Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). (c) Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy).

Subscriber: null; date: 21 January 2018

Paying for Healthcare

Abstract and Keywords

This chapter examines the sociopolitical aspects and the mechanisms of payment for healthcare in the United States. It begins with a historical background on the fragmented structure of insurance and payment, using Medicare as an example. It then describes the two most important prospective payment systems used by Medicare, one for inpatient care and the other for physician services, and argues that both systems are responsible for the continued increase in expenditures because they are not nearly as prospective as designed. The chapter also discusses the technical aspects of the two systems and the rising expenditures within the overall social and political context. Finally, it explains how payment is tied to disaggregation at three levels—payer, unit of payment, entity paid—and contends that the mechanisms it uses account for the uncontrolled increase in healthcare expenditures.

Keywords: payment, healthcare, United States, insurance, Medicare, inpatient care, physician services, healthcare expenditures, payer, unit of payment

I Introduction

Healthcare expenditures in the United States, whether measured as a percentage of gross domestic product (GDP) (17.7%) or as dollars spent per capita ($8,508), are nearly 50% higher than the countries with the next highest level of expenditures, the Netherlands (11.9% GDP) or Norway ($5,669 per capita).1 Recent research has placed blame for our high expenditures squarely on the prices we pay to providers,2 as well as our high administrative expenses, while ruling out other drivers of expenditures.3 The population in the United States is no older than that of other advanced, industrialized nations; in fact, the converse is true in many cases. In the United States we don’t utilize higher amounts of healthcare goods and services, with the possible exception of very costly technology and some specialty care; and we don’t substitute more intensive care for less intensive care any more than in comparable nations. Nor are the higher expenditures in the United States explained by higher quality; again, the converse is often the case. The simple truth is that in the United States we pay providers more money than is paid elsewhere.4

(p. 833) An ineluctable fact of payment is that one person’s expenditures are another person’s income. As a result, payment is an inherently political and social process.5 This chapter explores the manner in which this political and social process and the mechanisms of payment are linked and somewhat singular to the United States. In all industrialized nations, providers, particularly physicians, have enormous income-generating capacity, particularly when that power is wielded within and by the organizations of modern, advanced capitalism. The United States has no monopoly on the fact that providers are powerful. However, singular to the United States is an extremely fragmented insurance system in which each payer fends only for itself.6 Many other nations, like the countries of Western Europe, have multiple payers, but their payment is coordinated through some mixture of private and public power (often referred to as “all-payer” systems).7 No such coordination exists in the United States. As a result, compared with other nations, the payment equation, “expenditures equal incomes,” is tilted heavily toward the income side—physicians, hospitals, nursing homes, device and drug manufacturers, home health agencies, and so forth—as there exists no countervailing power on the payer side.8 The mechanisms of payment in the United States both reflect and reinforce this imbalance.

As described much more fully below, this structural feature of payment in the United States reinforces and is reinforced by several cultural elements and other structural aspects. Most particularly, in designing and implementing payment, administrators and politicians and much of the policy elite, particularly those influenced by health economics, explicitly try to avoid making the distributional choices that are the hallmark of payment in any nation. They seek to avoid politics by using technical means instead. Absent the use of collective means exercised in a payment system, one can expect growth in expenditures to continue to eclipse that of comparable nations.

The chapter begins by describing how the fragmented structure of insurance and payment became dominant and how it continues. It uses Medicare as the primary example because it is the largest single payer in the United States and because its methods are transparent and often adopted by private payers, while those of private insurers are proprietary and therefore largely unavailable. Using this example, the chapter describes the two most important prospective payment systems used by Medicare, the one for inpatient care and the other for physician services. The chapter makes the point, through examination of the technical apparatus of the payment systems, that the two systems are not nearly as prospective as designed, with the result that expenditures continue to rise. The chapter then connects the technical apparatus of the systems, and the resultant increasing expenditures, to the overall social and (p. 834) political context, particularly the eschewal of mechanisms to make explicitly the political distributional decisions at the heart of any payment system. Our systems use some mechanisms common to comparable nations’ payment systems, but when these elements are considered as a whole and within larger structure and context, one can see that payment in the United States is exceptional, as are our expenditures.

II History of Fragmented Structure of Payment

The nature of healthcare payment is inexorably linked to the nature of illness and healthcare insurance. Because illness is a probabilistic event, the risk that a particular individual will become ill cannot be predicted with confidence, while, because of the law of large numbers, confidence increases over larger classes of individuals. In the ideal, insurance involves the exchange of a sum certain, the premium, in exchange for certain services, likewise specified in advance, in the event of loss. Health insurance deviates from this ideal in the fact that both the probability and magnitude of loss can be altered after execution of the insurance contract. An insured can affect the probability of loss through behaviors increasing or decreasing the possibility of loss, a “problem” known as “moral hazard.” Healthcare providers can affect the existence of loss by defining and locating illness in individuals, and they likewise can affect the magnitude of loss through the services they deem warranted to treat a particular illness, phenomena discussed in economic literature as “supplier-induced demand.”9 Symmetrically, a healthcare insurer can affect the existence and magnitude of loss by denying services altogether or by curtailing their scope, amount, or duration. Neither the probability nor magnitude of loss is easily specified in advance, but, again due to the law of large numbers, the degree of uncertainty can be attenuated by risk pooling.

In industrialized countries other than the United States, this risk pooling is accomplished through either or both of governmentally created systems and social organizations, thereby creating systems grounded in social solidarity and social security. In Western Europe, with its strong tradition of solidarity and mutual aid, the risk of illness is pooled in the sickness funds, which cooperate voluntarily and with varying degrees of state coercion in collecting revenues, spreading risk, and paying providers. In national health system countries, like the United Kingdom and Scandinavia, national governments unite all citizens into a single risk pool. In Canada, Canadian Medicare brings the population together into the risk-pooling systems of the provinces and the territories, with some degree of federal subsidization. All advanced, industrialized countries socialize the risk-pooling function, and they pay with one voice, either as a single payer or as coordinated payers. Payment is “aggregated.”

By contrast, the United States is “the odd man out.”10 In the United States, risk pooling was, for present purposes, initially attempted by the creation of the Blue Cross and Blue Shield (p. 835) organizations.11 Unable to collect payment from individuals during the Great Depression, hospitals came together to form Blue Cross, effectively pooling among themselves the risk that a particular one of them would be unable to obtain payment from patients, who became subscribers. Shortly thereafter physicians followed suit with the formation of Blue Shield for payment for outpatient care. The world of provider-dominated, indemnity insurance was born.

Indemnity insurance constituted a method of solving the unpredictability of the occurrence and magnitude of illness in particular patients. In advance, no hospital or physician can identify which subscriber will access its services, and, vice versa, no subscriber can predict whether he or she will need the services of any provider, much less the identity of which one or ones. Because indemnity insurance gave each subscriber the right of access to all providers and, vice versa, granted to all providers the right of access to all subscribers, it provided a mechanism of linking all hospitals and all physicians to all subscribers.

A principal drawback of this approach was that it contained no means to control expenditures. Subscribers were indifferent to price because, aside from relatively insignificant deductibles or copayments, they paid nothing when they used care. Therefore, they had no reason to exercise countervailing power against providers’ ability to generate income for themselves, nor did provider-dominated insurers. Providers were “reimbursed” rather than “paid.”

To be reimbursed, most hospitals executed participation agreements with the Blue Cross companies, which stipulated that each hospital accepted assignment of subscribers’ right to indemnification of loss; to be reimbursed, hospitals merely had to document their “costs.” Reimbursement was therefore “cost-based” and “retrospective.” Physicians, by contrast, by and large refused to execute such participation agreements. For the most part subscribers paid cash on the barrel, paying for services at the time of delivery, and then sought indemnification from their insurers. Reimbursement was therefore “charge-based” and “retrospective.” The whole enterprise was a claims-paying enterprise, effectively operating as a blank check. Predictably, expenditures grew and grew as this form of health insurance became increasingly widespread—the payment, aka reimbursement, side of the payment equals incomes equation—and as healthcare, particularly medicine, increasingly became the technologically driven endeavor it has now become—the income-inducing side of the equation.

