A Flawed Case for Centrally Planned Health Insurance

Writing in the New York Times on October 21, economist Jason Furman recommended 5 books for people to read about the economy before next month’s election. He called them the 5 best books to read.

On one of them, Doug Irwin’s Free Trade Under Fire, I agree. I don’t know enough about 3 of them to have an opinion. But on one, We’ve Got You Covered, I disagree. If you read the book, make sure you read my critique of it.

Here it is, reprinted in full from the Hoover Institution’s Defining Ideas.

In their recent book, We’ve Got You Covered, two health economists make their case for ditching the current system of health insurance in favor of a government-financed, zero-premium insurance for basic coverage. They would allow people to buy supplementary insurance to expand their coverage. The two economists, Stanford’s Liran Einav and MIT’s Amy Finkelstein, are well-known contributors to the literature on health insurance.

With their breezy and humorous writing style, Einav and Finkelstein make what seems at first like a compelling case. They may sway many readers, especially those who don’t know the literature on health economics. But a careful look at their case for ditching our current health insurance and starting over with a centrally planned system uncovers serious omissions and some tensions between their own views. Two omissions are: any mention at all of health savings accounts, and any mention, with a one-sentence exception, of possible reforms of the supply side that would increase supply and reduce the price of health care. One major tension is on their view of the importance of co-payments and deductibles; moreover, they seem to misunderstand the way to measure the impact of co-payments.

The Arguments They Don’t Make 

You might be surprised to learn that the authors’ fundamental argument for a government-financed basic coverage for all is not that in our current system some people go uninsured. They write, “One in ten Americans lack formal health insurance coverage. But they are not uninsured.” How can that be? People without health insurance, they note, “receive a substantial amount of medical care and don’t pay for most of it.” They buttress their claim with strong evidence taken from the Oregon health insurance experiment, which Finkelstein was deeply involved in studying. In 2008, the Oregon government, constrained by its revenue, had decided to run a lottery to choose who would be added to Medicaid rolls and who would not. This is the kind of experiment that causes economists to rub their hands in glee. The reason: we get a large random sample of people who get the coverage and people who don’t, and we can track both their use of the medical system and their health.

One interesting result, the authors note, was that those who applied for the lottery but weren’t accepted received “almost four-fifths of the medical care they would get if they were insured, including primary care, preventive care, prescription drugs, emergency room visits, and hospital admissions.” They paid “only about twenty cents on the dollar for that medical care.” Thus, the authors’ claim that everyone is implicitly insured.

Nor do the authors argue that the US population would be noticeably healthier if everyone had the basic coverage that they prefer. Their evidence on this is threefold. First are the health care results of the Oregon experiment. Although those who won the lottery had depression rates 30 percent lower than those of the losers, there was “no difference in physical health measures like hypertension and high cholesterol.”

Second, they give evidence that “access to medical care is not a primary determinant of health disparities.” They start by retelling the “Tale of Two States,” told in the 1970s by health economist Victor Fuchs. Fuchs compared health outcomes in two states: Utah and Nevada. Utah’s residents ranked very high in health whereas Nevada’s residents ranked very low, with annual death rates that were 40 to 50 percent higher than in Utah. The differences, moreover, could not be attributed to medical care, income, schooling, or climate. The big difference was in personal behavior. Utahans smoked and drank much less than Nevadans. Not coincidentally, Nevadans had much higher death rates from lung cancer and cirrhosis of the liver. Subsequent studies of millions of Americans, the authors note, have found similar results. 

Their third piece of evidence is on the widely noted difference between Medicare spending per covered person in McAllen, Texas, and El Paso, Texas. Those in McAllen had similar characteristics to those in El Paso, but Medicare spending per person was twice the level of El Paso, with those in McAllen getting many more tests, hospital admissions, and even surgeries. But, the authors note, “McAllen residents were not enjoying better health than their counterparts in El Paso.”

The Patchwork and the Social Contract

What, then, is the authors’ argument for government-funded basic coverage? It’s a combination of three factors that interact. The first is the current patchwork provision of special coverages, mainly at the federal government level. The second is that people can lose coverage when they lose their job, which is often when they most need it, or when their income changes (if, for example, they are on Medicaid). The third is what they call the “social contract.” In laying out the patchwork, they go through the history of how the feds added coverages to respond to various emergencies of the day. A particular person’s hardship from lack of coverage would be publicized and various government officials, including President Reagan, would call for special coverage to handle such future situations. Thus, the patchwork.

Einav and Finkelstein argue that this government provision is evidence of our “social contract.” So why not, they ask, fulfill the social contract by having government provide basic coverage for everyone? They argue that for those who lack insurance, universal coverage would “provide financial protection against medical expenses, less anxiety, and greater peace of mind.” And for those who have formal insurance, universal coverage “would remove the omnipresent risk that their current coverage may disappear right when they need it most.”

Clearly, there is not literally a social contract. The closest I came to agreeing to something like a social contract was when I took a pledge upon becoming an American citizen. But that was about my obligations, not the US government’s or taxpayers’ obligations to me. So whence the social contract? They try to justify their claim by noting that in 1944 FDR asserted that health care is a basic human right. That’s a pretty thin basis for their belief. At least, though, the authors admit that the contract is “unwritten.”

