Saturday, February 24, 2018

Response to VerBruggen: Law and Liberty Forum On Opioids

Robert VerBruggen wrote the lead essay in the Law and Liberty forum on the opioid “crisis.”

I won’t summarize it, but you can read the essay to get some context. To me it’s just the standard narrative of the opioid epidemic: we loosened prescribing practices on opioids, which in turn created new addicts, which in turn led to lots of drug overdoses. I posted a debunking of this story here.
In that piece I tried to refute the narrative that VerBruggen is selling to his readers. I could just post a link to that piece and say, “Moving on.” But it might be more constructive to respond specifically to his piece.

The very first sentence of the essay gets us off to a bad start:
America has seen a veritable explosion of serious drug abuse over the past two decades.
Not it has not. Opioid abuse has been flat since at least 2002, even declining in recent years. Heroin use has increased in very recent years (we’ll get back to that later), but as I have pointed out repeatedly these are two different trends and should be treated very differently.  It makes little sense to combine heroin/fentanyl overdoses (a big risk applied to a small population) and prescription opioid overdoses (a small risk applied to a very large population) and call it “the opioid epidemic.” 

Maybe VerBruggen would take issue with the government statistics on prescription opioid abuse and addiction. Maybe there’s some kind of reporting bias that’s masking the trend or something. I’m fine with playing the “let’s doubt official government statistics because they might not be accurate” game. In fact, let’s play that game with the drug overdose statistics.

I argue here and here that it’s difficult to actually assign a cause of death to an individual body. I suspect that a lot of medical examiners are writing down “drug overdose” because it’s a handy explanation in some cases where the actual cause might be harder to discover (or inherently ambiguous). But don’t take my word for it. Do pick up a copy of the Karch’s Pathology of Drug Abuse textbook. I provide several excerpts in which he cautions the reader as to just how hard this problem is. In fact, I had to pare down my list of quotes for those posts because it was getting very long.

I have scoured the drug overdose data in great detail. It’s the same dataset from the CDC that VerBruggen is getting his numbers from. I find a lot of irregularities and a lot of indications of a spurious trend. “A body is a body, so there can’t be a spurious trend in death counts” one might be tempted to say. But the assigned cause of that death is malleable and subject to various guesses, mistakes and reporting biases. I think that is part of the story. Opioid prescriptions did something like triple from 1999 to 2015. If three times as many people are walking around with opioids in their bloodstreams, that’s three times as many opportunities for a cardiac arrhythmia or other mysterious sudden death to get marked down as a drug overdose. Or perhaps the death isn't "mysterious" in the sense of lacking an obvious cause, but rather has many contributing causes competing for the top spot of "underlying cause." Indeed, a typical drug decedent is sick and old, compared to the average user who is much younger and healthier. These CDC death records can list up to 20 contributing causes of death; if any of those are drug-related it is almost always labeled a "drug poisoning." You rarely see it going the other way, where "sleep apnea" is listed as the main/underlying cause with "other opioids" as a mere "contributing cause of death." I’m not literally claiming that the full rise is explained by miscoding deaths, but some large fraction likely is. I point out in this post that there were zero drug overdoses in certain states in 1999 (based on my filtering/counting which I think is defensible; see post for details). That seems implausible to me. It looks a lot more like they just weren’t looking for those things, until more recently when people (specifically medical examiners) caught on that this was happening. (Death codes changed over from the ICD-9 to the ICD-10 codes in 1999, which plausibly kicked off a spurious trend. As in, "Oh, there's a code for that now!")

VerBruggen takes the same approach that I criticized in a German Lopez piece: He starts with the total number of drug overdose deaths (“over 50,000”) and then tells his readers that 2/3rds involve some kind of opioids. Why start with an over-count, then force your readers to pull out a calculator (or open an Excel workbook) and multiply by 0.67? I suppose he sort of justifies this by saying, “In the official statistics that year, nearly two-thirds of drug overdoses involved an opioid of some kind—an undercount since many overdoses are not properly coded as opioid-related.” Alright, so miscoding causes of death is a problem? I agree. It’s throwing off official death statistics? I agree. But let’s count errors in both directions, or if we can’t estimate those errors admit to a little more skepticism.

By the way, the "over 50,000" thing is kind of a pet peeve of mine and it screams "sloppy." You can only get to "over 50,000" if you count suicides. See my workup of the 2015 data here. If we only count unintentional overdoses, the number is more like 44,000. About 15% are either suicides (mostly) or "undetermined intent" (a smaller proportion), assuming once again that the medical examiner was correct. Fifteen percent isn't a huge correction, but adding in ~8,000 deaths that have nothing to do with the problem you're writing about is hard to defend. You could tell a story such as "These suicides happened because the decedent was a miserable addict and couldn't take the shame anymore," but this is getting speculative. You don't get to count suicides as part of "the opioid epidemic" just because the physical/chemical cause of death was the same. Anyway, this is all slightly pedantic, because unintentional opioid overdoses are clearly high and rising.

VerBruggen introduces the book Dreamland by Sam Quinones, which is another presentation of the standard narrative (this time in book-length form). I found Dreamland to be frustrating and mostly useless for all the reasons my regular readers will be familiar with.
[I]f, by contrast, the abusers of opioids are mainly individuals who abused them from the beginning, then the epidemic is a strong indicator of the consequences of entirely legalizing drugs. It is a demonstration of what happens when drugs are in full supply, safely manufactured, and easily available to those seeking a high and at risk of addiction.
I’m sorry, but this is just wrong. The past 20 or 30 years are not an example of drug legalization. It’s not an example of relaxing prohibition, even on a specific class of drugs. Does anyone think they could just ask their doctor for opioids and get them? Of course not. Doctors are extremely suspicious of people who come to their offices complaining about pain, more so people directly requesting opioids. I hear media accounts of “pill mills” handing out too many pain pills with little oversight, but this is mostly a fringe phenomenon. Considering the number of severe chronic pain patients who can’t get opioids (some of whom commit suicide), considering that Cato was writing about an anti-opioid crack-down as early as 2005, considering that Jacob Sullum was writing about this paranoia as early as 1997, I think I’m on solid ground saying that obtaining opioids was never easy. (Sure, you could pick nits about how these are "libertarian" sources, but 1) the underlying information is still relevant and still stands and 2) who else would you expect to be writing about this kind of government overreach?) VerBruggen's presentation of the recent decades as a slackening of legal restrictions on recreational use of opioids is just wrong. Sure, the sheer tonnage prescribed increased by a factor of 3 or so. That doesn't mean it's easy to get your hands on some. 

(Anecdote: I was in the E.R. in a Columbus, OH hospital in 2006. There was a lady in the waiting room complaining vaguely about pain, presumably trying to get a prescription of opioids. This isn't much of an anecdote really, because everyone in medicine knows this kind of thing happens all the time. Are most of these people successful? If opioids are so freely available, why do they go to so much trouble?)

Much of the recent increase in opioid prescriptions is for acute pain after a surgery or after an accident. Maybe this mildly increases the number of “pills in circulation” (a term I’ve heard used to describe the passage of pills from legitimate patients to the black market). But what’s going on here? These unused pills are very dispersed, sitting a few per bottle in separate medicine cabinets in separate homes. Do a large fraction of these really make it into the hands of addicts? Do the patients themselves sell them? Do addicts have access to enough dispersed medicine cabinets to support a habit? (As in, can they visit the homes of enough friends and relatives, get access to their master bathrooms, and steal enough opioids to support a habit?) I’d like someone to better fill in this part of the story, because it seems pretty implausible to me. 

As to this: “…then the epidemic is a strong indicator of the consequences of entirely legalizing drugs.” No, of course it’s not. There is a long list of essentially harmless drugs that we could legalize. If there is any amount of substitution away from alcohol, tobacco, opioids, and cocaine (the biggest killers), it is worth legalizing those. I’m not sure to what extend “potential opioid users” and “potential psychedelic” or “potential dissociative anesthetic” users overlap, but there would surely be some people who are satisfied with the harmless substances and would thus not bother with more dangerous highs. I don’t understand how this sentence even made it into his essay, because in the very next part he begins to discuss marijuana legalization. This should have been an opportunity to pause and reflect on ways to divert people away from opioids by reducing the cost of other options.  It is unlikely that total intoxication would rise in a society of full legalization; rather people would choose the least costly forms of intoxication (least costly as measured in health and time costs as well as dollar costs).
[T]hose opposed to even the legalization of marijuana at this point have to admit that the public is not on their side. Support for marijuana legalization has crossed the 50 percent threshold...When it comes to legalizing hard drugs, the case is otherwise: Very few Americans are in favor.
Such surveys aren’t much good if people are constantly being misled by sensationalist drug journalism. Read Dan Baum's Smoke and Mirrors for a full account of this history. Every new drug "epidemic" turns into an overblown moral panic, usually before there are any vital statistics or any kind of  hard data supporting it. VerBruggen is careful and mild-mannered in his approach. He's hardly a sensationalist himself, and he has looked into the numbers. But his reporting on this story is still misleading.

