Wednesday, September 20, 2017

Libertarianism Is Mild-Mannered

Libertarian ideas seem strange and extreme to people who aren’t familiar with them. I’m going to argue in this post that they are not. Libertarianism is a mild-mannered ideology. From a few unimpeachable presumptions that no reasonable person would quarrel with, you can reach some extremely libertarian conclusions.

I often get the sense that libertarian policy proposals sound extreme-for-the-sake-of-being-extreme, at least to non-libertarians. It’s as if they think we’re courting controversy for its own sake, like we’re intentionally trying to be shock-jocks. “Hey, guys, watch me bite this bullet with a wolfish grin on my face!” And then I launch into an argument in favor of drug legalization or open borders or legalized organ markets. Some critics think these positions stem from an extreme deontology, a “Right must be done, or may the world burn” kind of ideology.

Not so. I’m much more of a utilitarian. If I hadn’t been convinced that economic freedom is actually good for society in some objective sense, I probably wouldn’t have become a libertarian. If some government policy demonstrates tremendous benefits, even if it significantly abridges freedom or costs a lot to implement, I would feel squeamish about ending it.

I favor legalizing heroin and cocaine because prohibiting these drugs causes enormous harm. Most of the heroin overdose deaths are due to prohibition, because in an illegal market you have to guess at the purity and even the identity of the substance you’re taking. Black market violence, which accounts for a substantial fraction of the murder committed every year, would basically disappear if we got rid of drug prohibition. We would have a lot fewer overdose deaths and murders if we legalized drugs. I’m not saying, “Legalize heroin, because freedom for its own sake trumps all other values.” This mild-mannered libertarian is saying, “Legalize heroin because it will save a lot of lives. I’m in favor of saving lives.”

I favor open borders because turning away an immigrant at the border potentially damns that person to a lifetime of third-world poverty. I’m against that. “That person” possibly means an innocent child who would make a fine American if given the chance. I don’t think we’d be flooded with hundreds of millions of new people. I think we’d increase the flow from a million or so immigrants a year to maybe two or three million. Housing prices and the existing job market (even the expanding job market) will attenuate any surge of immigrants, with or without any government border enforcement. I’m fine with screening for criminal backgrounds, so long as it isn’t too onerous. Once again, if the choice is “admit a questionable immigrant or damn him to third-world poverty for the rest of his life” we at least need to weigh these outcomes against each other. So, I admit “open the borders” sounds like shock-jock bullet-biting, but it’s not. It’s really a plea to allow willing landlords rent to willing tenants, to allow willing employers to hire willing workers, and to allow someone living in third-world misery to boost her family’s income by thirty-fold. And the majority of attempts to quantify the actual benefits of open immigration find that those benefits are truly enormous. Even the very most pessimistic estimates by the most anti-immigration economists (George Borjas) find that extremely liberalized immigration would only hurt high-school dropouts to the tune of ~10% of their income, while essentially everyone else benefits. Even if you think that's an unacceptable outcome, there is a potential deal to make everyone benefit with a tiny amount of redistribution. The ill-mannered shock-jock is the one who says, “My sense of national identity is more important than your family’s welfare, even after adding to that the welfare of my fellow countrymen who would rent, hire, and sell to you.”

I favor legalizing the sale of human organs because I don’t want thousands of people to die every year from preventable causes. Nobody is talking about selling the organs of unwilling donors here. I’m talking about allowing people to voluntarily sell their organs to desperate patients willing to buy them. I have an extra kidney, and you need an extra kidney. We can both benefit. I anticipate and appreciate all the “Yes, but”s. “Yes, but won’t some reckless individuals sell their kidneys for drugs and beer money?” “Yes, but if someone under economic duress is selling their kidney, isn’t that person being exploited?” “Yes, but the thought just gives me the heebie-jeebies.” (Frankly, I think this last “yes, but” is usually the real explanation, and the others are lame attempts to back-fit a rationalization to a visceral reaction.) All fine points well worth considering. But does it add up to thousands of lives saved each year? You can’t just blurt out a reason for not doing something. You have to somehow quantify it and weigh it against the alternative. I think the "moral" reservations are overblown. So what if someone uses their money on something you consider frivolous? Doesn’t someone else still get a kidney? So what if someone feels economically compelled to sell their kidney? Someone on the other side of that transaction gets to live (or at least gets to avoid the significant discomfort of dialysis). Mild-mannered libertarianism calmly deliberates on how to save the most lives and alleviate the greatest amount of human misery, even when the policy implications are uncomfortable and hard to swallow. It’s the shock-jock who blurts out the first answer that comes to mind and runs with every rationalization that sounds remotely plausible. Markets are the solution to waiting lists, so long as there is a usable supply of the thing being rationed. Raise the price and the shortage ends, whether you’re talking about wheat or kidneys.

There are dozens of other government programs with horrible consequences. Taxes on capital income. Underfunded pensions, which lead to fiscal catastrophes that completely paralyze the government. Underfunded long-term liabilities in general. Excessive regulation. Excessive control of medicine. Excessive control of pharmaceuticals. Misguided stimulus programs. "Anti-poverty" programs that expose poor working families to extremely high marginal tax rates. Mostly libertarians oppose these programs because they do tremendous harm, not because we’re willing to forgo an extremely beneficial program for the sake of freedom as an end unto itself.

I think most non-libertarians don't even realize that policy analysis is non-trivial. They just assume that their favorite government programs have all the benefits they imagine them to have, and anyone who opposes them does so because they don't want to pay (or their just plain evil for the sake of being evil, or they are too stupid to know better). It's easy to think we've identified all the appropriate "fix the world" levers, and we're just arguing about how hard to pull on them. Really, we're arguing about whether that lever actually helps at all or does positive harm. Mild-mannered libertarianism is trying to remind populism, conservatism, and progressivism that they've identified the wrong levers, or that they're trying to pull them in the wrong direction. It's hard sometimes to get over the sense of "do something"-ism, the notion that doing anything is better than doing nothing. But sometimes "nothing" is better than something. Society's problems are anti-inductive, after all. They thwart our best attempts to "do something."

"That guy just looks like a creep."

Do you ever look at someone like Donald Trump, Steve Bannon, or Donald Sterling and think, “Man, that guy just looks like a creep.”

Oh, you do? Well, congratulations then, because you’ve screwed up in the worst possible way. This kind of reaction misses the point by the widest mark imaginable. Remember the reason for the opprobrium leveled at these men. Recall the thing that we’re upset at them about. If you were to summarize, it would be something like: “We’re not supposed to judge people based on their outward appearance.” That is, you’re not supposed to make assumptions about someone based on how they look. Appearance doesn’t tell you what you need to know about a person’s moral worth, or value as an employee, or inner-most thoughts and secrets. It’s not much good to say, “Oh, it’s okay. Because I know from another source of information that these guys are really creeps. I’m just building on that starting point.” Imagine someone remarking on what a horrible person Michael Vick is, or Joaquin “El Chapo” Guzman, or Kim Jong Un, and then spewing a racial epithet at them. “Oh, no, it’s okay. I’m judging them based on their actual actions, and just building on that starting point.” I don’t think so. 

I get daily e-mails from the Washington Post that list various stories about “bad people” along with photos that make them look as ugly and unsympathetic as possible. They can’t just show a random pic of Trump. They have to show him with the worst possible grimace, or the most sinister-looking sneer. Using a person’s appearance to reinforce the fact that you don’t like them strikes me as being gratuitous and incredibly self-indulgent. It’s almost cheating, appealing to people’s visceral disgust reflex rather than appealing to a rational argument. I’m not just picking on the Washington Post, either. Every major media outlet does this, and I see this kind of garbage on my Facebook feed daily. The problem isn’t with the media, but with the consumers of media who eat this kind of thing up. (As I open my e-mail just now, I just got my daily list of headlines from the Post. I see two more photos of Trump. One looks roughly neutral, the other looks like it was deliberately chosen to make him look despicable. It would be interesting if they did a story that randomly chose a picture of Trump from an unbiased sample of images, or one that cycled through various pictures every few seconds.)

To be clear, this post is written partially to myself. I once caught myself thinking, “Man, Sterling just looks like a creep” and then, “Wait a minute…doesn’t thinking that completely freaking miss the point!?” Judging someone based on their appearance is a natural human reaction, and such judgments aren’t always inaccurate. But we should certainly strive to suppress these judgments. We should augment them with relevant information. We should also strive to discard “information” that viscerally confirms things we already want to believe. 

