The global spread of cannabis
Trends and patterns of abuse
Cannabis markets in the 1990s
Cannabis cultivation and production
Author: UNDCP Research Section
Creation Date: 1999/12/01
Research Section, Policy Development and Analysis Branch, United Nations International Drug Control Programme
Cannabis is by far the most widely cultivated, trafficked and abused illicit drug. Half of all drug seizures worldwide are cannabis seizures. The geographical spread of those seizures is also global, covering practically every country of the world. About 140 million people, 2.5 per cent of the world population, consume cannabis (annual prevalence) compared with 0.3 per cent consuming cocaine and 0.2 per cent consuming opiates. In the present decade, cannabis abuse has grown more rapidly than cocaine and opiate abuse. The most rapid growth in cannabis abuse since the 1960s has been in developed countries in North America, western Europe and Australia. Cannabis has become more closely linked to youth culture and the age of initiation is usually lower than for other drugs. An analysis of cannabis markets shows that low prices coincide with high levels of abuse, and vice versa. Cannabis appears to be price-inelastic in the short term, but fairly elastic over the longer term. Though the number of cannabis consumers is greater than opiate and cocaine consumers, the lower prices of cannabis mean that, in economic terms, the cannabis market is much smaller than the opiate or cocaine market. Given the present status of knowledge, the extent of cannabis cultivation and production is very difficult to measure or even estimate. The two most widely used estimates of cannabis production vary by a factor of 30, ranging from 10,000 tons to 300,000 tons. A process of triangulation, that is, linking production to consumption estimates, suggests that a reasonable order of magnitude for worldwide cannabis production would be about 30,000 tons.
Comparison in terms of spread
The production, trafficking and consumption of cannabis are far more widespread today than for any other illicit substance. The United Nations International Drug Control Programme (UNDCP) estimates that some 2.5 per cent of the global population may consume cannabis (annual prevalence), compared with 0.5 per cent consuming amphetamine-type stimulants, less than 0.3 per cent abusing cocaine and less than 0.2 per cent of the global population taking heroin. (1)
About half of all seizure cases of illicit drugs in the world concern cannabis (1996) (2) (see figure I). In terms of weight cannabis resin (hashish) seizures are almost 3 times as large as cocaine seizures and 30 times as large as heroin seizures. Global seizures of cannabis herb (marijuana) in terms of volume are almost 10 times as large as seizures of cocaine and almost 100 times as large as seizures of heroin (1996) (see figure II). Expressed in terms of dosages--to have a better direct basis for comparison--there is at least five times as much cannabis seized worldwide as heroin or cocaine (see table 1).
Table 1. Global seizures in tons and expressed as dosages, 1996
|Substance||Seizures (tons)||Effective doses (grams) a||Seizures expressed in billions of effective doses|
|Marijuana||3 015||0.350 b||0.0|
|Subtotal, cannabis||3 862||-||6.3|
Sources: UNDCP, annual reports questionnaire data; R. S. Gable, "Toward a comparative overview of dependence potential and acute toxicity of psychoactive substances used non-medically", American Journal of Drug and Alcohol Abuse, vol. 19, No. 30 (1993), pp. 263-191.
a The effective dosage indicated is the estimated median amount for an average adult human weighing 70 kilograms. Dosage amounts refer to the quantity of active substance administered; they do not include inactive material unless specifically mentioned.
b One half of a joint of a 700-milligram cannabis cigarette with a tetrahydrocannabinol content of at least 1.5 per cent.
c United States Drug Enforcement Administration (DEA) standard dosage unit.
d Transformation rate based on five lines of one fourth of a gram of cocaine containing 40 per cent cocaine and 60 per cent adulterants; street doses, according to the International Narcotics Control Board (INCB), are 100-100 milligrams, which would be equivalent to 1.6-3.2 billion doses.
e Pure heroin dose according to the DEA standard dosage unit; calculations have been adjusted for average purity of 40 per cent found in seizures; based on the INCB concept of average daily dose (0.3 gram), seizures would be equivalent to about 1 billion dosages; based on the INCB concept of street doses (0.1-0.5 gram), seizures would be equivalent to 0.06-0.3 billion doses.
Data also show that far more countries worldwide are affected by trafficking and abuse of cannabis than by other drugs such as heroin, cocaine or the various synthetic drugs. Almost all countries--183 out of 187 countries and/or territories reporting to UNDCP over the period from 1990 to 1996--reported seizures of cannabis herb or cannabis resin (see figure III). All the data therefore suggest that cannabis is the leading drug of abuse at the global level.
Global abuse trends
Cannabis originated in central Asia and spread across the globe over the last 5,000 years. It had reached all continents by the beginning of the twentieth century. Since cannabis abuse was not common among the major Powers at the beginning of the century, political interest in putting it under international control was limited. The first international drug control convention, the Hague Convention of 1912, included the opiates and cocaine, but not cannabis. Control of the international trade in cannabis only started with the 1925 Convention of the League of Nations, mainly as a consequence of a proposal by Egypt. (3) It took a further three and a half decades until cultivation of the cannabis plant was made subject to international control under the Single Convention on Narcotic Drugs of 1961. At that time significant levels of consumption and/or production were reported from countries in the Middle East/North Africa region (notably Egypt, Lebanon and Morocco), in south Asia (India, Nepal, Pakistan and Sri Lanka), in south-east Asia, in Latin America and the Caribbean (including Brazil, Jamaica and Mexico) and in a number of areas in west Africa.
By contrast, production and consumption of cannabis in what are, today, the largest markets--North America, Europe and Australia--were still very limited. That started to change in the 1960s, first in North America and later in both Australia and Europe. Rising demand led to increased cannabis production, first of all in the traditional cannabis-producing countries and more recently in the developed countries, where modern cultivation techniques have led to cannabis with a far higher tetrahydrocannabinol (THC) content.
Growing levels of production and consumption are also reflected in more trafficking and seizures. Cannabis seizures fluctuate from year to year. Nonetheless, the long-term trend clearly points upwards. Average annual seizures of cannabis plants and herb over the last decades were twice as high as a decade earlier and even more than six times larger than two decades earlier (see figure IV). Seizures of cannabis resin in the last decade were on average more than 3 times higher than a decade earlier and 15 times larger than two decades earlier (see figure V).
There was not only an overall increase in the amounts seized, but also a growing geographical spread of cannabis trafficking across the globe, which continued over the last decade. The number of countries reporting seizures of cannabis herb to UNDCP rose from 95 in 1986 to 143 in 1996; the number of countries reporting seizures of cannabis resin grew from 64 in 1986 to 79 in 1996.
Since seizures are only an indirect indicator of consumption and production trends and may simply reflect levels of enforcement, it is instructive to compare cannabis seizure and prevalence rates over a long period. Such an analysis, in the context of the United States of America shows a very high positive correlation. The more cannabis is on the market, apparently the higher the consumption and the higher, ceteris paribus, the likelihood of seizures. The correlation is particularly strong once annual seizure data are smoothed to reduce distortions due to individual large seizures in specific years. Cannabis seizure data for the United States, smoothed by a moving three-year average, compared with the 30-day cannabis prevalence data for senior high-school students, (4) show a high correlation coefficient (R 2 = 0.92) for the period from 1978 to 1997 (see figure VI). Such results come close to a perfect fit, suggesting that long-term trends in smoothed cannabis seizure statistics are a good reflection of consumption trends.
What apparently happened in the 1960s--and is still valid today--has been the entry of cannabis into mass youth culture, first in the United States, then in Europe and subsequently in the rest of the world, (5) supported by common trends in music and fashion, as well as a rapidly growing supply, turning the illicit cannabis trade into a multibillion dollar business. (6) In almost all countries, there is currently a concentration of cannabis consumption among teenagers and young adults, more so than for other drugs (see table 2).
The age of initiation is generally lower for marijuana than for drugs such as heroin or cocaine. In the mid-1990s, the average age of initiation for cannabis in Spain was 17 years, compared with 20.6 years for heroin and 22.3 years for cocaine. (7) In the United States the average age of initiation was 16.7 for cannabis, 19.1 for cocaine and 19.3 for heroin (1995). (8) The mean age of first use of cannabis in France is around 16 years (1996). (9) In Australia, the average age of initiation is 17.5 years for cannabis, 20 for heroin and 20.4 for cocaine (1995). (10)
Table 2. Importance of cannabis (lifetime prevalence) among 15- to 16-year-olds, compared with other drugs (percentage), in European Union countries, Australia, Canada and the United States of America
|Country or area, year of survey||All illegal drugs||Cannabis||Ampheta-mine-type stimulants||Ecstasy||LSD||Cocaine||Heroin|
(15- to 16-year-olds)
|51.3 a||47.2||7.9||4.6||11.9 b||3.7||4.1|
|United States, 1996
(tenth graders 16- year-olds)
|45.4||39.8||17.7 (4.4) c||5.6||9.4||7.1||2.1|
(15- to 17-year-olds)
|30.4||28.8||10.7 c||-||10.7 c||1.8||10.7 c|
|European Union countries, 1994/95; unweighted averages
(15- to 16-year-olds) d
Sources: European Monitoring Centre for Drugs and Drug Abuse Addiction, Annual Report on the State of the Drugs Problem in the European Union 1997 (Lisbon, 1997); National Institute on Drug Abuse, Monitoring the Future Study, vol. I (Rockville, Maryland, 1997); Canadian Centre on Substance Abuse, Canadian Profile 1997 (Ottawa, 1997); Centre for Behavioural Research in Cancer, "Australian Secondary Students' Use of Over-the-Counter and Illicit Substances in 1996" (November 1998).
a Estimate; based on the proportion of cannabis in all illegal drugs (92 per cent) according to the 1993 Australian household survey.
b LSD and other hallucinogens.
c All stimulants (including licit ones such as caffeine); number in brackets is for "ice" (methamphetamine) only.
d Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom of Great Britain and Northern Ireland.
