Evaluating alternative cannabis regimes (and follow-up comments)

ROBERT MacCOUN, PhD, Goldman School of Public Policy and Boalt Hall School of Law, University of California, Berkeley; RAND Drug Policy Research Center; PETER REUTER, PhD, School of Public Affairs and Department of Criminology, University of Maryland; RAND Drug Policy Research Center
*This paper is excerpted from MacCoun & Reuter (2001), with the permission of Cambridge University Press. The work was supported by the Alfred P. Sloan Foundation through a grant to RAND's Drug Policy Research Center.
^†See editorial, p. 98, this issue.
Correspondence: Robert MacCoun, Goldman School of Public Policy, University of California at Berkeley, 2607 Hearst Avenue, Berkeley, CA 94720-7320, USA
(First received 22 July 1999, final revision 14 December 1999, accepted 16 December 1999)

Outline

Background: Cannabis policy continues to be controversial in North America, Europe and Australia.

Aims: To inform this debate, we examine alternative legal regimes for controlling cannabis availability and use.

Method: We review evidence on the effects of cannabis depenalisation in the USA, Australia and The Netherlands. We update and extend our previous (MacCoun & Reuter, 1997) empirical comparison of cannabis prevalence statistics in the USA, The Netherlands and other European nations.

Results: The available evidence indicates that depenalisation of the possession of small quantities of cannabis does not increase cannabis prevalence. The Dutch experience suggests that commercial promotion and sales may significantly increase cannabis prevalence.

Conclusions: Alternatives to an aggressively enforced cannabis prohibition are feasible and merit serious consideration. A model of depenalised possession and personal cultivation has many of the advantages of outright legalisation with few of its risks.

Declaration of interest: Funded by the Alfred P. Sloan Foundation. The authors have no financial interest in the outcome of the research.

Cannabis is the cutting-edge drug for reform, the only politically plausible candidate for major legal change, at least decriminalisation (removal of criminal penalties for possession) and perhaps even outright legalisation (permitting production and sale). Compared with other drugs, the harms, physiological or behavioural, are less severe and the drug is better integrated into the culture. Throughout Western Europe and in the Antipodes there is pressure for reductions in the punitiveness of the marijuana regime.

This paper attempts to project the likely consequences of substantial changes in the basic legal regime for cannabis and to offer an assessment of those consequences. The best evidence on the effects of liberalising marijuana policy comes from The Netherlands which has experienced both decriminalisation and commercialisation, though without ever changing formal law. In this paper, we summarise and extend our analysis of the Dutch cannabis policy (MacCoun & Reuter, 1997).

The Dutch experience, together with those of a few other countries with more modest policy changes, provides a moderately good empirical case that removal of criminal prohibitions on cannabis possession (decriminalisation) will not increase the prevalence of marijuana or any other illicit drug; the argument for decriminalisation is thus strong. Making cannabis fully legal is likely to increase its use substantially because of promotion, particularly in the USA with its peculiar dedication to commercial free speech; that is possibly undesirable. An intermediate model can be devised which may be preferable to either legalisation or simple decriminalisation.

DUTCH CANNABIS POLICY
In compliance with international treaty obligations, Dutch law states unequivocally that cannabis is illegal. Yet in 1976 the Dutch adopted a formal written policy of non-enforcement for violations involving possession or sale of up to 30 g of cannabis - a sizeable quantity, since few users consume more than 10 g a month (probably 25-35 joints) (Cohen & Sas, 1998). In 1995, in response to domestic and international pressures, this threshold for possession was lowered to 5 g. Moreover, a formal written policy regulates the technically illicit sale of those small amounts in open commercial establishments; a 500 g limit on trade stocks was established in 1995. Enforcement against those supplying larger amounts is aggressive; in 1995 the Dutch government seized 332 tonnes of cannabis, about 44 per cent of the total for the European Union as a whole (Ministry of Foreign Affairs et al, 1995).

Between 1976 and 1986, a set of guidelines emerged stipulating that coffee-shop owners could avoid prosecution by complying with five rules: (a) no advertising; (b) no hard drug sales on the premises; (c) no sales to minors; (d) no sales transactions exceeding the quantity threshold; and (e) no public disturbances (Ministry of Foreign Affairs et al, 1995). In 1980, Ministry of Justice guidelines decentralised implementation, providing greater local discretion. As a result, enforcement became more lenient in Dutch cities, and somewhat stricter in smaller towns (Jansen, 1991). The effect is illustrated graphically in Dutch geographer A. C. M. Jansen's (1991) maps plotting cannabis coffee-shop locations in Amsterdam. He depicts nine locations in 1980, 71 in 1985 and 102 by 1988. (A location may correspond to more than one coffee shop.) Jansen notes that "the first coffee shops were usually situated in unattractive buildings in backstreets" (p. 69), but observes that over the course of the 1980s the shops spread to more prominent and accessible locations in the central city; they also began to promote the drug more openly.

