If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:

Or download the MP3


I- COVID Deaths vs. Overdose Deaths

According to Our World in Data, essentially a million people have died from COVID in the United States. Depending on your political persuasion you may think that this is an undercount or an overcount, but as we don’t have the time to get into all of that, this is the number we’re going to use. Not only is it probably about as accurate as one is likely to get without massive effort, particularly if you just want to compare the US with other western democracies, it also happens to be nearly identical to the number I’m going to use to start my discussion of drugs. As it turns out, about a million people have also died from overdosing on drugs in the US since 1999, which is generally when the current crisis is said to have started. Though again you can quibble about that number as well, but I’m going to assume that all different quibbles basically balance out.

So we’re left with 1 million people who died of COVID and 1 million people who died from overdosing, that’s our similarity, but what are the differences?

  • Our overdose statistics go back to 1999, while COVID has only been around since the beginning of 2020. Obviously on a day to day basis COVID is far more deadly.
  • On the other hand, with the exception of a small dip in 2018, overdose deaths just get higher and higher every year, while with COVID we expect the opposite, fewer and fewer deaths as time goes on. 
  • It’s nice to imagine that the United States could have locked down as hard as China and prevented nearly all deaths—though if you’ve seen the news, even China is having problems with that strategy now. But in reality everyone knew that, regardless of what we did, some COVID deaths could not be prevented. On the other hand, in theory, nearly all overdose deaths should be preventable. It’s extremely difficult to avoid getting Omicron. It seems significantly easier to just take fewer drugs.
  • Obviously the previous point is an oversimplification, but we assume at some point (nowish?) that COVID will just be treated like the flu. What is the analogy for overdosing? What is the minimum number of overdose deaths we should expect if things worked as they should? The 17,000 overdose deaths we had in 1999? The 10,000 we had in 1998? The 2,500 we had in 1980?
  • Finally, the biggest difference. COVID has gotten, and probably will continue to get, vastly more money and attention. This is the case even if we compare money and attention for COVID since 2000 with money and attention for overdoses since 1999. Which is not to say we haven’t spent a lot of money on the War on Drugs, much of it misspent, but even critics of the war only put it around a trillion dollars, and this is the cost going all the way back to 1971. While COVID spending is already closing in on $4 trillion

This post is dedicated to considering the drug crisis, and while we have been dealing with this crisis for several decades, I think the pandemic has definitely thrown many of the key issues into sharper relief. People would be very angry if COVID deaths just got worse despite everything we were doing, and yet that’s exactly what’s happening with overdose deaths. With COVID everyone is currently engaged in the exercise of deciding what level of danger is acceptable, are we trying to get it to the same level of the flu? Or is that too ambitious? Or perhaps not ambitious enough? And yet people don’t seem to be doing this with overdose deaths. No one can even imagine that we should be able to drive these deaths back down to their 1998 level or even their 1980 level. But 2022 does not seem all that different from 1998, and yet 10x as many people are overdosing, where does that order of magnitude increase come from? Is it entirely the fault of Purdue Pharma and Oxycontin? Or are there other factors? 

Most of all I want to consider, where do we go from here?

II- How “Technology” Contributed to the Increase 

When considering how the number of overdose deaths increased ten-fold in less than 25 years, I’d like to start by looking at the role of “progress” and “technology” in that increase. You may have noticed that both words are in scare quotes. This is an acknowledgement that I am using them in a more expansive fashion than most people. I nevertheless think that the designation and the grouping is accurate. As the “technology” case is easier to make, let’s start there.

The smoking gun here is fentanyl. To begin with it was first synthesized in a lab in 1960. Using technology which had only been invented in the 50s. Fentanyl is a product of modern technology, which not only didn’t exist, but was impossible to imagine more than 100 years ago. Of course, I understand why it was synthesized. The article I just linked to raves about its utility. Having a super potent opioid is perfect for all sorts of entirely legitimate ends, like anesthesia, and pain relief for terminal patients. But this potency, combined with its ability to be synthesized in a lab, make it perfect for the illegal drug trade as well. The potency makes it easy to smuggle and its ability to be artificially synthesized makes it hard to target the source. 