Tuesday, September 19, 2017

Coaching Someone On Their "Social Skills"

Imagine the following scenario, in which an introvert is trying to “coach” an extrovert on his life skills.
(Loud, bumpin’ party)
Introvert: Dude, you really need to head home, sit in a quiet room, and read a book.
Extrovert: What? But…I like hanging out here.
Introvert: (laughs) You think you do, but you need to take my advice on this. Just chilling out at home more often will make you so much happier.
Extrovert: I don’t know. I mean, I read stuff and sometimes it’s fun. Sometimes it’s even gripping and I can't put the book down. But just randomly grabbing a book and reading it? Whenever I’ve tried to make myself do that, it’s super boring.
Introvert: No, no, no, you’re just not trying hard enough. It’s like, you sit there staring at the book and scanning the words, but are you really reading it? Are you really giving it a chance?
Extrovert: Um…I feel like I have. Maybe you’re right, but isn’t it more likely that you just have very different preferences than me? Like, if I tried to do your stuff and you tried to do my stuff, we’d both be more miserable?
Introvert: (laughs again) Sure, sure. I just don’t feel like you’re hearing me. Whatever. I tried to help.
I have seen the reverse of this a few times. I’ve even been on the receiving end of it once or twice, even though I’m not exactly a shut-in. It's funny how people just assume you need the same kinds of social stimuli, and that you must be unhappy if you aren't getting it. No. Some people just prefer a little more quiet time, and some people prefer more company. Both are usually fine. Neither preference is necessarily pathological to indulge in.

(All that said, the quiet kid probably appreciates the invitation to hang out, if perhaps resenting the pressure to seek or enjoy it more than he does. The invitation without the "I'm coaching you to be happier" element would be ideal.)

I have heard some brash, abrasive loud-mouths criticize quiet people for their lack of social skills. I wanted to tell one such person that he lacks social awareness, that he just blurts shit out, that he isn't introspective enough to realize when he's mistaken, and that people find his bluntness to be off-putting. The quiet kid he's criticizing, on the other hand, is deeply introspective, is incisively analyzing every word and gesture in every conversation, is sometimes paralyzed by his social caution because he doesn't want to offend anyone or give off a "weird" vibe. The loud-mouth would probably do better to shut his mouth now and then, and the quiet kid would probably do better to speak up a little more. But the framing of the issue as "The quiet kid lacks social skills, while the outspoken person doesn't" really misses the mark.

I should clarify that I don't buy the binary model of introvert versus extrovert. It's really a continuum with a bell-shaped distribution. It's not bi-modal, with a peak on the "extrovert" side of the chart, another peak on the "introvert" side, and a bare spot in the middle. Some of the recent writing on introversion is almost a call to arms, suggesting that we self-identify as I-tribe and take back control from those brash, poor-mannered E-tribers. This kind of categorical self-identification can be unhealthy. Most people don't fall neatly into I-tribe or E-tribe, and most people have many "introvert" and many "extrovert" traits. Extroverts aren't generally brash and obnoxious, and introverts aren't generally shy and awkward. Rather, the combination of extroversion with a blustery personality can be obnoxious, just as the combination of introversion with social awkwardness can be infuriating to deal with. Extreme examples of both behaviors certainly exist, but it's important to remember that "extreme" isn't always pathological. 

Monday, September 18, 2017

Moral Outrage as a Deliberate Tactic

A while ago I wrote a post about moral outrage being a kind of commitment strategy. It's like staking out a position at the bargaining table and insisting that you won't budge.

This is surely a bi-partisan strategy, but I mostly see it as a tactic of the "social justice" leftists. I have no doubt that the people who bare their outrage are indeed displaying real anger. But most of the time when adults are dealing with other adults they have to attenuate their emotions. We all learn that we sometimes need to do this. So expressing outrage is a choice. People who do so are doing it deliberately.

I think it makes sense to point this out. It's perfectly fair to call people out when they are doing it. The result is usually that you become the target of the person's outrage. This reaction makes sense, even if the outrage is cultivated and deliberate. The person might sound like they are so unhinged that they can't think straight. But I imagine them saying, "Look, I have this super-weapon, and it's working for me. So if you attempt to disarm me, I'm going to use it on you." So when someone starts to denounce me for allegedly sympathizing with racists, or denounces me for allegedly being a sexist, I try to view it as an intentional, cynical debating tactic. When somebody starts cursing me out, not for taking a specific policy position but for insisting on an even-handed, thoughtful, outrage-free discussion (as happened to me on Facebook recently), it's simply hard to take such childish outbursts seriously. We need to look this in the eye and see it for what it really is. The response should be, "I see exactly what you are doing, and I'm simply not going to engage with such immature behavior."

(This old Slate Star Codex post is relevant. Scott responds to someone who argues that intentionally fighting dirty is justified. So at least in some cases the moral outrage we see is a carefully cultivated strategy.)

The "outrage as deliberate strategy" framing may be accurate even if the outraged party doesn't realize it. Evolution instilled emotions into us to serve specific purposes. Sadness to advertise to our community that we deeply regret a recent misfortune, as opposed to planning to opportunistically benefit from someone's untimely death. Happiness and merriment to bond with our allies and advertise that, indeed, they are our allies. Anger to warn our enemies that we will fight if slighted, even seek revenge (a costly endeavor in a one-off, forward-looking "cut-your-losses" sense) if attacked. Moral outrage warns the community that certain affronts will not be tolerated. Evolution gave us all this emotional baggage, but how we carry it is up to us to decide.

I said in my earlier post, linked to above, that baring your moral outrage is sometimes justified, and I still think that. If you're on the brink of some important policy change and you need to fight dirty just this one time to push it through and you're 100% confident that you're right about this one, fine. I understand. Just be careful that you don't get to used to doing this, so you think an angry outburst will win every argument. ("Hey, that was easy! That guy really shut his mouth. I'll remember this for next time.") Doing it too often invites blow-back, poisons your moral credibility, and backs you into a corner.

Thursday, September 14, 2017

Why It’s a Bad Idea To Discuss Salaries With Co-workers

Sometimes within a company there are large salary differences between individuals of similar job descriptions and seniority. Some commenters who have noticed this phenomenon take the most cynical possible view of it: Here is evidence that employers are screwing over their workers! If they’re willing to pay $100k a year for the more tenured engineer, how can they justify paying the similarly-qualified younger engineer only $50k?

Their solution to this perceived inequity is a pretty terrible one: Workers should supposedly be discussing salaries with each other and making comparisons with each other. The ones who think they are underpaid should demand raises.

I think this is a bad idea for lots of reasons. First off, how do you know that two employees are actually similar? People tend to be oblivious to their own flaws. Often times an incredibly smart and talented worker is completely unorganized. Often a lack of social skills makes an employee less valuable. We’re not all independent skilled craftsmen building widgets and selling them on the open market. Most of our jobs require some kind of coordination with other human beings. The inability to interact pleasantly with other people, to pick up on unspoken cues, to navigate bureaucracy, to be aware of power differentials, can really hurt you. I suspect there are a lot of extremely skilled programmers out there who are under-employed (even unemployed) because they can’t interact delicately with their customers (“customers” often being users of their software within the same company). As I like to say, “Computers do exactly what you tell them to do, with no appreciation of your actual intention. Some human beings have learned to adopt this nasty habit. We call them ‘computer programmers.’” The point is that some people are perfectly happy to take a specification from a customer, program it, and dust off their hands as if they were done.
Customer: “This isn’t working the way I expected. This software is useless.”
Programmer: “I did exactly what you told me. See the specs, bitch!”
Many of these kinds of disconnects can be fixed just by having people talk to each other, with the customer trying to understand what is technically feasible and the programmer trying to understand what the customer is actually trying to accomplish. I tried to explain my intention to one clueless programmer and he literally responded with, “Yeah, I don’t care what you’re trying to do.” To be fair, this guy was trying to be deliberately abrasive. He succeeded. I think some people literally take this attitude toward their work. To be perfectly clear, I am not just picking on computer programmers here. “I did exactly what you told me, so that should be a safe harbor.” versus “I took the time to understand your intention, and got you what you actually wanted.” This can be the difference between a low-value employee and a high-value employee, and it can differentiate two individuals who look the same on paper.

So that’s the first lesson. There are differences between employees that the employees themselves may not appreciate. There's another good reason to have a "don't talk about salary" norm or even an enforced rule.

Suppose two employees really are equally qualified, but one makes much more than the other. Should they be discussing this and demanding higher pay for the “underpaid” one? I still say “No.” If the junior engineer was willing to jump at his first job for $50k/yr, but that company needs to pay $100k/yr to keep its senior engineers from retiring or leaving, that’s none of your business. If you can find $100k/yr elsewhere, more power to you! Go for it! Find an offer and leverage it to get a pay raise, or pursue that other opportunity. That’s the real proof that you’re worth more as an employee: someone is actually willing to pay you more. Steven Landsburg makes this point especially well in this post. If someone is paid their fair market wage, they will either sit happily or calmly take another job elsewhere. However, if someone is currently working for their best possible employer, they will stay at their job and yell, kick, scream, and bite for higher pay and better perks. That company with a huge salary differential between similarly productive workers is probably paying appropriately. It would probably go out of business if it paid its young engineers $100k/yr, as it would if it tried to pay its older engineers $50k/year thus driving a mass exodus of firm-specific capital. Operating within a narrow set of constraints probably led this firm to a pay structure that looks inequitable to outside observers. This is similar to price discrimination, but in this case the buyer is paying different prices for indistinguishable goods. The important “keep your eyes on the prize” point here is that the company might not exist at all if it tried to pay everyone the same (even the same for employees with similar productivity).

