Quick Links: Ecstasy Statistics Neurotoxicity The User's Guide MDMA Technical FAQ
Strange Days at Monitoring the Future. (1/02/04)
(For more complete/recent statistics, visit MDMA Statistics.)
Is 'ecstasy' more dangerous than smoking crack? Today's teenagers apparently think so, according to the latest (2003) results of the US government funded Monitoring the Future Survey. Still, that was just an appetizer to the main event: A claim of huge reductions in 'ecstasy' use by high school students.
Usage levels have clearly declined from the 2000 spike, as confirmed by a modest reduction in 'ecstasy-related' emergency room visits in 2002. But the claimed giant downward dip in usage for 2003 seems a little...optimistic.
Monitoring the Future surveys students in 8th, 10th, and 12th grades each year. As a result, if you wait two years, you can check how the habits of a given age group have changed (what was the 10th grade group in 2001 became the 12th grade group for 2003.) So, I followed the data, and found something rather odd.
The below graph shows the number of new (first time) 'ecstasy' users for a given two-year period:
This leaves us with just one small question. Do we believe that the number of young people willing to try 'ecstasy' dropped by 90% in one year? There is no doubt that overall levels of use have declined, but this sudden absolute stop to young people experimenting with 'ecstasy' seems...well, a little too good to be believed. It's not, however, without precedent.
The Glue Sniffing Gap:
Examining the same type of shift for inhalants (solvents), we find something even stranger:
No, the graph isn't facing the wrong way. In the mid 90's, students suddenly became far less willing to admit to abusing inhalants as they grew older and got more anti-drug 'education'. The effect was so profound that for every two years older the students became, about a fourth of the students who had originally admitted to abusing inhalants would now claim to have never abused inhalants. I can't really blame them (even the 'druggies' make fun of the glue sniffers) but it brings up a fundamental question: What exactly is it that the Monitoring the Future data represents?
One of those hard lessons gained in research is that data means what it means. That is to say, if 26% of your sample group says they like being spanked with a wiffle-bat, it means that 26% of your sample group say that they like it. They aren't necessarily giving thoughtful or even honest answers. Researchers in human sexuality have long known that people lie about their sexual practices; in particular, women tend to under-report the number of sexual partners they've had when asked in interviews (but come clean when a polygraph ('lie detector') is used.)
What does this have to do with drug use among young people? While the number of people you've slept with might be embarrassing, drug use is criminalized and grounds for arrest, expulsion, revocation of college financial aid, etc. The result is a presumptive hesitancy to admit to drug use when the MTF people come to call. (I actually took part in an MTF survey years ago, and in spite of my best intentions, with classmates sitting next to me and teachers watching it was difficult to be honest.)
In particular, in the past few years anti-'ecstasy' propaganda has paid off for the government in the form of increasingly anti-'ecstasy' sentiments from students:
The most drug-war friendly interpretation of this correlation between increased anti-ecstasy indoctrination by the State and decreased reported use is that the 'education' efforts have succeeded in their goal of reducing use. To some extent that is no doubt the case, but given the suddenness and magnitude of the shift, coupled with the nonsensical changes in reported inhalant abuse over time, it seems very likely that perceived acceptability of using a given drug has a significant, even quite large impact on the number of students who will admit to drug use, even with the promise of anonymity.
Which leaves me asking: Do the Monitoring the Future surveys truly posses a high level of statistical validity, or is this another case of data being just data?