A generally observed pattern is the rapid rise in cannabis use among young people up to their early twenties and a decline thereafter. Canadian data from 1993 show almost a doubling in annual prevalence from the 15-17 age group to the 18-19 age group. In the latter group, cannabis prevalence is 3.2 times the general population average. (11) In Australia (1996), annual prevalence of cannabis use doubles from 13- to 14-year-olds to 16- to 17-year-olds (see table 3); between the ages of 12 (10.4 per cent) and 17 (48.2 per cent), annual prevalence of cannabis use quintuples. Among 17-year-olds cannabis use is 3.7 times the national average. (12) Similarly, United States household survey data (13) (1996) show that cannabis consumption (annual prevalence) rises sharply from the age of 12-13 to the age of 18-10. In the latter group, annual prevalence (30.1 per cent) is 3.5 times the average in the general population.
Table 3. Results of nationwide school surveys (lifetime prevalence) in European Union countries, Australia, Canada and the United States of America
|Country or area||13- to 14-year-olds||15- to 16-year-olds||
17- to 18-
|Selected European Union countries, weighted average a||6||16||26|
|Canada, 1995 (15- to 17-year-olds and 18- to 19-year-olds)||-||29||33|
|United States, 1995 (eighth-graders, tenth-graders and twelfth-graders, respectively)||20||34||42|
Sources: European Monitoring Centre for Drugs and Drug Addiction, Annual Report on the State of the Drugs Problem in the European Union (Lisbon, 1997); National Institute on Drug Abuse, Monitoring the Future Study (Rockville, Maryland, 1997); Canadian Centre on Substance Abuse, Canadian Profile 1997 (Ottawa, 1997); Centre for Behavioural Research in Cancer, Australian Secondary Students' Use of Over-the-Counter and Illicit Substances in 1996 (November 1998).
a Average, weighted by size of population; 13- to 14-year-olds: Austria, France, Greece, the Netherlands, Portugal and the United Kingdom; 15- to 16-year-olds: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom; and 17- to 18-year-olds: Austria, Belgium, France, Germany, Greece, Luxembourg, the Netherlands, Portugal, Spain and the United Kingdom.
The concentration of cannabis consumption among youth is also reflected in data from Europe. According to the 1996 British Crime Survey, (14) annual prevalence of cannabis among 16- to 19-year-olds (27 per cent) is three times as high as the national average (9 per cent for 16- to 59-year-olds). Data from the Amsterdam household survey (1994) (15) show prevalence rates of cannabis to rise from 4.7 per cent among 12- to 15-year-olds to 25 per cent among 20- to 24-year-olds, 2.5 times the average of cannabis abuse in Amsterdam (10.5 per cent) (see figure VII). Data from the Netherlands also indicate that cannabis is more of a youth drug than cocaine or heroin. Only Ecstasy (annual prevalence of 5.5 per cent among 20- to 24-year-olds) shows in relative terms an even stronger concentration among youth (4 times the average level of 1.4 per cent) than cannabis (2.5 times the average level of 10.5 per cent). For the European Union as a whole, (16) data suggest that lifetime prevalence of cannabis use among 13- to 14-year-old students quadruples from 6 per cent to more than 25 per cent of 17- to 18-year-old students (see table 3).
It may be interesting to note that overall levels of cannabis consumption in Europe are lower than in North America or Australia. The average lifetime prevalence of cannabis use for 15- to 16-year-olds in the European Union (1994-1995) is around 16 per cent, less than half the equivalent rate in the United States (34 per cent) (17) and only a third of the rate reported from Australia. About a quarter of all 17- to 18-year-olds have tried cannabis in the European Union, compared with more than 40 per cent of twelfth-graders (18-year-olds) in the United States and 55 per cent in Australia.
The spread of cannabis among 15- to 16-year-olds seems to have reached proportions similar to the United States in only two European countries: Ireland and the United Kingdom of Great Britain and Northern Ireland. The spread among students is also above the European Union average in the Netherlands, Spain and, to a lesser degree, Denmark, while the lowest levels of cannabis consumption in the European Union (5 per cent or less among 15- to 16-year-olds) are reported from two southern countries (Greece and Portugal) and the two northern countries (Finland and Sweden).
Though there will never be a definitive explanation for the drug abuse problem in all its dimensions, some key factors that have contributed to the spread of cannabis are, nonetheless, well established. In the 1960s, large segments of United States youth, often students, followed by young Europeans and Australians, began to experiment with cannabis as an expression of protest. Though the student movement gradually faded, the spread of cannabis consumption continued well into the late 1970s. Abuse had created a momentum of its own. The age of users starting to experiment with cannabis fell and by the mid-1970s initiation rates among the 12-17 age group started to exceed those for 18- to 25-year-olds, a trend that has not changed since (see figure VIII).
In the 1980s, the general climate in society changed, as did cannabis abuse, showing a clear downward trend among all age groups in the United States (see figures VIII and IX).
This trend was reversed in the 1990s, however. While general population surveys have revealed a general stabilization in the prevalence of cannabis use, it has definitely increased among the younger sections of society in the 1990s, both in terms of initiation (figure VIII) and prevalence (figure IX).
The general pattern of consumption trends in the United States can also be observed, to some extent, in other parts of the world. Most countries reported a rise in cannabis consumption over the period from 1960 to 1980, though the increase was less pronounced than in the United States. Following a period of stabilization in the 1980s, strong increases in a number of countries have been experienced again in the 1990s.
Developments in Europe are a good example in this regard. Having grown strongly since the late 1960s, cannabis consumption seems to have remained more or less stable in much of the 1980s, (18) though the picture varies from country to country and from city to city. (19) The overall stabilization of the 1980s was followed by marked increases in the 1990s.
Increasing consumption has been closely associated with youth culture. One recent study on the use of cannabis in France (20) --and the findings are true for most other European countries--has confirmed the ongoing process of integration and adaptation of cannabis use to changes in youth culture. Parallel to the increased popularity of synthetic drugs as part of the rave scene, there has been a dramatic increase in cannabis consumption in the European Union in the 1990s, leading to what in the United Kingdom has been termed the normalization of recreational drug use amongst young people (21) . According to the British Crime Surveys, annual prevalence of cannabis use among 16- to 19-year-olds rose from 18 per cent in 1991 (22) to 27 per cent in 1996; (23) lifetime prevalence of cannabis use in that age group rose from 26 to 35 per cent. Thus, one out of three youths has tried cannabis, one out of four young people has used it at least once over the last year, and 16 per cent of all youths used it in the month prior to the survey. (24) In the general population (16- to 59-year-olds) the annual prevalence of cannabis use increased from 5 per cent in 1991 to 9 per cent in 1996, bringing abuse levels in the United Kingdom up to those in the United States (see figure X).
Similar trends have been also reported from the Netherlands. Parallel to increased popularity of Ecstasy, cannabis strengthened its position as by far the most popular drug among youth. Monthly prevalence rates of cannabis use among pupils in the Netherlands (12- to 19-year-olds) rose from 2 per cent in 1984 to 3 per cent in 1988, reaching 11 per cent by 1996, a fourfold increase over the last decade. (25)
Monthly prevalence rates among 18- to 19-year-olds (18.1 per cent) in the Netherlands in 1996 were still slightly lower than in the United States (21.9 per cent in 1996), but were about a third of United States levels a decade ago (see figure XI). Monthly prevalence rates among 16- to 19-year-olds in the Netherlands (17.8 per cent) are already higher than in the United Kingdom (16 per cent). That comparison may not be conclusive, however, because in household surveys (such as the British Crime Survey), admission of illegal activities tends to be lower than in anonymous school surveys. According to school surveys investigating lifetime prevalence rates, cannabis use among 16- to 17-year-olds in the Netherlands rose from 13 per cent in 1988 to 34 per cent in 1996 (26) (that is, one out of three has been experimenting with cannabis).
Following only moderate increases in the 1980s, Germany also reported a strong rise in cannabis consumption in the 1990s. Overall prevalence of cannabis use among 18- to 39-year-olds rose from 4.1 per cent in 1990 to 7.7 per cent in 1995 (see figure XII). In the western Länder (states of the pre-unification Federal Republic of Germany), annual prevalence of cannabis consumption almost doubled, from 4.9 per cent in 1990 to 8.8 per cent in 1995; in the eastern Länder (the former German Democratic Republic), prevalence rates increased fivefold, from 0.7 per cent in 1990 to 3.5 per cent in 1995. (27) Overall levels of cannabis consumption in Germany (4.4 per cent among 18- to 59-year-olds in 1995) are slightly lower than the European Union average but higher than the pan-European average (including countries of eastern Europe). Nonetheless, compared with the Netherlands, the United Kingdom or the United States, the overall level of cannabis consumption in Germany is still rather low, about half the level in those countries.
