(This HTML file is around 65 megabytes large, so it may take a while to load. Alternatively, you can view the plots and data in the GitLab repository at https://gitlab.com/denktank/mdma-policy-mdmcda)

###-----------------------------------------------------------------------------
### Check for the required packages and set options and variables
###-----------------------------------------------------------------------------

if (!('ufs' %in% row.names(installed.packages()))) {
  if (!(packageVersion('ufs') >= "0.4")) {
    stop("You need to have R package `ufs` installed in at least ",
         "version 0.4. If that version is not yet on CRAN, you can ",
         "install it from GitLab with the `remotes` package and command:\n\n",
         "remotes::install_gitlab('r-packages/ufs');");
  }
}

producePlots <- TRUE;
produceConfidencePlots <- TRUE;
quiet = TRUE;
installPkgs = FALSE;

### Check for required packages
ufs::checkPkgs("DiagrammeR", install=installPkgs);     ### Imported by `mdmcda`
ufs::checkPkgs("DiagrammeRsvg", install=installPkgs);  ### Imported by `mdmcda`
ufs::checkPkgs("ggplot2", install=installPkgs);        ### Imported by `mdmcda`
ufs::checkPkgs("ggtext", install=installPkgs);         ### Imported by `mdmcda`
ufs::checkPkgs("justifier", install=installPkgs);      ### Imported by `mdmcda`
ufs::checkPkgs("knitr", install=installPkgs);          ### Imported by `mdmcda`
ufs::checkPkgs("kableExtra", install=installPkgs);     ### Imported by `mdmcda`
ufs::checkPkgs("openxlsx", install=installPkgs);       ### Imported by `mdmcda`
ufs::checkPkgs("yum", install=installPkgs);            ### Imported by `mdmcda`

ufs::checkPkgs("tidyr", install=installPkgs);          ### Used once for pivot_wider
ufs::checkPkgs("patchwork", install=installPkgs);      ### Combining ggplots
ufs::checkPkgs("sessioninfo", install=installPkgs);    ### For reporting session information

### Update `mdmcda` from GitLab; use development branch
ufs::quietGitLabUpdate("r-packages/mdmcda@dev", quiet=quiet);

if (knitr::is_html_output()) {
  knitr::opts_chunk$set(echo=TRUE);
} else {
  knitr::opts_chunk$set(echo=FALSE);
}

options(knitr.kable.NA = '');

knitr::opts_chunk$set(dev="png", 
                            dev.args=list(type="cairo"),
                              dpi=150,
                            comment="");

ufs::opts$ez$figSize(size = "a4",
                     fontSizeMultiplier = 1.3);

mdmcda::opts$ez$figSize(size = "a4",
                        fontSizeMultiplier = 1.3);

ufs::opts$set(ggSaveDPI = 150);
mdmcda::opts$set(ggSaveDPI = 150);

### Set regular expressions for extracting decision and criterion
### identifiers from the performance subtable filenames
mdmcda::opts$set(
  performanceTable_decisionRegex  =
    c("performance_subtable_for_(.*)_on_.*_by_all\\.xlsx$",
      "\\1")
);
mdmcda::opts$set(
  performanceTable_criterionRegex =
    c("performance_subtable_for_.*_on_(.*)_by_all\\.xlsx$",
      "\\1")
);

ggBaseSize <-
  mdmcda::opts$get('ggBaseSize');

### Set variable names
scorerId <- "all";
weightProfileName <- "meanWeights"
weightedEstimateName <- paste0(weightProfileName, "_weighted_estimate");
tempWeightedEstimateName <- paste0(scorerId, "_", weightedEstimateName);

### Set path with files
currentDir <- here::here();
dataPath <- file.path(currentDir,     "data");
workingPath <- file.path(currentDir,  "results");
scoredPST_path <- file.path(dataPath, "scored-PSTs");

### Set filenames
criteriaFile <-
  file.path(dataPath,
            "mdma-policy-mdmcda-fullCriteriaDf.xlsx");
postSetsFile <-
  file.path(dataPath,
            "postsets--(adjustments-after-collective-scoring-days).xlsx");
scenariosFile <-
  file.path(dataPath,
            "mdma-scenario-definitions.xlsx");
weightsFile <-
  file.path(dataPath,
            "mdma-criteria-weights.xlsx");

### Tell ufs::knitAndSave to always 'cat' the knitr chunk
options(ufs.knitAndSave.catPlot = TRUE);

### Set automatic figure numbering
ufs::setFigCapNumbering();

MDMA MDMCDA

Overview

This reproducible report documents the analyses done for the MDMA Policy Think Tank that was active in the Netherlands in 2019-2020 (see the Background tab for more details). This document loads the produced data and computes the results. The most recent version of this document is hosted at https://denktank.gitlab.io/mdma-policy-mdmcda, the corresponding Git repository is at https://gitlab.com/denktank/mdma-policy-mdmcda, the Open Science Framework project for this repository is https://osf.io/h58r6/, and as of November the 13th, a preprint will be available at https://doi.org/10.31219/osf.io/txy5z.

This document is organised in the tabs above:

Introduction: this tab with the introduction.
Background: details about the background of this initiative.
Method: details about the method we used.
Reading and preprocessing: mostly the code used to read and preprocess the data.
Input: the specifications and estimates by the experts that were the input for the analyses
Results: an overview of the main results.
First English-language manuscript: the figures and tables for the first English-language manuscript.
First Dutch-language report: the figures and tables for the first Dutch-language report.
Additional visualisations: more detailed visualisations.
Experts’ weights: visualisations showing the weights the individual experts assigned.
Confidence scores: visualisations showing confidence scores for each performance table.
Sensitivity analyses: sensitivity analyses to test the robustness of the results.
Exporting results: mostly the code used to export the results to spreadsheets.
Session information: the versions of R and the packages that were used, to enhance reproducibility.

App and website

There is an online app at https://mdmapolicy.com/app where you can specify your own policy model and weights for the outcomes. https://mdmapolicy.com (an alias of https://mdmabeleid.nl) is a website about this project where, as of the 13th of November, a Dutch-language report will become available.

`mdmcda` R package

The Think Tank used Multi Decision Multi Criteria Decision Analysis, which is an adaption of the MCDA procedure. This document and all analyses reported here use the mdmcda R package, available through https://r-packages.gitlab.io/mdmcda. Note that it is still under development, which is why this repository has been frozen as a registration in the Open Science Framework project at https://osf.io/h58r6/. If you run into any issues with reproducing these analyses, or if you would like to use this R package to conduct an MDMCDA in your own country, you can contact Gjalt-Jorn Peters (the package developer, who also wrote this reproducible document) through Twitter or email.

Background

Discussions about drug policy often polarize. Often, individuals’ morals, prejudices and emotions come to drive the discussion rather than theory- and evidence-based argument and trying to find the best solution.

To attempt to identify the best way forward without the process being derailed by idealistic and emotional arguments, formal decision-making instruments have been employed. A famous example is the application of Multi Criteria Decision Analysis to rank drug harms in the UK (Nutt, King & Phillips, 2010). This same approach can be used to not only look at drug harms, but also at the effects different drug policies have on such harms.

Røgeberg et al. (2017) did exactly this. They compared four regulatory regimes (absolute prohibition, decriminalisation, state control and free market) and used Multi Criteria Decision Analysis (MCDA) to compare their estimated effects on 27 criteria (e.g. “Reduce user harms”, “Enable medical use”, “Promote family cohesion”, “Reduce acquisitive crime”, and “Reduce economic sosts”) for alcohol and cannabis. For both drugs, state control emerged as the optimal regime.

However, a drawback of this approach is that in the process of estimating each regulatory regime’s performance on the criteria, the experts are aware of how the pattern of estimates impacts the relative performances of each regulatory model. This leaves the process relatively open to pre-existing biases.

In the Dutch think tank process, we aimed to eliminate such biases as much as possible. To do this, an R package was developed that implements Multi Decision Multi Criteria Decision Analysis (MDMCDA). The Method tab contains a description of how this relates to MCDA. In short: the scenarios (called “regulatory regimes” by Røgeberg et a. (2017) and “policy models” in the current exercise) are not predefined, but only defined afterwards. We split up the process in an “objective phase” and a “subjective phase”.

The objective phase in MDMCDA: expert input

In the “objective phase”, the first two steps were to establish:

the outcomes (“criteria” in MCDA terms, such as prevalence of MDMA use; health damage; MDMA-related organised crime; or damage to the enviroenment);
the policy instruments (‘decisions’ in MCDA terms, such as “legal status of MDMA possession”; “license required for MDMA sale to consumers”, “minimum age”); and
and the options that comprise each decison (‘alternatives’ in MCDA terms, such as, for “license required for MDMA sale to consumers”: “No license required”, “License required”, and “Not applicable”).

The next step was to estimate the effects of all alternatives within all decisions (all policy options within all instruments) on all criteria (outcomes). For example, if selling MDMA would require a license, how would that impact the prevalence of MDMA use, health damage, MDMA-related organised crime, and damage to the environment?

Once all these effects had been estimated, that concluded the more objective expert input part of the exercise.

The subjective phase in MDMCDA: scenarios and weights

The next two steps are mostly informed by personal or political preference and so are more subjective. These two steps are specifying the following two things.

The policy models: the scenarios that each consist of one selected option (alternative) for each policy instrument (decision). For example, under a prohibitive policy, MDMA sales to consumers would be illegal, so no minimum age would be defined; whereas under a regulatory policy, the minimum age can be set to 18 years.
The weights for the outcomes (criteria). For example, somebody might value user health very much, but international politics much lower; and somebody else might value crime reduction very highly, and economical outcomes very low.

Once these have been specified, it is possible to weigh all the estimates using the specified weights, and then compute the total scores for all scenarios given the selected alternatives that comprise them.

Characteristics of MDMCDA

This approach reduces bias because the experts estimate the effects of single alternatives (options) within decisions (instruments) on single criteria each time. Because at that point, the scenarios (policy models) have not yet been defined, and the criteria have not yet been weighted, it is not possible to change these estimates to make a given scenario “win”.

An second benefit is that defining the scenarios and specifycing the weights becomes independent of the estimates. It is wasy to change scenarios or weights afterwards and again compute how each scenario scores. In addition, it is possible to compose a scenario out of the highest-scoring alternative (option) within each decision (policy instrument), as well as compute the worst-scoring scenario.

A third characteristic of MDMCDA is instead of so-called “swing weighting”, global weighting is used (see the Method tab for details and rationale). This can be soon both as a disadvantage and an advantage, but one specific benefit is that this disentangles the “importance” and “impact” dimensions, which are combined in one weight when applying swing weighting. The impact is expressed in the estimates set in phase 1, whereas the importance is determined in phase 2.

cat('This document is the technical report with all results. It contains all the R code used to generate these results. The R code can be expanded by clicking the "Code" button that sometimes appears at the right-hand side of the page, such as just above this paragraph, which will show the `cat` command used to print this paragraph.');

This document is the technical report with all results. It contains all the R code used to generate these results. The R code can be expanded by clicking the “Code” button that sometimes appears at the right-hand side of the page, such as just above this paragraph, which will show the cat command used to print this paragraph.

To view underlying data or the R scripts themselves, check out the repository at https://gitlab.com/denktank/mdma-policy-mdmcda. This file and the other repository contents are frozen when a manuscript is submitted. These frozen public registrations are available at out Open Science Framework repository at https://osf.io/h58r6/.

Method

In regular Multi Criteria Decision Analysis (MCDA), a number of alternatives are crossed with a number of criteria as means to select the optimal alternative. The table obtained by crossing all alternatives with all criteria is called a performance table, and each cell contains an estimate. By weighting the criteria in terms of their importance the optimal alternative can be computed. Often, this weighting not only weighs each criterion’s importance, but simultaneously the relative magnitude of the impact the alternatives can have on each criterion, in which case the weighting is called swing weighting. Swing weighting also serves the purpose of calibrating differences in metric between criteria.

However, regarding substance policy, achieving consensus on a manageable set of alternatives, as well as on the weights to use, is challenging and perhaps impossible. For example, experts in legal systems and policing might define a “prohibition” policy very differently from experts in health and prevention. Similarly, experts in economy and tax policy might define a “free market” system very differently than experts in policing and prevention. Similar variation exists is preferred weighting: for example, criteria relating to crime and health may be weight differently by health professionals than by police officers.

This presents a problem for MCDA, which requires predefined alternatives. To respect these different viewpoints, a specific application of the MCDA procedure was used that works for multiple decisions as well as multiple criteria (Multi Decision Multi Criteria Decision Analysis, MDMCDA). MDMCDA enables establishing both the alternatives and the weights after the estimates in the performance table have been produced. In order to do this, it sacrifices the elegance of swing weighting. Specifically, MDMCDA assumes that each alternative can be broken down into separate decisions that each have “sub-alternatives” of their own.

For example, a “prohibiton policy” alternative may differ from a “free market” alternative in the decision whether sale to consumers is legal or not, as well as in the decision whether quality standards are defined. Each of these two decisions comprises a number of sub-alternatives: for example, sale to consumers can be legal, illegal, or only legal with a license; and quality standards can either be defined by a governmental body or not. A “prohibition policy” may consist of the alternatives that sale to consumers is illegal and no quality standards are defined; and a “free market” alternative may consist of the alternatives that sale to consumers is legal and quality standards are defined.