VerBruggen has two paragraphs on a cultural change in medicine regarding the treatment of pain, which led to the rise in opioid prescriptions. Thank goodness this happened. Some severe chronic pain sufferers have tried every option, and the only thing that works for them is high-dose opioids. VerBruggen to his credit gives a nod to this:

Few would deny that for many patients, pain is real and can devastate their quality of life.
What I think he fails to address (perhaps doesn't realize) is that any attempt to restrict opioid use will condemn some of these chronic pain patients to a life of uncontrolled pain. The false positive/false negative trade-off is fundamental. It is a sheer statistical fact. There is no way to side-step it. I wish that people in the opioid alarmist camp would state more clearly how comfortable they are with letting chronic pain patients suffer. It doesn't work to just say "Do better screening." We already have the most educated people in society (doctors!) making the decisions regarding who gets what medicine, based on detailed medical histories. Even if we're willing to spend a lot more of society's resources coming up with a better screening mechanism for sorting out addicts, we're already way past the point of diminishing marginal returns. There are two possible ways to reduce "unnecessary" opioid prescriptions: better screening ("our algorithm for predicting who's a pain patient vs. who's an addict is more accurate"), or a stricter cut-off ("for a given level of accuracy, change the threshold by which we label people 'true pain patients' vs 'recreational users faking it'"). The first is unlikely to work at all or to scale up if there even exists a better screening algorithm; the second cuts off more legitimate pain patients. My challenge for opioid alarmists: acknowledge the trade-off, and describe how comfortable you are with cutting off pain patients who are actually suffering. (Some people in high places are disturbingly comfortable with cutting off pain patients and making them "tough it out.")

VerBruggen cites a letter published in the New England Journal of Medicine that supposedly kicked off the change in attitudes toward opioids. Dreamland by Sam Quinones also makes repeated references to this letter, asserting or implying that the message in it was wrong. The letter states that only 4 out of 11,882 patients with no prior history of addiction got hooked on opioids after being administered. "We conclude that despite the widespread use of narcotic drugs in hospitals, the development of addiction is rare in medical patients with no history of addiction." Dreamland repeatedly presents this letter as some sort of mistake (VerBruggen's piece simply cites it without necessarily saying it was wrong, though that is the overall tilt of  his essay). The sample wasn't representative, it was from a time when opioids weren't readily available on the streets (really?), most of these patients were administered far milder doses than what's given today, etc. But guess what? The conclusion holds up extremely well. Only about 1% of opioid-naive post-surgery patients show any sign of misuse after being administered opioids; presumably the rate of addiction is far lower than this.

VerBruggen cites a crackdown on pill mills in Florida circa 2010, supposedly leading to a decrease in opioid overdoses. Again, I find this implausible given all the accounts of crack-downs on prominent, legitimate pain doctors going all the way back to the early 2000s. By what objective metrics was there a "crack-down", and when did it start, and what was the level of enforcement before the crack-down? Regression analysis? VerBruggen knows about these. It's an appropriate tool for quantifying the response to policy.

He has two paragraphs concerning the rise in heroin overdoses circa 2010 and the tainting with fentanyl, an extremely potent synthetic opioid (~50 times stronger than heroin). He does here point out that it's a very tiny fraction of former opioid users who switch to heroin, and that a heroin habit is much deadlier. Indeed, I've calculated that a heroin habit has an annual mortality rate of about 4% per year (admittedly this is based on probably imprecise or biased information about how many users there are and how many of them die). This again would have been a nice place for VerBruggen to insert some doubt about his proposal to keep drug prohibition in place. There is basically no market demand for fentanyl. Heroin users avoid it if they can, but in a black market it's impossible for them to know what they are getting. Jeff Miron makes this point in his Law and Liberty forum essay, but VerBruggen offers a pretty weak non-response to this point. Like I said in my response to Caulkins, this is absolutely the fault of drug prohibition. There is simply no way such a dangerous product would be sold, unlabeled and at such a dangerously high dosage, in a legal market. Prohibition advocates are responsible, but they keep trying to take a pass on this.

On the age distribution:
Opioid deaths have long been concentrated among the middle-aged but, in the past few years, young adults have been catching up thanks to heroin.
It's curious that he'd point this out, because it's something I noticed too. By the way, the average opioid death happens at about age 44, while the average heroin death happens at about age 38. By contrast the modal (peak-of-the-distribution) opioid abuser is in their 20s, with a mean age of 33. I think there are important policy implications here. Let's discuss risk factors, rather than just blandly issuing a blanket prohibition with all its nasty side-effects and a full-fledged black market. If most of the deaths are happening when people are older and more infirm, that suggests being cautious with people who have those risk factors. If you have a condition that makes breathing difficult (obesity, apnea), taking a medicine that slows your breathing may be dangerous. A targeted approach is called for, because clearly some populations are at great risk and some populations of users aren't. By the way, check out the chart in my link at the start of this paragraph, in a post I titled "Age Distribution of Legitimate and Non-medical Opioid Use". Notice anything? The age distribution of legitimate users more closely aligns with the distribution of deaths. This could imply that the deaths are coming from legitimate users, who occasionally (meaning very rarely) mix their opioids with other medications or alcohol, or simply take too much. Or it could simply be a product of older, sicker people being more susceptible to drug overdoses, or it could be a symptom of the "misdiagnosing the cause of death" problem I discussed above. Many interpretations are possible, but you may reach different policy implications depending on what you think the true explanation is.
And now we arrive at the key question: Are we talking here about pain patients—those who have inadvertently become addicted, so much so that they eventually take a fatal dose? Or are these folks non-medical users, who got hooked through deliberate misuse?
This is an excellent question, and kudos to VerBruggen for asking it.  Again, though, the policy implications are very different depending on how you answer this key question. Suppose it's the former. Well, as stated above it is a very tiny fraction of pain patients who even misuse their prescriptions, let alone become addicted or overdose. Are we willing to let 100 people suffer because one of them might take the drug recreationally? Or if we put it on the basis of avoiding a single incidence of addiction, are we willing to make even more people suffer (say, about 500 if these proportions are correct) to avoid a single addiction?

Or suppose it's the latter: prescriptions don't turn people into addicts, but existing addicts get a hold of loose pills from legitimate prescriptions and "doctor shop". In this case, restricting access to opioids seems even more deranged. "I'm sorry, you can't have the opioids that would spare you a few days or weeks (or months or years) searing pain, because someone who enjoys taking them might get a hold of them." I don't think such a policy would be fair or moral, and, make no mistake, fairness and morality are important considerations. I sometimes see this technocratic, quantitative-without-theory "public health" approach to the opioid issue that seems to want to decrease the body count at all costs. I don't think that's a worthy goal. You need some way of measuring and quantifying these costs, which requires a bit of theorizing.

He follows up that passage with a discussion of literature, suggesting that most drug abusers start out having had a legitimate prescription. None of the studies he cites establish causation, though admittedly some may be suggestive. Presumably most of these people started out having tried alcohol or coffee or marijuana, too. With something like ~85 million past-year users of legitimately prescribed opioids (the lifetime number would be much higher) and ~36 million people having abused prescription opioids at some point in their lifetime, these are huge populations.
To be absolutely clear, 41 percent of addicts’ having started out as medical patients with legitimate prescriptions is hardly a small percentage...
To be absolutely clear, at least 41 percent of almost any population will say they've had a legitimate opioid prescription at some point in their lifetime. And just over 10% will say they've abused opioids at some point in their lifetime. We're starting with very high base-rates, folks. So studying past use as a risk factor has some problems. Causation is not established here, only a mildly suggestive correlation, a post hoc kind of argument.
Doctors must prescribe these pills less often without denying relief to people who really do suffer from extreme pain, as they have already started doing. 
Once again, this is simply not possible. The false positives/false negatives trade-off is real. Let's draw out the curve and restrictionists can tell us what part of the curve they feel comfortable with. But let's stop pretending there is no trade-off to be negotiated. (Amusingly, VerBruggen lectures Miron about trade-offs regarding alcohol legalization in the follow-up essay, which we'll explore later.)
But since so much of the problem stems from blatantly illegal behavior rather than the misuse of legitimately prescribed drugs...
I'm actually not so sure. Once again, by several different measures there are no more opioid abusers, despite something like a tripling of prescriptions. An alternative explanation is that, of the 85 million or so legitimate opioid patients, some extremely tiny fraction of them occasionally slips up and mixes medications. Now, clearly many of these overdoses are from addicts or even non-addicted recreational users, but we don't know for sure in what proportions. This is another distinction that has important policy implications. If the rise in opioid deaths is mostly coming from legitimate pain patients, then what we're seeing is a roughly fixed risk (again, a tiny risk) of overdose simply being applied to a larger population. I don't think that's worrying, and should not spur a moral panic about drug use. It would be like worrying that there are "three times as many back surgery deaths" in an era when back surgery expands three-fold (or suppose driving or trampolines or ATVs or some other activity with a low-but-real risk expands three-fold). Anyway, the "much of" part in that phrase is a hedge, so I'll let it slide as technically true.

