activities: grounded on treaty provisions. Doubts remain regarding the attitude that INCB
will adopt regarding its role of collecting and compiling statistical returns on non-medical cannabis under
article 2 (9) (b).
On cannabis cultivation, there are two distinct possibilities. Article 2 (6) of the Single Convention
explains that “cannabis [is subject to the provisions] of article 28” but does not precise which paragraph,
opening the way for two compliance options, under either article 28 paragraph (1) or paragraph (2):
3.5.1. Option E1: article 2 (9) and 28 (1), unilaterally
With cultivation under article 28 (1). This path is the same regime of obligations applicable to the
cultivation of cannabis and opium poppy for medical and scientific purposes.
In the case of option E1, the manufacture, export, import, distribution of, trade in, use and possession of
cannabis would be exempt from the measures of drug control. In contrast, cannabis cultivation would remain
subject to the full extent of drug control obligations.
Besides considerations of international and European law, the burdensome system of regulations for
cultivation may be excessive or inappropriate for the regulation of a non-medical industry crop. The
appropriateness and viability of licensing schemes (and other obligations thought for the medical and
pharmaceutical sector) is an important consideration when considering needs for tailored transitional
measures in some jurisdictions such as socially-sensitive policies or policies aimed at containing the illicit
market by reintegrating small-scale and low-level drug offenders into licit schemes (Kilmer et al., 2021;
Rahwanji, 2019).
(i) Precedent
13 Form C distributed by INCB already includes a section in “Part II.B” where non-medical cannabis exempted under
article 2 (9) must be reported alongside cannabis for other purposes (Riboulet-Zemouli, 2022, pp. 52, 111).
There is no precedent.
(ii) EU legislation
As other lex lata approaches, Option E1 avoids the complex question of EU competencies. However, the
inconsistency between the legal regimes applied to cultivation and other activities may cause some issues,
and could be challenged.
(iii) Intermediary recommendation
In terms of policy coherence, adopting Option E2 instead of Option E1 appears more logical. Option E1 is
not recommended unless specific issues arise during the implementation of Option E2.
3.5.2. Option E2: article 2 (9) and 28 (2), unilaterally
With cultivation under article 28 (2). This exemption (no particular obligation) is similar to the regime
applicable to hemp cultivation under the Single Convention (discussed in Riboulet-Zemouli, 2022).
Although based on well-ascertained and conservative means of interpretation, this option can be difficult to
apprehend, because it relies entirely on treaty interpretation, which “is to some extent an art, not an exact
science” according to the ILC (1967, p. 218). Perhaps for this reason, some commentators have challenged
the applicability of this Option, particularly E2. Importantly, however, the interpretation of treaties is the
exclusive mandate of States, and not that of external institutions or stakeholders.
Nevertheless, the textual echo between the “industry” in article 2 (9) and “industrial” in article 28 (2)
invites favouring Option E2 over E1, vesting the former with more policy coherence and good faith than the
The waiver of direct obligation over cannabis cultivation (except a general obligation to avoid diversion
towards the illicit economy) allows Member States to adopt regulatory controls tailored to their specific
circumstances, including by relying on the use of contemporary approaches to regulatory oversight (e.g. use
of technology that did not exist in 1961, is not contemplated in the Convention, and could not be possible
with Option E1).
(i) Precedents
There are two precedents.
In Malta, the stated goal of the 2021 law is “to regulate the use of cannabis for purposes other than medical
or scientific purposes and to carry out work […] to implement harm reduction from the use of cannabis”
(Parliament of Malta, 2021) using treaty language present in Article 2 (9). Malta’s undefined “industry”
takes the form of the non-profit cannabis social club model. Malta also relies on article 28 (2): the
regulations prevailing for medical cannabis cultivation do not apply to non-medical crops (Jeanroy et al.,
Switzerland also appears to adopt this approach (Swiss Confederation, 2021; 2022). Though the name of the
regulations (pilot studies) suggests that Switzerland follows a “scientific purposes” approach (option F), a
close look at the law shows they are not:
the law regulates cannabis “for non-medical purposes,”
pilot studies conducted by authorised private sector entities (different industrial models),
the law is built within a robust public health and harm reduction approach,
no prescriptions, pharmaceutical rules, or licensing on cultivation are required,
no estimates are collected (although they are required for scientific purposes), but statistics on the
quantities of cannabis circulating within the pilot studies are collected (as article 2 (9) (b) requires).