Once all relevant decisions and all relevant “sub-alternatives” for each decision have been identified, each alternative (in the MCDA meaning of the word) can be defined in terms of which “sub-alternative” is selected within each decision. However, defining the alternatives is now no longer a precondition for producing the estimates in the performance tables (i.e. the estimates of the effects on each criterion). These estimates can now be produced on the level of the “sub-alternatives”. In other words, Multi Decision Multi Criterion Decision Analysis breaks down the MCDA procedure into a set of related sub-MCDA procedures, that are aggregated again once all performance sub-tables have been produced.

For example, in the example introduced above, the effects of whether sale to consumers is legal, illegal, or only legal with a license would be estimated for each criterion; and the effects of whether quality standards exist or not would also be estimated for each criterion. If the criteria would be, for example, “health” and “crime”, this would mean that for the first decision (legal status of sale to consumers), three estimates would be produced for the effects on health, and three for the effects on crime. For the second decision (whether quality standards exist), two estimates would be produced for the effects on health, and two for the effects on crime.

In total, therefore, ten estimates would be produced, in two performance sub-tables (each decision has its own performance sub-table). In a traditional MCDA, only four estimates would be produced: two alternatives (prohibition versus free market) crossed with two criteria (health and crime). As will be clear, when applying DMCDA as opposed to MCDA, the number of estimates to be produced is much higher. This is a disadvantage in terms of required effort, but is an advantage in terms of how complex estimating the effect on each criterion is. Breaking down each alternative into decisions and sub-alternatives means estimating the effects on each criterion requires less mental aggregation of component effects.

Often, criteria in an MCDA can be organized hierarchically. For example, the criterion “health” may in fact be a cluster of criteria “acute risks” and “long-term neurotoxicity”, and the criterion “crime” may be a cluster of criteria “organized crime” and “petty crime”. In MCDA, the full performance table is in practice split up by criterion cluster (i.e. by cluster of columns). For example, first the effects of each alternative on both criteria in the “health” cluster would be scored, and then the effects of each alternative on both criteria in the “crime” cluster.

Since swing weighting is performed for the effects on all alternatives on each criterion, this splitting up of the performance table by clusters of columns does not pose any problems. In MDMCDA, however, the performance table is also by each decision. Since in MDMCDA, the full performance table does not consist of alternatives crossed with criteria, but instead of decisions and their sub-alternatives crossed with criteria, the resulting performance table has many more rows than the performance table in conventional MCDA. During estimating each sub-alternative’s effect on the criteria in each cluster, therefore, the full performance table is also split by decision. This means the full performance table can be viewed as a grid of performance subtables: one for each criteria cluster-decision combination.

In our example, with two clusters (crime and health) that each consist of two criteria (organized and petty crime; and acute and long-term risk), and two decisions (legal status of sale to consumers and quality criteria), this results in four performance sub-tables to be filled with estimates. However, because swing-weighting simultaneously weighs each criterion’s importance and the relative magnitudes of effect of the alternatives, this requires all alternatives to be weighted simultaneously. Once alternatives are delineated into decisions and sub-alternatives, this is no longer possible.

Therefore, instead of calibrating the estimates using swing-weighting, global weighting is used (see also Monat, 2009). With global weighting, the metrics on which the alternatives effects on each criterion are estimated are established in advance. Specifically, in MDMCDA, the status quo represents a score of 0, and for each criterion, maximum deviation in the negative and positive directions are established. For example, if the status quo is relatively dismal, the positive extreme is set to 100, but the negative extreme to a smaller value, such as -20 (if things can get only around one fifth worse than they can get better). Another criterion may be scores from -100 to 50, if the status quo is already relatively desirable, but there is room for deterioration.

This approach fixes the scoring metrics such that scorers always have two anchor points (0 and either -100 or 100), and can score the estimates relative to those anchors. This eliminates the problem produced if a fixed scale would be used for all criteria (e.g. 0 to 100): in that case, the status quo is represented by a different score for each criterion, which introduces extraneous cognitive load for the scorers.

The magnitudes of the effect of the sub-alternative within each decision then manifest in the estimates scores. This has as second benefit that the weighting is restricted to the value one attaches to the criteria. In other words, this approach dissociates the more objective aspect of swing weighting fro its more subjective, moral aspect. That dissociation enables application of multiple weight profiles, enabling comparison of the effects of variations in criterion priority on total scores.

The terminology of MDMCDA is slightly different from the terminology of MCDA to retain unequivocal reference of ‘alternatives’ to the options comprising a decision. Therefore, in MDMCDA, each decision (e.g. legal status of sale to consumers) consists of two or more alternatives (avoiding the term “sub-alternatives”), and each specific configuration of alternatives (e.g. “prohibition”) is called a scenario. These scenarios, therefore, correspond to what in regular MCDA are the alternatives.

Reading and preprocessing

Reading data

###-----------------------------------------------------------------------------
### Read criteria, estimates, weights, scenarios, and post-sets
###-----------------------------------------------------------------------------

criteria <-
  mdmcda::read_criteria_from_xl(
    criteriaFile,
    showGraphs = FALSE);

estimates <-
  mdmcda::read_performance_tables(input = scoredPST_path);

weights <-
  mdmcda::read_weights_from_xl(input = weightsFile);

scenarioDefinitions <-
  mdmcda::read_scenarioDefinitions_in_columns_from_xl(
    scenariosFile
  );

postsets <-
  as.data.frame(openxlsx::read.xlsx(postSetsFile, sheet = 1));

###----------------------------------------------------------------------------
### Read labels
###----------------------------------------------------------------------------

### Read criterion labels
criterionLabels <-
  mdmcda::read_criterionLabels_from_xl(
    file.path(dataPath,
              "mdmcda-data--criterionLabels_en.xlsx")
  );
criterionLabels_NL <-
  mdmcda::read_criterionLabels_from_xl(
    file.path(dataPath,
              "mdmcda-data--criterionLabels_nl.xlsx")
  );

### Read decision labels
decisionLabels <-
  mdmcda::read_decisionLabels_from_xl(
    file.path(dataPath,
            "mdmcda-data--decisionLabels_en.xlsx")
  );
decisionLabels_NL <-
  mdmcda::read_decisionLabels_from_xl(
    file.path(dataPath,
            "mdmcda-data--decisionLabels_nl.xlsx")
  );

### Read decision descriptions
decisionDescriptions <-
  mdmcda::read_decisionDescriptions_from_xl(
    file.path(dataPath,
              "mdmcda-data--decisionLabels_en.xlsx")
  );

### Read scenario labels
scenarioLabels <-
  mdmcda::read_scenarioLabels_from_xl(
    file.path(dataPath,
            "mdmcda-data--scenarioLabels_en.xlsx")
);
scenarioLabels_NL <-
  mdmcda::read_scenarioLabels_from_xl(
    file.path(dataPath,
            "mdmcda-data--scenarioLabels_nl.xlsx")
);

### Read alternative labels
alternativeLabels <-
  mdmcda::read_alternativeLabels_from_xl(
    file.path(dataPath,
              "mdmcda-data--alternativeLabels_en.xlsx")
);
alternativeLabels_NL <-
  mdmcda::read_alternativeLabels_from_xl(
    file.path(dataPath,
              "mdmcda-data--alternativeLabels_nl.xlsx")
);

###-----------------------------------------------------------------------------
### Set orders for criteria, decisions, and scenarios
###-----------------------------------------------------------------------------

criterionOrder <-
  setdiff(names(criterionLabels),
          criteria$convenience$parentCriterionIds);
parentCriterionOrder <-
  intersect(names(criterionLabels),
            criteria$convenience$topCriterionClusters);

decisionOrder <- names(decisionLabels);
scenarioOrder <- names(scenarioLabels);

###-----------------------------------------------------------------------------
### Wrap the scenario labels for horizontal display
###-----------------------------------------------------------------------------

wrappedScenarioLabels <-
  unlist(lapply(scenarioLabels, function(x) 
    paste(strwrap(x, 10), collapse="\n")
  ));

wrappedScenarioLabels_NL <-
  unlist(lapply(scenarioLabels_NL, function(x) 
    paste(strwrap(x, 10), collapse="\n")
  ));

Criteria (outcomes of chosen policy)

The criteria (‘outcomes’) have a hierarchical structure to facilitate weighting them efficiently. They are clustered based on domain, such that each outcome can be considered an indicator of the cluster outcome. For example, prevalence in the general population, prevalence in vulnerable populations, and frequency and intensity of use by MDMA users are all indicative of MDMA use (the overarching cluster). Therefore, if a decision maker values MDMA use as an outcome, that cluster can received a high weight. The relative contribution of the three indicators can then be finetuned by setting the weights of the individual outcomes. Similarly, a decision maker who does not want MDMA use to play a large role when determining their policy can set the cluster weight to a low value or to zero. The latter would immediately take all three contained outcomes out of the total scores computed for the scenarios (policy models).

Estimates

In total, 2565 estimates were read. These estimates express estimated effects of 22 decisions (policy instruments), that together comprise a total of 95 alternatives (policy options), on a total of 27 criteria (outcomes).

Estimate post-sets

In total, 48 post-sets were loaded, pertaining to 10 decisions (policy instruments) and 11 criteria (outcomes).

Weights

To set weights, all experts submitted individually determined weights for each outcome and each outcome cluster on a scale from 0-100, by first setting the weight of the most important cluster and the most important outcome in each cluster to 100, and then setting the weight of the other clusters and outcomes according to their relative importance compared to the most important cluster or outcome (e.g. an outcome half as important as the most important outcome in the same cluster would get a weight of 50). These individually set weights will later be used to compute consensus weights that will be used in the rest of the procedures.

Weights were loaded from 16 scorers with identifiers ‘Scorer3’, ‘Scorer4’, ‘Scorer5’, ‘Scorer6’, ‘Scorer7’, ‘Scorer8’, ‘Scorer9’, ‘Scorer10’, ‘Scorer11’, ‘Scorer12’, ‘Scorer13’, ‘Scorer14’, ‘Scorer15’, ‘Scorer16’, ‘Scorer17’ & ‘Scorer18’.

Scenarios (“policy models”)

Scenarios are specific configurations of one selected alternative for each decision. The think-tank predefined 4 scenarios (policy models) with identifiers ‘repression’, ‘coffeeshop’, ‘adapted_coffeeshop’ & ‘free_market’.

Labels

Labels are read separately to enable easy translations to different languages.

Preparing data

In this section, the imported data are preprocessed. This can be inspected by expanding the code fragments.

Correcting weights

Initially, the “protection of the environment” criterion did not have a parent. However, the experts did not weigh this criterion taking that into account, causing the resulting mean weight to be too high. To correct this, this criterion was placed into its own cluster with the weight defined as the mean weight of all other clusters.

In addition, the experts specified weights for the “policy aligns with conservative values” and “policy aligns with liberal values”, but the Think Tank results should be apolitical, and so we override these to zero here.

###-----------------------------------------------------------------------------
### The 'planet' parent for the 'environment' criterion was added
### later; so we need to add a weight for it. As weight, we take the
### mean weight of all clusters, as agreed in the think tank meeting
###-----------------------------------------------------------------------------

planetWeight <-
  mean(
    weights$allWeights$weight[
      weights$allWeights$parentCriterion_id=="outcomes"
    ],
    na.rm=TRUE
  );

weights$individualWeights <-
  lapply(weights$individualWeights,
         function(dat) {
           dat[dat$id=="planet", "weight"] <- planetWeight;
           return(dat);
         });

weights$allWeights[weights$allWeights$criterion_id=="planet", "weight"] <-
  planetWeight;

###-----------------------------------------------------------------------------
### Set the weight of cultural values (liberal vs conservative) to 0 for the
### rest of the computations (because the think tank has no political/idealistic
### preference).
###-----------------------------------------------------------------------------

weights$allWeights[
  weights$allWeights$criterion_id=="cultural_values", "weight"] <-
  0;

Apply post-sets

Post-sets are individual estimates that were reconsidered by the think tank after the scoring days, or that were only completed after the scoring days.