Source: Germany, Ministry of Health, Representativerhebung 1995 (Bonn, 1995); and Bundeskriminalamt, Rauschgiftjahresbericht 1997 (Wiesbaden, 1998).
Though levels of cannabis abuse still differ substantially among countries in Europe--from low levels in the Nordic countries (Sweden and Finland) to high levels in the Netherlands, Spain and United Kingdom (the same countries that are also faced with high levels of consumption of amphetamine-type stimulants) (28) --overall trends in cannabis consumption seem to have been rather uniform across Europe in recent years. Rising levels of cannabis abuse during the period from 1994 to 1997 were reported to UNDCP by a number of European countries, including, as mentioned before, Germany, the Netherlands and the United Kingdom, as well as by Belgium, Finland, Norway, Sweden and a number of eastern European countries such as Bulgaria, Croatia, the Czech Republic, Hungary, Latvia, Lithuania, Poland, the Republic of Moldova, Slovakia and Ukraine. Most of the remaining countries of Europe either did not report or reported a stable situation.
A similarly uniform situation outside Europe was found in the Americas in 1997, where a majority of countries, including Argentina, Brazil, Chile, Colombia and a number of central American countries, as well as Canada and the United States, reported rising levels of cannabis consumption. Only the Mexican authorities reported a decline. In Africa, the Democratic Republic of the Congo, Ghana, Kenya, Malawi, Morocco, Sao Tome and Principe, Sierra Leone and South Africa reported rising levels of cannabis use in 1997, while falling levels were reported only by Nigeria. In Asia, rising consumption of cannabis was reported to UNDCP from central Asia (Uzbekistan), south-east Asia (Indonesia, Malaysia, Myanmar), east Asia (Republic of Korea) and west Asia (Jordan). In the latter two regions, the situation seems to be less uniform. The authorities in the Islamic Republic of Iran, Japan and the Syrian Arab Republic identified some decline. In most of Oceania, generally known for high levels of cannabis consumption, the authorities reported a stabilization in consumption patterns.
Cannabis herb and resin: regional trends
Insofar as seizures reflect trafficking and abuse, an overall upward trend in the 1990s at the global level can also be observed and has been particularly pronounced with regard to reported seizure cases. The total number of cannabis seizure cases reported to UNDCP more than tripled from 1986 to 1996 and in the 1990s clearly exceeded the numbers reported for heroin and cocaine (see figure XIII). In terms of quantities seized, cannabis resin seizures have generally shown an upward trend over the last decade. Between 1986 and 1996, seizures of cannabis resin (hashish) almost doubled (see figure XIV).
Figure XV, showing the development of seizures in different regions, suggests that trafficking in cannabis resin increased across all continents over the last decade. It shows the development of seizures in absolute numbers (quantities seized) and provides information on the relative importance of the various regions (the proportion of cannabis resin seizures in one region as a percentage of global cannabis resin seizures). Figure XV clearly reveals the rising importance of Europe, not only in absolute but also in relative terms. Almost a half of all cannabis resin seizures currently take place in Europe, up from a quarter a decade ago. One third of cannabis resin is seized in Asia and the Pacific. Africa and the Americas account for about one tenth each.
Cannabis herb seizure trends are less clear at the global level. Strong annual fluctuations make it difficult to identify clear patterns. If the analysis is based on a moving three-year average, the data suggest a downward trend in the late 1980s that was reversed in the 1990s (see figure XVI). This is similar to the pattern of seizure and abuse observed in the United States.
The regional breakdown of cannabis herb seizures (figure XVII) shows that after a general global decline in the late 1980s, there was a strong revival in North America and in Europe and a moderate increase in Asia and the Pacific in the 1990s. Such a pattern is also in line with trends emerging from a number of epidemiological studies.
Seizure data suggest that the strongest growth of cannabis herb trafficking occurred in Europe in the 1990s. In absolute terms, however, the overall cannabis herb markets are still dominated by North America, which accounted for half of the world cannabis herb seizures in 1995-1996.
a Preliminary data.
The interpretation of cannabis herb seizure data from South America is more difficult. Though South America accounted for nearly half of global seizures in 1985-1986, this had declined to only 14 per cent a decade later. It is unlikely, however, that the decline is a result of diminished consumption and epidemiological data indeed seems to indicate the opposite.
Surveys in Brazil, the largest country in the region, show that cannabis consumption has been rising strongly over the last decade, in particular in the 1990s (see table 4). In Bolivia, the annual prevalence of marijuana use among 12- to 50-year- olds in cities of more than 30,000 inhabitants quintupled from 0.6 per cent in 1992 to 3.1 per cent in 1998, far more than overall drug abuse, which rose from 3.6 per cent to 6.4 per cent. (29) Rising levels of cannabis abuse have also been reported by other countries in the region.
Table 4. Brazil: development of lifetime prevalence among students
(10- to 19-year-olds)
|All drugs (including solvents)||21.1||26.1||22.8||24.7|
Source: Centro Brasileiro de Informações, Universidade Federal de São Paulo, IV Levantamento sobre o Uso de Drogas entre Estudantes de 1 e 2 Gaus em 10 Capitais Brasileiras (São Paulo, 1997), p. 95.
Diminished South American seizures are probably the result of changing cultivation and interregional trafficking trends. In cases such as Colombia, there have been shifts from the cultivation of cannabis to coca and opium poppy over the last 15 years. (30) As far as trafficking is concerned, more cannabis is being produced in North America, which was the traditional market for South American cannabis herb. Smaller cannabis herb seizures in South America may also reflect changes in law enforcement targets, as the fight against cocaine trafficking has been given a higher priority in recent years.
Reported seizures from Africa are also unlikely to reflect actual trends in trafficking and abuse: the decline of seizures could well be a reflection of the poor and deteriorating status of law enforcement in many countries. A recent study by UNDCP on the situation in a number of African countries (31) suggested that cultivation, availability, trafficking and abuse of cannabis were actually higher than ever before.
Despite the South American and African cases, the general trend in global seizures of cannabis resin (figure XIV) and herb (figure XVI) is still clear: it has been going upwards in the 1990s. It is further supported by trend data, which are derived from Governments' annual assessments of whether consumption of particular drugs is rising, stable or declining. Figure XVIII shows these data for the three principal drugs. The net number of countries reporting increases in cannabis consumption fell in the 1980s, but has been increasing steadily in the 1990s.
All the available indicators--number of seizure cases, quantities seized and epidemiological surveys--as well as qualitative and ethnographic studies, point to the general conclusion that cannabis consumption is increasing across the world.
The size of any market, in economic terms, is determined by the quantity of goods sold (q) and the prices obtained (p). The aggregate turnover--defined as q * p--is used to identify the size of a market. Prices and quantities sold are, however, not independent of one another. The demand curve--reflecting the relationship between prices and quantities sold--is usually downward sloping, that is, low prices tend to be associated with high quantities sold and vice versa. The level of the demand curve itself is a function of income and other factors such as incentives (created through advertising and fashion) and disincentives, which in the case of illicit markets may be caused, for instance, by successful prevention campaigns. The supply curve--showing the relationship of prices and goods offered--is usually upward sloping. The level of the supply curve is a function of the underlying cost structure, determined, for instance, by the degree of direct access to resources and, in the case of illicit markets, of risk. The market will be cleared (that is, supply and demand are in balance) at the intersection of the demand and the supply curves. The present section will show that the various known components of the cannabis market fit fairly well with the basic market model outlined above. Available data reflect the rather decentralized nature of the cannabis market and the many actors involved on both the supply and the demand side, a situation that results in less influence of organized crime over the cannabis market than is the case in the heroin or the cocaine markets. (32) Such market structures and the de facto lower risks of imprisonment in many countries for cannabis-related offences than for those involving heroin or cocaine are factors contributing to the lower prices of cannabis. The lower prices, reflecting easy access and large-scale supply, may also be seen as one of the reasons for the global popularity of cannabis, clearly exceeding that of heroin or cocaine.
The average price of cannabis herb in developed countries, calculated as the (unweighted) average of the minimum and maximum prices identified for each individual country, was $12.5 per gram in 1995-1996, equivalent to 6 per cent of the price of a gram of heroin or 8 per cent of the price of a gram of cocaine (see table 5). The average price of a gram of cannabis herb in developing countries is far lower, on average about $1, less than 2 per cent of the price of a gram of heroin or cocaine. Cannabis herb prices in some of the main producing and trafficking countries of south and south-east Asia (Bangladesh, India, Sri Lanka and Thailand) as well as Africa (Ghana, Malawi, Namibia and South Africa) are particularly low--less than $0.1 per gram. They are also below average (less than $1 per gram) in the countries of the Caribbean and Central America.
Prices of cannabis resin are higher than prices of cannabis herb--by some 20 per cent, on average, in developed countries--though there are exceptions, probably reflecting different levels of THC. In India, Morocco and Pakistan, cannabis resin production is widespread and prices tend to extremely low--less than $1 per gram.