Over time, control over the Blue Cross and Blue Shield organizations was wrested from providers. If most people in the United States had been Blue Cross–Blue Shield subscribers, and if the Blues, each created under state law, had remained relatively united in the Blue Cross–Blue Shield national association, then this system of insurance could have exercised countervailing power against providers either to lower the price of care, or to control the volume and intensity of care, or to accomplish both, much as has happened elsewhere in other developed nations. However, in the United States this power failed to materialize for a number of reasons.

For one thing, starting in the 1930s and accelerating in the period after World War II, commercial insurance companies entered the market against the Blues. Unhindered by the obligation imposed on the Blues by state law to accept all comers and to charge a uniform, (p. 836) community-rated premium to all, the commercial insurance companies were able to cherry-pick employer-sponsored groups, which are relatively young and healthy, thereby leaving to the Blues the relatively bad risks. In turn, the Blues needed to increase their premiums, which put them at a competitive disadvantage relative to the commercial insurance companies, which cost them more relatively young, healthy groups, which necessitated increased premiums, and so on in a vicious circle known as the death spiral. The possibility of uniting the entire community within one enterprise fell further away.

For another thing, the entire community was never so united anyway. For a variety of reasons, described in Chapters by Jost and Hall (this volume), in the United States the organization of subscribers into risk pools fell largely to employers. In contrast to the rest of the advanced, industrialized world, other forms of social organization never existed or arose to take on the role of pooling potential patients into risk-bearing groups. Necessarily, given that the health insurance system was primarily employer-based, nonworking portions of the population—particularly the poor, the elderly and the young—were left out.

Nor was there in the United States a Bismarckian moment whereby government brought together increasingly fractured, squabbling groups of risk pools. Instead, government in the United States took on the role of a gap-filler, ameliorating, perhaps, the sharpest edges of a system that left out those not favored by the dominant employer-sponsored model. In 1965 Medicare and Medicaid, in a truly unique, almost revolutionary political moment,12 insured the elderly and some of the poorest of the poor. Over time other marginalized groups have been incrementally added: the disabled; those needing kidney dialysis; uninsured children in low-income families; and, under the Affordable Care Act (ACA), various categories of persons who have been uninsurable because of prior illness or who lack sufficient income to afford insurance, for example, the “working poor,” who fall outside of Medicaid’s traditional categories.

III Mechanisms of Fragmented Payment

Given this fragmentation of the social and political function of risk pooling, each particular payer was (and still is) left to its own devices in paying providers. This fragmentation of payment both spawned and reinforced a culture and social structure in which each payer acts only for itself and tries to push risk and cost onto others.

Medicare provides the most prominent example. The Medicare program packs the punch of the largest risk pool by far, and many payers adopt its methods, which are transparent. Medicare is also backed by the sovereign power of the United States; it is not just some private payer acting in the marketplace. However, despite these sources of power, from the very beginning Medicare has operated as only one payer among others, charged by Congress and its administrators to pay only its “fair share,” no less and no more. This continuing policy of “Medicare-only” results in structures of payment and administration designed to focus just on Medicare’s bottom line, not the wider impacts of its policies.13

(p. 837) Because of this narrow view, even Medicare’s pioneering move to “prospective” payment has failed to staunch the rise of expenditures in the United States, and prospective payment may have failed to control expenditures even to a great extent within the program itself (although this point is controversial, as indicated in section III.b. below).

In contrast to retrospective reimbursement, prospective payment is designed to impose a constraint on expenditures. To the extent that providers’ costs exceed the expected costs embodied in the prospective system, providers lose the difference. To the extent that costs are less than the expected norm, providers retain the surplus. Prospective payment thus is supposed to provide predictability and incentives to control costs.

Yet, a system remains prospective only if the payer or payers hold the line. If, instead, amounts that would be provider losses are forgiven by means of adjustments after the relevant accounting period—that is, “passed through”—then the system, to that degree, becomes retrospective. To a great extent, compared with other nations, that is the history of prospective payment in the United States.

Prospective payment, of course, relies in part on prediction of spending. Some of the difficulty of predicting expenditures is that the delivery of healthcare is relatively variable, subject to standardization to a degree much less than even relatively heterogeneous products such as custom-built houses. The cause of such heterogeneity is debated. One extreme position is that healthcare is so extremely differentiated, because no two patients are alike, that standardization is impossible. In contrast, another extreme position is that providers differentiate their services so as to escape the control imposed by standardization—that is, standardization could be easily accomplished. Most observers fall somewhere in the middle: It is relatively difficult to standardize healthcare.

Standardization, however, is the sine qua non of any payment system for which a goal is controlling expenditures. Take the modern hospital. We could standardize the care provided to all patients across all hospitals by paying for an average day. This “per diem” payment could be cost-based—the cost of the average day in all hospitals, derived by totaling all costs of all days and dividing that sum by the total number of days. However, the per diem need not be cost-based. A payer could decide that it can afford to spend a sum certain, say one million dollars, in an upcoming accounting period, divide that expenditure by the number of patient-days expected, and derive its per diem in that manner. The point is that any payment system must use some standardized unit of payment to impose some degree of control on expenditures, as discussed further in section III.a. below.

The particular units chosen for payment in the United States have generally been fairly discrete—they are episode- or activity-based. The two most significant mileposts in the transformation toward prospective payment were the creation of the Inpatient Prospective Payment System (IPPS) and payment to physicians based on the resource-based relative-value scale (RBRVS), a fee structure designed to be divorced from what physicians historically charged and based instead on the “value” of their services.14 For IPPS, the unit of payment is each inpatient episode of care, differentiated by the diagnosis that led to the stay. (p. 838) This categorization is accomplished though the use of diagnosis-related groups (DRGs), each of which is supposed to represent a homogeneous cluster of patients, the treatment of which does—or should, and the difference, discussed in section III.a. below, is crucial—command an identical use of resources, for example, “concussion,” “pneumonia,” “myocardial infarction.” For the RBRVS-based physician payment system, the unit of payment is the individual service furnished—for example, an initial office visit, a follow-up office visit, an injection, reading a CT scan—differentiated from other services by its “value,” as measured primarily by a work relative-value unit (RVU), a practice expense RVU, and a malpractice expense RVU (the cost of malpractice insurance). The two systems are designed to standardize the provision of healthcare to allow comparisons across each provider peer group, hospitals and physicians, and to impose a common payment across each, set in advance and free of provider control—in contrast to the provider-controlled payment methods, discussed previously in this section, with which Medicare started. A blank check no more.

For two principal reasons, things have not quite worked out as intended. The first is internal to the structures of the payment systems, while the second is outside in the systems’ interaction with the larger payment environment.

a. Inside

The two systems themselves have remained, to some extent, within the hands of providers, with the result that both IPPS and the RBRVS-based system have, again to some degree, functioned as pass-throughs.

With regard to IPPS, the story begins with the fact that the system was tied to the behavior of hospitals under retrospective, cost-based reimbursement. The DRGs form an ordinal ranking of the value of each hospital stay—more precisely each “discharge” because the “diagnosis” for payment is coded upon discharge—relative to the value of other discharges. The ordinal scale then has to be converted to actual payment dollars. To accomplish this task, each DRG is multiplied by the “base payment rate,” also known as the “standardized payment amount” or the “average standardized amount.” This “conversion factor” was derived from an actual, historical base period by calculating from that base period the average cost of all discharges in all acute-care hospitals in the United States, and then removing, through a process called “standardization,” other known costs that do not vary with diagnosis and over which a hospital has no control. Most important, these recognized costs include certain local variations, particularly local labor costs, as well as factors like teaching status, known to raise costs uniformly because patients at teaching hospitals are sicker than those at nonteaching facilities. This average cost, so standardized, thus represented—again, during an actual, historical base year—the average cost of the average discharge in the average acute-care hospital in the United States. Recognized variations are then added back in for each hospital.