Note: because of their regulations’ novelty, none of these two countries has yet submitted returns under
article 2 (9) (b). The Commentary explains that these returns are the compliance mechanism giving effect
to “the non-application of the full narcotics regime” (UN, 1973, pp. 73, 248).
(ii) EU legislation
The incorporation of the Single Convention into EU acquis facilitates the address of this Option. A
non-medical cannabis industry regulated by law, compliant under article 2 (9) of the Single Convention,
would certainly not be “without right” in the meaning of the Framework Decision.
In CJEU case law, the assessment of the “‘need for, and proportionality of’” morally-sensitive regulatory
measures (such as domestic cannabis regulations) relies “‘solely by reference to the objectives pursued’”
(cited in de Witte, 2013, p. 1572), corroborating the need for a strong public health and harm reduction
focus (already required under article 2 (9) (a) of the Single Convention).
As other lex lata approaches, Option E2 avoids the complex question of EU treaty-making competencies. By
relying on the core sovereign capacity of States to interpret treaties, it fundamentally anchors
decision-making processes at the domestic level. Interestingly, “industry” is one of the policy areas where the
EU only has “supporting competencies” to intervene in support, coordination or complement of member
States’ action, and harmonisation of laws or regulations is not necessarily required (articles 6 (b) and 173
TFEU). Similarly to all Options, the ascertainment of EU competencies requires more research.
Highlight: Treaty interpretation competences.
The interpretation of international treaties is the sole competence of States, as long as there is no formal
modification (lex ferenda).
In consequence, institutions such as the European Commission or the INCB do not have any specific
mandate or competence to challenge or contradict lex lata domestic reforms based on an
interpretation of treaty provisions in good faith under Options E.
(iii) Intermediary recommendation
Its transposition into domestic law requires minimal efforts: defining regulated uses as “other than
medical and scientific purposes” and embedding harm reduction and measures to prevent substance use
disorders within the law. Administrative costs to report to INCB, although limited, should be taken into
account, like the necessary expenditures in harm reduction and health services (these budgets are needed
in all Options, but they are a direct legal obligation under article 2 (9) (a)).
Pursuing this Option requires a background in international law, an understanding of complex principles and
obscure wording, and efforts to overcome preconceived ideas (e.g. the feeling that something must be there
somewhere to prohibit). Implementing this Option requires commitment and involvement in
argumentative efforts and, generally, a proactive defence of the model in the multilateral arena. In
particular, this Option calls for diplomatic efforts, proving good faith in efforts to reduce harms and abuse
using “other means” than “denaturing” under article 2 (9) (a).
The two carefully-designed precedents are encouraging: the two are in Europe, one in the Union. Option E2
is simple and presents few risks. It is immediately implementable, and the only option where it is possible in
good faith to remain in compliance with the text of the Single Convention (therefore with the 1988
Convention and EU acquis), using accepted rules of international law.
Option E2 is recommended in the case of a State acting individually.14
3.5.3. Option E3: article 2 (9) and 28 (2), multilaterally
With cultivation under article 28 (2), in a joint approach of like-minded Member States agreeing on a
shared interpretation. Sovereign domestic reforms are welcome “as long as it runs in parallel with
multilateral dialogue […] demonstrating a clear desire to resolve emerging challenges” (Jeanroy et al 2022,
p. 45). While Option E2 can be relied upon for action by a single country, it could also be leveraged by a
group of like-minded countries. In parallel to regulating independently on their territories, a group of States
could express their shared interpretation of article 2 (9) of the Single Convention via a non-binding Joint
Declaration, Statement, or Memorandum of Understanding…
Robert Kolb (2016, p. 131) describes such a situation where a group of states dissent to the treaty
interpretation of the majority, and agree together on a distinct interpretation: This interpretation is not
binding on the States having disagreed to it. However, it may bind the States agreeing to it.
Such a non-binding approximation of positions is to be preferred over unilateral moves, and could serve
valuable purposes for treaty discussions in the longer term. A Joint Declaration geared around a shared
interpretation of provisions within the Single Convention could also articulate claims of compliance under
human rights law (Option D). It could also reaffirm commitments to other aspects of international
cooperation in the field of drugs, as well as commitments not to jeopardise the domestic cannabis policies of
other States.