Companies have damn good reasons for discouraging employees from discussing salaries with each other. It breeds resentment. One employee with an unrealistic sense of their ability starts comparing himself to someone who is actually a more productive worker. The unqualified worker resents the other person’s higher pay and status. The more qualified worker resents the comparison. Perhaps someone has to explain to the low qualified employee why his pay is as low as it is, and it sounds like a personal attack. How could it not? “Here is what’s wrong with you… And we use these items as reasons to pay you less.” The less qualified employee resents having their problems stated out loud. If lack of emotional maturity is one of their flaws, the whole thing escalates. Everyone is probably happier not making explicit salary comparisons. Otherwise you get these obnoxious relative-status conflicts all the time. That would make for a pretty hostile work environment, one that almost nobody wants to work in.

I have read that if you introduce a bunch of chickens into a new pen, their stress levels are very high until the “pecking order” is established. When the relative status of every chicken is implicitly understood, stress levels then fall. “Everyone talk about salaries and ask for a raise” is like the community of chickens suddenly all saying, “Wait a minute, I’m getting a raw deal. Let’s re-evaluate. Constantly!” The analogy fails, of course, because employees aren’t confined to a pen. They are free to pursue a different pen if the existing pecking order is giving them a bad deal. 

So don't just ask for a raise. Ask for more responsibility. Ask to be trained for a better-paid position, or seek that training on your own. Ask if you are in the running for a management position, and how you can be slated for such a position if one opens up. Do constant continuing education, constantly be learning new skills, and stay up-to-date on trends in the industry you work in. Try to do things that will make you deserve better pay. Asking for more money for the same work is very risky. Remember that one of the possible answers to the request is "No" and think about the awkwardness of getting that response. This awkward truth, that you and your boss have different estimates of your value, must suddenly be stated out loud. You'd better be damn confident it will pay off before you try it.


Obviously this argument won't apply to all workers. Government employees, for instance, aren't necessarily being hired on an open market. See the Landsburg post linked to above. If you're in far-and-away the best possible job for you and have nowhere else to go, it doesn't make sense to seek a better deal somewhere else. Someone in this position might be able to coax a higher pay out of the same employer for the same work, but like I said it's risky.

In my case, it would be almost incoherent for me to barge into my bosses office, balls blazing, and demand a raise. I'm an actuary. The salary structure is laid out very clearly for actuaries, based on a combination of exams passed and years of experience. During the annual performance review, when raises for the next year are determined, the bosses themselves compare the salaries of similarly situated workers and try to close the gaps whenever necessary. I'm not saying everyone is in this kind of job, but many employers are diligently eliminating pay differentials between similar employees. The mistake is to think that a pay differential is completely arbitrary or always exists for cynical reasons.

Some people may try to shoe-horn this post into a discussion about the gender pay-gap (the one that shrinks to a few percentage points when you actually start controlling for pay-relevant factors). Please don't. That's not what I'm talking about here. 

Who Are My Foreign Readers?

I see traffic from Russia, Sweden, Brazil, France, Germany, and lots of other countries. Don't worry, I don't see IP addresses or anything, just rough traffic volume by country on a colored map. This makes me very curious. Who are you? Interested readers? Spammers looking for a mark? Americans reading the internet via a dark web browser routing through proxy servers in foreign countries? Feel free to completely ignore, or share something about why my blog interests you. I get almost zero feedback from my readers, so naturally I'm curious about who is reading and why. Thanks for listening.

Sunday, September 10, 2017

Ensuring America's Health

I recently read the book Ensuring America's Health by Christy Ford Chapin. Listen to her interview with Russ Roberts on Econtalk to get an idea of what the book is about.

(About 3/4 of the length of this entry is composed of excerpts from the book, all confined to the bottom of the post. Sometimes very long posts deter readers from starting. So feel free to consider the line 1/4 of the way down the page the "end", with the excerpts below that line being footnotes.)

Ensuring America’s Health is an interesting book because it shows the reader another way that medicine can be. When I say “medicine” I am mostly talking about how medical costs are financed. The book follows this evolution thoroughly. What exists now in America is not the only option, nor are the various European models the entire world of possibilities. (Singapore, anyone?) There were some obvious problems with early 20th-century medicine (again, regarding the financing rather than the hare-brained medical practices of the time). But as Richard Epstein puts it, “The study of human institutions is always a search for the most tolerable imperfections.”

The point is not to say, “Let’s return to medical financing as it was done in the 20th century.” The point is to say, “What can we learn from a different set of institutional arrangements than the ones we’re used to?” And “What arguments are made today that historical evidence can refute?” Americans in the early 20th century were much poorer than Americans today, and they mostly paid out of pocket for their health care. That doesn’t mean that out-of-pocket payments are the best possible policy today, but it certainly refutes the commonly held notion that Americans today are too poor to pay out-of-pocket. (All Americans? The average American? Everyone below the 30th percentile? All health expenses, or just large, infrequent ones? Claims about the unaffordability of healthcare are usually so imprecise as to be meaningless.)

Chapin starts by describing typical financing in the early 20th century. There was essentially no health insurance. Doctors would put up a shingle saying “Consultations from $1 to $10.” Since doctors knew who their patients were, they knew something about each patient’s ability to pay. They would often make house-calls. They probably knew what profession the head of the household worked in. They probably knew who was in the local country club and who was not. In a small community where everybody knows everybody, your material means are basically public information. With all this information, doctors were in a good position to practice price discrimination, the charging of different prices for the same service based on willingness to pay. Think student prices vs list prices for textbooks and software. Or senior discounts, which are offered not because seniors are poorer than average (far from it) but rather because seniors often have the free time to shop around for the best price. (There are sneakier versions of price discrimination, too.) A doctor’s office is an almost perfect opportunity to price discriminate. There are high fixed costs, like keeping the lights on in your office, paying staff, and keeping an inventory of medical supplies. (Admittedly this would all look very different in 1920 than it looks in your doctor’s office today, but the concept is the same.) The “marginal patient” is someone who you can profitably treat for $1, but if you charge $1 to each patient you will go broke and your office will shut down. The cost to treat the marginal patient is much lower than the average cost of treating a patient, in other words. The best strategy here is to charge high prices to people with high willingness to pay (the rich!) and offer lower prices to people with low willingness to pay (the poor, who often lack both willingness and ability to pay). Also helping the price discrimination equilibrium, doctors sell something that cannot be resold, and doctors often hold some degree of monopoly power. Resellability and robust competition can cause the price discrimination equilibrium to break down. (Economic exercise: Think about why this is the case.)

Doctors also were under a professional code of conduct that frowned upon (even shunned) doctors who failed to take care of the indigent. Chapin managed to speak to many doctors who had practiced in the 50s and 60s, some even in the 40s, before the era of Medicaid and Medicare. Many of them wrote off a huge amount of charitable medicine. They either did a lot of pro bono work or decided the better of trying to collect on patients who failed to pay up. Something like 30% of the medical service was provided completely free of charge. Keep in mind also that for medicine that wasn’t free, the rich paid more than the poor. The sliding scale doctor’s fee in effect created private income redistribution, or progressive taxation if you want to think of it that way. The poor patient just barely paid for the doctor’s time and the material cost of the bandages used on him; the rich patient paid the electricity bill, the phone bill, the nurse’s and secretary’s salary, and the other fixed costs of running a clinic. 

I'm suddenly reminded of someone who has complained to me about the exorbitant price he paid for a minor surgery. I listened sympathetically, but I wanted to say, "Sorry, Jack. The hospital marked you as a well-paid professional and decided to use you to cover its fixed expenses. Poorer patients get a better deal, as they should." If you're poor, you're going to pay just enough to barely cover your marginal expenses (or perhaps even less than that). If you're rich, you're going to be the one paying to keep the lights on and amortizing the cost of that new wing of the hospital.

The American Medical Association (AMA) was incredibly powerful, more so than it is today. They fought policy changes and legislation that would have changed the practice of medicine. More surprisingly, they fought private efforts to practice medicine under a different model than the one described above. They wanted to maintain physician control over the practice of medicine. They realized that if the practice of medicine were somehow corporatized or if insurance companies began covering the costs of medicine, then other parties would begin dictating how medicine was practiced. No more, "Okay, that'll be $10, because I say that's the price." Suddenly there would be an insurance company saying, "No, we won't pay $10. We'll pay $5. Or we'll place our customers in a different network that plays with us."

Health insurance became commonplace in the 1940s and 1950s, but there were serious problems with the market. Insurance customers wanted a lot more coverage at far lower prices than they were actually getting. (What else is new?) Insurance companies were characteristically conservative. They understood well the adverse selection problem and the moral hazard problem, both of which drive the cost of insurance to astronomical levels and thereby drive insurance premiums to unaffordable levels. Adverse selection means the tendency for sicker people to acquire insurance, knowing that they will be more likely to use it. ("I've got some of the warning signs for diabetes. Better get me a health insurance policy with a great big coverage limit.") Moral hazard means the tendency to incur more costs knowing that someone else is covering them. ("Well, since my insurer is paying for it, throw in a CAT scan and an EKG while you're at it.") Insurance customers felt like the premiums were always too high and the coverage terms were never quite generous enough. Insurance companies, on the other hand, constantly saw their costs running away from them and issued coverage restrictions and premium increases. Many insurers thought health insurance was fundamentally fraught with cost problems and never entered the market in the first place. Existing health insurers exited the market. Americans wanted an "affordable" health insurance with generous coverage, but insurers were unwilling to offer it.