Each year, the base payment rate is updated in a process that mixes expertise with politics. The “update factor” is proposed by the Department of Health and Human Services (HHS), subject to a recommendation from the Medicare Payment Assessment Commission (MedPAC), which is an expert body advising Congress, and then it is codified by Congress in legislation. Leaving aside the lobbying by the hospital sector, the updating process inevitably considers how hospitals are faring under IPPS, perhaps most saliently by considering “Medicare margins.”

(p. 839) An analogous point can be made with regard to the DRGs themselves. To begin with, the weights were initially calculated from hospital charges, which can be thought of as a complex set of list prices. To put the matter charitably, charges rarely have any rational relationship to actual resource use and are used simply to obtain greater revenue from payers15—so much for cleaving IPPS apart from provider-controlled retrospective reimbursement.

Further, the distinctions among the various DRGs are not airtight and the classification system can be gamed to maximize payment. Hence, twice the system has experienced “DRG creep,” in which the value of the weights has risen as providers became more familiar with the coding system and learned how to work it in their favor. This creep first occurred in the early years of implementation. It then occurred again, as actually predicted, after the weights were partially redesigned for fiscal year 2008.

Indeed, this redesign itself was sparked because a MedPAC study showed that the system was highly manipulable by providers. The study concerned the creation of specialty hospitals and allegations by community hospitals that the specialty hospitals were cherry-picking profitable cases. A particularly salient finding of the study was that specialty hospitals were formed in good part to take advantage of particularly lucrative DRGs and, moreover, to cherry-pick the less complex and therefore less expensive cases, with high margins, within those DRGs. This manipulation of the payment system was possible because the DRGs failed to capture so-called “within-DRG severity of illness.” Moreover, these differences were highly visible to, and therefore manipulable by, providers.16 While these issues have been somewhat ameliorated by the rebasing of, and by other changes made to, the weights and the classification system, these reforms were then followed by substantial DRG creep as hospitals became more familiar with the reformed classification system—the newly implemented Medicare severity-adjusted DRGs (MS-DRGs). Although the monetary value of some of this creep has been “taken back” from the hospital sector through a deduction in calculating the base payment update factor, the fact remains that the classification system itself, like the base payment rate, is somewhat controlled by providers. While perhaps on paper Medicare has controlled its inpatient expenditures more successfully than have other payers, it remains the fact that the key components of the system—the conversion factor and the MS-DRGs—render it far less prospective than designed.

The story regarding provider manipulation of the RBRVS-based system is much simpler. The RBRVS was developed by economists and colleagues at Harvard and was tested and refined by HHS and the Physician Payment Review Commission, Congress’s expert advisory body for physician payment at that time, the functions of which are now combined with oversight of other prospective payment systems in MedPAC. Two major purposes informed this work. First, the goal was to slow growth of expenditures. Between the enactment of Medicare in 1965 and the mid-1980s, expenditures for physicians’ services increased at an average annual rate of more than 13% and averaged 16% a year during the first half of (p. 840) the 1980s, thereby consistently outpacing growth in both GDP and total healthcare expenditures. Second, the system was explicitly designed to increase payments to generalists relative to specialists as a means to induce more medical graduates to enter primary care. To achieve both goals, the fee structure was designed to be totally independent of actual physician charges and thus provider control.17

Unfortunately, it did not stay so independent. Like the MS-DRGs, the RBRVS is an ordinal ranking of the value of different services relative to others, and like the DRG-based system, the RBRVS-based one has to be updated periodically to reflect changes in the relative-value units, particularly the work RVU. In this task, the Centers for Medicare and Medicaid Services (CMS), the agency within HHS that administers Medicare, relies heavily on the American Medical Association’s Relative Value Scale Update Committee (RUC). Substantial evidence shows that over time the dominance of RUC in the updating process was skewing the scale toward procedures and imaging, furnished most often by specialists, and thereby actually increasing the differential between payment for specialty and primary care. This adverse impact was (and still is) occurring because (1) RUC is dominated by specialists; (2) most requests for revaluing an RVU come from specialty societies; (3) the great majority of weights reviewed have been ones that were allegedly undervalued, with the result that review has most often led to an increase in value; (4) any increase in one weight diminishes the value of others because adjustments are “budget neutral”; and (5) CMS generally accepts the RUC’s recommendations.18

To redress the manner in which the updates to the weights have contributed to the increasing tilt toward specialists, particularly procedures and imaging, in 2006 MedPAC issued some fairly strong recommendations, all designed to reduce CMS’s reliance, in performing the update, on the medical specialty societies. Most important, MedPAC recommended that CMS form an expert panel of its own to advise it in its review of the RVUs. Under the recommendation, the panel, composed of experts in health economics and physician payment, would not “supplant” RUC but would “augment” it by providing an independent source of expertise. MedPAC also recommended that, separately from the congressionally required five-year reviews, CMS should, based on a number of factors MedPAC specified, identify values that might be ripe for reduction.19

CMS has adopted some of the recommendations, some of which were also codified in the ACA.20 Over the past several years, CMS has begun to identify “misvalued” services along the lines of criteria MedPAC recommended; it has obtained input from nonphysicians; and it has announced that it will conduct annual reviews instead of five-year ones. While CMS has failed to adopt MedPAC’s most important recommendation, the establishment of a review panel independent of specialty society control, it has complied with the ACA’s mandate to (p. 841) establish a formal process to review the work RVUs. The broadened process includes input from the medical directors of Medicare contractors, the opportunity for public nomination of codes to be reviewed, and contracts with consultants to validate the work RVUs.21 Nonetheless, MedPAC continues to express concern with reliance on the RUC,22 and it is not clear whether these changes will stem the RVUs’ drift toward widening the disparity between primary and specialty compensation.

The second aspect of provider manipulation of the RBRVS-based system concerns the substantial power of physicians, particularly specialists, to generate their own income by increasing the volume of care. To account for this “volume effect,” the RBRVS-based system employs a mechanism, the Sustainable Growth Rate (SGR) formula, which is an “expenditure target,” intended to take back any increase in aggregate expenditures resulting from increased volume. The SGR formula allows the volume of fee schedule services to grow at the same rate as per capita GDP.

In each year from 2002–2012, actual expenditures have exceeded the target. Congress, however, has intervened each year after 2003 to reduce the take-back for that year. The formula then compounds the shortfall in successive years—just like an interest rate compounds each year by adding to the base over which it is taken—and updates in any given year are based on what spending would have been had the full reductions occurred. The effects are now very substantial. For 2013, the take-back would have cut fees by a whopping 24%.23 Because that level of shock to providers is unreasonable, Congress has again suspended application of the SGR, therefore continuing a payment freeze.

This dynamic further skews the RBRVS away from cognitive services and toward procedures and imaging. Procedures and imaging are churned far more readily than time-intensive cognitive services because the latter, but not the former, are subject to the constraint that face time is needed, and there are only so many hours in the day. Moreover, this dynamic defeats the attempt to wrest expenditures away from a provider-dominated fee system. Nonetheless, because Congress has simply frozen pay, the RBRVS-based system, like IPPS, “on paper” has controlled expenditures for physicians’ services to a limited extent and relative to other payers.

All countries face three dimensions of aggregation-disaggregation: (1) whether payment is aggregated or disaggregated; (2) whether units of payment are aggregated or disaggregated; and (3) whether the entities paid are aggregated or disaggregated. Although the United States is distinctive in its disaggregation of payment, its mechanisms to address the other two dimensions bear some similarity to those used in comparative nations, although, importantly, as adapted to its particular structure, history, and culture. In contrast to disaggregation of payment, with regard to the two other dimension we’re talking questions of degree, even though as we will see, the sum of each part adds up to a whole that is very distinctive.

