VoxEU Column Health Economics

What should we do about cannabis?

More than one in five Europeans has taken cannabis at some point in their lives. This column explores the issues facing policymakers trying to deal with marijuana.

No serious commentator doubts that cannabis is potentially damaging to the user. Like tobacco, it is typically smoked and thus shares the potential for lung disease. Like alcohol, it affects reaction times and may raise the risk of road accidents. Cannabis has also been associated with cognitive impairment, deterioration in education performance (van Ours and Williams 2008), and psychotic illness (Arsenault 2004). Moreover, cannabis is often – albeit contentiously – seen as a causal gateway to more serious drug use (Kandel 2002). The question is what to do about it?

The age of users is key. Among recent cohorts, cannabis use tends to begin early or not at all. Figure 1 shows the pattern of prevalence by age in England and Wales depicting the proportion using the drug within the last year or at any time in the past. Prevalence reaches a peak by age 20 (with the risk of onset peaking two or three years earlier) and there is a subsequent decline – partly the result of tastes changing with age and partly an result of the lower rates of drug use in earlier birth cohorts. Age is critical. Early initiation into cannabis use is empirically (and perhaps causally) associated with heavier lifetime use (Pudney 2004) and with adverse developmental outcomes.

Figure 1. The age profile of cannabis prevalence (i) within the previous 12 months (ii) any time in the past (England and Wales Offending Crime and Justice Survey 2003)

Yet the existence of significant harm to the user is not sufficient to justify prohibition. Indeed, expert opinion suggests that direct harm to cannabis users is relatively modest compared to other illegal drugs – and legal ones, particularly alcohol and tobacco.

Figure 2, adapted from Nutt et al. (2007) summarises this in relation to the current UK classification of illegal drugs into classes A (with the heaviest penalties) through C (with the lightest). We could easily add other activities such as overeating and physical inactivity to emphasise further the weak relationship between something being legal and its potential for personal harm.

Figure 2. Subjective expert assessment of relative drug harms (adapted from Nutt et al. 2007)

Cannabis is formally illegal everywhere and all countries are required by the 1961, 1971 and 1988 UN conventions on narcotics to maintain a prohibitionist policy on possession and supply for non-medical use. Despite this policy, it is widely consumed in developed countries and it has been estimated that, for recent cohorts in the UK, the proportion of people ever having used cannabis by the age of 30 will be 60% or more (Pudney 2004). Given the ubiquity of cannabis, the modest level of perceived harm and the failure of prohibition to prevent widespread use, there is some discomfort with prohibitionist policy and many countries have liberalised their enforcement and penalisation regimes (see MacCoun and Reuter 2002), but all have so far stopped short of full legalisation.

Liberalisation of cannabis in the UK

A recent example of liberalisation was the UK reclassification of cannabis in 2004 from class B to class C, which reduced maximum penalties from 5 to 2 years’ imprisonment for possession, although the maximum penalty for supply was retrained at 14 years. Yet the new policing guidelines meant that cannabis possession effectively became a non-arrestable offence, dealt with by confiscation and warning rather than a criminal charge.

What effect did this have on cannabis use? As far as we can tell from passive observation, there was none at all. Figure 3 shows results from analysis of a series of annual surveys of 11-15 year old English school children. For each year, we predict the probability for a hypothetical child at ages 11-15 reporting:

  • having used cannabis
  • having ever been offered it;
  • seeing its use as “OK”.

Despite the predictions of a cannabis epidemic, Cannabis use, availability, and acceptability all show a continued decline through the period of reclassification.

Figure 3. Predicted age profile of prevalence, availability, and attitudes to cannabis by year (year-specific probit models for Survey of Smoking, Drinking and Drug Use among Young People, 2001-2007).

Segmenting the market

An obvious policy aim is to separate the markets for cannabis and harder drugs as far as possible. If cannabis and hard drugs share the same channels of retail supply, there is a risk that cannabis users will come under pressure to migrate to hard drugs, increasing the risk of long-term harm. The Dutch coffee shop policy can be seen as an attempt to segment the drugs market. Certain outlets are permitted to make small retail sales of cannabis (up to 5 grams) for consumption on the premises, with strict monitoring to prevent larger sales, underage consumption, and sale of harder drugs. In contrast, US courts (at the federal level, at least) penalise cannabis supply more heavily relative to hard drug supply than is typical in Europe. Although too weak to be anything more than suggestive, there is some evidence that the incentives created by this difference in relative penalties translates into a higher rate of cannabis/hard drug co-supply in the US than in England. Using matching methods to control for confounding factors, we find that regular cannabis users in the US seem to have significantly easier access to hard drugs than non-users of cannabis, while we find no significant difference in the English data.

