Causal Chains and Cost Shifting: How Medicare’s Rescue …bmayes/pdf/16.2mayes.pdf · Causal...

THE JOURNAL OF POLICY HISTORY, Vol. 16, No. 2, 2004.Copyright © 2004 The Pennsylvania State University, University Park, PA.

RICK MAYES

Causal Chains and Cost Shifting:How Medicare’s Rescue InadvertentlyTriggered the Managed-Care Revolution

The conventional wisdom on how managed care came to replacetraditional fee-for-service reimbursement as the nation’s dominantmode of health insurance is that enlightened businesses and theiremployers led the way in responding to the emergence of marketforces in health care in the 1990s.1 A common textbook treatmentof managed care’s ascendancy puts it this way: “Transformation ofthe health care delivery system through managed care has been drivenprincipally by market forces, and reinforced by government.”2 Theirony is that the opposite sequence of events is a more accurate por-trayal of what actually happened. As this article shows, the transfor-mation of America’s health-care system through managed care wasinitially triggered—albeit indirectly—by government actions andthen driven by market forces. In other words, before business behav-ior was a cause of managed care’s extraordinary growth, it was largelya response to and an unintended consequence of governmentpolicymaking: in this instance, Congress’s reform of Medicare in1983.

It is intuitively appealing to assume that the paradigm shift fromfee-for-service insurance to managed care was solely the result ofthe business community seeking to reduce costs by increasing mana-gerial control and market mechanisms. “Firms face a very clear in-centive structure: they must strive to maximize profits,” as PaulPierson and Jacob Hacker note. “This conclusion does not rest onassumptions about individual greed, but on the recognition thatmarket systems are powerful mechanisms for inducing decision-mak-ers to adopt profit-maximizing behavior.”3 But why, then, did themajority of businesses wait so long to begin switching their employ-ees en masse into managed care? Lawrence Brown has explained why

Angelia Fell

muse stampl

RICK MAYES 145

managed care did not thrive in the 1970s, despite concerted govern-ment action on its behalf.4 Nevertheless, why did employers stillnot begin a big switch to cheaper managed care by the early 1980sor at least by the mid-1980s at the latest? Managed care had been amandated alternative since 1974, a year after President Nixon signedthe HMO Act, which required businesses with more than twenty-five employees that already offered health insurance to make HMOsavailable to their employees.5

Business’ delayed transition to managed care suggests that ex-isting market incentives were necessary but not sufficient for induc-ing such a major paradigm shift in health insurance. Providers(doctors and hospitals) and patients greatly dislike managed care,relative to fee-for-service health insurance, because it both restrictspatients’ access to more expensive medical care provided by special-ists and limits physician autonomy. Employers, on the other hand,generally do not care about the specific form of health coverage theyprovide until its cost becomes a significant issue. Thus, there was nonatural incentive or tendency for employers and employees to switchfrom fee-for-service insurance to managed care.

As the following analysis shows, the critical catalyst for makingmarket incentives sufficiently appealing for this massive paradigmshift came as a result of change to another major actor in the Ameri-can health-care system, the single largest purchaser of hospital care:Medicare. Since Medicare and employers in the private sector pur-chase their medical care from the same hospitals and doctors, a dra-matic change to Medicare’s payment policy was bound to greatlyaffect (directly and indirectly) the cost-benefit calculations andpolicy decisions of private employers. “Medicare is the 800 poundgorilla,” observes David Abernethy, former senior Medicare special-ist and staff director of the House Ways and Means Health Subcom-mittee. “So when it slows its rate of expenditure growth, hospitals’overall rate of revenue growth slows; and that, in the end, puts thefinal pressure on private payers.”6

By examining the links in the causal chain between the reformof Medicare’s payment policy and the rise of managed care, we findthat government policymakers used prospective payment as a pow-erful tool to help balance the federal budget at the expense of health-care providers, especially hospitals.7 Instead of increasing the payrolltax for Medicare or making Medicare beneficiaries pay more for theirmedical care, government leaders increased less visible tax expendi-tures—tax revenue foregone—by precipitating a significant increase

MEDICARE AND THE MANAGED-CARE REVOLUTION146

in health insurance costs for businesses (see Christopher Howard’sThe Hidden Welfare State for more on this common government ap-proach to fiscal policy).8 Employers responded, in turn, by ditchingmore expensive fee-for-service insurance for their workers in favorof cheaper managed care. Ultimately, this linkage shows that noth-ing can transform an industry more quickly and profoundly than whenthe government—if it is an industry’s single largest customer—dra-matically alters how it pays for goods and services.9