The cumulative effect on these formal, quasi-formal and informal policies is to make cannabis readily available at minimal legal risk to interested Dutch adults. There are approximately 1200 coffee shops selling cannabis in The Netherlands ( Abraham et al, 1999, p. 93). Most offer an international variety of marijuana and hash strains of varying potency levels. Gram prices are 5 to 25 guilders ($2.50 to $12.50) (Kraan, 1994) compared with US figures of $1.50 to $15.00. It is possible that the Dutch marijuana is of higher potency and 'quality', a much less well-defined term but one which experienced smokers use. No data are available in either dimension. The continued high price of marijuana in The Netherlands probably reflects the aggressive enforcement against large-scale growers and distributors.

OUTCOMES
In MacCoun & Reuter (1997), we considered three key policy questions. Are levels of cannabis use higher in The Netherlands than in other Western nations? Did levels of cannabis use in The Netherlands increase following the 1976 depenalisation and subsequent de facto legalisation? And has the policy change weakened the statistical association between marijuana and use of other drugs? Here, we briefly summarise and extend those findings. In the final section, we clarify our interpretation of those findings and possible implications for US cannabis policy.

Prevalence of cannabis use in The Netherlands, USA, Denmark and Germany
At the very least, meaningful cross-sectional comparisons of drug use should be matched for survey year, measure of prevalence (life-time use, past-year use, or past-month use), and age groups covered in the estimate. Failure to meet these criteria has led to grossly discordant comparisons in which, for example, rates among 12- to 17-year-olds in one country are compared with those among 18-year-olds in another, all being called adolescents or teenagers (e.g. Associated Press, 3 October 1997 (P. Recer; published under different titles in various US newspapers); Los Angeles Times, 26 July 1998 (R. Housman; letter)).

Our 1997 paper presented 15 comparisons that met these criteria. Table 1 extends the list to 28: 16 comparisons to the USA, three to Denmark, two to West Germany, one to Sweden, one to Helsinki, one to France, and four to the UK. We identified 15 comparisons in 1997. Here we add 13 additional comparisons. Some of these pre-date that paper but were unknown to us at the time it was written. Two others (lifetime use among those 12 and older in Tilburg and Utrecht in 1995) were omitted from that study by an oversight. Including the latter increases the amount by which the US rates exceed those in Utrecht (from a 0.3% difference to a 1.4% difference) and especially Tilburg (from a 3.4% difference to a 7.9% difference). This does not change our substantive conclusion that "US rates are . . . similar to that of Utrecht, and higher than that of Tilburg" ( MacCoun & Reuter, 1997, p. 49). All but two occur in the 1990s, during the period we have characterised as de facto legalisation, not just depenalisation. Four contrasts compare national estimates from The Netherlands and the USA; three show negligible differences between the two countries (within sampling error), while the newest estimate (Abraham et al, 1999) suggests that US prevalence is much higher. This discrepant result may be attributable to the inclusion of older adults in the latter comparison, or due to some difference between the Centre for Drug Research (CEDRO) (household) and Trimbos-institut (school-based) national survey methodologies. Twelve comparisons involve US national data and a Dutch city. Six contrasts pair the USA with an estimate from Amsterdam - a large urban setting with a visible drug culture. American surveys indicate little difference, on average, between large metropolitan samples and the USA as a whole (Substance Abuse and Mental Health Services Administration, 1979-1999), but the estimates in Table 1 suggest that Amsterdam has a higher fraction of marijuana users than smaller Dutch communities. US rates are basically identical to those in Amsterdam and Utrecht, and higher than those in Tilburg ( Langemeijer, 1997; also see Abraham et al, 1999).

Table 1 Comparing cannabis use in The Netherlands and other nations

Trends in the prevalence of cannabis use
Did levels of cannabis use in The Netherlands increase following the 1976 depenalisation and subsequent de facto legalisation? We examined Dutch lifetime prevalence data from various sources between 1970 and 1996 (MacCoun & Reuter, 1997). (Past-month or past-year prevalence estimates would be more informative but are scarce, especially prior to 1986.) The data for ages 16-17 and 18-20 come from a periodic national school survey (de Zwart et al, 1997); the data for ages 16-19 in Amsterdam come from a periodic city survey (Sandwijk et al, 1995); and the trend line for 1970-1983 reflects a synthesis of various early estimates based on a multivariate analysis by Driessen et al (1989).

