In the last few days, television audiences in the USA have been hearing a great deal about IBM’s “Watson” computer system. The occasion of this publicity is Watson’s appearance as a contestant on the popular quiz show Jeopardy. IBM has emphasized Watson’s potential in the medical field:
Throughout this material, IBM’s spokespeople keep inviting us to imagine a near future in which Watson or systems like it will be found “in every doctor’s office.” What this phrasing suggests to me is a situation in which there are about as many doctors as there are now, those doctors are distributed in offices as they are now, and in those offices they examine patients who come to them as they do now. The state of affairs that this phrasing suggests is that these future offices will differ from their present-day counterparts in that Watson-like natural language processors will be installed to provide the patient with an “instant second opinion.”
A moment’s reflection will reveal that there is essentially no likelihood of such a scenario being realized, at least not in the USA. As soon as a machine is invented that is capable of giving a medical opinion that is of any value whatsoever, flesh-and-blood doctors will vanish from the lives of low-income patients forever. Once the machine is so improved that it can be trusted to give a sound diagnosis most of the time, with none but the trickiest cases requiring review by human doctors and none but a small percentage of those requiring active intervention to overrule the machine, the only patients who will ever meet their doctors will be the very wealthy and the scientifically interesting.
The parallel I would draw is with the institution of the “house call.” As recently as 40 years ago, it was so common for doctors to call on their patients at home that when people occasionally had to go to the doctor’s office to receive care, it was considered grounds for a radical overhaul of the healthcare system. Now, when a doctor does make house calls, it’s national news. I predict that 40 years from now, it will be as rare for a patient to visit a doctor for examination as it is today for a doctor to visit a patient at home.
What will the consequences of this change be for public policy? The central dilemma in technology policy is always the same, that there is little or no interval between the time when it is too soon to say what the effects of a development will be and the time when it is too late to do anything about that development. One thing we can say is that demand for medical doctors will drop dramatically, probably to 1% or less of the current per capita demand by 2050. Whether that means we will have only 1% as many doctors then as we do now, or that some larger number will share 1% of the income that doctors now collect, of course depends on a wide range of factors. Whichever way it goes, certainly no prudent investor would be interested in funding a new medical school at this time.
The cost of health care is a focus of much discussion in the USA, where it represents at least 1/7 of GDP. Eliminating doctors would change the way this spending breaks down, but would neither reduce demand for health care nor increase its supply. Moreover, many have argued that the reason health care costs so much more in the USA than in similar countries is that Americans do not really have a market for health care. Rather, employers pay for health insurance in order to avoid paying the corporate income tax. Since employers pay insurance companies money that would otherwise go to the taxman, they have little incentive to negotiate lower premiums; since insurers raise premiums when providers charge them more, they have no incentive at all to negotiate for lower prices. As long as the corporate income tax and its health-insurance deduction remain in place, US health care costs will continue to rise no matter how little money goes to doctors. Perhaps if the USA were to abolish the corporate income tax and replace it with a consumer-driven revenue source like the Value Added Tax, a consumer-driven health care system might emerge, but until then, technology cannot solve our problems.
Of course, unemployment is also a public policy problem. What happens to all the M.D.s whose degrees will become worthless in the years ahead? And what happens to public opinion when the appearance of a horde of jobless doctors makes it clear that education is no guarantee of employment? Marxism may be dead, but will it stay buried in a world where the owners of capital are the only economic group who lead lives secure enough to plan for their futures?