He launches into an exposition about "how the drug war works."
The federal system holds very few drug prisoners who weren’t involved in trafficking; in state prisons, only about one-quarter of drug prisoners were sentenced for possession...
I see a lot of this kind of stuff in drug reform debates, downplaying the number of people in prison. It's still a very  large number of people whose lives are ruined for no good reason. Even if it's a small percentage of the total population of prisoners, it's still big. I see a lot of people poo-pooing this as "not the solution to mass incarceration". Fair enough, but it would be a big step.
 In fact, arrests related to possession are more than quadruple those related to dealing. 
He says this in the context of denying that mere drug users are imprisoned at significant rates, but in fact this shows that these people are indeed harassed by our legal system. More so than incarceration statistics would show. Let's count this harassment in our calculus of whether the war on drugs is worthwhile. It is a cost. Let's add to it all the people who are needlessly harassed without getting arrested, and let's account for the forms of harassment that are far more severe than imprisonment.

VerBruggen seems favorable to marijuana legalization, though he raises some unnecessary notes of caution about a harmless substance.
 Marijuana legalization is not consequence-free, to be sure. It drains users of their initiative and, notwithstanding its reputation, it is habit-forming.
Citation, please? "Habit-forming" is a weasel phrase. It's not chemically addictive in the sense that opioids are, or in the (quite different!) sense that cocaine is. Anything can technically be "habit-forming."
It is not as addictive as opioids and seldom if ever kills those who smoke it.
Ugh. Just say "never." Sorry for seeming confrontational on something where he and I agree on the policy, but half-hearted endorsements of policies that are obviously good still bother me. If someone is overstating the costs of a policy, I'm going to correct them even if they still reach the same conclusion as I do.
Regarding treatment, we should have no illusions. Many addicts refuse help. We know of no treatment that is effective more than about half the time. And treatment is expensive. 
I mean this next part earnestly: Bravo, Robert! I've said something similar. Let's not pretend that "more money for treatment" will magically cure all addicts. I think his statement implicitly recognizes free will; an addict who "refuses help" is actively choosing to remain an addict. I think we should all admit that some large fraction of drug addicts are like this, in addition to there being a large population of addicts who desperately want to recover. By the way, VerBruggen should read Maia Szalavitz's book Unbroken Brain for a sort of contrarian take on addiction (I say "contrarian" but it's ultimately more plausible than the "drugs dominate the will" model of addiction).

In the same section a couple of paragraphs later, he seems favorable to drug courts. I am skeptical. See the link above to my post on "Drug Courts and Drug Treatment." Drug courts may be a (somewhat more) humane alternative to incarceration, but they are still oppressive and infantilizing to the people who attend them. Assuming we legalize substances that aren't inherently dangerous and limit drug courts to serious addicts who are actually causing problems (a tiny minority of drug users), I'm okay with them.
Addiction medications have proven to be highly effective, if far from 100 percent so. These include methadone, buprenorphine, and naltrexone.
Or how about let them buy pharmeceutical grade heroin, of a known dosage, unadulterated by super-opioids like fentanyl and carfentanil. Maintenance therapy works. It works by keeping the addict alive for the ten years or so (a typical tenure) that it takes to age out of their habit. Opioids don't do cumulative organ damage like cocaine or alcohol, so the addict can come out of this in good shape once they decide to clean up. Once again, VerBruggen could learn a lot by reading Maia Szalavitz. Anyway, he seems favorable to replacement therapy, so I'll count that as a win.

He also seems favorable to needle exchanges and safe injection facilities, and I applaud him for this. But how about we take these things out of the legal grey-zone they are stuck in? Just legalize them and put them in the plain light of day. The main driver of heroin overdoses is once again the fact that the user doesn't know what they are taking. (Also, recent release from prison is a huge risk factor; this is yet another way in which prohibition exacerbates the drug overdose problem.) Now, it's probably wise to direct heroin users to some kind of supervised facility even in a legal market in case they imprudently take more than they intended to. But this risk would be much lower in a fully legalized market, in which people buy their heroin from a pharmacist rather than a high-school drop out who doesn't even know what he's selling.

He ends his essay with the following flourish:
In short, the opioid epidemic has dispelled a lot of myths and quashed a lot of hopes regarding the War on Drugs. Legalizing drugs looks much less wise than it once did and, in any event, has little chance of gaining public support. A much narrower, but still powerful, set of reforms is in order.
Not really. In no meaningful sense has legalization been tried, and thus in no meaningful sense do the past 25 years tell us anything about the likely consequences of drug legalization. Drug users are not getting their heroin from legal, transparent supply chains. Recreational users are not discussing their intentions with pharmacists or doctors, who might caution them about various risk factors. Contra VerBruggen, there is most certainly a safe way to consume drugs, even the ones that are potentially dangerous or addictive.

Jeffrey Miron replies, and VerBruggen shoots back.

He invites us to turn back the clock a decade or two and ask a legalization advocate (his younger self-for example) what would happen if we dramatically expanded opioid prescriptions. He points out, probably correctly, that such an advocate probably would not have predicted the increase in mortality. He accuses Miron of issuing "a post hoc attempt to rationalize away the unexpected result of a disturbing national experiment." Experiment indeed, but once again the past 20 years was not even remotely an experimentation with drug legalization, for all the various reasons described above.

I don't quite understand VerBruggen's point with his hypothetical interrogation of a naive drug reformer. "He would have made a bad prediction, so we know his understanding of the world is totally wrong." Is that what he's implying? Let's play that game with VerBruggen. Given your narrative, wouldn't you predict that the number of prescription opioid users and addicts has increased dramatically in recent years? Does your narrative's failure to predict a completely flat trendline mean it's wrong? Not necessarily. But we should certainly bring all evidence to light and do a full appraisal of all competing theories, rather than place the spotlight on any one bad (hypothetical) prediction of one of the several competing narratives. Wouldn't his narrative imply that a crackdown would have reduced opioid-related deaths, when in fact they increased? Is his point about "new addicts versus populations of existing addicts" (stocks versus flows) the same kind of after-the-fact rationalizing that he criticizes?
Miron further writes that the U.S.’s experiences before 1914 (when opioids were legal) and during Prohibition support his point. Regarding the former, we don’t have good vital-statistics data from a century ago or more, and we also didn’t have as much disposable income to spend on drugs back then, so the comparison is limited. But it’s worth noting that the nation enacted drug restrictions at that time specifically because addiction was becoming a problem, dating back to morphine abuse following the Civil War and opium dens frequented by Chinese railroad workers around the same time.
VerBruggen accidentally catches  himself in a contradiction here. We "don’t have good vital-statistics data from a century ago", and yet "the nation enacted drug restrictions at that time specifically because addiction was becoming a problem." Did we really experience a problematic increase in addiction? How do we know if we didn't have good vital statistics? The impression of a massive society-wide problem is probably based on anecdotes and media accounts of drug-fueled rampages that still dominate, the same kind that sensationalist journalism still produces today. Surely there were real examples of extremely self-destructive, uncontrollable drug addicts. But without good vital statistics, there's no way of knowing how big the actual problem was or whether the government restrictions were justified.
The reality of Prohibition is rather messy, too. It certainly had plenty of bad effects, but recent research suggests it significantly reduced alcohol consumption and liver-cirrhosis deaths. It’s quite possible it saved lives on balance.
He links to a paper by Miron that estimates cirrhosis death rates possibly declined by 10-20% during prohibition. Presumably this is meant to embarass him, to hoist him by his own petard. A couple of points here. For one thing, Miron and his co-author Angela Dills (if I'm reading the paper correctly) are pushing back against claims of a much more extreme effect of alcohol prohibition on alcohol consumption and related problems. Notice how, nationwide and by-state, cirrhosis levels were trending down and (if I'm reading the graphs correctly) close to their minima already by 1920. Also, VerBruggen should read a lot more of Miron's papers. See (in Figure 1) how various estimates of alcohol consumption do not show such a steep decline, and most show a quick rebound after 1920. Once again, I'm fine with doubting national statistics from this era, but let's either doubt them or believe them consistently.