I’ve been careful to talk about overdose deaths in general, but when most people think about the drug crisis and overdosing on drugs they’re largely thinking of drugs in the opioid class, like heroin and prescription opioids like Oxycontin, or synthetic opioids, like fentanyl. And it is true that deaths from synthetic opioids (mostly fentanyl but excluding methadone) have increased 50 fold(!!) since 1999, with most of that increase coming since 2013. But deaths from cocaine have increased by 4 fold, while deaths from psychostimulants, which mostly refers to meth, have increased 30 fold in that period with most of that increase also coming since 2013. 

Though these latter two categories are less obviously stories of something created by technology getting out of hand, technology has still played a major role. 

If we start by looking at cocaine, it’s not immediately obvious why it’s gotten so much worse. Of course deaths from overdosing on cocaine have not increased at nearly the rate that deaths from meth and opioids have, but a 4x increase is still very significant. I murders or suicides or something similar had quadrupled recently then that’s all anyone would be talking about. And yet you probably haven’t heard anything about this increase. Even the books I read don’t spend any time on it. In part that’s probably because everything is going up. Even deaths from benzodiazepines are rising (a point we’ll return to) and in part it’s because the cocaine crisis started a long time ago, but as it turns out it also involved technology.

In the early 80s there was a glut of cocaine and in order to get rid of it dealers started turning it into crack. From Wikipedia:

Faced with dropping prices for their illegal product, drug dealers made a decision to convert the powder to “crack”, a solid smokeable form of cocaine, that could be sold in smaller quantities, to more people. It was cheap, simple to produce, ready to use, and highly profitable for dealers to develop.

The farthest back I’ve been able to find numbers is starting in 1968, and from then till now the low point of drug overdose deaths was 1980, just before this glut occured. As I’ve said I haven’t read much about the way that crisis unfolded. But what’s interesting is although there was a lot of attention on the “Crack Epidemic” it eventually dissipated, but the actual deaths from cocaine didn’t really go down, and the 90s were worse than the 80’s. In fact in 1999, when all the graphs start, it’s cocaine that’s the leading cause of death, not any of the various opioid categories. 

The important point is that it does appear to be an example of this same process of dealers discovering a new drug, or a new form of an old drug and coming up with innovative ways to sell and distribute it. A story that’s going to get repeated again and again. Which takes us to meth.

If you’ve been following my blog over the last few months I mentioned that I’ve been reading some books in preparation for this post, and The Least of Us by Sam Quinones makes some very interesting claims about meth and technology. The story goes something like this:

Back in the very beginning meth was made using what’s called the P2P method, and it gave off a “smell so rank” it could only be done far away from civilization by biker gangs like the Hells Angels, but sometime in the 1980’s the ephedrine recipe for meth was rediscovered, which was not only less smelly, but also an easier recipe to follow. At the time ephedrine was unregulated, so meth took off. One DEA agent said that between 2000 and 2004 he didn’t remember a single pot or heroin case, it was all meth. (To be clear he was stationed in California, not Appalachia.)

As you might imagine this only lasted for a while before the government responded and started cracking down on the availability of ephedrine. Initially production just moved to Mexico, but in 2008 Mexico banned it outright as well. In a perfect world this would have stopped the meth problem, but we live in a fallen world, and the War on Drugs, though not quite the unmitigated disaster many claim, has nevertheless proven to be an amazingly effective generator of negative second order effects. In this case rather than stop producing meth Mexican producers moved back to the P2P method. Given, by this point, the industrial scale of production, the smell was less of a concern then it had been back in the day, but it turned out that there was a different problem: P2P meth, unlike ephedrine based meth, basically causes people to go insane, or at least that’s what Quinones claims in his book. 

Here’s how one user described it: 

In 2009, out in Los Angeles, a man named Eric Barrera was a long-time user of crystal meth when one night he felt the dope change.

Eric is a stocky ex-marine who’d grown up in Oxnard, not far from Los Angeles. The meth he had been using for several years by then made him euphoric, made his scalp tingle; he grew talkative, wanting to party. But that night, in 2009, he was gripped with a fierce paranoia. His girlfriend, he was now sure, had a man in her apartment. No one was in the apartment, she insisted. Eric took a kitchen knife and began stabbing a sofa, certain the man was hiding there. Then he stabbed a mattress to tatters, and finally he began stabbing the walls, gripped by manic paranoia and looking for this man he imagined hiding inside. “That had never happened before,” he said, when I met him years later.

Eric was hardly alone. The new meth that had just begun to circulate in 2009 was different. Something had changed. Gang-member friends from his old neighborhood took to calling the new stuff “weirdo dope.” “Every bag of dope that I picked up after that,” he told me, “I hoped it would be euphoric like it was before. But the euphoria never came back. Instead I’d be up for days paranoid, wondering, Are they gonna raid the house?”