Table 5. Prices of plant-based drugs, 1995-1996
(United States dollars per gram)
|Country or area||Cannabis resin||Cannabis herb||Heroin||Cocaine|
|North Africa and the Middle East||0.8||10.0||5.5||3.0||3.0||3.0||29||165||76||119||149||134|
|Sub-Saharan Africa||5.7||23.0||(13.2) b||0.01||0.3||0.1||20||170||84||60||76||68|
Sources: UNDCP, annual reports questionnaire data; and price data compiled by the World Customs Organization.
a Unweighted average of the mean prices (mean between minimum and maximum prices) in each country.
b Based on two countries only.
In a number of cases it can be shown that low prices of cannabis coincide with high levels of abuse and vice versa. Australia, for instance, has very high levels of cannabis consumption (more than twice as high as in western Europe), while cannabis retail prices are only about half the average for developed countries. By contrast, Japan has low levels of cannabis abuse and prices that are about five times the average for developed countries. Within western Europe, countries such as Spain or the United Kingdom have high levels of cannabis consumption and low prices (about half the western European average), while some of the Nordic countries have rather low levels of abuse and high prices. In Finland and Norway, prices are about twice the western European average; in Sweden they are twice as high as in the United Kingdom or Spain. Both prices and abuse levels in France and Germany are close to the western European averages.
Some exceptions to the inverse correlation of prices and abuse can also be identified. In most cases, however, there are explanations for the apparent exceptions. Data for Portugal, for instance, indicate rather low levels of cannabis abuse, clearly below the western European average, going hand in hand with low levels of cannabis prices (half the western European average), but the per capita income of Portugal is also lower. Its gross domestic product (GDP) per capita (1995) is about half that of the United Kingdom and a third of that of Germany. (33) Thus the effective price of cannabis for the average consumer in Portugal is not low but rather high. Similarly, relatively high cannabis prices for Monaco are not necessarily a reflection of low levels of consumption, but could reflect consumers who are able to pay high prices.
Comparing prices in the United States and western Europe also reveals interesting results. Average marijuana prices, calculated as the average of the mean cannabis herb prices in western Europe ($7.5 per gram; range of $2 to $20) and the United States ($17.5 per gram; range of $3.5 to $31.5), would not fit the theory of a negative correlation between prices and levels of abuse. In this case, however, the price difference is to a large extent due to different categories of cannabis herb found on the United States market. Commercial marijuana is much cheaper than the high-THC variety, which explains the broad price range given by the United States authorities. Once the comparison is restricted to commercial marijuana, official United States publications (34) show a price range of $40 to $450 per ounce (that is, $1.4 to $16 per gram, or $8.7 on average), similar to the price range of $2 to $20 found in western Europe. If the price calculations were based on information obtained from 20 metropolitan areas in the United States, a gram of marijuana would cost on average $7 in the United States, (35) less than in western Europe ($7.5).
Another interesting question concerns the pattern of consumer reactions to changes in cannabis prices over time. Because most studies have been based on rather short time horizons, they have not provided much evidence of any significant price elasticities. For instance, one study showed that the price elasticity for annual prevalence of cannabis use in the United States amounted to just -0.06 (36) between 1988 and 1991, suggesting that cannabis consumption was virtually price-inelastic.
Such results may, however, change once longer periods are investigated. Although consumers hardly react to changes in cannabis prices in the short run, there nonetheless seems to be a strong inverse relationship between prices and abuse in the long run. Data show that, parallel with the increase in cannabis prices, demand for cannabis fell in the United States. Cannabis herb prices in the United States rose from between $250 and $650 per pound (mean: $450 per pound) (37) in 1983 to between $300 and $4,000 per pound (mean: $2,150 per pound) in 1995. (38) Adjusted for inflation, mean cannabis prices tripled between 1983 and 1995, while annual prevalence rates of cannabis use almost halved, from 15.9 per cent in 1982 to 8.6 per cent in 1996. Data for Germany also show a strong inverse relationship between cannabis abuse and cannabis prices. Wholesale prices of cannabis resin declined by almost two thirds between 1983 and 1997, while annual prevalence rose by almost 90 per cent between 1990 and 1995.
Relative importance of cannabis markets
In economic terms, the importance of a drug market is determined by the number of users of the particular drug, the quantities they use and the prices they pay for it. Even though the number of cannabis users is greater, the prices of cannabis are far lower than those of cocaine or heroin. As a result, cannabis markets are, in economic terms, smaller than those of heroin or cocaine.
The United States provides a good illustration. Though 78 per cent of all consumers of illicit drugs use cannabis (1995-1996), expenditure on cannabis is estimated to amount to only 12 per cent of total expenditure on illicit drugs (see table 6).
Table 6. United States of America: expenditure of drug users and drug abuse, 1995
|Expenditure of drug users||Drug abuse|
|Substance||Billions of dollars||Percentage of total||Annual prevalence||Percentage of total|
Sources: Office of National Drug Control Policy, What America's Users Spend on Illegal Drugs, 1988-1995 (Washington, D.C., 1997); and Substance Abuse and Mental Health Services Administration, Preliminary Results from the 1996 National Household Survey on Drug Abuse (Rockville, Maryland, 1997).
Similar patterns are also found in other countries. In the United Kingdom, for instance, cannabis sales were estimated to account for some $1.8 billion, equivalent to about 0.15 per cent of GDP (1996). Overall consumer expenditure on illicit drugs is estimated to amount to approximately $6 billion in 1996, equivalent to 0.5 per cent of GDP. Close to 90 per cent of United Kingdom drug users take cannabis. In economic terms the cannabis market is, however, still less than a third of the total United Kingdom drug market, reflecting low cannabis prices and thus overall low per capita expenditure on cannabis (see table 7). Nonetheless, the United Kingdom cannabis market is larger than the combined markets for amphetamine-type stimulants, cocaine and lysergic acid diethylamide (LSD), but it is smaller than the market for opiates, though there are about 20 times more cannabis than opiate abusers.
Table 7. United Kingdom of Great Britain and Northern Ireland: expenditure of drug users and drug abuse, 1996
|Substance||Expenditure of drug users||Drug abuse a|
|Billions of dollars||Percentage of total||Prevalence rates (percentage)||Estimated total number of drug abusers||Percentage of total|
|Opiates||2.3||36.0||0.4 c||144700 b||4|
|Total||6.3 d||100.0||10.0||3500000||- e|
Source: Chris Groom and Tom Davies, Office of National Statistics, "Developing a Methodology for Measuring Illegal Activity for the UK National Accounts", Economic Trends, No. 536 (July 1998), pp. 33-71.
a United Kingdom Office of National Statistics; estimates based on 1996 British Crime Survey data for recreational users and Department of Health data for problem drug users, using a multiplier of 3.5 for the extrapolation from users entering treatment services.
b Estimate based on 1996 British Crime Survey data and problem drug users identified through Department of Health data.
c Problem drug users only.
d Conservative estimate, assuming United Kingdom seizure rate of 20 per cent.
e Total does not add up to 100 per cent as there are many polydrug abusers.
The Netherlands is one of the few cases among developed countries where the cannabis market ($0.5 billion in 1995) is actually larger than the heroin and cocaine markets ($0.26 billion). Close to 60 per cent of overall drug expenditure goes into cannabis, (39) far more than in other developed countries. This is not surprising, given the fact that the Dutch cannabis market--in contrast to that of most other countries--is de facto decriminalized.
Main consumer markets for cannabis
As noted earlier, retail prices of cannabis herb in developed countries are, on average, 12 times higher than in developing countries. Even if similar levels of abuse were to be assumed, the cannabis markets in developed countries would be substantially larger as a consequence of the higher prices. It could be argued that high prices act as a disincentive and would lead to a reduction in the number of consumers, offsetting the direct impact on the turnover. However, prices can only be compared meaningfully if they are adjusted for differences in purchasing power. As GDP per capita in developed countries is on average some 15 times higher than in developing countries (1995), (40) the effective price of cannabis for consumers in developed countries is, on average,even lower than in developing countries. In other words, apparently higher nominal cannabis prices in developed countries do not act as a deterrent to consumption and also lead to the existence of far larger markets.
It may be possible to go a step further. Though methodologies used in drug surveys differ and results are not always directly comparable, available prevalence data suggest that the spread of cannabis among the general population in the developed countries of North America and western Europe and in Australia is actually more common than in developing countries. The main caveat to this argument is, of course, the well-known one that there are virtually no estimates, at least in quantifiable terms, of the extent of traditional cannabis consumption in developing countries.
Surveys conducted in the mid-1990s showed annual prevalence rates of cannabis use among youth and adults amounting to 13 per cent in Australia, 8.5 per cent in the United States, 7.5 per cent in Canada and about 5.5 per cent in the European Union (see figure XIX). Differences between countries in the European Union are, however, important. Spain and the United Kingdom, as well as Ireland and the Netherlands, show abuse levels above the European Union average, approaching those of Canada or the United States, while Austria, France, Germany and the Nordic countries have cannabis abuse levels below the European Union average.