Policy Feedback and Causal Chains

Social scientists often take a “snapshot” view of political life, ex-plains Paul Pierson. “How does the distribution of public opinionaffect policy outcomes? How do individual social characteristics in-fluence propensities to vote? . . . Disputes among competing theo-ries center on which factors (“variables”) in the current environmentgenerate important political outcomes.”10 But the significance of suchfactors, he points out, is “frequently distorted when they are takenfrom their temporal context.”11 So there is a strong case to be madefor shifting from snapshots to “moving pictures,” especially for study-ing events or phenomena that unfold over longer periods of time(often years).12 This is particularly true for studying sequences, ar-gues James Mahoney, in which “an event may trigger a chain of caus-ally-linked events that, once itself in motion, occurs independentlyof the institutions that initially trigger it. This sequence of events,while ultimately linked to a critical juncture period, may culminatein an outcome that is far removed from the original juncture.”13 Thetrick, Pierson adds, is to trace the chain of events and test how strongthe links are between them.14

Accordingly, this article contributes to a growing body of re-search that focuses on how public policies can be as much of aninfluence (independent variable) on political and private actors—through the political and economic feedback they generate—as theyare an outcome of them (dependent variable).15 The goal is to try toseparate the specific order of cause and effect, because sequenceanalysis is critical for causal analysis.16 A parallel goal with this typeof inquiry is to try to account for how individuals and institutionsrespond to changes in public policy, recognizing that governmentreforms often reconfigure incentives other than those originally in-tended. According to Pierson, “research on policy feedback has

RICK MAYES 147

stressed two arguments: that policy structures create resources andincentives that influence the formation and activity of social groups,and that policies affect processes of ‘social learning’ among majorpolitical actors.”17 For example, Hacker has shown that the parallelgrowth of public welfare programs (e.g., Social Security) and privatewelfare (e.g., health insurance) is a classic example of how “privatesocial benefits have ‘policy feedback’ effects that are not all thatdifferent from the policy feedback effects that are created by publicsocial programs.”18 In both instances, “major public policies consti-tute important rules, influencing the allocation of economic andpolitical resources, modifying the costs and benefits associated withalternative political strategies, and consequently altering ensuingpolitical development.”19

The paradigm shift from traditional fee-for-service health in-surance to managed care is a good example of how policy feedback,within a domino-like sequence of events, can result in a completelyunanticipated consequence. In this instance, the entire chain canbe summarized as the shifting of costs from one actor in the U.S.medical industrial complex to another—from the government to thehospital industry to privately insured patients and their employers.Congress’s seminal and financially necessary change to Medicare’sreimbursement scheme in 1983 (domino one) allowed Congress tosystematically reduce spending on hospital care in the latter half ofthe 1980s (domino two) in response to massive budget deficits. Hos-pitals reacted to reduced Medicare funding by increasing their costshifting to privately insured patients (domino three). By its very defi-nition, cost shifting was simply passed along the payment chain andcontributed significantly to large annual increases in private insur-ance premiums. Responding to the growing imperative for cost con-trol, employers logically switched more and more of their employeesinto much less popular—but also less expensive—managed-care plans(final domino). The remainder of the article is devoted to explainingthis linkage and the mechanisms that fostered it.

Origins of Prospective Payment: Trust Funds andMedicare’s “Crisis” Politics of Bankruptcy

As Theodore Marmor, Eric Patashnik, Jonathan Oberlander, JulianZelizer, and others have shown, the politics of Medicare policymakinghave often been waged under the auspices of “crisis-oriented” con-


cerns over the solvency of the program’s trust funds.20 When com-bined with even larger concerns over federal budget deficits thatemerged in the mid- to late 1980s, Marmor notes, Medicare was foundto be uniquely vulnerable to major programmatic change in a waythat its companion program in Social Security, Old Age and Survi-vors Insurance (OASI), never was.21

The structure of Medicare’s financing with its two trust funds iscentral to the program’s direct effect on the U.S. federal budget.When Congress passed Medicare in 1965, adding it to Social Secu-rity, the public health insurance program was comprised of two partswith separate financing arrangements. Part A, the Hospital Insur-ance (HI) Trust Fund, pays for beneficiaries’ hospital costs. It is fi-nanced from a 2.9 percent payroll tax. Part B, the SupplementaryMedical Insurance (SMI) Trust Fund, pays for beneficiaries’ physi-cian and outpatient expenses. It is financed by general tax revenuesand premiums paid by Medicare beneficiaries. Because Part A is fi-nanced by a payroll tax, it can conceivably go bankrupt by payingout more in expenditures than it receives in tax revenue. Part B,however, is immune to such threats (for all intents and purposes)because its partial funding from general tax revenues operates as an“open pipeline” to the Federal Treasury.