Chapin tells this narrative of customers asking for cheap, generous health insurance and insurers refusing to offer it. I think she could have made the point more explicit: insurance customers wanted a free pony. They basically wanted something that was economically unfeasible. When the private market didn't deliver, that should have been a signal that the thing they "wanted" was in reality something they were unwilling to pay for. This was a classic case of professed preferences being different from revealed preference. Saying you want something versus actually being willing to shell out for it are two different things.

When the private insurance market fails to deliver the goods, demand for political solutions begin to emerge. It was bi-partisan: Truman, Eisenhower, Kennedy, Johnson, and Nixon all had expansions of the medical welfare state on their agendas. The AMA did a lot of red-baiting to counteract these initiatives. They fought all of these attempts to socialize medicine, referring to any change in medical policy as a step toward communism. That's not to say they were wrong to fight any of the specific "reforms"  in health care finance. It's just that they lost the public's trust and kind of ended up looking like ogres. The were tone-deaf to the average American, and ultimately they blew their moral authority. To be fair to the AMA, the average American wanted a free pony full of medicine (and frankly it still does). The AMA may have been right in every single policy fight. Right or wrong, their rhetoric made them look unsympathetic. People stopped seeing them as a legitimate authority, and politicians started ignoring their lobbying efforts.

Policy holders complained about the skimpy medical coverage offered in the private market. When political pressure forced insurers to offer more "generous" coverage provisions, those same policyholders complained about the high premiums, which naturally were necessary to cover those generous coverage provisions. Insurers, conservative by nature, were nervous about offering expanded coverage at any feasible price.

Senior citizens were routinely denied health insurance policies, and for good reasons that are familiar to anyone in the insurance industry. In order to be "insurable" in the traditional sense, a risk needs to be "fortuitous from the point of view of the insured." (Fortuitous means "unpredictable", not "fortunate" as it's often used in casual speech.) The poor health experienced by seniors is extremely predictable. The proper way to finance predictable expenses is through saving and/or borrowing, not through an insurance policy. You might want to insure against the possibility that you will have an especially rough decade in your 60s or 70s. But everyone should count on having to use more and more healthcare as they get older. If the typical person has one surgery in their 50s, two in their 60s, and a hip replacement in their 70s, and some chemotheraby in their 80s, and a steady increase in the frequency of office visits over that entire span, then these are predictable expenses, not insurable in the traditional sense. Insurance should not be used to finance this absolutely typical trajectory. In fact it's a waste of money to finance these expenses with insurance, because that incurs the added expenses of claims handling and all the other functions an insurance company performs. Also, the adverse selection and moral hazard problems are looming large. The particularly unhealthy seniors will seek coverage, knowing they will make more use of it. And once insured, even a typical or especially healthy senior might say, "Well...since Blue Cross is paying, maybe I'll do two checkups every year." Insurance companies thoroughly understood these basic concepts of finance, risk, incentives, and probability. But the general public did not. They demanded a free pony. And they demanded expanded coverage for a sympathetic demographic: the elderly. Medicaid, directed specifically at people who lacked the means to pay for healthcare, might be defensible. But Medicare, directed generally at the elderly, was not. A financially unsound insurance policy doesn't suddenly make sense because the government underwrites it. In fact, it tends to make the problem worse, because government programs respond only weakly to financial constraints, and bad government programs never go away. Free markets are disciplined by profits and losses. An unsound insurer can offer overly generous coverage for a little while, but eventually goes insolvent, or the very high premiums drive away all the customers.

Insurance faces some fundamental constraints. I see the political demands for government solutions as attempts to ignore these constraints, not to simply dampen one constraint at the expense of intensifying another. Premiums must cover the cost of paying claims and the expenses related to running an insurance company. A government program that provides health insurance is subject to the same basic "money in equals money out" constraint. Premiums must reflect the relative costs of providing insurance to individual policyholders. Otherwise the good risks don't bother purchasing insurance. If you just charged everyone the overall average price, this is too high a price for the best risks, so they often go without (although particularly risk-averse individuals might buy it anyway). You get the dreaded death spiral, whereby premiums increase, driving away the low-risk individuals, which further increases the average cost of insurance, driving up premiums higher, ad infinitum. (Perhaps not literally ad infinitum, but at least until you reach an equilibrium where far too few people are insured and many of those who are insured are paying way too much.) Finally, the entity paying for medicine has to have the ability to say "No." Otherwise insurance policy holders will ask for the moon. You have to have coverage limits and coverage exclusions, and someone has to push back when unreasonable claims are submitted. These constraints do not disappear when government takes over the insurance sector or issues edicts to the private market. You can try to trade off one constraint for another, but they don't go away.

It would be unfair of me to expect Ensuring America's Health to conduct a thorough cost-benefit analysis of health insurance policy based on these insurance concepts. Chapin's book is an excellent historical account, with events laid out logically and in chronological order. And she actually does a very good job of introducing insurance concepts and even some basic economics. She does so mostly in a "positive" rather than "normative" sense. As in "This is why insurers behaved as they did, this is why insurance customers' demands weren't met by the private market" versus "This is how we ought to design a working health insurance system..." But in my head I kept saying, "The public is asking for the laws of economics to be repealed. Chapin should tell her readers why this is impossible."

The very last chapter of the book brings us up to the present day. It is objective and even-handed. I don't get the sense that Chapin supports or objects specifically to Obamacare. But I found it slightly irksome that she calls it "comprehensive reform". I think Obamacare didn't qualitatively change the health insurance market. It just doubled down on all the bad things we were already doing. (For example, intensifying the link between insurance and employment, intensifying coverage mandates, intensifying restrictions on actuarially sound pricing. It did, unintentionally, lead to more cost sharing because unreasonably high premiums led people to choose higher deductibles, so a minor win there!) But that's just me being picky. I'm not going to consider that a valid criticism of her excellent book. In this last section, she points out that doctors are very comfortable with insurance being the first-dollar payers of all medical costs. I think many hospitals and private practices see this as simply the way things are and support any policy that  helps them manage their revenue streams more predictably. The Affordable Care Act was probably good for them from that point of view.


Below are some passages I highlighted while reading.

From page 14:
Around the turn of the twentieth century, a typical office sign read “Consultations – from $ 1 to $ 10, Cash.” 10 To formulate the patient's bill, doctors used a sliding fee scale, which not only took into consideration services performed but also the ability to pay...Professional ethics obliged doctors to care for all sick individuals, regardless of ability to pay, and accounting ledgers indicate that charity work was a regular activity for most practitioners. Even in urban areas – where, away from the community's watchful eye, social mores frayed – physicians found serving in charitable hospitals desirable for training students and for acquiring experience managing unusual and interesting cases. Sliding fee scales compelled wealthier patients to subsidize at least some of the physician's charitable care; however, they also encouraged the best doctors to focus on engaging a wealthy clientele.
 Page 15:
As efficacious medical treatments and Joseph Lister's standards of cleanliness helped transform hospitals into more appealing institutions of care, administrators relied on doctors, not only to continue providing charitable services but also to attract middle- and upper-class patients who could deliver much-needed revenue.
 This should give you an idea of how medicine was practiced and how much discretion physicians had. And the following should tell you how much power they held (page 24):
Moreover, doctors organized through local medical societies frequently colluded against “unethical” practitioners by refusing to refer patients to them. AMA leaders made explicit their intention to professionally harm any physician who failed to uphold the embargo against group practice and insurance, as, for example, in this 1934 JAMA editorial: "The young physician who is tempted by the offer of some commercial agency to enter into such schemes of combinations should bear in mind that he thereby jeopardizes his entire future in the practice of medicine and sacrifices the medical birthright for which he has already paid six or seven years of his life." Although many practitioners were willing, even eager, to participate in multispecialty groups and prepaid programs, the example of doctors who lost their licenses or hospital privileges convinced most physicians to shun such plans.
 In other words, the AMA sanctioned doctors who accepted insurance arrangements the AMA did not approve of. They also sanctioned models of medical care that they didn't approve of. Chapin points out that they sowed the seeds of government intervention by constraining the market for medicine:
Yet even as AMA leaders successfully implemented their economic vision, they simultaneously, by undermining market modernization, paved the way for the very types of governmental intercession they were so keen to avoid.
The AMA got hit with an anti-trust lawsuit (page 26):
 The month of July 1938 presented extraordinary challenges to AMA political and economic clout. One week after the national conference, AMA leaders, fearing that proposals for government-funded insurance were gaining ground, met with the Technical Committee to offer a compromise. AMA representatives agreed to support each of the committee's recommendations – all of which they had previously opposed – as long as policymakers dropped their advocacy of compulsory health insurance. Technical Committee members refused the deal. Then, on July 31, the Department of Justice filed an antitrust suit against the AMA. Persecution of group doctors had finally caught up with them. AMA officials had collaborated with Washington, D.C. hospitals to deny admitting privileges to physicians associated with the Group Health Association (GHA).
Chapin describes how the health insurance model creates perverse incentives to overcharge (page 28):
Rather than establishing direct financing relationships with physicians, insurers had to send indemnity payments to subscribers. This compensation structure preserved individual physician-patient financing and allowed doctors to set their own fees. However, indemnity policies drove up service prices because they attenuated the fee-setting restraints placed on doctors when patients paid the entire bill from their own pocket. Physicians could more readily rationalize bill padding when a nameless, faceless company was supplying part of the payment. Sliding fee scales exacerbated the situation. Because doctors were accustomed to setting fees according to the patient's ability to pay, many practitioners charged insured patients higher prices than uninsured patients. Despite this obvious glitch, AMA leaders maintained that “cash (indemnity) benefits only will not disturb or alter the relations of patients, physicians and hospitals.”
Under the "indemnity payments to subscribers" model, the doctor charges whatever the heck he wants, and the insurance company has to pony up. An alternative model is that the insurer negotiates with the provider over what will be paid for and what will not, what fees will be paid for which services, which providers are in the network, etc. The AMA's fears about loss of physician sovereignty are coming true. Inevitably insurers will begin to demand some control over how medicine is provided, given that they have to pay for it. They won't tolerate doctors and hospitals saying, "This is the cost. Shut up and pay it." forever.

