Kip’s Korner

Kip Sullivan, JD, is Minnesota's local health policy expert, also recognized nationally for his expertise and substantial contributions to the health policy literature. Using the evidence he offers perspectives that are refreshing and unfortunately all too rare in today's health policy climate. Kip Sullivan is an active member of Physicians for a National Health Program.
Kip Sullivan, JD, is Minnesota’s local health policy expert, also recognized nationally for his expertise and substantial contributions to the health policy literature. Using the evidence he offers perspectives that are refreshing and unfortunately all too rare in today’s health policy climate. Kip Sullivan is an active member of Physicians for a National Health Program.

Patient turnover in ACOs destroys accountability

An experiment under the ACA: newly-minted Affordable Care Organizations (ACOs) will save Medicare money.  Affordable sounds good.  Care sounds good.  But how can an organization be held “accountable” for services given to a patient population when it loses one third of its patients and physicians in one year? Read more.

What “failed” in Vermont was not single-payer. Part I of III,(1/20/2015)

By Kip Sullivan, JD

On December 17, 2014, Vermont’s Governor Peter Shumlin released a report to the Vermont legislature in which he announced he would not propose a plan to finance universal coverage. His rationale was that the taxes he would have to raise to achieve universal coverage could “shock” the Vermont economy. He said his focus for the next several years would be getting rid of the fee-for-service system in Vermont and replacing it with payments that shift insurance risk onto large agglomerations of clinics and hospitals known as “accountable care organizations” (ACOs).

Although Shumlin studiously avoided using the label “single payer” in explaining his decision to abandon universal health insurance, his announcement was nevertheless almost universally described as a repudiation of the single-payer solution. This characterization of what it was Shumlin abandoned is false. The proposal Shumlin described in his December 17 report was not a single-payer system. It was a system in which taxes would replace premium payments as the insurance industry’s primary source of revenue, and spending would be slightly restrained with a cap on annual growth in total spending of 4 percent.

Not surprisingly, when Shumlin estimated what it would cost to achieve universal coverage while leaving the multiple-payer system in place, he discovered he needed more tax revenue than he thought politics would allow him to raise.

In this email, I wilI describe Shumlin’s proposal in some detail. In the February and March emails I will discuss the commentary on Shumlin’s decision by opponents and proponents of single-payer. That commentary raised several important issues that anyone who cares about universal coverage needs to understand. Those issues include whether a federal law known as the Employee Retirement Income Security Act is an impediment to a state single-payer, and whether state-level single-payer systems that don’t incorporate Medicare can cut health care costs.

On May 26, 2011, Shumlin signed Act 48, a law which was widely described as a single-payer law but which in fact merely authorized Shumlin to propose a plan for universal coverage. Act 48 did not use the label single-payer. Rather, it spoke only of universal coverage and set forth several principles that Shumlin had to observe in the course of developing his universal coverage plan. Not one of the principles called for replacing the insurance industry with a state agency as the sole insurer of Vermont residents.

In December of last year, three-and-a-half years after Act 48 was passed, Shumlin released his proposal and his rationale for not financing it. His 57-page report is available here The enormous appendices to the report are available here

Shumlin’s report (by which I mean the report plus the appendices) devotes massive amounts of ink to coverage issues and relatively little to the structure of the system he is proposing. As a result, it is easy to determine who is covered for what and difficult to determine what system Shumlin was proposing.

Shumlin’s proposed coverage can be stated simply: Beginning in 2017, he would cover all Vermont residents who are not insured by Medicare or TRICARE (the program for the military) with good coverage for traditionally covered services (physician, hospital, etc.); he would not cover long-term care, adult dental care, nor adult vision and hearing services. Vermont residents would pay 6 percent of the cost of the covered services out of pocket and taxes would cover the other 94 percent.

By contrast, the system Shumlin proposed is difficult to divine. The reader has to piece together what system he was proposing from comments spread throughout the report. Here are the essential elements:

  • Shumlin would replace premium payments by tens of thousands of employers and individuals as the primary source of insurance industry revenue with tax revenues paid out by the Green Mountain Care Board (GMCB).
  • He would hire Blue Cross Blue Shield of Vermont (BCBS) to function as a third party administrator (TPA) for the GMCB. BCBS would be to the GMCB what a TPA is to self-insured corporations – an entity that does not bear insurance risk but which carries out many other insurance functions, notably processing claims and reimbursing clinics, hospitals, and pharmacies.
  • BCBS would negotiate contracts with Vermont’s three ACOs. BCBS would be authorized to shift insurance risk onto the three ACOs and off the insurance industry by paying the ACOs first via “shared savings” contracts and eventually via total capitation (that is, a set fee per patient per year, which is no different from a premium payment).
  • BCBS would also negotiate contracts with all other providers remaining in the non-ACO (“fee-for-service”) sector.
  • BCBS would annually bring all the contracts it negotiated to the GMCB for approval, and once the GMCB approved them, BCBS would dole out tax dollars according to those contracts.
  • The GMCB and BCBS would somehow “work together” to ensure that the total cost of all the contracts did not grow by more than 4 percent annually.

