Ezra Klein on the health of nations

In the American Prospect, Ezra Klein compares five countries’ health care systems (hat tip to Mark Thoma). in spite of his general defence of Canada, there are a few areas where I think he gets it wrong and I have added in some comments in those places.

The Health of Nations
Here’s how Canada, France, Britain, Germany, and our own Veterans Health Administration manage to cover everybody at less cost and with better care than we do.

By Ezra Klein
Issue Date: 05.08.07

Medicine may be hard, but health insurance is simple. The rest of the world’s industrialized nations have already figured it out, and done so without leaving 45 million of their countrymen uninsured and 16 million or so underinsured, and without letting costs spiral into the stratosphere and severely threaten their national economies.

Even better, these successes are not secret, and the mechanisms not unknown. Ask health researchers what should be done, and they will sigh and suggest something akin to what France or Germany does. Ask them what they think can be done, and their desperation to evade the opposition of the insurance industry and the pharmaceutical industry and conservatives and manufacturers and all the rest will leave them stammering out buzzwords and workarounds, regional purchasing alliances and health savings accounts. The subject’s famed complexity is a function of the forces protecting the status quo, not the issue itself.

So let us, in these pages, shut out the political world for a moment, cease worrying about what Aetna, Pfizer, and Grover Norquist will say or do, and ask, simply: What should be done? To help answer that question, we will examine the best health- care systems in the world: those of Canada, France, Great Britain, Germany, and the U.S. Veterans Health Administration (VHA), whose inclusion I’ll justify shortly.

Putting aside the VHA, America’s annual per person health expenditures are about twice what anyone else spends. That actually understates the difference, as our 45 million uninsured citizens have radically restricted access to care, and so the spending on the median insured American is actually quite a bit higher. Canada, France, Great Britain, and Germany all cover their entire populations, and they do so for far less money than we spend. Indeed, Canada, whose system is the most costly of the group, spends only 52 percent per capita what we do.

While comparing outcomes is difficult because of various lifestyle and demographic differences in the populations served, none of the systems mentioned betray any detectable disadvantage in outcomes when compared with the United States, and a strong case can be made that they in fact perform better. Here, however, I largely restrict myself to comparisons of efficiency and equity. With that said, off we go.

Oh, Canada!

As described by the American press, Canada’s health-care system takes the form of one long queue. The line begins on the westernmost edge of Vancouver, stretches all the way to Ottawa, and the overflow are encouraged to wait in Port Huron, Michigan, while sneering at the boorish habits of Americans. Nobody gets to sit.

Sadly for those invested in this odd knock against the Canadian system, the wait times are largely hype. A 2003 study found that the median wait time for elective surgeries in Canada was a little more than four weeks, while diagnostic tests took about three (with no wait times to speak of for emergency surgeries). By contrast, Organisation for Economic Co-operation and Development data from 2001 found that 32 percent of American patients waited more than a month for elective surgery, and 5 percent waited more than four months. That, of course, doesn’t count the millions of Americans who never seek surgery, or even the basic care necessary for a diagnosis, because they lack health coverage. If you can’t see a doctor in the first place, you never have to wait for treatment.

Canada’s is a single-payer, rather than a socialized, system. That means the government is the primary purchaser of services, but the providers themselves are private. (In a socialized system, the physicians, nurses, and so forth are employed by the government.) The virtue of both the single-payer and the socialized systems, as compared with a largely private system, is that the government can wield its market share to bargain down prices — which, in all of our model systems, including the VHA, it does.

Given physician incomes, which are much higher than in Europe, it is hard to argue that this constitutes bargaining down. This is because, by being next door to the US, we have to pay compensation that is reasonably equivalent. Cost control through a single payer can have some downsides, too – a government can come to power and underfund the system, and Canada has less spanky, techno-gadgets than in the US (this may or may not be a good thing, and anyway, access to gadgets in the US is limited to those with gold-plated coverage).

A particularly high-profile example of how this works is Canadian drug reimportation. The drugs being bought in Canada and smuggled over the border by hordes of lawbreaking American seniors are the very same pharmaceuticals, made in the very same factories, that we buy domestically. The Canadian provinces, however, bargain down the prices (Medicare is barred from doing the same) until we pay 60 percent more than they do.

I don’t think this latter point on drugs is correct. I do not think there is a coherent bulk purchasing scheme in place in any province. Canada does have the Patented Medicine Prices Review board, which caps the price of new pharmaceutical drugs, and we have a legacy of generic production (despite the former practice of compulsary licensing having been kneecapped by the Mulroney government as a chip to get the Canada-US Free Trade Agreement). Drugs are the fastest growing part of the health care system, and there has been a lack of political will to take on Big Pharma in a big way.

