Boekhout van Solinge, Tim (1997), Drugs in France: Prevalence of use and drug seizures. In: Korf, D.J. & H. Riper (Eds.) (1997), Illicit drugs in Europe. Proceedings of the Seventh Annual Conference on Drug Use and Drug Policy. Amsterdam: Universiteit van Amsterdam, Faculteit der Rechtsgeleerdheid.
© Copyright 1997 Tim Boekhout van Solinge. All rights reserved.


Drugs in France

Prevalence of use and drug seizures

Tim Boekhout van Solinge


Statistics play an important part in national drug policies. They are widely used to evaluate the success or failure of a country’s policies. National figures are also used for cross-national comparison, but often without the original sources or the methodology being known. This paper raises questions about the reliability of available drugs data, using France as a case in point. How sound are these data and how far are they representative of the phenomena they are supposed to clarify? Two types of data are examined: prevalence data on drug use (specifically, numbers of heroin addicts) and data on drug seizures (specifically heroin). The paper investigates the reliability of these French data and the apparent discrepancy between the amounts of drugs the authorities claim are being used and the quantities that are being seized.


The available French data will be used here to illustrate some methodological difficulties in arriving at reliable drug prevalence and drug seizure data. Such problems obviously exist in many countries. My principal reason for choosing France is that my experience with these data enables me to critically analyse it.

Prevalence of use: drug addicts in France

Few sound figures are available on the prevalence of drug use in France. Estimates range from 50,000 to several million drug users (Padieu, 1994). Such differences relate in part to the definition being applied. Often no clear definition is even given, causing ambiguity about whether figures refer to the number of people that have ever used drugs (lifetime prevalence), recent users (past year) or current users (past month). As a case in point, the research institute Baromètre Santé recently claimed that lifetime prevalence of cannabis use in a given age group dropped 40% within two years’ time (CFES, 1995). Now while lifetime prevalence may vary slightly from one year to another, a decline in the order of 40% within one age group is out of the question, because someone who has ever used cannabis will always remain in the statistics. This anomaly was ignored here and the figures have been presented uncritically as ‘facts’. Another problem is that many prevalence statistics are rendered useless by failing to specify the type of drug. A variety of figures are also in circulation on the numbers of addicts in France. Again, the figure arrived at depends largely on one’s definition of a drug addict and the types of drugs involved.

In 1995 a statistical handbook of drug data was published entitled Drogues et toxicomanies, indications et tendances (Carpentier and Costes, 1995). It put the number of heroin addicts in France at 160,000. This number has since been widely cited and has come to be the official estimate. It seems worthwhile to analyse just how this figure was calculated. The usual sources used for estimating numbers of addicts are the police and justice system and/or the health and treatment system. This produces a high degree of uncertainty, since not all addicts are known to these two systems. The estimate of 160,000 heroin addicts derives from the treatment system. In the original publication it was warned that it is a minimum estimate, being based on those known to the treatment system (Carpentier and Costes, 1995). This is often subsequently forgotten in government publications. There is great uncertainty as to what proportion of the population of heroin addicts is known to a system. Figures vary, but some sources put this at no more than 50% for the treatment system (Boekhout van Solinge, 1996).

Even within the group known to the treatment centres, estimation causes problems too. The first is the anonymity guaranteed by French drug legislation to a person entering the treatment system. This means that clients are not formally registered as heroin addicts. However positive the intentions, it does make it virtually impossible to get exact figures. Other methods of estimation are therefore used. In France this is the `demographic theory of stationary populations’ (Costes, 1992), a general theory applied to stable populations with zero population growth. It calculates the total population by multiplying the number of newborn babies by the average life expectancy. When it is applied to estimate drug addicts, newborn babies are replaced by heroin addicts in treatment for the first time, and life expectancy becomes the average duration of heroin addiction. The total number of addicts would thus be the number of new addicts times the average duration of addiction.

The estimated figure for newcomers to the treatment system is 20,000 and the presumed average duration of addiction is 8 years, yielding 160,000 heroin addicts. This very simple calculation shows us how the addict population is calculated in France. Both assumptions involved are extremely speculative. First, the presumed average addiction of eight years seems dubiously low, and it turns out to be derived from one small survey (INSERM, 1993). Second, the figure of 20,000 new addicts is open to question. In view of the required anonymity, it is impossible to know how many addicts enter the treatment system every year. The only source for estimating this is a survey carried out each November (SESI, 1995), and the only certitude it gives is the number of newcomers that month, which came to 3,800 in the year in question. This is then multiplied by 5.3 to get the number on an annual basis. Why multiply by 5.3 and not by 12? That is because the specialized institutions, the only branch of the system with figures for both the whole year and for November, came up with a factor of 5.3. This multiplier has subsequently been applied to the treatment system as a whole. Notwithstanding all the guesswork involved here, this method furnishes the official estimate of the number of heroin addicts in France.

Drug seizures

The next question is how many of the drugs bound for the market are being seized, and how many drugs are on the market. The answer is likewise uncertain, but it is often assumed that approximately 10% of the drugs in circulation are being intercepted, a figure also applied by Interpol. This estimate has no scientific basis, so that we might speak of the `10 per cent myth’. The French police have even been known to suggest that 25% is being intercepted, a figure for which evidence is also lacking.