Scholars disagree over whether policymakers, particularly Waysand Means Chairman Wilbur Mills, designed Medicare to be insu-lated from regular political debate or, rather, to encourage it. Marmorand Oberlander argue that the “bankruptcy crises” that have repeat-edly erupted over Medicare are a perverse outcome, unintended bythose who designed the program to be a vehicle for smoothly andeffectively achieving national health insurance via incrementalsteps.22 Conversely, Patashnik and Zelizer see a certain institutionallogic to Medicare’s design that, they argue, has “served a valuablesocial purpose by periodically forcing policymakers to engage in ahealthy examination of one of the nation’s largest and most expen-sive social programs.”23

Either way, financial problems with Medicare arose soon afterthe program began operation. According to a report submitted tothe Senate Finance Committee in the spring of 1966, the system forpaying hospitals “contains no incentives whatsoever for good man-agement and almost begs for poor management.”24 Robert Ball, com-missioner of the Social Security Administration during Medicare’sdevelopment and implementation, agreed: “After-the-fact reimburse-ment for hospital costs clearly was flawed, and within a couple of

RICK MAYES 149

years I and other government officials were calling for some form ofprospective payment.”25

The core of Medicare’s problem stemmed from its lack of cost-containment incentives; hospitals were neither penalized for costincreases nor rewarded for finding ways to control them.26 “Medi-care gave hospitals a license to spend,” notes Rosemary Stevens. “Themore expenditures they incurred, the more income they received.Medicare tax funds flowed into hospitals in a golden stream, morethan doubling between 1970 and 1975, and doubling again by1980.”27 Medicare’s formula for hospital reimbursement invitedabuse, because it operated on a “cost+ 2 percent basis” for all ser-vices. Since the 2 percent was a percentage of costs (and added byCongress to reflect the added nursing costs for Medicare patients),28

it amounted to an open-ended proposition by offering a small bonusfor every cost increase. Consequently, Oberlander explains, Medi-care “quickly acquired a reputation, as chairman of the Senate Fi-nance Committee Russell Long put it, as a ‘runaway program,’ animage only reinforced by much higher than expected costs in thekidney dialysis benefit added to Medicare in 1972.”29

First Domino: Development, Passage, and Phase-in ofMedicare’s New Hospital Payment System

By the advent of President Ronald Reagan’s first year in office in1981, Medicare was predicted to go bankrupt by as early as 1987 or1988.30 Interfund borrowing from Medicare’s HI trust fund ($12.4billion) and the Disability trust fund ($5.1 billion) to the OASI trustfund exacerbated an already deteriorating financial situation forMedicare.31 Moreover, the structural concessions that policymakershad made to Medicare’s design in 1965, so the program could finallyovercome the AMA’s political opposition, led to a very lucrative butultimately unsustainable system for paying hospitals and doctors.32

Ironically, Reagan’s new Republican administration, with itsideological emphasis on pro-market policies and downsizing the fed-eral government, created a unique political context for a Medicarereform proposal that involved increased government regulation. AsOberlander points out, “fiscal exigency simply overwhelmed ideol-ogy. . . . Given the administration’s short-term goals for reducingdomestic spending, a market approach to Medicare reform was not


viable.” Thus, “federal regulatory authority over medical providersconsequently had to be strengthened.”33

In rationalizing Medicare’s hospital reimbursement scheme,policy learning came by way of applied federalism. In the 1972 So-cial Security Amendments, Congress authorized the Department ofHealth and Human Services (HHS) to conduct statewide experi-ments with different forms of hospital reimbursement.34 By 1982, itwas monitoring nine individual state experiments. One in New Jer-sey looked particularly promising with its novel use of diagnosis-related groups (DRGs), designed in the early 1970s by Robert Fetterand John Thompson at Yale University.35 The definitive report onNew Jersey’s experience began by explaining that DRGs grew out ofacademic “efforts in the 1970s to define a hospital’s product and theuse and costs of resources essential to produce it.”36