Obviously the question of what makes this meth different is a big one. And Quinones didn’t have a definitive answer. There seem to be three potential explanations. The first is that the P2P method is prone to contamination from the industrial chemicals used in the process and this contamination is what causes the paranoia. The second possible explanation is that meth comes in two different forms d-methamphetamine and l-methamphetamine. The P2P method produces both in equal quantities. Separating the two is difficult, but according to Quinones, Mexican producers have figured out how to do it. But what if they’re sloppy? It’s possible that if you’re taking a significant amount of l-meth, at the level of an addict, that it might bring paranoia as a side effect. And the third possibility is just some other difference in the P2P process, something we haven’t figured out yet. 

For my part neither of the first two explanations seems particularly compelling. The old ephedrine based meth was made under all sorts of conditions by all sorts of people and yet it reliably produced euphoria? While the P2P meth, now being made on basically an industrial scale, uniformly produces paranoia? Still this is the explanation Quinones seems to lean towards. The other explanation, that the change comes down to an inclusion of the other isomer, makes somewhat more sense to me, given that it’s specific to the new process, but l-meth has been studied a fair amount, and is used as in a variety of medicine and there’s nothing to indicate that it causes paranoia. Though as I pointed out addicts are probably taking a lot more than what any study has used, and there is that old saying that the dose makes the poison. But if I had to make a prediction I think I would assign the highest probability to it being some third thing we haven’t figured out yet, though it would get just a plurality of the probability, not a majority. 

This whole business of meth going from somewhat manageable to causing insanity is not something I’ve seen mentioned anywhere else. So perhaps Quinones is exaggerating the problem. But then again, as I pointed out in the very beginning, lots of things about the drug crisis don’t get nearly the attention they deserve, so it wouldn’t surprise me at all to discover that this is a thing and Quinones was the only one dedicated enough to document it. If it is a real thing, it seems like a thing we really ought to get to the bottom of. 

As you can see, technology has done a lot to create and sustain the drug crisis. So much of the story of the crisis is a story of improving technology and distribution methods. Both of Quinones’ books, The Least of Us, and before that Dreamland have large sections that are all about logistics, and improving those logistics, sometimes through better personnel management, sometimes through improved distribution, and sometimes through technology. Though of course in a broad sense improved HR, and improved distribution are also technological advances, ones they’ve borrowed from business. At this point, of course. Mexican drug operations are basically big businesses, ones that are built around taking some chemicals as input and using them to create a profitable output. Businesses that are largely agnostic about which chemicals go in, and which drugs come out. Under this model it made perfect sense to switch to the P2P method for meth. And it made even more sense to replace heroin production and distribution with fentanyl production and distribution. If you were in business and you could replace hundreds of farmers and truckers with a few chemists and just a couple of truckers, you would count that as progress. And indeed it is, which takes us to:

III- How “Progress” contributed to the Increase  

In Quinones’ book Dreamland (see my original discussion here) he puts forth three developments which combined, in perfect storm fashion, to create the opioid crisis. The first, and best known was the introduction of Oxycontin by Purdue Pharma, the second was the development of a sophisticated heroin distribution system running from Northwest Mexico into the US. And the third was an ideological shift in the way the medical profession viewed pain.

As you may have noticed from the title I decided to split this post into two parts, and we’ll discuss Oxycontin and Purdue in the second part. I’ve already discussed the Mexican logistical revolution as much as I’m going to (which is not to say that my coverage has been comprehensive or even adequate, more just that I ran out of time and space.) Accordingly the only thing remaining is to discuss the way the treatment of pain changed. But in doing so I don’t want to just discuss changes in the treatment of pain. I want to look at changes in the way we do everything.

While I won’t be discussing Purdue Pharma, just yet, I do want to spend a small amount of time talking about Arthur Sackler. As I mentioned in my review of Empire of Pain Arthur Sackler was not one of the Sacklers who owned Purdue—those were his brothers—but this fact does not absolve him of all guilt for our current situation, because while Arthur didn’t have any part in the creation of Oxycontin he created the playbook his brothers used to market it. Arthur’s own fortune was made through the marketing of Valium, a benzodiazepine. And what do you know, if we look at our chart of overdose deaths benzos have a category all of their own, and somehow, despite not benefiting from Mexican innovations in logistics, or being involved in pain management, deaths in this category have also increased a staggering amount since 1999: 10 fold, so more than cocaine, but less than meth and fentanyl. Now many of these deaths, particularly since 2014 have involved people who combined fentanyl and benzos, but eyeballing the chart, it looks like benzos went from around 1200 to around 6000 in the years from 1999 to 2011 before combining it with an opioid really took off. Why would that be? We think we know why opioid overdoses increased so dramatically but why did all other categories of overdosing also increase at the same time? I would opine that it all goes back to Arthur Sackler and Valium, and then just a little bit farther still, back to an idea.