Single-payer systems are also better at holding down administrative costs. A 2003 study in The New England Journal of Medicine found that the United States spends 345 percent more per capita on health administration than our neighbors up north. This is largely because the Canadian system doesn’t have to employ insurance salespeople, or billing specialists in every doctor’s office, or underwriters. Physicians don’t have to negotiate different prices with dozens of insurance plans or fight with insurers for payment. Instead, they simply bill the government and are reimbursed.

Fee-for-service payment to doctors is itself problematic (with a few exceptions). It would be better to have physicians on salary, as they are in the small number of community health centres (which are a much better model of primary care, involving nurses and other professionals), or on some sort of capitation model, where they are paid for caring for a specific population (and thus have an incentive for preventive measures).

The downside of a single-payer system in the Canadian style is that it constructs a system with a high floor and a low ceiling. If you don’t like the government’s care options, there’s no real alternative. In this, Canada is rare. As we’ll see with both France and Germany, other countries are able to preserve a largely nationalized system with universal access while allowing private options at the upper levels.

There are alternatives. If doctors do not practice as part of the public system, they can see patients privately. It’s just that when public health insurance was introduced, after some initial opposition doctors realized that this was a good deal – they got paid, which was not always the case before. Canadians can also go to the US to get private treatment. But these are expensive options available only to the few.


It’s a common lament among health-policy wonks that the world’s best health-care system resides in a country Americans are particularly loath to learn from. Yet France’s system is hard to beat. Where Canada’s system has a high floor and a low ceiling, France’s has a high floor and no ceiling. The government provides basic insurance for all citizens, albeit with relatively robust co-pays, and then encourages the population to also purchase supplementary insurance — which 86 percent do, most of them through employers, with the poor being subsidized by the state. This allows for as high a level of care as an individual is willing to pay for, and may help explain why waiting lines are nearly unknown in France.

Canada also has supplementary health insurance, typically provided by employers, to cover eye exams, supplemental therapies (massage, physio, etc), drugs, dental, and out-of-country care. This part (about 30% of the total) most resembles the US private-insurance-through-employer-benefits system.

France’s system is further prized for its high level of choice and responsiveness — attributes that led the World Health Organization to rank it the finest in the world (America’s system came in at No. 37, between Costa Rica and Slovenia). The French can see any doctor or specialist they want, at any time they want, as many times as they want, no referrals or permissions needed. The French hospital system is similarly open. About 65 percent of the nation’s hospital beds are public, but individuals can seek care at any hospital they want, public or private, and receive the same reimbursement rate no matter its status. Given all this, the French utilize more care than Americans do, averaging six physician visits a year to our 2.8, and they spend more time in the hospital as well. Yet they still manage to spend half per capita than we do, largely due to lower prices and a focus on preventive care.

That focus is abetted by the French system’s innovative response to one of the trickier problems bedeviling health-policy experts: an economic concept called “moral hazard.” Moral hazard describes people’s tendency to overuse goods or services that offer more marginal benefit without a proportionate marginal cost. Translated into English, you eat more at a buffet because the refills are free, and you use more health care because insurers generally make you pay up front in premiums, rather than at the point of care. The obvious solution is to shift more of the cost away from premiums and into co-pays or deductibles, thus increasing the sensitivity of consumers to the real cost of each unit of care they purchase.

This has been the preferred solution of the right, which has argued for a move toward high-deductible care, in which individuals bear more financial risk and vulnerability. As the thinking goes, this increased exposure to the economic consequences of purchasing care will create savvier health-care consumers, and individuals will use less unnecessary care and demand better prices for what they do use.

Problem is, studies show that individuals are pretty bad at distinguishing necessary care from unnecessary care, and so they tend to cut down on mundane-but-important things like hypertension medicine, which leads to far costlier complications. Moreover, many health problems don’t lend themselves to bargain shopping. It’s a little tricky to try to negotiate prices from an ambulance gurney.

Health is also a public good. We do not want people to stay at home due to the out-of-pocket cost of care when they really should see a doctor. The costs of dealing with people who just need to be turned back home are not that high relative to the costs that would occur if they sick stayed home and got worse, at which point the treatment is likely to get more expensive and the health outcomes worse.

