Tuesday, March 06, 2007

Lies, Damned Lies, Statistics, and 'Prohibition Works!'

Those attempting to defend the status quo to the media, usually those working in government or enforcement, frequently quote statistics that give the misleading impression that prohibition is working – when the exact opposite is true.

It is hard to think of another area of social policy where the waters are so muddied by statistical chicanery. This sort of misleading evidence, attempting to dress up failure as success, usually comes in one of five varieties. Below is Transform's handy pocket guide to identifying and challenging them:

1. Localised success – (cherry-picking part 1)

Example: “cocaine/heroin production in Colombia/ Afghanistan has fallen this year.”

These stats may well be true (they may not, but let’s assume they are). However, crucially, local production is irrelevant in a global market, as falls in production in one region will quickly be made up by rises in another. This pattern has been observed repeatedly in regional shifts in production of coca, opium and cannabis – so frequently that it has become known in official shorthand as ‘the balloon effect’ (if you squeeze a balloon on one side, it expands on the other). The key point here is that the trend in global production has always kept pace with global demand, which has risen steadily over the past 4 decades (see: why prohibition can never work p….). Illegal drug markets are not confined by geographical boundaries, and localised successes should not be allowed to disguise larger scale systematic failure to control global production. This is the worst form of cherry-picking. Keep the focus on the bigger picture – using official national and international statistics that are not in dispute .

2. Short term success – (cherry-picking part 2)

Example: “street drug dealing fell by 10% in the last 6 months in Birmingham”.

Again this may well be true – but short-term changes often mask longer-term trends. They can also be due to (non-policy related) external factors, changes in statistical collection or methodology, and sometimes a marginal change can be within statistical error parameters. This sort of cherry picking can also be countered by bringing the focus back to the bigger picture statistics on the failure of the policy nationally and internationally (be careful to make sure the blame is on the policy makers, not those who are implementing it – the police do their job as best they can, it just happens to be an impossible one). Also remind policy makers that it is the policy of prohibition that created the crime and illegal markets in the first place.

3. Process success

Examples: “We have set up a new agency, appointed a new Tsar, instigated a partnership project with Jamaican police, invested millions in a, b and c, announced ambitious new targets on x, y and z” etc. etc

These are an age-old exercise in distraction. Policy must be judged on outcomes, not inputs or process indicators. Challenge policy makers on their record: the outcomes of the policies they are supporting. Don’t let them get away with announcing yet more headline-grabbing new (sp)initiatives. Have the new changes made any difference to the bigger picture on supply, availability, crime, problematic use? The problems with prohibition are fundamental and cannot be solved with superficial tweaks to policy which at best will marginally reduce the harms created by the policy in the first place, and more likely will cost government and taxpayers more money for no results. This policy is essentially unchanged for two generations.

4. Success relative to previous disasters

Example: “crack use has fallen since last year”

When compared to a policy as disastrous as heavy-handed enforcement and large-scale incarceration, almost any change in intervention will start to look like progress. A good example is the improved outcomes from coercing drug-using offenders into abstinence-based ‘treatment’ as opposed to sending them to jail. The point here is that imprisonment is so expensive and counterproductive that literally any alternative would produce better results – burning the money, giving offenders juggling lessons, ANYTHING.

The crack example can also illustrate the important point that drugs come in and out of fashion largely independently of policy and law. Prevalence of one drug may fall after an epidemic (crack use in the US being a recent example) whilst another simultaneously rises (in the US this has been methamphetamine). Again the way to avoid this is to focus on the longer-term bigger picture –overall drug use has risen steadily for decades – especially of the most problematic drugs. If a ‘stabilisation’ has been ‘achieved’, this may be sold as a success but most likely simply reflects a saturated market demand. The UK government has, for example, been claiming success in the stabilisation of heroin use in the UK over the past 4 or 5 years – It needs to be pointed out that it has stabilised at the highest level in UK history, the highest level in Europe, and a level approximately 1000% higher than in 1971.

5. Success on completely meaningless indicators

Examples: ‘volume of drug seizures is up’, ‘number of dealers jailed has increased!’, ‘ we have ‘smashed’ record numbers of drug gangs’ etc.

These are measures that reflect the level of expenditure on enforcement and the size of the illegal market. They rarely, if ever, translate into the policy outputs that prohibition is striving for – i.e. reduced drug production, supply and use (let alone reduced harm). They sound great in the media; catching baddies, intercepting nasty drugs etc – but it gives the misleading impression of success when in reality the opposite is true. Again, challenge people using these sorts of statistics to show what impact they are having on meaningful indicators and keep to the bigger picture. Do not let statements from officials such as talking about ‘x quantities of drugs prevented from reaching the streets’ go unchallenged when they fully aware that such seizures have no impact on overall supply and that drugs are cheaper and more available than ever - a fact not disputed by the Home Office.

This guide is an extract from Transform's forthcoming guide: 'After the War on Drugs, Tools for the Debate'

For some further discussion on the dubious science of prohibition see Ben Goldacre's bad science Guardian column on Prohibition vs the Gold Standard :

Why are drugs such a bad science magnet? Partly, of course, it’s the moral panic. But more than that, sat squarely at the heart of our discourse on drugs, is one fabulously reductionist notion: it is the idea that a complex web of social, moral, criminal, health, and political problems can be simplified to, blamed on, or treated via a molecule or a plant. You’d have a job keeping that idea afloat.

1 comment:

chrisbx515 said...

I have noticed that even when sensible evidence based facts are used in blogs and debates that the level of understanding of the issues as a whole are very low. Unless we can get the knowlegde to misguided or uninformed individuals then the government will be able to continue in any way it wants to which will not be to even come close to ending prohibition, why should they we have not given them enough reasons to ....yet.