This system is obviously not a single-payer system. A system in which the insurance industry’s premium revenue stream is replaced with tax revenues is not a single-payer system. We need a name for such a system. I propose we call it a single-trough system: All the insurance companies sup at the public trough. The system Shumlin refused to finance was a single-trough system, not a single-payer system. Shumlin’s proposed system bears a striking resemblance to Medicare’s Medicare Advantage program in which Medicare overpays insurance companies. It bears no resemblance to a single-payer system such as Canada’s or the system described by Minnesota Senator John Marty’s single-payer legislation.

The only element listed above that was not explicitly stated in Shumlin’s report is the hiring of BCBS as the GMCB’s TPA. Shumlin did not identify BCBS as the company he wants to hire. He used the label “designated entity” to describe the TPA. I substituted BCBS for “designated entity” because it makes the picture of what Shumlin is proposing clearer, and because there is some evidence that BCBS is who Shumlin has in mind. Shumlin refers obliquely to BCBS in a footnote (footnote 32, p. 23). Moreover, William Hsiao (the co-author of a 2011 report to the Vermont legislature recommending Shumlin’s single-trough system) and colleagues identified BCBS as the obvious choice for the TPA in an article published in Health Affairs. “Blue Cross Blue Shield of Vermont … would be a natural contractor for the single-payer system,” Hsiao wrote (Health Affairs, July 2011, p. 1233).

Not surprisingly, in describing how he calculated the total cost of his proposed system, Shumlin indicated he assumed none of the savings of a single-payer system. He stated that he assumed no reduction in administrative expenses for doctors and hospitals; he did not calculate any savings from reduced administrative costs in the insurer sector; he stated he would not assume any savings in payments to clinics and hospitals; and he said not a word about reduced drug costs that single-payers typically achieve by negotiating with drug companies.

Instead, Shumlin made it clear the only savings he anticipated achieving under his proposal would be a 1-to-2 percent reduction caused by his proposed 4-percent limit on annual growth in total spending. Because 4 percent would be less than the 5 to 6 percent rate of growth his actuaries predicted would otherwise occur, Shumlin thinks his growth cap would cut Vermont’s $5 billion annual spending bill by $75 million.

The oddest feature of Shumlin’s proposal is his expectation that Vermont’s three ACOs will either replace Vermont’s insurance industry or merge with it as the ACOs assume more and more insurance risk and gradually morph into insurance companies. Here is an example of a statement from the report that suggests that’s what Shumlin has in mind:

Certain functions currently performed … by insurance carriers warrant further analysis, as they might reasonably transfer to ACOs as providers [Shumlin likes to call ACOs “providers”] assume more responsibility and capability for managing medical risk in the future. These include:

  • Care management and care coordination
  • High-cost case management
  • Chronic illness management
  • Pre-authorization for referrals and drugs
  • Pre-certification for certain types of care
  • Utilization management .

In addition, as ACOs evolve, there would be a need for continued examination of the appropriate state role in overseeing the degree to which those entities assume insurance risk, and the degree to which insurers could reduce their need for financial reserves as risk is transferred to ACOs.” [p. 22, my emphasis]

This scenario – insurance companies ceding more and more of their functions to ACOs – raises complex questions about how the GMCB, through BCBS, is supposed to pay insurance companies and ACOs if their division of labor is changing every year. Shumlin made no attempt to answer these questions. It is a good bet that the reason he made no effort to estimate any savings from the odd system he is proposing is that neither he nor anyone else can predict what the insurance industry and the ACO sector will look like even a few years from nor whether the new ACO-insurer hybrids will be more or less expensive than the existing system of insurers and ACOs.

What little evidence we have at this date on ACOs indicates they raise rather than lower health care spending. This is true as well of other managed care ideas Shumlin endorses in his report.

Is it any wonder, then, that Shumlin felt he couldn’t afford to pay for universal coverage under his system? He proposed to funnel tax dollars into the bloated insurance industry to pay for relatively rich coverage for all Vermonters. As if that weren’t expensive enough, he proposed to impose the cost of ACOs and other managed care fads on top of that.

