You’re Going to Need a Bigger Plan

The author of the much ballyhooed Mercatus Center study of the costs of “Medicare for All”, Charles Blahous, takes to the pages of the Wall Street Journal to discuss his study. He clears up an important point near the beginning:

While such large amounts of money are difficult to comprehend, my cost estimate is essentially a lower bound. Medicare for All’s actual price tag would likely be even higher. My projection generously assumes the plan would succeed in lowering prescription-drug costs and that administrative costs would somehow be less than half what they are among private insurers.

IMO that the federal government would do something it has been extremely reluctant to do heretofore, using its monopsony power to bargain for lower prices for pharmaceuticals, is unlikely in the extreme. Let’s put it another way: if the federal government were willing to do that, we might not have the problem we presently do.

How much would M4A reduce administrative costs? Why should we assume that its administrative costs would not only be lower than those of private insurance companies but lower than those for Canada’s plan? A JAMA study of some years ago found that Canada’s administrative costs were around half ours. Using the rule of thumb that everything is more expensive here (we pay more to build a foot of bridge or road than anywhere else in the world), I think that two-thirds of present costs could be achievable. That presents a problem for M4A’s advocates who count on bigger savings to defray the costs of their plan.

He then proceeds to the even greater problem:

Most important, it assumes Medicare for All would successfully cut all health-care provider payments down to Medicare’s reimbursement rates, which are more than 40% lower than private insurance rates—and even below providers’ costs of delivering services. Moreover, it assumes that Medicare for All will somehow do all this without disrupting the availability and quality of health care.

Saying that another way, proponents are either saying that everyone in the health care sector should take about a 20% pay cut or Medicare reimbursement rates should fill in the gap which would increase costs even more.

For me here’s the real sticking point:

Part of the cost increase from Medicare for All would naturally come from covering those who are currently uninsured. But the proposed legislation would also expand coverage of specific benefits such as dental, vision and hearing, and greatly increase demand for health services that are already insured, through its stipulation that “no cost-sharing, including deductibles, coinsurance, copayments, or similar charges, be imposed on an individual for any benefits.”

The more of a person’s health care is paid by insurance rather than out of pocket, the more health-care services he tends to buy, regardless of quality and effectiveness. Providing first-dollar coverage for a range of health-care services would therefore be a powerful force driving additional health-care spending. Although Medicare for All proponents believe the administrative efficiencies of single-payer insurance would reduce national health-care costs, my research found the opposite—specifically, that the added costs associated with increased coverage far surpass not only the savings attainable from lower administrative costs, but also the savings potentially gained from swapping brand name drugs for generics.

Some have seized on a scenario in my estimates showing a slight decline in projected total public and private health expenditures under Medicare for All. But that decline, relative to current projections, relies on an assumption that Medicare for All would immediately and dramatically cut provider payment rates by roughly 40%. Without such cuts, Medicare for All would drive national health costs further upward, and the federal price tag would be $38 trillion during its first 10 years.

To the best of my knowledge no other plan in the world other than British National Health imposes such a stipulation. In France cost-sharing is about 10%, in Germany 16% (higher than here), and in Switzerland, the country with per capita health care spending closest to our own, about 30%. In other words we would be breaking completely new ground. Throw any assumptions based on the experience with social insurance in other countries out the window. It should also be noted that the costs of a bureaucracy do not scale linearly but more closely to n log n.