Routine activities and drug trafficking: the case of the Netherlands
Graham Farrell *
Center for CrimePrevention Studies, Room 1314, School of Criminal Justice, Rutgers State University of New Jersey', 15 Washington Street, Newark, NJ 07102, USA
Received I April 1997~ received in revised form I August 1997 accepted I August 1997
Like the air we breathe, sometimes things are such an accepted and mundane part of our experience that we fail to recognise their fundamental importance. This is a most certainly the case where the relationship between licit trade and illicit drug trafficking is concerned. When variations in illicit drug trafficking and consumption are analyzed, it is possible that we often look too hard for drug policy-related scapegoats. when at least a significant part of the variation is due to other factors. This paper attempts to build an empirical case study of the Netherlands as a specific illustration. While opinion regarding the validity and potential implications of the case study will almost certainly remain divided, it is hoped that it will stimulate recognition of the potential for further contributions to the analysis of international drug policy through the utilization or the routine activities approach. 1998 Elsevier Science B.V. All rights reserved.
1. The case of the Netherlands
The volume of shipping passing through Dutch ports is both greater and more concentrated than that anywhere else in Europe, Relative to its size, the Netherlands has larger trucking fleets than those of neighbouring countries. Since these are the main modus operandi for international drug trafficking then, in line with routine activity theory (Cohen and Felson 1979; Felson 1994), it could be expected that illicit drug trafficking into the Netherlands would be high, relative to neighbouring countries. Previous research has found drug prices in the Netherlands to be lower than elsewhere. This may, at least in part, be explained by variations in the flow of licit trade and commercial activities. However, such an innocuous statement may have broad-ranging implications. It is possible that variations in European drug trafficking routes and markets are falsely attributed to drug policies when they are primarily driven by variations in licit routine activities.
One simple observation often made about the illicit drug situation in the Netherlands is that it has lower drug prices and greater availability of illicit drugs than its European neighbours; consequently, it serves as a distribution hub in different ways. The deduction often made--which forms the policy issue discussed later in this paper is that this is due to differences between Dutch drug policy and that of its neighbours. What this paper suggests is that the deduction falsely attributes too great a role to policy relative to competing explanations. This is riot to make a naive statement that policy makes no difference, but rather, to observe that there are additional important factors that warrant consideration and that these are often overlooked.
2. Origins of the present paper: variation in European drug prices
To a large extent, the present paper was driven by previous findings related to variations and trends in illicit drug trafficking and prices in Europe. That research suggested that both the wholesale and retail prices of' heroin and cocaine in the Netherlands were lower than in other European countries, noting that
"the Netherlands seems to be the cheapest country in Europe in which to make retail purchases of' heroin. and possibly for cocaine, though prices are only available for a few years. In the Netherlands, prices may be driven more by the generally more lenient Dutch drug law enforcement policy (Leuw 1991), so that more suppliers are competing for the same consumers, thereby bringing prices down. The problem area here is that the policy might be expected to impact differentially upon wholesale and retail prices since, as Leuw (1991) describes, the law enforcement emphasis is at the wholesale rather than the retail level. We do not have an explanation for this anomaly." (Farrell et al., 1996)
It was also concluded that
"Several areas for future research have been highlighted,... (including) further specific investigation of why Dutch wholesale as well as retail prices appear to be lower".
It was not the fact that Dutch retail prices were lower that was really of interest. That might be expected since, as discussed later, law enforcement at the retail level is 'generally more lenient' than that in other European countries. It was the lack of explanation for the finding that wholesale prices were also lower that left the authors stumped. This anomaly does not seem to be explained by Dutch law enforcement policy since that focuses upon the higher level distribution end of the illicit drug trade. Further investigation seemed warranted and is the basis for the present paper. The first step taken here was to analyze the data further to test whether the suggested differences in prices for both drugs at both levels were statistically significant. In short, they were. This was tested using a Marin-Whitney U-test, the results of which are found at tile end of this paper. Both cocaine and heroin proved significantly cheaper in the Netherlands compared to the other European countries of the study. This applied at both the retail and wholesale levels. While the data was relatively fragmentary, the strength of' the findings and their consistency across drug types and years, using a data set collected annually over a period of 14 years, suggests that the pattern is sufficiently robust to warrant further investigation.