The lesson to be learned from Shumlin’s abandonment of “a publicly financed health care system,” as Shumlin put it, is not that “Vermont’s single payer health care plan failed,” to quote a typical comment from the blogosphere Rather, the lesson is that financing universal coverage while leaving today’s multiple-payer system in place is very difficult and probably impossible.

Any attempt to achieve universal coverage, be it through a single- or multiple-payer system, will require a shift from premium payments to tax revenues. That will not be easy even for single-payer proponents. Opponents of single-payer programs will brandish the tax increase and conceal the fact that a huge burden of premium and out-of-pocket payments will be lifted off the backs of employers and individuals. But by lowering the revenue needed to achieve universal coverage, single-payer legislation makes the task of countering scaremongering less difficult. Shumlin’s refusal to endorse single-payer made this task far more difficult than it had to be.

What “failed” in Vermont was not Single-Payer
Part II of III
February 17, 2015

Our January newsletter contained an article I wrote describing the health care system Vermont Governor Peter Shumlin said he supported but refused to finance. Shumlin announced his decision on December 17, and described his preferred system in detail in a December 30 report available here. The conclusion I presented in our January newsletter is that Shumlin essentially proposed substituting one multiple-payer system for another. He proposed a system in which an insurance company (probably Blue Cross Blue Shield of Vermont) would negotiate contracts with three hospital-clinic chains known as “accountable care organizations” (ACOs) under which the ACOs would gradually accept more and more insurance risk until they were indistinguishable from insurance companies.

In this installment of our three-part series on what happened in Vermont, I elaborate on my description of ACOs as insurance companies .

An honest definition of “ACO” would be:
“A chain of hospitals and clinics which gradually accepts more and more insurance risk and, as it does, assumes more and more of the functions of an insurance company until  it is indistinguishable from an insurance company.” But ACO proponents don’t say this. &After all, this definition begs an obvious question: What is the point of replacing the current iteration of the insurance industry with another?

ACO proponents dodge that question by offering definitions of ACO that are so vague they make rational discussion difficult. Here is a typical definition of “ACO”:
“ACOs consist of providers who are jointly held accountable for achieving measured quality improvements and reductions in the rate of spending growth. Our definition emphasizes that these cost and quality improvements must achieve per capita improvements in quality and cost, and that ACOs should have at least limited accountability for achieving these improvements while caring for a defined population of patients.”(Mark McClellan et al., “A national strategy to put accountable care into practice,” Health Affairs 2010;29:982-990).

This definition contains the ingredients common to virtually all ACO definitions, notably, language depicting a (poorly defined) group of providers being “held accountable” (by unidentified means by unidentified parties) for “measured improvements” (measured at an unknown cost to providers and the measurer) in the “cost and quality” of (unidentified) health care services delivered to a (vaguely) “defined population.”

To understand that ACOs are simply insurance companies dressed up as something new, it helps to know the history of the ACO concept.

The “ACO” label was invented at the November 9, 2006 meeting of the Medicare Payment Advisory Commission (MedPAC), a commission created by Congress to give Congress advice on how to run the Medicare program. The commission had been instructed by Congress (via the Deficit Reduction Act of 2005) to propose an alternative to the now-infamous Sustainable Growth Rate (SGR) formula. The SGR is the algorithm that ties annual growth in total spending on Medicare Part B (the part that insures the elderly and the disabled for physician services) to growth in the national economy. The SGR was supposed to induce the nation’s 800,000 doctors to order fewer services for Medicare patients, and if they didn’t, then average fees paid by Medicare would be reduced to ensure total spending on Part B services stayed under the growth limit.

The SGR hasn’t worked for obvious reasons. First, the SGR assumes Part B costs are rising because the average doctor orders too many services. There is no evidence that the typical doctor orders services patients don’t need. The evidence indicates, rather, that due to multiple factors, many of which are outside physician control, overuse and underuse are common, and underuse is far more common than overuse. Second, even if overuse is as rampant as Congress thinks it is, and even if all overuse were due entirely to physician stupidity or greed, punishing the entire US physician work force for the errant decisions of a portion of that work force can’t work. Physicians in one part of the country cannot control what physicians elsewhere do.

In its instructions to MedPAC to propose alternatives to the SGR, Congress made it clear it wanted MedPAC to propose breaking up the national physician work force into smaller pools. As a MedPAC staff member put it at MedPAC’s October 5, 2006 meeting, “The report [from MedPAC to Congress] must discuss disaggregating the current national (SGR) target into multiple pools using five alternatives: Group practice, hospital medical staff, type of service, geographic area, and physician outliers.” (p. 3). The theory motivating Congress to ask for this report was that an SGR-like limit would be more likely to induce doctors to order fewer services if the limit were applied to smaller pools.