The trend line implies that, among Dutch adolescents, cannabis use was actually declining somewhat in the years prior to the 1976 change and that the change had little if any effect on levels of use during the first 7 years of the new regime. Unfortunately, we lack data on the stringency of enforcement in the years immediately prior to the change in law, though the trend lines are fairly smooth and declining for at least 6 years prior to 1976.

In the 1984-1996 period, which we characterise as a progression from depenalisation to de facto legalisation, these surveys reveal that the lifetime prevalence of cannabis in Holland has increased consistently and sharply. For the age group 18-20, the increase is from 15% in 1984 to 44% in 1996; past-month prevalence for the same group rose from 8.5% to 18.5% (de Zwart et al, 1997). Is this an effect of the emergence of de facto legalisation?

Two comparison series offer insight: the US Monitoring the Future annual survey of high-school seniors (Bachman et al, 1998), and an annual survey of Oslo youth, aged 15 to 21 (Norwegian Ministry of Health and Social Affairs, 1997). The USA and Norway both strictly forbid cannabis sales and possession, and aggressively enforced that ban throughout the period. Note that because the Oslo survey has a broader age range, these estimates are more meaningful for comparing trends over time than absolute differences in prevalence in any given year.

The two comparison series behave very differently from the Dutch series, and from each other until 1992. The US rates increase until 1979 and then fall steadily and substantially until 1992, while the Oslo figures increase sharply only until 1972, and then fluctuate around a flat trend until 1992. Interestingly, during the period 1992 to 1996, all three nations have seen similar large increases, as have Canada (e.g. Adlaf et al, 1995) and the UK (Table 1). This weakens the hypothesis that the Dutch increases from 1992 to 1996 are attributable to Dutch policies per se; the fact that comparable increases occurred in nations with such different legal risks highlights the important role of non-policy influences that are only poorly understood. Nevertheless, the increases in Dutch prevalence from 1984 to 1992 provide the strongest evidence that the Dutch regime might have increased cannabis use among the young. As is seen in Fig. 1, this was a period in which use levels were fairly flat in Oslo and declining in the USA. Available estimates also suggest flat or declining use during this period in Catalunya, Stockholm, Hamburg and Denmark (Hartnoll, 1994), Germany as a whole (Reuband, 1992), Canada (Adlaf et al, 1995) and Australia (Mugford, 1992). Thus, unlike the widespread post-1992 rises, the 1984-1992 escalation seems (almost) uniquely Dutch. In only one other location was cannabis use clearly increasing during this period - Helsinki, where lifetime prevalence doubled among 15-year-olds between 1988 (5%) and 1992 (10%) (see Hartnoll, 1994).

Fig. 1 Estimated lifetime prevalence of cannabis among 18-year-olds in the USA, The Netherlands, Amsterdam and Oslo.

We are not claiming that the increases circa 1984-1992 are solely attributable to coffee-shop commercialisation, nor that commercialisation is synonymous with coffee-shop transactions. Commercialisation also involves the heightened salience and glamorisation (in the youth-cultural sense) that results from widespread, highly visible promotion - the veiled references to cannabis in shop signs and advertisements, but also the explicit depictions in counter-cultural media ads, postcards and posters.

The gateway association
Has the Dutch policy change influenced the statistical association between marijuana and use of other drugs? Though American hawks argue that more lenient cannabis policies might lead to greater levels of hard-drug use, a central rationale for the 1976 Dutch legal change was the notion that separating the soft- and hard-drug markets might actually weaken any gate-way effect (Ministry of Foreign Affairs et al, 1995). Dutch policy may have had some success in separating these markets. Most Dutch cannabis users obtain that drug through either coffee shops or friends; few buy from street dealers. According to the 216 experienced Amsterdam cannabis users interviewed by Cohen & Sas (1998), hard-drug sales at coffee shops are quite rare; only four reported that cocaine could be purchased, and only one knew of heroin sales at a shop. Among past-year cannabis users aged 18 and older in The Netherlands as a whole, 48% cite coffee shops as their place of purchase; only 0.7% report purchases from strangers on the street (Abraham et al, 1999). Fewer than 2% of past-year cocaine users report buying cocaine at coffee shops.

In Amsterdam, as in the USA, almost all hard-drug users have used cannabis, but the vast majority of cannabis users have not used hard drugs. In both countries the surveys underestimate the number who frequently use cocaine or heroin and who almost certainly used marijuana. This reduces the denominator and numerator for calculating the percentage of marijuana users who went on to these other drugs; since the numerator is much smaller, this reduces the estimated rate below the true value. However, the problem holds in both nations and, since the Dutch are seen as doing a better job of integrating their addicts into the household population, may be less severe for The Netherlands than the USA. Only 22% of those aged 12 and over who have ever used cannabis have also used cocaine (Cohen & Sas, 1996). This compares to a figure of 33% for the USA. For heroin, the corresponding figures are 4% for Amsterdam and 3% for the USA - statistically identical.