I read the following passage thinking, "Ugh. Here is the original sin."
Long-standing Internal Revenue Service (IRS) rules permitted businesses to count employee fringe benefits as tax write-offs. Thus, when the War Labor Board issued a wage freeze during World War II, employers could simultaneously decrease their tax burden and attract scarce labor by furnishing health insurance.
This is it, folks. This is why health insurance is tied to employment. It's a historical accident. The resulting policy is terrible.

Chapin concludes the chapter covering the 30s and 40s:
The rules and regulations that AMA leaders attached to medical prepayment produced a costly and inefficient insurance company model. Although there existed many possible ways of configuring health care financing and delivery, AMA heads settled on a faulty model because they believed it offered doctors the best prospects for safeguarding professional power.
 On the inadequacies of the private health insurance market:
The elderly and chronically ill, who were generally unable to obtain insurance because of their employment status and the financial risks of covering them, relied on family resources or charity for medical care. Moreover, even insured individuals could amass large medical bills. Commercial insurance policies usually covered only a portion of catastrophic costs. Blue Cross provided more generous service benefits; however, the nonprofit's focus on first-dollar costs meant that seriously ill subscribers ran up considerable debt after their allocation of covered hospital days expired.
The problems of adverse selection and moral hazard rear their ugly heads (page 46):
Just as underwriters had forecast, insurance company arrangements caused medical service costs and, concomitantly, policy premiums to escalate swiftly. During the late 1940s, right as health insurance gained firm footing among the populace, medical care cost surges began outpacing price increases in all other goods categories in the Consumer Price Index.
In other words, any serious insurance professional could have told you this would have happened. The sick, with very high healthcare costs, will be more likely to acquire insurance (the adverse selection problem). And, once  insured, patients will tend to seek more care than they otherwise would (the moral hazard problem).

AMA leaders complain (rightly in my opinion) about weird, unreasonable expectations of insurance customers. As an actuary, I can attest that the average customer has very odd ideas about insurance. From page 48:
“The man on the street,” groused AMA official Frank Dickinson, “has been lulled into believing that he is being robbed if he does not get at least one claim check from his health insurance plan almost every year.” Dickinson, who served as director of the AMA's Bureau of Medical Economic Research, observed that “most car owners deem their collision insurance ‘comprehensive,’ even though they must pay the first $50 of a collision claim.”
Private insurers tried to educate their customers and potential customers about the underlying drivers of insurance costs (page 49):
Industry trade associations dispatched speakers to civic organizations, women's groups, and businesses to lecture policyholders on the merits of catastrophic or major medical coverage. Covering small claims, contended industry representatives, produced a system of “trading dollars” that only drove up premium prices. Pamphlets distributed to consumers argued that the purpose of insurance was to protect against large, unforeseeable financial risks, such as those posed by serious accidents or grave diseases.
Chapin compares the "insurance" model to the prepaid medical group model (page 52):
Proposals to strengthen prepaid group policies failed, and they remained a small market niche. Yet the mere existence of prepaid groups – which commonly provided more generous coverage at lower costs than insurance companies – highlighted inefficiencies in the broader market. Indeed, during the 1970s, policymakers would return to the prepaid group idea under the refurbished label of “Health Maintenance Organizations.”

 The medical industry realized it would have to expand coverage voluntarily or else government would force its hand (page 58):
During the 1950s, almost every meeting or publication of private medical groups featured speeches and articles calling on doctors and insurers to prevent federal interference in health care by expanding coverage. An insurance executive summed up the degree to which political purposes had shaped the goals of private interests: '[S] omehow and in some way the base of insurance coverage for protection against disease must be broadened. The Utopia is, of course, that we may evolve a system of total and comprehensive medical care … we should make every effort to achieve this goal by voluntary means instead of by compulsion.'  To thwart legislative initiatives, not only did voluntary interests have to cover more citizens, they also had to reconceptualize the function of insurance. Stripped-down policies that covered only a portion of hospital costs were no longer feasible – nothing less than all-inclusive benefits would satisfy the objectives laid out by policymakers.
You can sort of get around the adverse selection problem by issuing "group rates" to all the employees working for a certain employer (for example; there are other conceivable ways of grouping for the sake of group rating). But group rating isn't exactly a perfect solution. Page 62:
Because group benefits were linked to employment, women were less likely to have coverage. Only one-third of women worked for wages during the 1950s and of those, two-thirds worked part-time or seasonally. Although employers increasingly offered workers coverage for dependents, the practice linked women's health care access to a male breadwinner. Additionally, as benefits were related to one's position in the corporate hierarchy, African Americans, the working poor, and rural laborers often lacked insurance. A 1955 survey found that although 80 percent of families with annual incomes over $ 5,000 had insurance, only 40 percent of families with earnings below $ 3,000 and 45 percent of rural residents owned medical coverage.
On the AMA's dysfunction (page 67):
Through these battles, physician leaders frittered away political authority with inflexible arguments that often failed to advance beyond red-baiting. Many of the AMA's difficulties, including its inability to formulate an effective long-term political strategy, were rooted in organizational characteristics. The leadership was drawn from an elite group of physicians who were largely detached from the concerns of rank-and-file doctors. Moreover, two governing bodies shared and thus bickered over responsibility for association policies. When the leadership did find consensus on an issue, that decision was then ground through a sprawling bureaucracy that chipped away at the original objective until only a withered fragment remained. In combination with the association's political tactics, the organizational structure undermined member cohesion. Numerous political fights aggravated members with differing ideological opinions, while the association's decentralized, democratic structure bred conflict.
More on the AMA being out-of-touch with its constituents (page 81):
Complaining about the “big-shot specialists” who issued “communiqu├ęs from the Chicago Citadel,” one member observed that “[ n] o one knows better what the common people need and desire, what local handicaps to health services there are, than the little family medic – who is never asked to express his opinion.” Grassroots sentiments seldom found a hearing before national leaders because years of moving up the ranks through county, state, and then national office required physicians to impress already-established officials. Consequently, the AMA's leadership seemed inflexible and aged. With an average age of fifty-nine, most delegates were nearing retirement.
And they were prescient enough to poison their future moral credibility, too (page 88):
Particularly in the South, black physicians were refused admission to the AMA; indeed, most of them would not have bothered to apply.
The following paragraph gives a pretty good summary (page 94):
Once the AMA designated the insurance company model the only suitable form of medical prepayment, physicians ceded every opportunity to move the health care economy toward more efficient and inexpensive care. AMA leaders went through the motions of expressing interest in economic matters. They created councils and committees to study insurance and appointed representatives to confer with insurers. They also knew what they did not like: any third-party financing that might infringe upon physician autonomy or pay. This reactionary, backward-looking stance of AMA leaders left to insurance companies the primary responsibility for developing the health care market and also created political difficulties as policymakers looked for ways to make medical services more accessible.
 Chapin describes an episode near the turn of the century where financially sound insurance companies acquired insolvent and shady insurance companies that were unable to cover their liabilities. These insurers went beyond just covering medical costs. They actually took on the duty of improving public health by educating their policyholders (page 99):
Both the LIAA and ALC sponsored advertisements, placed magazine articles, and published pamphlets emphasizing the industry's financial stability and spotlighting the peace of mind that families with insurance enjoyed. Representatives from both associations worked closely with state regulators to develop laws designed to drive out fraudulent and marginal companies. Northeastern firms banded together to assume the policies of more than thirty companies that were unable, because of financial stress or fraud, to fulfill subscriber obligations. The largest industry enterprise, Metropolitan Life Insurance of New York, initiated a decades-long program that, in an effort to protect policyholders and often at the behest of state regulators, regularly purchased failing companies. Leading firms also launched public health campaigns that simultaneously boosted the industry's image while serving customer needs and reducing policyholder mortality rates due to infectious ailments. A multicompany bureau was established to provide subscriber medical exams and send representatives to discuss any identified health problems with the policyholder's doctor. Metropolitan dispatched nurses to tend to sick policyholders, primarily by instructing their families on hygiene and caretaking duties. The company carried out a massive health education drive that, through leaflets and advertisements, informed the general public about topics ranging from the “care of children” to “flies and filth.”
The following quote is a word of caution from insurance companies to congress. It is basically a warning that government is not magic (page 106):
Government reinsurance of health insurance plans would introduce no magic into the field of financing health care costs. … Reinsurance does not reduce the cost of insurance. Reinsurance does not make insurance available to any class of risk … not now within the capability of voluntary insurance to reach.
A brief note on the competitive nature of the market in 1950:
By the early 1950s, approximately eight hundred commercial companies offered some form of medical insurance.
 Chapin tries to correct a mistaken narrative of the history of 20th century health insurance (page118):
Scholars who assume that commercial companies cherry-picked healthy employee groups out of the broader insurance pool and left nonprofits to cover higher-risk individuals have somewhat mischaracterized the industry's behavior. 93 By 1960, commercial companies sold approximately 40 percent of their hospitalization policies to individuals. 94 Moreover, during the 1950s, about three-fifths of HIAA companies took steps to limit the quantity of subscriber renewals they would refuse due to declining health. More than seventy firms introduced policies that were guaranteed renewable for life – though, of course, premium prices could rise. 95 In sum, to undermine arguments in favor of government intervention in their field, commercial leaders had to demonstrate their industry's social responsibility and pursuit of the public interest; they attempted to accomplish this mission by insuring a significant portion of the individual-purchase market, including some high-risk subscribers. At the end of the 1950s, when reformers ramped up their campaign to supply aged citizens with health insurance through Social Security, HIAA firms employed extraordinary measures to combat the legislation. Insurance companies began permitting workers with policies supplied through their employer to retain coverage upon retirement.
Attempts to standardize care for the sake of keeping medical costs reasonable were resisted by local doctors who resented the "formula." Today you sometimes hear the war-cry "I don't practice cook-book medicine." This apparently has a long history:
Physicians who jealously guarded their control over plan operations branded the MIA [Medical Indemnity of America] “a grab for power” by national leaders. A frustrated NABSP [National Association of Blue Shield Plans] representative commented that it “was as if General Mills tried to do a national business while each of its local plants made cake-mixes by its own formula.”
I'm suddenly remembering that the acronyms made the book hard to read in parts. You can easily lose track of who's doing what to whom and why.

