Wednesday, November 27, 2019

Cannabis Legalization Doesn't Drive Up Violence

Many states have begun to legalize recreational marijuana. Some people, most notably Alex Berenson, have claimed that this is leading to deteriorating trends in certain vital statistics post legalization. He points to the apparent increases in violent crime in Colorado, Oregon, and Washington. I first became aware of this earlier this year, when Berenson published a piece in the New York Times warning about the supposed dangers of marijuana. Malcolm Gladwell and Tyler Cowen amplified the signal, in my opinion far too credulously. It's come to my attention again. Berenson's recent debate with Jacob Sullum (my write-up of that discussion here) led me to do some digging. To get a good idea of Berenson's position, it's worth listening to this Joe Rogan podcast, in which Joe hosts a sort of discussion/debate between Berenson and Dr. Michael Hart. It's a good source to understand the nature of Berenson's claim. His take is that marijuana use causes psychosis in some people; I think he admits that only a small fraction of people are really at risk and most normal people don't  have anything to worry about. Nevertheless, in his telling, increasing the incidence of cannabis use will cause some small fraction of people to turn psychotic. Psychosis is an enormous risk factor in the commission of violence; he mentions a 20x figure in the Joe Rogan podcast, as in people with psychosis are 20 times as likely to commit violent crime. So more cannabis -> more psychosis -> more violent crime.

So how does one decide whether cannabis legalization increases violent crime? The naive first approach would be to look at rates of violent crime before and after legalization. If violence increases post legalization, that's at least suggestive. But of course that doesn't tell you anything about causation. Violence crime rates increase and decrease for lots of reasons. If violent crime is increasing in states that legalize, we need a control group to compare them to. We need to establish the counterfactual, as in what would have happened if not for cannabis legalization. Also, as with anything we measure, there is random noise in these statistics. That's not to say we don't know how many murders actually happened in Colorado in a given year. Even supposing we count those accurately and to a body (and I think the counts are quite precise), the process that generates those body counts is random. Many of these violent crimes are crimes of passion. Someone gets angry in the moment and assaults or murders someone; the same person on a slightly better day wouldn't have become a murderer. Someone fires a gun with the intent to kill, but misses their victim. A murderer experiences a delay in their plans when their victim does something unexpected, giving the would-be murderer time to rethink things. So the actual realized crime rate, literally the count of crimes divided by the population, is a somewhat noisy measure of the thing we're really interested in, the underlying propensity of the population to commit violent crime. So we need error bars. We need confidence intervals to tell us if an X% increase in crime rates is something that could come about by random chance or something that indicates that the underlying drivers of crime have actually changed.

There have been some very good studies that find appropriate control groups and compute the confidence intervals, and I'll get to these later. There seems to be some contention about whether those controls are done appropriately. Perhaps the statisticians, who are often accused of being themselves biased in favor of legalizing cannabis, are using statistical magic to make a real effect disappear. No doubt this is a very real problem in some academic disciplines. There are generally many different model specifications to choose from. Just a few nobs to turn and flips to switch on your regression analysis leads to a combinatorial explosion of possible models. It's always possible that the modeler has done a large number of these and cherry-picked the one that gave him the desired result. I think the appropriate way to deal with this is for the statisticians and critics to come together and discuss the various choices in model design, decide which specifications are more plausible or appropriate, settle on a constrained space of model specs, and openly report the sensitivity of the results to various specs.

For the moment, I'll start with crude methods.

The Raw Data

Below I have plots of violent crime rates by state for the states that  have legalized cannabis. The data all comes from here. (Simply click on Summary (SRS) Data With Estimates and download the estimated_crimes.csv file. That's what I'm using to build these plots.) There are four kinds of violent crime tracked here: aggravated assault, homicide, rape, and robbery. If Berenson's story is correct, we're probably only interested in aggravated assault and homicide. It's unlikely that paranoid psychosis-induced violence would have an effect on the other categories of violence.

Start with aggravated assault, which is more frequent than murder and thus has a stronger signal, statistically speaking. Notice that I'm color-coding based on the year that cannabis was legalized. Colorado and Washington legalized in 2012. Oregon, Alaska, and DC legalized in 2014. California, Nevada, Massachusettes, and Maine legalized in 2016, and Vermont legalized in 2018. The data is only through 2018, so we only get one post-legalization data point for Vermont. (If the figure is overlapping the stuff on my side-bar, note that you can click directly on the figure to enlarge it for a clearer view.)


It's certainly true that some of these states have seen increases in violence, as measured by this statistic. The post-legalization trends for AK, CO, WA, and OR are up. But DC is down. CA, MA, ME, and NV are down, despite the trend for the nation as a whole (marked "US") being slightly up. California has been declining for two decades, despite having medical marijuana since 1996. Many think that medical legalization in California was de facto legalization, as it was permissible for a doctor to recommend it for almost anything. (I don't fully understand this, but doctors in California could issue "recommendations" that weren't officially "prescriptions", but as long as you secured such a recommendation you could get cannabis from a dispensary.) Other states placed tight constraints on what conditions were acceptable for medical cannabis, so the connection between medical cannabis and rates of recreational use is less clear. Note that Colorado and Washington initially decrease before coming back up. This is not a unambiguous "legalization leads to violence" story. Now let's look at homicide rates.

Again, not an unambiguous story. Homicide certainly increased in Oregon, but then in 2018 it came back down to where it was 15 years ago. California is declining? DC is essentially flat. Certainly, Colorado, Alaska, and Washington look like they're up, but it would take some cherry-picking to tell an anti-legalization story.

Let's take a longer view. In the above, I'm purposefully focusing on the year 2000 and later so we can compare post-legalization to recent trends. What if we go all the way back to 1979, where the data begins? Violence was famously much higher in the 80s and 90s. As Stephen Pinker documents in his book Better Angels of Our Nature, violence has been declining for a very long time. Take a look at the recent trends in light of a historical perspective. Here is assault:


And here is homicide.


Other than assault in Alaska, the historical perspective makes the recent trends look puny. Don't get me wrong, it would be horrible if a state's violence rate went back to what it was in the 80s and 90s. I don't want be accused of making the argument: "The effect gets smaller when I rescale my y-axis." But we should try to appreciate the effect of cannabis on crime, whatever its magnitude, in context with other drivers of crime and other policy considerations. It is widely believed that drug prohibition is a major driver of violent crime because of black market violence, and the so-called "crack epidemic" of the 80s and 90s saw high levels violence between rival drug gangs. This was mostly not pharmacologically induced violence. This paper by Jeff Miron suggests that variation in the level drug enforcement explains something like half the variation in the murder rate. Note that's a regression analysis, and statements such as "explains half the variance" are statements about correlation, not causation. But the "causal" story is reinforced by the work of Paul Goldstien, who actually looked at all the murders in New York city from 1984 to1988. He coded them as non-drug related, pharmacologically induced violence, economic crimes to support a drug habit, and black market violence. Violence related to drug commerce was the dominant category. (See the discussion in the Miron paper, which also lists the Goldstein papers in the References.) If that's even close to correct, we should be far more concerned about violence related to drug trafficking than the pharmacologically induced violence that Berenson is warning us about.

Maybe actual use rates post legalization changed more in some states versus others? We don't care about legalization per se unless it drives up cannabis use and leads to these social problems that the critics are worried about. I looked this up, too. It's not exactly well organized, but SAMHSA tracks this by state and publishes reports. They have "past year" and "past month" marijuana use. I'm using past month, because presumably the higher-frequency use is what's going to track with excessive use and thus violence, assuming that causal connection exists.  (Here is one year's data. I could not find a file where this was all in one place for all years. Also, some years have this in a pdf, while others had it in an Excel spreadsheet. I had to awkwardly key values from a pdf into my own spreadsheet for those years. Government agencies need to get a lot better about how they disseminate their data!) The SAMHSA reports compute by-state use rates using two years at a time, so, for example, 2016 on my charts is really 2016-2017. I couldn't find anything in the 2004-2008 time period, but one report had the data for "2002-2003". So we at least have one data point from the period well before legalization. Here is what I saw.