(p. 842) All payment has to utilize some unit of payment, which can be relatively aggregated—“global”—or relatively disaggregated—“activity-based.” As discussed in this section previously, hospitals could be paid for discrete services, for a DRG or based on a global budget. The first unit of payment is the most disaggregated, the second method less so while remaining activity-based, and the global budget stands at the extreme side of this aggregation-disaggregation dimension. Payment based on relatively global units are divorced from the quantity and types of services provided—a hospital is paid on a global budget or for a day of care regardless of what is actually done—while episode-based payment is connected much more closely with activity—“discharge-based” payment, as indicated above, is dependent on what is actually done because the coding of the discharge into a DRG is based, in good part, on what services were furnished during the inpatient stay.

The entity being paid can likewise be relatively aggregated or disaggregated. A single radiologist, for example, could be paid for the particular services he or she provides, or the radiologist could be paid by an individual salary. Alternatively, radiology could be compensated based on a contract price paid to an entire radiology group servicing an entire hospital. Analogously, a hospital can be paid separately from an after-care facility like a nursing home; or the hospital and the after-care facility could be aggregated into one entity paid through a “bundled payment” among them. Indeed, a major reason for the failure of the IPPS to constrain overall federal expenditures is that as financial pressure has been put on inpatient stays, hospitals have shifted care to more lucrative sites, paid outside IPPS, such as hospital outpatient departments, ambulatory surgery centers, rehabilitation facilities, and so forth.

No level of aggregation or disaggregation is perfect.24 Smaller units of payment allow payers to monitor and control more of what is performed. For example, if one is concerned that a physician on salary, or a hospital on a global budget, will tend to provide fewer or less intensive services to all patients or to select patients who require fewer or less intensive services, one can instead pay on a fee-for-service, or per diagnosis, basis. However, providers paid on more discrete units of payment have incentives to increase volume of care. Providers also have an incentive to unbundle services because they can make more money by shifting costs to others or increasing volume. Hospitals paid under IPPS have incentives to discharge patients “quicker and sicker” to shift costs of care to whatever provider ends up with the patient next, for example, a nursing home; the hospital’s payment for the MS-DRG remains the same regardless of the duration of the inpatient stay. Hospitals likewise have incentives to unbundle services to get paid “twice”—for example, move to an outpatient department, testing that would otherwise occur after inpatient admission, because that way the hospital obtains both inpatient and outpatient payment. Conversely, if one is concerned that a physician paid by fee-for-service, or a hospital per diagnosis, is going to churn volume, thereby increasing expenditures and possibly harming patients, then one can put the physician on salary, or the hospital on a budget (but then it is difficult to monitor activity).

(p. 843) By and large payment in the United States has tended toward disaggregation on all three dimensions—fragmented payers use relatively disaggregated units of payment furnished to relatively disaggregated providers. While the ACA and other more recent policy initiatives, such as accountable care organizations (ACOs) and a variety of demonstration projects discussed in the chapter by Saver (this volume), represent attempts to bundle payment across the continuum of care, the fact remains that payment in the United States is still disaggregated on all dimensions.

b. Outside

The second major reason that neither IPPS nor the RBRVS-based system is as prospective as designed is due to the fact that Medicare’s payment policy is “Medicare-only.” As a result Medicare’s payment “is assessed relative to the costs of treating Medicare beneficiaries, and [MedPAC]’s recommendations address a sector’s Medicare payments, not total payments.”25 In other words, Medicare sets its rates with indifference to the effect on other payers or the healthcare system more generally. As noted, this “Medicare-only” policy has allowed Medicare to control its expenditures relative to other payers.26

This fact has led some to charge that Medicare has shifted costs to other payers,27 while others lay the finger of blame on the failure of private payers to control their expenditures.28 Regardless of the fact this argument has clear implications for the question whether Medicare is paying “enough,” this “push-pull” debate obscures the fact that the continuing escalation of expenditures is a joint product of all of Medicare and private payers—they’re all in it together because the problem is fragmented payment.29 IPPS and the RBRVS-based system have allowed Medicare to control its expenditures but only relative to other payers, not absolutely and not relative to comparable nations. Whether the failure to control overall system expenditures stems from Medicare’s pushing other payers’ expenditures up, or whether the continuing escalation of payment is caused by other payers’ pulling Medicare’s expenditures up, either or both effects push and pull expenditures in the United States to ever higher levels. Without coordinated payment—that is, with the continued fragmentation of payment—the push-pull will simply continue.30

(p. 844) IV Social and Political Context of Payment: The Denial of Politics

So far we have seen that the singular, extreme degree of disaggregation of payment in the United States prevents the assertion of countervailing power against providers; it also comes at the price of significantly higher administrative costs, perhaps as much as a third or fourth more than other nations, because payment methods are neither standardized, nor (except for Medicare’s) are they transparent, both of which cause significant transactions costs.31

The reasons that payment in the United States is singularly disaggregated across all dimensions are, obviously, very complicated, but one can say that the relative disaggregation across the board comports with a number of aspects of social and political context in the United States. The most obvious one is that fragmentation of payers and disaggregated units of payment are mutually reinforcing. A hospital, most saliently, cannot be paid on a global budget, as has been traditional in many comparable countries, because payment on a global budget requires coordination among payers. In turn, the fragmented culture and structure tend to support methods of payment that drill down as far as possible to each element of care so as to ensure that no payer pays anything more than the costs of its subscribers32—paying anything more would benefit another payer, likewise competing to pay just its costs and nothing more.33

To enable each payer to pay just its costs alone, payment must necessarily be cost-based, and “accuracy” is a primary goal of health services research and healthcare policy and administration.34 Yet, if payment “accurately” reflects costs, then costs just pass through and a prospective system is prospective no more. The search for accuracy thus takes the form of a search for measuring “real” or “true” or “efficient” costs. This search boils down to finding some means to evaluate which costs are eligible for payment. This evaluation, in turn, requires the use of some hypothecation, here: how a hypothetical provider would perform in a hypothetical market free of the “distortions” that characterize the actual markets that exist in healthcare. The hypothecation can define “efficiency” to be the result of competition relative to a chosen “yardstick,” which serves as a proxy when markets fail—for example, competing around the average cost of the average hospital—or “efficiency” can be thought (p. 845) about in standard neoclassical terms as allocative efficiency.35 Either way, “accuracy” simply means that payers, like those in charge of Medicare, are deciding, through “accurate” determinations of what is “real,” the level of resources to inject into the healthcare sector: “The word ‘real’ thus has an unintended double meaning: (1) ‘real’ in the sense that a phenomenon occurs independently of measurement; (2) ‘real’ in the sense that the phenomenon will be recognized as ‘happening’ by the system and built into rates.”36

In other nations payment is much less focused on accurately measuring costs and is structured to enable explicit political decisions regarding, first, how much a society can afford to spend on healthcare, as opposed to bridges, highways, schools, defense, and so forth, and, second, the allocation of standards of living to providers.37 These decisions are made through collective mechanisms, which may mix private and public elements and many of which, from the perspective of the United States, too much resemble collective bargaining.38 As examples, France applies its DRG weights within overall budget priorities and expenditure targets set for each sector by its Parliament and Ministry of Health, and then by states and regional health authorities, which contract with hospitals, and with rates adjusted for failure to meet expected aggregate expenditures39; Germany uses DRGs in combination with volume-adjusted “revenue budgets” negotiated between peak associations, primarily the sickness fund and hospital associations40; in decentralized Switzerland, cantons use DRGs within the context of an overall budget negotiated between the hospital and sickness fund associations subject to approval of the relevant canton.41 These mixed uses can accomplish diverse purposes, like monitoring activity within a budgeting system; allocating funds in internal markets; distributing aggregate federal spending among states or regions in a regionalized system; paying for patients who cross regional or national boundaries.42 These (p. 846) ends all involve distributional decisions, revealing how payment is a political and social process.