Higher estimates of cannabis prevalence than in the European Union or the United States were reported to UNDCP, from 1992 to 1995, by Papua New Guinea (23.5 per cent), Brazil (15 per cent), Ghana (12 per cent) and the Federated States of Micronesia (11 per cent) (41) and in 1996 by Sierra Leone (12 per cent). Not all of these estimates were based on systematic nationwide surveys. Results are thus not directly comparable. (42) Most other developing countries and countries of eastern Europe reported annual cannabis prevalence rates of less than 3 per cent. Prevalence rates for some Latin American countries are above that benchmark. Annual prevalence of cannabis use in Bolivia was 3.1 per cent in 1998 (12- to 50-year-olds in cities of more than 30,000 inhabitants). (43) Government estimates for annual prevalence in Guatemala (1997) were equivalent to 3.3 per cent for the general population and 5.2 per cent among those aged 15 and above. (44) Estimates based on systematically conducted national household surveys showed annual prevalence rates of cannabis use of 4.1 per cent in Colombia (for those over 12 years of age; 1996) and 4 per cent in Chile (for 12- to 64-year-olds; 1996), (45) which are also rather high rates for developing countries.
The largest number of cannabis users of any developing country in absolute terms is probably found in India, which has a long history of traditional cannabis use. Extrapolating results of various Indian studies undertaken at the state level over the period from 1989 to 1992, the total number of monthly cannabis abusers may range from 0.4 to 1.7 per cent of the male population above the age of 10, (46) equivalent to between 1.4 million and 5.7 million people. Except perhaps for the situation in Brazil (for which only annual but no monthly prevalence estimates are available), all other estimates of cannabis use in developing countries show current use of clearly less than 1 million people per country. The Indian studies suggest, however, that prevalence of cannabis use is lower than in the United States, where monthly prevalence covered about 10 million people in the mid-1990s, 5 per cent of those aged 12 years and above.
The basic picture does not change if lifetime prevalence is used as an indicator. The highest lifetime prevalence rates of cannabis use among the youth and adult population have again been reported from Australia and the United States (31 per cent), followed among developed countries by Canada (23 per cent) and the European Union (average of 15 per cent) (see table 8). The lower ratio of lifetime to annual prevalence among European Union countries (3:1) compared with the United States (4:1) reflects the fact that widespread cannabis abuse in western Europe is a more recent phenomenon than in North America.
Table 8. Cannabis annual and lifetime prevalence estimates for Australia, the United States of America, Canada and the European Union
|Country or area, (age group), year of survey||Annual prevalence (percentage)||Lifetime prevalence (percentage)|
|Australia (14 and above), 1995||13.0||31.0|
|United States (12 and above), 1995||8.4||31.0|
|Canada (15 and above), 1994||7.4||23.1|
|European Union estimate, 1995 a||5.3||15.0|
Sources: European Monitoring Centre for Drugs and Drug Addiction, Annual Report on the State of the Drugs Problem in the European Union (Lisbon, 1997); Substance Abuse and Mental Health Services Administration, Preliminary Results from the 1996 National Household Survey on Drug Abuse (Washington, D.C., 1997); Commonwealth Department of Health and Family Services, National Drug Strategy Household Survey 1995 (Canberra, 1996); and Canadian Centre for Substance Abuse, Canadian Profile 1997 (Ottawa, 1997).
a Estimate based on data from Belgium, Denmark, Finland, France, Germany, Spain, Sweden and the United Kingdom and estimates for Austria and the Netherlands.
Developing countries usually have lifetime prevalence rates of less than 10 per cent (apart from Jamaica, which has 29 per cent), (47) which makes them significantly lower than those of Australia, North America and the European Union. For South American countries, lifetime prevalence rates of cannabis use in the early 1990s amounted to 8.3 per cent in Peru, 7.3 per cent in Guatemala, 6.1 per cent in Panama, 5.3 per cent in Colombia and 2 per cent in the Dominican Republic. (48)
In more recent years, higher lifetime prevalence rates of marijuana use have been reported by the Inter-American Drug Abuse Control Commission (based on data from 1995-1996), as follows: 16.7 per cent in Chile, 9.2 per cent in Colombia, 3.9 per cent in Costa Rica and 4.1 per cent in Ecuador. Lifetime prevalence rates of around 3 per cent have been reported for the northern states in India and levels of around 7 per cent for the southern states (1989-1991). (49) In Pakistan, the number of abusers of charas (cannabis resin) was estimated to be 950,000 (1993), equivalent to some 2.4 per cent of the population (15 years of age and above). In Africa, some studies undertaken in the early 1990s suggest that the abuse levels may be lower than in the United States or western Europe. One study in South Africa found that 13 per cent of adult males in urban areas reported having tried cannabis, but only 5 per cent in tribal areas; one survey conducted in the state of Lagos found that 5 per cent of the sample had used cannabis. (50)
One way of validating the general patterns emerging from these very heterogenous prevalence estimates is to link them to trafficking patterns. Seizure data are primarily a reflection of trafficking and of law enforcement activities. As trafficking activities are ultimately geared towards selling drugs to consumers, however, they may be used, to a certain extent, to identify underlying consumption patterns, in particular in geographical areas where transit trade does not play an important role. This applies in particular to North America and Europe, as cannabis trafficked within each area is primarily for domestic consumption.
Such analysis of seizure data (see table 9), clearly shows the importance of North America for marijuana (53 per cent of global seizures in 1995-1996) and of Europe for hashish (49 per cent of global seizures in 1995-1996). In terms of overall cannabis seizures (herb and resin), North America is higher than Europe, indirectly confirming the basic patterns reflected in the prevalence data.
Table 9. Geographical distribution of seizures of cannabis compared with other plant-based drugs, 1995-1996 averages (Percentage by region)
|Item||Cannabis herb (marijuana)||Cannabis resin (hashish)||Cannabis herb and resin||Heroin||Cocaine|
|Amount seized (tons)||3029.3||953.5||3982.8||29.7||304.5|
|North America a||53.1||6.8||42.0||5.5||47.8|
|Asia and the Pacific||6.1||32.9||13.0||57.1||0.1|
Source: UNDCP, annual reports questionnaire data.
a Includes Mexico.
The largest seizures of cannabis herb take place in Mexico, the United States and the countries of the European Union. Together with Colombia and South Africa, two important suppliers of cannabis herb, the European Union and the countries of the North American Free Trade Agreement account for 70 per cent of all marijuana seizures worldwide (1995-1996).
The countries of the European Union, as well as Morocco and Pakistan, are the countries most severely affected by trafficking in cannabis resin. Together they account for almost 85 per cent of global seizures of cannabis resin (see table 10). While European Union countries are large consumers of hashish, Morocco and Pakistan are important producer countries. There is a greater geographical concentration in trafficking of hashish than in trafficking of marijuana.
Though there have been some shifts in the positions of individual countries over the last few years, most countries identified in table 10 (based on 1995-1996 averages) have been strongly affected by cannabis consumption, transit trade or production over the last decade. In most instances, countries are faced with two if not all three dimensions of the problem simultaneously. Thus, Canada, Colombia, Ghana, India, Jamaica, Mexico, Morocco, Pakistan, Paraguay, South Africa, Thailand, the United States and the European Union are all affected by domestic production, trafficking and consumption of cannabis.
Table 10. Cannabis seizures, 1995-199
Cannabis herb (marijuana)
Cannabis resin (hashish)
|Country or area||Tons seized||Country or area||Tons seized|
|United States of America||
|United States of America||25.9|
|Iran (Islamic Republic of)||14.5|
Source: UNDCP, annual reports questionnaire data.
There is little reliable information on the extent of cannabis cultivation. Though cannabis is the most widely abused illicit drug, actual knowledge of the extent of production is much more limited than for other narcotic plants. UNDCP has received production estimates from many countries across the globe, reflecting the wide spread of cannabis cultivation. This does not guarantee, however, that reported estimates are reliable or comparable. Available cultivation and production estimates, as will be shown in the present section, have to be interpreted with a large degree of caution. There are several reasons for this:
(a) An apparent reluctance to undertake rigorous investigation of the extent of illicit cannabis cultivation;
(b) Unresolved problems in identifying cannabis cultivation by means of modern technologies such as remote sensing;
(c) Different varieties of cannabis, including the wild species, and the new high-THC plants, (51) which make estimates complicated and subject to potentially large measurement errors;
(d) Definition problems with regard to the area under cultivation;
(e) Definition problems with regard to the cannabis yield. For some countries, the yield is given as the weight of the fresh plant, for some as the weight of the dry plant, while for others it is the weight of the fresh herb or that of the dried herb or of the resin.
Spread of cultivation
Data available to UNDCP make it clear that cultivation of cannabis is far more widespread across the globe than cultivation of any other illicit crop (see figure XX). Between 1990 and 1996, 120 countries reported illicit cannabis cultivation in their territory. This means that cannabis grows in at least two thirds of all countries in the world, compared with fewer than 20 per cent in which opium poppy is cultivated and fewer than 4 per cent in which coca leaf is cultivated.
Origin of cannabis production
One indirect indicator for establishing production patterns is the origin of seizures. For 1996, Interpol identified 79 source countries (that is, countries from which cannabis was exported), spread across Africa (25), Europe (23), Asia and the Near and Middle East (18), as well as the Americas (13), (52) clearly reflecting the global dimension of cannabis cultivation and trafficking.
