What can we learn from the Dutch cannabis coffeeshop system?

Robert J. MacCoun
Addiction (2011) 106(11):1899-910
November 2011

what-can-we-learnIn 1976 the Netherlands adopted a formal written policy of non-enforcement for violations involving possession or sale of up to 30 g of cannabis. The ‘gateway theory’ has long been seen as an argument for being tough on cannabis, but interestingly, the Dutch saw that concept as a rationale for allowing retail outlets to sell small quantities. Rather than seeing an inexorable psychopharmacological link between marijuana and hard drugs, the Dutch hypothesized that the gateway mechanism reflected social and economic networks, so that separating the markets would keep cannabis users out of contact with hard-drug users and sellers.

application-pdfDownload the paper (PDF)

This essay examines what the available data can tell usabout how these retail coffeeshop sales may have influenced cannabis use and its consequences. Three types of indicators are examined: the prevalence and intensity of cannabis use; market indicators, including prices, purity, seizures and enforcement; and treatment data as a partial indicator of harms to users. Other types of consequences (for public safety, public order, economic productivity, family life, health and personal enjoyment) are not examined because they are so difficult to quantify and because they pose such severe causal identification problems.

This essay does not seek to judge the Netherlands or critique its internal policy decisions—which as we will see are undergoing significant change. Rather, the goal is to see what other jurisdictions might learn about potential policy options and outcomes by drawing upon the Dutch experience and their energetic efforts to document their policies and outcomes. There are daunting analytical challenges in making cross-national comparisons of drug policies and outcomes, but if we want to identify more effective policies we need to make comparisons across jurisdictions, and it is surely better to make provisional judgments than provincial ones.

Results: The available evidence suggests that the prevalence of cannabis use among Dutch citizens rose and fell as the number of coffeeshops increased and later declined, but only modestly. The coffeeshops do not appear to encourage escalation into heavier use or lengthier using careers, although treatment rates for cannabis are higher than elsewhere in Europe. Scatterplot analyses suggest that Dutch patterns of use are very typical for Europe, and that the ‘separation of markets’ may indeed have somewhat weakened the link between cannabis use and the use of cocaine or amphetamines.

Conclusions: Cannabis consumption in the Netherlands is lower thanwould be expected in an unrestricted market, perhaps because cannabis prices have remained high due to production-level prohibitions. The Dutch system serves as a nuanced alternative to both full prohibition and full legalization.



Whether the Dutch should go forward with some variant of this coffeeshop model in the future is uncertain, and Dutch citizens may feel it is none of our business anyway. However, the Dutch experience is potentially informative for people in other countries who are debating whether to legalize cannabis sales.

The best available evidence paints a nuanced picture. Dutch citizens use cannabis at more modest rates than some of their neighbors, and they do not appear to be particularly likely to escalate their use relative to their counterparts in Europe and the United States. Moreover, there are indications that rather than increasing ‘the gateway’ to hard drug use, separating the soft and hard drug markets possibly reduced the gateway.

However, the Dutch experience also raises some cautionary notes. There are several lines of circumstantial evidence that the Dutch retail system increased consumption, especially in its early years when coffeeshops were spreading, open to 16-year-olds and advertised more visibly than they do today. If so, this increase occurred in a hybrid system in which high-level enforcement probably served to keep prices from dropping the way they might in a full-scale legalization scheme.

Many people look to the Netherlands as a model for what might happen if cannabis were legalized, but what the Dutch have done is quite different, and far more nuanced, than the kind of full-scale legalization that is usually debated in other countries. The Dutch system is ambiguous by design, and it is ambiguous in ways that give officials leverage over prices and sales in ways that might be far harder to achieve in a full-scale legalization regime.