(i) Precedent
There is no precedent.
Highlight: Joint Declarations shape changes.
There is no precedent of joint declarations mentioning explicitly article 2 (9) of the Single Convention.
Nonetheless, the “Joint Statement” released on 15 July 2022 by Germany, Luxembourg, and Malta
(Gouvernement luxembourgeois, 2022) may account for an early foundation for Option E3. Indeed, the
Joint Statement’s title references “the regulation of cannabis for non-medical and non-scientific uses
–terminology present only in article 2 (9) of the Single Convention. Furthermore, the object and purpose of
the Convention is interpreted as a concern for the health and welfare of humankind, leading to base
cannabis-related reforms and policies on public health objectives.
The gradual build-up of a package of interpretive elements, via a series of Joint Statements of this kind,
can be a positive way to progressively and constructively reinforce Option E3’s acceptance.
14 States taking action on their own should keep in mind the Guiding Principles Applicable to Unilateral Declarations
of States Capable of Creating Legal Obligations of the ILC (2013, pp. 162–166).
(ii) EU legislation
Insofar the intention is not to conclude a binding instrument, and its scope does not interfere with EU
exclusive competence, there is no particular dispositions for the conclusion between States of non-binding
Memoranda of Understanding, Joint Declaration, Joint Statements, etc. (Erlbacher, 2019, p. 1662).
(iii) Intermediary recommendation
Option E3 is recommended if several States agree on a shared interpretation of treaty provisions. It is also
recommended as a costless first step to future joint action. Without being binding, it has legal value in
establishing a clear position, proving good faith and advancing the debate constructively.
3.6. Option F: Direct regulation as an experiment (scientific purposes)
This option proposes that domestic regulations “formulate a scientific research objective for the legalisation
scheme” such as “a comprehensive, population-wide cohort study in which all individuals purchasing
cannabis are automatically enrolled into a study examining the inter-generational health effects of long-term
cannabis use” (Fultz et al., 2017). The applicable provisions of the Single Convention would be the same
measures as the ones for medical cannabis, for all activities.
This option does not require changes in treaty texts or ratification status. Importantly, it may also not require
domestic legal reforms since many drug laws already include protocols for conducting such scientific
experiments under dispositions which comply with the Conventions’ obligations.
In addition, the extent and type of regulatory measures mandated under the Convention for scientific
experiments are highly unfit for non-medical use. Under the scientific experiment scheme foreseen under
Option F, the full extent of drug control measures would apply:
medical prescriptions upon dispensation,
medicalisation (including for people with no use disorder),
general requirements of a pharmaceutical supply chain,
comprehensive estimates and statistical returns on all activities,
in addition to the inappropriate measures for cultivation outlined in section 3.5.1 (Option E1) and
potentially added barriers at the domestic level (complex application and validation procedures for
experiments), etc.
These administrative constraints are likely to incur elevated costs. They also represent significant barriers to
entrepreneurship and access to legally-regulated markets, likely to indirectly favour the continuation of
illicit markets (FAAAT 2021; Kilmer et al., 2021; Rahwanji, 2019).
From the legal standpoint, analysing whether the meaning of “medical and scientific” could be understood as
“public health purposes,” van Kempen and Fedorova (2019a, pp. 40–46, 218; 2019b, p. 272) show that it
does not hold the test of good faith, and is therefore to be discarded. Even under an approach of
harmonisation with positive human rights obligations, they find that “there are still various objections with
this solution.” Whether domestic cannabis regulations can reasonably be considered a scientific purpose
remains doubtful and difficult to sustain in good faith. Case law from the ICJ concur:
“a State cannot manipulate a ‘scientific’ loophole as a way to breach a key treaty provision, in this case to
camouflage a commercial whaling operation as scientific research” (Lines and Barrett, 2018, p. 450).
(i) Precedent
There is no precedent.
Switzerland and the Netherlands did adopt laws framing regulatory trials with cannabis. However, these laws
are not related to this option since none comply with the Convention’s obligations for “scientific purposes,”
which are equivalent to the requirements for medical purposes. Switzerland’s experiments are framed under
Option E2 (see section 3.5.2.(i)), while the Dutch wietexperiment embedding into the Conventions’
framework remains unclear.