While cannabis cultivation takes place in at least two thirds of all countries, it is still interesting to note the areas of concentration. An analysis of the origin of seizures significantly reduces the number of significant producers. Interpol statistics for 1996, for instance, showed that out of 772 tons of cannabis resin seizures, (53) the origin could be traced back mainly to Morocco (some 400 tons) and to Afghanistan and Pakistan (260 tons), indicating that the two regions are by far the most important sources of cannabis resin supply.
The analysis of the origin of seizures also suggests that the cannabis herb production is less concentrated than cannabis resin production. Total cannabis herb seizures reported to Interpol amounted to 2,458 tons in 1996. (54) The largest suppliers of cannabis herb (1996) are apparently located in Mexico and the United States (origin of more than 1,500 tons seized), Colombia (more than 500 tons), Cambodia and Thailand (more than 60 tons) and Nigeria (25 tons), as well as in the Democratic Republic of the Congo, Ghana, Jamaica and South Africa.
Trends in cannabis plant seizures
Seizures and eradications of the cannabis plant can also be an indirect indicator of production. Seizures of plants will only occur in countries where production takes place. While the end products (herb and resin) may be trafficked across countries, the plant itself is usually seized at or close to the production site.
The identification of trends based on cannabis plant seizures is, however, complicated by the fact that some Member States report the number of cannabis plants eradicated, while others report the weight of the plants eradicated. Data in figure XXI are shown in terms of the number of cannabis plants seized, as well as in terms of weight, using available information on seizure weight and applying a conversion ratio of 1 pound per plant, as used in publications of the Drug Enforcement Administration and the Office of National Drug Control Policy of the United States. (55) The overall trends reflected in cannabis plant seizures, based on such a transformation ratio, show a decline in the late 1980s, some stabilization in the early 1990s and a strong increase thereafter (see figure XXI). Such trends are basically in line with seizure data for cannabis end products and consumption trends reported by Member States to UNDCP.
As reflected in figure XXI, the overall trends follow very much the seizures in unit terms. Smaller conversion rates for units--reflecting lower yield figures of cannabis reported from Europe (averages of 30 (56) -100 (57) grams per plant)--would result in overall lower levels of cannabis plants seized and less of a decline in the second half of the 1980s, but the trend line for the 1990s would still move upwards.
While the overall trends in seizures of cannabis plants and end products are rather similar, geographical patterns differ. The end products are seized mainly in the consumer markets of North America (about 50 per cent of herb) and Europe (about 50 per cent of resin). The largest seizures of the cannabis plant, by contrast, have been reported from Africa (61 per cent of all seizures from 1990 to 1996) and Asia (28 per cent) (see figure XXII). Insofar as changes in seizures reflect underlying shifts in production, data suggest that cannabis production increased strongly in Africa in the 1990s (from 11 per cent of global seizures between 1984 and 1989 to 61 per cent between 1990 and 1996), reflecting, inter alia, strong eradication efforts by Egypt and South Africa. The importance of the Americas, by contrast, declined, mainly as a result of lower seizures and eradications in South America. While almost half of all seizures of the cannabis plant had taken place in the Americas from 1984 to 1989, that proportion dropped to 8 per cent between 1990 and 1996.
There are indications that domestic cannabis production in several developed countries is increasing. From negligible levels a few decades ago, the United States authorities estimate that 25 per cent of marijuana in the country currently originates in the domestic market (1997). (58) Similarly, tentative estimates for a number of European countries (France, (59) Germany (61) and United Kingdom (63) ) based on seizure data, intelligence reports and investigations among drug consumers suggest that from negligible levels a decade ago, close to a quarter of the cannabis herb market (which in Europe, however, is smaller than the cannabis resin market) could already come from domestic sources. The proportion could be even higher--close to half of all production--in the Netherlands. (65) Cannabis resin, by contrast, is still almost exclusively imported into Europe, mostly from Morocco but also from Afghanistan and Pakistan.
Global production estimates
While knowledge of the various aspects of cannabis cultivation discussed so far appears to be fairly substantial for what is, after all, an illegal activity, knowledge of the extent of cannabis production is close to non-existent. Most available numbers and estimates of the extent of cannabis cultivation and production are highly speculative and in many cases contradict each other. It is nonetheless desirable to try and establish some likely magnitudes.
The United States International Narcotics Control Strategy Reports (INCSR) estimated global cannabis production to be 36,800 tons in 1989, falling to 10,300 tons in 1997. It is unlikely, however, that global production declined by 70 per cent, in particular when viewed in the light of increased production in the United States (71) and all the evidence of increase adduced above. The INCSR estimates reflect actual declines in Latin America, but they do not account for offsetting increases in cannabis production in other parts of the world. The INCSR only give cultivation figures for Latin America. If these were extrapolated to the global level using INCSR production estimates, the total cannabis cultivation figure would amount to about 15,000 hectares. This would mean that the extent of cannabis cultivation was less than a tenth of opium poppy or coca cultivation worldwide--an utterly unlikely proposition.
Another way to estimate the extent of global cultivation has been to base such estimates on seizure statistics and to assume a certain interception rate. UNDCP statistics show seizures of 847 tons of cannabis resin and 3,015 tons of cannabis herb in 1996. It is usually taken for granted that actual production is significantly larger than seizures and a 1:10 ratio is frequently applied. Total production would then be 8,500 tons of cannabis resin and 30,000 tons of cannabis herb. Estimates of cannabis resin of around 8,500 tons have some credibility, as they can also be arrived at by means of another line of reasoning. (72) Similarly, estimates of around 30,000 tons of cannabis herb may not be as arbitrary as they appear to be at first sight.
In the calculations so far, seizures and eradications of the cannabis plant itself are not included. Seizures of cannabis plants in terms of weight totalled 3,095 tons in 1996. In addition, Member States reported seizures and eradications of 247 million cannabis plants in 1996, excluding eradications of wild (low-potency) "ditchweed" in the United States. Transforming the plant eradication figures into one common weight measure--taking an average yield of 1 pound per plant--the weight of eradicated and seized plants would have been equivalent to some 115,000 tons in 1996 (see also figure XXI). Adding cannabis herb and resin seizures, it might be argued that overall global cannabis production could be in the magnitude of half a million tons in the mid-1990s.
Such magnitudes are not prima facie implausible. Half a million tons of cannabis would be equivalent to 0.1 per cent of the global wheat harvest or 7 per cent of the global tobacco harvest. (73) Such estimates would also be in line with estimates provided by Member States to UNDCP in the 1990s. According to those reports, the global area under cultivation and wild growth of cannabis could be around 840,000 hectares. It could range between 670,000 hectares (three times the extent of opium poppy or four times that of coca cultivation) and 1,850,000 hectares (7 times the level of global opium production or 10 times that of global coca cultivation). The broad range is due mainly to changing perceptions of Governments as to the extent of the area under cultivation or wild growth. Using the estimate of 840,000 hectares and taking average yields of 660 kilograms per hectare (1995-1997) used in the INCSR, annual production of cannabis could be more than half a million tons. Taking lower yields for areas of wild cannabis into account (36 kilograms per hectare), production would still amount to some 320,000 tons.
While not implausible, such magnitudes do not appear to be very realistic, because they do not tally with estimates of global consumption. Consumption estimates in the United States, (74) the United Kingdom (75) and France, (76) based on the annual budget spent on cannabis products by consumers, suggest average consumption per consumer (all consumers, including those taking it just once or twice a year) to amount to 75 to 220 grams per year. Assuming an average consumption of 100 grams of cannabis, equivalent to between 100 and 200 joints a year (the estimate for cannabis consumers in the United Kingdom), the above production estimate of 500,000 tons would suggest a global supply of cannabis for some 5 billion people, almost as much as the current world population of 5.8 billion. Taking the slightly lower estimate of 320,000 tons of cannabis and assuming that the average consumer was taking 220 grams per year (high consumption estimates), at least 1.4 billion people, or 24 per cent of the global population, would be consuming cannabis. That is unlikely and would clearly exceed UNDCP prevalence estimates, which, based on the reports of Member States, put the global number of cannabis users in the 1990s at 140 million (2.4 per cent of the total world population).
There are, of course, a number of potential sources of error relating to cultivation estimates, yield estimates and transformation ratios for converting units into weight measures. Evidence of these has been cited throughout the foregoing text and a few additional examples should serve to illustrate the complexity of the estimation process. For instance, official estimates of the extent of cannabis in Kazakhstan, including wild growth, were 132,000 hectares in 1992 (UNDCP, annual reports questionnaire). In 1997, the authorities in Kazakhstan estimated 320,000 hectares. (77) Preliminary results of the 1998 UNDCP survey of illicit crops in Kazakhstan basically confirmed the latter estimates, concluding that the area covered by cannabis was almost 330,000 hectares. However, the survey also revealed that only some 300 hectares (0.1 per cent) were actually under cultivation. (78) The situation in the Russian Federation seems to be similar. Estimates provided by the Russian authorities of the areas covered by cannabis (cultivated and wild-growing) range from 140,000 hectares (79) (actually detected) to more than 1 million hectares. (80) The large extent of wild-growing cannabis is probably the best explanation of why central Asia and the Russian Federation, though constituting the largest areas in the world covered by cannabis, do not play a major role in the supply of cannabis to countries outside the Commonwealth of Independent States.