Indeed, the mechanics of payment in the United States effectively obstruct the rendering of explicit distributional choice. Clearly, as in all systems using a DRG-like unit of payment, there have been winners and losers under IPPS, but the “rough justice”43 from the use of averages is much rougher in the United States than elsewhere. IPPS is a remarkably top-down national payment system used across a vast and diverse landscape. Even France, with its Cartesian culture, does not use its version of DRGs—its patient classification system (PCS)—in such a fashion. Its PCS units are applied to expenditures allocated regionally. Further, its PCS is much more highly differentiated than are the MS-DRGs; and certain services like chemotherapy and radiotherapy, which would significantly reduce within-PCS homogeneity if bundled within the PCS, are unbundled and paid separately. The combined result of these features is that the averages generated through these mechanics are much more “granular” and therefore sweep far less broadly than the averages used in the United States, in a trade-off between attaining a much more tailored system at the cost of greater administrative complexity and enhanced ability of hospitals to game the system.44 By contrast, our use of less differentiated MS-DRGs to allocate an annual national expenditure occurs at such a dizzying height—national flat rates per MS-DRG—that no matter how many adjustments for local variations are allowed one still gets at best a Hubble-like view of the distributional effects being wrought.45 Therefore, comparatively, the manner in which we use DRGs in the United States serves to preclude the very political distributional choices that are at the center of any payment system.

In the United States, politics is a dirty word and something to be avoided. As a result, we lack collective mechanisms such as exist in other nations to obtain coordination, either voluntary or compulsory, and to resolve disputes; the United States is left with payment policy as its only basic tool. The problems with the SGR and its predecessor, the Volume Performance Standard (VPS) aptly illustrate this point. In their use of the VPS, creators thought they were modeling methods used in British Columbia in the 1980s.46 However, the designers failed to understand the difference between imposing an aggregate target or (p. 847) cap and negotiating one, and moreover, they missed the fact that to succeed, a target has to be used in conjunction with the other methods then deployed in British Columbia47; most notably, bilateral negotiations among parties who, locked together for the long haul, have to compromise48; and control on the supply side, as British Columbia in those years was, in draconian fashion, eliminating current physician stock by refusing to grant, or severely restricting geographically, billing numbers for some physicians, and it was actively preventing the diffusion of capital into and within the ambulatory care sector, instead of reimbursing it, as we do, and (worse still) instead of spreading its recovery, as does the RBRVS system, across a relative-value scale.

Because these forms of collective action are too “political” to be used in the United States, our nation denies itself the use of these varying systemwide, “macrolevel” forms of controlling utilization through control of investment in capital, including human capital. As Professor Joseph White quips, “to paraphrase the movie Field of Dreams, if you do not build it they cannot come.”49 Other nations have learned that it is far easier politically to control capacity at the outset than to deny its use after the fact because the latter deprives investors of expected returns and patients of expected benefits. To control expenditures, it is better to nip those expectations in the bud.50 Such “regulation,” however, would be anathema to the avoidance of politics in the United States. Instead, as a weak substitute the attempt to control utilization occurs at the point of service, the “microlevel,” through the use of financial incentives and aggressive utilization review at the point of service—when the barn is already on fire.

Given the eschewal of overt politics, the fixation on accuracy of payment at a microlevel comes laden with the technocratic wish to be saved from politics by reliance on technical answers to (non)political questions of what payment should consist of. Both IPPS and the RBRVS-based systems were developed with such goals in the background. Although IPPS’s standardized payment amount was based on actual, historical costs, it was expected that the historicity of the base payment rate would fade with time. The aspiration was that it could be updated annually in a scientific manner devoid of politics (or anything else) such that the payment system would reflect only the efficient payment of cases calibrated by their severity. The hope for the RBRVS was similar. It supposedly derived from the scientific analyses of the value of work—the time, mental effort and judgment, technical skill and physical effort, and stress associated with potential risk for a patient—combined with supposedly objective measures of practice expense, including liability insurance. With that foundation in place, scientific and “accurate” updates could preclude extraneous considerations such as politics.51

(p. 848) One reads the mountains of technical reports, however, and wonders how anyone can seriously conclude that politics is eliminated. For example, as discussed in section III.a. above, each year MedPAC recommends an annual update to the IPPS base payment rate. At times it has utilized what is called the “discretionary adjustment factor.” The elements canvassed have included at various times: beneficiaries’ access to care (e.g., number of hospitals and beds, growth of specialized services, volume of services, hospitals’ access to capital); quality of care (e.g., safety indicators, satisfaction measures, readmission rates); new technologies; changes in site of care; coding behavior; changes in productivity; provider costs; Medicare margins; total margins; projected costs and margins; costs of “relatively efficient hospitals”; prior underpayments or overpayments; various “one-time factors”; and the general healthcare environment. These considerations are stirred in a pot—a report—and out pops a recommendation, such as the following.

In considering its update recommendation, the Commission has struck a balance between a number of competing factors. On the one hand, average total Medicare margins are negative (–5 percent in 2009 and projected to reach –7 percent in 2011). On the other hand, our update framework indicators (access to care, including supply and service volume; quality of care; and access to capital) are positive. Furthermore, the negative Medicare margins are due at least in part to the lack of private financial pressure for cost containment, and the set of hospitals identified as efficient have a median Medicare margin of about 3 percent. On the basis of these circumstances, the Commission contemplated an update of 2.5 percent.

However, two additional considerations led the Commission to its recommended update of 1 percent. For inpatient services, the Commission and others have documented past and ongoing overpayments resulting from changes in documentation and coding after implementation of MS-DRGs in 2008. Current law does not allow full recovery of past overpayments and no action has been taken to stop the ongoing overpayments. The Commission believes that all overpayments should be recovered and that the most urgent step is to stop the ongoing overpayments. To accomplish this objective, the Commission would reduce the ongoing overpayment by 1.5 percentage points—that is, the difference between its contemplated update of 2.5 percent and its recommended update of 1 percent. This adjustment would account for 1.5 percentage points of the 3.9 percent adjustment needed to fully prevent accumulation of further overpayments.52

Behind technical names like “discretionary update factor” lies a very simple fact: This is simply a very, very global budget. It is a political judgment how much money should be devoted to Medicare hospitals nationwide for acute inpatient care. This extremely macrolevel judgment is then drilled down to the individual hospital and to the individual case through the complex mechanisms of IPPS—and not in a well-tailored fashion at all.

Finally, consonant with this focus on microlevel control, as well as the technocratic wish embodied in relatively disaggregated units of payment, is the American penchant to control expenditures through financial incentives as a means of behavioral control.53 Payment (p. 849) becomes primarily a means to obtain some other end rather than the allocation of income and choice of the amount of expenditures. One can see this concept in the adoption of the DRGs, as just one example among many, as the following illustrates.

Payment on the basis of a per-case rate for each DRG is intended to create specific financial incentives that encourage hospital management to adopt desirable methods of controlling the cost of care. It was hoped that hospital management, facing a separate payment rate per discharge for each DRG, would have strong incentives to: (1) improve productivity; (2) use less expensive inputs where possible; (3) influence physicians to reduce the length of stay, limit the volume of inpatient services, and use a less expensive mix of services to treat each patient; (4) specialize in treating types of cases the hospital can produce efficiently; and (5) adopt cost-reducing technologies, while avoiding cost-increasing technologies.54

One can also see this concept in the latest fad, value-based purchasing—discussed further in the chapter by Saver in this volume that rests on an alleged scientific base that desired outcomes can be scientifically derived and that those ends can be achieved through a scientifically tailored use of financial incentives.55 IPPS, RBRVS-based payment, and like schemes “promise instrumental control akin to the actions of a skilled watchmaker. When the clock is running too fast, the watchmaker intervenes to slow it down. When the clock runs too slowly, the task is to speed it up. It is all a matter of clockwork.”56 This very strong emphasis on payment as a means of behavioral control is, relatively, another example of American exceptionalism in payment.57

In sum, payment in the United States is characterized by disaggregation at all three levels—payer, unit of payment, entity paid—and it uses mechanisms that are failing to control expenditures: methods focused on costs and accuracy that actually obscure distributional effects; the use of payment as the sole lever to control utilization without utilizing collective, macrolevel methods; and the resultant attempt to control price and utilization through use of microlevel, finely tuned, financial incentives—all in the technocratic wish to displace overt political decisions regarding levels and allocation of expenditures and provider incomes. Taken together, these relative, comparative differences combine to render the (non)system of payment in the United States exceptional, singular too in its result, our ever-rising expenditures. We are stuck, as if in mud.58