Thus, although the Dutch have failed to eliminate the statistical association between cannabis and hard-drug use - we estimate that the probability of cocaine or heroin use among those in Amsterdam who have never tried cannabis is essentially zero - it is possible that they have weakened it, at least for heroin. Also, only 6% of cannabis users had used cocaine more than 25 times; only 2% were current (past-month) users. Just 2% of cannabis users had used heroin more than 25 times; less than 1% were current users. Note, however, that the alleged gateway is a function of both the number of people who have tried marijuana and the probability of cocaine use given marijuana use. Any increase in the former component (the prevalence of marijuana use) might offset reductions in the latter component (the probability of moving on to cocaine use), and it is possible that Dutch commercialisation has had such an effect. From the perspective of breaking the gate-way link, a regime that tolerates home cultivation of small quantities (as in Alaska and South Australia) might be more effective than the coffee-shop model.

The basis for continued cannabis prohibition
The case for continued prohibition of non-medical uses of marijuana rests primarily on four possible harms: (a) marijuana's role as a gateway to other drugs of known dangerousness, a role generally believed to be unrelated to its legal status; (b) the health consequences and impact on adolescent development; (c) behaviour when intoxicated; and (d) the difficulty of quitting. We think none of these turns out to be very substantial; in particular, the gate-way effect (which has seven possible interpretations) has probably been greatly overstated.

Our judgement, based on review of the research literature, is that at present the primary harms of marijuana use (including those borne by non-users) come from criminalisation: expensive and intrusive enforcement, inequity, shock to the conscience from disproportionate sentence and a substantial (though generally non-violent) black market. Certainly the drug itself causes damage: it generates accidents causing harm to both the user and others; regular use by adolescents may adversely affect development; it may have some substantial impact on the prevalence of cancer among frequent users; a non-trivial share of users has difficulty quitting when they wish to and see their lives as somewhat harmed because of their dependence. But the adverse consequences of criminalisation, at least with current US enforcement, seem more substantial.

The available evidence suggests that removal of the prohibition against possession itself (decriminalisation) does not increase cannabis use. In addition to the Dutch experience from 1976 to 1983, we have similar findings from analysis of weaker decriminalisations (with fines retained for the offence of simple possession of small quantities) in 12 US states (Single, 1989) and South Australia and the Australian Capital Territory (Hall, 1997; McGeorge & Aitken, 1997). The fact that Italy and Spain, which have decriminalised possession for all psychoactive drugs, have marijuana use rates comparable to those of neighbouring countries provides further support. This prohibition inflicts harms directly and is costly. Unless it can be shown that the removal of criminal penalties will increase use of other more harmful drugs, perhaps because of the signal of lessened disapproval, it is difficult to see what society gains.

Decriminalisation is normatively flawed (why does sale remain illegal?) and still leaves the harms of black markets. However, the removal of the sales prohibition has more complex effects. We believe that it would generate larger increases in marijuana use as a result of promotion by the legal suppliers. Promotion could not be effectively limited in the US commercial marketplace for a product which, with no therapeutic goal, would be provided in conventional commerce rather than through doctors and pharmacies. Recent experience with legalised gambling, as well as the difficulty of suppressing cigarette promotion, added to the post-World War II erosion of repeal's liquor controls, all suggest that legal commercial interests are likely to weaken regulatory efforts. This is especially plausible for marijuana, whose harms are relatively slight, hence complicating the task of defending stringent regulation against the efforts of a legal industry. If, even with relatively tight regulation, The Netherlands saw a large increase in marijuana prevalence, US legalisation might lead to very high prevalence rates indeed. The increase in marijuana use would have to be weighed against the reduced intrusiveness of the state, reduction of black markets and possible substitution of marijuana for alcohol, which might be net health enhancing.

Other regimes between decriminalisation and commercialisation are possible. For example, the state of Alaska permits home production for own consumption and gifts to others. The impact on prevalence is difficult to determine (Segal, 1990) but it may be an appropriate compromise between the excess of commercialisation and the barren rights of decriminalisation. Our purpose here is not to choose an optimal regime but only to suggest that available evidence provides a basis for a reasonable debate about the likely consequences of regime changes.

CLINICAL IMPLICATIONS
* It is important to distinguish depenalisation from commercialisation when assessing alternatives to drug prohibition.
* The elimination or steep reduction in penalties for cannabis possession does not appear to influence cannabis prevalence.
* Legal or quasi-legal commercial sales of cannabis may produce significant increases in cannabis prevalence.