Note that cannabis use trends upward even prior to legalization, but it definitely looks like there is a post-legalization increase. Unfortunately, there just aren't enough post-legalization data points for the states that legalized in 2014 and 2016 (and none yet for VT, which legalized in 2018). Now, one could seize on this and say, "Ah ha! Cannabis use increased more in Colorado, Oregon, and Washington, and those are the states with the clearest pattern of post-legalization increases in violence!" But it also looks like DC and Nevada saw a stark increase in cannabis use, and these jurisdictions bucked the "increasing violence" trend. California saw increasing rates of cannabis use for a decade and a half, but its rates of violent crime were declining over this period. Crudely speaking, this is a period of declining rates of violence and increasing rates of cannabis use.

For comparison, here are some non-legalizing states. It looks like cannabis use rates are increasing in many states that did not legalize. So the increase in the general trend for the US is not simply being driven by legalizing states.

Maybe "past month use" isn't what we care about? Maybe it's only excessive, daily use that leads to psychosis and violence, and if we had that measure it would track with the violence rates? Maybe, but I would at least expect "past month use" and "daily use" to be highly correlated.

Of course, this is all very crude. Eyeballing graphs and computing raw statistics is dangerous. One is likely to fool oneself and spot patterns that aren't really there. We need a more rigorous analysis that controls for confounders and computes confidence intervals. I'm only showing this because some anti-pot crusaders obsess over the raw data. The statisticians who do a more rigorous analysis are sometimes accused of being biased, using statistical tricks to make the effect disappear. With the above plots, I'm merely trying to show that crude methods, like eyeballing the raw trends, don't tell an unambiguous "more cannabis -> more crime" story.

Studies Using Synthetic Controls

I found two good papers on the cannabis-violence question, both of which use synthetic controls. The idea here is that you find a set of "control" states that behaved like the "treatment" state prior to legalization. Actually, if I'm reading this right, the approach uses a weighted average of all "untreated" states, picking a set of weights that best matches the legalizing state's trends prior to legalization. (One paper uses a subset of non-legalizing states, which it calls "donor states".) Having established these weights, one can compute what the trend for the legalizing state should have been if it hadn't legalized, then compare that to what actually happened. One of these papers discusses legalization of medical cannabis, the other recreational legalization.

The paper on medical cannabis is called Joint Culpability: the Effect of Medical Marijuana Laws on Crime. Here is a link to the full paper. Medical marijuana laws are not the same as laws that fully legalize recreational marijuana, but surely they increase the availability of marijuana. Some states had lax prescribing guidelines, such that you could get a prescription for almost condition. In fact, anti-cannabis fuddy-duddies like to complain that medical cannabis was just a wedge issue designed to make recreational cannabis more acceptable. As described above, many of them complained that California's medical cannabis law was de facto legalization, considering the ease with which one could obtain a doctor's recommendation. Supposing these cannabis critics are correct, legalizing medical marijuana should have led to an increase in recreational use. It seems likely that there would be some leakage from legitimate prescriptions to recreational users. (Based on some personal anecdotes, I know that this happens to some extent.) This paper cites several studies that quantified the effect of medical marijuana on overall use; they suggest the increase is in the 10%-20% range. If our hypothesis is that an increase in recreational marijuana use leads to more violence, the experience of medical cannabis states is a useful test.

Generally the paper finds no statistically significant effect one way or the other on violent crime or property crime. But if one fixates specifically on California, there does appear to be a significant reduction (a ~20% reduction) in both types of crime. I'm happy to call California an outlier and just declare this a null result. But the result is not favorable to the "cannabis use causes violence" story. See the comparison of synthetic controls versus actual experience in California.

The second paper is a student thesis by one Paul Lalonde titled Reeform Madness: Crime in the Changing Landscape of Marijuana Regulation. (I love the puns built into the titles!) It limits its analysis to Colorado and Washington. That makes sense. The crime data it uses is through 2016, so that doesn't give it much time to test the states that legalized in 2014 and 2016. It would be interesting to see this analysis redone in the next couple of years when more states have had a longer experience with full legalization.

By the way, if anyone is tempted to dismiss this as "some kid's term paper", note that Ben Hansen, got a similar result. Hansen is an academic economist with publications in the "synthetic controls" literature, notably this one, a working paper about the effects of cannabis legalization on auto fatalities. As far as I can tell he hasn't published on the violence question.

Reeform Madness generally finds that post-legalization violence rates in Colorado and Washington are slightly lower than the synthetic controls, and property crimes are slightly higher, but the results are not statistically significant. (Read the paper for the exact methodology; they're using some test statistics I'm not familiar with. Anyway, it's clear that they're not hitting the traditional "p < 0.05" threshold.)


In sum, while it's true that Colorado and Washington saw increases in violent crime post legalization, states that look like Colorado and Washington saw comparable increases. I'm quite happy to call this a null result or declare it inconclusive. But certainly this doesn't paint a picture of a post-legalization spike in violence, even for those states that definitely did see increases in murder and assault.

Back-of-the-Envelope Sanity Check

I want to do a quick computation just to get an idea of what magnitude of effect Berenson is suggesting. Past-month cannabis use in Colorado was 10.4% of the population in 2011/2012 and 16.4% in 2016/2017. Let's subtract the two and say that 6% of the population turned from non-smokers to smokers because of legalization. The murder rate increased from 3.0 per 100k in 2011 to 3.4 per 100k in 2016, a 13% increase. For 6% of the population to drive a 13% increase in murder at the population level, this must mean that "past month cannabis use" is a huge risk factor for murder. It apparently increases your chance of becoming a murderer by 216% ( = 13% / 6% )! Even making some more conservative adjustments, this is a huge effect. Suppose that in addition to creating new users, legalization doubled the intensity of use among exiting users. So let's say effectively 16.4% of the population are "new users", adding the actual new users to the existing users who are now doing double-duty. And suppose we only blame pot legalization for half the increase in the murder rate (so call it 6.5% instead of 13%). We're still saying 16.4% of the population caused a 6.5% increase in the murder rate, so this population's murder rate must have increased by 40%. These seem like implausibly large figures to me. The people who are attributing even small population-level increased in violence to cannabis legalization are actually positing a huge effect of marijuana's propensity to cause violence. Of course, most of them will say, "No, no, you won't be 40% more likely to commit murder, or any more likely for that matter. It's all due to a small subset of people who can't handle it, have a bad reaction, and turn psychotic." But the smaller we make the "susceptible" population, the larger the effect on violence we must posit. 

There is nothing special about Colorado; I get similar absurdly large "effect sizes" when I look at Oregon and Washington, depending on the end-points. Of course, as implied by the discussion above, doing this exercise for some states would yield negative effects of legalization on crime.

I think that you can probably find news stories about people who smoked way too much, got overly excited or paranoid, and did violence to someone. I even think it's reasonable to say that cannabis use was the cause, in a but for sense, of some of those violent acts. (I'm reminded of an old post at Slate Star Codex in which Scott Alexander describes a colleague's frustration with marijuana. It seems that cannabis use interfered badly with his treatment of some of his schizophrenia patients. This is part of Berenson's story: cannabis use is particularly bad for people with preexisting mental illness, even if it's benign for most users.) But these are exceedingly rare. Maybe it's reasonable to trumpet these stories as a cautionary tale for people with risk factors, but blaming them for population-level increases in violent crime is just implausible. It's also inconsistent with the crime data from various states. 

I hope that someone with more time and better resources than myself picks this up and expands upon this literature. I think there is a lot more work that can be done. A suggestion to the people who are doing this kind of work, it might be helpful if you showed the full details of your analysis, say in an R Markdown or a Jupiter notebook. That way you can really answer your critics and say, "Which step in the analysis do you disagree with? Run it yourself and see if the result changes." That way, when someone furrows their brow at charts like the ones copied above, there is no question as to how they were generated. Give the critics a chance to respond with specific criticisms, and give yourself a means to "shut them up" if they're just blowing smoke. 

Tuesday, November 26, 2019

If It Weren't For the Mississippi River, We Could Have Kept All That Fresh Water!

Imagine you’re watching the Mississippi river flow into the ocean with a very confused friend. He’s shaking his head at the “waste” he sees. All that fresh water, just flowing into the ocean! Why, if it weren’t for this river, there’d be plenty of water to satisfy all household and industrial needs for fresh water. “Damn this river! If only we could…well…dam this river! Okay, we already do that. I mean, if only we could fill in the entire thing with earth so our precious water would stop escaping into the ocean.”

