'Drug misuse: findings from the 2012 to 2013 CSEW' was published a couple of days ago. CSEW stands for Crime Survey England & Wales, the more accurate name for what used to be called the British Crime Survey (BCS).
The main purpose of the CSEW is to collect experiences of crime that don’t get into the reported crime figures as a means of dispelling all the stories about unreported crimes making the official police figures unreliable. At the same time, it asks about perceptions of crime so it can compare what people think is happening with what is actually happening.
The CSEW is done by face to face interviews and, since 1996, there has been an additional optional survey on drug use done by the participant entering information directly into the surveyor’s computer to preserve anonymity.
The key findings from the latest survey, if you only look at the media coverage and government statements, are that drug use has declined since 1996 and that a lot of people use nitrous oxide (a ‘legal high’ but not a ‘NPS’ (new psychoactive substance) because it has been around a very long time). This is the first time nitrous oxide has been included and I’m not sure if the idea was to try and get an indication of how many people are willing to use psychoactive substances as long as doing so does not break the law.
The first thing to say about the CSEW is that it is wrong. It has always been wrong. This is very easy to demonstrate. If you go to the data and look at Table Ext_03 'Estimates of numbers of illicit drug users, 16 to 59 year olds', you will see that the CSEW reports that there are 27,000 people in England and Wales who used heroin in the past year. If you add methadone and go for the top of the estimate range you get to 86,000. But the most recent figures from the National Drug Treatment Monitoring System (NDTMS) show that just under 160,000 people were ‘clients’ of the treatment services because of their use of opiates (Table 2.3.1, page 9) So, the CSEW is, is very wrong when it comes to opiate use. It is more wrong than that at first appears because the 86,000 will include people who used an opiate just once in the last year and they are not the sort of problem users who seek treatment.
There is a fuller explanation of the flaws in the CSEW in this previous blog so I won’t repeat them.
The important question, though, is whether the CSEW (and the BCS before it) has been consistently wrong throughout its life. If the errors are consistent, it may be useful as a guide to trends in substance use amongst its narrow survey group. That is significant, this year, because the government latched onto the downward trend reported by the survey. The Home Office minister responsible for drugs, Jeremy Browne, called it "really positive news" but Browne is a LibDem whose party officially favours a change in drug laws so he has to walk a tightrope between his coalition duties and his party loyalty. I have no doubt that the Home Secretary, Theresa May, and the Prime Minister will, at some point, cite these figures as proof that current policy works.
Actually rather than talking about a longer term downward trend the talk has been of the year on year fall because, of course, the government wants to claim credit for what it has done since 2010. Both the 16-59 and 16-24 lines in the report show falls over the past couple of years but, of course, the 16-59 line includes 16-24 year olds. Separating out just the 25-59 year olds is quite complex. Figure 3.1 on page 18 gives a breakdown into seven age bands but I found that too confusing to be useful. I used the CIA World Factbook latest estimates of population to get at the approximate split into just two age bands. Then I’ve used total population numbers for each year since 1996 (taken with some trepidation from Wikipedia) and applied the same proportions throughout to get the approximate age splits for each year.
I plotted ‘any drug use’ in the past year adding my 25-59 estimate to the 16-24 and 16-59 figures from the report.
Then I did the same for any Class ‘A’ use in the past year
Then, any cocaine use in the last year
In each case, there has been no real change since 2009/10 for the 25-59 year old group.
But, most significantly, I then took any drug use other than cannabis and found
That suggests that any trend is drug use since 1996 is almost entirely the result of changes in the use of Cannabis sativa, marijuana, by people in the 16-24 age band.
And a plot of cannabis alone for the three age bands shows this.
I’m not as good a mathematician as I’d like to be so I’d be very happy if someone tried the same exercise. I’ll gladly send the spreadsheet I worked on if anyone wants to contact me.
I don’t want to over analyse the above given that I’ve used approximations for the population splits and, as I said before, the CSEW is just plain wrong. What seems clear, however, is that the figures show a small upward trend in the 25-59 age range.
I still do not know what this report means but I am absolutely clear about what it doesn’t mean. It doesn’t mean current policy is working.
You can send comments via the contact page but please be sure to say what blog entry you are commenting on.