[U]nlike in other countries, sellers of health-care services in America have considerable power to set prices, and so they set them quite high. Two of the five most profitable industries in the United States — the pharmaceuticals industry and the medical device industry — sell health care. With margins of almost 20 percent, they beat out even the financial sector for sheer profitability. The players sitting across the table from them — the health insurers — are not so profitable. In 2009, their profit margins were a mere 2.2 percent. That’s a signal that the sellers have the upper hand over the buyers.
I don’t agree that insurers are being bullied as buyers. If we’re going to bring up the financial sector, a better analogy would compare pay differentials between revenue-generating traders (providers) and the back office clerical and IT workers (insurers), rather than assume some common baseline of industrial profitability. The health care providers actually (try to) improve health; the insurers (are supposed to) support that primary effort. But overall, the story Klein tells here is broadly consistent with many other explanations of high prices in US health care.
What will solve that problem? Probably not health care reform, though regulators will struggle mightily to impose some discipline via IPAB and other entities. Followers of Clayton Christensen think pure technological innovation may wildly succeed where an oft-captured regulatory system is failing. Farhad Manjoo provides some empirical support for their hopes:
As computers get better, we’ll need fewer humans across a range of specialties. Look at mammography: One of the main ways radiologists can improve their breast diagnoses is by “double reading.” When two radiologists independently examine a collection of mammograms, the number of cancers detected increases substantially. A study published in 2008, however, found that a radiologist who uses ImageChecker can skip the second reading: A computer and a human are just as good as two humans.
[T]he doctors who are the juiciest targets for automation might not be the ones you’d expect. They’re specialists . . ., the most highly trained, highly paid people in medicine. It’s precisely these factors that make them vulnerable to machines. By definition, specialists focus on narrow slices of medicine. They spend their days worrying over a single region of the body, and the most specialized doctors will dedicate themselves to just one or two types of procedures. Robots, too, are great specialists. They excel at doing one thing repeatedly, and when they focus, they can achieve near perfection. At some point—and probably faster than we expect—they won’t need any human supervision at all.
Robots and automation are already taking on prominent roles in wars, factories, and political campaigns. The type of pattern recognition common to some medial specialties may be natural to them, particularly as electronic medical records and digitization take hold. Of course, an all purpose “physician robot” would be a much harder endeavor. In the context of a discussion of rationing, one health law textbook suggests that a mapping of possible interventions “would require rigorous scientific information on each of the almost 10,000 diagnostic entries in the International Classification of Diseases (9th ed.) (known as ‘‘ICD-9’’) and for each of the 10,000 medical interventions listed in the AMA’s Common Procedural Terminology (known as ‘‘CPT’’ codes).” ICD-10 has about 7 times more codes than ICD-9. But just as chess was once considered a field impenetrable to artificial intelligence, and now has been mastered by some computers, so too might medicine itself become subject to the exponential growth in information processing characteristic of mature digitized industries. It’s becoming clear that “the variety of jobs that computers can do is multiplying as programmers teach them to deal with tone and linguistic ambiguity.”
So will technology save us from ever-increasing health care costs? I’m not optimistic, because politics and economics are a constraint on all these developments. The same patterns of patronage and tribute that make comparative effectiveness research such a hard sell in the US may well restrain technology adoption. Just as specialists dominate the RUC, they can probably find ways to slow the adoption of technological substitutes for their hard-won expertise. As Umair Haque has observed, “In a neofeudal polity, patronage replaces meritocracy. ‘Success’ for an organization, coalition, or person is to become a client of a powerful patron, pledging your services (soft and hard, informal and formal), in perpetual alignment with the patron’s interests.” We’ll see many physicians in coming years invest time and effort in technological innovation, and others devoted to deterring its spread in order to protect current income streams.
At this point, you’re probably expecting me to side decisively with the technologists as heroes. But I can’t do so. I don’t buy an economic model premised on incentivizing innovation by setting off a race among radiologists (or, more realistically, financiers) to be the first to patent the machine that can replace all the other radiologists. Rather, I think the real foundation for radically productive innovation in this and other fields is a baseline of social support and commitment to retraining for professionals who could be displaced by the technology. I’m not saying, “pay radiologists what they make now, forever.” Rather, I’m trying to articulate a variant of a “guaranteed basic income” argument for those who invest heavily in learning about science, technology, and medicine. This baseline of educated users, improvers, and evangelizers of technology is the foundation of any venturesome economy. As Amar Bhide has explained,
[T]he different forms of innovation interact in complicated ways, and it is these interconnected, multilevel advances that create economic value. . . . To state the proposition in the terminology of cyberspace, innovations that sustain modern prosperity have a variety of forms and are developed and used through a massively multiplayer, multilevel, and multiperiod game.
We may well find that in decades to come, machines can do the jobs of radiologists and pathologists much better than people can. But if that transition occurs, it’s important to recognize how much current specialists invested to attain their skills, how hard they presently work to maintain a high level of medical skill in this country, and how future innovations may well dry up if people feel that those on STEM career paths are utterly vulnerable to being “kicked to the curb” once a machine does their job slightly better. Not only is “sole inventorship” a myth; we often fail to appreciate the complex educational and service apparatus necessary for innovation to take place. As Alperovitz and Daly have shown, any system that grants 93% of its gains to 1% of the people is an ongoing instruction in the economic futility of the efforts of the vast majority of its citizens.