Prior to looking at why the price differences exist, the significance of illicit drug prices may need some brief explanation. Prices are important because basic economics suggests that lower prices are due to increased availability and thereby serve to increase demand. In relation to illicit drugs, law enforcement efforts aim to reduce availability (supply), thereby increasing prices and reducing demand (less people purchase the same good at a higher price).
While the empirical analysis of prices was one factor, the second trigger that initiated this paper was the response of a colleague to the previous findings. A US colleague, Marcus Felson, inquired in a teasing rhetorical way whether the
Netherlands had any large ports, knowing full well that it did. Much of Felson's work for over two decades has related to the impact of routine activities upon crime trends and the follow-up investigation into this insight led to the present paper,
In order to provide a backdrop for the ensuing analysis and discussion, the next section outlines key aspects of drug policy in the Netherlands.
3. Dutch drug policy summarized
Dutch policy considers drug use it public health issue and distinguishes between 'hard' and 'soft' drugs based on the 1976 Opium Act. This i~ a distinction that is not formally made in the UN conventions but is an operational concept in many countries and one understood and accepted by all but the most innocent or extreme.
The key aims of Dutch policy are to concentrate enforcement efforts upon high-level traffickers, to separate the 'hard' and 'soft' drug markets at the retail level and to minimize the risks to health and life of illicit drug users (Leuw 1991; Leuw and Haen Marshall, 1994; Horstink-Von Meyenfeldt, 1996). In practice, law enforcement resources are directed towards high level traffickers rather than low-level dealers ~Wd users, thereby allocating scarce resources to the problem perceived as more important. This has the benefit of reducing the strain upon the criminal justice system of possession- related arrests. As such, i-he policy aims to eliminate any possibility of a 'stepping-stone' or 'gateway' effect (soft drug users moving to hard drugs) by separating the cannabis market from those for heroin, cocaine and the amphetamine-type stimulants. This is most obviously manifest in the form of the registered coffee shops where small amounts of cannabis can be purchased for personal use. Purchasers avoid the risks of mixing with the less desirable purveyors of other illicit products.
Critics represent Dutch policy as being lax on users and encouraging cannabis use. They imply, by omission of specifics, a general encouragement of all drug use including heroin and cocaine. The 1993 INCB report states that, in the opinion of' the Board, the cannabis policy transgressed the parameters of the international conventions (INCB, 1993). In response, the Dutch prime minister stated that the policy had been successful and would be maintained. Polak and Lap (1994), their critique of the report of the board in the International Journal of Drug Policy suggested that "Greater value is attached to the verbatim text of the International Conventions than to the results that have been attained in the various nations as a consequence of different interpretations of the text of the treaty."
Subsequent to the Board's report however, the coffee shop policy has been tightened up and the number of coffee shops reduced. Horstink-Von Meyenfeldt (1996) notes that "the Netherlands needs to be critical about the effects and results of the soft drug policy. It seems that the percentage of hard and soft drugs used in relation to the total Dutch population is more or less the same as the statistics in Germany and Switzerland, although neither country has a coffee shop policy"
In 1992, cannabis use among Dutch youth (age 12 - 18) was that I 3.6'V,, had 'ever used' it and 6.5% had used it in the past month (Zimmer and Morgan 1995). This compares to American youth (age 12 -17) of whom 11.7% had ever used cannabis, and 4.9% in the last month, and American high school seniors of whom 35.3% had ever used cannabis and 15.5% had used cannabis in the last month (Zimmer and Morgan 1995). In Amsterdam in 1994, 6.4% of all adults used cannabis in the past month (Cohen and Sas, 1996), compared to a figure of 4.7% of the whole US population in 1995 and 8.2'V~) of youths age 12 17, representing a doubling in rates for youth since 1992 (SAMHSA, 1996). Note however that the latter figures are not strictly comparable, since levels of use in Amsterdam are almost certainly higher than the Dutch average. In short, levels of cannabis use in the Netherlands and the USA seem to be characterized by their similarities rather than their differences, despite widely differing policies.