There were other ambitious state experiments in hospital pay-ment reform in New York, Massachusetts, Connecticut, Rhode Is-land, Maryland, and Washington.37 But President Reagan’s newSecretary of HHS, Richard Schweiker, came from Pennsylvania and“religiously summered at the Jersey shore,” according to RobertRubin, assistant secretary of Planning and Evaluation at HHS from1981 to 1984, who was principally involved in the political negotia-tions between the administration and key members of Congress overDRGs.38 “Being in the health care field as a Senator and Congress-man for twenty years, Schweiker was well known, so it wasn’t un-usual for him to hear about these kinds of things,” adds Rubin.“Actually there were two books on DRGs; he had carefully read bothof them and had underlined them. He and I talked about his ques-tions at some length and he became convinced that DRGs made themost sense.”39 Jack Owen, vice president of the American HospitalAssociation (AHA) beginning in 1982, previously had run the NewJersey Hospital Association for twenty years and was instrumentalin securing the cooperation of New Jersey hospitals in the state’sexperiment with DRGs. As vice president of the AHA, he urgedSecretary Schweiker to adopt New Jersey’s innovative form of reim-bursement and then persuaded Senator Bob Dole that the AHAwould support the move to a prospective payment system.40

Building on New Jersey’s model, albeit without a formal evalu-ation indicating whether or not the state’s experiment worked,41 thefoundation for Medicare’s new prospective reimbursement methodwas DRGs. As a patient classification system, DRGs sorted patientsinto groups according to medical condition (Table 1). Medicare’s

RICK MAYES 151

payment for any specific DRG, Louise Russell explains, “is the samefor every patient in a given group, regardless of how long the patientstays in the hospital or what else is done during the stay.”42 Thecrucial features of Medicare’s prospective payment system “are thatpayment is prospective—rates are set before services are delivered—and that a single lump-sum rate pays for the entire hospital stay. . . .If the hospital can take care of the patient for less than the fixedrate, it keeps the profit. If not, it absorbs the loss.”43

Changing Medicare payment from a retrospective to a prospec-tive system was revolutionary.44 Previously, doctors and hospitalsoften charged patients different rates for the same procedure basedon their ability to pay or how long they stayed. This meant that thesame hospital procedure would often cost Medicare twice or eventhree times as much in one location as compared to another. “Thecost reimbursement model from before 1983 was insanity. On theface of it, it encouraged people to do more, it paid them to do moreand not in any particularly rational way,” according to Sheila Burke,a key staff member on the Senate Finance Committee at the timeand Senator Bob Dole’s chief of staff. “Going to DRGs, therefore,had all the right things going for it politically and conceptually. . . .In effect, you could say to the average member of Congress, whotended to not want to get into the minutiae of Medicare policy be-

Table 1. Selected Diagnosis-Related Groups, 1988

DRGNumber Title 106 Coronary bypass with cardiac catheterization 127 Health failure and shock 176 Complicated peptic ulcer 236 Fractures of hip and pelvis 317 Admit for renal [kidney] dialysis 433 Alcohol or drug abuse or dependence, left hospital

against medical advice 470 Ungroupable 474 Tracheostomy

Source: Louise Russell, Medicare’s New Hospital Payment System: Is It Work-ing? (Washington, D.C.: The Brookings Institution, 1989), 10.


cause it was one of the more boring aspects of their lives, ‘Why shouldit [a particular hospital service or procedure] cost anything differentbetween L.A. and San Francisco or San Francisco and Chicago, orChicago and Detroit?”45

The government’s fiscal priorities so dominated the develop-ment of Medicare’s PPS in 1982 and 1983 that there existed littleinterest-group influence or congressional and media debate. Mostmembers of Congress did not understand exactly how the PPSworked.46 They voted for it, however, because Medicare was ap-proaching insolvency and because congressional leaders piggy-backedthe plan onto even more vital Social Security legislation that had topass for monthly OASI checks to continue uninterrupted.47 Attach-ing Medicare reform to critical Social Security reform was a purelyopportunistic, but effective, decision.48 A veto-proof bill emerged,largely immune from interest-group influence due to its sheer urgency.49

The PPS’s four-year phased-in approach had profound effectson hospital administration. In order to cushion their transition fromtraditional cost-reimbursement to a prospective system, first-yearDRG payments were based on each hospital’s historical costs. Hos-pitals quickly reduced their patients’ average length-of-stay50 and inthe process, according to Robert Coulam and Gary Gaumer, reapedhuge windfall payments: “Widely conceded ‘overpayment’ in the firstyear of PPS created a situation in which margins were increasing asexpenses per case were dropping, due to large reductions in lengthof stay. This not only made the first year a somewhat aberrant inter-vention, but also armed most hospitals with an unanticipated sourceof disposable funds, and probably altered expectations as well.”51

Hospital administrators transformed their medical records de-partments—where accurate coding of patient records determined howmuch hospitals got paid or whether they got paid at all—with morepersonnel and improved technology. The cliché of choice became“PPS brought medical records out of the basement.”52 In addition,hospitals that had a teaching mission or served a disproportionateshare of poor patients successfully persuaded Congress to have Medi-care pay them more generous DRG rates because of their specialstatus.53 Significant change ensued. The Medicare hospital paymentreforms “were the most drastic and far-reaching changes in Federalhealth policy since the passage of Medicare itself,” notes David G.Smith.54 In 1984, Michael Bromberg, executive director of the Fed-eration of American Hospitals (FAH), said as much in his testimonybefore Congress: “The Medicare law that brought us prospective