You’re so taken aback by the silliness of his comment that you momentarily forget that flowing water is useful, in that it provides relatively cheap transportation among other things. Your more immediate thought is that the water will find its way to the ocean, Mississippi or no Mississippi. It’s no good to blame this particular river. Even supposing someone could snap their fingers and instantaneously replace all that water with dirt (perhaps a confused environmentalist with absolute power, like Thanos with the infinity stones), the rains would still come. Without the mighty river and its tributaries, the rain water would simply sit on the land and temporarily cause a lot of flooding. But water flows downhill. It eventually finds its way to the ocean. It would first follow whatever gentle gradients present themselves on the landscape. Eventually the flowing water would carve out a new set of creeks, streams, tributaries, and rivers. A new river would eventually establish itself, because the natural forces of evaporation, condensation, gravity, and erosion compel it. Blaming this particular river makes little sense.

I hope reading this gives you some sense of the frustration I felt reading Fentanyl, Inc.: How Rogue Chemists Are Creating the Deadliest Wave of the Opioid Epidemic by Ben Westhoff. The subtitle is an excellent summary of what the book is about. Don’t get me wrong. It’s a great book, well worth reading. But there were many “Dam this river!” moments in the book. He spends a little too much time fixating on particular individuals and companies that produce synthetic drugs. Fentanyl, for example, is produced in clandestine labs. If you shut them down, others would pop up to feed the market demand. It's synthesized using precursor chemicals that are sold by a few legitimate Chinese chemical companies. If you shut those down or made the precursors illegal, clandestine labs would synthesis those precursors or maybe even discover alternative means of synthesizing fentanyl. I'm sure he realizes this, and he might even issue this caveat once or twice (if he did, I missed it).

Generally, the book is about synthetic drugs and how they are produced and trafficked. Governments are always banning chemicals as soon as some people figure out that they’re fun, and chemists are always coming up with clever synthetic analogues that copy the pharmacology of those chemicals. Sort of. Unfortunately, sometimes the synthetic analogues are not as safe as the originals. MDMA, aka Ecstasy, is basically non-toxic. Purported harms, such as the claim that it turns your brain to Swiss cheese, were either entirely untrue or gross exaggerations. Nevertheless, “do something” politicians can’t seem to help themselves when it comes to banning fun, and MDMA was outlawed basically everywhere. Synthetic analogues of MDMA, often called “Molly”, started popping up, and partiers started dropping dead at raves. The author claims that this is what piqued his interest in the topic. He knew about the rave scene; he's a journalist whose regular beat is music. It was almost unheard of for people to actually die from MDMA use, but every major rave was seeing a death or two. What the hell was going on? It turns out that ravers were using synthetic drugs of unknown origin, often with deadly consequences. Real MDMA was hard to come buy. It seemed that synthetic "Molly" had mostly replaced it.

To take another example, synthetic marijuana (sometimes called "Spice" or "K2") was briefly popular. It was sold in head shops, often packaged as potpourri and labeled "not for human consumption." Obviously such packaging will not give the user appropriate dosing information, and concentration of the active chemical might vary dramatically from one producer to another, or even from one batch to another by the same producer. To make things even worse, it takes a while for Spice to "kick in", so some users would take hit after hit before they felt anything and ended up taking way too much. Initially, users described synthetic weed as producing a cannabis-like high. I actually first heard about it from a friend who said she used it, probably in 2009, and she described it to me in those terms. However, unlike weed it made some users extremely agitated and even psychotic. One bad batch seems to have caused a minor zombie apocalypse. In this particular instance, thirty-three users wandered dazed around a public park and required medical attention. A friend of mine who is an EMT described having to manhandle a guy who was whacked out on synthetic cannabis, warning the several people listening that they shouldn't take it. This whole thing is unfortunate, because legal cannabis would have none of these problems. It's relatively easy for users to appropriately dose (when smoked, not necessarily when eaten) and extremely non-toxic. Wagging our collective fingers at the chemists who synthesized it or the head shops that sell it makes little sense. This is a policy problem. Given a policy of drug prohibition, someone is going to fill this market niche. We will inevitably experience all these problems with new, experimental drugs that are less safe than the drug whose effects they are trying to copy.

Westhoff did some very deep reporting on the Chinese chemical industry. I am impressed by his success in infiltrating those companies, and his sheer balls. He was often traveling in parts of China that he didn't know, with an unreliable GPS, in a vehicle driven by someone who was suspicious of his motives. It certainly took guts to do what he did. He claims to be the first western journalist to actually visit any of these businesses, and for all I know that's true. He sometimes posed as an interested buyer and spoke to sales reps. Remember, these are legitimate companies selling precursor chemicals, not fentanyl itself. Westhoff discovered that these companies were making enormous shipments of chemicals whose only known use was the production of fentanyl. Again, there is a lot of implied finger-wagging. Westhoff discusses whether the people working for these companies knew what they were actually selling, speculating that the sales reps probably didn't but the managers probably did. The undertone of this "Do you know what you're selling?" discussion is that, if they did know, they were morally responsible for the damage caused by fentanyl. The book doesn't come out and say this, but there's no other reason to discuss the company's knowledge of how its chemicals are used. Once again, it's kind of ridiculous to think that we wouldn't have this fentanyl problem if it weren't for these particular Chinese companies deciding to be malfeasors. This is another policy problem. With policy such as it is, with heroin being illegal for recreational use, someone is going to manufacture fentanyl and ship it to the United States. Westhoff does a good job of mapping out particular streams and tributaries in the supply chain, but this mighty river will flow even if we dam those up. The underlying forces that keep it flowing don't change just because you shut down a Chinese firm or two. Westhoff even concedes that the same chemical expertise exists in India, and law enforcement in that country has discovered clandestine operations.

There is a long section about the academic literature on synthetic drugs. Often a new drug begins life in the pages of academic journals, and then some enterprising chemist figures out that these substances might make fun club drugs. Westhoff recounts the tale of a young man who dies as a result of using one of these synthetic drugs. He interviews both the academic who invented it and the father of the man who died. The academic points out that this kind of work serves a purpose, and that he can't be held responsible for someone's inappropriate use of a scientific discovery. The father, who obviously blames the scientists for his son's death, responds with something like "Fine, if that helps him to sleep at night." Westhoff isn't necessarily taking sides, but he is suggesting that there might be moral culpability for the scientist who synthesize these drugs and giving voice to people who take this view. Again I found myself longing for some kind of economic analysis, some kind of suggestion that these things happen because of the incentive structure created by drug prohibition, not because particular people discovered particular chemicals.

The final chapter of the book is about harm reduction. Westhoff discusses the various means that ravers use to stay safe. There are some organizations (such as DanceSafe) that show up at raves with drug testing kits so the ravers know what they're taking. Sometimes these organizations will try to set up booths. Rave organizers don't care for this and sometimes turn them away. Because of the absurd RAVE Act (sponsored by Joe Biden, whose record on drug policy is just abysmal), rave venues could be seized by police if they can show the organizers knew there would be drug use on the premises. Some organizers quite reasonably believe that allowing drug testing booth is an admission that you know drug use is going on. Pretending it doesn't exist probably puts them on firmer legal ground, but it doesn't make the drug use go away, and it makes harm reduction that much harder. Luckily, there are organizations that show up to these raves without announcing themselves and secretly distribute testing kits, effectively performing a kind of guerrilla harm reduction.

The book also covers safe injection facilities, where IV drug users can come to inject. There are volunteer nursing staff who can administer naloxone to users who overdose. These facilities tend to have good safety records, but they are often forced to operate underground. Knowingly hosting a lot of active drug users is illegal, but they are lobbying various local governments to figure out the boundaries of the law and how drug laws will actually be enforced.