LIMITATIONS
* Cross-national surveys differ with respect to language, question wording, sampling, and other details that may bias inferences about relative cannabis prevalence.
* Cross-national differences in cannabis prevalence do not necessarily reflect national differences in cannabis policies.
* Existing sources of data on national cannabis use permit only weak forms of causal inference; observed correlations might be spurious.

REFERENCES
Abraham, M. D., Cohen, P. D. A., van Til, R., et al (1999) Licit and Illicit Drug Use in The Netherlands, 1997. Amsterdam: Centre for Drug Research (CEDRO). [Context Link]

Adlaf, E. M., Ivis, F. J., Smart, R. G., et al (1995) The Ontario Student Drug Use Survey: 1977-1995. Toronto: Addiction Research Foundation. [Context Link]

Bachman, J. G., Johnston, L. D. & O'Malley, P. M. (1998) Explaining recent increases in students' marijuana use: impacts of perceived risks and disapproval, 1976 through 1996. American Journal of Public Health, 88, 887-892. [Context Link]

Boekhout van Solinge, T. (1997) Cannabis in France. In Cannabis Science: From Prohibition to Human Right (ed. L. Böllinger). Frankfurt am Main: Peter Lang.

Cohen, P. & Sas, A. (1996) Cannabis Use, a Stepping Stone to Other Drugs? The Case of Amsterdam. Amsterdam: CEDRP Centrum voor Drugsondersoek, Universiteit van Amsterdam. [Context Link]

Cohen, P. & Sas, A. (1998) Cannabis Use in Amsterdam. Amsterdam: CEDRP Centrum voor Drugsondersoek, Universiteit van Amsterdam. [Context Link]

de Zwart, W. M., Stam, H. & Kuipers, S. B. M. (1997) Kerngegevens: Roken, drinken, drugsgebruik en gokken onder scholieren vanof 10 Jaar. Utrecht: Trimbos-instituut. [Context Link]

Driessen, F. M., Van Dam, G. & Olsson, E. B. (1989) De ontwikkeling van het cannabisgebruik in Nederland, enkcle Europese landen en de VS sinds 1969. Tijdschrift voor Alcohol, Drugs en Andere Psychotrope Stoffen, 15, 2-14. [Context Link]

European Monitoring Centre for Drugs and Drug Addiction (1996, 1997) Annual Report on the State of the Drugs problem in the European Union. Lisbon: EMCDDA.

Hall, W. (1997) The recent Australian debate about the prohibition on cannabis use. Addiction, 92, 1109-1115. [Context Link]

Hartnoll, R. (1994) Multi-city Study: Drug Misuse Trends in Thirteen European Cities. Co-operation Group to Combat Drug Abuse and Illicit Trafficking in Drugs (Pompidou Group). Strasbourg, France: Council of Europe Press. [Context Link]

Hibell, B., Andersson, B., Bjarnason, T., et al (1997) The 1995 ESPAD Report: Alcohol and Other Drug Use Among Students in 26 European Countries. Stockholm: Swedish Council for Information on Alcohol and Other Drugs, and the Council of Europe Pompidou Group.

Jansen, A. C. M. (1991) Cannabis in Amsterdam: A Geography of Hashish and Marijuana. Muiderberg, The Netherlands: Coutinho. [Context Link]

Kraan, D. J. (1994) An economic view on Dutch drugs policy. In Between Prohibition and Legalization: The Dutch Experiment in Drug Policy (eds E. Leuw & L. H. Marshall), pp. 283-310. Amsterdam: Kugler. [Context Link]

Langemeijer, M. P. S. (1997) Prevalence of drug use: A comparison of three Dutch cities. In Illicit Drugs in Europe (eds D. Korf & H. Riper). Amsterdam: SISWO. [Context Link]

MacCoun, R. J. & Reuter, P. (1997) Interpreting Dutch cannabis policy: Reasoning by analogy in the legalization debate. Science, 278, 47-52. [Context Link]

MacCoun, R. J. & Reuter, P. (2001) Drug War Heresies: An Agnostic Look at the Legalization Debate. Cambridge: Cambridge University Press, in press. [Context Link]

McGeorge, J. & Aitken, C. K. (1997) Effects of cannabis decriminalization in the Australian Capital Territory on university students' patterns of use. Journal of Drug Issues, 27, 785-793. [Context Link]

Ministry of Foreign Affairs, Ministry of Health, Welfare and Sport, Ministry of Justice, et al (1995) Drugs Policy in The Netherlands: Continuity and Change. Rijswijk, The Netherlands: Ministry of Justice. [Context Link]

Mugford, S. (1992) Licit and illicit drug use, health costs and the crime connection in Australia: public views and policy implications. Contemporary Drug Problems, 19, 351-385. [Context Link]

