Beyond the substances: a multi-criteria analysis of spread of drugs in Europe

Background: Statistics on drugs are mainly broadcasted as the prevalence of single substances users, sometimes only in some cases there is an effort to present the combined use of several substances. From a policy perspective such multidimensional information has two main shortcomings: within the substances the prevalence of users is not informative about what the frequency of use is; among the substances the multidimensional information does not permit ranking (i.e. we do not know how the prevalence of a specific substance can be compared with the prevalence of another substance).

Purpose: This paper proposes the use of a new technique, the SMAA-PROMETHEE (Corrente et al. 2014), to estimate a composite index of spread of drugs, by the frequencies of use of nine different illegal substances reported on the European School Survey Project on Alcohol and other Drugs (ESPAD) from 33 European countries for 20 years (1995-2015).

Methods: The ESPAD database involves nearly 500,000 adolescents aged 16 or over and contains detailed information on the quantity and frequency of use of legal and illegal psychoactive substances. The major strength of a composite index is its capability to summarize complex multidimensional realities into one dimension. This allows an easier interpretation, assessment, communication and comparison, than a battery of many separate indicators. Nevertheless, composite indicators may send simplistic and misleading messages, if the construction process is not transparent and/or lacks conceptual and statistical principles (OECD, 2008). Compared with other aggregation techniques, SMAA-PROMETHEE has three main advantages: it allows to analyse ordinal ranked data (as it is the case of frequencies of use reported in surveys on drugs); it is non-compensatory, which means that an individual cannot compensate a lower value in one dimension with a higher value in another; it produces rankings with Monte Carlo generation of weights, avoiding any arbitrary intervention of researchers.

Results: Our estimates allow an unprecedented assessment of country-level trends in spread of drugs. Furthermore, the synthetic information allows to evaluate the drugs distribution by means of inequality measures, both within and a between the European countries. To this regard, over time we find a decreasing inequality between countries and an increasing inequality within countries. In line with the global trends detected in the distribution of socio-economic characteristics, our results reveal a convergence in the country-level spread of drugs, which is coming at the cost of increasing differences in the spread of drugs within countries in Europe.