It may be partly the logic underpinning the policies and partly the findings of empirical studies, that most European countries (i.e. Germany, UK. Spain and Portugal) and countries such as Canada and Australia, have begun to adopt similarly pragmatic policies which come under the general rubric of harm reduction. The notable exceptions are probably France and Sweden, but for the most part European countries operate within the general framework of prohibition according to the international conventions, while emphasizing elements of' drug control which impact upon the risks to the life and health of illicit drug users.
4. Does policy influence prices?
Of course drug policy can influence illicit drug prices, but why, specifically, might Dutch drug policy make heroin and cocaine cheaper on average at both the wholesale and retail levels than in the rest of Europe? Less aggressive and punitive law enforcement activity in retail markets might be expected to result in a fall in retail prices since there is less risk involved in dealing (the classic statement on how search time and risk affect purchasing and prices is made by Moore (1973)). However, focusing law enforcement efforts upon the wholesale distributors might increase wholesale prices (reduced supply causes prices to increase). The coffee shop policy might be expected to result in lower cannabis prices in a competitive market. In addition, if the coffee shop policy effectively separated hard and soft drug markets, then it might reduce demand for hard drugs, causing prices to fall as a result. It is difficult to see how the coffee shop policy would influence the price of heroin and cocaine, at the wholesale level.
While this is at best a cursory discussion and does not delve into price elasticities, it gives some crude indicators of what might be expected. An explanation for lower heroin and cocaine retail prices might be derived from policy. However the same is not true for wholesale prices, which, if anything, might be expected to be higher than elsewhere. This contradicts the empirical evidence of lower Dutch wholesale prices for cocaine and heroin. The lack of explanation of the level of wholesale prices might also shed some doubt upon the possible explanations that were mooted to explain lower heroin and cocaine retail prices.
This contradiction and doubt are important, particularly since a competing or supplementary hypothesis is offered below which is supported by a not unreasonable rationale combined with fairly clear-cut empirical evidence.
5. A routine activities approach
This section outlines the theory on which this paper is based. The most compelling extant explanation of trends in crime rates for many crimes (burglary, car crime etc.) is Cohen and Felson's routine activity theory (Cohen and Felson, 1979; Felson, 1994). They explained the long-standing paradox of why crime rates in developed countries had risen in the second half of the 20th century during the time that there had been a general amelioration ofthe so-called socio-economic 'root causes' of crime. The essence of the explanation was that as living standards had improved, more consumer goods became available to be stolen and lifestyles and technologies changed in ways that facilitated criminal victimization.
Routine economic and social activities may play an important role in explaining a lot of variations in patterns and trends of illicit drug trafficking and use. This is neither a particularly new nor complex notion. Most people have a hunch that vast increases in global trade, the political integration of the Eastern block and trade liberalization and trade agreements that facilitate cross-border commerce, have helped illicit drug traffickers as well as other forms of organized crime. They provide greater camouflage, while increased use of precursors and demand for drugs for medical purposes provide an increased trade volume and number of channels from which these can be illicitly diverted. Likewise, the technological advances in computerization and the globalization and expansion of computerized financial transactions, help traffickers launder illicit proceeds in the global financial system. These may explain some changes in the illicit drug industry over time, but the specifics of such arguments are seldom teased out. This is particularly true in terms of cross-sectional rather than longitudinal analysis. There may be many implications for international drug policy that remain to be uncovered through following these hunches through to their logical conclusions. While the present paper is a specific case study, a more wide-ranging exposition of these themes is given in a complementary paper elsewhere (Keh and Farrell, 1998).