(ii) EU legislation
Undertaken under pre-existing domestic laws (already in compliance with all levels of drug legislation), such
an experiment would not a priori face significant hurdles at the EU level. As other lex lata approaches,
Options F avoids the complex question of Union treaty-making competence share. There is large-enough
room for Member States to engage in such a scientific experiment under domestic legislation.
(iii) Intermediary recommendation
Because of the obligations it conveys, unfit for a non-medical market, burdensome for governments and,
more importantly, due to the difficulty to reconcile it with the core principle of good faith in the
interpretation and implementation of international law, Option F is not recommended.
Establishing an industry limited to internal borders seems the most straightforward solution. It presents
advantages, for example, to allow a smooth and unaltered development of a domestic cannabis industry.
In contrast, such a protectionist approach risks “insulating national policies from transnational rights that
companies and citizens derive from the EU legal order” (De Witte, 2013, p. 1557). It would also oversee the
structural impact and codependency of transnational markets on the continent, whether licit or illicit.
Consequently, the oversight and regulation of commercial exchanges within the European single market
(between the Member States having regulated) should be contemplated –act rather than suffer.
Indeed, non-medical cannabis products are “goods” within the economic community (as case law suggests).
As such, they are subject to the general rules of the internal market like any good, in particular, unhindered
trade across borders. No piece of EU law expressly prohibits imports and exports of licit non-medical
cannabis. Nevertheless, individual States can rely on article 36 TFEU to impose the following:
prohibitions or restrictions on imports, exports or goods in transit justified on grounds of public morality, public
policy or public security; the protection of health […].”
Such prohibition or restriction under article 36 has not yet been applied to non-medical cannabis, since it has
never been lawfully marketed in the Union. However, some Member States have relied on it to restrict or
prohibit imports of hemp products (e.g. CJEU, 2003) or medicinal cannabis.
Therefore, in the case of an EU State lawfully commercialising cannabis on its territory, other Member
States not regulating should make derogations under article 36 TFEU to continue prohibiting the
imports of cannabis.
On the other hand, after several Member States have regulated, they may not be in a position to prohibit
imports and exports between themselves. The prohibition of imports can almost only be justified if there
is a similar ban for all trade within the territory (CJEU, 1979). Otherwise, prohibiting imports after
having regulated domestic markets risks being considered a discriminatory measure under article 36:
“[…] Such prohibitions or restrictions shall not, however, constitute a means of arbitrary discrimination or a
disguised restriction on trade between Member States.”
In addition to article 36 TFEU, numerous other derogations to free trade motivated by public interest have
been established by CJEU case law (mandatory requirements; Craig and de Búrca, 2011, pp. 677–689).
Article 36 is likely to prevent EU Member States that have regulated at the domestic level from prohibiting
imports. Therefore, in order to apply coherent and proportional measures to regulate influx of cannabis,
others than a ban on imports, States may want to opt for controlling cross-border trade using the mandatory
requirements (Armstrong, 2015, pp. 518–531; Klamert et al., 2019, pp. 506, 513, 533).
Highlight: Prohibition remains an available option.
Licit cannabis products are goods. As such, their legalisation will inevitably, although progressively, bring them
under the general frameworks related to trade in goods in the EU.
As time passes and an increased number of jurisdictions allow licit cannabis, controlling cross-border trade will
remain possible: countries that do not legalise can continue to prohibit imports, by declaring it under article 36
TFEU. And countries having legalised could apply mandatory requirements.
Generally, however, it will be for each EU State on a case-by-case basis to justify that the restrictions to trade that
they apply are not incoherent, disproportional, or discriminatory.
4.1. Multi-tiered trade relationships
Considering three hypothetical EU Member States:
- State α and State β have regulated their non-medical cannabis industry in compliance with
international and EU law
- State γ decided to maintain prohibition within its borders.
Without a doubt, State γ will remain fully entitled to continue prohibiting imports of non-medical
cannabis in its territory originating from States α or β.
In contrast, case law suggests that, even if desired, State α may not be able to prohibit imports from, or
export to, State β (de Witte, 2013, pp. 1575; see CJEU; 2001). State α would nonetheless remain able to
apply more circumcised and proportional measures like quotas or requirements on labelling, advertising,
prior authorisation mechanisms, and so on (Klamert et al., 2019, pp. 518–522, 535–537; see also section
2.3.4). Any restriction should apply equally to all market players (domestic and foreign). Upon becoming
licit, any limitation to the free movement of cannabis goods and services between two legal countries
must be strictly necessary, suitable and appropriate.