It may seem impressive that 5,000 kilograms of heroin are being seized in the European Union every year (Table 1), but the figures for cocaine (Table 2) are three to four times higher. Does this mean much more cocaine is used, or is so much more being seized than heroin? The answer is unknown, because little is known about cocaine consumption. Few cocaine users enter the treatment and/or justice systems, the usual sources of prevalence data.

How much do we know about heroin consumption, actually? It is often presumed we know a lot, and this is based on the assumption that most heroin addicts are known to the health and/or judicial system. Taking the estimate of 160,000 French heroin addicts, we could make a rough estimate of heroin consumption on a yearly basis. We will assume a minimum daily dosage of 0.25 gram, which derives from a study in Amsterdam. It is a minimum estimate, because compared to other cities not much heroin is used in Amsterdam. Many heroin addicts receive methadone there, and the quality of the heroin is high — factors which reduce the intake of heroin. We may therefore presume that the average daily consumption in France is higher than in Amsterdam.

Table 1. Quantities of heroin seized in the European Union.
Belgium 107
Germany 1,438
France 327
Italy 1,358
Netherlands 570
Portugal 55
Spain 672
United Kingdom 547
Other EU countries 317
Total 5,391
Source: Europol/Europol Drugs Unit (1995)

Applying the figure of 0.25 gram nevertheless to the presumed 160,000 French addicts yields a total daily heroin consumption in France of 40 kilograms, or 15,000 kilos on an annual basis. The quantity of heroin seized in recent years was 400 to 500 kilos per year (Boekhout van Solinge 1996). That is only 3% of the rough minimum yearly consumption, far below the 10% generally assumed or the 25% claimed by the French police. And, as noted, the total annual consumption may well be much higher, meaning that far less than 3% of the heroin is being intercepted. Another indication that this is indeed the case is that quantities of drugs seized are continually rising, but prices have not increased. If seizures constituted a significant portion of the total, prices would go up in accordance with the law of supply and demand. Even gigantic seizures of cocaine, for example, have no influence on price trends. Much to the contrary, prices in Europe are declining everywhere.

Table 2. Quantities of Cocaine Seized in the European Union.
Belgium 1,221
Germany 1,332
France 1,625
Italy 1,345
Netherlands 3,433
Portugal 1,893
Spain 4,454
United Kingdom 2,248
Other EU countries 171
Total 17,722
Source: Europol/Europol Drugs Unit (1995)


This paper has described the methodological difficulties in generating prevalence data on drugs, illustrating the extreme unreliability of existing data. Available data from France were used as an example to support this general hypothesis. It should be repeated that the estimates used here are by no means sound, for very little is known about numbers of drug addicts, the amounts being consumed, and the ratio of quantities seized to total quantities of drugs on the market.

In view of the importance usually attributed to drugs data, especially in political debates on drugs policy, researchers should emphasize the high degree of uncertainty in the available data and the need to improve data collection methods. A first step might be to incorporate such issues into research programmes and to conduct more specific methodological research on them.


Boekhout van Solinge, T. (1996) ‘Cannabis in Frankrijk (Le cannabis en France)’ In: Cohen, P. and Sas, A. (1996) (eds) Cannabisbeleid in Duitsland, Frankrijk en de Verenigde Staten, Amsterdam: Centre for Drug Research (CEDRO), University of Amsterdam.

Boekhout van Solinge, T. (1996) Heroïne, cocaïne en crack in Frankrijk. Handel, gebruik en beleid (L’héroïne, la cocaïne et le crack en France. Trafic, usage et politique). Amsterdam: Centre for Drug Research (CEDRO), University of Amsterdam.

Carpentier, C., Costes, J.M. (1995) Drogues et toxicomanies. Indications et tendances, Paris: Délégation Générale à la Lutte contre la Drogue et la Toxicomanie (DGLDT) & Observatoire Français des Drogues et des Toxicomanes (OFDT).

CFES (1995) Baromètre Santé 93/94. Paris: Editions CFES.

Costes, J. M. (1992) Pour une estimation du nombre de toxicomanes. Paris: Ministère des Affaires Sociales, de la Santé et de la Ville.

Europol/Europol Drugs Unit (1995) Drugs Seizures Statistics. The Hague: Europol/EDU.

INSERM (1993) Base de données en toxicomanie. Toxicomanes consultant dans les institutions spécialisées 1991-92, Institut National de la Santé et de la Recherche Médicale (INSERM).

Observatoire Géopolitique des Drogues (1994) Etats des drogues, drogue des Etats. Paris: Hachette.

Observatoire Géopolitique des Drogues (1995) Géopolitiques des drogues 1995. Paris: La Découverte.

Observatoire Géopolitique des Drogues. La Dépêche Internationale des Drogues, monthly publication in English, French and Spanish.

Padieu, R. (1994) L’information statistique sur les drogues et les toxicomanes, Paris: La Documentation Française.

Schiray, M. (ed) (1992) Penser la drogue, penser les drogues. II – Les marchés interdits de la drogue. Paris: Editions Descartes.

Service des Statistiques, des Etudes et des Systèmes d’Information (SESI), (1995) La prise en charge des toxicomanes dans les structures sanitaires et sociales en novembre 1993 (`enquête de novembre’). Paris: Ministère des Affaire Sociales, de la Santé et de la Ville.