6. Trade and transportation flows
The present application of' the routine activity approach to drug trafficking looks at flows of trade into Europe. The Netherlands is the home of Europe's largest port - Rotterdam. The volume of trade moving annually through Rotterdam is far greater than that in any other European port. Fig. I shows trade volumes moving through European seaports (there may be some double counting here since it both imports and exports goods). However, the Netherlands is also the home of Amsterdam, giving it two of the largest ports in Europe. When combined it is evident that the Netherlands is the nation that has by far the largest international trade movements by sea. Fig. 2 shows the volume of maritime trade by country.' Figs. I and 2 deliberately use different and complementary data sources.
Other things being equal, it would be expected that the Netherlands would experience a greater volume of illicit drug trafficking by sea cargo than its neighbouring countries. It is simply a side-effect of economic and commercial trade patterns. It may also be that factors such as the intensity or concentration of larger trade flows necessitate more speedy processing, which lends further advantages to illicit trafficking in high volume trade areas. From this expectation it would be expected that local markets would have greater availability and lower prices, since they are the closest and most accessible. It might also be expected that the prevalence of illicit drug consumption would be commensurately higher.
There was a statistical break in the data for France. In previous editions of' the UN yearbook, France had only a relatively minor volume of recorded shipping. In more recent volumes, the figure is far greater. This statistical break in the data would suggest either an error or a change in the way the French data is reported or recorded.
A perfect correlation between licit and illicit trade would not be expected. This is a simplified explanation to which there Lire many exceptions: variation in licit trade is just one, albeit important, facilitator of illicit drug trafficking. Others. such as language and family ties to drug producing areas may also contribute, but the present analysis concentrates upon variations~ in trade. Similarly, the present analysis seems particularly appropriate for cocaine and heroin which are imported to Europe, whereas amphetamines are allegedly produced both within the Netherlands and in neighbouring European countries.
As well as trafficking heroin and cocaine into the Netherlands, distributors need to get them to other European markets. Once in the EC, overland shipment is easier and cheaper than shipping. Linked to the volume of shipping, it is no surprise to find that the Netherlands has one of the largest trucking trades in Europe, after Germany and France, which are both larger and more populous countries (Fig. 3). Moreover, when seen in relation to its size, the Netherlands ranks only 8th in Europe by population (Fig. 4) and 17th in Europe by area (Fig. 5)2 the Netherlands has by far the most concentrated international maritime trade and commercial trucking transportation in Europe. As well as the small land area suggesting a 'funneling effect' upon trafficking, population density might be a crude indicator of' potential illicit demand in a country. It is easier to distribute illicit drugs in densely populated areas where there are more potential customers passing by and others providing cover. Population densities for countries where the data was available are shown in Table I with an index to show the density in other countries relative to the Netherlands. The results are probably not as well known as might be expected. The author was surprised to find that the Netherlands is by far the most densely populated country in Europe, around double the density of the UK and Germany, four times that of France and 20--30 times that of most of Scandinavia.
7. Expected illicit market size
Using population density as a crude measure of' potential access to and availability of illicit demand and licit maritime trade volume as a crude measure of opportunity for illicit supply, a third measure, which is here termed expected illicit market size, can be calculated. Countries with few opportunities for illicit supply and unattractive consumer markets would be expected to have little illicit trafficking and consumption-small markets. Countries with lots of' opportunity for illicit supply and a dense consumer market providing lots of potential demand and cover, would be expected, other things being equal, to have lots of illicit drug trafficking. Other countries will have a variable mixture, perhaps high potential supply and low potential demand or vice versa, and both would be expected to have medium-intensity illicit drug markets. Table 2 shows expected levels of illicit trafficking calculated as the multiple of the population density index relative to the Netherlands (Table 1) and a licit trade supply index relative to the Netherlands (derived from but not shown in Fig. 2). Data is shown for the nine countries for which both data points were available. In fact, expected market size is not evenly spread between countries, with some with high, some with medium and some with low expected levels. It is highly skewed towards the bottom end. This is because of an interaction effect--two big scores multiplied together make a disproportionately big score. Using these crude indicators of' potential illicit supply and demand it would be expected that levels of illicit drug use in the Netherlands might be over three times those of Germany, five times those of Italy, 20 times those of Spain and many times more than those of other countries. The fact that they are not is a point returned to later.