One other reason the authors add to make their case for basic coverage is the fact that $40 billion per year is spent on health care for the uninsured but not by the uninsured. About 80 percent of that is government funded. They say that this is big potatoes. Is it? In recent years, health care spending in the United States has exceeded $4 trillion annually. So $40 billion is only about 1 percent of health care spending. Does it really make sense to have the $40 billion tail wag the $4 trillion dog?

The Authors’ Lack of Coverage

It’s slightly ironic that in a book titled We’ve Got You Covered, there are important issues in health economics that the authors don’t cover at all, issues that matter for an overall evaluation of the health care system.

For example, one increasingly popular innovation in health care purchase in the past few years is health savings accounts (HSAs). With HSAs, people save before-tax money the way they do with an Individual Retirement Account and can then spend it tax-free on health insurance, co-payments, and deductibles. Many people have used these to cover expenses that otherwise would have been difficult to cover. Would HSAs solve all of the problems that Einav and Finkelstein point to? Of course not. But they would help. It was disappointing, therefore, that the authors didn’t even mention HSAs.

Also, after reading every page, including every one of the authors’ extensive (and impressive) footnotes, I could find no mention, other than one sentence, of government restrictions that make the supply of health care less than otherwise and prices higher than otherwise. Three such are certificate of need (CON) laws at the state level, federal restrictions on immigration of doctors, and restrictions that prevent non-doctors from doing things that are now done only by doctors.

CON laws are laws that require those who want to build a hospital or a surgical facility, to name only two, to get permission from a state government agency. Existing providers often show up to contest the permission and often succeed. One of the most extreme CON laws is in Illinois. In his 2008 book, Code Red, Northwestern University health economist David Dranove points out how badly Illinois CON laws have hampered the building of hospitals. As a result, he notes, Illinois hospitals are located “where Illinoisians lived in the 1950s.” The CON restrictions, by restraining supply, also limit competition and elevate prices.

Similarly, immigration restrictions prevent tens of thousands of doctors from moving to the United States, making doctors’ fees higher than otherwise.

Also, pharmacists, who usually know much more about drugs than doctors do, are typically prevented from prescribing drugs. If they were able to do so legally, as they are in many countries, people could often skip an expensive trip to the doctor, saving time and money. Think about someone who trains for years to be a pharmacist. If he doesn’t work in a compounding pharmacy, what is his main job, besides warning customers about side effects? Counting. That’s a huge waste of resources.

Fortunately, the state governments of Idaho, Colorado, and Montana are taking steps toward allowing pharmacists to prescribe medicine.

Strikingly, though, because they are so focused on insurance, Einav and Finkelstein cover none of this.

Co-Pays and Deductibles

As recently as the 1970s, we didn’t have good empirical evidence showing that requiring  patients to pay a co-pay would influence the amount of medical care that people demanded. But Einav and Finkelstein point out that we now have scads of such evidence. One might think, therefore, that they would advocate having at least modest co-pays as part of their basic coverage. That would cause people to spend less money, presumably cutting out the marginal uses. Surprisingly, though, the authors oppose co-pays.

Why? They argue that when other governments have introduced co-payments into their systems, they find themselves backing down and eliminating co-payments or having a sometimes-long list of exemptions from co-payments. Also, they argue, in countries that have modest co-payments, the restraint on spending is modest. Yet in their later discussion of Medicare, they note that because Medicare will pay for whatever doctors and hospitals provide, “why not run a few extra tests and scans to be on the safe side, or send the patient to a specialist when in doubt?” It seems to this health economist (for two years I was the senior economist for health policy with President Reagan’s Council of Economic Advisers) that co-pays could restrain Medicare spending a fair amount.

That might be a judgment call, but what isn’t a judgment call is the case that health economists make for co-pays. It is not that the insurer or, with government programs, the government, will collect a lot of money from modest co-payments. Rather, it’s that those co-payments could result in substantial savings from the extra tests and scans and the specialist visits that otherwise would occur. The authors point to one UK report that found that the cost of collecting co-pays “could outweigh the revenue collected from them.” But this is a bogus way of judging. Even if that claim were true, it leaves out the major source of cost saving: the extra medical care not provided.

The “Basic Coverage” Empty Shell

In a chapter titled “A Shack, Not a Chateau,” the authors advocate “very basic” coverage. That way, government spending wouldn’t be as large as otherwise. Later they state that basic coverage “should include primary and preventive care, specialist, outpatient, emergency room, and hospital care.” The hard part, they write, is in choosing what more to cover. To avoid tackling the issue directly, they advocate an overall health care budget to pay for the basic coverage and call for making tradeoffs within that budget.

I wonder whether the authors worry about whether the same kind of political forces that ended co-pays in other countries or that added patchwork coverages in this country would lobby successfully to have various coverages included under “basic,” with the effect being higher spending, possibly even higher than it would have been. They don’t seem concerned.

The Road Not Taken

At the end of the book, the authors tell the story of Stan Brock, who founded Remote Area Medical to provide free health care to underserved areas in the United States. Then they tell of Tommy Douglas, who, as premier of Saskatchewan in 1947, introduced government-funded hospital insurance. (I saw him speak at a New Democratic Party convention in Winnipeg in 1969, by the way.) They clearly prefer the Tommy Douglas option and must have their reasons. But it would have been nice to know what those reasons are.

What if we combined three things: (1) the reforms I suggest above that would reduce health care prices, (2) health savings accounts, and (3) an expanded role for private voluntary charity? It would be interesting to see how that would compare.