### Do replacements
estimates$multiEstimateDf <-
  mdmcda::set_postsets(
    multiEstimateDf = estimates$multiEstimateDf,
    postsets = postsets,
    coder = scorerId
  );

replacing estimate for the effect of alternative 2 for decision b2b_sale_legal_status on criterion costs_financial_crime, which was NA, with 5.
replacing estimate for the effect of alternative 3 for decision b2b_sale_legal_status on criterion costs_financial_crime, which was NA, with 5.
replacing estimate for the effect of alternative 4 for decision b2b_sale_legal_status on criterion costs_financial_crime, which was NA, with 5.
replacing estimate for the effect of alternative 5 for decision b2b_sale_legal_status on criterion costs_financial_crime, which was NA, with 15.
replacing estimate for the effect of alternative 2 for decision consumer_sale_legal_status on criterion state_revenues_vat, which was NA, with 80.
replacing estimate for the effect of alternative 3 for decision consumer_sale_legal_status on criterion state_revenues_vat, which was NA, with 100.
replacing estimate for the effect of alternative 4 for decision consumer_sale_legal_status on criterion state_revenues_vat, which was NA, with 60.
replacing estimate for the effect of alternative 2 for decision consumer_sale_legal_status on criterion state_revenues_tax, which was NA, with 80.
replacing estimate for the effect of alternative 3 for decision consumer_sale_legal_status on criterion state_revenues_tax, which was NA, with 100.
replacing estimate for the effect of alternative 4 for decision consumer_sale_legal_status on criterion state_revenues_tax, which was NA, with 60.
replacing estimate for the effect of alternative 1 for decision crime_priority on criterion criminalisation_of_users, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision crime_priority on criterion criminalisation_of_users, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision crime_priority on criterion criminalisation_of_users, which was NA, with -50.
replacing estimate for the effect of alternative 1 for decision crime_priority on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision crime_priority on criterion international_trafficking_mdma, which was NA, with 20.
replacing estimate for the effect of alternative 3 for decision crime_priority on criterion international_trafficking_mdma, which was NA, with 20.
replacing estimate for the effect of alternative 2 for decision international_strategy on criterion organized_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision international_strategy on criterion organized_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 4 for decision international_strategy on criterion organized_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 5 for decision international_strategy on criterion organized_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 6 for decision international_strategy on criterion organized_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision international_strategy on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision international_strategy on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 4 for decision international_strategy on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 5 for decision international_strategy on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 6 for decision international_strategy on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision international_strategy on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision international_strategy on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 4 for decision international_strategy on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 5 for decision international_strategy on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 6 for decision international_strategy on criterion international_trafficking_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 1 for decision legal_age_enforcing on criterion liberal_values, which was NA, with -25.
replacing estimate for the effect of alternative 1 for decision legal_age_enforcing on criterion conservative_values, which was NA, with -25.
replacing estimate for the effect of alternative 1 for decision licenses_for_selling on criterion organized_crime_mdma, which was NA, with -25.
replacing estimate for the effect of alternative 2 for decision licenses_for_selling on criterion organized_crime_mdma, which was NA, with 50.
replacing estimate for the effect of alternative 1 for decision licenses_for_selling on criterion organized_crime_other, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision licenses_for_selling on criterion organized_crime_other, which was NA, with 30.
replacing estimate for the effect of alternative 2 for decision ontneming on criterion criminalisation_of_users, which was NA, with 0.
replacing estimate for the effect of alternative 1 for decision pricing_restrictions on criterion international_image, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision pricing_restrictions on criterion international_image, which was NA, with 0.
replacing estimate for the effect of alternative 1 for decision production_legal_status on criterion small_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision production_legal_status on criterion small_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision production_legal_status on criterion small_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 4 for decision production_legal_status on criterion small_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 5 for decision production_legal_status on criterion small_crime_mdma, which was NA, with 0.
replacing estimate for the effect of alternative 1 for decision quality_management_sanctions on criterion international_image, which was NA, with 0.
replacing estimate for the effect of alternative 2 for decision quality_management_sanctions on criterion international_image, which was NA, with 0.
replacing estimate for the effect of alternative 3 for decision quality_management_sanctions on criterion international_image, which was NA, with 0.

### Also correct the parent element for environmental protection
estimates$mergedConfidences$parentCriterion_id <-
  ifelse(estimates$mergedConfidences$parentCriterion_id == "outcomes",
         "planet",
         estimates$mergedConfidences$parentCriterion_id);

Correct alternative values for possession status

For decision (instrument) possession status, the alternative values (policy option values) were misspecified in the performance sub-tables: the values were ‘0’, ‘1’, ‘2’, ‘4’ & ‘5’ (3 was omitted). This code corrects this by shifting alternatives 4 and 5.

To verify this, the tables below were produced.

kableExtra::kable_styling(
  knitr::kable(
    table(
      estimates$multiEstimateDf[
        estimates$multiEstimateDf$decision_id=='posession_status',
        c('alternative_label', 'alternative_value')]
    )
  )
);

	0	1	2	4	5
Bezit is niet toestaan	0	27	0	0	0
Bezit is toegestaan	0	0	0	0	27
Gebruikershoeveelheid gedogen	0	0	27	0	0
Gebruikershoeveelheid legaal, grootbezit gedoogd	0	0	0	27	0
Niet van toepassing	27	0	0	0	0

estimates$multiEstimateDf[
  estimates$multiEstimateDf$decision_id=='posession_status',
  'alternative_value'] <-
  
  ifelse(
    estimates$multiEstimateDf[
      estimates$multiEstimateDf$decision_id=='posession_status',
      'alternative_value'] < 3,
    estimates$multiEstimateDf[
      estimates$multiEstimateDf$decision_id=='posession_status',
      'alternative_value'],
    as.numeric(estimates$multiEstimateDf[
      estimates$multiEstimateDf$decision_id=='posession_status',
      'alternative_value']) - 1
  );

cat("\n\nFrequencies of each option after correction:\n\n");

Frequencies of each option after correction:

kableExtra::kable_styling(
  knitr::kable(
    table(
      estimates$multiEstimateDf[
        estimates$multiEstimateDf$decision_id=='posession_status',
        c('alternative_label', 'alternative_value')]
    )
  )
);

	0	1	2	3	4
Bezit is niet toestaan	0	27	0	0	0
Bezit is toegestaan	0	0	0	0	27
Gebruikershoeveelheid gedogen	0	0	27	0	0
Gebruikershoeveelheid legaal, grootbezit gedoogd	0	0	0	27	0
Niet van toepassing	27	0	0	0	0

Weighting

The individual weights are now averaged, and then rescaled such that the most important outcome and cluster again had a weight of 100 (i.e. to the same metric used by the individual scorers). All outcome weights are then divided by 100 and multiplied with the weight of the corresponding clusters.

After this procedure, the most important outcome in each cluster has the weight of its cluster, and the other outcomes’ weights decrease proportionally. Finally, every weight is divided by the sum of all weights and multiplied by 100, so that every final weight expresses the relative contribution of the corresponding outcome in each model’s final scores.

Again, the visualisations can be inspected in the second tab.

###-----------------------------------------------------------------------------
### Aggregate weight estimates
###-----------------------------------------------------------------------------

weightsMeansAndSDs <- mdmcda::weightsMeansAndSDs(weights);

###-----------------------------------------------------------------------------
### Combine with criteria tree to accumulate over the hierachy. Note that
### because data.tree objects use R6, the criteria$criteriaTree object is
### updated in `criteria`, so we only need to store weightsMeansAndSDs
###-----------------------------------------------------------------------------

weightsMeansAndSDs <-
  mdmcda::combine_weights_and_criteria(
    weightsMeansAndSDs,
    criteria,
    weightCols = c(raw = 'weight_mean_proportion',
                   rescaled = 'weight_mean_rescaled_proportion')
  );

###-----------------------------------------------------------------------------
### Compile weight profile
###-----------------------------------------------------------------------------

weightProfiles <-
  mdmcda::create_weight_profile(weightsMeansAndSDs = weightsMeansAndSDs,
                                criteria = criteria,
                                profileName = weightProfileName);
weightProfileNames <- names(weightProfiles);

###-----------------------------------------------------------------------------
### Add weights and weighted estimates to multiEstimateDf
###-----------------------------------------------------------------------------

estimates$multiEstimateDf <-
  mdmcda::weight_multiEstimateDf(
    multiEstimateDf = estimates$multiEstimateDf,
    weightProfiles = weightProfiles,
    scorer = scorerId
  );

### Remove the `scorerId` from the name of the weighted estimates because we
### didn't have multiple estimators (just the Think Tank estimates)
estimates$multiEstimateDf[, weightedEstimateName] <-
  estimates$multiEstimateDf[, tempWeightedEstimateName];

###-----------------------------------------------------------------------------
### Export final average weights for use in Shiny app or other purposes
###-----------------------------------------------------------------------------

openxlsx::write.xlsx(
  weightsMeansAndSDs,
  file.path(workingPath,
            "mdmcda-data--weightsMeansAndSDs.xlsx"),
  overwrite = TRUE
);

### Note that the critera cluster weights are in column "rescaled_product"
### (criteria clusters can be recognized because their parent is the root
### element (the element with parentId "-"), in this case, "outcomes").
### The weights for the criteria themselves (i.e. the weights making up the
### weight profile we use in this analysis script) are in column
### "rescaled_total_percentage".

### Also export weight profiles directly
weightprofiles_asDf <-
  mdmcda::write_weightProfile_to_xl(
    weightProfiles,
    file.path(
      workingPath,
      "mdmcda-data--weightProfiles.xlsx"
    )
  );

###-----------------------------------------------------------------------------
### Compute the total scores for each alternative in each decision
###-----------------------------------------------------------------------------

scores_per_alternative <-
  mdmcda::compute_scores_per_alternative(
    multiEstimateDf = estimates$multiEstimateDf,
    weightProfiles = weightProfiles
  );

###-----------------------------------------------------------------------------
### Compose best and worst scenario
###-----------------------------------------------------------------------------

bestAlternatives <-
  mdmcda::compute_best_alternatives(
    scores_per_alternative=scores_per_alternative,
    ignoreRegex = "^0$"
  );

worstAlternatives <-
  mdmcda::compute_worst_alternatives(
    scores_per_alternative=scores_per_alternative,
    ignoreRegex = "^0$"
  );

### From these data frames, extract the vectors to store in the object with
### scenario definitions. Sometimes, multiple options score the same. In these
### cases, select the first option for the scenario definition.

bestScenario <-
  as.numeric(gsub("^.*\\s(\\d+)$",
                  "\\1",
                  bestAlternatives$alternative_id));
names(bestScenario) <- bestAlternatives$decision_id;

worstScenario <-
  as.numeric(gsub("^.*\\s(\\d+)$",
                  "\\1",
                  worstAlternatives$alternative_id));
names(worstScenario) <- worstAlternatives$decision_id;

###-----------------------------------------------------------------------------
### Define scenario with tweaks to make it more acceptable to list of scenarios
###-----------------------------------------------------------------------------

xShopScenario <-
  bestScenario;

xShopTweaks <-
  c('posession_status' = 3,
             ### User quantity is legal, higher quantities condoned
    'advertising' = 1,
             ### Advertising not allowed
    'consumer_sale_legal_status' = 4,
             ### Analogous to pharmaceutical law
    'legal_age' = 2,
             ### Age limit is 18 years
    'export_status' = 1,
             ### Export is illegal
    'healthpromotion_responsible_government' = 4
             ### Both governments
    );

xShopScenario[names(xShopTweaks)] <-
  xShopTweaks;

###-----------------------------------------------------------------------------
### Add these three new scenarios to the list of scenarios
###-----------------------------------------------------------------------------

scenarioDefinitions <-
  c(scenarioDefinitions,
    list(optimal_scenario = bestScenario,
         worst_scenario = worstScenario,
         x_shop = xShopScenario));

###-----------------------------------------------------------------------------
### Generate a weighted estimate dataframe with all scenarios and the selected
### alternatives
###-----------------------------------------------------------------------------

### Create dataframe for the weighted estimates
weightedEstimates <-
  mdmcda::build_weighted_estimate_df(
    multiEstimateDf = estimates$multiEstimateDf,
    criterionOrder = criterionOrder,
    decisionOrder = decisionOrder,
    scenarioOrder = scenarioOrder,
    scenarioDefinitions = scenarioDefinitions,
    scorer = scorerId,
    setMissingEstimates=0
  );

### Actually weigh the estimates
weightedEstimates <-
  mdmcda::weight_estimates_by_profile(weighted_estimate_df = weightedEstimates,
                                      weight_profiles = weightProfiles);

### Add parent criterion identifiers
weightedEstimates$parentCriterion_id <-
    criteria$convenience$parentCriterionIds_by_childId[
      as.character(weightedEstimates$criterion_id)
    ];

###-----------------------------------------------------------------------------
### Total scores per scenario
###-----------------------------------------------------------------------------

scoresPerScenario <-
  mdmcda::scores_by_scenario(weightedEstimates = weightedEstimates,
                             estimateCols = weightedEstimateName);

### Add labels
scoresPerScenario$label_en <-
  scenarioLabels[scoresPerScenario$scenario_id];
scoresPerScenario$label_nl <-
  scenarioLabels_NL[scoresPerScenario$scenario_id];

### Sort in scenarioOrder
scoresPerScenario <-
  scoresPerScenario[match(scenarioOrder, scoresPerScenario$scenario_id), ];

### Get English and Dutch clean versions
scoresPerScenario_en <-
  scoresPerScenario[, c("label_en", weightedEstimateName)];
scoresPerScenario_nl <-
  scoresPerScenario[, c("label_nl", weightedEstimateName)];

### Round the scores
scoresPerScenario_en[, weightedEstimateName] <-
  round(scoresPerScenario_en[, weightedEstimateName]);
scoresPerScenario_nl[, weightedEstimateName] <-
  round(scoresPerScenario_nl[, weightedEstimateName]);

### Pretty column names
names(scoresPerScenario_en) <- c("Policy model", "Overall score");
names(scoresPerScenario_nl) <- c("Beleidsmodel", "Totaalscore");