Want to talk about the quality of vital statistics? Let's get into it! In the book Drug War Heresies, the authors discuss the inconsistent (or just consistently awful) quality of drug overdose and abuse statistics across the developed world. It's often hard or impossible to do inter-country comparisons. There is a long discussion of drug policy shifting in Italy, from a lax policy to a crack down with a law enforcement approach and back to a lax policy with a treatment/harm-reduction approach. But all of this was done based on shifting political powers and ideologies; there was virtually no good data supporting these movements. For that matter, are vital statistics any good in the United States in the present day? I have serious reservations, which I discuss at length above. Once again, our survey data seem to indicate that you can triple the amount of opioids prescribed without increasing the number of addicts or the number of non-addicted casual users. If these data are accurate, they significantly refute VerBruggen's whole narrative. If they are not, then we need to adopt a skeptical stance toward the statistics and admit that the policy implications are unclear.

Even for solid, unimpeachable data, the policy implications for some measured trend are usually ambiguous. They depend on your assumptions about what is generating those data, your value system, how much weight you put on various costs, etc. VerBruggen is taking the sheer, brute fact of a rising death rate and trying to draw policy implications from this. You can't get an ought from an is.
What about the fact that efforts to control prescription-pill abuse since 2010 seem to have backfired, driving addicts to dangerous alternatives like heroin and Fentanyl? This is something I discussed in my original piece. It does illuminate the need to consider current addicts when introducing new controls, but it doesn’t suggest loosening controls to begin with is a good idea. By 2010, prescription-opioid overdoses had roughly quadrupled in a decade, the clear result of increased, not limited, supply.
This is a dodge, and a pretty lame one at that. I've said this before: those fentanyl deaths would not happen in a legal market, because there is no real demand for the stuff. A dealer in a legal market trying to sell fentanyl as heroin would be sued for fraud. Such legal options are taken off the table in a black market.

Drug Interactions: Something VerBruggen Misses Entirely

Something else I've written about at length, but which VerBruggen misses entirely, is the degree to which these "drug overdoses" are actually multi-drug interactions. I have detailed stats on this for the figures from 2014, 2015, and 2016. I actually think it's quite misleading to call these "drug overdoses" because such a small proportion of them are single-drug overdoses. Most involve multiple substances. From 2016 (the most recent year for which this detailed data are available), about a third of these involved benzodiazepines. About 13% involved alcohol. About 9% involved antidepressants. Only about 14% of prescription opioid overdoses involved only a single substance. This is a big deal, because once again it illuminates a specific risk factor for policy to target, rather than suggesting blanket prohibition. Once again, prohibition removes a lever of control over this specific problem, but legalization would restore this control. A drug user who has to stand before a pharmacist and ask for recreation-grade opioids could receive a lecture about mixing. The pharmacist could spot someone purchasing multiple potentially interacting substances and warn them against mixing, or (with a little bit of corporate or government intervention) perhaps they could cross-reference their current purchase with recent purchases. Or perhaps they could issue "do not mix" warnings as a matter of policy and print warning labels and recommended dosages on the packaging. VerBruggen mentions "harm reduction" in a favorable light, but supply-side prohibition takes most of the obvious solutions off the table.

Opioid Alarmists Have Under-Theorized the Problem

VerBruggen is side-stepping philosophy. He is trying to derive policy implications from brute facts: Opioid prescriptions tripled, and then opioid-related mortality tripled (or perhaps quadrupled or more). But I don't think it's possible to avoid these sticky questions of moral philosophy. Even ducking that, the exact policy prescription depends on the exact mechanism that is causing the death statistics to rise.

VerBruggen tells us that the death rate due to opioids rose from six per 100,000 in 1999 to 16 per 100,000 in 2015. Okay. Is 16 per 100,000 a lot? Is it too much? Was six per 100,000 acceptable? Did going from 0.00006 to 0.00016 cross some threshold at which drug prohibition suddenly becomes a good idea? I don't know, and VerBruggen doesn't really tell us. Alarmists like to present the numbers in a way that makes them look big and scary, but every time I've seen these numbers presented as a risk, as some kind of rate per user, they don't look scary at all. Sixteen per 100k of total population? This number would be only very slightly bigger if you divided by the number of total prescriptions in the U.S.; last I checked there were something like 200 million prescriptions annually (compared to ~330 million total people). When I try to place a dollar estimate on this risk, it seems comparable to what a rational person might pay to avoid a week or so of nagging pain. Even supposing that you divide by a much smaller number, like the number of recreational opioid users or the number of opioid addicts, it doesn't look so bad. It's nothing like the ~4% per year mortality rate for heroin users (from that cause of death alone!).

Even the act of dividing by total users and expressing this ratio as the risk per user raises sticky philosophical questions, and I address these in my response to Caulkins. If, say, 15 out of every 100,000 people has a gene sequence that will cause them to drop dead if they ingest opioids (and supposing it's impossible to test for this, thus making it a true risk rather than a known quantity), then such a risk calculation represents something meaningful. Then again, if you can choose whether to indulge a foolish impulse to snort or inject your Oxycontin (and, more to the point, do so repeatedly until you have an unmanageable addiction), then dividing by total users is a pointless mathematical exercise. Risk is about unknown possibilities happening by random chance. People can choose by their own free will whether they are exposed to a fatal drug interaction or overdose. (At the very least they can initially choose, and even without this caveat there is some element of choice in the decision to continue using or to abstain.)

By what standard should we judge the rise in death rates? From a pure libertarian standard, anything that anyone does with their freedom, so long as they are only harming themselves, is acceptable. From an economic rationality standard, rational actors weigh the costs and benefits of their actions using the best information available to them (they rationally acquire new information until the cost of acquiring more information becomes prohibitive). Thus any action taken is justified according to the actor's own cost-benefit calculus (ignoring externalities, or assuming these are somehow internalized).These both give similar results. Under the libertarian standard, people are morally at liberty to do whatever self-harm they choose to indulge. Under the economic rationality standard, any action taken is recommended by the person's own cost-benefit calculus, and we should be reluctant to second-guess this because third parties cannot observe the person's preferences. You don't have to accept either standard to conclude that prohibition is a good idea, but I should point out that these are the standards that most people apply to most realms of life: choosing a romantic partner, choosing friends, choosing a church, choosing what to read, choosing what to eat, etc.

Perhaps a slightly paternalistic standard is appropriate, or even a harshly paternalistic standard: people aren't fully rational, because they underestimate certain costs. We are justified in using force to stop them. The economics on this are pretty clear. Absent making heroic and stilted assumptions, it is essentially impossible to deter self-harm in a way that produces a net benefit to society. The harder you hammer self-harmers, the greater the total harm to society. The penalty you impose (either directly on users or indirectly by targeting suppliers) grows much faster than the benefit gained by successfully deterring users. (Don't take my word for it. Take Gary Becker's word for it. Or better yet, read his actual argument.)

What about a "fairness" standard of judgment? Suppose restricting access to opioids means that people who really need them can't get them. I think this is terribly unfair. You'd have to be able to deter a lot of abusers for every legitimate pain patient to make this worthwhile, and by all accounts it's the other way around. (See above, regarding the very small proportion of legitimate patients who go on to become abusers.)

What about a "personal responsibility" standard? Is it fair to punish manufacturers and dealers because some of their customers do stupid things and kill themselves? (Again, it's become too easy to accidentally overdose because of the fentanyl phenomenon, which is a product of prohibition. That aside, it takes some phenomenally stupid patterns of drug use to actually kill yourself.) I don't think so.

How about a "public health" standard, where any increase in illness or mortality is bad and should be decreased at any price, no matter how high. I said above that this is a deranged and immoral standard, but as far as I can tell it's the only one that's consistent with VerBruggen's ultimate policy recommendation to continue supply-side prohibition. (And even then, only given his assumptions about the effectiveness of supply-side prohibition, which are pretty implausible.)