A wiser approach is to seek to separate cost-effective care from unproven treatments, and align the financial incentives to encourage the former and discourage the latter. The French have addressed this by creating what amounts to a tiered system for treatment reimbursement. As Jonathan Cohn explains in his new book, Sick:

In order to prevent cost sharing from penalizing people with serious medical problems — the way Health Savings Accounts threaten to do — the [French] government limits every individual’s out-of-pocket expenses. In addition, the government has identified thirty chronic conditions, such as diabetes and hypertension, for which there is usually no cost sharing, in order to make sure people don’t skimp on preventive care that might head off future complications.

The French do the same for pharmaceuticals, which are grouped into one of three classes and reimbursed at 35 percent, 65 percent, or 100 percent of cost, depending on whether data show their use to be cost effective. It’s a wise straddle of a tricky problem, and one that other nations would do well to emulate.

Great Britain

I include Great Britain not because its health system is very good but because its health system is very cheap. Per capita spending in Great Britain hovers around 40 percent what it is in the United States, and outcomes aren’t noticeably worse. The absolute disparity between what we pay and what they get illuminates a troublesome finding in the health-care literature: Much of the health care we receive appears to do very little good, but we don’t yet know how to separate the wheat from the chaff. Purchasing less of it, however, doesn’t appear to do much damage.

What’s interesting is that many of the trade-offs that our health-care system downplays, the English system emphasizes. Where our medical culture encourages near-infinite amounts of care, theirs subtly dissuades lavish health spending, preferring to direct finite funds to other priorities.

This sort of national prioritizing is made easier because Great Britain has a socialized system, wherein the government directly employs most of the providers. Great Britain contains costs in part by paying doctors through capitation, which gives doctors a flat monthly sum for every patient in their practice. Since most patients don’t need care in a given month, the payments for the healthy subsidize the needs of the sick. Crucially, though, the fixed pool of monthly money means doctors make more for offering less treatment. With traditional fee-for-service arrangements, like ours, doctors gain by treating more. The British system, by contrast, lowers total costs by lowering the quantity of prescribed care. As University of San Francisco professors Thomas Bodenheimer and Kevin Grumbach write, “British physicians simply do less of nearly everything — perform fewer surgeries, prescribe fewer medications, and order fewer x-rays.”

That may sound strange, but it also means that society pays for fewer of those surgeries, fewer of those medications, and fewer of those X-rays — and as far as we can tell, the English aren’t suffering for it. Indeed, a 2006 study published in The Journal of the American Medical Association found that, on average, English people are much healthier than Americans are; they suffer from lower rates of diabetes, hypertension, heart disease, heart attack, stroke, lung disease, and cancer. According to the study’s press release, the differences are vast enough that “those in the top education and income level in the U.S. had similar rates of diabetes and heart disease as those in the bottom education and income level in Great Britain.”

Great Britain’s example proves that it is possible to make economy a guiding virtue of a health system. We could do that on the supply side, through policies like capitation that would change the incentives for doctors, or on the demand side, by making patients pay more up front — or both, or neither. Americans may not want that system, in the same way that the owner of a Range Rover may not want a Corolla, but we should at least recognize that we have chosen to make health care a costly priority, and were we to decide to prioritize differently, we could.


The German system offers a possible model for those who want to retain the insurance industry but end its ability to profit by pricing out the sick and shifting financial risk onto individuals. The German system’s insurers are 300 or so different “sickness funds” that act both as both payers and purchasers for their members’ care. Originally, each fund covered only a particular region, profession, or company, but now each one has open enrollment. All, however, are heavily regulated, not for profit, and neither fully private nor publicly owned. The funds can’t charge different prices based on age or health status, and they must continue covering members even when the members lose the job or status that got them into the fund in the first place. The equivalent would be if you could retain membership in your company’s health-care plan after leaving the company.

The move toward open enrollment was an admission that interfund competition could have some positive effects. The fear, however, was that the funds would begin competing for the healthiest enrollees and maneuvering to avoid the sickest, creating the sort of adverse selection problems that bedevil American insurance. To avoid such a spiral, the government has instituted exactly the opposite sort of risk profiling that we have in the United States. Rather than identifying the unhealthy to charge them higher rates, as our insurers do, the government compels sickness funds with particularly healthy applicants to pay into a central fund; the government then redistributes those dollars to the funds with less-healthy enrollees. In other words, the government pays higher rates to sickness funds with unhealthy enrollees in order to level the playing field and make the funds compete on grounds of price and efficiency. In this way, the incentive to dump the sick and capture the well is completely erased. The burdens of bad luck and ill health are spread across the populace, rather than remaining confined to unlucky individuals.