###-----------------------------------------------------------------------------
### Compute the scores and plots for all scenarios
###-----------------------------------------------------------------------------

scenarioScores <-
  lapply(
    names(scenarioDefinitions),
    function(x) {
      return(
        mdmcda::scenario_overview(
          scenario = scenarioDefinitions[[x]],
          scenarioLabel = scenarioLabels[x],
          multiEstimateDf = estimates$multiEstimateDf,
          estimateCol = weightedEstimateName,
          decisionOrder = decisionOrder,
          decisionLabels = decisionLabels,
          criterionOrder = criterionOrder,
          criterionLabels = criterionLabels,
          alternativeLabels = alternativeLabels,
          parentCriterionIds_by_childId =
            criteria$convenience$parentCriterionIds_by_childId,
          parentCriterionOrder = parentCriterionOrder,
          scoreBarchart_criteria_args = list(yLab = "Total score per outcome\n(sum of weighted estimates)",
                                             fill = "black"),
          scoreBarchart_decisions_args = list(yLab = "Total score per instrument\n(sum of weighted estimates)",
                                              fill = "black")
        )
      );
    }
  );
names(scenarioScores) <- names(scenarioDefinitions);

### Dutch versions
scenarioScores_nl <-
  lapply(
    names(scenarioDefinitions),
    function(x) {
      return(
        mdmcda::scenario_overview(
          scenario = scenarioDefinitions[[x]],
          scenarioLabel = scenarioLabels_NL[x],
          multiEstimateDf = estimates$multiEstimateDf,
          estimateCol = weightedEstimateName,
          decisionOrder = decisionOrder,
          decisionLabels = decisionLabels_NL,
          criterionOrder = criterionOrder,
          criterionLabels = criterionLabels_NL,
          alternativeLabels = alternativeLabels_NL,
          parentCriterionIds_by_childId =
            criteria$convenience$parentCriterionIds_by_childId,
          parentCriterionOrder = parentCriterionOrder,
          scoreBarchart_criteria_args = list(yLab = "Score",
                                             fill = "white"),
          scoreBarchart_decisions_args = list(yLab = "Score",
                                              fill = "white")
        )
      );
    }
  );
names(scenarioScores_nl) <- names(scenarioDefinitions);

###~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
###
### Determine order for displaying the criteria within their clusters. This is
### based on the decreasingly sorted scores in the optimal model.
###
### This should not be deleted, but only needs to be run once and then the
### results are copy-pasted at the top.
###
###~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

# criteriaOrder <-
#   unlist(
#     lapply(
#       parentCriterionOrder,
#       function(clusterName) {
#         res <-
#           scenarioScores$optimal_scenario$byCriterion[
#             criteria$convenience$childCriteriaIds[[clusterName]],
#             ,
#             drop=FALSE
#           ];
#         return(
#           res$criterion_id[
#             order(
#               res[, weightedEstimateName],
#               decreasing = TRUE
#             )
#           ]
#         );
#       }
#     )
#   );
# dput(criteriaOrder);

###~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
###
### End of commented out fragment that should be preserved
###
###~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

###-----------------------------------------------------------------------------
### Create performance tables for each scenario
###-----------------------------------------------------------------------------

performanceTables <-
  lapply(
    unique(weightedEstimates$scenario_id),
    function(currentScenario) {
      return(tidyr::pivot_wider(
        weightedEstimates[weightedEstimates$scenario_id==currentScenario,
                         c("decision_id",
                           "criterion_id",
                           weightedEstimateName)],
        id_cols="decision_id",
        names_from="criterion_id",
        values_from=tidyselect::all_of(weightedEstimateName)
      ));
    }
  );
names(performanceTables) <- unique(weightedEstimates$scenario_id);

Input

Outcomes and anchors

Outcome tree

cat("\n\n<div style='width: 70%'>\n\n");

cat(
  mdmcda::plot_criteria(
    criteria,
    labels = criterionLabels,
    show_weights = FALSE,
    renderGraph = FALSE,
    returnSVG = TRUE,
    outputFile = file.path(workingPath,
                           "criteria-tree-without-weights.pdf")
  )
);

cat("\n\n</div>\n\n");

Anchors of each outcome

for (currentCriterion in criterionOrder) {
  
  ufs::cat0("\n\n##### ", criterionLabels[currentCriterion], "\n\n");
  
  if (producePlots) {
  
    ufs::knitAndSave(criteria$anchoringGraphs[[currentCriterion]],
                     path = workingPath,
                     figWidth = 5,
                     figHeight = 5,
                     figCaption = paste0("Outcome anchors for outcome ",
                                         criterionLabels[currentCriterion]));
    
  }
  
}

Use frequency and intensity by users

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning in sprintf(suffix, counter): one argument not used by format ''

Figure 1: Outcome anchors for outcome Use frequency and intensity by users.

Use in vulnerable populations

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 2: Outcome anchors for outcome Use in vulnerable populations.

Prevalence (general population)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 3: Outcome anchors for outcome Prevalence (general population).

Quality of information about MDMA

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 4: Outcome anchors for outcome Quality of information about MDMA.

Health damage

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 5: Outcome anchors for outcome Health damage.

Stigmatization of users

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 6: Outcome anchors for outcome Stigmatization of users.

Social benefits

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 7: Outcome anchors for outcome Social benefits.

Health benefits

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 8: Outcome anchors for outcome Health benefits.

Social damage

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 9: Outcome anchors for outcome Social damage.

Shift to other drugs

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 10: Outcome anchors for outcome Shift to other drugs.

Organized crime (MDMA-related)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 11: Outcome anchors for outcome Organized crime (MDMA-related).

Criminalisation of users

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 12: Outcome anchors for outcome Criminalisation of users.

MDMA-related small crime

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 13: Outcome anchors for outcome MDMA-related small crime.

Organized crime (not MDMA-related)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 14: Outcome anchors for outcome Organized crime (not MDMA-related).

Criminals exploitating vulnerable groups

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 15: Outcome anchors for outcome Criminals exploitating vulnerable groups.

International trafficking of MDMA

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 16: Outcome anchors for outcome International trafficking of MDMA.

State revenue (other)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 17: Outcome anchors for outcome State revenue (other).

State revenue (VAT)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 18: Outcome anchors for outcome State revenue (VAT).

Costs related to pollution (dumpings)

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 19: Outcome anchors for outcome Costs related to pollution (dumpings).

Costs related to health damage

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 20: Outcome anchors for outcome Costs related to health damage.

Costs related to crime

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 21: Outcome anchors for outcome Costs related to crime.

Economic boycotts

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 22: Outcome anchors for outcome Economic boycotts.

International image of the Netherlands

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 23: Outcome anchors for outcome International image of the Netherlands.

International legal countermeasures

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 24: Outcome anchors for outcome International legal countermeasures.

Environmental damage

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 25: Outcome anchors for outcome Environmental damage.

Policy consistent with conservative values

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 26: Outcome anchors for outcome Policy consistent with conservative values.

Policy consistent with liberal values

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 27: Outcome anchors for outcome Policy consistent with liberal values.

Performance (sub)tables

These are the performance subtables that were scored. There is one for each combination of a policy instrument (with its various policy options) and outcome cluster (with its child outcomes).

performanceSubtables <-
  mdmcda::all_performanceSubtables_from_estimates(
    estimates = estimates,
    criteria = criteria,
    alternativeLabels = alternativeLabels,
    parentCriterionOrder = parentCriterionOrder,
    criterionLabels = criterionLabels,
    decisionOrder = decisionOrder,
    decisionLabels = decisionLabels,
    humanReadableOnly = TRUE
  );

for (parentCriterion_id in parentCriterionOrder) {
    
  ufs::cat0("\n\n#### Outcome cluster: ", criterionLabels[parentCriterion_id], " {.tabset .tabset-pills}\n\n");
  
  for (decision_id in decisionOrder) {

    ufs::cat0("\n\n##### Instrument: ", decisionLabels[decision_id], "\n\n");
    
    ### Clear names of first two columns
    names(performanceSubtables[[decision_id]][[parentCriterion_id]])[1:2] <-
      "";

    cat(
      kableExtra::kable_styling(
        knitr::kable(
          performanceSubtables[[decision_id]][[parentCriterion_id]],
          row.names=FALSE
        )
      )
    );
    
  }
}

Outcome cluster: Use (prevalence & patterns)

Instrument: Possession

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Possession	Not applicable	0	0	0
Possession	Possession is illegal	10	10	10
Possession	User quantity is condoned	0	0	0
Possession	User quantity is legal, higher quantities condoned	-15	-30	-10
Possession	Possession is legal	-20	-40	-20

Instrument: Packaging

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Packaging	Not applicable	0	0	0
Packaging	Plain packaging with prevention message	5	5	5
Packaging	Only plain packaging	0	0	0
Packaging	Only prevention message	-15	-5	5
Packaging	No restrictions	-20	-10	-10

Instrument: Advertising

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Advertising	Not applicable	0	0	0
Advertising	Advertising not allowed	0	0	0
Advertising	Packaging advertising only	-5	-10	0
Advertising	Age limited advertising only	-10	-30	-15
Advertising	Business to business advertising only	0	0	0
Advertising	Advertising allowed	-20	-30	-15

Instrument: Sales across companies

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Sales across companies	Not applicable	0	0	0
Sales across companies	Illegal	0	0	0
Sales across companies	Regulated	0	0	0
Sales across companies	Analogous to commodity law	0	0	0
Sales across companies	Analogous to pharmaceutical law	0	0	0
Sales across companies	Allowed	0	0	0

Instrument: Sales to consumers

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Sales to consumers	Not applicable	0	0	0
Sales to consumers	Not allowed	0	0	0
Sales to consumers	Regulated	5	-5	20
Sales to consumers	Analogous to commodity law	-10	-20	-20
Sales to consumers	Analogous to pharmaceutical law	0	0	0
Sales to consumers	Allowed	-20	-30	-30

Instrument: Age limit

		Use in vulnerable populations
Age limit	Not applicable	0
Age limit	No age limit	-20
Age limit	Age limit is 18 years	0
Age limit	Age limit is higher than 18 years	0

Instrument: Punishable

		Use in vulnerable populations
Punishable	Not applicable	0
Punishable	Nobody punishable	-10
Punishable	Seller punishable	20
Punishable	Seller and buyer punishable	20

Instrument: Legal requirements for selling

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Legal requirements for selling	Not applicable	0	0	0
Legal requirements for selling	No license required	-25	-20	-20
Legal requirements for selling	License required	25	0	20

Instrument: Price restrictions for consumer sale

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Price restrictions for consumer sale	Not applicable	0	0	0
Price restrictions for consumer sale	No price restricions	-20	-10	-10
Price restrictions for consumer sale	Minimum pricing	20	10	0

Instrument: MDMA quality rules (QA)

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
MDMA quality rules (QA)	Not applicable	0	0	0
MDMA quality rules (QA)	No quality definition/requirements	0	0	10
MDMA quality rules (QA)	Quality definition/requirements in place	-10	-20	20

Instrument: Sanctioning QA rules

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Sanctioning QA rules	Not applicable	0	0	0
Sanctioning QA rules	None	0	0	0
Sanctioning QA rules	Mild	0	-10	10
Sanctioning QA rules	Severe	-10	-20	20

Instrument: Monitoring

		Use frequency and intensity by users
Monitoring	Not applicable	0
Monitoring	None	0
Monitoring	Selective	0
Monitoring	Extensive	20

Instrument: Health promotion (HP) funding

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Health promotion (HP) funding	Not applicable	0	0	0
Health promotion (HP) funding	No funding	-5	-5	-20
Health promotion (HP) funding	Minimal funding	0	0	0
Health promotion (HP) funding	Substantial funding	0	0	20

Instrument: Control prevention policy

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Control prevention policy	Not applicable	0	0	0
Control prevention policy	None	0	0	0
Control prevention policy	Weak	0	0	0
Control prevention policy	Strong	0	0	0

Instrument: Health promotion perspective

		Use in vulnerable populations	Use frequency and intensity by users
Health promotion perspective	Not applicable	0	0
Health promotion perspective	Abstinence	0	-5
Health promotion perspective	Harm reduction	5	20

Instrument: Government responsible for HP

		Use frequency and intensity by users
Government responsible for HP	Not applicable	0
Government responsible for HP	No government	-10
Government responsible for HP	Regional government	0
Government responsible for HP	National government	0
Government responsible for HP	Both governments	0

Instrument: Production of MDMA

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Production of MDMA	Not applicable	0	0	0
Production of MDMA	Illegal	0	0	0
Production of MDMA	Regulated	-20	-20	20
Production of MDMA	Analogous to commodity law	-40	-40	20
Production of MDMA	Analogous to pharmaceutical law	0	0	20
Production of MDMA	Allowed	-40	-40	10

Instrument: Export status

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Export status	Not applicable	0	0	0
Export status	Export is illegal	0	0	0
Export status	Export is legal	0	0	0

Instrument: International strategy

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
International strategy	Not applicable	0	0	0
International strategy	Compliant	0	0	0
International strategy	Violation	0	0	0
International strategy	Based on condoning	0	0	0
International strategy	Inter se	0	0	0
International strategy	Exceptional position	0	0	0
International strategy	Adjustment of treaties	0	0	0

Instrument: Priority crime fighting

		Prevalence (general population)	Use frequency and intensity by users
Priority crime fighting	Not applicable	0	0
Priority crime fighting	Low	0	0
Priority crime fighting	Selective	0	0
Priority crime fighting	High	10	-10