At MedPAC’s next meeting (the November 2006 meeting), the commissioners discussed one of those pooling ideas: The idea of pools built around hospitals. To assist them in their discussion, the commission invited Elliot Fisher, a professor of medicine at Dartmouth, to present a study of what he called the “extended hospital medical staff” (EMHS). Fisher used an arbitrary formula to “connect” doctors to hospitals. His algorithm produced 5,000 EMHS’s – 5,000 pools of doctors “connected” to nearby hospitals – across the country. During the discussion between commission members and Fisher that followed Fisher’s presentation, MedPAC’s chairman Glenn Hackbarth referred to Fisher’s EMHS’s as “accountable organizations.” A few minutes later, Fisher replied, “I love your notion of accountable organizations. It’s exactly the right thing we want to create” (p. 311 of the transcript A few minutes after that another commissioner referred to “this accountable – whatever we’re calling it” (p. 356).

Even though MedPAC had no evidence that the “accountable something” could cut Medicare costs, the commission agreed to flesh out the concept and recommend it to Congress. In its June 2009 report to Congress, MedPAC made it clear it agreed with the congressional theory that Medicare’s problem is ignorant or greedy doctors ordering too many services and that ACOs could be an effective antidote to this problem. “The ACO’s role is to create a set of incentives strong enough to overcome the incentives in the FFS [fee-for-service] system to drive up volume without improving quality,” the commission wrote. “The degree to which ACOs will succeed in counterbalancing the current incentive for volume growth is uncertain. However, there is no uncertainty in the need to create a new set of incentives. The current unrestrained FFS payment system has created a rate of volume growth that is unsustainable.”[My italics] 42)

One month after that November 2006 MedPAC meeting, Health Affairs published on its website a paper by Fisher and colleagues which summarized Fisher’s presentation to MedPAC (“Creating Accountable Care Organizations: The Extended Hospital Medical Staff: A new approach to organizing care and ensuring accountability.”) Health Affairs, the most influential health policy journal in the country, subsequently published numerous papers by Fisher and other ACO proponents.

This combination of events – congressional interest in solving its SGR problem, MedPAC’s endorsement of the ACO concept, and promotion of the concept in the pages of the Health Affairs – catapulted the ACO to the center of America’s never-ending debate about how to solve the health care crisis. By 2009 the ACO fad was so hot ACOs were written into all three of the bills that would together form the Affordable Care Act (ACA). The ACO provisions required CMS to create ACOs within Medicare. These provisions became law when President Obama signed the ACA in March 2010. By 2010, insurance companies across the country had started their own ACOs to “serve” the non-elderly.

By 2010, Congress, Fisher, MedPAC and the rest of the health policy elite were no longer defining an ACO as hospital-centric. Instead, the ACA defined an ACO as an entity with control over enough primary care doctors (either via contract or ownership of their clinic) to serve 5,000 Medicare enrollees or 15,000 non-elderly patients. By 2010 ACO advocates had also fleshed out their vision of how ACOs were to be paid. They would be gradually weaned off the fee-for-service system and slowly forced to accept more and more insurance risk until they bore all insurance risk, or so much risk they had to set aside reserves and otherwise function like an insurance company.

For example, a 2010 paper by Steven Shortell, Mark McClellan (CMS administrator and FDA director under George W. Bush), and Fisher proposed that ACOs go through three stages of “growth” characterized by increasing exposure to the risk of financial loss and the promise of financial gain and by acceptance of more burdensome “quality” reporting requirements (see discussion of “tier three” ACOs at p. 1296, Shortell et al., here ). Tier three ACOs would be reimbursed by “full capitation.”

“Full capitation” is managed-care speak for “premium payment.” “Capitation” means paying a set fee per person in advance that does not change depending on the person’s medical needs. In exchange for that payment, the entity receiving the payment is under a legal obligation to pay for all necessary medical services needed by the person on whose behalf the capitation payment was made. That’s precisely what a premium is. We pay premiums in advance to entities we call “insurance companies.” In exchange the insurance company is legally obligated to pay for all necessary medical services needed by the person on whose behalf the premium payment was made.

Thus, when the Vermont legislature convened in 2011 to debate legislation that became Act 48 (legislation that called for universal coverage), the ACO was a hot health policy fad, and its advocates were unabashedly recommending that ACOs be paid in a manner that shifts insurance risk onto them. In February 2011, a report previously commissioned by the Vermont legislature , written by William Hsiao and two colleagues, recommended that the legislature enact a “single-payer system” that somehow incorporated ACOs, and these ACOs should be paid by capitation.