RICK MAYES 153

payment for the first time has clearly given us incentives 180 de-grees different from any we have ever had, and we have responded.”55

In 1985, Bromberg reiterated his claim that Medicare’s “ProspectivePayment System is the most effective cost containment program everenacted, successful beyond anyone’s expectations.”56

Everyone was initially pleased with Medicare’s PPS. Medicare’srate of expenditure growth slowed dramatically in 1985, with Part Apayments to hospitals providing the bulk of the program’s reduc-tions (Table 2). At the same time, hospitals profited handsomely.Their positive Medicare margins—which reflected the total amountof Medicare inpatient payments they received relative to the totalinpatient costs they incurred treating Medicare patients—were al-most 15 percent in 1984 and 1985 (Fig. 1). Such large MedicarePPS margins helped to offset hospitals’ regular losses on both Med-icaid and charity care patients, which left them with an average over-all profit margin of slightly more than 5 percent (Fig. 1).

For the first time ever, though, Medicare’s new method of reim-bursement separated hospitals into financial “winners” and “losers.”57

Each year’s average Medicare PPS margin masked an enormousamount of variation around the mean. Even with a positive overallaverage of almost 15 percent Medicare PPS margins in the early years,hundreds of hospitals had significantly higher margins. Meanwhile,there were hundreds of hospitals that either were so inefficient orhad such an unpredictable mix of Medicare cases (often small ruralhospitals) that they still managed to lose money on their Medicarepatients (see Table 3).

Ultimately, Medicare’s new PPS was a huge but not an immedi-ate change for the hospital industry. The four-year phase-in periodallowed hospitals to make minor adjustments and technological cod-ing improvements, which significantly increased most hospitals’ fi-nancial margins and created the term “DRG creep.”58 The vastmajority of hospitals found themselves much better off financiallyduring the early years of Medicare’s new PPS than they were undercost reimbursement.59 But the AHA’s Jack Owen knew that the goodfinancial times would not last. He believed that Congress would cometo view hospitals’ sizable profits as potential budgetary savings: “Itold my member hospitals to put their money in the bank. . . . ‘Itwon’t continue,’ I said. ‘You’re going to get reduced.’”60


Tab

le 2

. Tot

al M

edic

are

and

Nat

iona

l Hea

lth E

xpen

ditu

res

(in b

illio

ns),

198

0�87

Yea

r

Nat

iona

l

All

Med

icar

eH

ealt

hPe

rcen

tage

Med

icar

e P

erce

ntag

eH

ospi

tal

Perc

enta

geEx

pend

itur

esC

hang

eEx

pend

itur

es C

hang

ePa

ymen

tsC

hang

e

198

0$2

49.1

—$3

6.4

—$2

5.4

— 1

981

288.

615

.9%

43.7

20.0

%30

.620

.3%

198

232

3.8

12.2

51.2

17.3

35.7

16.5

198

335

6.1

10.0

58.1

13.5

39.9

11.8

198

438

7.0

8.7

64.8

11.5

44.5

11.7

198

5^42

0.1

8.5

69.8

7.8

47.1

5.7

198

645

2.3

7.7

75.8

8.5

49.2

4.6

198

7 4

92.5

8.9

82.0

8.2

51.3

4.2

’80–

83—

12.7

*—

15.0

*—

16.2

*’8

4–87

—8.

5*—

9.0*

—6.

5*

Sour

ce:

ProP

AC

, Med

ica r

e a n

d th

e N

a tio

nal H

ealth

Ca r

e Sy

stem

, R

epo r

t to

th e

Co n

gres

s (W

a shi

ngto

n, D

.C.:

June

199

1), 1

2, 1

11.

* A

nnua

l ave

rag e

inc r

e ase

ove

r th

e fo

ur-y

e ar

peri

od^

Fir

st fu

ll y

e ar

of M

edic

a re ’

s PP

S in

ope

rati

on

RICK MAYES 155

Second Domino: Medicare Policy’s IncreasingSubordination to Budget Policy

The mutual admiration between Congress and the hospital industryover the success of the PPS deteriorated when Congress turned toMedicare in 1986 as a means of addressing the nation’s growing bud-get deficits.61 The same Michael Bromberg, who just a year earlierhad effusively praised federal policymakers, now accused Congressand the Reagan administration of operating in “bad faith” and vio-lating the PPS “contract.”62 Hospitals started withholding requestedfinancial information from Congress concerning their finances, par-ticularly their overall margins. They concluded that politicians only

Source: ProPAC, Medicare and the American Health Care System, Report to theCongress (Washington, D.C., June 1995 and 1996), 55 and 68, respectively.