I like harm reduction, but I need to issue a gripe on this point. Harm reduction advocates need to start talking seriously about legalization. Legalization accomplishes everything that harm reduction is trying to do. It eliminates the problem of users not knowing what substances they're taking, or in what dose they are taking them. Overdoses might still happen because someone miscalculates their tolerance, but the problem is dramatically reduced. Moreover, supposing there are still serious overdose risk factors under a legal regime, in a free market someone will figure this out and administer some kind of paid service to mitigate those risks, possibly provided gratis with purchase of the drug. The phenomena of needle exchanges, guerrilla harm reductionists at raves, and safe injection facilities all demonstrate that there is a massive willingness to supply these services when the market price is zero! (In fact, the market price is quite negative, considering that at least some of these people are exposed to legal penalties.) One sees a similar pattern with psychedelics. Michael Pollan details his experience with psychedelic "shamans" or guides in his book How to Change Your Mind. Not only do these people give away the product for free, they offer guidance to the psychonaut while they are tripping. This is quite labor-intensive, as some of these psychedelics last for hours. Clearly there is an enormous willingness to supply these accommodation services for drug users. Allow legal companies to provide these services for a fee, and we won't need to rely entirely on volunteers. I really don't understand what the harm reductionists have in mind, at least the ones who stop short of legalizing the production and sale. An industrial-scale factory that synthesizes these drugs using modern quality control, audited by third parties, perhaps regulated by the FDA, would solve most of these problems at the front-end. We won't have to rely on volunteers instituting half-measures, with their hands being bound by the frontier of the legal grey-zone they operate in. Are these harm reductionists afraid of the increase in drug use that would probably arise from a legal market? Are they leftists who tend to be skeptical of "big business"? (This is my preferred explanation for what's happening.) Are they secret legalizers who are just seizing on harm reduction as the incremental step that's available right now? (If that's the explanation, maybe I shouldn't be writing any of this.) Incidentally, this is a river that might not flow if the particular individuals didn't exist. Some of these harm reduction practitioners are extremely dedicated individuals with a very wide reach and rare moral courage. There isn't necessarily an economic force providing a "supply of harm-reduction volunteers" as there is a force servicing "the market demand for heroin." It seems like the former is much more reliant on rare individuals; the latter continues no matter how many dealers or manufacturers we arrest.

Apologies to Ben Westhoff, who really did write an excellent book. I hate to do a review where I talk about the book I would have written and the topics I would have covered. Fentanyl, Inc. isn't a book about drug legalization, and it isn't quite fair or me to insist that it should have been. But I do think that an economic understanding of the war on drugs will enlighten the reader as to what's going on and present some context for why these things are happening. What is the point, after all, of studying a massive social problem if we aren't going to talk about policy solutions? If it weren't for the fact that massive  numbers of people are overdosing on fentanyl, this wouldn't even be a book. It's fine to have a "just the facts" presentation of what's happening, and the book mostly achieves this. Let's explore the contours of all those streams and tributaries that feed this mighty river. It's good to know about the world that we live in as it actually exists. But let's also explore the forces that dump all this water on the land in the first place.

Sunday, November 24, 2019

Jacob Sullum's Debate with Alex Berenson

Jacob Sullum recently debated Alex Berenson at the Soho Forum. The proposal formally under debate was
Except for laws prohibiting the sale of drugs to minors and driving while impaired, all laws that penalize drug production, distribution, possession, and use should be abolished, along with special "sin" taxes on drugs.
But it's actually a wide-ranging debate/discussion of drugs and drug policy, not necessarily aimed specifically at answering this question. (The debate was a tie by the official vote, which surprised me. Sullum clearly had the better arguments and the facts on his side. Perhaps he would have won if the proposal was milder? The vote would have been more favorable to him if the proposal had been for a relaxation of drug laws or abolition of laws against particular drugs. I could actually see myself being a stickler for the literal language of the proposition and voting for Berenson,  thinking, "Well, Sullum is right about prohibition generally, but he didn't defend this proposition.")

I will try to discuss the major themes of the debate in the paragraphs below. The format of the debate was two 15-minute intros, two 5-minute responses, a Q&A session between the two participants and with the audience, and finally a 5-minute closing remarks, followed by the vote. I will try to integrate the discussion, in which a particular topic may have been discussed several minutes or an hour apart on the actual timeline (as one participant responds to the other several minutes later when it's their turn to speak, or when an audience member brings up a topic discussed an hour ago). I won't hesitate to add my own commentary or judgment. (It's my blog, anyway.)

Berenson repeatedly claims that drug use is a "crime of risk," similar to driving drunk or texting while driving. His point is that you are potentially causing random harm by taking drugs, so we need to punish all users, even the ones who don't happen to cause any harm. He says, "You can drive 50 miles drunk, but get pulled over a few feet from your house." The point being that even though you in fact made it home okay, you were risking people's lives along the entire way. I think he's trying to say that taking drugs is like playing a game of Russian Roulette (though, and I think he'd admit this, you'd have to have a gun with dozens or hundreds of chambers and load some less lethal bullets to make that analogy work). When someone uses drugs, it's like they cast some dice in a cosmic game of chance, and they get harmed very badly if it comes up snake-eyes.

I think this "random risk" framing is absolute bunk, and I've said so before. Initiating drug use and continuing to use is a choice, at least initially. Contrary to some of the scary claims made to fool school children and impressionable adults, a single dose cannot cause addiction. Certainly it won't cause a physical dependence. That requires a pattern of repeated drug use over time, which (again, at least initially) is a deliberate action. Let me start with a hypothetical scenario in which Berenson's framing is right. Suppose some random fraction of the population had a gene that made their brain chemistry interact badly with cocaine (or some other drug), such that those people are driven violently psychotic after a single dose. It might make sense to call that a "crime of risk," because taking drugs in that world more closely resembles a game of chance, unless we have some way of knowing who does and doesn't have the gene. But Berenson's framing does not match the real world. People have agency. Some fraction (a small fraction) of drug users become addicts and engage in compulsive, repeated drug use. While it's technically true that they would never become an addict if they never took that first dose, it's a harder sell to claim that that first dose causes their addiction. Sure, it's a first link in a causal chain. It's a necessary condition. But one has to account for the vast majority of users who don't develop problems. It simply isn't useful to abstract away people's agency (the thing that makes them human beings!) and treat their behaviors as "risks" that "cause" their other behaviors. What about a compelling video game? Or a significant other with toxic personality traits? Some fraction of people get addicted to gaming in ways that damage their lives, and some people find themselves unable to leave an unhealthy relationship. Is reading subversive literature a "crime of risk" because some fraction of people who do it will engage in violence? In such cases of political radicalization, it really is the literature that causes the person to become violent. It's far more useful to think of these behaviors as the willful actions of goal-directed, intelligent people, not random bit-flips in a robot whose machinery has been compromised.

(For an example of deliberate decision-making in drug addicts, see for example Carl Hart's research on crack addicts. He found that crack addicts were able to easily resist a free dose of crack if they were offered some other prize, such as a few dollars in cash or a gift card.)

Berenson has another major theme. He repeatedly says, "The pharmacological effects of drugs are the same whether they are legal or not." This is an incredibly weaselly kind of statement. It is literally true and utterly misleading at the same time. Yes, if someone is holding a syringe containing a dose of heroin, and suddenly the Supreme Court strikes down drug laws as unconstitutional (for once fulfilling their long-neglected constitutional duty), that dose of heroin will have the same pharmacological effects whether it's injected a moment before or a moment after the court's decision. So the statement is literally true. But it is a useless statement, because the legal status determines the distribution of drugs that are sold on the market. A harshly enforced drug war means that heroin and fentanyl will be preferred by dealers and traffickers. It's easier to hide and smuggle a single vial of fentanyl or carfentanyl than it is to traffic the equivalent dosage of heroin. And it's easier to smuggle pure heroin than it is to traffic large quantities of pain pills. Drug users prefer a user-friendly dose, while traffickers prefer to minimize (optimize anyway) their legal risks and maximize the number of effective doses per shipment. This is the classic observation that people prefer to drink beer and wine, but most alcohol produced during prohibition was hard liquor. It's not so much that the drug users' preferences change, nor is it true that distributors decide what users will consume and the user just has to accept it. It's rather that the economics change under prohibition. Traffickers could still ship beer during prohibition, but at a price that consumer aren't willing to pay.