A word of caution needs to be exercised here. The interaction effect between supply and demand represented in the expected market size indicator is only a possibility that is raised here. It Is tentative, the data is incomplete, and the constituent measures crude. Perhaps most telling is the fact that even without considering this indicator, the absolute differences in licit trade volumes are sufficient to be of interest in and of themselves. Even without a possible multiplicative effect of supply and demand, the ensuing discussion is warranted.
The present paper arose out of previously published work on heroin and cocaine prices, combined with an established theoretical perspective.
At a general level, even the brief analysis contained here suggests there may be a large scope for the analysis of routine activities to inform international drug policy. The more specific but tentative conclusion is that, if enforcement responses were equally good in all European countries, then the seaborne flow of illicit drugs into the Netherlands would be expected to be great relative to that of other countries. There may be important policy implications related to the case study.
The effect of greater supply of illicit heroin and cocaine to the Netherlands would be lower wholesale and retail prices, in turn leading to greater demand. This is for the simple reason that both the volume and concentration of maritime and land trade in Europe is at its greatest in the Netherlands and these are important modus operandi for traffickers (Farrell et al., 1996). The significance of air transportation might be included in a more sophisticated model than is developed here: it would probably modify downwards the routine flows into the Netherlands compared to elsewhere. However, on a crude score based on shipping trade, the Netherlands beats other countries hands down for the cover this gives to trafficking. Even with incomplete data, other European countries seemed characterized by the similarity of the low camouflage scores rather than their differences. It is known that bulk shipments tend to come by ship rather than overland and so would even greater proportion of drugs by weight would be expected to pass through the Netherlands prior to distribution. If fact, then one conclusion is that if illicit drug wholesale and retail prices were linked solely to the opportunity for trafficking, then Dutch prices are far higher than what would be expected. From this it might be infered that Dutch law enforcement, concentrated as it is upon the wholesale-level market, is driving up prices far more effectively (relative to expectations) than its neighbours. Perhaps the Dutch are not so soft on drugs as some policy-obsessive commentators and institutions might like us to believe. This possibility cannot be stated as a conclusion since there is a clear need for additional empirical evidence and measurement from different angles.
If the present tentative analysis is correct, there could be important implications for international drug policy. These would require further investigation than is given here, but this does not preclude them being flagged. Firstly, since it is expected that the flow of heroin and cocaine is greater into the Netherlands, the policy of deliberately segregating the hard and soft drug markets might be a pragmatic mean~ of reducing the illicit consumption of' hard drugs , As well as a means of reducing, the harm deriving from these drugs, the market separation policy may be a means of reducing the demand for hard drugs through reducing the 'gateway' effect. Secondly, the antithesis of this would mean that removing this policy, via, say, international political pressures, could lead to an Increase in the level of' illicit use of' hard drugs. Thirdly, the analysis would suggest that enforcement efforts concentrated in Dutch ports could stop a disproportionate amount of drug trafficking. but the practical problem that the sheer volume of trade imposes (in 1990 it was already nearly 300 million tonnes annually through Rotterdam alone) would make this nigh impossible to implement. Even if law enforcement effort were doubled, and even if' this doubled the risk of' trafficking. which it probably would not,' the overall risk of trafficking by sea might still be insignificant.
The analysis, while conceptually simple, might have further implications. It may shed some light upon the priorities that are accorded to different factors in decision-making related to national and international drug policy. The International Narcotics Control Board, for example, seems obliged to take a primarily legalistic stance in relation to delineating the parameters of the conventions, rather than assessing the pro's and con's of policies in their specific contexts. Just as local cultural contexts need to be considered in many instances of' international policy making, so might economic and social contexts and conditions need to be considered in others. It would be against the basic principles and moral purpose of the United Nations to insist on the Netherlands changing its policies to meet legal requirements if these changes could in fact prove detrimental to human health and well-being. A polite description of' such would be that it is cutting off the nose to spite the face.