Early on Arthur and his brothers worked in an asylum, where the insane languished in appalling conditions. Being reform minded they looked for some way to help these unfortunate people. The story of their various experiments is too long to go into here, but eventually they discovered that, to quote from The Empire of Pain: 

When they injected forty patients who had been diagnosed as schizophrenic with histamine, nearly a third of them improved to a degree where they could be sent home. Some patients who had not responded to any other course of treatment did respond to histamine.

“There was a sense, in their press clippings, that this trio of brothers at a mental hospital in Queens might have stumbled upon a solution to a medical riddle that had bedeviled societies for thousands of years. If the problem of mental illness originated in brain chemistry, then perhaps chemistry could provide the solution. What if, in the future, the cure for insanity was as simple as taking a pill?”

The fact that schizophrenics improved when given histamine was obviously wonderful news. On the other hand, the idea that all mental issues, large or small, could be fixed with a pill, was a dangerous overreach. Nevertheless they took this idea and ran with it. While the Sacklers didn’t do much to discover new cures, like the one they’d stumbled on with histamine, their zeal, and in particular Arthur’s, led them to become experts at marketing chemicals. A strategy which relied on this idea that just taking a pill was all it took to cure what ailed you.

You might imagine that the next step in this story was applying the strategy to Oxycontin. But actually the next step was applying the idea to pain management in general. Oxycontin didn’t create a revolution in the ideology of pain management, a revolution in the ideology of pain management created the conditions necessary for Oxycontin. The revolution in the treatment of pain management is a long story, and this post is already long, but I came across this comment over on Marginal Revolution from a doctor which sums up the situation pretty well:

I’m an anesthesiologist, so I do all my narcotic “prescribing” via syringe these days. Before that, I was an internist, writing lots of prescriptions. I was doing this up til the mid-90’s, when we started hearing about the supposed “epidemic” of untreated, severe pain. Lots of actors involved in that little drama: pharma; Big Nursing looking to demonize “uncaring” physicians for their own ends; inter alia. Anyone remember “pain is the 5th vital sign”? I sure do. There was relentless pressure to make sure that no one, ever, faced a quantum of untreated pain. Suddenly, pain surveys and other forms of government coercion became part of the water we swam in. Getting a reputation as an “undertreater” of pain could have serious professional consequences.

Is anyone surprised that the pharmaceutical industry responded to this milieu? And that government piled on through its enforcement arms in HHS? If you tell the public for a couple of decades that everyone is entitled to a pain-free existence (not the actual message sent, but often the message received), then don’t be surprised at the disaster that results.

Presumably the connection between that original assertion of the Sacklers (and to be fair I’m sure it wasn’t just them) and this situation should be obvious: If you can cure something as obviously bad as pain with a single pill why wouldn’t you? But once you start thinking along these lines, why would you limit it to only things which are legal? If you can take some drug and it makes all your problems go away why wouldn’t you?

I understand there are other factors involved. Drugs are addictive. Wicked companies have marketed them with lies and distortions. There are all the advancements in distribution and logistics I mentioned previously. But along side all of that, and perhaps preceding it, is the idea that we can use progress to solve all of the old problems. Anxious? Take a Valium. In pain? Take Oxycontin. Not enjoying life as much as you think you should be? Take meth. 

Because the thing is, that as much as we might want to blame Oxycontin for creating a drug crisis, which came out of nowhere in 1999, deaths from drug overdosing have gone up every year since 1990. In the last 30 years no matter what drug you look at, and no matter when you decide to start looking, everything is going up. My argument is that this phenomenon is yet another unforeseen side effect of progress, one that’s going to keep getting worse. Can anything be done? We’ll answer that question next time.


I didn’t want to split this in two, but things have been extraordinarily crazy, and to add to the craziness, we’ve decided to move. If you want to help with the expense of that consider donating