What about a utilitarian standard? What counts in our calculus of costs and benefits? Surely pleasure counts. The notion that people use these substances and continue to use them because they enjoy them is totally lost on drug warriors. How do we begin to count the costs imposed on people involved in the supply chain? Inconvenienced motorists who are stopped and harassed by police. Terrifying SWAT raids on residences, which more often than not turn out to be innocent. Multi-decade prison sentences, often for low-level dealers. Lopsided gender ratios in heavily-policed neighborhoods, leading to a breakdown of family structure. Not to mention the budgetary cost of employing law enforcement and imprisoning drug dealers. The human cost is truly enormous, even if we set aside our moral revulsion and adopt a strict cost-benefit calculus, even if we convert "man-years languishing in prison" to a bland dollar figure. VerBruggen wants it to continue, so apparently he thinks it's worth it. I wish he'd tell us how he got to this answer, because I'm stumped.
And Andrew I. Cohen made a more philosophical case against continuing the War on Drugs, even if it’s focused on dealers rather than users, emphasizing the seriousness of deploying state power and writing that I have “not quite shown us why such a stark measure is justified here.” On that we can agree to disagree; the incredible rise of opioid-involved fatalities that ensued when we loosened controls on opioids shows, to me, that the state does need to be involved here. And while we can have a discussion about how to treat low-level dealers, I believe criminal sanctions and not just civil fines are certainly needed to deter high-quantity traffickers.
Apparently he's endorsing all the horrible things we currently do to drug dealers? I'm left scratching my head here. Nothing in his main essay or his response comes anywhere close to justifying current policy.

I was hoping that this forum would lead me to the crux of my disagreement with these "thoughtful prohibitionists." Unfortunately, I'm still confused. Prohibition of any kind does not appear to survive any kind of cost-benefit analysis, based on any serious attempt to quantify and weigh costs and benefits. VerBruggen seems like an incredibly thoughtful journalist. I hope he will reconsider his support for supply-side prohibition. Despite his arriving at good policy prescriptions in a lot of areas specifically regarding drug policy, I think he is deeply misguided on this point.

Wednesday, February 21, 2018

So You Want To Be Kept As a Pet?

[This will not be one of my more thoughtful posts.]

When people ask to be protected from competition from foreigners, I imagine they are really saying to the rest of society, "Keep me as a pet." The request for protection comes in two forms: import restrictions (protecting them from people toiling in their home country and shipping us the goods), and immigration restrictions (protecting them from people literally crossing the border and "taking" their job). It's like saying, "I can't compete, and I'm unwilling to take the pay cut necessary to keep doin' what I'm doin'. Please protect me from people who can do the same thing I can do only better and cheaper." It's basically asking the rest of society to subsidize the lifestyle you've grown comfortable with so you don't have to adjust to a changing world.

It's like asking the rest of society to "adopt" your factory or office building, pump money into it (even though it's become economically irrelevant or wasteful), and turn it into one of those historic villages where actors wander around trying not to break character while interacting with the tourists. Of course this isn't literally what happens. The propped-up office or factory surely produces some economically meaningful output, which props up the illusion that it's a viable company. But the unfettered economics suggest that the firm should close, and the workers and capital employed by that firm should go into other productive ventures. Propping up these dying businesses halts progress. The churn is sometimes painful, but people do adjust when the inevitable change finally comes. Like Deirdre McCloskey says, economic change is win-win-win-win-win-win-win-lose. The wins outweigh the losses, but eventually you do experience that loss and you adjust, perhaps entering an industry that didn't exist five years ago. To halt the losses is to throw out all those wins, all because a sympathetic-looking interest group asked to be coddled. In the reductio ad absurdum, we're all still toiling farmers plus maybe the rare skilled tradesman. Thank goodness we didn't get stuck there.

Law and Liberty Forum on Opioids: My Reaction to Caulkins

A few months ago, Jeffrey Miron (who has been something of an e-mail pen pal) asked for my commentary on this essay he wrote for Law and Liberty. It's a response to another essay on the same site by Robert VerBruggen. (In my opinion, the VerBruggen piece is incredibly wrong-headed and his narrative is wrong in some pretty basic ways. I'll respond directly to his piece at another time, but I think my post from last September  still holds up well.)

There are three other responses, and I'll try to get to each of them in time. For this post, I'll focus on this one by Johnathan Caulkins.

Caulkins pushes back against the argument that most people who try drugs, even hard drugs, do so responsibly and aren't harmed by them. He recasts the problem from being "proportion of problem users" to "proportion of total use that is problematic." As in, most who people try cocaine don't get hooked. But if you look at the proportion of incidents of cocaine use, or the proportion of cocaine going to problem users, it's very high. Probably a majority by his estimates.

In 1994, Jim Anthony and colleagues published what is still one of the most widely-cited estimates of what proportions of people who ever try various drugs go on to become dependent.[2] Based on data collected between 1990 and 1992 by the National Comorbidity Survey, their estimates for the three major “hard” drugs varied from 11.2 percent for stimulants (which includes methamphetamine but also weaker amphetamine-type-stimulants) to 23.1 percent for heroin. I’ll focus on the proportion for cocaine (16.7 percent) since cocaine was then the most widely used hard drug.
 The 16.7 percent figure does not mean that at any given time five people are enjoying cocaine for every one that is harmed by its use. People who become dependent often suffer through 10 or 20 years of dependence, whereas most of those who do not become dependent use for much less than a decade, and often only quite briefly. So the proportion of days-of-use that pertains to people struggling with dependence is much greater than 16.7 percent.
He dissects some survey data about how many times cocaine users have used during their lifetime. The result of his back-of-the-envelope calculation appears troubling at first glance:

[T]he odds for the average person who tries cocaine are an expectation of three days of misery per day of harmless fun.
Sounds like a pretty bad deal, huh? The implied lesson is that cocaine is more dangerous than it appears according to "addiction per user" ratios.
Thus a naïve interpretation of Anthony et al.’s “capture ratio” is that trying cocaine is like playing Russian roulette, with just one chance in six of disaster. But after recognizing that happy use is transitory and harmful use is long-lasting, the odds are effectively reversed. It is more akin to playing roulette with bullets in five of the pistol’s chambers, not one.
Pretty damning, right? I had just recently written about this topic based on the SAMHSA drug survey data. I had the first-blush common-sense reaction that most drug users don't get hooked and don't persist in their drug use. Caulkins is inviting us to flip the numbers by using a "problematic use per incident of drug use" ratio rather than a "problematic use per user" ratio. I think his analysis is wrong for some basic reasons, and his re-casting to per incident is a mistake.

Admittedly this is getting philosophical; I'm not accusing Caulkins of making a factual or mathematical error. But, as Daniel Dennett says, "There is no such thing as philosophy free science. There is only science whose philosophical baggage has been taken on board without examination." So let me briefly play the role of the probing, groping TSA agent and see if Caulkins has inadvertently snuck something past us. Let's examine away.

Free Will

First of all, doing something that's potentially habit-forming is not like a game of Russian roulette. There isn't a flipping coin, tumbling die, roulette wheel, or spinning barrel of a six-shooter inside our heads. Human beings are sentient. They consciously decide whether to take risks or avoid them. They consciously (or unconsciously) weigh costs and benefits. The person who gets hooked on cocaine makes a series of decisions. An initial decision: "Hmm. I've heard this thing has a bad reputation. I'll try it anyway." As Caulkins himself concedes, most people make it through this step unscathed. There is a subsequent decision to use again: "That felt really good, I think I'll repeat." Or (again, far more typical): "No thanks." Somewhat paradoxically, a really good first experience can lead to a total swearing off. Drug users often quite rationally recognize that a continued dalliance with the pleasant substance might result in a habit that's hard to control. A sort of "That was good. Too good." reaction. Someone has to really indulge repeatedly and quite deliberately to turn it into a bad habit.

See my post on Unbroken Brain for more of these details on drug addiction. Drug users are mostly rational; they don't get ensnared in the "chemical hooks" of the substances they imbibe. I'll admit that it would be pig-headed of me to try to ignore (in my argument) the temptations imposed by chemical dependence and the fact that some people find these temptations irresistible. I think it's equally pig-headed to ignore the fact that most people with a chemical dependence do in fact overcome their addictions and get their lives in order. They choose to do so. They decide to make the change, despite the temptation. I'm not denying the existence of drug addiction or ducking the point here. I'm just trying to put some proper context around the phenomenon of drug addiction.

Caulkins Dismisses Too Many Causal Users

Caulkins presents a useful table showing a breakdown of how many times "lifetime users" of cocaine have actually used. 29% only once or twice; 16% three to five times; 15% six to ten times. (The survey asked on how many days they used, not how many times they used; an evening-long coke-binge in which you bumped 20 times counts as one day of use.) So fully 60% of "lifetime users" have only used it ten or fewer times. As long as Caulkins would grant that the "not even once" propaganda is overblown nonsense, I'll grant that we might want to ignore people who have only touched the stuff a few times. This population was likely never "at risk" because they never used persistently enough. (Then again, see my caveat above about drug users rationally swearing off something that's "too good" after only one or a few uses. I have heard second-hand stories about people doing this, so it can't be too uncommon. Such persons might even describe themselves as having once been "dependent.")