By contrast, reports of cannabis cultivation in countries of south and south-east Asia focus on cultivation areas that have been detected and eradicated, which means only a few hundred hectares per country. Given the large domestic consumption and exports from these subregions, the actual extent of cultivation is probably considerably larger. Strongly differing estimates have also been received from various countries in Africa. The authorities in Malawi estimate the area under cannabis cultivation to be about 120,000 hectares (1996), making the country Africa's largest producer and one of the world's largest ones, which is very unlikely. Official estimates from South Africa were about 80,000 hectares for 1994-1995, (81) but were lowered thereafter--for methodological reasons--to levels ranging from 1,000 to 2,000 hectares (1996-1997). (82) Official Moroccan estimates put cultivation at 50,000 hectares (1997), up from 11,500 hectares in 1986. (83) Other sources, however, estimate the area under cultivation at 80,000 (84) -90,000 hectares. (85) Differing estimates have likewise been received for countries in Latin America. Estimates reported by Mexico put cultivation (before eradication) at approximately 32,000 hectares (1995); (86) the United States estimated the area under cultivation in Mexico for the same year (before eradication) at less than 19,000 hectares.
Apart from cultivation estimates as such, yield figures have a major impact on the final production estimates. In some cases, potential errors are exacerbated by differences in yield reporting. This may be due to different ways of measuring the output of the cannabis plant based on the weight of the cannabis plant as such, the weight of the foliage, the weight of marijuana (cannabis herb) and the weight of hashish (cannabis resin).
Cannabis yields per hectare can range from a few kilograms (in areas of wild growth) to several thousand kilograms per hectare. For cultivated areas, the average of 660 kilograms per hectare (1995-1997) used in the United States INCSR is a reasonable measure. By contrast, the yield figures from South Africa, reporting 2,120 kg/ha, and Chile, reporting 8,000 kg/ha (1995), seem to be based on measuring the weight of the plant as a whole and are thus not directly comparable to the yields of the cannabis end products. But even for one and the same country, available yield figures may differ substantially. In Kazakhstan, for example, potential yields are close to 450 kilograms of cannabis herb per hectare, while the actually harvested amount could be as low as 4.8 kg/ha. (87) Morocco offers another example: INCSR estimates for Morocco used a yield of 4.5 kilograms of cannabis resin per hectare in the late 1980s, (88) while the United States Drug Enforcement Administration (DEA) estimates in the mid-1990s assumed that cannabis resin yields in Morocco ranged from 20 to 800 kg/ha. (89) The German Bundeskriminalamt reports that cannabis resin yields in Morocco usually range from 10 to 45 kg/ha.
If seizures are used as an indirect indicator for the extent of production, another potential error can occur when transforming the eradicated or seized plants into actual weights of the end product. Yields per plant at the global level are probably lower than transformation ratios of 1 pound, that is, 453.59 grams per plant, applied in the United States. In the United Kingdom, for instance, an average transformation ratio of 100 grams of cannabis per plant is frequently applied. (90) Thus, an estimate based upon the average amount of cannabis plants and products seized between 1990 and 1996 and a transformation rate of 100 grams per plant would yield a total seizure figure of 11,500 tons per year, instead of the 115,000 tons mentioned above.
Since not all production is seized, the actual production figure has to be larger than 11,500 tons. Analysis of the extent of demand (91) suggests that consumption, and therefore production, is unlikely to be more than 30,000 tons. The latter figure would be in line with the UNDCP estimate of about 140 million people consuming cannabis (30,000 tons divided by 220 grams per user per year).
The present paper has tried to draw together the information and data that can be aggregated to form a global picture of the production, trafficking and abuse of cannabis. It shows that cannabis is by far the most widely cultivated, trafficked and abused illicit drug. Half of all seizure cases worldwide are cannabis seizures. The geographical spread of those seizures is also global, covering practically every country in the world. Two and a half per cent of the world's population, about 140 million people, consume cannabis (annual prevalence), compared with 0.3 per cent consuming cocaine and 0.2 per cent consuming opiates.
In terms of trends and patterns, the abuse of cannabis has grown more rapidly than the abuse of cocaine and opiates in the present decade. Over the last four decades, the most rapid growth in cannabis abuse has been seen in the developed countries of North America, Australia and western Europe. This phenomenon has been linked to cannabis being more closely integrated into youth culture, with an age of initiation that is usually lower than for other drugs. Regional trends are different for cannabis resin and herb. For resin, about a quarter of all seizures worldwide took place in Europe in the mid-1980s; the proportion had increased to half of all seizures by the mid-1990s. For cannabis herb, half of all seizures in the mid-1990s were in North America, up from a quarter in 1985-1996. In the mid-1980s, half of all cannabis herb seizures took place in South America. That proportion declined to 14 per cent in 1995-1996. The decline did not represent lower consumption in South America, which actually increased over the same period; it only represented less export of cannabis herb.
The analysis of cannabis markets has shown that low prices coincide with high levels of abuse, and vice versa. This is particularly true if prices being compared are adjusted to reflect local purchasing power. With regard to elasticity, cannabis appears to be price-inelastic in the short term. In the longer term, however, there is some evidence that demand for cannabis may be fairly elastic. Though the number of cannabis consumers is greater than consumers of opiates and cocaine, the lower prices of cannabis mean that, in economic terms, the cannabis market is much smaller than the opiate or cocaine market. Thus, in the United States, 78 per cent of all drug consumers take cannabis, yet the market for it represents only 12 per cent of the total market for illicit drugs. In the United Kingdom, 89 per cent of drug consumers take cannabis, but the cannabis market represents only 29 per cent of the total illicit drug market. The Netherlands is an exception to this pattern: cannabis represents about 60 per cent of the total illicit drug market, clearly a result of the de facto decriminalization of cannabis. In geographical terms, the main consumer markets for cannabis are Australia, North America and western Europe, with annual prevalence rates of 13 per cent, 8 per cent and 5 per cent respectively (1995 data; population 15 years and above). In most developing countries, annual prevalence rates are usually well below five per cent.
Given the present status of knowledge, the extent of cannabis cultivation and production is very difficult to measure and estimate. The wide geographical spread of cannabis cultivation can be documented: between 1990 and 1996, 120 countries reported cannabis cultivation, compared with 6 countries reporting coca and 35 reporting opium poppy cultivation. The extent of this cultivation, and consequently total cannabis production, is much more difficult to assess. Estimates vary by a factor of 30, from the United States INCSR figure of 10,000 tons to the figure of 300,000 tons derived by aggregating country data received by UNDCP in the annual reports questionnaires. The only way to find a reasonable order of magnitude within these widely differing estimates is to link production to consumption estimates. Such an attempt at triangulation suggests that 30,000 tons would be a reasonable estimate of annual cannabis production worldwide.
3. Abuse of hashish was widespread in Egypt. Following a report by an Egyptian doctor in 1868 on the effects of and accidents caused by hashish, cultivation of cannabis was forbidden in 1884. Yet the authorities reported that 30-60 per cent of all insanity cases remained related to hashish abuse. The Egyptian argument for international control was based upon the assertion that the domestic ban on cultivation was offset by international trafficking of cannabis from Lebanon and India, often organized by European traffickers (League of Nations, Records of the Second Opium Conference, Geneva, November 17th, 1924-February 19th, 1925, vol. I, Plenary Meetings; text of the debates, "Hashish: Proposal of the Egyptian Delegation that hashish should be included in the list of narcotics with which the Conference has to deal", Geneva, 1925, p. 13; Observatoire géopolitique des drogues, Atlas mondial des drogues (Paris, Presses universitaires de France, 1996), pp. 54 and 55).
4. United States of America, Department of Health and Human Services, National Institute on Drug Abuse, National Survey Results on Drug Use from the Monitoring the Future Study, 1975-1997 (Washington, D.C., 1997).
12. Australia, National Drug Strategy: Household Survey Report 1995 ..., p. 32, and Centre for Behavioural Research in Cancer, Australian Secondary Students' Use of Over-the-Counter and Illicit Substances in 1996 (November, 1998), p.16.
18. R. Hartnoll, "A Multi-City Network as an Epidemiological Information System", in European University Institute, Florence, Policies and Strategies to Combat Drugs in Europe: The Treaty on European Union: Framework for a New European Strategy to Combat Drugs? Georges Estievenart, ed. (Dordrecht, Martins Nijhoff, 1995), pp. 261-175.
21. F. Measham, R. Newcombe and H. Parker, "The normalization of recreational drug use amongst young people in North-West England", British Journal of Sociology, vol. 45, No. 2 (June 1994), pp. 287-311.
22. J. Mott and C. Mirrlees-Black, Self-reported Drug Misuse in England and Wales: Findings from the 1992 British Crime Survey, Research and Planning Unit Paper 89 (London, Home Office Research and Planning Unit, 1995), p. 42.
32. One indirect way to measure the importance of large-scale organized crime in trafficking is to calculate the ratio of significant seizure cases (a proxy for the importance of large-scale trafficking activities that are usually carried out by organized crime) and the overall number of seizure cases (a proxy for the overall market size). Based on significant seizures reported by countries to the World Customs Organization (1997) and the total number of seizure cases reported by Member States to UNDCP, the ratio of significant seizure cases to all seizure cases for cocaine-related products is about 1:10, the ratio for opiates (including heroin) is about 1:140, while the ratio for cannabis is more than 1:300. This suggests a strong involvement of organized crime in cocaine trafficking and a comparatively low involvement of organized crime in cannabis trafficking. The latter is characterized more by the involvement of a large number of individuals or small trafficking groups.