The system works well enough that even though Germans are allowed to opt-out of the sickness funds, they largely don’t. Those with incomes of more than $60,000 a year are not required to join a sickness fund; about 10 percent of these citizens purchase private insurance and .02 percent choose to eschew coverage entirely. The retention of a private insurance option ensures that Germans have an escape hatch if the sickness funds cease providing responsive and comprehensive coverage; it also clears a channel for experimentation and the rapid introduction of new technologies. And the mix of private-public competition works to spur innovation: By 2005, Germany had spent $21.20 per capita wiring its system with health-information technology; America, meanwhile, had spent a mere 43 cents per capita, and most U.S. hospitals still have no systems to speak of.

What the German system has managed to achieve is competition without cruelty, deploying market forces without unleashing capitalism’s natural capriciousness. They have not brought the provision of health care completely under the government’s control, but neither have they allowed the private market, with its attendant and natural focus on profits, to have its way with their health system. It’s a balance the United States has been unable to strike.

The Veterans Health Administration

The mistreatment and poor conditions at the Walter Reed Army Medical Center were a front-page story recently, and they were rather conclusive in showing the system’s inadequacy. But don’t be confused: Walter Reed is a military hospital, not a VHA hospital. Poor reporting inaccurately smeared the quietly remarkable reputation of the best medical system in America.

Over the last decade or two, the VHA system has become a worldwide leader in both the adoption and the invention of health-information technology, and it has leveraged its innovations into quantifiable gains in quality of care. As Harvard’s Kennedy School noted when awarding the VHA its prestigious Innovations in American Government prize:

[The] VHA’s complete adoption of electronic health records and performance measures have resulted in high-quality, low-cost health care with high patient satisfaction. A recent RAND study found that VHA outperforms all other sectors of American health care across the spectrum of 294 measures of quality in disease prevention and treatment. For six straight years, VHA has led private-sector health care in the independent American Customer Satisfaction Index.

Indeed, the VHA’s lead in care quality isn’t disputed. A New England Journal of Medicine study from 2003 compared the VHA with fee-for-service Medicare on 11 measures of quality. The VHA came out “significantly better” on every single one. The Annals of Internal Medicine pitted the VHA against an array of managed-care systems to see which offered the best treatment for diabetics. The VHA triumphed in all seven of the tested metrics. The National Committee for Quality Assurance, meanwhile, ranks health plans on 17 different care metrics, from hypertension treatment to adherence to evidence-based treatments. As Phillip Longman, the author of Best Care Anywhere, a book chronicling the VHA’s remarkable transformation, explains: “Winning NCQA’s seal of approval is the gold standard in the health-care industry. And who do you suppose is the highest ranking health care system? Johns Hopkins? Mayo Clinic? Massachusetts General? Nope. In every single category, the veterans health care system outperforms the highest-rated non-VHA hospitals.”

What makes this such an explosive story is that the VHA is a truly socialized medical system. The unquestioned leader in American health care is a government agency that employs 198,000 federal workers from five different unions, and nonetheless maintains short wait times and high consumer satisfaction. Eighty-three percent of VHA hospital patients say they are satisfied with their care, 69 percent report being seen within 20 minutes of scheduled appointments, and 93 percent see a specialist within 30 days.

Critics will say that the VHA is not significantly cheaper than other American health care, but that’s misleading. In fact, the VHA is also proving far better than the private sector at controlling costs. As Longman explains, “Veterans enrolled in [the VHA] are, as a group, older, sicker, poorer, and more prone to mental illness, homelessness, and substance abuse than the population as a whole. Half of all VHA enrollees are over age 65. More than a third smoke. One in five veterans has diabetes, compared with one in 14 U.S. residents in general.” Yet the VHA’s spending per patient in 2004 was $540 less than the national average, and the average American is healthier and younger (the nation includes children; the VHA doesn’t).

The VHA’s advantages come in part from its development of the health-information software VistA, which was created at taxpayer expense and is now distributed for free to any health systems that wish to use it. It’s a remarkably adaptive program that helps in virtually every element of care delivery, greatly aiding efforts to analyze symptoms and patient reactions in order to improve diagnoses and treatments, reduce mistaken interventions, and eliminate all sorts of care redundancies.

The VHA also benefits from the relative freedoms of being a public, socialized system. It’s a sad reality that in the American medical system, doctors make money treating the sick, not keeping patients well. Thus, we encourage intervention-based, rather than prevention-based, medicine. It’s telling, for instance, that hospital emergency rooms, where we handle traumas, are legally required to treat the poor, but general practitioners, who can manage conditions and catch illnesses early and cheaply, can turn away the destitute.