Hsiao et al. claimed that a single-payer system would cut Vermont’s costs by 15 percent over ten years, and a “single payer system” plus “capitated” ACOs would cut Vermont’s costs by 25 percent. Hsiao et al. claimed, as MedPAC had earlier, that ACOs would save money by reducing overuse. “Moving Vermont towards an integrated delivery system under the ACO model will also yield savings in terms of reduced overuse and duplication of service and tests,” they wrote (p. xiii).

PNHP’s Vermont chapter and its national office worked hard to keep ACOs out of Act 48. They asked me to testify before the Vermont legislature to rebut the claims made for ACOs by Hsiao et al. In my testimony before the Vermont Senate health committee, which I provided one month after Hsiao et al.’s report was published, I stated there was no evidence to support any claim that ACOs could save money, much less 10 percent of Vermont’s total health care bill.

Our efforts failed. Vermont’s legislature passed Act 48 with language in it that Governor Shumlin and other ACO proponents interpreted as endorsing ACOs. Shumlin signed the bill in May 2011.

Act 48 did not use the label “ACO.” Instead it called for “integrated delivery systems” (IDSs) that would be paid with “global budgets” to cover “a defined population” for “a defined period of time” (see the definitions of “global payment” p. 18 and “integrated delivery systems” pp. 19-20 of Act 48 definition of IDS should ring a bell. It sounds almost exactly like the “definition” of ACO. The only difference is the use of the phrase “global budget” in Act 48’s description of IDSs. But a “global budget” for an IDS is simply a year’s worth of premiums (or, if you prefer managed-care speak, capitation payments) for a “defined population.” In sum, Act 48’s definition of “IDS” is identical to the standard definition of ACO.

It isn’t clear from Shumlin’s December report how quickly he would shift insurance risk to Vermont’s ACOs. But it is clear he intends to shift risk. His report refers frequently to ACOs (see for example the reference to “our three ACOs” p. 9) and to his expectation that Vermont’s insurance industry will gradually cede the role of bearing insurance risk to “our three ACOs.” And the report states that as this transfer of risk occurs, Vermont regulators will have to require that ACOs set aside reserves and otherwise regulate ACOs as if they were insurance companies.

There is no reason to assume that Shumlin’s proposed system will be any more efficient than Vermont’s current multiple-payer system. It seems reasonable to expect that premium-collection costs will drop under such a system because insurers and ACOs won’t have to collect premiums (they will be paid via taxes in periodic installments). But Shumlin’s system will leave billing, marketing, lobbying, and utilization review costs in place, and, moreover, it will impose new requirements on insurers, ACOs and state agencies (pay-for-performance schemes and interoperable electronic medical records, for example) that will drive administrative costs up.

A “single-payer” that pays multiple ACOs via capitation/premium payments/global budgets is an oxymoron. Shumlin’s proposal is not a single-payer proposal. It is the antithesis of a single-payer proposal. It bears a much greater resemblance to the Affordable Care Act. Like the ACA, it injects more taxes into the multiple-payer system. And like the ACA, it seeks to “reform” the multiple-payer system with ACOs and other managed care fads. It is little wonder Shumlin decided he couldn’t afford universal coverage under such a system.


What “failed” in Vermont was not Single-Payer
Part III of III
May 1, 2015

The first two installments of this series focused on Vermont Governor Peter Shumlin’s characterization of the proposal he abandoned last December. We saw that Shumlin proposed an expensive three-ACO system supervised by a third-party administrator; he did not call his proposal a single-payer plan; and he refused to claim that his proposal would achieve any of the savings single-payer systems achieve.

In this last installment I want to focus on a claim made by a prominent single-payer proponent about Shumlin’s proposal that was not accurate and which added confusion to an already muddled debate about what happened in Vermont. This proponent stated accurately that what failed in Vermont was not a single-payer, and then offered the wrong rationale for his statement. He said Shumlin’s proposal was not a single-payer because Shumlin hadn’t gotten permission from the federal government to roll Medicare, Medicaid, and self-insured employers into his proposed program. I have been organizing for and writing about universal health insurance since 1986, and about single-payer systems since 1989. I have never heard this argument before.

Vermont single-payer advocates responded to this unorthodox claim shortly after it was made. I agree with their response, and would only criticize it for being too brief and for saying nothing about Shumlin’s infatuation with ACOs.

In this essay I will focus my remarks on the claim that a state program that does not incorporate Medicare and Medicaid cannot be called a single-payer. The question of whether states can include employees of self-insured employers in any type of program, single- or multiple-payer, is complex enough to justify a separate article. A federal law called the Employee Retirement Income Security Act (ERISA) is sometimes interpreted as prohibiting states from taxing self-insured employers to finance any public health insurance programs, even a small multiple-payer program like MinnesotaCare. I’ll write an article about ERISA in the next few months. It will conclude:

  1. The law on this issue has not been settled because no court has ever ruled on this issue (that’s because no one has ever sued to overturn a state single-payer system, for the obvious reason that no state has ever enacted a single-payer system), but;
  1. legal experts generally agree states can insure employees of self-insured employers through single- or multiple-payer programs.