Berenson apparently doesn't understand this. He implausibly insists that drug potency is driven by consumer demand. I think this is a rather pathetic dodge. He even acknowledges that alcohol prohibition caused distributors to specialized in more potent spirits, but insists that alcohol is somehow different. His explanation is that alcohol is "bulky", being dosed by the ounce, while other substances are dosed by the gram or milligram. This is irrelevant. Law enforcement routinely make seizures of many hundreds of pounds or even tons of cocaine or heroin. Traffickers certainly want to reduce the bulk of those shipments to reduce the chances of seizure, or to increase the bang-for-the-buck for a given size shipment. Fentanyl is about thirty times as potent as heroin by weight, so a one ton shipment of heroin can be reduced down to something that fits into a suitcase. (Alternatively but less plausibly, that one ton shipment can be made thirty times as profitable by trafficking in the more potent version of the drug.) The substitution of fentanyl for heroin is definitely driven by supply-side considerations and not by consumer demand. (There seems to be some debate about whether heroin users don't really want fentanyl at all or whether some of them have learned to use it carefully and thus prefer it. But it's clear that most of them would rather have heroin.) I feel obligated to call this stuff out whenever I see it, because Berenson is obviously wrong. Drug prohibitionists surely want to deny their responsibility for the carnage they've caused, and Berenson certainly tries to deflect blame here. But the huge spike in heroin and fentanyl deaths is 100% the fault of prohibitionists like Berenson. It is the absolutely predictable result of prohibition, which causes high and uncertain potency in the drug supply. That problem does not exist in the legal market, where alcohol and pharmaceuticals are clearly labeled. I think adults are morally responsible for the predictable consequences of the policies they advocate, and it's important to not allow them to wriggle out of their culpability. I have a hard time taking Berenson seriously here.

More important than potency is the reliability and predictability of the dosing. If drug users knew the exact dosage of whatever they were using, it wouldn't actually matter at all whether it was fentanyl or heroin. Professional pharmaceutical labs could measure out some particular dose of heroin or some equivalent dose of fentanyl. It's not the potency of fentanyl that makes it dangerous. In fact it's not inherently dangerous. The fentanyl patch is incredibly effective medicine, because it metes out a safe dose of the drug over time. The sheer potency of the drug itself is not relevant. It's the combination of high potency and imprecise dosing that makes it so dangerous. I believe Sullum points out that with alcohol, you know what you're getting because it's right there on the label. That becomes less true when we're dealing with a black market, where the traffickers are often amateurs and the products aren't meaningfully labeled.

Berenson points out that the "iron law of prohibition", that prohibition tends to increase potency, does not seem to apply to marijuana. I've dealt with that claim in a previous post. It appears to be true, but that doesn't mean that the "iron law" didn't apply to alcohol or doesn't apply to cocaine, heroin, and fentanyl. It makes sense to me that a marijuana supply that is constantly being disrupted by law enforcement won't be able to consistently produce high quality, high-dose strains of cannabis. Shutting down a grow operation is likely to destroy the institutional expertise that created that strain, and that knowledge is unlikely to be preserved and passed on. This is somewhat speculative; I don't know exactly why the economics of weed prohibition are different from the economics of heroin prohibition. But there is little doubt that traffickers prefer less bulky product, all things considered.

In another example of missing the point, Berenson responds to Sullum's long list of violent or corrupt actions by police officers. In his opening speech, Sullum mentions instances of drug dealers or innocent people being killed or injured by police raids. He also mentions several instances of corruption by drug cops. Berenson responds by claiming that Sullum is creating a distraction by focusing your attention on "an unfortunate few" police crimes. He's trying to cast these as isolated incidents committed by a few bad apples. Sullum's actual point is that drug law enforcement is inherently violent and corrupting of our institutions. Many of the violent actions by police officers aren't actually mistakes. They intentionally burst into residences where they think drug trafficking is taking place, often intentionally killing the dog or holding the family, often including children, at gunpoint while ransacking the house. Sullum doesn't make the following argument (unless I missed it), but drug policing intrinsically requires subterfuge and underhanded methods, because (unlike crimes with a victim) neither the seller nor the buyer is interested in cooperating with police. Police must rely on informants, who are often themselves criminals, whose criminal activities are known to and overlooked by police. Sometimes these informants are explicitly granted legal privileges in order to catch and prosecute drug traffickers who commit exactly the same offenses. As Sullum points out, some of these informants don't actually exist. (Ahem, Fuzzy Dunlop. Yes, it's fiction, but it happens in real life, too.) As the book Smoke and Mirrors by Dan Baum chronicles in great detail, drug policing has eroded our constitutional rights. Courts repeatedly ruled in favor of plainly unconstitutional searches and police surveillance, because for whatever reason they found drug enforcement so compelling. There are enormous constitutional carve-outs made explicitly in the "service" of making drug policing more feasible. This is clearly a consequence of trying to persecute a voluntary activity. Berenson's response to Sullum on this point is far too dismissive.

The beginning of Berenson's opening statement is rather bizarre. He explains that legalizers have been making the same arguments for decades, at one point saying that Sullum's speech could have been made 30 years ago. Legalizers have repeatedly made these same arguments as events have been unfolding to refute them, in Berenson's telling. He bemoans that drug warriors have been focusing on tactical battles without bothering to publicly explain why prohibition is a good idea. He says that prohibitionists have conceded the philosophical high ground to the legalizers. Supposedly the successful stigmatization of drunk driving, smoking, and pregnant drinking demonstrate the wrong-ness of legalization arguments. He points to the recent opioid crisis as an example of expanded access to recreational drugs leading to social harms, basically repeating (or implying without clearly stating) the standard narrative of the opioid epidemic. I've argued against that narrative in great detail on this blog, so I won't rehash it here. When Sullum gets a chance to respond, he points out that stigmatization is not prohibition! With the exception of drunk driving, the problems Berenson mentions were entirely a result of changing perceptions and norms. Population levels of smoking or pregnant drinking have fallen because of better information about risk and changing attitudes. I think Berenson is responding to a certain attitude among some legalizers who think there should be no stigma whatsoever regarding drug use. I'm sure some people hold this view. The libertarian position is that drugs should be legal, but private individuals can hold whatever beliefs or moral judgments they like about other people's hobbies and lifestyles. (I think Sullum shares this view, though I don't recall him saying so during the debate.) Anyway, this nice diversion actually makes a very libertarian point, which is that you don't need prohibition to massively change people's attitudes or behaviors regarding risk. I want to tell Berenson that prohibitionists have not "conceded" the philosophical high ground. Legalizers have taken it from them.

Berenson points out that legalizers have successfully framed the conversation to be about drug users, rather than the people around them who are hurt by the bad behavior of those users. He highlights a legitimate worry. There are certainly adults who use drugs and neglect their children or poison their relationships with other family members. Three point here in response. One, I don't think anyone has been allowed to forget about "the children." Helen Lovejoy political economy utterly dominates our politics. Two, I think Berenson would be hard pressed to find a penalty that deters drug users without harming their children. Penalties for drug use usually entail time in jail or humiliating drug court protocols. Any such penalty would make it hard to find steady work, and a drug conviction on your record is huge red flag to employers. Berenson seems hip to decriminalization, so maybe he's thinking we'd mostly avoid legal penalties for the users themselves. But there is no way to deter drug use without some kind of unpleasant penalty being applied to the user. Maybe we publicly cane drug users, then send them home and back to work? I think Berenson just hasn't thought this through. By the way, someone who is neglectful or abusive of their children is probably going to be non-responsive to the small risk of facing legal penalties for drug use. Three, alcohol. Alcoholics often neglect or abuse their family members, and it's legal. And that's a good segue to the alcohol discussion.

After both opening statements, Sullum responds to Berenson by saying that everything he says about the harms of "drugs" applies equally or even more intensely to alcohol. I think this is a really awkward point for prohibitionists, and no one has adequately answered it. Some just shrug their shoulders and say "Well, we're stuck with alcohol." Sometimes they argue (implausibly) that rates of heroin and cocaine use would be comparable to alcohol if we legalized these other substances, and the social harm from these drugs would grow in proportion. Berenson does a version of this. He attempted to claim that alcohol is less dangerous than the illegal drugs, saying "the reason there are so many harms related to alcohol is that so many people use alcohol." This isn't as straightforward as he implies; some 97 million people used opioids in 2015, with the vast majority of that use being medical. Certainly fewer users than alcohol, but the same order of magnitude. While I think it's a myth that recreational opioids were readily available over the past couple of decades, if Berenson believes this story he's got a mystery to explain: Why so few addicts? Why no increase in addiction or abuse rates?

Sullum challenges Berenson on the addictive potential of opioids, claiming the vast majority of medical users don't develop any kind of problem (which is certainly true). Berensons claims that if you gave everybody in the room a course of opioids, some would get addicted. Sullum conducts an impromptu poll of the audience, first asking how many had used prescription painkillers (which got a very large show of hands) and then asking how many had some kind of problem after their prescription ran out. Out of the huge number of people raising their hands to the first question, only a single person raised their hand to this second question. Not exactly a scientific poll, but 1) Berenson brought up the scenario of "giving everyone this room a course of opioids" and 2) Sullum has the relevant population-level statistics handy.