This case study is far from perfect. It illustrates the need for others at the very least. Sweden could be a potential candidate: relatively low rates of' illicit drug use and trafficking in Sweden are sometimes attributed to the strong prohibitionist line of the national drug policy. It might be interesting to empirically examine whether higher prices and the alleged lower rates of' drug use were simply because of a lower volume of licit trade (providing little cover fir trafficking) entering a relatively small and dispersed consumer market with few customers and prospects for distribution. Other possibilities for developing routine activity-type analysis might include longitudinal analysis of variation in trade flows to see if' they are tracked to different areas by drug trafficking. Are fast changes in licit trade flows tracked by changes in illicit trade flows? How do the patterns differ for other types of illicit trade and transnational crime?
9. Concluding comment
The author is not so naive as to expect that readers will wholly embrace the case study presented here. Analysts of the Dutch case may be too politically entrenched for some to consider examining it afresh from different angles. However, it is hoped that the present analysis may shine some light upon, or stimulate debate relating to, the issues and processes involved in producing geographical variations in illicit drug trafficking and consumption. Whether or not the reader agrees with the potential implications of the case study, it is hoped that this does riot detract from a recognition of the potential for the wider application of the routine activities approach. Its apparent simplicity belies its importance, and it may have the potential to inform a range of analyses related to international illicit drug trafficking and policy, and it is in this vein that the present paper is presented.
1. land area data for Belgium, Luxembourg, the Czech Republic and Slovakia were not given in UNDP 1995.
2.For the Netherlands licit trade > index population density index = 100
3. if' law enforcement is already concentrated upon the wholesale level, then additional resources introduced here would probably experience diminishing marginal returns (risk imposed per unit of' investment).
The author would like to thank an anonymous International Journal of Drug Policy (IJDP) reviewer for comments on an earlier draft.
Cohen LW, Felson M. Social change and crime rate trends: a routine activities approach. Am Sociol Rev 1979:44(4):588 608.
Cohen P, Sas A. Cannabis use, a stepping stone to other drugs? The case of' Amsterdam. University of' Amsterdam: CEDRO, Centre tor Drug Research (internet version), 1996.
Farrell G, Mansur K, Tullis M. Cocaine and heroin trafficking and Prices in Europe 1983 1993. Br J Criminol 1996;36(2):255 81.
Velson M. Crime in E'veryday Life: Insights and Implications for Society. Thousand Oaks. CA: Pine Forge. 1994.
Horstink-Von Meyenfeldt L. 'The Netherlands: tightening up of the cafes policy. In: Dorn N, Iepsen J, Savona E, editors. European Drug Policies and Enforcement. London: Macimlian, 1996.
Keh, Farrell G. 1998. Drug trafficking in the global village. Transnat Org Crime 1997 (in press).
INCB. Report of' the International Narcotics Control Board for 1992. New York: United Nations. 1993.
Leuw E. Drugs and drug policy in the Netherlands. In: Tonry M, editor. Crime and Justice: a Review of Research. Chicago: Chicago University Press, 1991.
Leuw E, Haen Marshall 1. Between Prohibition and Legalization: the Dutch Experiment in Drug Policy. New York: Kugler. 1994.
Moore IT. Discrimination on the effective price of heroin. Am Econ Rev 1973;2:270 7.
Polak F, Lap M. A view on The Board. Int J Drug Policy 1994~3:147 55.
SAMHSA (Substance Abuse and Mental Health Services Administration), Preliminary Estimates 1rom the 1995 National Household Survey on Drug Abuse. Office of Applied Studies: Advance Report Number 18, August 1996 (Internet version).
UNDP. Human Development Report. Oxford: Oxford University, 1995.
Zimmer L, Morgan IP. Exposing Marijuana Myths: a Review of the Scientific Evidence. New York: Open Society Institute, 1995.