Instrument: Maximum penalty

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Maximum penalty	Not applicable	0	0	0
Maximum penalty	Retain present sanction	0	0	0
Maximum penalty	More severe sanction	0	0	0

Instrument: Confiscation

		Use in vulnerable populations	Prevalence (general population)	Use frequency and intensity by users
Confiscation	Not applicable	0	0	0
Confiscation	Unchanged	0	0	0
Confiscation	Intensify	0	0	0

Outcome cluster: User health

Instrument: Possession

		Health damage	Social damage	Social benefits	Shift to other drugs	Stigmatization of users
Possession	Not applicable	0	0	0	0	0
Possession	Possession is illegal	-20	-50	-10	-30	-40
Possession	User quantity is condoned	0	0	0	0	0
Possession	User quantity is legal, higher quantities condoned	-20	0	10	0	10
Possession	Possession is legal	-20	0	10	0	10

Instrument: Packaging

		Health damage	Social damage	Social benefits	Quality of information about MDMA
Packaging	Not applicable	0	0	0	0
Packaging	Plain packaging with prevention message	5	0	0	50
Packaging	Only plain packaging	0	0	0	0
Packaging	Only prevention message	-10	-5	0	50
Packaging	No restrictions	-20	-10	10	0

Instrument: Advertising

		Health damage	Social damage	Social benefits	Stigmatization of users
Advertising	Not applicable	0	0	0	0
Advertising	Advertising not allowed	0	0	0	0
Advertising	Packaging advertising only	-10	-10	10	10
Advertising	Age limited advertising only	-30	-30	30	40
Advertising	Business to business advertising only	0	0	0	0
Advertising	Advertising allowed	-30	-30	30	50

Instrument: Sales across companies

		Health damage	Quality of information about MDMA	Stigmatization of users
Sales across companies	Not applicable	0	0	0
Sales across companies	Illegal	0	0	0
Sales across companies	Regulated	10	10	10
Sales across companies	Analogous to commodity law	10	10	10
Sales across companies	Analogous to pharmaceutical law	20	20	10
Sales across companies	Allowed	5	5	10

Instrument: Sales to consumers

		Health damage	Social damage	Health benefits	Social benefits	Quality of information about MDMA	Stigmatization of users
Sales to consumers	Not applicable	0	0	0	0	0	0
Sales to consumers	Not allowed	0	0	0	0	0	0
Sales to consumers	Regulated	-5	-5	0	10	20	20
Sales to consumers	Analogous to commodity law	-20	-20	0	20	10	30
Sales to consumers	Analogous to pharmaceutical law	0	0	10	0	20	10
Sales to consumers	Allowed	-30	-30	0	20	10	40

Instrument: Age limit

		Health damage	Social damage	Social benefits	Shift to other drugs	Stigmatization of users
Age limit	Not applicable	0	0	0	0	0
Age limit	No age limit	-10	-10	10	0	0
Age limit	Age limit is 18 years	0	0	0	-10	-5
Age limit	Age limit is higher than 18 years	0	0	0	-20	-10

Instrument: Punishable

		Health damage	Social damage	Social benefits	Shift to other drugs	Stigmatization of users
Punishable	Not applicable	0	0	0	0	0
Punishable	Nobody punishable	-10	-10	10	0	0
Punishable	Seller punishable	10	10	0	0	0
Punishable	Seller and buyer punishable	20	-20	0	-10	-10

Instrument: Legal requirements for selling

		Health damage	Social damage	Quality of information about MDMA
Legal requirements for selling	Not applicable	0	0	0
Legal requirements for selling	No license required	-15	0	0
Legal requirements for selling	License required	20	10	40

Instrument: Price restrictions for consumer sale

		Health damage	Social damage	Social benefits	Shift to other drugs
Price restrictions for consumer sale	Not applicable	0	0	0	0
Price restrictions for consumer sale	No price restricions	-20	-10	10	0
Price restrictions for consumer sale	Minimum pricing	50	10	-10	-10

Instrument: MDMA quality rules (QA)

		Health damage	Health benefits	Quality of information about MDMA	Stigmatization of users
MDMA quality rules (QA)	Not applicable	0	0	0	0
MDMA quality rules (QA)	No quality definition/requirements	10	0	0	0
MDMA quality rules (QA)	Quality definition/requirements in place	40	10	70	20

Instrument: Sanctioning QA rules

		Health damage	Health benefits	Quality of information about MDMA
Sanctioning QA rules	Not applicable	0	0	0
Sanctioning QA rules	None	10	-10	10
Sanctioning QA rules	Mild	20	10	20
Sanctioning QA rules	Severe	50	20	40

Instrument: Monitoring

		Health damage	Quality of information about MDMA	Stigmatization of users
Monitoring	Not applicable	0	0	0
Monitoring	None	-30	-50	-20
Monitoring	Selective	0	0	0
Monitoring	Extensive	30	50	20

Instrument: Health promotion (HP) funding

		Health damage	Social damage	Health benefits	Social benefits	Quality of information about MDMA	Stigmatization of users
Health promotion (HP) funding	Not applicable	0	0	0	0	0	0
Health promotion (HP) funding	No funding	-30	-5	0	0	-50	-20
Health promotion (HP) funding	Minimal funding	0	0	0	0	0	0
Health promotion (HP) funding	Substantial funding	30	15	5	15	50	20

Instrument: Control prevention policy

		Health damage	Social damage	Health benefits	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
Control prevention policy	Not applicable	0	0	0	0	0	0	0
Control prevention policy	None	0	0	0	0	0	0	0
Control prevention policy	Weak	0	0	0	0	0	0	0
Control prevention policy	Strong	0	0	0	0	0	0	0

Instrument: Health promotion perspective

		Health damage	Social damage	Health benefits	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
Health promotion perspective	Not applicable	0	0	0	0	0	0	0
Health promotion perspective	Abstinence	-30	-10	-5	-10	-20	-60	-25
Health promotion perspective	Harm reduction	15	10	5	10	0	10	10

Instrument: Government responsible for HP

		Health damage	Social damage	Quality of information about MDMA
Government responsible for HP	Not applicable	0	0	0
Government responsible for HP	No government	-10	-10	-20
Government responsible for HP	Regional government	0	0	0
Government responsible for HP	National government	0	0	0
Government responsible for HP	Both governments	0	0	0

Instrument: Production of MDMA

		Health damage	Health benefits	Social benefits	Quality of information about MDMA	Stigmatization of users
Production of MDMA	Not applicable	0	0	0	0	0
Production of MDMA	Illegal	0	0	0	0	0
Production of MDMA	Regulated	-10	0	20	50	30
Production of MDMA	Analogous to commodity law	-20	0	10	50	30
Production of MDMA	Analogous to pharmaceutical law	0	10	30	30	20
Production of MDMA	Allowed	-20	0	10	10	30

Instrument: Export status

		Health damage	Social damage	Health benefits	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
Export status	Not applicable	0	0	0	0	0	0	0
Export status	Export is illegal	0	0	0	0	0	0	0
Export status	Export is legal	0	0	0	0	0	0	0

Instrument: International strategy

		Health damage	Social damage	Health benefits	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
International strategy	Not applicable	0	0	0	0	0	0	0
International strategy	Compliant	0	0	0	0	0	0	0
International strategy	Violation	0	0	0	0	0	0	0
International strategy	Based on condoning	0	0	0	0	0	0	0
International strategy	Inter se	0	0	0	0	0	0	0
International strategy	Exceptional position	0	0	0	0	0	0	0
International strategy	Adjustment of treaties	0	0	0	0	0	0	0

Instrument: Priority crime fighting

		Health damage	Social damage	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
Priority crime fighting	Not applicable	0	0	0	0	0	0
Priority crime fighting	Low	10	10	5	0	5	10
Priority crime fighting	Selective	0	0	0	0	0	0
Priority crime fighting	High	-20	-15	-5	-20	-40	-40

Instrument: Maximum penalty

		Social damage
Maximum penalty	Not applicable	0
Maximum penalty	Retain present sanction	0
Maximum penalty	More severe sanction	-10

Instrument: Confiscation

		Health damage	Social damage	Health benefits	Social benefits	Shift to other drugs	Quality of information about MDMA	Stigmatization of users
Confiscation	Not applicable	0	0	0	0	0	0	0
Confiscation	Unchanged	0	0	0	0	0	0	0
Confiscation	Intensify	0	0	0	0	0	0	0

Outcome cluster: Crime

Instrument: Possession

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Possession	Not applicable	0	0	0	0	0
Possession	Possession is illegal	-100	-100	0	0	-50
Possession	User quantity is condoned	0	0	0	0	0
Possession	User quantity is legal, higher quantities condoned	50	50	-20	-30	20
Possession	Possession is legal	50	50	-20	-30	20

Instrument: Packaging

		Organized crime (MDMA-related)
Packaging	Not applicable	0
Packaging	Plain packaging with prevention message	-20
Packaging	Only plain packaging	-20
Packaging	Only prevention message	0
Packaging	No restrictions	-20

Instrument: Advertising

		MDMA-related small crime	Organized crime (MDMA-related)
Advertising	Not applicable	0	0
Advertising	Advertising not allowed	0	0
Advertising	Packaging advertising only	10	10
Advertising	Age limited advertising only	30	30
Advertising	Business to business advertising only	10	10
Advertising	Advertising allowed	40	40

Instrument: Sales across companies

		Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA
Sales across companies	Not applicable	0	0	0
Sales across companies	Illegal	0	0	0
Sales across companies	Regulated	60	-10	-10
Sales across companies	Analogous to commodity law	80	-10	-80
Sales across companies	Analogous to pharmaceutical law	60	-10	-10
Sales across companies	Allowed	80	-10	-80

Instrument: Sales to consumers

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Sales to consumers	Not applicable	0	0	0	0	0	0
Sales to consumers	Not allowed	0	0	0	0	0	0
Sales to consumers	Regulated	50	40	20	-20	-10	30
Sales to consumers	Analogous to commodity law	50	50	60	-60	-20	50
Sales to consumers	Analogous to pharmaceutical law	50	40	20	-20	-10	30
Sales to consumers	Allowed	50	50	60	-60	-20	50

Instrument: Age limit

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Criminals exploitating vulnerable groups
Age limit	Not applicable	0	0	0	0
Age limit	No age limit	50	50	10	0
Age limit	Age limit is 18 years	30	30	10	-10
Age limit	Age limit is higher than 18 years	20	20	0	-10

Instrument: Punishable

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Criminals exploitating vulnerable groups
Punishable	Not applicable	0	0	0	0
Punishable	Nobody punishable	50	50	30	20
Punishable	Seller punishable	50	30	20	10
Punishable	Seller and buyer punishable	30	10	10	10

Instrument: Legal requirements for selling

		MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	Criminals exploitating vulnerable groups
Legal requirements for selling	Not applicable	0	0	0	0
Legal requirements for selling	No license required	50	-25	0	10
Legal requirements for selling	License required	40	50	30	20

Instrument: Price restrictions for consumer sale

		MDMA-related small crime	Organized crime (MDMA-related)	Criminals exploitating vulnerable groups
Price restrictions for consumer sale	Not applicable	0	0	0
Price restrictions for consumer sale	No price restricions	0	0	0
Price restrictions for consumer sale	Minimum pricing	-20	-10	-10

Instrument: MDMA quality rules (QA)

		MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)
MDMA quality rules (QA)	Not applicable	0	0	0
MDMA quality rules (QA)	No quality definition/requirements	0	0	0
MDMA quality rules (QA)	Quality definition/requirements in place	20	20	-10

Instrument: Sanctioning QA rules

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Sanctioning QA rules	Not applicable	0	0	0	0	0	0
Sanctioning QA rules	None	0	0	0	0	0	0
Sanctioning QA rules	Mild	0	0	0	0	0	0
Sanctioning QA rules	Severe	0	0	0	0	0	0

Instrument: Monitoring

		MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Monitoring	Not applicable	0	0	0	0	0
Monitoring	None	0	0	0	0	0
Monitoring	Selective	10	10	10	10	10
Monitoring	Extensive	20	20	20	20	20

Instrument: Health promotion (HP) funding

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Health promotion (HP) funding	Not applicable	0	0	0	0	0	0
Health promotion (HP) funding	No funding	0	0	0	0	0	0
Health promotion (HP) funding	Minimal funding	0	0	0	0	0	0
Health promotion (HP) funding	Substantial funding	0	0	0	0	0	0

Instrument: Control prevention policy

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Control prevention policy	Not applicable	0	0	0	0	0	0
Control prevention policy	None	0	0	0	0	0	0
Control prevention policy	Weak	0	0	0	0	0	0
Control prevention policy	Strong	0	0	0	0	0	0

Instrument: Health promotion perspective

		Criminalisation of users
Health promotion perspective	Not applicable	0
Health promotion perspective	Abstinence	0
Health promotion perspective	Harm reduction	15

Instrument: Government responsible for HP

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Government responsible for HP	Not applicable	0	0	0	0	0	0
Government responsible for HP	No government	0	0	0	0	0	0
Government responsible for HP	Regional government	0	0	0	0	0	0
Government responsible for HP	National government	0	0	0	0	0	0
Government responsible for HP	Both governments	0	0	0	0	0	0