But Caulkins takes this way too far. He points out that anti-tobacco activists ignore people who have smoked on fewer than 100 occasions. That makes perfectly good sense to me with respect to tobacco. But consider someone who used cocaine 20 times. That could be one coke-fueled outing every weekend for the better part of half a year, or every other weekend for the better part of a year. Someone in that category could be said to have dabbled significantly. And someone who does imbibe with that kind of frequency might develop a mild "dependence" or at least a craving for the habit. That's not really frequent enough to do serious cumulative damage (long-term cocaine use damages the heart, among other things). But they might be represented in the "16.7%" figure that Caulkins cites. I think Caulkins is loading his figures by trying to dismiss the all but 14% who have used on 100 or more days in their life. The 40% who have used 10 or more times are fully in play, in my opinion.

From the paper that the 16.7% figure comes from:
[D]ependence was assessed whenever participants reported at least several occasions of extramedical drug use, under the assumption that even as few as six occasions might be sufficient for development of drug dependence, but that drug dependence would be extremely rare or improbable among persons who had used the drug no more than several times.
There Are Gradations of "Dependence"

Caulkins invites us to imagine worst-case scenarios:
People who become dependent often suffer through 10 or 20 years of dependence, whereas most of those who do not become dependent use for much less than a decade, and often only quite briefly.
"Often?" How often? Half the time? Once in every ten? Dependence is just like any other social problem. There is a distribution of severity, with the most severe instances being the least common. The better part of the 16.7% are probably people who remember using a little too much, or perhaps remember a few genuine problems caused by their drug use which quickly prompted them to stop. Most people who become full-fledged addicts age out of it by their late 20s or early 30s. A decade is a typical tenure for someone who's already become an addict, according to various other sources I've read (Unbroken Brain, High Price, sorry I don't have specific academic references handy for this "stylized fact").

His comment about "three days of misery per day of harmless fun" is more than a little bit hyperbolic. No doubt some addicts are completely miserable. But I'll bet that many of the people who strictly meet the criteria for "dependence" still at least somewhat enjoy their habit, even if they recognize it's bad and wish they would stop.

Selection Bias

It's worth keeping in mind (as Caulkins quite appropriately reminds us about halfway down the page) that this data comes from within a regime of drug prohibition. The sample of individuals who imbibe in a prohibition regime is very different from the sample who would imbibe under full legalization. These are people who are disproportionately likely to be risk-takers. By definition, they are people who choose to break the law. We are constantly inundated with information about how dangerous and addictive these substances are. Pause and think about what kind of person ignores these warnings and imbibes anyway.  People who have impulse control problems are going to be over-represented in this sample of the population. People who don't generally have their lives together (unmarried, marginally employed, no dependents or perhaps neglectful of their existing dependents) will be over-represented here. If you have a normal job and family life, certain patterns of drug use are ruled out of the question. If you look at a population where these things are missing, you're going to see a disproportionate number of addicts and persistent drug users. Of course most people who have been users (even of hard drugs) are not dysfunctional, but any population of illegal users is going to have disproportionate numbers of dysfunctional adults. You can't simply apply numbers from this population to the general population and speculate that it's a reasonable estimate for what would happen under full legalization.

The Substances Themselves Differ Under Prohibition Versus Legalization

Bolivian Indians chew coca leaf all day long. They do not inexorably escalate to powdered cocaine or crack. Presumably this is closer to the model of "legal cocaine use." Or look at another class of stimulants. Compare attention deficit disorder medications to methamphetamine. They are substantially the same substances (in fact government surveys and death statistics count them in the same category!), but school  children with ADD prescriptions spend significant portions of their day (every day) under their influence. They don't inexorably escalate to smoking or injecting methamphetamine.

Under legalization, there would likely be some coca tea drinkers and perhaps leaf chewers (lozenges? nasal sprays? tinctures?). But few would escalate to pure powdered cocaine. We likely would not have many more intense users than we currently have. More likely, we'd fill in the lower-dose-but-more-frequent-use left tail of the distribution, which full-fledged prohibition chops off. It's doubtful that the right-tail of intense frequent use would expand much if at all. You might get the occasional tea drinker who occasionally mixes his brew strong enough to get a mild buzz, much like the caffeine buzz you'd get from a tall cup of Starbucks.

The distribution of "days used in lifetime" would probably expand rightward, putting more people in the categories of more frequent use. At that point, we could talk about dropping people who used on 100 or fewer days. But I think that kind of data-censoring is inappropriate given the regime the data comes from.

Adjusting for Implausible Results

Look at the paper that the 16.7% figure comes from. See Table 2 on the 8th page of the document. So supposedly 4.9% of past psychedelic users developed a dependence? 9.1% of marijuana users? Some kind of "bullshit implausibility adjustment factor" needs to be applied here. Psychedelics and marijuana don't cause physical dependence or withdrawal. Any perceived dependence is psychological, and no more concerning than an "addiction" to video games. Maybe these drugs were the vehicle by which some people chose to harm their lives, but it would be unfair to blame the drugs for the problems of people with poor impulse control or other unrelated problems. I made this point on my "Persistence of Drug Use" post (linked to above).

I think what's going on here is that people are recalling their drug use as a "youthful indiscretion". Perhaps many of them are embarrassed about their former habit and recall it as being more harmful than it really was. Supposedly there were quality control checks in place to get accurate measures of "dependence" according to the DSM III definition (read the paper for details). But the psychedelic and marijuana numbers indicate, to me at least, that some kind of misreporting is creeping in. People who are asked about their drug use, years later when they are older and wiser, likely misreport how bad it was.

Picking the Relevant Base for "Exposure"

I'm an actuary, so I'm keenly aware of the problem of "picking the relevant exposure base." If I have a population of 1000 cars, all else equal it will have twice the accidents as a population of 500 cars. If I have a sample of 1000 cars for 2 years, all else equal there will be twice as many accidents as 1000 cars for 1 year. In fact "car-years" is a standard unit of exposure. Then again, I could pick "households" as my basis for exposure. Some households have an old beater than never gets driven plus two or three cars for regular use. The old beater isn't as exposed to risk as the others. Not all car-years are created equally, but then again neither are all households created equally. Perhaps I could use "miles driven." A car that drives twice as many miles, all else equal, will have twice as many accidents. But highway miles are safer than city miles. So maybe "equivalent highway miles driven", something that re-casts all miles driven to an equivalent number of highway miles. Or maybe I just use "vehicle-years" and adjust each individual exposure for risk factors: the guy who drives 6000 miles and the guy who drives 1000 miles each gets one "car-year" of exposure, but the first guy gets a factor of six adjustment when I calculate his accident risk.

There are different ways of doing this, some equivalent to others. But I think using "incidents of use" or "days of active use" as the exposure for "risk of addiction" stacks the deck in a way that a "per lifetime user" basis does not. Likewise, most casino-goers don't have to worry about developing a gambling addiction. But if you recast your base as "per dollar gambled", you'll find a much larger proportion (maybe a majority?) are being gambled by people with gambling problems. If you're trying to assess a priori risk, you want an exposure base that causes the risk to rise linearly as the exposure rises. It would be silly to use, say, "dollars of insurance claims paid" as my exposure base, because this restricts us to automobiles that have already been involved in accidents. Likewise, the problem with addiction is that if you do it a little too much, you will become "captured" and end up doing it a lot too much. The if the exposure base for the social problem you are trying to measure (be it auto accidents or drug addiction) skyrockets when a problem occurs, it's a bad exposure base.

I'll applaud Caulkins for raising an interesting point about what basis to use, but I don't think it's at all clear that the "per days of use" basis is the relevant one. It depends on the question you're trying to answer. "I'm offered cocaine for the first time. Should I try it?" I think the "per lifetime user" basis is the relevant one for answering this question. "I've tried before, and I have the opportunity to acquire some tonight. Should I?" Maybe the "per use" basis starts to look more relevant for this kind of question, especially for the tenth or 20th offering. I think the "per days of use" basis comes dangerously close to being a tautology. Caulkins cuts off the left tail of the distribution (too much so, I argued above) by claiming that those infrequent users aren't really exposed to addiction. Then, having censored the data to only include the right tail of the distribution, he argues that most of this cocaine use is done by people with addiction issues. Of course he brings in data on what fraction of lifetime users experienced dependence (the 16.7% figure cited above), so it's not literally a tautology. But if frequent, persistent use is part of the definition of dependence, we're still trapped inside a tautology. As in, "Let's define persistent, repeated use as problematic. Oh my goodness! Low and behold, most drug use is problematic!" Breaking the tautology depends on the independence of the "drug dependence" question and the "persistent use" question. If these are strongly linked by definition, as I suspect they are, we're stuck in tautology world.

By the way, this is hard. I struggled with the issue of "what exposure base to use" in a previous post. Suppose I want to know the mortality of cocaine users. Are the "past year users" all at risk? Or just the "past month users" who presumably have a more serious and persistent habit? Let me just reiterate that I am not at all faulting Caulkins for raising this issue.