Norwegian Ministry of Health and Social Affairs (1997) New Trends in Drug Abuse in Norway.[Context Link]

Reuband, K. H. (1992) The epidemiology of drug use in Germany: basic data and trends. In Drug Addiction Treatment Research: German and American Perspectives (eds G. Bühringer & J. J. Platt), pp. 3-16, Malabar, FL: Krieger. [Context Link]

Sandwijk, J. P., Cohen, P. D. A., Musterd, S., et al (1995) Licit and Illicit Drug Use in Amsterdam II: Report of a Household Survey in 1994 on the Prevalence of Drug Use among the Population of 12 Years and over. Amsterdam: Department of Human Geography, University of Amsterdam. [Context Link]

Segal, B. (1990) Drug-taking Behavior among School-aged Youth: The Alaska Experience and Comparisons with Lower-48 States. New York: Haworth Press. [Context Link]

Single, E. (1989) The impact of marijuana decriminalization: an update. Journal of Public Health Policy, 10, 456-466. [Context Link]

Substance Abuse and Mental Health Services Administration (1979-1999) National Household Survey on Drug Abuse. Rockville, MD: US Department of Health and Human Services. [Context Link]

Follow-up comments

The British Journal of Psychiatry (2001) 179: 175-177
© 2001 The Royal College of Psychiatrists

Comparative cannabis use data

Centre for Drug Research, University of Amsterdam, Postbus 94208, 1090 GE Amsterdam, The Netherlands

D. J. Beukenhorst

Centraal Bureau voor de Statistiek (CBS), Heerlen, The Netherlands

MacCoun & Reuter (2001) examine alternative legal regimes for controlling cannabis availability and use. They claim that the Dutch experience (the coffee shop system with decriminalisation of purchase, followed by "commercial promotion") significantly increases cannabis use prevalence. They conclude, however, that primary harm comes more from criminalisation than from decriminalisation. They base their conclusions on the comparison of cannabis use data from The Netherlands and from other countries. Rightly, they warn that "meaningful cross-sectional comparisons of drug use should be matched for survey year, measure of prevalence... and age groups covered in the estimate". They forget that the comparisons should also be matched for type of geographical area. Comparing Flatbush to New York City would have limited relevance, even if correctly matched for age group etc. To compare cannabis use in New York City with somewhere else, one would have to look for a similar area, both in address density and in variation of population and lifestyle. Amsterdam could be compared to San Francisco, because these cities are very similar in size and cultural characteristics, but not to New York City, a metropolis over 10 times as large, or to the USA as a whole. Such comparisons are wrong and without meaning.

We agree with MacCoun & Reuter that decriminalising cannabis merits serious consideration. But we disagree with their observations on "commercialisation". In this letter we will turn most of our attention to the epidemiological material the authors base their conclusions on.

MacCoun & Reuter focus entirely on cannabis prevalence (assuming that a lower prevalence is better than a higher one) without considering whether this is the most relevant issue; the social and legal consequences of the use of cannabis could be considered at least as important. But given that a comparison of prevalence figures is a useful first step towards informed comparisons, we propose that the conclusion of MacCoun & Reuter that the commercial type of Dutch coffee shop system increases cannabis prevalence is based on statistically ill-founded comparisons of Dutch prevalence figures with those in other Western nations.

MacCoun & Reuter compare cannabis prevalence figures of a Dutch city or nationwide with prevalence figures from the USA or other Western nations. Differences are summed and averaged, resulting in (among others) a mean Dutch—US difference and a mean Dutch—European difference. This is statistically erroneous for reasons we supply below.

First, in 16 cases a Dutch city is compared with a nation (UK, USA, Sweden, etc). By doing this, MacCoun & Reuter presuppose that prevalence rates are the same all over The Netherlands. This is incorrect: in our 1997 national survey we found large geographical differences between locations with different address densities, a measure of urbanisation. For example, lifetime prevalence of cannabis use in Amsterdam (address density > 3000/km²) was 36.7%, the average national prevalence was 15.6% and average prevalence in rural areas (address density <500/km²) 10.5%. Correct international comparisons can be made, but have to be between comparable geographical or urban areas. Despite the sensitivity MacCoun & Reuter demand for correct comparisons, nationwide US figures (260 million inhabitants, including major metropolitan areas) are compared with the small Dutch city of Tilburg (165 000 inhabitants).

Second, comparisons are arbitrarily selected. For example, replacing prevalence figures for Amsterdam (the city most often chosen in MacCoun & Reuter's comparisons) with figures for Rotterdam changes the outcomes of the average difference in cannabis prevalence between the Dutch and other systems.