Suppose States α and β wish to commerce, but their territories are not adjacent. In that case, it is unlikely that
transit would be allowed via State γ, except of course in the case of its express authorisation (art. 36 TFEU;
art. 74, 76, and declaration in the Final Act of the Schengen Implementing Convention; see appendix 2).
When States α and β trade in non-medical cannabis across their border, the reliance on some baseline pieces
of EU legislation with general application could be contemplated. For instance:
The provisions of Directive 2001/95/EC on general product safety should be considered (European
Parliament, 2010) in the absence of harmonised EU regulation.
Regulation (EU) 2019/515 “on the mutual recognition of goods lawfully marketed in another
Member State establishes a simplified voluntary procedure. Appreciated by all stakeholders (Jan,
2020), this tool could arguably become a required process in a newly-created trade segment. It would
not be discriminatory, at least not in early stages of the development of cross-border cannabis trade.
4.2. Cannabis-specific EU trade law
Besides general internal market rules, there is one obligation and some commitments directly affecting the
licit trade in cannabis.
The only obligation under EU acquis on cross-border trade in cannabis is contained in article 76 (3) of the
Schengen Implementing Convention. It mandates States to “notify each other of the measures taken in
order to monitor the legal trade.”
Tourism involves the free movement of people, and freedom of services. But there are commitments of EU
Member States, with legal ties, against so-called drug tourism (Deutscher Bundestag, 2022b, p. 5).
Besides non-legislative acts, drug tourism is mentioned in case law (sections 2.3.3 and 2.3.4) and echoed in
the Implementing Convention (provisions mentioned in the paragraph above). A requirement exists for States
to take measures to avoid drug tourism which “concerns both the maintenance of public order and the
protection of the health of citizens” (CJEU, 2010, §65).
Cannabis has been accessible to any traveller in Dutch coffee-shops for decades. The few incidents that arose
are disturbances in localised, dense, touristic areas… not issues affecting other countries’ territories
(Monshouwer et al., 2011; Pereira and de Paula, 2017; Wen et al., 2019). So far, it has been possible to
restrict access to non-medical cannabis only to residents. However, this exceptional derogation to the
freedom of services, justified partly by the fact that only one Member State allowed access, may not last
forever (section 2.3.4).
Banning cannabis tourists risks continuing to offer a market segment to criminal actors. Rather than
discriminatory restrictions on access, States’ approach should be twofold. First, internally, adopt sound
regulations, learning from the Dutch experience, to prevent unintended impacts of cannabis tourism on the
local population. Kang and McGrady (2020) call “policy makers and industry professionals [to] engage in
continual conversations on how to plan and manage this new tourism segment for community and state
Second, vis-a-vis neighbouring States (which constitutes the core meaning of the dispositions on drug
tourism in EU acquis), it is desirable to take diligent actions to prevent any unwarranted trade (exports) by
tourists. Monitoring the phenomenon, and education and awareness-raising among travellers, are essential.
Followed by Canada and the United States, this approach has raised no particular concerns so far (Canadian
Government, 2021; US Embassy in Canada, n.d.). Authorities are coordinated and provide extensive
information and guidance to travellers, including in airports, train stations, etc., warning:
“It is illegal to take cannabis across the Canadian border, whether you are entering or leaving the country. You
could be charged with a criminal offence if you try to travel to other countries with any amount of cannabis in
your possession. This applies to all countries, whether cannabis is legal there or not” (Government of Canada,
Public policies could encourage “multimodal tourist attractions (combining cultural, agro-, or eco-tourism,
with therapeutic/health tourism or with leisure-led tourism)” (FAAAT, 2021, p. 49) to prevent the apparition
of cannabis-only centred tourism practices.
Highlight: A world of cannabis tourism
Cannabis has centuries-old links with travel and tourism. Today, all jurisdictions having reformed its access
contemplate cannabis tourism in some form or another. Banning tourism or prohibiting foreigners from accessing
legal cannabis risks perpetuating an illicit market or creating a grey area in the nascent legal economy.
Tourism related to cannabis should be allowed but framed within the existing touristic offering (to avoid
massification and unsustainable practices), publicly monitored, and accompanied with targeted education and
awareness-raising campaigns for foreign visitors.