It's Hard to Deter Self-Harm

Suppose I'm wrong about everything and cocaine use really is the three-to-one game of Russian roulette that Caulkins thinks it is. Does that support the notion of drug prohibition? Of course not.

The problem with drug prohibition is that it eats its own tail. It requires the implausible dueling assumptions that drug users are irrational enough to ignore the risks of drug use, but rational enough to be deterred by legal penalties and the paltry price increase imposed by prohibition. ("Paltry" because the full price includes all those nasty risks in addition to the actual dollar price tag.) If you actually try to model this out by stating your assumptions clearly, you find that it doesn't work. Someone who is willing to play a 3-to-1 game of Russian roulette is someone who is not going to be deterred by a legal slap on the wrist, an increase in the market price (even a severe one), or the search costs required to find a dealer. If the bulk of the "cost" of cocaine use is embodied in the inherent pharmacological cost (risk of addiction and self-harm from continued use), then drug prohibition is unlikely to meaningfully deter these users. He cites the example of marijuana legalization leading to a massive increase in daily usage, but this is a distraction. Marijuana is not harmful or addictive, so it's actually plausible that prohibition causes significant deterrence. The legal penalties, higher market price, etc., are a significant component of the total cost in the case of marijuana. Not so in the case of cocaine, if we're to believe Caulkins' estimates of the risk of addiction. Make whatever assumptions you like about drug users, but keep those assumptions consistent. They're irrational? Cool, I can buy that. Then they won't be rationally deterred by anti-drug laws. They're rational after all? Cool, then their drug use must be a rational decision that you simply fail to understand. They rationally respond to legal sanctions while irrationally responding to the pharmacological risks? No, now you are confused. Specify what the demand curve looks like, but once you've done so stick with it and spell out the implications.

Caulkins wrote a very thoughtful essay, and it has given me quite a lot to think about. I just don't buy his bottom line (about cocaine, anyway). I'll try to respond to other essays in the Law and Liberty forum as I have time.

Addendum

I felt a need to respond to this part of his essay:

Second, I concede that prohibition harms many people, probably more than it helps. However, it harms most of them only modestly, whereas some whom it protects benefit enormously.
This strikes me as a pretty blithe dismissal of the suffering caused by drug prohibition. Are we shifting back to a per-person exposure base? I guess if the median person "harmed by drug prohibition" is the casual user who can't score any, or (to use an even broader base) the tax payer saddled with the bill for a useless and ineffective drug war, this statement is literally true. But just as there is a thick right tail to the distribution of cocaine-related harms, there is a thick right tail to the distribution of drug-war related harms. That is, there are infrequent but severe cases of harm that likely dominate the total harm, by any reasonable accounting. Let's fixate on the innocent people whose homes are unnecessarily raided, the people languishing in prison because they triggered a mandatory minimum over an arbitrary weight limit, the families destroyed by the incarceration of their loved ones, the communities destroyed because incarceration has imposed a lopsided male/female ratio, the people overdosing on heroin or cocaine tainted with fentanyl (yes, that is the fault of drug prohibition, as much as the drug warriors would love to take a pass on this one).

Pardon me for dwelling on this, but what a lopsided comparison. Read the second sentence in that excerpt again. Are we comparing the median person "harmed" (implied by his use of the phrase "most of them") to the very most extreme cases of "drug abuse averted" (implied by his use of the phrase "some of whom")? If we do a proper cost-benefit analysis, weighing all costs against all benefits, Caulkins would have a very hard time justifying cocaine prohibition.

Saturday, February 17, 2018

Thomas Sowell’s Farewell Letter to His Secretary


In Man of Letters, Thomas Sowell publishes many of his personal letters.  It is a very engaging read, and it gives you a real flavor for  his thinking and his influences.

One letter is to his secretary Beverly, who recently quit (retired?). It's clear from Sowell's very heartfelt letter that he is sad to see her go:
I am of course very sorry to lose a very good secretary. But I have also gotten to know you somewhat over the past year or so, and if I may consider myself a friend, then as a friend I think you may have made the best decision. Just this past weekend I expressed my concern to my wife that you seemed to be making the job far harder on yourself than it needed to be, partly by trying to shape my decisions instead of simply getting me the information that I needed to make my own decisions. She suggested that I take you to lunch and air our different conceptions of the work. But, by the time I reached the office on Monday, you had made your decision.
Emphasis mine. I think this is a common employer-employee dynamic. The employee is trying too hard to shape the decision-making (beyond their actual mandate to do so), while the employer just wants the necessary information to make a decision. Sometimes it’s even cynical. The employee tries to influence the employer toward the decision that will require the least effort and headache (for the employee). The employer senses this and has to push back through the employee’s manipulation and stonewalling. Sometimes it’s sheer ego, as in the employee thinks they know better and wants to be the boss. And of course sometimes the employee does know better, and the boss’s boneheaded decision really does blow up in everyone’s face even though s/he tried to warn him.

This is a slightly different variation of something I wrote about in a recent post. It’s not specific to work relationships, either. I think it could be at play in any power dynamic (parent-child) or even between equals (partners in a firm or project). I feel like I’ve been on both sides of this conflict.

Some Quick Advice


Download a couple of good podcasts that you want to listen to. You’re reading my blog at this moment, right? Surely there is some podcast that’s just as good. Or maybe an audiobook or some talks or lectures on Youtube. Or maybe even some music. Got it? Awesome. Your opportunity cost for doing household chores is now very close to zero. You can wash dishes or clean the cat litter or declutter the surfaces in your home or dust or do laundry or clean bathrooms. Your spouse will appreciate it, and you’ll feel productive. You might even feel good about doing it. I usually end up enjoying the feeling when I’m immersed in a productive task, even something simple like house work. The "switching cost" or "activation energy" (getting started in other words) is sometimes rough, but once something gets started it's not that bad. It doesn't feel like work. Go forth! Sometimes I even take my own advice on this one. 

The Power of Mutual Knowledge

There’s a puzzle I first encountered on Steven Landsburg’s blog “The Big Questions.” It involves an island of 100 blue-eyed and 100 brown-eyed natives being visited by a foreigner. There is a strictly observed religious tradition to never talk about anyone else’s eye-color, and to commit ritual suicide within a day if you ever discover your own eye color. (There are no reflective surfaces on the island.) But of course everyone can see everyone else’s eye color. Everyone with blue eyes knows there are at least 99 blue-eyed people and 100 brown-eyed people, just as everyone with brown eyes knows there are at least 99 brown-eyed people and 100 blue-eyed people. They just don’t know their own eye color. A foreigner (who happens to have blue eyes) arrives by boat, spends several months visiting and learning their ways, then sails away. Just as he leaves, he says, “Well, how interesting that there would be blue-eyed people in this part of the world!” And he sails off.

At first glance, he didn’t tell them anything. “Of course, everybody already knew that there are blue eyed people on the island! The foreigner’s statement adds no information.” But if you work through the puzzle, you discover the surprising result that everyone commits ritual suicide on the 100th day. It's a subtle story about mutual knowledge slowly creeping in and eventually having horrendous consequences. (Note that Landsburg is making a very different point than I am.)

I have two dueling thoughts on this. My first thought is, “This is way too complicated for anyone to actually figure out. Nobody is smart enough to actual work this out and figure out their own eye-color. In the real world, everyone would be safe.”

My second thought is, “Social life is unimaginably more complex than a simple rule about eye-color and ritual suicide. Of course people are constantly working out complex implications of mutual knowledge. Of course blurting shit out makes people uncomfortable. It may only 'reveal' information that everybody knows. But it reveals that everybody knows that everybody knows that everybody knows, ad infinitum.”

Imagine saying something unflattering about a coworker. “Everyone in this room knows you’re not qualified.” Everyone, including the accused, might already know, and everyone might suspect that everyone else already thinks it. But plausible deniability has been taken away. Now every time this coworker looks someone in the eye, he’ll see shame staring back at him. The boss, who was willing to tolerate the under-performer out of pity, doesn't have plausible deniability when someone asks, "How can you keep him on your team?" The coworker who was picking up the under-performer's slack feels emasculated if he continues. Everyone could live with the uncomfortable truth before it became mutual knowledge. It doesn't have to be such an obvious accusation, either. More in line with the puzzle, it could be a snippy comment about someone not carrying his weight. It's obvious enough who the target was, so mutual knowledge seeps in. 