Moreover, patients are transient, so early investments in their long-term health will offer financial rewards to other providers. And which HMO wants to be known as the one that’s really good at treating diabetes? Signing up a bunch of diabetes patients is no way to turn a profit.

As Longman details, the VHA suffers from none of these problems. Its patients are patients for life, so investing early and often in their long-term health is cost-effective; the system was set up to deal with the sick, so the emphasis is on learning how to best manage diseases rather than avoid the diseased; and the doctors are salaried, so they have no incentives to either over- or undertreat patients. Moreover, the VHA is not only empowered to bargain down drug costs; it also uses formularies (lists of covered drugs), and so is actually empowered to walk away from a pharmaceutical company that won’t meet its offer.

The results have been clear. “Between 1999 and 2003,” writes Longman, “the number of patients enrolled in the VHA system increased by 70 percent, yet funding (not adjusted for inflation) increased by only 41 percent. So the VHA has not only become the health-care industry’s best quality performer, it has done so while spending less and less on each patient.” Pretty good for socialized medicine.

The goal of health care is to get everyone covered, at the lowest possible cost, with the highest possible quality. But in the United States, there is another element in the equation that mucks up the outcome: Our system seeks to get everyone covered, at the lowest possible cost, with the highest possible quality, while generating the maximum possible profits. Within that context, the trade-offs and outcomes all seem to benefit the last goal, and so we tolerate 45 million uninsured Americans, unbelievably high prices, and a fractured system that lacks the proper incentives to deliver high-quality care.

This makes it hard to move toward a preventive system, as Canada has, because preventive medicine pays less. It makes it hard to address moral-hazard issues wisely, as the French have, because it’s unprofitable to insure diabetics, and less profitable still to make their care essentially free. It makes it hard to institute the cost savings that Great Britain has, because with less money flowing into the system, there would be far less profit to be made. It makes it hard to harness market forces while protecting against individual risk, as Germany has, because insurer business models are predicated on shifting risk to employers and individuals, and profits are made when insurers can keep that risk from being shifted back onto them. And it is impossible to implement the practices that have so improved the VHA, because doing so would require a single, coherent health system that stuck with its members through their life cycles rather than an endlessly fractured structure in which insurers pawn off their members as they grow old, ill, or unemployed.
I would quibble that Canada does not have “a preventive system”. Some new efforts towards “public health” have been made in recent years but they have been rather small. We still basically have a doctors and hospitals universal system topped up, to varying degrees depending on the province, with some home care, residential care, and drug coverage. In BC, community care efforts like residential and home care, which could be considered preventive, have been cut back, as have supplemental health therapies like physiotherapy, massage and chiropractic, as have basic eye exams.

That’s not to say that there’s no room for profit within the American health-care system, but that it’s time the discussion stopped focusing on how to preserve the interests of moneyed stakeholders and started asking how to deliver the best care, for the lowest cost, at the highest quality — to every American. Such a system will probably still have private insurers (at least at the high end of care), pay enough to encourage pharmaceutical innovation, and allow for choice and competition and market pressures. But it will take as its guiding principle the health of the populace, rather than that of the providers. That, in the end, is what all the model health-care systems have in common. Except ours.

In a blog post, Klein follows up on Canada vs US:

… Canada is often considered a fairly mediocre system. Yet, of the 38 studies examined, 14 showed clear advantaged for Canadian patients, five suggested US care was superior, and the remainder were mixed. The studies showing the Canadian systems superiority found effects both on income — low-income Americans with breast or prostate cancer do much worse than low-income Canadians with the same conditions — and care effectiveness. For conditions like kidney failure or cystic fibrosis, Canadian care was simply better. You can pick through the tables with all the results here.

It’s not that the data shows unbelievable advantages for Canada, to be sure. As the authors conclude, “although Canadian outcomes were more often superior to US outcomes than the reverse, neither the United States nor Canada can claim hegemony in terms of quality of medical care and the resultant patient-important outcomes.” The question raised is slightly different: How can we possibly countenance a system that costs twice as much as the Canadian system but delivers slightly worse care? Even assuming diminishing returns, our expenditures should result in care outcomes at least 20% or 30% better than Canada’s. Instead, they’re about 5% worse, but cost around 187%. Does it sound like we’re getting a good deal?

One comment

Leave a Reply

Your email address will not be published. Required fields are marked *