* *
The elements outlined above are applicable independently of the Option (A–F) chosen as long as trade
takes place between two States where non-medical cannabis is lawfully marketed, without prima facie breach
of treaty obligations, and without transiting via (or impacting otherwise) those States that retain prohibition
in their territory.15 These elements do not constitute a definitive legal statement. Yet, they suggest that the
panorama of the legal status of trade in licit non-medical cannabis in the EU may not be impossible, and
is certainly not as difficult to apprehend as it can seem at first.
The (short) discussion of trade-related aspects of non-medical cannabis in this section is barely
scratching the surface of a complex topic, in large parts still understudied. In particular, analyses of the
difficult question of trade between EU and non-EU States are highly desirable.
15 Option F could be an exception: special regulations for the trade in products used in research settings might exist.
Countries have choices to comply with their international obligations. Constructive and non-disruptive
approaches which protect and promote the rule of law and international peace and stability must be favoured.
In this regard, it is essential to dialogue and adopt an incremental approach.
This is why the recommended path to follow is based on four steps:
Step 1: Decriminalise personal cultivation, possession and use
As discussed in section 1.2, decriminalisation is a consensual policy, recommended by a wide variety of
stakeholders in the international arena, common in the EU, and with positive impacts ranging from
improvements for health and human rights to savings in State budgets.
Variations and adjustments to domestic decriminalisation schemes remain possible for Member States.
Chosen modalities have different impacts and benefits, and should be carefully weighted (Fernández Ochoa,
2022; Kilmer et al., 2021).
Action in this direction should be taken as early as possible.
Step 2: Regulate a non-medical cannabis industry under current legal framework
–Option E
Option E under article 2 (9) of the Single Convention can supplement decriminalisation, in the short-term.
It provides the most suited basis for a malleable regulated industry, adapted to local contexts and
specificities: pilot studies of dispensing models, cannabis social clubs, farmers’ market-type, more classical
commercially economic schemes, or (ideally) a mixed-model economy.
State taking action on its own (E2): Regulations under Option E2 optimise existing legislation whilst
avoiding challenging the validity of a Single Convention that States have accepted for decades. This Option
presents few risks, enabling a seamless shift away from perceived obligations to ban cannabis, without
affecting other commitments or the treaty itself. It provides valuable margins for governments to
manoeuvre and allows claims of double compliance under both the Single Convention and international
human rights law (Option D) without invalidating or ignoring any treaty provision.
Collective approach (E3): States might also consider taking joint actions under Option E3: several States
sharing the same interpretation under Option E publicise it via a non-binding Declaration or Memorandum,
also reaffirming trust in international cooperation, human rights, and commitment to respecting the policy
choices of other States. Countries acting in such a way could progressively establish commercial
exchanges between themselves. Relying on the procedure of mutual recognition of goods under Regulation
(EU) 2019/515 could assist this process.
At this stage, political and legal actions continue to be the full sovereign competence of States, and do not
call for the involvement of European institutions –as long as reforms do not alter the rules or framework of
the Union. The involvement of EU institutions is recommended at the following stage.
Step 3: Engage in dialogues with WHO, INCB, and the European Commission
Efforts at the domestic level should not be isolated from the United Nations system. WHO and INCB are
the two international bodies mandated under the Conventions to which article 2 (9) seems to talk. The
requirements of harm reduction and prevention of substance use disorders under article 2 (9) (a) would
benefit from guidance and assistance, including sharing good practices, at the WHO level. The reporting
requirements under article 2 (9) (b) would benefit from a frank and open dialogue with the INCB.
A domestic regulation of cannabis complying with international law and EU acquis under Option E2 or E3
falls entirely outside of the area of competence of the European Commission at the stage of its design
(step 2). While the Commission must remain foreign to the sovereign design and adoption of legislative
provisions on non-medical cannabis regulations, it could be most needed to help States implement,
follow-up, and coordinate after their domestic policy reforms. The Commission could also have an important
role to play to mediate between States with different legal approaches to cannabis.
This step should be entered only after step 2 has been completed.
Step 4: Initiate long-term talks towards a future regular amendment
In the longer-term, action should be prepared or anticipated.
As cannabis law reforms multiply, the likelihood of an amendment of the Single Convention will augment.
To ad