This lovely passage describes how mere economics can constrain your social conscience (page 146) :
Neglect of basic tenets of insurance also contributed to Blue Shield's anemic pecuniary condition. Nonprofit leaders chided for-profit companies for hesitating to underwrite health insurance and then entering the field with an abundance of caution. Yet the commercial industry's behavior reflected a rational fear of the costs associated with the insurance company model and a more sophisticated understanding of actuarial principles. Describing how early nonprofit plans determined the financial reserves necessary to support policy offerings, one administrator proudly explained that “there were no statistics or actuarial bases, it was merely an idea.” At the time, commercial insurers also lacked dependable actuarial data; however, nonprofit leaders evinced a more reckless, seat-of-the-pants attitude. Even at the end of the 1950s, once commercial insurers had begun constructing reliable morbidity tables, nonprofit administrators continued to lean too heavily on “whims and fancies” rather than statistical formulations.
In other words, you can have extremely good intentions, but if you run out of money you won't do anyone any good.

Chapin has a good passage on  how markets and prices are supposed to function, and how government and even private insurance can screw with the price signal (page 156):
Markets discipline through price. In properly functioning markets, prices convey information about supply and demand to consumers and producers. If demand increases for a particular service, prices rise, thus signaling to consumers to buy prudently. Price increases also encourage providers to enter the market, which, by raising service supply, drives down prices. In government-directed systems, officials hold down prices by setting agency or regional budgets. Under the insurance company model, the liability of paying for care was transferred from patients to insurance organizations, thereby altering the perception of prices in the minds of consumers, sending an artificial signal that they had been lowered. Meanwhile, insurance companies lacked effective means of regulating physicians and hospitals. Consequently, neither markets nor the government controlled provider fee setting or service supply.
On how insurance induces insureds to use unnecessary medicine (page 158):
An investigation conducted by Michigan Blue Cross during the early 1950s determined that one-fifth of subscriber hospital days were unnecessary and 28 percent of admissions involved some element of “faulty use.” 18 Later in the decade, a nationwide study found that on an annual basis, per one hundred individuals, uninsured persons spent an average of seventy days in the hospital while insured patients racked up one hundred days.
Third-party funding in combination with doctors’ traditional reliance on sliding fee scales produced a ready vehicle for physician bill padding. Because doctors had always taken the patient's ability to pay into account when calculating fees, many practitioners inflated service prices for insurance subscribers. Bill padding diminished the value of indemnity policies. For example, a 1950s Indiana investigation revealed that approximately half of the indemnity policyholders under study were charged higher fees than uninsured patients.
A short passage summarizing how medical insurance changed the way medicine was financed (page 191):
Depending on the stability and affluence of their customer base, physicians wrote off between 10 and 40 percent of patient bills as uncollectible. Because of the difficulty and expense of collecting patient fees and because some insurance subscribers chose to pocket indemnity checks rather than pay their medical bills, a number of physicians, flouting the preferences of AMA leaders, lobbied insurance companies for direct compensation.  By the mid-1960s, physicians received between 30 and 70 percent of their income directly from insurance companies. These direct financing linkages severed doctor-patient monetary relationships and became crucial conduits through which insurance companies intensified cost containment measures.
Chapin summarizes the nationalization of medicine in the 1960s to present (page 234):
Spiraling costs drove insurers and, after Medicare's passage, government officials to attempt to close the institutional gap that separated them, the financiers, from doctors and hospitals. Third parties implemented cost containment measures, which entailed insurers wading into the medical delivery process to supervise physician work and regulate provider remuneration. This process unfolded gradually and generated substantial conflict as organized physicians still retained a good deal of influence, which they wielded in political fights and in battles against third-party payers. Nonetheless, the very outcome that AMA leaders initially sought to prevent had come to fruition – physicians were now bound within a corporate structure of their own, inadvertent making.
And the absurd cost inflation that it lead to (page 235):
By 1970, physician fees were growing at twice the rate of average consumer price increases while hospital charges were rising five times more rapidly. The “massive crisis” and “soaring” costs of a system “teetering on the brink of disaster” dismayed voters and policymakers and triggered a litany of questions about the shortcomings of U.S. health care.
Read the whole thing.

Wednesday, September 6, 2017

The Costs of Drug Use: Overdose Risk

I wanted to do a little back-of-the-envelope calculation to show how much drug users are "paying" for their habit. One of the risks of drug use is that you will overdose and die. I want to quantify that, just to get a rough feel for the magnitudes.

To do this, I'll need to use a concept that makes some people squeamish: the statistical value of a human life. When making certain policy decisions, it is necessary to use this concept. Economists measure this by figuring out how much people are willing to pay for safety features that avoid, say, a 1 in a million chance of death. Or they figure out the pay differentials between risky jobs and similar jobs that are less risky. They come up with numbers such as, "People will pay $6 to avoid a 1 in one million chance of death, therefore the statistical value of a human life is $6 million, according to their own valuation." You can argue that the true number should be higher or lower, but you can't avoid putting some kind of reasonable number on the value of a human life. Carried to its logical limit, the "human life has infinite value" position would have us spending all of society's wealth on ridiculous safety features. I'm going to use the smallest value in the above link for the sake of my calculations below: a human life is worth roughly $6 million. If valuing a life is too hard a concept to get your head around, think of it as "It's worth about $60,000 to avoid a 1% chance of death" or "It's worth about $6,000 to avoid a 1/1,000 chance of death."


There were about 4.5 million users of cocaine in the past year, and there were about 5,000 deaths last year. This comes to a 0.11% death rate. Converted to a dollar figure, that's about $6,600 (= 0.11% * $6 million). So the average cocaine user is effectively willing to spend $6,600 on his habit on this consideration alone. Of course, you have to add to this the cost of acquiring the drug, the opportunity cost of whatever else the person would have done with his time, and health problems due to cocaine use that fall short of actual death. I'm not making the case that cocaine use is destructive so we have to stop these people. On the contrary, I'm making the case that it will be very hard to stop someone who is this motivated to indulge their habit.