Sullum and Berenson converse about the relative dangers of alcohol and other drugs, with Sullum listing off some statistics on addiction rates per active user. He points out that the rate of addiction for prescription opioids per user (about 2%) is well below the rate for alcohol drinkers (~8%). Meth is slightly less addictive by the same measure. He concedes that cocaine and heroin are more addictive than alcohol (though not, in the case of cocaine, by a large margin). He points out that the substance that absolutely tops the list is tobacco. I think it's perfectly fair to point out that legal substances are more addictive and more damaging than illegal ones. Sullum also points out that it's not really addiction per se that we should be worried about. It's the harms from drug use that should concern us. Opioids don't do any kind of cumulative damage to the body, whereas alcohol damages the liver and tobacco damages the lungs (among other organ systems). Someone could be an opioid addict for decades and not suffer any harms from it. The only real danger is overdose and communicable diseases in the case of IV drug use, problems made much worse under prohibition. So contra Berenson, the legal recreational drugs are more harmful than at least some of the illegal ones. (Some illegal stimulants, like cocaine and meth, cause heart damage with excessive use. But Bolivian Indians chew coca leaf, and school children can ADHD medications, which are basically analogs of meth, around the clock for years, all without damage or escalation of use. So clearly the substances themselves can't be blamed. Other illegal drugs, like many hallucinogens and dissociative anastetics, are almost completely non-toxic.) I think Sullum is on solid ground in claiming that drug laws are arbitrary, meaning (as he clarifies) that legal status doesn't correlate with health risks or social problems. Berenson challenges his use of the term "arbitrary", but I think Sullum does an excellent job of defending his framing.

There is an interesting exchange about stoned versus drunk driving. Sullum points out that there might be some substitution away from alcohol and toward cannabis. Given that stoned driving doesn't seem to be as dangerous as drunk driving, the public health consequences of cannabis use might be a plus. Berenson is having none of this; he insists that stoned driving is just as dangerous as drunk driving, and cites some statistics about positive tests for cannabis among motorists in states that have legalized. Sullum points out that those studies don't actually distinguish between a mere positive test for cannabis metabolites and active impairment, because cannabis metabolites tend to stay in your system for a long time. Berenson insists that you can distinguish between active and passive metabolites.  I had heard that this was not true, though admittedly my source was probably a Sullum piece at Reason. Sullum points out that he did a lot of research on this and actually spoke with the experts, many of whom are concerned about stoned driving, and that they tend to agree that there just isn't a universal threshold for impairment with THC (as there arguably is with the alcohol). He wrote a feature piece (I presume this one) based on his findings. Sullum mentions controlled trials on driving courses, where impairment is measured in a very scientific manor, and contrasts these with population-level trends in accidents, which are multi-factorial and it's hard to tease out causation (and thus blame cannabis legalization). Interestingly, Berenson later confronts a question about declining violence in a time of drug law relaxation by saying that "crime is multi-factorial." He apparently agrees that social problems can be multi-factorial, but is quick to seize on any time series trend or state-level data that is favorable to his story.

I have seen attempts to "control for confounders" by comparing legalizing states to their non-legalizing neighbors. I discussed one such study here. There are several problems with these studies, including that the neighboring states are not good controls for the legalizing states (Idaho is like Colorado, except for legalization? Really?). Also, the proportion of the population turning from non-users to users is small, ~7% of the adult population. Blaming even a small increase in accidents on cannabis posits an implausibly large effect on their driving ability. A 3% increase in accidents attributed to cannabis implies that this 7% of the population has a 43% increase in their accident frequency! That is huge. I'm a pricing actuary; I would love to know about a risk factor that's this strong, but those are quite rare. More plausibly, there's little or no population-level effect of cannabis on accidents. This study, which used a synthetic control (basically creating "synthetic Colorado and Washington" out of other parts of the United States) and comparing it to legalizing states' actual experience, found no significant difference between legalizing states and non-legalizing states. Berenson mentions statistics from the Colorado and Washington DOI, but it's not clear if he's talking about the raw numbers (which are meaningless by themselves) or a rigorous attempt to tease out causality. My impression was that Sullum gets the better of this exchange, and Berenson's claims are highly speculative.

During the Q and A with the audience, Berenson clarifies his point about why/which drugs should be considered dangerous. He states that drugs which cause psychotic violence are legitimate targets for prohibition. He claims that this consideration absolves alcohol, which doesn't cause psychosis except in late-stage alcoholism, even while admitting in the same sentence that alcohol causes violent behavior. People fear psychotic violence because it's more random than other kinds of violence. That's true enough, but Berenson validates this irrational fear by saying, "...and they should!" I think this is bizarre. If alcohol causes  more violence and more socially destructive behaviors, I don't think it matters that the violence is technically not in the "psychotic" category. Anyway, it's not at all clear that alcohol-induced violence is categorically different from other kinds of pharmacologically  induced violence. I'm sure the spouse or child of an abusive alcoholic doesn't feel comforted by such distinctions. Rational policy should focus on objectively quantifying and minimizing harm, not fanning the flames of an irrational panic. I understand his point: People feel (irrationally) like they can control their exposure to non-random violence. They can avoid the violence of the black market by avoiding black market activities, they can avoid certain people whom they deem to be violence-prone. But they can't avoid exposure to the random whacked-out crack head or terrorist (to name another form of random violence toward which the public is irrationally obsessed). It might be defensible for public policy to put a thumb on the scale in favor of this preference, but Berenson seems to be dramatically discounting non-psychotic violence (or perhaps dramatically over-inflating the importance of psychotic violence as compared to other kinds).

Berenson says around the 36-minute mark that cannabis is "extremely neurotoxic", which would seem to imply permanent harm. But  he quickly follows up by saying it can cause "Extreme psychosis after a single use, but not permanent psychosis." I don't think he knows what neurotoxic means. (Is sleep neurotoxic, given that it causes vivid hallucinations?) To get a sense of what he's talking about, he describes someone who eats too many cannabis-infused edibles and get admitted to the ER because they think their friends are aliens. To be sure, this happens. Edibles can take hours before their effects kick in, so it's easy to take too much. Naive users will eat some, feel nothing, eat more, repeat, and by the time it kicks in they've taken way too much. But, also to be sure, the effects are temporary and (as he concedes by saying "...not permanent psychosis") no permanent damage results from these "overdoses." Psychosis seems like a hysterical term for someone who takes too many edibles and freaks out, even if it's technically/clinically accurate.

Around the 1:04 mark, Berenson claims that marijuana causes some users to become "extremely violent." Earlier, around the 39-minute mark, he attempts to blame a recent uptick in violence on cannabis legalization. This is just so contrary to almost everyone's experience with marijuana use. It tends to make people mellow and lazy, not violent and agitated. Putting together all of his comments, I think Berenson is claiming that some small fraction of people with pre-existing risk factors who take very large quantities can become violent or psychotic. But given the rarity of these risk factors, it's implausible that cannabis is responsible for a population-level increase in violence. It seems very unfair to punish everyone for cannabis use given that a very small fraction of people, using in a very irresponsible and deliberate manner, will experience a problem. Sullum at one point asks if there's really no way to distinguish between the problem users and the normal users who won't experience any problems. The later comprise the vast majority of users, so it seems unfair to impose restrictions or criminal penalties on them for the behavior of a small minority.

At several points during the discussion, incarceration is mentioned. A questioner from the audience asks about incarceration rates. Jacob Sullum points out that drug offenders actually aren't a huge proportion of the prison population. Some people who make the legalization argument exaggerate the magnitude of the problem, sometimes even claiming drug laws as the solely responsible for America's outlier incarceration rates. Sullum clearly points out that we'd have to be less punitive on violent crime to make a real dent in the incarceration rate. (Read John Pfaff's book Locked In; it's an excellent source of information on this point.) For sure we imprison a lot of people for drug laws, and if you think drug laws are unjust than even one is too many. But there's no question some people exaggerate these numbers; ending prohibition would leave the vast majority of our prisoners behind bars. Berenson says, "Yes, that's all true." And he's quick to add some numbers of his own, claiming there are only a dozen or so people in Pennsylvania prisons for marijuana offenses.