(p. 850) V Conclusion

Healthcare works best when payers, patients, and providers are joined together for the long haul in integrated pools such that there is continuity of care, incentives to invest in prevention, and stability in revenue generation and payment (two sides of one equation). This chapter has discussed how a very fragmented payment system feeds on itself and becomes entrenched in culture and institutions; and this fragmentation enables cost-shifting, volume generation, shifting of sites of care, risk selection, upcoding, and the like. This behavior is then countered by payers’ moves to rebundle services, even to rebundle providers into entities like ACOs (and numerous other examples), in a constant game of move, countermove.59 However, these efforts are likely to be successful only in a marginal way because they are swimming upstream against so much fragmentation elsewhere—the multitude of private payers (e.g., over one million ERISA plans, over seven thousand exchange plans60), which operate with different rules, along with a multitude of public payers (e.g., Medicare, fifty-one state Medicaid plans, Children’s Health Insurance Program (CHIP), which likewise operate with different rules. In principle, under highly regulated circumstances the finance side can remain somewhat disaggregated—like in Germany or moreover, in Switzerland and the Netherlands—while the payment side is coordinated.61 However, in the United States, at least right now, aggregating the payment side is unlikely to occur.

When compensation is short term as part of constantly changing relationships—like the ideal widget of simple neoclasssical models— then no one has any long-term incentives, much less long-term bonds that we describe with terms like “solidarity,” “loyalty,” and “commitment.” In the world of fragmented payers using fragmented units of payment given to fragmented providers, it is very difficult to solve the problem of collective action necessary to “cross-subsidize”—now a pejorative term in the United States—social goods like education and the maintenance of standby capacity like trauma units.

The story of payment in the United States, recounted here, is a bleak one. The expenditure fire in the United States continues to burn out of control. Regardless of rhetoric about “bending the cost curve,”62 the structure and culture of payment remain the same. Furthermore, (p. 851) recent evidence exists that growing consolidation in both the insurance and provider sectors, as well as increased vertical integration between them, is correlated with enhanced prices paid to both levels.63 Without some radical change, business will continue as usual, if not get worse.

However, some glimmer of change is illuminating the horizon. Under the ACA’s authority to experiment with innovative payment methods, CMS has granted permission to include Medicare spending in an overall per capita expenditure cap to control volume as part of Maryland’s all-payer rate-setting system for hospital services—currently the only such system in the United States.64 Vermont’s single-payer system remains on the books,65 although as of March 2015 the governor has announced that he cannot now put forward the required public financing plan because it would require a substantial increase in taxes.66 Massachusetts, within its statewide universal health insurance scheme, is experimenting with payment methods that would bring all providers within an overall expenditure target.67 Finally, given the mounting evidence that price is the main driver of expenditures, and given the fact that nothing else seems to be working, even some members of the policy elite who are generally supportive of markets in healthcare have started to discuss seriously all-payer state rate setting.68 While it is unclear whether these efforts can spread to other states,69 we can at least say that a whiff of aggregating payment, at least at the state level, is in the air.

Notes:

(1) Organization for Economic Co-Operation and Development, Health Data 2013—Frequently Requested Data, http://www.oecd.org/health/health-systems/oecdhealthdata2013-frequentlyrequesteddata.htm (accessed June 2, 2014).

(2) Gerard F. Anderson et al., It’s the Prices, Stupid: Why the United States Is So Different from Other Countries, 22 Health Aff. 89 (2003); Carlos Angrisono et al., Accounting for the Cost of Health Care in the United States, Mckinsey Global Institute (2007); Alexis Pozen & David M. Cutler, Medical Spending Differences in the United States and Canada: The Role of Prices, Procedures, and Administrative Expenses, 47 Inquiry 124 (2010); Paul B. Ginsburg, Reforming Provider Payment—The Price Side of the Equation, 365 New Eng. J. Med. 1268 (2011); Miriam J. Laugesen & Sherry A. Glied, Higher Fees Paid to US Physicians Drive Higher Spending for Physician Services Compared to Other Countries, 30 Health Aff. 1647 (2011).

(3) See, e.g., Dante Morra et al., US Physician Practices Versus Canadians: Spending Nearly Four Times as Much Money Interacting with Payers, 30 Health Aff. 1 (2011); Lawrence P. Casalino et al., What Does It Cost Physician Practices to Interact with Health Insurance Plans?, 28 Health Aff. w533 (2009); Stephanie Woolhandler et al., Costs of Health Care Administration in the United States and Canada, 349 New Eng. J. Med. 768 (2003).

(4) See, e.g., Uwe Reinhardt, Divide et Imperia: Protecting the Growth of Health Care Incomes (Costs), 21 Health Econ. 41 (2012).

(5) See, e.g., Robert G. Evans, Coarse Correction—And Way Off Target, 22 J. Health Pol., Pol’y & L. 503, 504 (1997); Uwe E. Reinhardt, Resource Allocation in Health Care: The Allocation of Lifestyles to Providers, 65 Milbank Q. 153 (1987).

(6) See, e.g., Theodore R. Marmor, American Health Care Policy and Politics: Is Fragmentation a Helpful Category for Understanding Health Reform Experience and Prospects?, in The Fragmentation of U.S. Health Care: Causes and Solutions 343, 352 (Einer R. Elhauge ed., 2010).

(7) See, e.g., Joseph White, Implementing Health Care Reform with All-Payer Regulation, Private Insurers, and a Voluntary Public Insurance Plan 5 (May 3, 2009), http://www.ourfuture.org/files/JWhiteAllPayerCostControl.pdf (accessed June 3, 2014).

(8) See, e.g., Reinhardt, Divide; Bruce C. Vladeck & Thomas Rice, Market Failure and the Failure of Discourse: Facing Up to the Power of Sellers, 28 Health Aff. 1305 (2009); Jonathan Oberlander & Joseph White, Public Attitudes Toward Health Care Spending Aren’t the Problem; Prices Are, 28 Health Aff. 1285 (2009).

(9) See, e.g., Roger Feldman & Michael A. Morrisey, Health Economics: A Report on the Field, 15 J. Health Pol., Pol’y & L. 627, 640–641 (1990).

(10) Brian Abel-Smith, Who Is the Odd Man Out?: The Experience of Western Europe in Containing the Costs of Health Care, 63 Milbank Q. 1 (1985).

(11) For the history of the Blue Cross–Blue Shield organizations, see Sylvia A. Law, Blue Cross: What Went Wrong? (1976) see generally Paul Starr, The Social Transformation of American Medicine 290–334 (1982).

(12) See generally Theodore R. Marmor, The Politics of Medicare (2000).

(13) See, e.g., David M. Frankford, The Complexity of Medicare’s Hospital Reimbursement System: Paradoxes of Averaging, 78 Iowa L. Rev. 517 (1993).

(14) The discussion of Medicare’s IPPS and RBRVS-based system draws heavily from Sara Rosenbaum & David M. Frankford, Law and the American Health Care System ch. 12 (2012); Frankford, Complexity; David M. Frankford, Measuring Health Care: Political Fate and Technocratic Reform, 19 J. Health Pol., Pol’y & L. 647 (1994); Jonathan Oberlander, The Political Life of Medicare ch. 5 (2003); and David G. Smith, Paying for Medicare: The Politics of Reform (1992).

(15) See, e.g., Uwe E. Reinhardt, The Many Different Prices Paid to Providers and the Flawed Theory of Cost Shifting: Is It Time for a More Rational All-Payer System?, 30 Health Aff. 2125 (2011) [hereinafter Reinhardt, The Many Different Prices Paid to Providers]; Uwe E. Reinhardt, The Pricing of U.S. Hospital Services: Chaos Behind a Veil of Secrecy, 25 Health Aff.57 (2006) [hereinafter Reinhardt, The Pricing of U.S. Hospital Services].