* PPS Margin = (hospitals’ total inpatient Medicare payments – hospitals’ totalinpatient Medicare costs) ÷ hospitals’ total inpatient Medicare payments

Hospitals' Inpatient Medicare (PPS) Margin and Overall Margin,* 1984-92

-5

0

5

10

15

84 85 86 87 88 89 90 91 92

Year

PPS Margin

Overall Margin

Fig 1. Hospital�s inpatient Medicare (PPS) margin and overall margin,* 1984�92.

Perc

ent

Year


1984

1985

1986

1987

1988

1989

1990

1991

1992

14.5

14.

0 9

.5 6

.6 3

.9 1

.4 -

.05

-2.

4 -

1.0

16.8

18.

8 3

2.3

39.

8 4

6.1

51.

8 5

6.7

60.

8 6

0.0

Perc

enta

ge o

fH

ospi

tals

Los

ing

Mon

ey o

n th

eir

Med

icar

e Po

pula

tion

Tab

le 3

. Hos

pita

ls� I

npat

ient

Med

icar

e (P

PS

) �P

rofit

� M

argi

n* a

nd P

erce

ntag

e of

Hos

pita

ls w

ith O

vera

ll M

edic

are

Loss

es, 1

984�

92

Hos

pita

ls’ I

npat

ient

Med

icar

e (P

PS)

Mar

gin*

Sour

ce:

ProP

AC

, Med

ica r

e a n

d th

e A

mer

ica n

Hea

lth C

a re

Syst

em,

Rep

o rt t

o th

e C

o ngr

ess

(Wa s

hing

ton,

D.C

., Ju

ne 1

994

and

1996

), 5

8 a n

d 68

.*

PP

S M

a rg i

n =

(h

ospi

tals

’ to

tal

inpa

tie n

t M

edic

a re

paym

ents

– h

ospi

tals

’ to

tal

inpa

tie n

t M

edic

a re

c ost

s) ÷

hosp

ita l

s’ t

ota l

inpa

tie n

t M

edic

a re

paym

ents

RICK MAYES 157

meant to use the information to justify further Medicare rate reduc-tions. By 1988, members of Congress and the hospital industry wereopenly trading angry accusations of lying, fraud, and deceit.63

Congress ignored the hospital industry’s complaints and annu-ally reduced the rate of increase to Medicare’s DRG payments. Ac-cording to Bill Gradison, the ranking Republican on the House Waysand Means Subcommittee on Health until 1993, the key to Congress’sability to extract huge savings from Medicare was the budget recon-ciliation process.64 Leon Panetta, chair of the House Budget Com-mittee in the 1980s, observed that the reconciliation process “scaredthe hell out of” the hospital and other industries.65 Moreover, “pro-viders, particularly hospitals, were always viewed by Congress as aneasier target than doing anything that would have ever affectedMedicare beneficiaries,” according to Rick Pollack, executive vicepresident of the AHA. “I don’t know how many people on the Hillwould be up front in admitting this, but they would sort of have ahole in the budget target to reconciliation . . . and they’d save thePPS update factor to be the last thing to be determined and say, ‘Ok,we gotta save a billion bucks over three years, so let’s just make thistweak to Medicare’s payment system or that tweak.’ At the end ofthe day, it was legislated in the back rooms and it was all a budgetnumber.”66

Many policymakers, including senior staff and members of Con-gress, have admitted to using prospective payment for larger budget-ary purposes. According to Lisa Potetz, senior hospital analyst atProPAC from 1984 to 1989 and senior Medicare analyst on the Sen-ate Finance and House Ways and Means Committees from 1989 to1995, congressional leaders came to view prospective payment as avaluable and effective tool for reducing the deficit.67

Adjusting hospital payment rates as part of the budget recon-ciliation process had a noticeable impact on Medicare’s financialcondition. According to the CBO, it reduced the “growth rate ofreal [Part A] spending from 5.4 percent annually between 1980 and1985 to just 1 percent annually between 1985 and 1990.”68 AsPatashnik notes, “This extended the HI Trust Fund’s projected dateof exhaustion from 1991 in the 1981 Trustees report to 2005 in the1991 report.”69 Robert Reischauer, CBO director from 1989 to 1995,explains why it was so attractive (politically and fiscally) for Con-gress to manipulate Medicare payment policy for larger budgetarypurposes: “Medicare was the cash cow! There is a very simple reasonfor this and that is that Congress could get credited for deficit re-


duction without directly imposing a sacrifice on the public. . . . Andto the extent that the reduction actually led to a true reduction inMedicare services, it would be difficult to trace back to the Medi-care program or to political decision-makers.”70