Instrument: Production of MDMA

		Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Production of MDMA	Not applicable	0	0	0	0
Production of MDMA	Illegal	0	0	0	0
Production of MDMA	Regulated	50	-10	-10	-10
Production of MDMA	Analogous to commodity law	90	-50	-80	-10
Production of MDMA	Analogous to pharmaceutical law	20	-10	-10	0
Production of MDMA	Allowed	90	-60	-90	-10

Instrument: Export status

		Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA
Export status	Not applicable	0	0	0
Export status	Export is illegal	0	0	0
Export status	Export is legal	50	-20	-40

Instrument: International strategy

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
International strategy	Not applicable	0	0	0	0	0	0
International strategy	Compliant	0	0	0	0	0	0
International strategy	Violation	0	0	0	0	0	0
International strategy	Based on condoning	0	0	0	0	0	0
International strategy	Inter se	0	0	0	0	0	0
International strategy	Exceptional position	0	0	0	0	0	0
International strategy	Adjustment of treaties	0	0	0	0	0	0

Instrument: Priority crime fighting

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Priority crime fighting	Not applicable	0	0	0	0	0	0
Priority crime fighting	Low	0	-20	-50	20	0	10
Priority crime fighting	Selective	0	0	0	0	20	0
Priority crime fighting	High	-50	-20	20	-20	20	-20

Instrument: Maximum penalty

		Criminalisation of users	MDMA-related small crime	Organized crime (MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Maximum penalty	Not applicable	0	0	0	0	0
Maximum penalty	Retain present sanction	0	0	0	0	0
Maximum penalty	More severe sanction	-20	25	10	20	-20

Instrument: Confiscation

		MDMA-related small crime	Organized crime (MDMA-related)	Organized crime (not MDMA-related)	International trafficking of MDMA	Criminals exploitating vulnerable groups
Confiscation	Not applicable	0	0	0	0	0
Confiscation	Unchanged	0	0	0	0	0
Confiscation	Intensify	25	60	60	20	-10

Outcome cluster: Financial costs and benefits

Instrument: Possession

		Costs related to crime
Possession	Not applicable	0
Possession	Possession is illegal	0
Possession	User quantity is condoned	0
Possession	User quantity is legal, higher quantities condoned	10
Possession	Possession is legal	10

Instrument: Packaging

		State revenue (VAT)	State revenue (other)	Costs related to health damage
Packaging	Not applicable	0	0	0
Packaging	Plain packaging with prevention message	70	70	-20
Packaging	Only plain packaging	80	80	-30
Packaging	Only prevention message	90	90	-40
Packaging	No restrictions	100	100	-50

Instrument: Advertising

		State revenue (VAT)	State revenue (other)	Costs related to health damage
Advertising	Not applicable	0	0	0
Advertising	Advertising not allowed	0	0	0
Advertising	Packaging advertising only	30	30	-10
Advertising	Age limited advertising only	60	60	-20
Advertising	Business to business advertising only	10	20	0
Advertising	Advertising allowed	100	100	-25

Instrument: Sales across companies

		State revenue (VAT)	State revenue (other)	Costs related to pollution (dumpings)	Costs related to crime
Sales across companies	Not applicable	0	0	0	0
Sales across companies	Illegal	0	0	0	0
Sales across companies	Regulated	10	75	60	5
Sales across companies	Analogous to commodity law	10	100	60	5
Sales across companies	Analogous to pharmaceutical law	10	50	80	5
Sales across companies	Allowed	10	100	60	15

Instrument: Sales to consumers

		State revenue (VAT)	State revenue (other)	Costs related to health damage
Sales to consumers	Not applicable	0	0	0
Sales to consumers	Not allowed	0	0	0
Sales to consumers	Regulated	80	80	-20
Sales to consumers	Analogous to commodity law	100	100	-25
Sales to consumers	Analogous to pharmaceutical law	60	60	-15
Sales to consumers	Allowed	100	100	-25

Instrument: Age limit

		State revenue (VAT)	Costs related to health damage	Costs related to crime
Age limit	Not applicable	0	0	0
Age limit	No age limit	100	-30	0
Age limit	Age limit is 18 years	90	-20	-10
Age limit	Age limit is higher than 18 years	70	-10	-20

Instrument: Punishable

		State revenue (VAT)	Costs related to health damage	Costs related to crime
Punishable	Not applicable	0	0	0
Punishable	Nobody punishable	0	0	0
Punishable	Seller punishable	-5	10	-20
Punishable	Seller and buyer punishable	-5	10	-20

Instrument: Legal requirements for selling

		State revenue (VAT)	State revenue (other)	Costs related to pollution (dumpings)	Costs related to health damage	Costs related to crime
Legal requirements for selling	Not applicable	0	0	0	0	0
Legal requirements for selling	No license required	100	100	0	-60	0
Legal requirements for selling	License required	100	100	50	-30	-10

Instrument: Price restrictions for consumer sale

		State revenue (VAT)	State revenue (other)	Costs related to health damage
Price restrictions for consumer sale	Not applicable	0	0	0
Price restrictions for consumer sale	No price restricions	100	100	-40
Price restrictions for consumer sale	Minimum pricing	70	70	-20

Instrument: MDMA quality rules (QA)

		Costs related to health damage	Costs related to crime
MDMA quality rules (QA)	Not applicable	0	0
MDMA quality rules (QA)	No quality definition/requirements	0	0
MDMA quality rules (QA)	Quality definition/requirements in place	30	20

Instrument: Sanctioning QA rules

		Costs related to health damage	Costs related to crime
Sanctioning QA rules	Not applicable	0	0
Sanctioning QA rules	None	-10	0
Sanctioning QA rules	Mild	0	-10
Sanctioning QA rules	Severe	10	-20

Instrument: Monitoring

		Costs related to health damage	Costs related to crime
Monitoring	Not applicable	0	0
Monitoring	None	-20	-20
Monitoring	Selective	0	0
Monitoring	Extensive	20	20

Instrument: Health promotion (HP) funding

		Costs related to health damage
Health promotion (HP) funding	Not applicable	0
Health promotion (HP) funding	No funding	-50
Health promotion (HP) funding	Minimal funding	0
Health promotion (HP) funding	Substantial funding	30

Instrument: Control prevention policy

		Costs related to health damage
Control prevention policy	Not applicable	0
Control prevention policy	None	0
Control prevention policy	Weak	-10
Control prevention policy	Strong	-20

Instrument: Health promotion perspective

		Costs related to health damage
Health promotion perspective	Not applicable	0
Health promotion perspective	Abstinence	-20
Health promotion perspective	Harm reduction	20

Instrument: Government responsible for HP

		Costs related to health damage
Government responsible for HP	Not applicable	0
Government responsible for HP	No government	-50
Government responsible for HP	Regional government	0
Government responsible for HP	National government	0
Government responsible for HP	Both governments	0

Instrument: Production of MDMA

		State revenue (VAT)	State revenue (other)	Costs related to pollution (dumpings)	Costs related to health damage	Costs related to crime
Production of MDMA	Not applicable	0	0	0	0	0
Production of MDMA	Illegal	0	0	0	0	0
Production of MDMA	Regulated	10	90	70	30	70
Production of MDMA	Analogous to commodity law	10	100	70	25	80
Production of MDMA	Analogous to pharmaceutical law	10	80	80	35	60
Production of MDMA	Allowed	10	100	70	0	100

Instrument: Export status

		State revenue (VAT)	State revenue (other)	Costs related to pollution (dumpings)	Costs related to health damage	Costs related to crime
Export status	Not applicable	0	0	0	0	0
Export status	Export is illegal	0	0	0	0	0
Export status	Export is legal	0	0	0	0	0

Instrument: International strategy

		Costs related to crime
International strategy	Not applicable	0
International strategy	Compliant	0
International strategy	Violation	-10
International strategy	Based on condoning	0
International strategy	Inter se	10
International strategy	Exceptional position	-25
International strategy	Adjustment of treaties	0

Instrument: Priority crime fighting

		State revenue (other)	Costs related to pollution (dumpings)	Costs related to health damage	Costs related to crime
Priority crime fighting	Not applicable	0	0	0	0
Priority crime fighting	Low	0	0	0	20
Priority crime fighting	Selective	0	0	0	0
Priority crime fighting	High	20	20	-35	-40

Instrument: Maximum penalty

		State revenue (other)	Costs related to crime
Maximum penalty	Not applicable	0	0
Maximum penalty	Retain present sanction	0	0
Maximum penalty	More severe sanction	10	-20

Instrument: Confiscation

		State revenue (other)	Costs related to crime
Confiscation	Not applicable	0	0
Confiscation	Unchanged	0	0
Confiscation	Intensify	100	-80

Outcome cluster: International politics

Instrument: Possession

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Possession	Not applicable	0	0	0
Possession	Possession is illegal	0	0	0
Possession	User quantity is condoned	-30	0	0
Possession	User quantity is legal, higher quantities condoned	-40	-5	-30
Possession	Possession is legal	-55	-10	-40

Instrument: Packaging

		International image of the Netherlands
Packaging	Not applicable	0
Packaging	Plain packaging with prevention message	-5
Packaging	Only plain packaging	-10
Packaging	Only prevention message	-20
Packaging	No restrictions	-30

Instrument: Advertising

		International image of the Netherlands
Advertising	Not applicable	0
Advertising	Advertising not allowed	0
Advertising	Packaging advertising only	-20
Advertising	Age limited advertising only	-20
Advertising	Business to business advertising only	-20
Advertising	Advertising allowed	-30

Instrument: Sales across companies

		International image of the Netherlands
Sales across companies	Not applicable	0
Sales across companies	Illegal	-10
Sales across companies	Regulated	0
Sales across companies	Analogous to commodity law	-30
Sales across companies	Analogous to pharmaceutical law	-20
Sales across companies	Allowed	-30

Instrument: Sales to consumers

		International image of the Netherlands	International legal countermeasures
Sales to consumers	Not applicable	0	0
Sales to consumers	Not allowed	0	0
Sales to consumers	Regulated	-40	-15
Sales to consumers	Analogous to commodity law	-60	-15
Sales to consumers	Analogous to pharmaceutical law	-15	0
Sales to consumers	Allowed	-60	-15

Instrument: Age limit

		International image of the Netherlands
Age limit	Not applicable	0
Age limit	No age limit	0
Age limit	Age limit is 18 years	20
Age limit	Age limit is higher than 18 years	20

Instrument: Punishable

		International image of the Netherlands
Punishable	Not applicable	0
Punishable	Nobody punishable	0
Punishable	Seller punishable	10
Punishable	Seller and buyer punishable	10

Instrument: Legal requirements for selling

		International image of the Netherlands
Legal requirements for selling	Not applicable	0
Legal requirements for selling	No license required	-80
Legal requirements for selling	License required	-50

Instrument: Price restrictions for consumer sale

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Price restrictions for consumer sale	Not applicable	0	0	0
Price restrictions for consumer sale	No price restricions	0	0	0
Price restrictions for consumer sale	Minimum pricing	0	0	0

Instrument: MDMA quality rules (QA)

		International image of the Netherlands
MDMA quality rules (QA)	Not applicable	0
MDMA quality rules (QA)	No quality definition/requirements	0
MDMA quality rules (QA)	Quality definition/requirements in place	20

Instrument: Sanctioning QA rules

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Sanctioning QA rules	Not applicable	0	0	0
Sanctioning QA rules	None	0	0	0
Sanctioning QA rules	Mild	0	0	0
Sanctioning QA rules	Severe	0	0	0

Instrument: Monitoring

		International image of the Netherlands
Monitoring	Not applicable	0
Monitoring	None	0
Monitoring	Selective	10
Monitoring	Extensive	20

Instrument: Health promotion (HP) funding

		International image of the Netherlands
Health promotion (HP) funding	Not applicable	0
Health promotion (HP) funding	No funding	-30
Health promotion (HP) funding	Minimal funding	-20
Health promotion (HP) funding	Substantial funding	0

Instrument: Control prevention policy

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Control prevention policy	Not applicable	0	0	0
Control prevention policy	None	0	0	0
Control prevention policy	Weak	0	0	0
Control prevention policy	Strong	0	0	0

Instrument: Health promotion perspective

		International image of the Netherlands
Health promotion perspective	Not applicable	0
Health promotion perspective	Abstinence	10
Health promotion perspective	Harm reduction	0

Instrument: Government responsible for HP

		International image of the Netherlands
Government responsible for HP	Not applicable	0
Government responsible for HP	No government	0
Government responsible for HP	Regional government	0
Government responsible for HP	National government	0
Government responsible for HP	Both governments	-10

Instrument: Production of MDMA

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Production of MDMA	Not applicable	0	0	0
Production of MDMA	Illegal	0	0	0
Production of MDMA	Regulated	-60	0	-60
Production of MDMA	Analogous to commodity law	-90	-10	-90
Production of MDMA	Analogous to pharmaceutical law	-60	0	-60
Production of MDMA	Allowed	-90	-10	-90

Instrument: Export status

		International image of the Netherlands	Economic boycotts	International legal countermeasures
Export status	Not applicable	0	0	0
Export status	Export is illegal	0	0	0
Export status	Export is legal	-90	-10	-90