You could think of other examples. You're in a group of friends, two of whom have an obvious mutual crush, and perhaps another friend in the group is jealous. Maybe everyone knows this dynamic exists, and maybe everyone suspects that everyone else knows. But blurting it out would be really uncomfortable. Even someone who indirectly hinted at it (perhaps with a light joke or teasing) might be scolded or shamed for creating an awkward moment. If you don't viscerally feel the discomfort of this scenario, think about how the group might split into factions. The jealous rival might feel compelled by shame to avoid the flirting couple. Other friends might feel compelled to choose between factions. Even when everybody knows and everybody suspects that everybody knows, everyone still has plausible deniability. 

In the same vein, merely stating that "some people" have cynical attitudes and do illicit activities may implicate you. In an alternative version of the puzzle given above, there is a society of 100 couples. Every husband cheats on his wife, and every wife knows about every infidelity of every other woman's husband (just not her own). In this version, she must murder her husband within 24 hours if she figures out he is a cheater. By the same logic as the blue-eyed and brown-eyed islander story, if some incautious outsider blurts out what everyone already knows, something awful happens. On day 100, all the cheating husbands die. 

This isn't about trivial matters of social faux pas and embarrassment. Dictators don't like crowds, because crowds tend to turn into angry protests. And these reveal to the world that, yes, everyone else is dissatisfied with the status quo. It's hard to maintain the fiction of a "100% approval rating" or a "bountiful harvest" in the light of this kind of public demonstration. Nicolae Ceausescu was brought down when people started chanting at a public speech and he lost control of the audience. The Arab Spring seems like another example of this. Why wouldn't Hosni Mubarak just sit in office and hold power? Why not just ignore the protesters and wait it out, like American presidents do all the time? I think this "mutual knowledge" dynamic is at play and cracks the armor of a dictatorship much more than it does in a democracy.

I am stealing some of these ideas about mutual knowledge from a Steven Pinker book, though at this point I couldn't even tell you which one. The Blank Slate? Or maybe it was How the Mind Works.

Are there other good examples of this dynamic at work?
________________________________


Think about an island with one blue-eyed and one brown-eyed person. On this island, the foreigner’s statement would cause the blue-eyed person to discover his eye-color. The blue-eyed person knows that the other person’s eyes are brown. Knowing he must be the blue-eyed person, he commits ritual suicide. The brown eyed person, seeing this, realizes that he must have brown eyes, or the blue-eyed person wouldn’t have discovered his eye-color and killed himself. “If I had blue eyes, he would have waited a day..”

Now think about an island with two blue-eyed and two brown-eyed people. The blue-eyed people know there’s at least one blue-eyed person. The brown-eyed people know there are at least two blue-eyed people. The foreigner’s statement might first cause each blue-eyed person to think, “Oh, he’s talking about that blue-eyed person. If that blue-eyed person sees three brown-eyed people he’ll commit ritual suicide within 24 hours. If not…” So when the blue-eyed person doesn’t commit ritual suicide within 24 hours, the other blue eyed person says, “Uh, oh. He was talking about both of us!” This is symmetric. They both commit ritual suicide. The brown-eyed people have worked this out, too, and so they know there were 2 blue-eyed people, not three. This allows them to work out that they must both have brown eyes.

Now think about an island with 3 blue-eyed and 3 brown-eyed people…work this one out yourself. By induction, this process keeps going. “On the Xth day, all X blue eyed and all X brown-eyed people commit ritual suicide.” And all because one loud-mouth visitor blurted something out.  

Or think about it this way. Obviously if there's only one blue eyed person on the island, the foreigner's statement that there's a blue-eyed person reveals that person's eye color to him. Ritual suicide.
Given this, if there are two blue-eyed people, the foreigner's statement will reveal that there's at least one blue-eyed person. On day 2, each blue eyed person works out that the other sees a blue-eyed instead of a brown-eyed person and commits ritual suicide.
Given this, if there are three blue-eyed people, after day 2 each blue-eyed person works out that there must be three blue-eyed people.
And so on. There's no magic number where this induction stops working.

Thursday, February 15, 2018

The Wikipedia Test

The "illusion of explanatory depth" confuses us into thinking we understand things at a deeper level than we really do. Simple stuff like "How does a tiolet work" or "How does a bicycle work" tends to stump us when we're asked specific questions about the mechanisms. Same goes for political topics and things on the news.

A decent test of your understanding is to look up the Wikipedia (or good ole' encyclopedia) entry for a topic that you have strong opinions about, and see if there's anything that's mind-bogglingly surprising to you. If you're finding a lot of surprises, and they seem to check out (check references! The Wikipedia is fallible!), then you probably didn't understand the topic as well as you thought.

I remember reading the Dakota Access pipeline Wikipedia page and being floored by the extent to which the builders had received voluntary easements. Apparently, to a very large number of people whose properties were affected, this was a pretty unobjectionable project (given appropriate compensation). I wish the people waxing wroth on my Facebook feed would have gone through this exercise. It might not have made them "pro-pipeline", but it would have made them re-think whether this was the world's greatest injustice.

Or do "the Google test". Simply look up the first few Google hits. Maybe search for "best arguments for/against..." Again, if there are a lot of surprises here, consider that maybe you need to do some reading, because you didn't understand your topic so well after all. I recently found that it was very easy to get the canonical list of supposed "non-neutrality" transgressions by internet service providers. I also found that this list completely falls apart when you look at the examples in any detail.

Maybe I'm mistaken about these topics. But the exercise of doing some research (even cursory research) on the topic that excites you is bound to yield some interesting surprises. Pick something that's been in the news, something you think you understand well, and start digging.

Latent Knowledge and Maps of Knowledge

Bryan Caplan’s latest book, The Case Against Education, is very engrossing. Also depressing. The most depressing piece of the book is the section that describes just how much we forget. When tested even months after the final exam, people seem to have lost most of what they learned.

I’m skeptical. I’ll start by describing my experience with the actuarial exams. These are four-hour exams that people typically spend four or five months studying for. (The pass rate is something like 50% for a good sitting.) There is a broad syllabus covering, say, 15 to 25 papers or textbook chapters. I would be able to work my way through the entire syllabus maybe four or five times in my 4+ months of study.

On the first pass through, it literally feels like you learn nothing at all. You read over the paper and an associated study guide, look at some practice problems, and go “Huh?” Almost nothing sticks. It’s hard to conceive of any test that would pick up the meager knowledge-gain of this first pass-through. Maybe you could vaguely detect that students pick up a concept or two on this first pass. Thoroughly confused, you move on to the next paper, and so on until you’ve done a first pass for everything in the syllabus.

On the second pass, you think, “Oh, this looks mildly familiar. But I don’t remember what the hell any of this is about.” But something magical happens. You say, “I understand it now.” You do a little bit better on some of the practice problems. Then you move on to the next paper, for which you find yourself having a similar “Ah ha” experience.

So plainly it isn’t possible that I learned nothing at all on the first pass-through. I would have loved to skip straight to the “Ah ha” of the second pass. I would have paid a fortune to skip that painful, humbling, slogging first pass. But plainly I had to go through this step. Clearly I was learning something that allowed the second pass to be more profitable.

I think much of what we learn and then forget is like this. I couldn’t necessarily pass any of my college or grad school physics exams. But I can pick it up again if I ever need to. More to the point, I can pick it up quickly and easily without a first slogging pass through it. I wonder how important this “latent knowledge-building” truly is. I’ve learned subject matter that I had never studied in school, so it might not be all that important after all. Surely someone has studied this concept. I wonder if Caplan came across any research on it? It seems like some critics of his book have brought it up, but I’m not sure Caplan has referred to research exploring/debunking this latent knowledge theory of education.

In addition to this latent knowledge, maybe we retain "maps of knowledge" after we forget the bulk of the subject matter. I may not remember how to do all varieties of calculus problems, but I know whether some problem I come across calls for an integration or a Lagrange multiplier. I can look up the appropriate textbook chapter. (BTW, I may have that rare one job in ten thousand that ever calls for these things, and then only rarely, and even then I use a computer to do it for me after setting up the problem.) It probably helps to "know your way around a topic." Then again, to Caplan's recent Econtalk exchange with Russ Roberts, you could probably design a test for this. "Which technique is most appropriate for this problem..." "Which textbook would you reach for if presented with this problem..."

None of this impugns Caplan's overall thesis. I still think we're all over-schooled, we forget too much, and we waste time on silly or useless topics. Caplan's guess that education is 80% signalling is probably a good estimate. Those actuarial exams I mentioned? 90% useless. It's just another long vetting process. "Can you pass these exams? (Stamps forehead with "Grade A".) Awesome, you get to be an actuary! Can you pass these exams, too? (Stamps forehead with "Grade AA".) Cool, you get an even better job as an actuary!" Or maybe I'm wrong and the latent knowledge and knowledge maps are really important. I just get the strong sense that most of the official actuarial
 syllabus is rarely or never used.