This is all fine, and it's important to set the record straight and get the numbers right. If some legalizers are exaggerating the costs of prohibition, prohibitionists should correct them. (In this case Sullum, a legalizer, is setting the record straight and rebuking some of his fellow travelers.) But it's curious that prohibitionists never seem to argue in defense of incarceration. They never say, "We don't actually incarcerate that many people for drug offenses, but we should!" Berenson himself says during the debate that the sheer number of cannabis users is so large (~40 million per year) that we're not going to arrest our way out of that problem. But if we're taking prohibition seriously as a policy, shouldn't we? Isn't it awkward to argue that we should have laws on the books that we don't actually enforce? Berenson seems to be pushing some form of decriminalization. When asked by the moderator, he clarifies that nobody should go to jail for smoking cannabis. He equivocates on harder substances; under his scheme cocaine and heroin users could face some kind of prison time. But he makes it clear that he prefers some kind of drug court for most users and criminal penalties for dealers and growers. An audience member, a mother whose son went to prison for growing marijuana, asks if her son's sentence was just. Berenson doesn't say a clear "No"; he indicates that some kind of criminal penalty is appropriate in that kind of case. Sullum points out the deep moral confusion in "decriminalization" regimes, that have strict penalties on dealers but little or no penalties for the users. He says that the people causing the actual problem are the ones who use drugs irresponsibly. Typically "aiding and abetting" carries a lower penalty than the crime itself. (I recognized this as Lysander Spooner's argument in his essay Vices Are Not Crimes.) I think that this is an extremely awkward point for prohibitionists, similar to the point about alcohol. I think they could claim that, as a practical matter, it's easier to go after a single source (a large cartel or drug gang) than a lot of smaller targets. But that's morally dubious. It reminds me of the Somali legal system as described in The Law of the Somalis. If a family member of yours murders someone, your family is obliged to turn him over to the family of the murder victim for punishment, otherwise the wronged family can kill someone else from your family in the murderer's place. One can see how this can lead to a good equilibrium, in which families discipline their own members and don't aid criminals in their midst in escaping justice. But the society has to pre-commit to actually punishing an innocent person to make this work. There is something deeply morally depraved about punishing innocent parties because it's "more practical" than punishing the actual wrongdoers. It's one of many compromises our society has made to make drug policing possible. If we tolerate this kind of thing, that leaves us all deeply compromised.

There are a few other interesting parts. Berenson at one point denies that marijuana has any medical benefits, though he quickly clarifies that CBD oil has been shown effective for certain kinds of seizures. He's all for that if it works, he says. Sullum concedes that some legalizers have exaggerated the medical benefits of marijuana, which is certainly true enough, but that he's at least slightly more impressed by the promise of cannabis. He mentions the nausea relieving effects for chemotherapy patients and (I think) wasting syndrome. Berenson is insistent that cannabis is not effective for pain control, because it has failed clinical trials. My own feeling is that marijuana is probably effective medicine for some conditions, but in ways that cannot be measured in randomized controlled trials. A common effect of a cannabis high is that every little nagging discomfort is magnified. As a friend once put it to me, "I was so sensitized, I could feel my socks." As in, we get habituated and inured to these minor physical sensations and annoyances, but smoking weed brings them back into focus. My initial response when I heard that some people used cannabis as a pain reliever was "No way!" But pain is a very subjective thing, and so is pain relief. Some people get no pain relief from opioids, while to others they are a godsend. It makes sense that marijuana would have a similar hetergeneity in its effects. In fact, it's even plausible that a majority people in a randomized controlled trial will feel their pain is exacerbated by cannabis, but that some fraction of them get relief. The study could yield a null result, even though it's effective for some and not for others, and even if the participants know whether or not its working for them. Berenson is actually hip to this kind of argument. Responding to a point about a study that found cannabis doesn't increase violent tendencies, he says (and I'm paraphrasing), "When you screen for people who don't have psychotic tendencies and give them a controlled dose of cannabis, you don't get violent behavior. That's not how cannabis causes violence." It's when you let cannabis loose in the world and those people with psychotic tendencies take large doses that you get violence, in his telling. He's appealing to the notion that people have heterogeneous responses to cannabis when it comes to the negative effects. The same principle applies to the positive effects. For something subjective like pain relief or the use of cannabis to control nightmares (to mention an anecdote I heard from someone I know), it's plausible that the drug has a large positive effect on some people, but not a consistent enough effect to pass a clinical trial. I think Berenson is too dismissive here and has too constrained a concept of "medicine."

Something that frustrates me about these debates on drug policy is that the prohibitionists never actually bother to defend prohibition. Berenson opens by expressing this exact frustration, but even he fails to mount a defense. That's not to say these people don't present evidence and give moral or logical arguments for prohibition. And I'm not saying that Berenson "fails to mount a defense" because his arguments are bad. I'm saying this: even supposing we grant all of Berenson's points, how do we add it all up? How do we know whether legalizing marijuana is a good idea or not? He claims that violent crime is up in the states that legalized first, and this is consistent with his story that excessive marijuana use in some people causes psychosis (and psychosis is an enormous risk factor for violence). Suppose the causal arrow is as he says: legalizing marijuana causes more violent crime. Okay. How much more? And how much is acceptable? How much do we weigh the enjoyment that people get from these substances? Berenson seems to acknowledge that drug users genuinely enjoy their drug habits, but it isn't clear what kind of policy/decision-weight he places on it. It's really not enough to point out that some social problems worsen after drugs are legalized (according to official public health statistics), even if one solidly establishes a causal link. To do so is only to compute some entries on the "cost" side of the ledger. We have to weigh these costs against the benefits. This is not a mere exercise in computing statistics and performing causal inference to yield an effect size. One inherently requires some kind of philosophy, a moral philosophy, to weigh costs against benefits. Drug prohibition fundamentally robs us of our bodily autonomy. It puts people in cages when they haven't violated anyone's rights (in Sullum's formulation, which I find very compelling). It reassigns ownership of our bodies to the government. One could do a cost-benefit analysis by converting man-years of incarceration to a dollar figure and comparing it to the avoided harms of drug-related social problems, also converted to dollars. Frankly I find this kind of social engineering a little bit creepy. I favor legal gay marriage, and for that matter interracial marriage, for reasons that have little to do with such a cost-benefit computation. I would do so even if a rigorous analysis went the other way. There's something to be said for freedom being a value unto itself. It's not just a general meta-rule that tends toward social optimization. (Though it is that, too. People who don't value freedom in and of itself often fail on their own terms. Policy proposals are often bad by objective measures that any reasonable person cares about. The kind of dry cost-benefit analysis I'm describing in this paragraph is still important, because we need to be able to engage intellectually with people who don't hold our values.)

I hasten to add that such a computation almost certainly favors legalization of all classes of drugs. We have very good theoretical reasons to doubt the wisdom of drug prohibition. Prohibition raises the cost of a good, because users face legal penalties and supply-side interdiction raises the price, search-costs, and harms of the substance. So users buy less of it. When demand for a product is inelastic, doubtless an accurate description of addictive drugs, then prohibition necessarily increases the total costs of drug use. There are fewer users as we increase the penalties, but the costs paid by the continuing users increase faster than the costs saved by deterring users. The paper The Economic Theory of Illegal Goods: the Case of Drugs by Gary Becker et al. is the definitive treatment of this topic. I did a write-up of it here, and here is a link to the original paper, which is quite readable. Even if you suppose that drug related externalities are very large, those "externalities" are almost invariable behaviors that are already criminalized. There are already criminal penalties, so the costs of these behaviors is already internalized. Drug users already contemplate them and build them into their decision to use. (If you are going to respond by suggesting that drug users don't respond rationally to the prospect of criminal penalties, then my follow up will be to point out that drug prohibition is a series of criminal penalties. If anyone is conceding that these don't work in general, that's a major concession to the anti-prohibition side of this argument.)

Consider also the supply side. Prohibition of a good with inelastic demand also increases revenues for dealers. Quoting from Tyler Cowen and Alex Tabarrok's economics textbook (a fuller version of the quote is excerpted here), "When the demand curve is inelastic, increasing the price increases the seller's revenue." There's no such thing as "push harder" or "let's just be smarter about this." The more you push up the price, the better funded the drug traffickers, and the better equipped they are to hire mules, invest in law enforcement evasion, bribe officials, they engage in technological innovation (such as fentanly-producing labs which, once again, makes smuggling easier). See the supply and demand curves in this article by Benjamin Powell for a good graphical description of what's happening.