(16) See MedPAC, Report to the Congress: Physician-Owned Specialty Hospitals (Mar. 2005).

(17) See, e.g., William C. Hsiao et al., Results and Policy Implications of the Resource-Based Relative-Value Study, 319 New Eng. J. Med. 881 (1988); William C. Hsiao et al., Results, Potential Effects, and Implementation Issues of the Resource-Based Relative Value Scale, 260 JAMA 2429 (1988); Physician Payment Review Commission, Medicare Physician Payment: An Agenda for Reform 4–5, 20–21, 35 (Mar. 1987).

(18) See MedPAC, Report to the Congress: Medicare Payment Policy ch. 3 (Mar. 2006) [hereinafter MedPAC March 2006 Report]; see also MedPAC, Report to the Congress: Medicare and the Health Care Delivery System 12–17 (June 2011).

(19) See MedPAC March 2006 Report, at 142–148.

(20) ACA § 3134, Pub. L. No. 111-148 (2010).

(21) See, e.g., Dep’t Health & Human Services, Center for Medicare & Medicaid Services, Medicare Program; Revisions to Payment Policies Under the Physician Fee Schedule, Clinical Laboratory Fee Schedule & Other Revisions to Part B for CY 2014, 78 Fed. Reg. 43,282, 43,285, 43,302–304 (July 19, 2013).

(22) See, e.g., Letter from Glenn M. Hackbarth, Chairman, MedPAC to Marilyn Tavenner, Administrator, CMS 6–10 (Aug. 30, 2013), http://www.medpac.gov/documents/083013_MedPAC_PartB_COMMENT.pdf (accessed June 3, 2014).

(23) See, e.g., MedPAC, Report to the Congress: Medicare Payment Policy 97, 259 (Mar. 2014).

(24) For a comprehensive discussion of the trade-offs among payment methods, see R. P. Ellis & M. M. Miller, Provider Payment Methods and Incentives, in Health Systems Policy, Finance, and Organization 322 (Gúy Carrin ed., 2009); see also Randall P. Ellis & Thomas G. McGuire, Provider Behavior Under Prospective Reimbursement: Cost Sharing and Supply, 5 J. Health Econ. 129 (1986).

(25) MedPAC, Medicare Payment Policy: Report to the Congress 30 (Mar. 2011) [hereinafter MedPAC, March 2011 Report].

(26) See, e.g., MedPAC, March 2011 Report, at 51–57.

(27) See, e.g., Allen Dobson, The Cost-Shift Payment “Hydraulic”: Foundation, History, and Implications, 25 Health Aff. 22 (2006).

(28) See, e.g., Chapin White & Vivian Yaling Wu, How Do Hospitals Cope with Sustained Low Growth in Medicare Prices?, 49 Health Servs. Res. 11 (2014); Austin B. Frakt, How Much Do Hospitals Cost Shift?, 89 Milbank Q. 90 (2011); MedPAC, March 2011 Report, at 51–57; Jeffrey Stensland et al., Private-Payer Profits Can Induce Negative Medicare Margins, 29 Health Aff. 1045 (2010).

(30) This article does not address the more general debate concerning the use of markets to finance healthcare. Switzerland, for example, like the United States, has multiple payers, and they compete on some dimensions of healthcare delivery and finance, while unlike the United States payment is aggregated, or, in other words coordinated.

(31) See, e.g., text and notes at note 3; White, Implementing.

(32) Cf. Elizabeth Kilbreth, Paying by the Rules: How Eliminating the Cost Shift Could Improve the Chances for Successful Health Care Reform, 35 J. Health Pol., Pol’y & L. 177 (2010) (cost-shifting is corrosive of social solidarity needed to create consolidated insurance).

(33) For a general discussion of the relationship between adoption of a payment technology like DRGs and social and political context, see, e.g., Thomas D’Aunno et al., Conclusions: The Global Diffusion of Casemix, in The Globalization of Managerial Innovation in Health Care 346 (John R. Kimberly et al. eds., 2008); John R. Kimberly, DRGs in Western Europe: Lesson and Comparisons in Managerial Innovation, in The Migration of Managerial Innovation: Diagnosis-Related Groups and Health Care Administration in Western Europe 340 (John R. Kimberly & Gérard de Pouvourville eds., 1993); see also Frankford, Measuring; David M. Frankford, The Medicare DRGs: Efficiency and Organizational Rationality, 10 Yale J. Reg. 273 (1993).

(34) See Frankford, Measuring; see also Joseph White, Cost Control After the ACA, Sept./Oct. 2013 Pub. Admin. Rev. S24, S28–S29.

(35) See, e.g., Andrew Street et al., DRG-Based Hospital Payment and Efficiency: Theory, Evidence, and Challenges, in Diagnosis-Related Groups in Europe: Moving Towards Transparency, Efficiency and Quality in Hospitals 93–94 (Reinhard Busse et al. eds., 2011). For scathing criticism of initial use of average costs in England, as in the United States, as a yardstick to set rates, see Andrew Street & Alan Maynard, Activity Based Financing in England: The Need for Continual Refinement of Payment by Results, 2 Health Econ., Pol’y & L. 419 (2007).

(37) See, e.g., Naoki Ikegami & Gerard F. Anderson, In Japan, All-Payer Rate Setting under Tight Government Control Has Proved to be an Effective Approach to Controlling Costs, 31 Health Aff. 1049 (2012).

(38) See, e.g., Wilm Quentin et al., Hospital Payment Based on Diagnosis-Related Groups Differs in Europe and Holds Lessons for the United States, 32 Health Aff. 713 (2013).

(39) See, e.g., Zeynep Or & Martine Bellanger, France: Implementing Homogeneous Patient Groups in a Mixed Market, in Diagnosis-Related Groups in Europe, at 221; Martine M. Bellanger & Philippe R. Mossé, The Search for the Holy Grail: Combining Decentralised Planning and Contracting Mechanisms in the French Health Care System, 14 Health Econ. S119 (2005).

(40) See, e.g., Alexander Geissler et al., Germany: Understanding G-DRGs, in Diagnosis-Related Groups in Europe, at 243.

(41) See, e.g., Hervé Guillain, Casemix in Switzerland, in The Globalization of Managerial Innovation in Health Care, at 176. See also International Profiles of Health Care Systems, 2013 (Commonwealth Fund Study, Sarah Thomson et al. eds.), at 122.

(42) See, e.g., Diagnosis-Related Groups in Europe; The Globalization of Managerial Innovation in Health Care; The Migration of Managerial Innovation; see also Frankford, Measuring.

(43) Rick Mayes & Robert A. Berenson, Medicare Prospective Payment and the Shaping of U.S. Health Care (2006).

(44) See, e.g., Or & Bellanger, France; see also Conrad Kobel et al., DRG Systems and Similar Patient Classifications Systems in Europe, in Diagnosis-Related Groups in Europe, at 37. For a fuller comparison of the different construction and usage of DRG-like PCSs, compared with the United States, see, e.g., Quentin et al., Hospital Payment.

(45) See, e.g., Bruce C. Vladeck, Medicare’s Prospective Payment System at Age Eight: Mature Success or Midlife Crisis?, 14 U. Puget Sound L. Rev. 453, 479–480 (1991). This view was shared by Judith Lave, who was the director of HCFA’s Office of Research from 1980 to 1982, the period in which IPPS was developed. See Judith R. Lave et al., A Proposal for Incentive Reimbursement for Hospitals, 11 Med. Care 79, 85–89 (1973) (proposing the use of a case-adjusted per-case reimbursement system with cost-sharing between the government and a hospital for deviations from the per-case rate); Judith R. Lave, Hospital Reimbursement Under Medicare, 62 Milbank Memorial Fund Q. 251, 254–256 (1984) (recommending the immediate abandonment of the use of national rates); Judith R. Lave, The Impact of the Medicare Prospective Payment System and Recommendations for Change, 7 Yale J. on Reg. 499, 521–527 (1990) (recommending various changes to reduce the reliance upon national rates).

