Overdose Deaths in the USA and Portugal

/ bs king

I’d like to take a break from posting on all cause mortality in the US and its relationship to COVID deaths to post about a different subset of deaths. Drug overdose deaths are unequivocally on the rise this year, going from about 78,056 between May 2019 and April 2020 to 100,306 in between May 2020 to April 2021. Since I am obsessed with state level data, here’s the CDC breakdown of where the jumps occurred:

This was of interest to me because (as I have posted previously) my brother has an unfortunate amount of experience with opioid addiction, and has written a book about it. With the opioid crisis in the US continuing to worsen, it’s no surprise he’s started trying to look at other countries to see how they have addressed things and if anything they’ve done has made a difference.

Enter Portugal.

If you read anything about our handling of drugs, particularly any of the cases around drug legalization, Portugal is bound to get cited. The basic narrative goes something like this: in 2001 Portugal decriminalized the consumption of all drugs. Most critics thought this would lead to worsening drug abuse/deaths, but instead drug related deaths fells, incarceration rates fell, and drug use was not impacted. Here’s an advocacy organization making this case. Here’s Time Magazine making this case. Here’s the Cato Institute making this case. Here’s the Obama White House complaining about everyone who’s making this case.

When I first heard about Portugal, I’ll admit I took this all at face value. After all, even the White House resorted to kind of nitpicking the statistics. But a few weeks ago I stumbled on to this Twitter thread from Lyman Stone that left me with more questions than answers. He started by pointing to the Global Health Data Exchange, where you can pull all sorts of mortality data for all sorts of countries. First, here’s Portugal’s deaths from drug use disorders:

Not bad. Looking at the source data, nothing for 2000 is available, so we skip from 1999 to 2001 and we do see a drop between those two years. So far so good. However, Lyman pointed out that drug use disorder deaths are one of two types of deaths filed under “substance use disorder deaths”. The other is alcohol related deaths. Here they are plotted together:
So….the same year the drug related deaths started to drop, alcohol deaths went up by an approximately equal amount. So what happened here? Did people start taking alcohol less seriously when drugs were legalized, or did some deaths that previously would have been called drug deaths get filed as alcohol deaths? Or is it all a coincidence? I actually don’t know. I couldn’t find anyone addressing this. It does appear Portugal has a legitimately alarming rate of alcohol consumption, so any explanation appears somewhat plausible. What we do know is that other countries nearby did not see the same increase in alcohol related deaths, even as they saw their drug deaths decrease:
If you lump them all together, we see that Spain has seen a decrease (though is higher) and Portugal has been steadily increasing since 1990. If you use the DALY’s (those account for age), things look even worse for Portugal:
I went to Wikipedia to see if I could get any general insight on the topic, and the opening paragraph caught my eye. Per Wiki, the purpose of the policy change in 2001 was actually to focus on HIV deaths. There isn’t a solid citation for this so I’m not sure if this is true, but I will note other articles tend to say this was “part of a broader health initiative” so it seems plausible. If it is true, then it’s actually pretty confusing what happened next:
Throwing in a few comparison countries (including the US), we note that every other country had a precipitous drop in HIV deaths in the late 90s. Portugal has had a much slower decline. Note the order of magnitude as well: right now Portugal appears to have over twice as many people dying of HIV as they do of substance use. I have no idea what happened here, but we did not see this in Spain or Italy.

Wiki did add to the drug overdose death mystery as well, noting that in this 2010 paper it stated that “the subsequent increase has been attributed by local informants to a shift in measurement practices, namely an increase in the number of toxicological autopsies performed (from 1,166 in 2002 to 2,805 in 2008), which increased the probability that people would be found to have drugs in their bodies at death.” To note: I have no idea if Portugal is now testing more people than most countries or fewer. This is why international comparisons are hard.

The same paper actually clarifies a different question that I had. Lyman Stone brought up the fact that despite decriminalizing everything in 2001, in 2008 the laws were amended to recriminalize large amounts of drugs. Not being very familiar with drugs/amounts, I wasn’t clear how much a “large amount” was. It turns out it’s a “10 day supply” which is apparently defined as “0.1 g heroin, 0.1 g ecstasy, 0.1 g amphetamines, 0.2 g cocaine or 2.5 g cannabis”. Now obviously there is not a super standardized measurement for heroin, but based on a bunch of Googling (that may have permanently flagged my account somewhere) I found that in Ithaca, New York one gram of heroin is typically divided in to 20 bags. A bag is $20, or a bundles for $180. So two bags of heroin would land you back in criminal territory.

This clarifies why the incarceration/conviction rate went up, as I assume 2 bags may not be a super unusual occurrence.

So basically, here are the questions I’m left with: why did Portugal experience such a spike in alcohol deaths right after they decriminalized the consumption of drugs? Why haven’t their HIV deaths come down like most other countries? Did the recriminalization really only catch dealers, or were a lot of regular users caught up in this as well? If anyone knows any good sources on any of this, I would appreciate hearing from you!