253
Views
3
CrossRef citations to date
0
Altmetric
Research Article

Cannabis cultivation and detection: A comparative study of Belgium, Finland and Denmark

, , , &
Pages 203-215 | Published online: 24 Jan 2013
 

Abstract

Research on cannabis cultivation has identified several factors associated with a grower's likelihood of detection by law enforcement. However, these studies are difficult to compare, as they drew from different data sources and methods, and have focused on only one geographical location. This article revisits the issue of detection using a large sample of cannabis cultivators recruited in three countries: Belgium (n = 659), Denmark (n = 560) and Finland (n = 1296). Respondents were recruited in the context of a self-reported online survey conducted successively in each country between 2006 and 2008. Multivariate analyses suggest several country-specific similarities and differences. Importantly, the Finnish growers reported being arrested significantly more often than Belgians or Danes. The probability that Finnish growers would be arrested increased with time spent on growing, the size of the cultivation site and when respondents did not work alone. In Denmark, the risks increased with the size of the cultivation-related network, but decreased when respondents started growing later in life. In Belgium, no cultivation-related characteristics were associated with detection. The results indicate that the risks of apprehension for cannabis cultivation are typically country-specific. These findings are discussed in the context of country-specific policies in regards to cannabis.

Notes

Notes

1. In this study, both ‘detection’ and ‘arrest’ have the same meaning. As the main outcome variable asks respondents whether they have been arrested, it is often referred to as such.

2. Market size estimates are difficult to locate, although Kilmer and Pacula's (Citation2009) best estimates for 2005 indicate that the Belgium, Denmark and Finland retail cannabis markets were 40,900, 19,000 and 11,300 kg, respectively. To gauge the current market, we use the most recent (complete) figures reported by EMCDDA (2006) for Finland and EMCDDA (2008) for Belgium and Denmark.

3. Unfortunately, determining the extent to which these seizures were recorded as drug offenses is difficult, but the figures do provide a context in which to gauge the various markets.

4. Age is excluded as a covariate (in the final analyses) because less than 7% of Belgians responded.

5. MCMC is a stochastic process that produces parameter estimates by obtaining a posterior distribution using information from the variables included in the analysis (Gilks, Richardson, & Spiegelhalter, Citation1996). Using the PROC MI command in SAS 9.2, the MCMC procedure imputed values for missing data until reaching a monotone missing pattern.

6. Following Allison (Citation2002, Citation2009) and Schafer's (Citation1997) recommendations, models were developed including predictors, covariates and dependent variables in the analyses, as well as highly correlated auxiliary variables.

7. Five imputed data sets were generated using the MCMC procedure, then, using the five imputed data sets, an additional imputation was conducted using regression techniques. This does not impute a sixth data set, but rather performs the regression techniques (that have a monotone missing pattern) on the five imputed data sets. As a result, the final analysis includes five separate imputed data sets.

8. PROC MIANALYZE is a useful technique for examining and making inferences about parameter estimates and the effects of missing data. Using the algorithms developed by Rubin (Citation1987), SAS 9.2 (SAS Institute, Cary, NC) computes within and between-variance for the datasets and produces a ‘total variance’ score; when the imputed models are correctly identified, SAS provides consistent parameter estimates and their standard errors (Horton & Lipsitz, Citation2001). To verify the validity of these estimates, MI output provides approximations for the fraction of missing information (FMI) and its influence on the interpretation of parameter estimates (Horton & Lipsitz, Citation2001; Schafer, Citation1997).

9. The authors wish to thank one of the anonymous reviewers of Drugs: Education, Prevention, and Policy for encouraging us to address this issue directly in the article.

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.