Advocates of prohibition are using a very shallow application of economic theory: when you increase the price of something, you get less of it. That's true enough as it is, but presumably we care about the social costs of drug use, not drug use per se. Taking the demand-side and supply-side economics of drug prohibition together, it's almost certainly a terrible policy. One needs to make heroic assumptions to rescue it.

Monday, November 18, 2019

Legitimate Pain Patients Can Look a lot Like Addicts

This is a point I made a while ago in another post, but here it is again. There is no clear way to reliably distinguish between a real pain patient and an opioid addict who is seeking to use pain medicine recreationally. They show up to their doctor, they describe their various ailments, and they explain that a certain medicine give them relief (perhaps even specifying which opioid worked for them). Even supposing it initially were the case that pain patients all exhibit Behavior X while opioid addicts don't, people are smart. They learn to game systems. It will become very clear very quickly that doctors respond better to patients who exhibit Behavior X, and recreational users will learn to fake it. This doesn't mean doctors should start seeing Behavior X as a telltale sign of addiction. It's not! It's a behavior of real pain patients that recreational users have learned to mimic! Treating Behavior X as a sign of addiction will lead to lots of genuine pain patients being denied opioids. Any such screening process is bound to be anti-inductive, meaning the harder we try to understand it the harder reality will try to thwart our efforts.

See this recent Slate Star Codex in which Scott Alexander describes several patients who were treated by their doctors like drug addicts. He makes clear that these are stylized examples, composites created to protect the identities of his patients. But plainly this kind of thing happens all the time. It's tragic. Sometimes the patient is far more of an expert in his/her own medical needs than their doctor, having fine tuned their own care in a trial-and-error process. It's pretty clear that many doctors are shirking their responsibility to treat their patients because of legal hazards. My own summary of Alexander's post is that there is no avoiding the false positives/false negatives trade-off. Any net you build to catch addicts is going to ensnare some legitimate patients, who perhaps make the "mistake" of learning how to fine tune their own medications and ask their doctor for those medicines by name.

I read a lot of literature about drug policy, so I routinely come across "opioid epidemic" stories. It irritates me when people include throw-away lines about how we've "prescribed too many opioids." Even authors who are very sympathetic to pain patients, who are perhaps also sympathetic to drug decriminalization and harm reduction strategies, often preface with something like, "While it's certainly true that we've prescribed too many opioids in recent decades..." They may even point the finger at Purdue Pharma or the Sackler family as the corporate malefactors in the opioid crisis.*  I want to push back really hard on this concession and say: No, we did not prescribe "too many" opioids. We were prescribing too little in the 80s, then, mercifully, attitudes toward pain treatment changed. For the reasons described in the top paragraph, there is no clear discriminator that neatly separates pain patients from addicts. The phrase "too many" seems to imply that the addicts and the pain patients were easily separable ex ante, and we could have just positively identified the addicts and sent them home without prescriptions. It's only "too many" in an ex post sense, in that we knew only after the fact when someone was a poor candidate because they in fact became an addict or had a fatal overdose. If there was no ex ante way to discriminate, then it wasn't a mistake to prescribe opioids, and in no meaningful sense did we prescribe "too many."

For these reasons, I think we should basically give opioids to anyone who wants them. Any kind of screening process will inevitably leave some desperate pain patients untreated. Besides, we now know the consequences of denying prescription painkillers to recreational users; recent crackdowns and supply restrictions have led to a massive surge in the use of illicit heroin and fentanyl and a huge increase in overdose deaths. Restricting access to the relatively safe, standardized, factory-produced opioids isn't helping addicts. It's driving them to more lethal behaviors.

It is very clear that it was too difficult for chronic pain sufferers to get pain relief until the late 1990s, and even afterwards there was a lot of law enforcement scrutiny and even prosecutions of pain doctors. When pain patients are forced to taper off their opioids, perhaps because their doctor feels compelled to follow the CDC's misguided guidelines or perhaps because their doctor was targeted by the DEA, they often commit suicide. This is a history of too many restrictions on opioid prescribing.

I think drug reform will necessarily be a joint effort of libertarians and left-leaning harm reductionists. I want to say something to the left-leaning members of this coalition: stop falling for this "corporate malefactor" narrative of the overdose crisis. It makes you look impressionable. Producers and distributors are not morally responsible for the misuse of their products, and anyway they are not causally responsible for the recent surge in heroin and fentanyl overdoses. This ideology that lays blame on producers denies agency to patients (and more generally to consumers of non-pharmaceutical products). By the way, making money is not a sin. The profit motive doesn't inherently taint every human transaction in which money changes hands. If we as a society want to accomplish something, the best way to ensure it gets done is to allow someone to make a living doing it. There is no way for society to have a safe, legal supply of drugs if we're going to turn around and sue every manufacturer and distributor out of existence. I want to ask, "What exactly do left-wing harm reductionists have in mind?" Shall we keep the production and distribution of opioids in the black market, but have enough supervised injection facilities to accommodate drug users? There won't be any legitimate opioid manufacturers if we don't let them legally earn some kind of return on their investments. How do we accomplish harm reduction without a legal supply of opioids? Should the government sue every manufacturer and pharmacy because some small fraction of patients end up harming themselves? What kind of communication should opioid manufacturers be able to have with doctors and patients? Will any attempt to communicate information be interpreted post hoc as "deceptive advertising"? If manufacturers believe that the risk of addiction has been wildly overestimated, and they can support their beliefs with published research (which in fact tends to find low addiction rates for opioid patients), should they be forbidden from arguing their case? Should chronic pain sufferers be denied such a public advocate just because profit supposedly impugns their motives? Should the government's anti-opioid crusades (as carried out by the CDC and the DEA) have no counterweight, except for the unpaid volunteers who are motivated enough to push back? Should pain specialists (who also make a living supplying opioids to users) be subject to prosecution because some small fraction of their patients are abusing? Any one of these doctors with a large enough patient base probably "knows" that some of their patients are abusing, by sheer chance. It's just that they've made the decision to treat pain using some kind judgment about who is or isn't likely to be a real pain patient, applying some kind of plausibility threshold, and yielding a "treat" or "do not treat" decision. Should the authorities be able to second-guess their threshold? Any application of the principle that "these doctors should have been more discriminating" leads to pain sufferers being denied relief. I see a lot of opprobrium aimed at parties who make money, and it has very little to do with their moral blameworthiness. Frankly, it has more to do with the "bigness" of those parties and anti-corporate economic populism. The public loves to hate a corporate villain. Some sloppy media organizations and journalists have served the people what they want to hear. Don't fall for it. "News" stories that get the most clicks and media narratives that become accepted wisdom among "very serious people" aren't necessarily true.

______________________________________________
*There is even a push to blame the pharmacies who fill the prescriptions. This game of blaming the intermediary is absurd, considering that their only role is to actually fill a prescription that a doctor and patient decided was necessary. It should not be the pharmacy's business to object to such decisions. The bigness of Walmart, CVS, and Walgreens makes them compelling targets for a government shakedown of for left-wing moral opprobrium. That doesn't mean the opioid crisis is their fault. By all means, allow them to check for conflicting prescriptions and warn about drug interactions. But it makes no sense to blame these intermediaries for the sheer tonnage of opioids they handled.

I wasn't sure where to insert this in the main body of my post, so I'll put it down here. I know someone on Facebook who is at Cato and writes a lot about drug policy and the opioid crisis. He often shares stories about the opioid crisis. One of his friends, a lawyer, often pipes up to say that he sees cases in his practice of "obviously" negligent prescribing behaviors. He sues doctors for prescribing to people who "shouldn't have gotten them" (in someone else's estimation). His comments make it sound like these are open-and-shut, slam-dunk cases of inappropriate prescription. Obviously, this guys description of what he does for a living is going to be self-serving. I think it's hard to appreciate the ex ante principle, that it can be impossible to determine ahead of time whether or not something is a mistake. He has the advantage of filtering for instances where some medical misadventure has in fact taken place, and he get to craft a narrative post hoc about how obviously this patient should never have been prescribed opioids. I really hope he's losing a lot of these cases, because it's crap like this that makes it harder for pain patients to get relief. There are costly errors in both directions, but only the false positives become visible to him. This isn't even a government policy, just an out-of-control tort system that compels doctors to practice defensive medicine. Like I've said before, libertarians should spend some time thinking about how bad institutions can arise and damage society, even when it has nothing to do with a central government. If bottom-up civil society can mimic the harmful policies of a centralized government, we should worry about that, too.