Summary of this installment

I begin by reviewing the custom of calling the traditional Medicare program a single-payer program. That program covers only 10 percent of the American population and does not have all the features of an ideal single-payer program. Should we call it a single-payer? The vast majority of single-payer proponents, and many neutral observers, think we should. I will demonstrate that they are correct – that there is no rational reason to refer to the traditional Medicare program as anything other than a single-payer.

Next I’ll ask you to consider what label you would use to describe national legislation that left Medicare and Medicaid intact but which insured the rest of the population through a one-payer system like the traditional Medicare program. Approximately 110 million Americans are enrolled in Medicare and Medicaid out of a population of 320 million. That leaves 210 million, or two-thirds of the population, not enrolled in either Medicare or Medicaid. I’ll summarize the evidence indicating the savings such a system could achieve are almost as much as the savings that could be achieved by a system that included Medicare and Medicaid, and the reason for that is precisely because multiple payers have been replaced by one payer, and that one payer has been given the power to reduce the prices charged by the providers of medical goods and services.

Then I’ll ask, If we call the traditional Medicare program which insures only 10 percent of the population a single-payer, and if a one-payer system that covers two-thirds of the US population can save nearly as much as a system that includes the entire population, why wouldn’t we call such a system a single-payer?

Next I’ll pose the same question at the state level using Minnesota as an example, and I’ll reach the same conclusion. We will see that a Minnesota single-payer that does not incorporate Medicare and Medicaid will cover almost three-fourths of the population and, as was the case at the national level, will achieve nearly the same amount of savings as a single-payer that covered the state’s entire population of 5.4 million.  There is, in short, no rational reason to describe a one-payer system for “only” the non-poor and non-elderly at either the state or national level as anything other than a single-payer.

I’ll conclude with a brief comment on the claim that it will be impossible or extremely difficult for states to get permission to roll Medicare and Medicaid into their single-payer program. The correct statement is that it will be difficult, but it is not impossible.

Before I dive in to these issues, let me put this entire nomenclature issue in perspective. I recognize that ultimately what we all care about is an efficient and humane system of universal coverage. The label we use for such a system is merely a means to an end; we care far more about achieving that system than what we call it. But winning the fight for that system will require crystal clear communication with the public and with our allies. Having a clear definition of what a single-payer is and is not will be essential to clear communication. If we let opponents or proponents of single-payer turn the label into mush, our job will be much more difficult.

The traditional Medicare program is called a “single-payer” even though it covers only 10 percent of Americans

Let’s begin by examining the widespread practice of calling the traditional Medicare program (the original, non-Medicare-Advantage arm of Medicare) a single-payer. Out of 50 million Medicare enrollees, approximately 30 million, or about 10 percent of the US population of 320 million, are in the traditional fee-for-service Medicare program. Here is a typical example of the widespread habit of referring to the traditional Medicare program as a single-payer:

“The good news is that, in fact, a large-scale single-payer system already exists in the United States and its enrollees love it. It is called Medicare. [T]he program has been a resounding success since its introduction 48 years ago. Medicare should be expanded to cover all Americans.”

In this example, the author, Senator Bernie Sanders (D-VT), is using the traditional Medicare program the way single-payer advocates have over the years – as an example of a home-grown, single-payer program Americans hold in high regard. PNHP members and other single-payer proponents routinely refer to HR 676, the national single-payer bill, as a “Medicare for All” program.

We categorize the traditional Medicare program as a single-payer knowing full well it is not the ideal single-payer – it doesn’t cover the entire population, it doesn’t have authority to set budgets for the nation’s hospitals, and since 2003 it has been expressly forbidden by law from negotiating prices with drug companies. But it has the most essential feature of a single-payer – one payer (the Center for Medicare and Medicaid Services) bypasses insurance companies and pays doctors and hospitals directly. It also as well as other features commonly associated with single-payers, namely, the authority to set limits on payments to hospitals (via “diagnosis related groups”) and physicians (via a negotiated fee schedule). Ergo, we call it a single-payer, even though we all know it covers only 10 percent of the US population and is not endowed with the ability to set budgets for hospitals or negotiate with drug companies.