35. National Institute on Drug Abuse, Epidemiologic Trends in Drug Abuse (Washington, D.C., June 1997), p. 55; and Office of National Drug Control Policy, Pulse Check: National Trends in Drug Abuse (Washington, D.C., Summer 1997), pp. 29-33.
39. R. Van der Werf, "Registration of illegal production in the national accounts of the Netherlands", joint OECD/ECE/Eurostat Meeting on National Accounts, Paris, 3-6 June 1997 (Voorburg, Statistics Netherlands, 1997), quoted in Trimbos Institute (Reitox Focal Point), National Report 1997: the Netherlands, draft version, p. 61.
42. They have also been subject to major changes in subsequent periods. Annual prevalence estimates reported from Brazil (1995-1997) ranged, for instance, from 2 to 21 per cent (recalculated for 15-year-olds and above) and for Ghana from 0.1 to 12 per cent (recalculated for 15-year-olds and above), reflecting different extrapolation techniques and different concepts of defining drug abusers (UNDCP, annual reports questionnaire data).
47. J. M. Jutkowitz, "Drug prevalence in Latin America and Caribbean countries: a cross national analysis", Drug Education, Prevention and Policy, cited in World Health Organization, Cannabis: A Health Perspective and Research Agenda (Geneva, 1997), p. 10.
51. See detailed discussion of these issues in articles by K. Szendrei, "Cannabis as an illicit crop: recent developments in cultivation and product quality" and C. Bone and S. J. Waldron, "New trends in illicit cannabis cultivation in the United Kingdom of Great Britain and Northern Ireland",
in the present issue of the Bulletin on Narcotics.
52.Interpol, Worldwide Traffic in Cannabis (Lyon, August 1997), p. 2.
53.UNDCP statistics show slightly higher numbers: seizures of 847 tons of cannabis resin and of 3,015 tons of cannabis herb in 1996. The difference can be explained by the larger number of countries contained in the UNDCP data set.
55.United States of America and Mexico, México y Estados Unidos ante el Problema de las Drogas ..., p. 75.
56.Germany, Bundeskriminalamt, Rauschgiftjahresbericht 1997 (Wiesbaden, 1998).
57.Chris Groom and Tom Davies, Office of National Statistics, "Developing a methodology for measuring illegal activity for the UK national accounts", Economic Trends, No. 536 (July 1998), pp. 33-71.
58.UNDCP, annual reports questionnaire data.
59.One survey among drug users in France (1996) indicated that about 25 per cent of cannabis herb might originate from domestic sources. The French cannabis herb market, however, is much smaller than the cannabis resin market. Hydroponic cultivation is known, but it is still the exception. Ninety-seven per cent of French cannabis cultivators produce for their own consumption, only 5 per cent for mainly commercial purposes. Only a few cultivators plant more than 100 plants a year.
60.R. Ingold M. Toussirt, Le Cannabis en France, Paris 1998, p. 38.
61.Domestic cannabis production in Germany is gaining in importance, though it is apparently less developed than in the Netherlands or the United Kingdom. Approximately 67,000 plants were seized in 1997, a quarter more than in 1996. There were more than 2,800 cases of illegal cultivation of cannabis detected in Germany (1997), of which 29 cases (1 per cent) concerned cultivation of cannabis in greenhouses. Compared with the Netherlands or the United States, the number of indoor cannabis operations, however, is still small. The United States authorities detected 2,900 indoor operations in 1997, 100 times more than the German authorities. The average cannabis yield reported by the authorities in Germany is around 28 grams per plant, less than in the United Kingdom (100 grams) or in the United States (1 pound), though similar to the yields reported from the Netherlands (22 grams). Applying the average yield figures to the number of plants seized in Germany in 1997, the data suggest that plants eradicated (reflecting domestic production) would have yielded some 1.9 tons of cannabis herb. Total cannabis seizures amounted to 4.2 tons of herb and 7.3 tons of resin in 1997.
62.UNDCP, Annual Reports Questionnaire Data.
63.United Kingdom law enforcement authorities reported to the meeting of Heads of National Drug Law Enforcement Agencies, Europe, held in Vienna in September 1998, that cultivation of cannabis using hydroponics and artificial lighting had been increasing in recent years. The number of cannabis plants seized in the United Kingdom, as reported to UNDCP, doubled over the period from 1994 to 1996. According to Home Office data, the United Kingdom police seized some 115,600 plants in 1996, which--applying an average weight of 100 grams per plant (the conversion rate usually applied in the United Kingdom)--would be equivalent to about 11.6 tons of cannabis herb. In addition to 46.1 tons of cannabis resin, seizures of herbal cannabis by the customs authorities (reflecting imports) amounted to 30.5 tons in 1996.
64. C. Groom and T. Davies (Office for National Sttistics), "Developing A Methodology for Measuring Illegal Activity for the UK National Accounts", Economic Trends, July 1998, p. 33-71.. " " -
65.Most hydroponic cultivation of cannabis in Europe takes place in the Netherlands.
66. United Kingdom, H.M. Customs and Excise, An Assessment of Global trends in Drug Trafficking and Production, January 1998, p. 88. -
67.Bundeskriminalamt, Rauschgiftjahresbericht 1997, Wiesbaden 1998.
68.UNDCP, Annual Reports Questionnaire Data.
69.Bundeskriminalamt, Rauschgiftjahresbericht 1997, Wiesbaden 1998.
70. Bundeskriminalt, Rauschgiftjahresbericht 1997, Wiesbaden 1998. " "
71.On this point, see R. Clayton, Marijuana in the "Third World": Appalachia, USA, Studies on the Impact of the Illegal Drug Trade, vol. 5 (Boulder and London, UNU/UNRISD, Lynne Rienner Publishers, 1995).
72.Europe accounts for about a half of all cannabis resin seizures in the world. Based on the origin of seizures, Interpol estimates that about 70 per cent of cannabis resin seized in Europe could originate in Morocco. Production in Morocco is considered by some sources to range from more than 2,000 tons (United States, INCSR) up to levels about twice as high (German Bundeskriminalamt). Taking the estimate of 2,000 tons of illegal Moroccan cannabis resin exports to Europe, total cannabis resin imports into Europe would amount to about 2,900 tons; based on the higher production estimate of Morocco, total imports into Europe could amount to up to 5,700 tons. If Europe accounts for half the global cannabis resin market, global cannabis resin production should be between 5,800 and 11,400 tons. An estimate of 8,500 tons would thus fall well within that range.
73.Food and Agriculture Organization of the United Nations, quoted in Statistiches Jahrbuch für die Republik österreich 1997.
74.BOTEC Analysis Corporation, Marijuana Situation Assessment, study prepared for the Office of National Drug Control Policy (Cambridge, Massachusetts, 1994), p. 17.
75.Chris Groom and Tom Davies, "Developing a methodology ...", p. 42.
76.Rodolphe Ingold and Mohamed Toussirt, Le Cannabis en France ..., p. 38.
77.K. Szendrei, "Kazakhstan, a unique environment for illicit crop eradications: An analysis for consideration by UNDCP" (Vienna, 1997), unpublished report; "Central Asian republics: regional report situation" (Vienna, UNDCP, 1994), unpublished report.
78.Office for Drug Control and Crime Prevention, Central Asia crop survey (Vienna, 1999).
79.Russian Federation, "The destruction of wild narcotic plants and illicit cultivation: problems and prospects" (Conference paper No. 7) (Moscow, 1997), p. 1, in UNDCP, Proceedings of the International Conference on Drug Control Cooperation with the Russian Federation, Moscow, 16-17 April 1997.
80.Russian Federation, "Illicit drug trafficking and counter measures in the Russian Federation" (Conference paper No. 4) (Moscow, 1997), p. 5, in UNDCP, Proceedings of the International Conference on Drug Control Cooperation ... .
81.South African Police Service, National Crime Investigation Service, The Illicit Drug Trade as a National and International Threat (Pretoria, November 1995), p. 6.
82.UNDCP, annual reports questionnaire data.
84.United States of America, Department of State, Bureau for International Narcotics and Law Enforcement Affairs, International Narcotics Control Strategy Report, 1997 (Washington, D.C., 1997).
85.Information provided to UNDCP by the German Bundeskriminalamt.
86.UNDCP, annual reports questionnaire data.
87.Office for Drug Control and Crime Prevention, Central Asia crop survey ... .
88.United States of America, Department of State, United States International Narcotics Control Strategy Report (Washington, D.C., 1991).
89.Drug Enforcement Administration (DEA) estimates of the mid-1990s on the situation in Morocco suggested a yield of 2-8 tons of plant material per hectare and 10-100 kilograms of cannabis resin per ton of plant material. This meant yields ranging from a minimum of 20 kilograms up to 800 kilograms per hectare (see DEA, Drugs in Africa (Washington, D.C., 1997)).
90.Chris Groom and Tom Davies, "Developing a methodology ...", p. 46.
91.For an interesting example of such a process of triangulation, see Advisory Committee on Illicit Drugs (Australia), Cannabis and the Law in Queensland: A Discussion Paper (Brisbane, July 1993).