(46) See Paul B. Ginsburg & Philip R. Lee, Defending U.S. Physician Payment Reform, 8 Health Aff. 67 (1989); see also Smith, Paying for Medicare, at 198.

(48) As Joe White remarked concerning negotiations and arbitration in Canadian provinces, “Kicking and screaming in many cases, both sides have been dragged into recognizing that they have to live together, that rules are necessary, and neither side should be able to impose its will.” Joseph White, Competing Solutions: American Health Care Proposals and International Experience 69 (1995).

(49) Joseph White, Targets and Systems of Health Care Cost Control 24 J. Health Pol., Pol’y & L. 653, 659 (1999).

(50) See, e.g., Morris L. Barer et al., It Ain’t Necessarily So: The Cost Implications of Health Care Reform, 13 Health Aff. 88, 95 (1994).

(51) See, e.g., Frankford, Measuring; Smith, Paying for Medicare, at 4, 11, 71–72, 84, 86, 100, 106, 232–233, 246–247. See generally James A. Morone, American Political Culture and the Search for Lessons from Abroad, 15 J. Health Pol., Pol’y & L. 129 (1990); Gary A. Belkin, The Technocratic Wish: Making Sense and Finding Power in the “Managed” Medical Marketplace, 22 J. Health Pol., Pol’y & L. 509 (1997). For a discussion of the scientism in such endeavors, see David M. Frankford, Scientism and Economism in the Regulation of Health Care, 19 J. Health Pol., Pol’y & L. 773 (1994).

(52) MedPAC March 2011 Report, at 60. See generally Frankford, Complexity, at 618–635.

(53) As Bill Glaser wrote, “policy analysis in health care finance has specialized in technical economics rather than in the construction of politically feasible structures.” William A. Glaser, The Politics of Paying American Physicians, 8 Health Aff. 129, 131 (1989); see also William A. Glaser, Designing Fee Schedules by Formulae, Politics, and Negotiations, 80 Am. J. Pub. Health 804 (1990). Quite arguably, Glaser tended to paint too pretty a picture of consensus rather than the assertion of power. Compare, e.g., William A. Glaser, Doctors and Public Authorities: The Trend Toward Collaboration, 19 J. Health Pol., Pol’y & L. 795 (1994), with, e.g., Victor G. Rodwin, Physician Payment Reform: Lessons from Abroad, 8 Health Aff. 76 (1989).

(54) Prospective Payment Assessment Commission, Report and Recommendations to the Department of Health and Human Services 16 (Mar. 1, 1990). For criticism of the heroic assumptions behind such behavioral control, see Frankford, The Medicare DRGs.

(55) See generally Sandra J. Tanenbaum, Pay for Performance in Medicare: Evidentiary Irony and the Politics of Value, 34 J. Health Pol., Pol’y & L. 717 (2009).

(56) David M. Frankford, Managing Medical Clinicians’ Work Through the Use of Financial Incentives, 29 Wake Forest L. Rev. 71, 71–72 (1994).

(57) As Joel Cantor wrote to me, this difference “is like European football compared to American football. The rules of the latter are so complex that they require a host of experts, instant replays, and so-forth, to know what is going on. In European football, one simply has to get goals.”

(58) See, e.g., David Chinitz & Victor G. Rodwin, What Passes and Fails as Health Policy and Management, 39 J. Health Pol., Pol’y & L. 1113 (2014); see also Mark Stabile et al., Health Care Cost Containment Strategies Used in Four Other High-Income Countries Hold Lessons for the United States, 32 Health Aff. 643 (2012).

(59) See generally David M. Frankford, The Normative Constitution of Professional Power, 22 J. Health Pol., Pol’y & L. 185 (1997).

(60) Robert Wood Johnson Foundation, Health Insurance Exchange Compare Dataset, http://www.rwjf.org/en/research-publications/find-rwjf-research/2014/03/breakaway-policy-dataset.html (accessed June 4, 2014).

(61) For a truly illuminating discussion of how carefully calibrated such a system must be, see Tsung-Mei Cheng, Understanding the “Swiss Watch” Function of Switzerland’s Health System, 29 Health Aff. 1442 (2010).

(62) For a critique of the alleged cost controls in the ACA, see White, Cost Control. For contrary arguments that price cuts in the ACA will have spillover effects to reduce expenditures of other payers, see Chapin White, Cutting Medicare Hospital Prices Leads to a Spillover Reduction in Hospital Discharges for the Nonelderly, 49 Health Servs. Res. 1 (2014); White & Wu, How Do Hospitals; Chapin White, Contrary to Cost-Shift Theory, Lower Medicare Hospital Rates for Inpatient Care Lead to Lower Private Payment Rates, 32 Health Aff. 935 (2013).

(63) See, e.g., Austin B. Frakt et al., Plan-Provider Integration, Premiums, and Quality in the Medicare Advantage Market, 48 Health Servs. Res. 1996 (2013); see also Laurence C. Baker et al., Vertical Integration: Hospital Ownership of Physician Practices Is Associated with Higher Price and Spending, 33 Health Aff. 756 (2014); Leemore Dafney et al., Paying a Premium on Your Premium? Consolidation in the US Health Insurance Industry, 102 Am. Econ. Rev. 1161 (2012); Martin Gaynor & Robert Town, The Impact of Hospital Consolidation—Update, Robert Wood Johnson Foundation Synthesis Report (2012), http://www.rwjf.org/content/dam/farm/reports/issue_briefs/2012/rwjf73261 (accessed June 8, 2014); Ann S. O’Malley et al., Rising Hospital Employment of Physicians: Better Quality, Higher Costs?, Center for Health System Change, Issue Brief 136 (2011); http://www.hschange.com/CONTENT/1230/1230.pdf (accessed June 10, 2014); William B. Vogt & Robert Town, How Has Hospital Consolidation Affected the Price and Quality of Hospital Care?, Robert Wood Johnson Foundation, Research Synthesis Report No. 9 (2006).

(64) See, e.g., Rahul Rajkumar et al., Maryland’s All-Payer Approach to Delivery-System Reform, 370 New Eng. J. Med. 493 (2014).

(65) See, e.g., Ashley M. Fox & Nathan J. Blanchet, The Little State That Couldn’t Could? The Politics of “Single-Payer” Health Coverage in Vermont, J. Health Pol., Pol’y & L, published ahead of print February 19, 2015, doi: 10.1215/03616878-2888381, http://jhppl.dukejournals.org/content/early/2015/02/12/03616878-2888381.abstract (accessed March 16, 2015).

(66) See, e.g., Peter Shumlin, “This Is Discouraging News,” Rutland Herald, December 21, 2014, http://www.rutlandherald.com/article/20141221/OPINION06/712219919/1018/OPINION (accessed March 16, 2015).

(67) See, e.g., Robert E. Mechanic et al., The New Era of Payment Reform, Spending Targets, and Cost Containment in Massachusetts: Early Lessons for the Nation, 31 Health Aff. 2334 (2012).

(68) See, e.g., Paul B. Ginsburg & L. Gregory Pawlson, Seeking Lower Prices Where Providers Are Consolidated: An Examination of Market and Policy Strategies, 33 Health Aff. 1067 (2014). Ginsburg, Reforming; Robert Berenson, Unchecked Provider Clout in California Foreshadows Challenges to Health Reform, 29 Health Aff. 699 (2010); see also Joseph P. Newhouse, Assessing Health Reform’s Impact on Four Key Groups of Americans, 29 Health Aff. 1714, 1723 (2010).

(69) Compare, e.g., Robert Murray, The Case for a Coordinated System of Provider Payments in the United States, 37 J. Health Pol., Pol’y & L. 679 (2012), with, e.g., Mark Pauly & Robert Town, Maryland Exceptionalism? All-Payers Regulation and Health Care System Efficiency, 37 J. Health Pol., Pol’y & L. 697 (2012).