Medicare’s new payment system successfully restrained theprogram’s rate of expenditure growth. “Though Medicare’s cost sav-ings may not have been impressive on an international scale,”Oberlander explains, “compared to the inflationary American pri-vate insurance market they were downright remarkable.”71 One re-sult of Medicare’s major cost savings, however, was that more andmore hospitals lost money on their Medicare patients (particularlythose with complicated diagnoses), largely because they did not re-strain their cost growth. Hospitals’ costs-per-case increased at anaverage annual rate of 8.6 percent between 1986 and 1992, morethan twice the rate of general inflation.72 With increasing financialpressures, hospitals began to include operational efficiency measuresand program closures to try to save money at the margins. But morethan anything else, according to Stuart Altman, former chair ofMedicare’s Prospective Payment Assessment Commission (ProPAC),hospitals felt pressure to increase revenue by cost shifting to privatepayers.73

Third Domino: Hospitals’ Increased Cost Shifting toPrivately Insured Patients

By the late 1980s, the majority of hospitals were losing money ontheir Medicare patients (Table 3).74 According to ProPAC, whilehospitals’ overall cost growth returned to its “historical rate through-out the remainder of the 1980s, Medicare’s PPS margin steadily fell,dropping below zero in 1990 and to -2.4 percent in 1991.”75 AsAltman suggests, hospitals responded largely by turning to privatelyinsured patients to make up for these losses, as well as for an increas-ing share of their Medicaid losses and unreimbursed charity care(Table 4).76

Both the business community and commercial insurers had beenaware of cost shifting long before the PPS’s implementation,77 butthis form of cross-subsidization had traditionally remained modestenough to avoid open conflict.78 From 1984 to 1993, however, theaverage annual increase in the per capita cost of private health in-surance for medical services was 22.7 percent more than the rate of

RICK MAYES 159

Table 4. Hospitals� Overall Payment-to-Cost Ratios* by Payer, 1980�92

Year Total Medicare Total Medicaid Total Private Payer

1980 0.96 0.91 1.121981 0.97 0.93 1.121982 0.96 0.91 1.141983 0.97 0.92 1.161984 0.98 0.88 1.161985 1.01 0.90 1.161986 1.01 0.88 1.161987 0.98 0.83 1.201988 0.94 0.80 1.221989 0.91 0.76 1.221990 0.89 0.80 1.271991 0.88 0.82 1.301992 0.89 0.91 1.31

Source: S. Guterman, J. Ashby, and T. Greene, “Hospital Cost GrowthDown,” Health Affairs 15 (Fall 1996): 137, Exhibit 4.

* Medicare’s payment-to-cost ratio in this table contrasts with the PPS mar-gins in Table 3 and Figure 1. This is attributable to the broader scope ofthe payment-to-cost ratio, which reflects payments and costs for all Medi-care services (inpatient and outpatient acute care, medical educationprograms, and hospital-based post-acute care). Payments for outpatientservices, medical education, and post-acute tend to be below reportedcosts because of the use of fee schedules, discounts from cost-based rates,and payment limits. In addition, the payment-to-cost ratio reflectsMedicare’s share of all hospital costs, whereas the PPS margin is calcu-lated using only Medicare-allowable costs, which are believed to be 3–5percent lower.124 See Guterman, Ashby, and Greene, “Hospital CostGrowth Down,” 139 n. 14.

increase in the per capita cost of Medicare beneficiaries for the sameservices.79 Private payers’ payment-to-cost ratio peaked at 131 per-cent in 1992 (see Table 4). And smaller businesses were particularlyvulnerable to the negative effects of cost shifting because they werefar less able to obtain strong bargaining positions in negotiating theirhealth insurance contracts.80

In more competitive markets without a large government pres-ence, David Drake argues, cost shifting of this magnitude would not


and could not occur.81 But as current CMS administrator and formerpresident of the FAH, Tom Scully, observes, medical providers aresecond only to defense contractors in their dependence on govern-ment payments, which provide approximately 40 percent of theirtotal revenues.82 In addition, hospitals basically treat all patientsalike. Hence, it is difficult, if not impossible, for them to separatepublic and private patients into various parts of the hospital thatmight have different cost structures.83