Prescription Painkillers

There are about 85 million legal prescription pain medicine users in the past year. (I'm inducing this by subtraction; the SAMHSA survey asks about "any use" and "illicit use". Obviously the "any use" number is bigger. I am subtracting the "illicit use" number from it.) There were 13,000 deaths from prescription opioids like Oxycontin and hydrocodone (I'm adding methadone deaths to "other opioid" deaths from CDC death data, avoiding overcounting). This gets us about a 0.015% chance of death, or 1.5 deaths for every 10,000 legal users. The "value-of-a-human-life" calculation comes to roughly $900. That's what I'm paying in terms of "risk of death" by becoming an opioid patient. Have you ever suffered severe pain? That probably sounds like a bargain to avoid even a week of nagging pain, let alone months of severe pain. The $900 figure is likely too high, as I suspect most of the deaths come from particularly incautious people doing incredibly stupid things, such as taking way too much or mixing opioids with alcohol, benzodiazepines, and other medications. Avoiding these behaviors can drop the chance of overdose to essentially zero.


There were roughly half a million heroin users last year. There were 12,345 heroin-related deaths, but that number is probably too low. Sometimes people who are buying "heroin" actually get fentanyl or something stronger. Sometimes the overdose is labeled a "heroin" death on the death certificate and sometimes it is labeled a "synthetic narcotics" death. If I count the total of heroin and synthetic narcotics deaths (avoiding double-counting those in both categories) I get 18,425. Taking these figures literally, a heroin user has a 3.69% chance of death. He is effectively paying $221,000 for his habit by this consideration alone. And once again, you have to add in the actual dollar costs, opportunity cost of time spend acquiring and using, other health issues, etc. Suppose there are actually a lot more heroin users than the SAMHSA survey picks up. After all, a lot of these people are homeless and won't be reachable by a household survey, and a lot of survey-takers might be hesitant to tell someone from the federal government that they used heroin in the past year. Fine, inflate the number of users by a factor of ten. You still get that a heroin user is effectively "spending" $22,100 on his habit, again on the "chance of death" consideration alone. Whatever we're doing to deter heroin users, it has to be on this order of magnitude or they won't even flinch at it. The chance of being arrested or imprisoned will have to be very large. Drug interdiction will have to push the cost of heroin to stratospheric levels (this has been a failure, as heroin is relatively cheap these days). Otherwise the deterrence imposed by drug prohibition will be a drop in the bucket compared to what they are willing to pay.

All of this reinforces a point I've made in previous posts: people have a very high willingness to pay for their drug habits. You have to threaten them with something comparable to their "willingness to pay" in order to get them to stop, otherwise they will simply laugh off your attempt. And "threatening" means occasionally carrying out the threat, otherwise it's no threat at all.

Feel free to quibble with these numbers, as they are back-of-the-envelope. I'll list some of my own reservations. Risk is heterogeneous. (Trust me, I'm an actuary.) You can't just divide the number of drug-related deaths by the number of users to calculate a generic "death risk." Some of these users are very foolish risk-takers and suffer a much higher chance of death-by-overdose. Some of these users are extremely cautious. People who use prescription painkillers as prescribed have nothing to worry about. People who use cocaine once and never develop a habit probably don't need to worry (although there's always the chance of a sudden cardiac arrhythmia). People who get injections of morphine in a hospital setting likewise have no need to worry. (It's a relevant comparison because morphine is chemically similar to heroin, it's just that the fat-solubility of heroin makes it cross the blood-brain barrier more quickly). In other words, most normal, casual users are easier to deter than my numbers above imply, but hard-core problematic drug users are going to be much harder to deter. But it's the problem users who we want to deter, and the casual users who we shouldn't so much worry about. The "risk is heterogeneous" quibble means drug prohibition makes even less sense.

You can come up with other quibbles, for sure. Maybe the $6 million figure is way too high for someone who is destroying their body with drugs. Maybe the SAMHSA figures are biased low because people are unwilling to confess to their drug use on a government survey. (Careful with this critique; this would mean that the risks of drug use calculated above are overstated. But if any drug warrior wants to argue that drug use is a lot less harmful than my calculation implies, I will listen with amusement.) Maybe the death totals collected by the CDC are overstated or understated. Determining a cause of death is a human decision, after all, and subject to a great deal of error. The definitive textbook on drug pathology is just riddled with warnings about how hard it is to actually diagnose a cause of death.

Drug use data from the SAMHSA survey is here (large pdf, ~3,500 pages of charts). You can look up deaths by drug type at the CDC website using their WONDER database. This take a little work, but it's not hard. I am excluding suicides from the death totals.

The Prison Lobby Believes The War on Drugs Is a Failure

The private prison industry* and public sector prison guard unions tend to lobby for stricter drug laws and fight efforts to relax those drug laws. This is strong evidence that drug prohibition does not work, at least in the estimation of the prison lobby. Let me explain.

Supposedly drug prohibition prevents crime by penalizing drug use. Drug users behave irrationally and commit assault and property crimes, the thinking goes. Or they commit various property crimes to support their drug habits. Sure, if you're going to enforce drug laws you're going to have to put some people in prison. You don't have much of a "law" if you're not going to enforce it. But we get less crime overall, because we deter so much drug use. For the price of locking up a few dealers, we deter a lot of drug-fueled zombies from breaking windows, assaulting people, and neglecting their children. That's the rationale for drug prohibition in a nutshell.

The prison lobby doesn't buy it. If they did, they'd be eager to fill their beds with the drug-fueled maniacs that will follow after drug legalization. They would either lobby for or perhaps decline to oppose reforms that relax (repeal?) our drug laws. Maybe I'm wrong for some reason. Maybe the prison lobby doesn't have a strong analytical team and they are simply mistaken about the effect of relaxing drug laws (plausible). Or maybe drug offenders fill up prisons now, with the benefits of drug deterrence coming in the far-off future, such that a net present value calculation makes the short-term gain beneficial to the prison industry (far less plausible). Or maybe the prison lobby is genuinely lobbying in favor of the public interest (even less plausible; see footnote).

If you want to read something empirical on the deterrent effect of drug prohibition, I recommend Jeffrey Miron's Drug War Crimes or Lies, Damned Lies, and Drug War Statistics by Matthew Robinson and Renee Scherlen. Both include time series data showing that drug use doesn't really budge in response to various policy tweaks, such as tightening or softening drug enforcement. If the prison lobby thinks that the effect of drug prohibition on actual drug use is so small as to be irrelevant, they are probably right. My guess is they either implicitly understand this or they did a little research and figured it out.


*Do read the Snopes piece. It appears that we don't know how much the private prison lobby is spending specifically to fight drug law reforms. A quote from an SEC filing was taken as advocacy, when in reality the quote was simply warning its shareholders that relaxing drug laws will hurt revenues. This actually more clearly makes my point than if they were simply lobbying for a law. They could lobby for strict drug laws but claim they privately held the public's best interests at heart. Instead they are saying flat out that relaxing drug laws could reduce their profits by making it harder to fill prison beds.

Here is the SEC filing, an annual report called the 10-K.
Our ability to secure new contracts to develop and manage correctional and detention facilities depends on many factors outside our control. Our growth is generally dependent upon our ability to obtain new contracts to develop and manage new correctional and detention facilities. This possible growth depends on a number of factors we cannot control, including crime rates and sentencing patterns in various jurisdictions and acceptance of privatization. The demand for our facilities and services could be adversely affected by the relaxation of enforcement efforts, leniency in conviction or parole standards and sentencing practices or through the decriminalization of certain activities that are currently proscribed by our criminal laws. For instance, any changes with respect to drugs and controlled substances or illegal immigration could affect the number of persons arrested, convicted, and sentenced, thereby potentially reducing demand for correctional facilities to house them. Legislation has been proposed in numerous jurisdictions that could lower minimum sentences for some non-violent crimes and make more inmates eligible for early release based on good behavior. Also, sentencing alternatives under consideration could put some offenders on probation with electronic monitoring who would otherwise be incarcerated. Similarly, reductions in crime rates or resources dedicated to prevent and enforce crime could lead to reductions in arrests, convictions and sentences requiring incarceration at correctional facilities.
Emphasis mine. And later in the report it says:
Our inmate transportation subsidiary, TransCor, is subject to regulations promulgated by the Departments of Transportation and Justice. TransCor must also comply with the Interstate Transportation of Dangerous Criminals Act of 2000, which covers operational aspects of transporting prisoners, including, but not limited to, background checks and drug testing of employees; employee training; employee hours; staff-to-inmate ratios; prisoner restraints; communication with local law enforcement; and standards to help ensure the safety of prisoners during transport. We are subject to changes in such regulations, which could result in an increase in the cost of our transportation operations.
It is important to note that the report isn't advocating for or against any particular policy, just warning its shareholders about potential threats and costs. Let's not heap scorn upon them for telling it like it is. Still, it's interesting that they see drug testing of employees as a likely net cost. If the logic of drug prohibition is correct, wouldn't removing those drug-addled employees lower their costs? Apparently even cheap, private efforts to reduce the costs of drug use don't pay off. Expensive public efforts to eradicate drug use, even more so.