Third, MacCoun & Reuter state that the lifetime prevalence of cannabis in The Netherlands has increased consistently and sharply in the age group 18-20, stating: "the increases... provide the strongest evidence that the Dutch regime might have increased cannabis use among the young". This finding is based on school survey data (lifetime cannabis use in 1984: 15%, in 1996: 44%). Again, the choice of figures that are compared is crucial. Moreover, the Dutch school survey data of the age group 18-20 is an extremely biased selection of this age cohort. The school survey takes place in some primary schools, but mostly in secondary educational institutions, that are designed for 12- to 18-year-olds. However, some persons remain much longer in this system for a variety of reasons but they are atypical for the age group in general. They bias the school survey estimate for this age group.

More suitable figures are given by Statistics Netherlands (CBS) and by the Centre for Drug Research (CEDRO), and reflect a much more moderate increase or no increase at all. Statistics Netherlands measures cannabis use prevalence in a national representative sample. For the age group 18-20 lifetime cannabis use remains at the same level over time (17% in 1989, 19% in 1990, 18% in 1991, 20% in 1992 and 14% in 1993; data from D. J. B.). Using CEDRO data, we are able to produce trend data for the city of Amsterdam for the same age group 18-20: in 1987 lifetime cannabis use was 34%, rising to 44% in 1997. This is a rather modest increase in cannabis use, very similar to the slowly rising consumption levels of other European and US measurements. The 18- to 20-year-olds in the samples fromAmsterdam are randomly selected from the citizen registry, and represent the age group much better than 18- to 20-year-olds still attending school. The ‘dramatic’ increase that MacCoun & Reuter hypothesise in Dutch cannabis use in the period 1984-1996 (as reflected in the same age group) does not exist.

Finally, the most serious flaw develops by creating a series of ‘absolute’ differences between Dutch and other data, and averaging them. MacCoun & Reuter create the suggestion that too large or too small differences will be averaged and thereby, in the form of an ‘average’ difference, become more reliable. The opposite is true. If pears can not be compared to apples, their ‘differences’ can not be used for normal mathematical computations.

EDITED BY MATTHEW HOTOPF

Declaration of interest  None. The views expressed by D. J. B. are the author's own and do not necessarily reflect the policies of CBS.

REFERENCES

Abraham, M. D., Cohen, P. D. A., van Til, R. J., et al (1999) Licit and Illicit Drug Use in The Netherlands. Amsterdam: Centre for Drug Research (CEDRO).

MacCoun, R. & Reuter, P. (2001) Evaluating alternative cannabis regimes. British Journal of Psychiatry, 178, 123-128.

Substance Abuse and Mental Health Services Administration (1997) National Household Survey on Drug Abuse: Population Estimates 1997 . Rockville, MD: US Department of Health and Human Services.

Authors' reply

Richard & Rhoda Goldman School of Public Policy, University of California, 2607 Hearst Avenue, Berkeley, CA 94720-7320, USA

P. Reuter

School of Public Affairs and Department of Criminology, University of Maryland, USA

EDITED BY MATTHEW HOTOPF

We thank Abraham et al for their comments, but they have misrepresented our paper, and we find their arguments either misleading or unconvincing.

Abraham et al complain that 16 of our 28 statistical comparisons contrast a Dutch city with a national estimate from the USA or another nation, suggesting that we "pre-suppose that prevalence rates are the same all over The Netherlands". We made no such presupposition. As we clearly stated in our article: "American surveys indicate little difference, on average, between large metropolitan samples and the USA as a whole... but the estimates in Table 1 suggest that Amsterdam has a higher fraction of marijuana users than smaller Dutch communities. US rates are basically identical to those in Amsterdam and Utrecht, and higher than those Tilburg". We then note that "unfortunately, many of the available contrasts between The Netherlands and her European neighbours suffer from the same weakness, comparing rates for an entire nation as a whole to those in the largest city of another nation". And we state that the contrasts where the Dutch rates are higher are mostly "attributable to comparisons limited to Amsterdam". We conclude that "Dutch rates are somewhat lower than those of the USA but somewhat higher than those of some, but not all, of its neighbours. Amsterdam's level of marijuana use is comparable to that of the USA".

Abraham et al further complain that our comparisons were "arbitrarily selected". In fact, our 1997 Science article included every Dutch cannabis prevalence rate for which we could find a reasonable international contrast matched by year, age range and type of prevalence. Our recent update in the British Journal of Psychiatry added another 13 comparisons. We welcome further comparisons but a fair reading of both papers makes it clear that we attempted to be exhaustive, given the limited availability of Dutch drug prevalence data in English-language sources. (Indeed, where possible we had Dutch-language sources translated.) In any case, we emphasise that we drew no policy conclusions from these static cross-sectional comparisons. That portion of our article was an attempt to correct grossly misleading comparisons of Dutch and US rates in the American media.