Instrument: International strategy

		International image of the Netherlands	International legal countermeasures
International strategy	Not applicable	0	0
International strategy	Compliant	0	0
International strategy	Violation	-50	-30
International strategy	Based on condoning	-40	-10
International strategy	Inter se	-20	0
International strategy	Exceptional position	-40	-30
International strategy	Adjustment of treaties	-50	-30

Instrument: Priority crime fighting

		International image of the Netherlands	International legal countermeasures
Priority crime fighting	Not applicable	0	0
Priority crime fighting	Low	-50	-20
Priority crime fighting	Selective	0	0
Priority crime fighting	High	20	0

Instrument: Maximum penalty

		International image of the Netherlands
Maximum penalty	Not applicable	0
Maximum penalty	Retain present sanction	0
Maximum penalty	More severe sanction	30

Instrument: Confiscation

		International image of the Netherlands
Confiscation	Not applicable	0
Confiscation	Unchanged	0
Confiscation	Intensify	30

Outcome cluster: Environmental protection

Instrument: Possession

		Environmental damage
Possession	Not applicable	0
Possession	Possession is illegal	0
Possession	User quantity is condoned	0
Possession	User quantity is legal, higher quantities condoned	0
Possession	Possession is legal	0

Instrument: Packaging

		Environmental damage
Packaging	Not applicable	0
Packaging	Plain packaging with prevention message	20
Packaging	Only plain packaging	10
Packaging	Only prevention message	10
Packaging	No restrictions	-20

Instrument: Advertising

		Environmental damage
Advertising	Not applicable	0
Advertising	Advertising not allowed	0
Advertising	Packaging advertising only	0
Advertising	Age limited advertising only	0
Advertising	Business to business advertising only	0
Advertising	Advertising allowed	0

Instrument: Sales across companies

		Environmental damage
Sales across companies	Not applicable	0
Sales across companies	Illegal	0
Sales across companies	Regulated	0
Sales across companies	Analogous to commodity law	0
Sales across companies	Analogous to pharmaceutical law	0
Sales across companies	Allowed	0

Instrument: Sales to consumers

		Environmental damage
Sales to consumers	Not applicable	0
Sales to consumers	Not allowed	0
Sales to consumers	Regulated	0
Sales to consumers	Analogous to commodity law	0
Sales to consumers	Analogous to pharmaceutical law	0
Sales to consumers	Allowed	0

Instrument: Age limit

		Environmental damage
Age limit	Not applicable	0
Age limit	No age limit	0
Age limit	Age limit is 18 years	0
Age limit	Age limit is higher than 18 years	0

Instrument: Punishable

		Environmental damage
Punishable	Not applicable	0
Punishable	Nobody punishable	0
Punishable	Seller punishable	0
Punishable	Seller and buyer punishable	0

Instrument: Legal requirements for selling

		Environmental damage
Legal requirements for selling	Not applicable	0
Legal requirements for selling	No license required	0
Legal requirements for selling	License required	0

Instrument: Price restrictions for consumer sale

		Environmental damage
Price restrictions for consumer sale	Not applicable	0
Price restrictions for consumer sale	No price restricions	0
Price restrictions for consumer sale	Minimum pricing	0

Instrument: MDMA quality rules (QA)

		Environmental damage
MDMA quality rules (QA)	Not applicable	0
MDMA quality rules (QA)	No quality definition/requirements	0
MDMA quality rules (QA)	Quality definition/requirements in place	50

Instrument: Sanctioning QA rules

		Environmental damage
Sanctioning QA rules	Not applicable	0
Sanctioning QA rules	None	-20
Sanctioning QA rules	Mild	30
Sanctioning QA rules	Severe	50

Instrument: Monitoring

		Environmental damage
Monitoring	Not applicable	0
Monitoring	None	0
Monitoring	Selective	20
Monitoring	Extensive	30

Instrument: Health promotion (HP) funding

		Environmental damage
Health promotion (HP) funding	Not applicable	0
Health promotion (HP) funding	No funding	0
Health promotion (HP) funding	Minimal funding	0
Health promotion (HP) funding	Substantial funding	0

Instrument: Control prevention policy

		Environmental damage
Control prevention policy	Not applicable	0
Control prevention policy	None	0
Control prevention policy	Weak	0
Control prevention policy	Strong	0

Instrument: Health promotion perspective

		Environmental damage
Health promotion perspective	Not applicable	0
Health promotion perspective	Abstinence	10
Health promotion perspective	Harm reduction	0

Instrument: Government responsible for HP

		Environmental damage
Government responsible for HP	Not applicable	0
Government responsible for HP	No government	0
Government responsible for HP	Regional government	0
Government responsible for HP	National government	0
Government responsible for HP	Both governments	0

Instrument: Production of MDMA

		Environmental damage
Production of MDMA	Not applicable	0
Production of MDMA	Illegal	0
Production of MDMA	Regulated	20
Production of MDMA	Analogous to commodity law	30
Production of MDMA	Analogous to pharmaceutical law	50
Production of MDMA	Allowed	10

Instrument: Export status

		Environmental damage
Export status	Not applicable	0
Export status	Export is illegal	0
Export status	Export is legal	60

Instrument: International strategy

		Environmental damage
International strategy	Not applicable	0
International strategy	Compliant	0
International strategy	Violation	0
International strategy	Based on condoning	0
International strategy	Inter se	0
International strategy	Exceptional position	0
International strategy	Adjustment of treaties	0

Instrument: Priority crime fighting

		Environmental damage
Priority crime fighting	Not applicable	0
Priority crime fighting	Low	-20
Priority crime fighting	Selective	0
Priority crime fighting	High	20

Instrument: Maximum penalty

		Environmental damage
Maximum penalty	Not applicable	0
Maximum penalty	Retain present sanction	0
Maximum penalty	More severe sanction	20

Instrument: Confiscation

		Environmental damage
Confiscation	Not applicable	0
Confiscation	Unchanged	0
Confiscation	Intensify	30

Outcome cluster: Ideological values

Instrument: Possession

		Policy consistent with liberal values	Policy consistent with conservative values
Possession	Not applicable	0	0
Possession	Possession is illegal	-100	100
Possession	User quantity is condoned	0	-30
Possession	User quantity is legal, higher quantities condoned	50	-90
Possession	Possession is legal	80	-100

Instrument: Packaging

		Policy consistent with liberal values	Policy consistent with conservative values
Packaging	Not applicable	0	0
Packaging	Plain packaging with prevention message	50	40
Packaging	Only plain packaging	30	0
Packaging	Only prevention message	70	-20
Packaging	No restrictions	-40	-80

Instrument: Advertising

		Policy consistent with liberal values	Policy consistent with conservative values
Advertising	Not applicable	0	0
Advertising	Advertising not allowed	-50	50
Advertising	Packaging advertising only	20	-20
Advertising	Age limited advertising only	60	-40
Advertising	Business to business advertising only	10	-10
Advertising	Advertising allowed	50	-50

Instrument: Sales across companies

		Policy consistent with liberal values	Policy consistent with conservative values
Sales across companies	Not applicable	0	0
Sales across companies	Illegal	-50	50
Sales across companies	Regulated	40	40
Sales across companies	Analogous to commodity law	0	-30
Sales across companies	Analogous to pharmaceutical law	-20	45
Sales across companies	Allowed	-10	-50

Instrument: Sales to consumers

		Policy consistent with liberal values	Policy consistent with conservative values
Sales to consumers	Not applicable	0	0
Sales to consumers	Not allowed	-100	100
Sales to consumers	Regulated	100	-20
Sales to consumers	Analogous to commodity law	50	-80
Sales to consumers	Analogous to pharmaceutical law	-20	0
Sales to consumers	Allowed	-20	-100

Instrument: Age limit

		Policy consistent with liberal values	Policy consistent with conservative values
Age limit	Not applicable	0	0
Age limit	No age limit	-50	-80
Age limit	Age limit is 18 years	70	30
Age limit	Age limit is higher than 18 years	-50	80

Instrument: Punishable

		Policy consistent with liberal values	Policy consistent with conservative values
Punishable	Not applicable	0	0
Punishable	Nobody punishable	-25	-25
Punishable	Seller punishable	70	50
Punishable	Seller and buyer punishable	-20	70

Instrument: Legal requirements for selling

		Policy consistent with liberal values	Policy consistent with conservative values
Legal requirements for selling	Not applicable	0	0
Legal requirements for selling	No license required	-30	-100
Legal requirements for selling	License required	70	70

Instrument: Price restrictions for consumer sale

		Policy consistent with liberal values	Policy consistent with conservative values
Price restrictions for consumer sale	Not applicable	0	0
Price restrictions for consumer sale	No price restricions	-40	-60
Price restrictions for consumer sale	Minimum pricing	20	40

Instrument: MDMA quality rules (QA)

		Policy consistent with liberal values	Policy consistent with conservative values
MDMA quality rules (QA)	Not applicable	0	0
MDMA quality rules (QA)	No quality definition/requirements	-50	-30
MDMA quality rules (QA)	Quality definition/requirements in place	50	30

Instrument: Sanctioning QA rules

		Policy consistent with liberal values	Policy consistent with conservative values
Sanctioning QA rules	Not applicable	0	0
Sanctioning QA rules	None	-50	-80
Sanctioning QA rules	Mild	20	-20
Sanctioning QA rules	Severe	50	60

Instrument: Monitoring

		Policy consistent with liberal values	Policy consistent with conservative values
Monitoring	Not applicable	0	0
Monitoring	None	-70	-70
Monitoring	Selective	40	70
Monitoring	Extensive	70	40

Instrument: Health promotion (HP) funding

		Policy consistent with liberal values	Policy consistent with conservative values
Health promotion (HP) funding	Not applicable	0	0
Health promotion (HP) funding	No funding	-20	20
Health promotion (HP) funding	Minimal funding	0	0
Health promotion (HP) funding	Substantial funding	20	-40

Instrument: Control prevention policy

		Policy consistent with liberal values	Policy consistent with conservative values
Control prevention policy	Not applicable	0	0
Control prevention policy	None	20	-20
Control prevention policy	Weak	-20	20
Control prevention policy	Strong	-50	50

Instrument: Health promotion perspective

		Policy consistent with liberal values	Policy consistent with conservative values
Health promotion perspective	Not applicable	0	0
Health promotion perspective	Abstinence	-80	100
Health promotion perspective	Harm reduction	80	-50

Instrument: Government responsible for HP

		Policy consistent with liberal values	Policy consistent with conservative values
Government responsible for HP	Not applicable	0	0
Government responsible for HP	No government	-100	-100
Government responsible for HP	Regional government	0	0
Government responsible for HP	National government	0	0
Government responsible for HP	Both governments	0	0

Instrument: Production of MDMA

		Policy consistent with liberal values	Policy consistent with conservative values
Production of MDMA	Not applicable	0	0
Production of MDMA	Illegal	-100	100
Production of MDMA	Regulated	100	20
Production of MDMA	Analogous to commodity law	60	-80
Production of MDMA	Analogous to pharmaceutical law	40	0
Production of MDMA	Allowed	-10	-100

Instrument: Export status

		Policy consistent with liberal values	Policy consistent with conservative values
Export status	Not applicable	0	0
Export status	Export is illegal	-30	30
Export status	Export is legal	30	-30

Instrument: International strategy

		Policy consistent with liberal values	Policy consistent with conservative values
International strategy	Not applicable	0	0
International strategy	Compliant	-50	70
International strategy	Violation	25	-70
International strategy	Based on condoning	40	-50
International strategy	Inter se	50	-40
International strategy	Exceptional position	60	-50
International strategy	Adjustment of treaties	70	-60

Instrument: Priority crime fighting

		Policy consistent with liberal values	Policy consistent with conservative values
Priority crime fighting	Not applicable	0	0
Priority crime fighting	Low	-20	-60
Priority crime fighting	Selective	50	-40
Priority crime fighting	High	-40	80

Instrument: Maximum penalty

		Policy consistent with liberal values	Policy consistent with conservative values
Maximum penalty	Not applicable	0	0
Maximum penalty	Retain present sanction	0	0
Maximum penalty	More severe sanction	-20	60

Instrument: Confiscation

		Policy consistent with liberal values	Policy consistent with conservative values
Confiscation	Not applicable	0	0
Confiscation	Unchanged	0	0
Confiscation	Intensify	20	80

Predefined policy models

The think tank predefined a number of policy models.

for (scenario_id in c("coffeeshop",
                      "adapted_coffeeshop",
                      "free_market",
                      "repression")) {
  
  cat("\n\n#### ", scenarioLabels[scenario_id], "\n\n");

  cat(
    mdmcda::scenario_alternative_table(
      scenarioDefinition = scenarioDefinitions[[scenario_id]],
      alternativeLabels = alternativeLabels,
      decisionOrder = decisionOrder,
      decisionLabels = decisionLabels
    )
  );
}

Coffeeshop model

Decision	Alternative
Possession	User quantity is condoned
Packaging	Plain packaging with prevention message
Advertising	Advertising not allowed
Sales across companies	Illegal
Sales to consumers	Regulated
Age limit	Age limit is 18 years
Punishable	Nobody punishable
Legal requirements for selling	License required
Price restrictions for consumer sale	No price restricions
MDMA quality rules (QA)	No quality definition/requirements
Sanctioning QA rules	Not applicable
Monitoring	Extensive
Health promotion (HP) funding	Substantial funding
Control prevention policy	Weak
Health promotion perspective	Harm reduction
Government responsible for HP	Both governments
Production of MDMA	Illegal
Export status	Export is illegal
International strategy	Based on condoning
Priority crime fighting	High
Maximum penalty	Retain present sanction
Confiscation	Unchanged

