(QuasLacrimas is not back from hiatus. You’ll notice this post is written as though addressed to RCAFDM, and implicitly alludes to his previous work rather than recapitulating it for my reader.When I’m not using twitter I tend to write unjustifiably long blog comments. When they get disastrously long, I spare the OP my tl;dr and put it on my own blog instead. When I’m really going off – two posts.)
I. This was a great essay, going beyond settling the relatively simple (!?) error Noah Smith made to summarize your previous conclusions and hammer in some of the bespoke distinctions. It also brought into sharp relief the triangular relation between [explaining health costs] x [explaining illusions about health costs] x [explaining the sentiment that rising health care costs are onerous].
> …is it really surprising we’d increasingly direct our resources to higher-order desires…?
This sentence in particular crystallized why I agree 100% with your model of healthcare costs (i.e., US costs dictated by same formula that determines other OECD costs, not an exception to them) but am still not sanguine about health care expenditures. Because: who’s the “we”, kemosabe? If the US had a ultralibertarian system where you paid à la carte in gold coins, then obviously “we” would be choosing to purchase more and better medical procedures. If the US had single-payer, we/our representatives would be choosing to allocate more medical care to citizens and budgeting the requisite funds for it.
But what we have is a mess and I don’t think anyone feels that they are choosing to “purchase more health”. If I understand correctly, you’re guessing that the whole phenomenon of rising healthcare expenditures is sort of like “people choose en masse to buy Lasik instead of glasses” — and if most health spending were out of pocket, this would be the most reasonable interpretation, and I would be perfectly okay with it. But it seems to me that the standard-of-care locked in by insurance contracts, medicare/-caid regulations, tort precedents, and first-responder obligation-to-treat has come to cover increasing amounts of tests, diagnoses and interventions as procedures are routinized and medical tech advances. So you get employees struggling to pay for insurance plans whose coverage they consider “the same”, you get Congress struggling to pay for “the same” coverage for old people…
(Voluntarily uninsured people wanting to buy, or facing a mandate to buy insurance is a variation on the same theme. It’s not like buying your car where you’re thinking about all the places you could go if you could drive, they just want the same access to the healthcare system they previously had, but the organizing logic of standard-of-care makes it increasingly unfeasible to let healthy people opt out of the buffet menu.)
This could be the simplest way of putting the intuition I’m sketching out: by definition US healthcare expenditures are rising because healthcare costs are rising net of healthcare prices (close enough?). I’m also willing to concede (for the sake of argument, and because it’s plausible) US doctors are performing more procedures: more billable activities are taking place. But can we can say what the patients are buying more of? More years of life? Superior physical condition? Delivery from particular diseases? The procedures themselves aren’t consumption per se, that’s not the consumer’s goal, procedures can’t plausibly be the object of a higher desire (MRI tourism??) and can’t by themselves explain why healthcare would be a strongly superior good. And if Americans are buying the same health outcomes but paying more for them, that suggests to me either the price levels or productivity growth in healthcare has been mismeasured. (I’d guess if it now takes three procedures to get the same result as 20 years ago, that’s a negative productivity shock to the medical sector, but no doubt there are technicalities I don’t understand.)
II. Tacitly of course I’m assuming the finding that all the extra spending in the richest OECD countries doesn’t improve health outcomes! Quite likely there is evidence pointing the other way as well; perhaps the truly massive scale of the white mortality uptick has been masked by a simultaneous improvement in mortality from all other causes besides deaths of despair? (Big if true. — Or maybe we should model higher-order medical purchases as more like bungee-jumping, and the smack high is the kind of thing we’re buying and the mortality is ¯\_(ツ)_/¯ ) Conceivably you have a whole list of budget items like pills and SRS and …? that, taken together, paint a picture of these finicky higher-order desires. (Such a list wouldn’t strike me as unthinkable now that I’m aware HIV, a single virus!, can account for 8% of consumer Rx spending by itself.)
But so far as I know there is no clear positive case that shoveling money into the medical furnace empirically reduces mortality and morbidity, which agrees with anecdata and a scattering of suggestive facts like [% spending in last N weeks of life] and [marginal QALY-benefit of some screenings at or below discomfort of procedure]. I was one of the idiots who thought (back in 2010) the ACA was going to put a barefooot doctor on every ghetto block to hand out preventative medicine that was going to like, save money on… okay it’s embarrassing even to think about how it was supposed to work, but that’s another example of a discrete expansion of consumption for the sake of a certain good (disease prevention) that never actually materialized. I am also under the impression that administrative share of health costs are rising (but I vaguely recall you poo-pooing that last year). These suggest to me that hospitals are selling a credibility-good, and credibility-goods are inherently prone to principle-agent problems and signaling spirals as incomes rise, so the level of spending is explained by US incomes but *not* by consumer’s desire to purchase better healthcare outcomes.
The credibility-good thesis restated with respect to its policy implications: it is possible US healthcare costs are increased by structural problems in the US health sector, yet the difference between US and non-US healthcare costs is not caused by structural features present in the US system and absent in non-US systems, but rather by higher US incomes triggering structural features present in US and non-US systems that predispose the sector to cost increases (or, if you prefer: to negative productivity shocks).
III. By the way: the most straightforward, directly plausible, and convenient-to-model way to understand the claim that healthcare — or for that matter education or journalism — sell credibility-goods is to say these sectors involve fiduciary relationships stemming from information asymmetries: you pay the doctor because he knows whether you’re sick and how dangerous it is and you don’t, but by the same token when the doctor declares a certain course of treatment necessary and cost-effective for the price, you as patient have a limited ability to evaluate the truth of the matter.
(If you knew how to tell what diseases you have and how effective the alternative treatments are, you wouldn’t need to pay the doctor for his expertise to begin with. Ceteris paribus for ignorance and education, at least when the students make their own enrollment decisions.)
In a market in credibility-goods —— if patients trust their doctors unconditionally, prices rise; if each diagnosis requires seeking a second and third opinion to authenticate the authority of the first diagnosis, costs rise; if the doctors need to send costly signals of credibility, costs rise; if they need to extensively document and justify each diagnosis, productivity falls; if civil courts punish experts’ misdiagnoses and failures as breaches of fiduciary duty, the experts start to insist their patients pay for diagnostics and treatments previously classed “optional and slightly excessive”. When people are poor, the fact that they cannot afford even the most reliably effective treatments helps to impose price discipline on the whole system of credibility-goods, but the more expertise they can afford to buy, the more scope there is for the principal-agent problem to infect market dynamics.
I mention this to foreshadow my next post, which will consider an alternative (and rather more speculative) interpretation of credibility-goods, with different implications for why they’re getting expensive.