Why do we do that? First of all, because Medicare is the nation’s most efficient insurance program, public or private, precisely because it is a single-payer. It has extremely low overhead costs (1 to 2 percent of expenditures versus 15 to 20 percent for insurance companies and roughly 10 percent for self-insured employers) and relatively low physician and hospital costs. Secondly, referring to Medicare as the prototype of what we want is an honest but quick and easy way to communicate to the public what we have in mind for the entire country. The entire country is familiar with the Medicare program. By what moral or logical principle should urge Senator Sanders and all other single-payer supporters to stop calling the traditional Medicare program a single-payer and thereby deny ourselves an effective method of communicating our vision to the public? I can’t think of one.

The practice of dividing health care systems into single-payers and multiple-payers began three decades ago and is now widespread. Like all durable typologies, the single-versus-multiple-payer typology is widely used because it serves a useful purpose. The traditional Medicare program indisputably falls onto the single-payer side of that typology. No rational person would place it on the multiple-payer side.

It is well worth debating how Medicare differs from the ideal single-payer so that we’re clear on what an ideal single-payer looks like. It is not worth debating whether traditional Medicare is a single-payer. That accomplishes nothing other than to confuse a debate many people already find difficult to follow.

Can a nation or state enact a single-payer bill if it doesn’t incorporate Medicare and Medicaid?

If single-payer proponents and neutral observers have for years referred to Medicare as a single-payer even though it covers only 10 percent of the American population, why would anyone declare that state legislation that sets up a one-payer insurance program for a majority of the state’s population is not “a bona fide single-payer”?

Let’s assume the legislation we’re debating is not Vermont’s multiple-ACO bill or the Minnesota Health Plan we support in this state, but a single-payer bill at the federal level that cuts Medicare and Medicaid out of the national program. Let’s imagine, to be more precise, that HR 676 is amended in committee to keep Medicare and Medicaid as separate programs. That would mean the single-payer administrator under HR 676, the Department of Health and Human Services, would insure “only” the 210 million Americans who are not in Medicare or Medicaid.

(Here is how I calculated the 210 million figure. Our current population is about 320 million. Medicare insures roughly 50 million people, Medicaid insures roughly 70 million, and of this total of 120 million about 10 million are “dual eligibles” – they participate in both programs and thus have to be subtracted from the 120 million total to derive 110 million. Subtracting 110 million from 320 million yields 210 million.)

Let’s assume you are czar of the single-payer movement and it’s up to you to tell the rest of us whether HR 676 is still a single-payer if it covers “only” 210 million people. What would you say?

I hope you would say HR 676 is obviously a single-payer that will save a bundle of money precisely because it is a single-payer, and your only criticism of the amended version is that it won’t save quite as much money as it would have if it had covered all 320 million of us. To understand why, consider the ways in which single-payers save money. To oversimplify slightly, research indicates roughly half the savings achieved is due to reduced administrative costs, and half due to reduced prices paid to clinics, hospitals, drug companies, and equipment manufacturers. Within the category of administrative savings, roughly half is achieved within the insurer sector and half within the clinic-hospital sector.

Now let’s ask how insuring “only” 210 million people instead of 320 million would affect each type of savings. The administrative savings would be slightly less because the administrative savings in both sectors – the insurer and the provider sectors – would be slightly less. Clinics and hospitals would have to bill three programs – Medicare, Medicaid and the new single-payer for everyone else – rather than one. And HHS and state Medicaid programs would have to incur the cost of determining when people are eligible for the three programs. But these costs are a minor portion of the 30 to 35 percent of total health care spending devoted to administration. (Exactly how minor is impossible to say because the research on the components of total administrative costs does not tell us what portion is due to the cost of determining eligibility for Medicare and Medicaid.)

But aside from this diminution of administrative savings, a single-payer for 210 million Americans would be nearly as efficient as the ideal national single-payer, and even more efficient than the current traditional Medicare program. It could negotiate hospital budgets with hospitals, fee schedules with doctors, and prices with drug companies and equipment manufacturers. Because all three programs would be run by HHS, these negotiations could be unified into one process.

In short, there would be no rational reason to assert that a single-payer for 210 million Americans is not a single-payer.

Now let’s pose the same question at the state level. Is there any reason to say that a state cannot or should not refer to a one-payer insurance program for the non-elderly and non-poor as a single-payer program? Let’s take Minnesota as an example. The proportion of our population that is not enrolled in Medicare or Medicaid is slightly higher than the national proportion – about 4 million out of 5.4 million, or about three-fourths. Our state single-payer bill, the Minnesota Health Plan (MHP) bill, would cover “only” those 4 million people unless and until we got permission from the federal government to roll Medicare and Medicaid into the MHP program. Let’s assume for the sake of argument we know now that Minnesota will never ever get permission to do that, and that the MHP will “only” cover 4 million people from now till eternity. Should we call such a program a single-payer?