An extreme form of market segmentation is full legalisation, which gives governments the freedom to impose taxes and regulate the market through controls on advertising, consumption in public places, minimum purchasing age and the physical characteristics of the product itself. The last of these is an interesting possibility, since the levels of the main psychoactive ingredient, tetrahydrocannabinol, and another possibly anti-psychotic component, cannabidiol, are believed to be relevant to the risk of psychotic illness and might conceivably be subject to controls in a legalised market. There is some (contested) evidence from the tobacco and alcohol markets that non-price controls of various kinds have significant restraining effects on consumption (Evans et al 1999, Saffer and Chaloupka 2000, Saffer and Dave 2002, Fidler and West 2010).

The big questions hanging over any move to a legal regulated market are: the set of consumer responses to (presumably) lower market price and increased availability; the size of the residual black market at realistic levels of excise tax; and finally the effectiveness of health messages when not backed by prohibition. I suspect that the black market is a serious constraint on our ability to impose large excise taxes – in the tobacco market where domestic production is virtually nil, it is estimated that smuggled goods make up 20% or more of the market (HMRC 2008). On the other hand, separation of the health message from government may increase its effectiveness, since doctors are much more trusted by young people than politicians (Torrey-Purta et al. 2001).

Natural experiments

The disappointing conclusion – which contrasts with the apparent certainty displayed by some commentators – is that the available evidence is too weak for us to be at all certain about the consequences of radical policy change, as Kilmer et al. (2010) have shown for the legalisation proposals currently under consideration in California. But it is certainly reasonable to entertain the possibility that a legalised, carefully regulated cannabis market could produce better social outcomes than our present system of half-hearted, partially-effective prohibition.

In my view, the best way to begin putting policy on a better footing is to allow more variation in policymaking, including the legalisation option. This could be done if domestic supply and consumption of cannabis were removed from the international drug prohibition treaties, while retaining the ban on international trade in the drug. At present we are limited to decriminalisation unaccompanied by the instruments of regulation available for legal markets, so that the potential benefits of a non-prohibitionist approach are largely precluded. The removal of cannabis from the UN treaty structure would pass the responsibility for cannabis policy back to national governments, with freedom of action to pursue independent policies. Some will choose to stick to the status quo, others will choose decriminalisation or legalisation. In doing so, we will certainly have the chance to learn more about the effects of policy.


Arseneault, L, Cannon, M, J Witten, R Murray (2004), “Causal association between cannabis and psychosis: examination of the evidence”, British Journal of Psychiatry, 184:110–7.

Evans, WN, MC Farrelly, and E Montgomery (1999), “Do workplace smoking bans reduce smoking?”, American Economic Review, 89:728-747.

Fidler, JA and R West (2010), “Changes in smoking prevalence in 16-17-year-old versus older adults following a rise in legal age of sale: findings from an English population study”, Addiction 105, 1984-1988.

HMRC (2008), Tackling Tobacco Smuggling Together.

Kandel, DB (2002), Stages and Pathways of Drug Involvement, Cambridge University Press.

Kilmer, B, JP Caulkins, RL Pacula, RJ MacCoun, and P Reuter (2010), “Altered State? Assessing How Marijuana Legalization in California Could Influence Marijuana Consumption and Public Budgets”, Santa Monica: RAND Corporation.

MacCoun, RJ and P Reuter (2001), Drug War Heresies. Learning from Other Vices, Times and Places, Cambridge University Press.

Nutt, D, LA King, W Saulsbury, and C Blakemore (2007), “Development of a rational scale to assess the harm of drugs of potential misuse”, The Lancet, 369:1047-1053.

Pudney, SE (2010), “Drugs policy: what should we do about cannabis?”, Economic Policy, 25:165-211.

Torrey-Purta, J, R Lehmann, H Oswald, and W Schulz (2001), Citizenship and Education in Twenty-eight Countries: Civic Knowledge and Engagement at Age Fourteen. Amsterdam: IEA.

van Ours, JC and J Williams (2008), “Why parents worry: initiation into cannabis use by youth and their educational attainment”, Journal of Health Economics, 28:132-142.

210 Reads