We are taken to task for using the Dutch school survey data from the Trimbos Institute, rather than data from Statistics Netherlands or the CEDRO Amsterdam survey. As noted below, we did in fact report CEDRO estimates. But the 1990s Amsterdam trends mentioned by Abraham et al are not relevant to our commercialisation thesis; as we explained in our article, the dramatic growth in cannabis commercialisation in Amsterdam occurred between 1980 and 1988 and almost every Western nation saw increases in cannabis use after 1992 for reasons apparently unrelated to drug policy.

We are delighted to learn of the national Statistics Netherlands estimates, which as far as we can tell have not been cited previously in the English-language literature - although the search engine on their website produces no statistics for "drugs", "drug", "cannabis" or "marijuana". But now we are puzzled as to why a 1997 paper by Marieke Langemeijer announcing CEDRO's own national survey stated that "The implementation of the national survey means that finally, The Netherlands will have a decent source of data that serves multiple purposes among which the basic information for health care, prevention, education and drug policy. Hopefully, it is the beginning of a high quality drug research tradition". Similarly, a CEDRO press release of 14 April 1998 stated that "figures for the entire country will soon no longer have to be based on local surveys since a national study on drug use in The Netherlands is currently being carried out by CEDRO". Moreover, neither the CEDRO nor the Trimbos researchers mention these data in their English-language monographs on Dutch drug use trends.

Our Fig. 1 showed that during the 1984-1992 period the Trimbos lifetime prevalence estimates rose even more steeply for the age 16-17 group than for the age 18-20 group. This clearly undermines the concern raised by Abraham et al about a selection bias involving older students, but at any rate, that criticism misses the point. Sampling biases of the Trimbos school survey do not preclude its use for studying trends over time. Moreover, our trend analysis compared it to age 18-20 trends from the US Monitoring the Future school survey. The Trimbos researchers state that their survey was designed to permit comparisons to that US survey (see Plomp et al, 1991: 11).

Abraham et al complain that we averaged non-comparable estimates, but fail to mention that we grouped our estimates so that ‘city v. nation’ averages and ‘nation v. nation’ averages were presented separately. We think our averaging was well within contemporary standards of meta-analysis, but no matter - we presented the raw data so readers could decide for themselves. At any rate, no conclusions of our work hinged on these averages - indeed, we did not even include them in our presentation of these data in our forthcoming book, Drug War Heresies (MacCoun & Reuter, 2001b).

Abraham et al suggest that our alleged inattention to the geographical issue undermined our inferences about the effects of commercialisation. On the contrary, the fact that cannabis prevalence is higher in Amsterdam is quite consistent with our hypothesis. During the 1980s, when we contend the commercialisation effect occurred, various estimates suggest that over a quarter of all Dutch cannabis coffee shops were in Amsterdam, yet Amsterdam accounted for only about 5% of the total Dutch population. As late as 1997, Abraham et al (1999) reported that last-year users from the highest-density Dutch addresses were more likely to cite coffee shops as their cannabis source than were users from low-density Dutch addresses.

As we stated in the article, the evidence for our commercialisation hypothesis was indirect and at best purely correlational, though we noted that it is consistent with evidence on gambling, tobacco and alcohol marketing. Moreover, the quasi-legal status of the Dutch system, which tends to keep prices high, almost surelyunderstates the likely commercialisation effects of full legalisation. Given weak data, our inferences may well be wrong, but we think the comments of Abraham et al shed little light on that question.

REFERENCES

Abraham, M. D., Cohen, P. D. A., van Til, R. J., et al (1999) Licit and Illicit Drug Use in the Netherlands. Amsterdam: Centre for Drug Research (CEDRO).

Langemeijer, M. (1997) The prevalence of illicit drug use in the general population and in schools, as monitored by a number of different methods. In Invitational Conference on Monitoring Illicit Drugs and Health: Final Report pp. 11-21. Utrecht: Trimbos Instituut.

MacCoun, R. & Reuter, P. (1997) Interpreting Dutch cannabis policy: reasoning by analogy in the legalization debate. Science, 278, 47-52.

MacCoun, R. & Reuter, P. (2001a) Evaluating alternative cannabis regimes. British Journal of Psychiatry, 178, 123-128.

MacCoun, R. & Reuter, P. (2001b) Drug War Heresies: Learning From Other Vices, Times, and Places. Cambridge: Cambridge University Press.

Plomp, K. N., Kuipers, H. & van Oers, M. L. (1991) Smoking, Alcohol Consumption and the Use Of Drugs by Schoolchildren from the Age of 10. Amsterdam: VU University Press.

Trimbos Instituut (2001) Fact Sheet: Cannabis Policy, Update 2000. http://www.trimbos.nl/indexuk.html