Getting a fix on figures

aug14Just how closely linked are drug misuse and acquisitive crime statistics?

‘Drinking sensibly, never doing drugs – is this the age of the young puritan?’ asked the Guardian last month. ‘Why drugs are no longer cool: teenagers are internet addicts while their parents snort cocaine,’ offered the Telegraph.

The Health and Social Care Information Centre’s latest figures on falling rates of drug and drink consumption among secondary school pupils (see news story, page 5) made national headlines, just the latest of countless reports over the last few years that seem to confirm that younger people are slowly turning away from drugs – or older-established drugs anyway. While services can struggle to keep up with the growing list of new psychoactive substances, as well as higher rates of image and performance-enhancing drug use, problem heroin and crack use does seem to be increasingly confined to an older, entrenched population, many of whom started using in the ‘heroin epidemic’ of the 1980s when new supply routes meant more drugs coming into the country at a time of mass unemployment.

It’s no secret that the dramatic increases in funding for drug services that came with the advent of the NTA was in part driven by a desire to keep a lid on crime figures, but few attempts have been made to properly map out the link between acquisitive crime rates and problem drug use until the Home Office’s new report, The heroin epide­mic of the 1980s and 1990s and its effect on crime trends – then and now.

According to the document, the national peak of the epidemic was ‘probably’ between 1993 and 2000, while crime peaked between 1993 and 1995. As well as a comprehensive review of existing research literature, the report used police force area-level comparisons of the ‘Addicts Index’ and recorded crime data from 1981 to 1996, alongside modelling the number of heroin/crack users and their offending. 

The police area comparisons showed that ‘different types of theft generally peaked together within an area’ but ‘the timing and size of these peaks varied across areas and was highly correlated with heroin use’, concluding that about 40 per cent of the national rise in the highest-volume crime types – such as burglary and theft from vehicles – from 1981 to the peak could be ‘attributed to rises in the number of heroin users’. The modeling approach, meanwhile, found that ‘heroin/crack use could account for at least half of the rise in acquisitive crime in England and Wales to 1995 and between a quarter and a third of the fall to 2012, as the ‘epidemic cohort aged, received treatment, quit illicit drug use or died’.

However, the document makes it clear that – despite the wide body of evidence drawn together by researchers – the ‘hidden’ nature of the study group ‘means that robust data remain sparse’. The paper is also careful to stress that other factors – most obviously unemployment – also play a significant role, and while peaks in acquisitive crime levels matched the timescales of heroin epidemics in England, the US and parts of Europe, there were also regional exceptions.

The lack of high-quality data means that the fundamental questions of whether opiates/crack caused the crimes committed by the people taking them, and whether the peaks in drug use correlated with peaks in crime may never be answered ‘definitively’, it states. However, ‘on causality, the evidence gathered here shows that opiate/crack use almost certainly generated additional offences, but quantifying this precisely remains challenging’.

The best summary, it concludes, is to demonstrate the existence of an ‘epidemic narrative’ that fits the facts. This is that drug epidemics produced a cohort of users, and a steady rise in crime during the 1980s – during which most of England and Wales ‘remained relatively unaffected by the epidemic’ – then increased ‘very rapidly in the 1990s as every police force area except Merseyside reached its peak of opiate/crack use’. Then, once ‘all susceptible individuals had been “exposed”, the number of new users probably decreased just as quickly as it had risen’ and crime fell – at first quickly as ‘less-recalcitrant’ users quit in large numbers, and then more slowly.

‘The cohort was not homogeneous,’ it states. ‘Many (perhaps most) did not become either long-term addicted or prolific criminals and some were offenders before using opiates or crack. While many probably had the clustering of crime risk factors that c

ould have marked them out for a criminal career in the absence of the epidemic, the cohort probably also included a number of individuals whose only crime risk factor was a suscep­tibility to peer influence at a time when heroin use was spreading in their area. For the first group, heroin use may have accelerated and extended an existing criminal career and for some of the second group heroin may have kick-started a criminal career.’

Perhaps mindful of the potential media reaction, the paper also clearly spells out that the impact was on crime volumes rather than overall harm, which is largely driven by violent and sexual offences. ‘The most important caveat though, is that this narrative does not imply that opiate/crack use was the sole factor driving crime trends,’ it states. ‘Many factors are likely to have been important and interactions may also be crucial.’

Among the policy implications, it says, are that as the number of heroin and crack users continues to fall, it will continue to be at a relatively slow pace, as many older users will have been in and out of services for years, and ‘focusing resources on the most important individuals may be the key.’

The other main policy conclusion, despite shifting drug trends, remains the  importance of preventing a future epidemic, it stresses. ‘Evidence shows epidemics do not strike all areas simultaneously and there is a lag between epidemic start and the moment it becomes visible on treat­ment or criminal justice datasets. Local-level monitoring is therefore crucial.’

Report at

We value your input. Please leave a comment, you do not need an account to do this but comments will be moderated before they are displayed...