Of course we should. The MHP has all the important features of a single-payer – one payer reimburses doctors and hospitals directly (it bypasses insurance companies), and the one-payer has the authority to negotiate hospital budgets with hospitals, fee schedules with doctors, and prices with drug companies and equipment manufacturers. For those reasons, it would cut almost as much waste out of the Minnesota system as a national single-payer would.

A state single-payer can never save as much as a national single-payer (even if the state rolls Medicare and Medicaid into its program) because states don’t wield the negotiating clout with doctors, hospitals, drug companies and manufacturers that the federal government can. But asserting that a state single-payer cannot negotiate with as much power as the federal government is very different from asserting that a state cannot enact a single-payer. The latter claim is false. That would be like saying low-dose aspirin is not aspirin. Of course it is. It’s just not as strong as a higher dose.

One other disadvantage state-level single-payers have compared with a national single-payer is that the former must incur the administrative costs associated with billing the insurers of non-residents who seek treatment within the state with the single-payer program. This is a bigger problem for very small states such as Vermont. But even small states can cut their administrative costs drastically with a single-payer. Even employers with work forces that are tiny compared with state populations enjoy dramatic reductions in administrative costs when they tell their insurance companies goodbye and switch to self-insuring.

Finally, it is worth nothing that states have much more control over their Medicaid programs than they do over Medicare. That means those states, like Minnesota, could make the administration of their Medicaid and single-payer-for-the-nonelderly-nonpoor programs nearly uniform.

Can states get permission from the feds to incorporate Medicare and Medicaid into their single-payer programs?

For the sake of argument, I have assumed that states cannot get permission from the federal government to incorporate Medicare and Medicaid into their single-payer programs. Now I’m going to drop that assumption because it’s just not accurate. The accurate assumption is that getting permission to roll Medicare and Medicaid into state programs has always been possible and remains so today. In fact, Section 1332 of the Affordable Care Act specifically requires HHS, the agency in charge of “waivers” to permit states to incorporate Medicare and Medicaid into state programs, to create a process by which states can obtain both Medicare and Medicaid waivers simultaneously. For those of you interested in this issue, I urge you to read this interesting paper about the origins of Section 1332 by a former member of former Senator Ted Kennedy’s staff

The process states must go through to get waivers is not easy, and will no doubt be harder for states that wish to enact single-payer systems. I personally find it hard to imagine that a Secretary of HHS appointed by a president hostile to single-payer will approve of the necessary waivers any time in the near future. But we have reason to be optimistic over the longer haul. The current health care system is its own worst enemy, and the Affordable Care Act has serious problems. Reality is going to push more and more politicians into the single-payer camp or, at minimum, into the neutral camp.

Final thoughts

Let me close with a slightly unorthodox question: Should a state that enacts a single-payer system refuse to seek waivers for Medicare and Medicaid? I think there is. Once we debunk the notion that a state single-payer that doesn’t include Medicare and Medicaid cannot save money, our minds should be open to that question. Here’s my worry. My concern about rolling Medicare and Medicaid into a Minnesota single-payer program is that the fate of our single-payer would no longer hinge solely on the changing opinions of the Minnesota legislature and governor but would rely on the changing opinions of Congress and the president as well.

Through its control over the Medicare budget and roughly half of the Medicaid budget, Congress and the president could inflict stress on a Minnesota single-payer that included the Medicare and Medicaid programs. This could happen even if the Minnesota legislature and governor were doing their best to ensure the entire program was adequately funded and was being efficiently run. In that event, the public would have a hard time knowing who to hold accountable.

Property taxes offer a good analogy. Voters often don’t know whether to be upset with their local city, county or school board when property taxes go up, or with the state legislature. That’s because all four levels of government have a hand in determining property taxes.

Do we want to create a similar problem in health care? I’m not sure we do.




Why Obamacare can’t lower costs

PNHP Latest News, May 14th, 2014

How to Think Clearly about Medicare Administrative Costs

Journal of Health Policy, Politics and Law, February 2013

Counterpoint: Managed Care Needs a Better Watchdog

Star Tribune, January 2013

Two thirds of Americans support Medicare for All
PNHP Blog, 2011

Testimony to Vermont Senate Health Committee on the problems with Accountable Care Organizaitons
March 2011

History and Definition of the Accountable Care Organization, Health Policy’s Latest Fad
PNHP California Blog, October 2010

Why Minnesota Health Care Report Card is Flawed
Star Tribune, November 2010

Physician report cards are like No Child Left Behind -they don’t improve health outcomes
Star Tribune, January 2008