Adapted coffeeshop model

Decision	Alternative
Possession	User quantity is condoned
Packaging	Plain packaging with prevention message
Advertising	Advertising not allowed
Sales across companies	Regulated
Sales to consumers	Analogous to pharmaceutical law
Age limit	Age limit is 18 years
Punishable	Nobody punishable
Legal requirements for selling	License required
Price restrictions for consumer sale	Minimum pricing
MDMA quality rules (QA)	Quality definition/requirements in place
Sanctioning QA rules	Mild
Monitoring	Extensive
Health promotion (HP) funding	Substantial funding
Control prevention policy	None
Health promotion perspective	Harm reduction
Government responsible for HP	Both governments
Production of MDMA	Analogous to pharmaceutical law
Export status	Export is illegal
International strategy	Based on condoning
Priority crime fighting	High
Maximum penalty	More severe sanction
Confiscation	Intensify

Free market model

Decision	Alternative
Possession	Possession is legal
Packaging	No restrictions
Advertising	Advertising allowed
Sales across companies	Allowed
Sales to consumers	Allowed
Age limit	Not applicable
Punishable	Nobody punishable
Legal requirements for selling	No license required
Price restrictions for consumer sale	No price restricions
MDMA quality rules (QA)	No quality definition/requirements
Sanctioning QA rules	None
Monitoring	None
Health promotion (HP) funding	No funding
Control prevention policy	Strong
Health promotion perspective	Not applicable
Government responsible for HP	No government
Production of MDMA	Allowed
Export status	Export is legal
International strategy	Exceptional position
Priority crime fighting	Selective
Maximum penalty	Retain present sanction
Confiscation	Unchanged

Repression model

Decision	Alternative
Possession	Possession is illegal
Packaging	Not applicable
Advertising	Advertising not allowed
Sales across companies	Illegal
Sales to consumers	Not applicable
Age limit	Age limit is 18 years
Punishable	Not applicable
Legal requirements for selling	Not applicable
Price restrictions for consumer sale	Not applicable
MDMA quality rules (QA)	Not applicable
Sanctioning QA rules	Not applicable
Monitoring	Extensive
Health promotion (HP) funding	No funding
Control prevention policy	Strong
Health promotion perspective	Abstinence
Government responsible for HP	No government
Production of MDMA	Illegal
Export status	Export is illegal
International strategy	Compliant
Priority crime fighting	High
Maximum penalty	More severe sanction
Confiscation	Intensify

Final weights

cat("\n\n<div style='width: 70%'>\n\n");

cat(
  mdmcda::plot_criteria(
    criteria,
    labels = criterionLabels,
    renderGraph = FALSE,
    returnSVG = TRUE,
    outputFile = file.path(workingPath,
                           "criteria-tree-with-weights.pdf")
  )
);

cat("\n\n</div>\n\n");

Results

These three tabs contain the main results. Note that additional details are available in the “Additional visualisations” tab.

Performance of the models

Plots

if (producePlots) {

  mainResultsFigure <-
    mdmcda::scoreBarchart_scenariosByCluster(
      weightedEstimates = weightedEstimates,
      estimateCol = weightedEstimateName,
      parentCriterionOrder = parentCriterionOrder,
      parentCriterionLabels = criterionLabels,
      scenarioOrder = scenarioOrder,
      scenarioLabels = wrappedScenarioLabels
      ### , sortByScore = TRUE                ### Uncomment to sort by score
    ) +
    ggplot2::labs(title = NULL,
                  x = NULL,
                  y = "Total score (sum of weighted estimates)") +
    ggplot2::guides(fill = ggplot2::guide_legend(nrow = 2)) +
    ggplot2::theme(
      plot.margin = ggplot2::margin(l=.5, r=.5, t=0, b=.5, unit = "line"),
      panel.grid.major.x = ggplot2::element_blank(),
      panel.grid.minor.x = ggplot2::element_blank(),
      panel.grid.major.y = ggplot2::element_line(size=.1, color="grey")
    );
  
  ufs::knitAndSave(mainResultsFigure,
                   path = workingPath,
                   figWidth = 10,
                   figHeight = 5,
                   dpi = 600,
                   figCaption = "Main results figure (English version)");

  mainResultsFigure_NL <-
    mdmcda::scoreBarchart_scenariosByCluster(
      weightedEstimates = weightedEstimates,
      estimateCol = weightedEstimateName,
      parentCriterionOrder = parentCriterionOrder,
      parentCriterionLabels = criterionLabels_NL,
      scenarioOrder = scenarioOrder,
      scenarioLabels = wrappedScenarioLabels_NL
      ### , sortByScore = TRUE                ### Uncomment to sort by score
    ) +
    ggplot2::labs(title = NULL,
                  x = NULL,
                  y = "Totaalscore (som van gewogen schattingen)") +
    ggplot2::guides(fill = ggplot2::guide_legend(nrow = 2)) +
    ggplot2::theme(
      plot.margin = ggplot2::margin(l=.5, r=.5, t=0, b=.5, unit = "line"),
      panel.grid.major.x = ggplot2::element_blank(),
      panel.grid.minor.x = ggplot2::element_blank(),
      panel.grid.major.y = ggplot2::element_line(size=.1, color="grey")
    );
  
  ufs::knitAndSave(mainResultsFigure_NL,
                   path = workingPath,
                   figWidth = 10,
                   figHeight = 5,
                   dpi = 600,
                   figCaption = "Main results figure (Dutch version)");
  
}

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning in sprintf(suffix, counter): one argument not used by format ''

Figure 28: Main results figure (English version).

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 29: Main results figure (Dutch version).

Tables

###-----------------------------------------------------------------------------
### Scores per cluster and per scenario
###-----------------------------------------------------------------------------

clusterScores <-
  mdmcda::criteriaCluster_df(
    weightedEstimates = weightedEstimates,
    estimateCol = weightedEstimateName,
    parentCriterionOrder = parentCriterionOrder,
    parentCriterionLabels = criterionLabels_NL,
    scenarioOrder = scenarioOrder,
    scenarioLabels = scenarioLabels_NL
  );

for (i in scenarioOrder) {
  
  mdmcda::cat0("\n\n#### ", scenarioLabels_NL[i], "\n\n");
  
  print(
    kableExtra::kable_styling(
      knitr::kable(
        clusterScores[clusterScores$scenario_id==i,
                      c('cluster_label',
                        'meanWeights_weighted_estimate')],
        row.names=FALSE,
        col.names=c("Cluster", "Score")
      )
    )
  );

}

Hoogst mogelijke scores

Cluster	Score
Criminaliteit	4222.3637
Ideologische waarden	0.0000
Financiele kosten en baten	1811.9481
Milieubescherming	1373.6109
Internationale politiek	-554.8366
Gebruik (prevalentie & patronen)	657.7565
Gezondheid van de gebruiker	5758.9394

X-Shop model

Cluster	Score
Criminaliteit	3814.2134
Ideologische waarden	0.0000
Financiele kosten en baten	1617.5156
Milieubescherming	1107.7508
Internationale politiek	-264.8586
Gebruik (prevalentie & patronen)	699.6037
Gezondheid van de gebruiker	5725.0663

Gesloten cannabisketen model

Cluster	Score
Criminaliteit	3384.9998
Ideologische waarden	0.0000
Financiele kosten en baten	1449.6162
Milieubescherming	1107.7508
Internationale politiek	-200.3435
Gebruik (prevalentie & patronen)	816.9417
Gezondheid van de gebruiker	4161.7869

Coffeeshop model

Cluster	Score
Criminaliteit	2248.8003
Ideologische waarden	0.0000
Financiele kosten en baten	745.4951
Milieubescherming	310.1702
Internationale politiek	-168.7658
Gebruik (prevalentie & patronen)	551.5716
Gezondheid van de gebruiker	1840.4343

Vrije markt model

Cluster	Score
Criminaliteit	1852.9857
Ideologische waarden	0.0000
Financiele kosten en baten	956.0448
Milieubescherming	132.9301
Internationale politiek	-1064.5670
Gebruik (prevalentie & patronen)	-2241.8658
Gezondheid van de gebruiker	-1879.1651

Repressie model

Cluster	Score
Criminaliteit	245.836069
Ideologische waarden	0.000000
Financiele kosten en baten	-496.170673
Milieubescherming	487.410332
Internationale politiek	87.493071
Gebruik (prevalentie & patronen)	4.885243
Gezondheid van de gebruiker	-3107.784375

Laagst mogelijke scores

Cluster	Score
Criminaliteit	-1229.0828
Ideologische waarden	0.0000
Financiele kosten en baten	100.2161
Milieubescherming	0.0000
Internationale politiek	-455.2087
Gebruik (prevalentie & patronen)	-949.5807
Gezondheid van de gebruiker	-4718.4598

Performance of each alternative

Plots in detail

if (producePlots) {

  alternativePerformancePlots <- list();

  for (currentDecision in decisionOrder) {
    
    ufs::heading(decisionLabels[currentDecision],
                 " {.tabset .tabset-pills}",
                 headingLevel = 5);
    
    alternativePerformancePlots[[currentDecision]] <- list();
    
    for (currentAlternative in names(alternativeLabels[[currentDecision]])) {
      
      ufs::heading(alternativeLabels[[currentDecision]][[currentAlternative]],
                   headingLevel = 6);
      
      alternativePerformancePlots[[currentDecision]][[currentAlternative]] <-
        
        mdmcda::scoreBarchart_criteria_for_singleDecision(
          multiEstimateDf = estimates$multiEstimateDf,
          estimateCol=weightedEstimateName,
          decision_id=currentDecision,
          alternative_value = currentAlternative,
          criterionOrder = criterionOrder,
          criterionLabels = criterionLabels,
          parentCriterionOrder = parentCriterionOrder,
          parentCriterionIds_by_childId = criteria$convenience$parentCriterionIds_by_childId,
          decisionLabels = decisionLabels,
          alternativeLabels = alternativeLabels
        );
      
      ufs::knitAndSave(
        alternativePerformancePlots[[currentDecision]][[currentAlternative]] +
          ggplot2::theme(
            plot.margin = ggplot2::margin(l=4, r=1, t=1, b=1,
                                          unit = "line")
          ),
        path=workingPath,
        figCaption = paste0(
          "Performance for alternative ",
          decisionLabels[currentDecision], ": ",
          alternativeLabels[[currentDecision]][[currentAlternative]]
        )
      );
      
    }
  }

}

Possession

Not applicable

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning in sprintf(suffix, counter): one argument not used by format ''

Figure 30: Performance for alternative Possession: Not applicable.

Possession is illegal

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 31: Performance for alternative Possession: Possession is illegal.

User quantity is condoned

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

Figure 32: Performance for alternative Possession: User quantity is condoned.

User quantity is legal, higher quantities condoned

Warning: Using ragg device as default. Ignoring `type` and `antialias` arguments

Warning: one argument not used by format ''

MDMA Policy MDMCDA (Multi Decision Multi Criteria Decision Analysis)

MDMA Policy Think Tank

14:42:46 on 2021-07-04 CEST (UTC+0200)

MDMA MDMCDA

Overview

App and website

mdmcda R package

Background

The objective phase in MDMCDA: expert input

The subjective phase in MDMCDA: scenarios and weights

Characteristics of MDMCDA

Method

Reading and preprocessing

Reading data

Criteria (outcomes of chosen policy)

Estimates

Estimate post-sets

Weights

Scenarios (“policy models”)

Labels

Preparing data

Correcting weights

Apply post-sets

Correct alternative values for possession status

Weighting

Input

Outcomes and anchors

Outcome tree

Anchors of each outcome

Use frequency and intensity by users

Use in vulnerable populations

Prevalence (general population)

Quality of information about MDMA

Health damage

Stigmatization of users

Social benefits

Health benefits

Social damage

Shift to other drugs

Organized crime (MDMA-related)

Criminalisation of users

MDMA-related small crime

Organized crime (not MDMA-related)

Criminals exploitating vulnerable groups

International trafficking of MDMA

State revenue (other)

State revenue (VAT)

Costs related to pollution (dumpings)

Costs related to health damage

Costs related to crime

Economic boycotts

International image of the Netherlands

International legal countermeasures

Environmental damage

Policy consistent with conservative values

Policy consistent with liberal values

Performance (sub)tables

Outcome cluster: Use (prevalence & patterns)

Instrument: Possession

Instrument: Packaging

Instrument: Advertising

Instrument: Sales across companies

Instrument: Sales to consumers

Instrument: Age limit

Instrument: Punishable

Instrument: Legal requirements for selling

Instrument: Price restrictions for consumer sale

Instrument: MDMA quality rules (QA)

Instrument: Sanctioning QA rules

Instrument: Monitoring

Instrument: Health promotion (HP) funding

Instrument: Control prevention policy

Instrument: Health promotion perspective

Instrument: Government responsible for HP

Instrument: Production of MDMA

Instrument: Export status

Instrument: International strategy

Instrument: Priority crime fighting

Instrument: Maximum penalty

Instrument: Confiscation

`mdmcda` R package