It is very difficult to estimate the size of any market. Even something like the number of new cars registered is not an absolute indicator of the number of new cars purchased because of the practice of ‘pre-registering’ by dealers. If you can’t be completely certain about market size with a product where every sale has to reported to a regulatory body then estimating the size of the market for illegal products is, clearly, open to many errors.
Nonetheless, the drug use part of the Crime Survey for England & Wales (CSEW) (still, sometimes, referred to as the British Crime Survey (BCS)) is still used by many people and organisations as the basis for policy planning and implementation.
I’ve written before about the flaws in this survey’s results and now a new report from the USA attempting to get a better estimate of drug use has shown up the flaws in the US equivalent of the CSEW and, thus, enables me to write my second ‘I told you so’ piece in two days.
A major difference between the US and the UK is that, whereas it is left to amateurs like me to critique the data relied on by the UK government, in the USA, it was the Office of National Drug Control Policy (ONDCP) that realised the estimates it had based on survey data were flawed and, therefore, commissioned a report from the RAND Corporation to try and improve the quality of the information it uses.
The summary of what I wrote in July 2012 is that there is a very obvious difference between the number of users of heroin reported by the CSEW and the much larger number of people seeking treatment for problems with heroin use.
The new report ‘What America’s Users Spend on Illegal Drugs: 2000-2010’ takes results of testing on people arrested throughout the USA and works back from that to estimate population level usage. I’m not qualified to go into whether this procedure, which is much more complex than you’ll think from my simple distillation, is any more accurate but it certainly produces some very different results.
This piece in USA Today co-authored by Beau Kilmer, Co-director of RAND Drug Policy Research Center, points out that the household survey run by the National Survey on Drug Use and Health (NSDUH) finds only about 60,000 daily or near daily users of heroin whereas the RAND report puts the number at closer to 1 million.
The methodology is more fully described by Jacob Sullum in Forbes who calls it an ‘impressive headache-inducing feat, but one that is subject to “great uncertainty”’.
There is always a lot to take away from lengthy reports on drug use and, it seems, it is not just different people who will focus on different aspects. In USA Today, Kilmer’s piece (co-authored by Jonathan Caulkins) is headlined ‘Hard drugs demand solid understanding’ but a week later, again in Forbes and this time as the single author, Kilmer writes under the headline ‘U.S. Cocaine Consumption Declines by Half, While Marijuana Use Jumps’.
Since cannabis use is at the forefront of drug policy considerations in the USA at the moment, it is interesting to look at that claim in a little more detail. The RAND report points out that its figures for cannabis use are much more closely based on the NSDUH because, it claims, people are more willing to disclose marijuana use than other drugs.
I have two problems with the reports results for cannabis use. First, it provides figures from 2000 to 2010 but states, in a note under figure 5.2 on page 50, that ‘The 2000–2003 marijuana estimates are not perfectly comparable to the later years because of changes in survey questions and methods’. Given that the total use figures show a doubling from 2000 to 2010 I have concerns that including figures that are incomparable is simply asking for misleading claims to be made.
My second problem is that, as far as I can tell, the report applies a consistent factor to adjust the NSDUH figures for under-reporting but it is very clear that attitudes to cannabis have changed considerably even in the last five years that Kilmer focuses on so, I would guess, there has been a change in the willingness of survey participants to disclose their use of cannabis. Though these surveys are set up to be completely anonymous, you’d have to be naïve not to assume that many users have little trust in the claims of anyone in authority.
This, for me, is the single most important aspect of attempts to quantify the prevalence of cannabis use. If the apparent increase in prevalence is actually a reflection of an increased willingness to declare use then, as decriminalisation/regulation/legalisation progresses, there will be an increase in prevalence as a result the removal of the perception that the survey cannot be trusted to remain anonymous.
When the backlash to more realistic approaches to cannabis use comes it will rely on ‘evidence’ that the new policies have brought about a surge in consumption. But, at the same time, since I’m guessing about the effect of attitude to willingness to be truthful in surveys, it would be equally wrong for reformers to dismiss any increase in use as an artefact of the surveying process.
Regardless of the great amount of work that has gone into this report, it remains a simple fact that estimates of drug use, whether heroin, cannabis or a NPS, are inherently wrong and anyone making claims based on representing such estimates as cold hard fact should be viewed with suspicion.
I can’t close without mentioning that, as the RAND report's title states, this is an attempt to quantify the sales value of the drug market in the USA. For me, estimating the number of users of a substance as a basis for estimating the numbers who will find their use problematic and require support is one thing. And, for cannabis, any under-reporting of prevalence is more than offset by the over-perception of the numbers of users whose use becomes problematic making the survey flaws less of a worry in terms of actual support. Estimating the monetary value of substance use, however, looks like an attempt to protect enforcement budgets.
I fully expect prohibitionists and drug aunts to misquote the report’s findings to support their entrenched positions. And I fully expect to be back to say ‘I told you so’ for a third time, soon.
You can send comments via the contact page but please be sure to say what blog entry you are commenting on.