Things That I Think Are Horrible

It's always interesting to see the reaction to various news events, like the recent Neo-Nazi demonstration and counter-demonstrations in Charlottesville. The response is appropriate. Neo-Nazis really are terrible and they support horrible things. The world would be quite a lot worse off if they got their way. That said, I always feel some unease about the "shouting-down" method of confronting horrible things. The Neo-Nazi demonstrations were meant to show, "We're a big, unruly mob and we won't be reasoned with." The counter-demonstrations were meant to show, "Same here, but our mob is bigger." Of course this works fine until it doesn't. You can't always be so certain that the bigger mob will be the one on the side of the angels. That's why we sometimes  need to step back from our emotions and engage in the marketplace of ideas, even when we don't like what other people are selling.  The "our mob is bigger" method risks shutting down unpopular but good ideas before they even get a fair hearing.

I'd also like to point out that there are things in this country a hell of a lot worse than Neo-Nazis. Sure, the Nazis would be a big problem if they were more numerous. But (let's face it) they're a tiny fringe group that has been banned from polite society. Membership in such a group will lose you most of your friends in most places in America. Horrible ideas that have no chance of gaining a foothold are a lot less scary than horrible ideas that are entrenched in official government policy.

The War on Drugs is worse. The murder rate is substantially higher than it would be if we repealed drug prohibition. Blood-borne pathogens are a huge problem because of the prohibition on the legal sale of clean needles, and the repeal of this prohibition or the introduction of (often illegal-but-tolerated) needle-exchange programs reduces rates of HIV and hepatitis transmission dramatically. Draconian restrictions on prescription opioids prevent pain sufferers from getting the only medicine that works for them. Millions of innocent people are harassed, hundreds of thousands arrested, and tens of thousands jailed or imprisoned on drug charges. Militaristic, no-knock S.W.A.T.-style raids, which are primarily used to serve drug warrants, terrorize 50,000 residences a year, often holding the family at gunpoint, killing the family pet, and occasionally killing an innocent person (or for that matter, killing someone whose "crime" was selling marijuana to a willing buyer). Drug overdose deaths are a lot more common than they would be in a world without prohibition. Many of the heroin overdoses (to take the deadliest drug in recent years) are caused by rapidly fluctuating drug purity or adulteration with fentanyl or something stronger. Basically these overdoses happen because people don't know what they're buying, and they inject something that's maybe ten times stronger than what they thought. Despite all this, it is still respectable to approve of drug prohibition.

Immigration restrictions are worse. Condemning someone to third world poverty when they could otherwise have escaped is unbelievably cruel. Concerns about a "race to the bottom" for wages are grossly exaggerated. Global open borders would likely double world GDP, by our best estimates. And yet our government imposes draconian restrictions on immigration, and it's seen as respectable to approve of these restrictions.

I could go on. There are ideologies that I think are enormously harmful and government policies that implement some of these ideologies, but where it's harder to tell a "human costs" story than in the case of drug prohibition and immigration restrictions. Labor market restrictions. Prohibitions on selling your organs, something that costs us tens of thousands of lives a year. The high implicit marginal tax rate caused by various "anti-poverty" entitlement programs. The nationalization of medicine and schooling. Taxes on capital. Regulation of the market in every single industry (obviously, some are worse than others).

These things are all serious problems. I think people who support these policies are deeply misguided and mostly don't have a leg to stand on. But I would get nowhere by simply shouting curses into their faces and telling them what awful people they are. It wouldn't make much sense for me to get in someone's face and say, "You support X, huh? You're a Nazi!" No, I'd have to talk to these people. I would need to start by showing them some basic respect, like they are human beings or something. Apparently that approach also works on racists.

Sunday, September 3, 2017

Drug Use Isn't the "Cause" of Social Problems

I was going to write a post making a philosophical point about what it means for something to "cause" something else. It is very hard to identify the cause of a human being's behavior, because the cause is buried somewhere deep inside their skull where we can't observe it. In intelligent creatures, behaviors have a lot of internal, unobservable causes. To kick off this philosophical post, I'll start with some gossipy bullshit about people I know on Facebook.

Sometimes I'll post something on Facebook arguing for the end of drug prohibition. Someone I know will occasionally pop up with some bland skepticism, we'll argue for a little while, and he'll cordially suggest we discuss it over a few drinks. I find this maddening. This is someone who puts people in prison for drug charges, and he's effectively saying, "Let's get together for highballs and discuss how the stupid plebs can't handle their shit. I'll bring the cigars."

I'm sure a person in his position could tell me about all the criminals who did stupid shit after using way too many drugs way too often. It would be easy for someone in law enforcement or the healthcare profession to get the impression (wrongly) that the drugs themselves are causing the social problems, rather than the drug use and social problems being symptoms of some other dysfunction. Could poor impulse control cause both, perhaps? Could poor impulse control be exacerbated by drug use, which leads to other problems? Is it possible that drugs are no problem at all for people with normal impulse control? Isn't it likely that people with impulse control problems don't respond to the threat of legal sanctions? And doesn't this render the entire enterprise of drug prohibition moot, given that the deterrent effect is probably small or nil?

I want to recite to my friend a summary of all the social ills caused by excessive alcohol consumption. Tens of thousands of traffic fatalities each year. Comparable numbers of cirrhosis deaths and alcohol poisonings. Sick days. Job loss and unemployment. Domestic abuse. Child neglect. Excited delirium. All because of the alcohol you're holding in your hand right now.

Such an accusation would be hysterical. If someone occasionally indulges with some top-shelf scotch or with a hand-crafted micro brew, that certainly doesn't mean he's to blame for all the social problems cause by excessive drinkers. Those scary statistics about tens of thousands of cirrhosis deaths and traffic fatalities are in no way a function of my moderate drinking, one might legitimately protest. And such a defense would be perfectly fair.

Moderate drinking is not the cause of social problems. Drinking in general is not the cause of social problems in general, or even alcohol-related problems in general. You can't simply aggregate "all drinking" and make it responsible for "traffic fatalities related to drunk driving" and "cirrhosis of the liver". Most drinkers cause no social problems at all. A few cause enormous harm. And so it is with other substances, too. Sometimes it is best to remove temptation. I have known people who deliberately don't drink and don't socialize when others are drinking, because in the past they have gone ape-shit after a few drinks. They know that one drink will become ten, and they will do property damage by the end of the night, if nothing else. I admire people who realize when they need to be tied to the mast. But likewise I know people who can sip on a beverage all night and never have a problem.

Cocaine kills a lot of people (~5,000 in 2014), but there are millions of cocaine users. Most of them are safe about it. Heck, many of them are one-time-only users. But a few of them persist recklessly, day in and day out, and destroy their hearts. It's not that the millions of users are all rolling dice in some game of chance, where death and survival are determined entirely by The Fates. Most users have essentially zero chance of dying from their habit, but a few indulge so stupidly and recklessly that they kill themselves. Suppose you tracked down some Bolivians who chew coca leaf as a low-dose stimulant, as is tradition in that culture. You try to accuse them of killing 5,000 Americans. They would scratch their heads and scowl at you. "You see," you continue, "the chemical that you chew daily is the same one that millions of Americans snort, in super-concentrated powdered form, into their noses. Most of them don't have any problems, but a few of them develop nasty habits and a small fraction of them die from it!" They shrug at you, ignore your idiotic hyperbole, and keep chewing coca into old age.

You can make exactly the same point about other substances. Methamphetamines are basically the same chemicals as drugs that treat ADHD. ICD-10 (cause of death codes found on death certificates and in CDC data files) uses the same code for both, and the annual SAMHSA survey on drug abuse uses the same category for both (see here, page 8 for details). Some people are on ADHD medication 24 hours a day, and some people take very high doses of methamphetamine in a way that instantly crosses the blood-brain barrier (smoking, snorting, injecting). But it would be ludicrous to accuse ADHD patients of causing the skin lesions, poor dental health, and heart problems suffered by chronic methamphetamine abusers. An ADHD patient could rightly protest, "I use my medication as prescribed. The result is that you have an upstanding, well-functioning citizen that you otherwise would have lost to an intractable attention-span problem. What some other idiots do with their bodies is not on me." This would be a perfectly fair defense of an Adderall patient, or for that matter of a moderate methamphetamine user.

You can do this same exercise with any other substance. There are 200 million legal prescriptions of opioid painkillers each year. The death rate per prescription is less than 1 in 10,000. If you were suffering from chronic or acute pain, you would probably say, "I like those odds!" More so if you realize that you can choose willingly whether to take stupid risks, like taking way too much or swallowing pills with alcohol and benzodiazepines (to name two of the biggest risk factors for overdose). The decision to take a substance does not automatically subject you to the base mortality rate for users of that substance. For that you have to do something foolish and risky, and usually you know if you're doing it or not. Statistics that merely divide the number of drug-related problems by the number of total users (such as "X% of drinkers become alcoholics" or "Y% of heroin users become addicts" or "Z% of prescription opioid users eventually overdose on their painkillers") are completely meaningless. They deny any agency to the human beings who suffer from these problems, just as they deny agency to the ones who resist temptation and the ones who indulge moderately.

It's simplistic to generically blame "drugs" or "alcohol" for social problems. Any fool can see there are specific risky behaviors and specific solutions. To blame drugs in general would be to miss the point. Issuing a general prohibition is unlikely to get at the root causes of the problem.