Antonio Costa, the Executive Director of the UN Office on Drugs and Crime (UNODC), and Frederick Polak, a Dutch psychiatrist have been engaged in an interesting (and worldwide) debate about Dutch cannabis policy. Polak challenges Costa to answer the question why cannabis use in the Netherlands is lower than in many neighbouring countries despite the free availability of cannabis in coffee shops for adults over 18 years. After a recent visit to Amsterdam, Costa is claiming that cannabis use in that city is three times higher than anywhere else in Europe. Is this true?

The first round of the debate took place at the International Drug Policy Reform Conference of the Drug Policy Alliance (DPA) in New Orleans in December 2007. The point made to Costa was that the Netherlands effectively had free cannabis availability, but average to low (European) levels of use. Costa dodged the question and accused the Netherlands of "poisoning the rest of Europe" with its amphetamine production. After a formal protest by the Dutch Government, Costa had to write a letter to apologize for his remarks. (See: the audio registration of Question & Answer with Dr. Antonio Maria Costa at the DPA conference; UNODC director (lamb) addresses the DPA (slaughter) on the Transform Drug Policy Foundation Media Blog and Mr. Costa Meets the Opposition in the Drug War Chronicle).

The second round was at the NGO forum at the Commission on Narcotic Drugs (CND) in Vienna in March 2008. Again Costa dodged the question, quoting erroneous figures about coffee shops and a distortion of the current debate in the Netherlands. He then closed the discussion. (See: UNODC and the NGO forum: "yet again sir, you do not answer my question" on the Transform Drug Policy Foundation Media Blog, and Silenced NGO Partner on the Hungarian Civil Liberties Union blog).

The third round was at the International Harm Reduction Conference in Barcelona in May 2008. In the meantime Costa traveled to Amsterdam where he made an official visit to one of the Dutch coffee shops (the Dampkring according to the ENCOD bulletin) and a controlled injection site, and met with local authorities. On the UNODC website an article was published, Why should we care about cannabis?, sounding alarm bells about cannabis use.

After Costa’s speech in Barcelona, Polak had the opportunity to challenge him again. This time, he reformulated his question, asking Costa what he learned during his visit in the Netherlands. Mr. Costa answered that the city of Amsterdam “is characterized by rates of drug addiction – I am referring to cannabis use –  three times greater than anywhere else in Europe.” UNODC is preparing a discussion paper on this issue, to be published soon, which concludes: “availability causes use.” (See: Polak's question: Round 3 on the Hungarian Civil Liberties Union blog and the HCLU video below)

Is Costa's claim true?

The latest available figures on cannabis prevalence in Amsterdam are in the national drug use survey Licit and illicit drug use in the Netherlands 2001. The trend in lifetime use of cannabis in Amsterdam was 36.7 per cent in 1997 and 38.1 per cent in 2001, according to the survey. According to the State of the drugs problem in Europe, the 2007 Annual Report of the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) "it is conservatively estimated that cannabis has been used at least once (lifetime prevalence) by more than 70 million European adults, that is on average nearly a quarter (22 per cent) of all 15–64-year-olds." (See: The changing picture of cannabis use in Europe).

When last year use - more figures for estimating regular use - is considered "estimates suggest that more than 23 million European adults have use cannabis in the last year, producing an average figure of about 7 per cent of all 15–64-year-olds," according to the EMCDDA. The comparable figures in the Dutch national survey about Amsterdam are in 'last month continuation' (continuation rates indicate what proportion of lifetime users of a drug used this drug in the last year and the last month respectively), which fell from 22.1 per cent in 1997 20.6 per cent in 2001. (See: Cannabis use in the Netherlands in seven samples). This may be the source of Costa's claim.

However, as is explained in the EMCDDA report 'last month use' is an even more reliable measurement for regular use: "last month prevalence will include people using cannabis more regularly, although not necessarily in an intensive way." In Amsterdam 'last month use' fell from 8.1 per cent in 1997 to 7.8 in 2001. The average in Europe is "on average about 4 per cent of all 15–64-year-olds," according to the EMCDDA. That would mean that in Amsterdam cannabis would "only" be two times higher, if you choose to compare different surveys with different methodologies and target populations.

Somebody should explain Costa that the situation in Mondovì, the small town in Piedmont where he was born, is somewhat different than Amsterdam. Cannabis use in a cosmopolitian urban centre as Amsterdam is likely to be higher than in a backwater in Piedmont due to all kinds of demographic, social and cultural factors. Comparing drug use in both places is comparing apples with pears, just as comparing the figures about Amsterdam with the figures from national European surveys.

When looking at national prevalence figures the situation is completely different. According to the EMCDDA annual report last month prevalence in the Netherlands as a whole is 3.3 per cent. That is lower than in Spain (8.7 per cent), Italy (5.8 per cent), the United Kingdom (5.2 per cent), France (4.8 per cent), the Czech Republic (4.8 per cent), Luxembourg (4 per cent), Austria (3.8 per cent) and Germany (3.4 per cent). In other words, in Europe as a whole last month prevalence is generally higher than in the Netherlands. (See: Last month prevalence of drug use among in nationwide surveys among the general population). A 2005 survey among young outgoing people in inner city pubs in Amsterdam showed a last month prevalence of 22 per cent, down from 24 per cent in 2000. The survey had a low response rate (26 per cent). (See: The Netherlands: Drug Situation 2006, Report to the EMCDDA by the Reitox National Focal Point, February 2006). Maybe Costa stumbled on this figure but comparing it to national prevalence surveys in other countries would be a methodological crime.

The real question

It remains unclear on which figures Costa is basing his claim, let alone the methodological problems involved in comparing different surveys in different set and settings. Somebody in Vienna should stand up and explain the difference between interpreting statistics and manipulating them. The forthcoming UNODC discussion paper will no doubt conclude that there is a correlation between availability and use. You don’t have to be a rocket scientist to understand that such a correlation is obvious.

A 2004 comperative study by Peter Cohen, Craig Reinarman and Hendrien L. Kaal, The Limited Relevance of Drug Policy: Cannabis in Amsterdam and in San Francisco – published in the peer reviewed American Journal of Public Health – tried to answer a somewhat related question: whether criminalization constrains use or whether decriminalization increases it. The study compared the career use patterns of representative samples of experienced cannabis users in two cities with many similarities but with different drug control regimes: Amsterdam, the Netherlands (decriminalization), and San Francisco, California (criminalization).

The study concluded that “the findings did not support claims that criminalization reduces cannabis use and that decriminalization increases cannabis use. Moreover, Dutch decriminalization does not appear to be associated with greater use of other illicit drugs relative to drug use in San Francisco, nor does criminalization in San Francisco appear to be associated with less use of other illicit drugs relative to their use in Amsterdam. Indeed, to judge from the lifetime prevalence of other illicit drug use, the reverse may be the case.”

If criminalization or decriminalization do not appear to have a significant effect on use  as the study concludes, the debate can finally move on to the question of what is the most effective and humane policy to reduce cannabis related harm while securing human rights, avoid insane incarceration rates etc.. The real question should not just be whether there is a correlation between availability and use, but also whether there is a correlation between criminalization and availability.

Read also:

• Decriminalization of cannabis, in the March 2008 issue of Current Opinion in Psychiatry.
• We should be careful about cannabis policy, Peter Sarosi, Drug Reporter, May 23, 2008