Evidence of hospitals’ extensive use of cost shifting even camein the form of confession. In a written reply to a series of questionsposed by the Senate Labor and Human Resources Committee, theAmerican Hospital Association admitted that hospitals routinelyshifted some of their costs to privately insured patients, who thenpaid inflated bills.84 James Mongan—currently CEO of PartnersHealthCare in Boston and formerly president of Massachusetts Gen-eral Hospital, senior staff member of the Senate Finance Commit-tee, and deputy assistant secretary for Health in the Carteradministration—argues that hospitals have to cost shift to privatepayers or risk bankruptcy.85 Michael Bromberg of the FAH, whichrepresents the nation’s investor-owned, for-profit hospitals, admit-ted that hospitals regularly increase charges to private patients tocompensate for reduced reimbursement for public patients.86 AndRick Pollack, senior vice president of the AHA, argues that costshifting was standard operating procedure for most hospitals untilmanaged care made it increasingly difficult by negotiating signifi-cant discounts with medical providers.87

With the growth of cost shifting and the directly associated in-crease in insurance premiums,88 businesses concluded that their ben-efit-cost ratio for involvement in health insurance had fundamentallychanged.89 The problem of escalating private insurance premiumswas not new, but cost shifting greatly exacerbated it and contributedsignificantly to unprecedented annual premium increases—often inexcess of 20 percent—in the late 1980s (Fig. 2).90 Hewitt Associ-ates, a benefits consulting firm, identified cost shifting as the singleleading source of health plan premium increases in 1987 and 1988.91

Similar studies in the early 1990s found cost shifting to be the singlelargest factor in the rise of private health insurance premiums, morethan increased utilization, technology improvements, and deduct-ible erosion combined.92 According to Rashi Fein, “as Medicare (andMedicaid) tightened their reimbursement policies, they paid hospi-tals less than the hospitals believed was a fair share of total hospital

RICK MAYES 161

expenses. Hospitals reacted by increasing charges to other payers,especially to commercial insurance carriers, in order to cover theshortfall in total receipts. In turn, private insurers had to raise theirpremiums in order to, as they would put it, ‘subsidize’ patient careonly partly paid for by government.”93

Economists debate the technical dynamics of cost shifting,94 butarguably what is most important about the concept is the extent towhich employers believed it largely explained their rapidly increas-ing health-care costs and then subsequently made their health in-surance decisions based on their beliefs. Due in part to ProPAC’sreports on Medicare in the late 1980s and early 1990s, which re-peatedly maintained that the phenomenon was large and growing,

0

5

10

15

20

25

30

1984 1985 1986 1987 1988 1989

Year

PRate of

Change in

Premium

Sources: KPMG and A. Foster Higgins and Co., Inc., Survey of Employer-Sponsored Health Plans, 1981–1991; Health Insurance Association ofAmerica surveys.

Fig 2. Average annual percent increase in fee-for-service health insurancepremiums, 1984�89.


cost shifting became the dominant explanation among employersfor the rapid increase in private health insurance premiums.95 Rep-resentative John Dingell (D-Mich.) even initiated hearings in 1991before the House Energy and Commerce Committee to investigateHumana’s and other hospitals’ “controversial practice of cost shift-ing.”96 David McFadden argued that the notorious “$7 aspirin,” likethe $200 military toilet seats from the decade before, had becomethe infamous symbol of cost shifting.97 Malcolm Gladwell, a leadinghealth-policy journalist at the time, maintained that cost shiftingwas the biggest reason for the rapid increase in private health insur-ance costs.98

Employers’ concerns over cost shifting peaked following thepublication of an influential report to the Healthcare FinancialManagement Association in the summer of 1992 by Lewin/ICF, apolicy consulting firm headed by Allen Dobson, formerly the direc-tor of HCFA’s Office of Research (1981–88). In the Lewin/ICF re-port, rising health-care costs “were being allocated unevenly becausesome stakeholders are better than others at insulating themselvesfrom paying their fair share of the costs,” according to Dobson andHFMA president, Richard Clarke. “Stakeholders with significantpurchasing power and those who purchase coordinated care are mov-ing to protect themselves from what they believe is an untenablesituation, leaving others, particularly small business and non-grouppurchasers, to fend for themselves. An important aspect of this in-terplay of stakeholders is the cost-shifting phenomenon.”99 As Dob-son and Clarke argued, businesses and employers were essentiallypaying a “sick tax” to cover the additional costs that providers wereshifting to them.100

As chairman of the Prudential Insurance Company and head ofthe Business Roundtable’s Health Care Task Force between 1988and 1994, Robert Winters had a unique vantage point from whichto observe cost shifting, given his position as the head of a majorcompany that sold health insurance policies. The BusinessRoundtable is an association of chief executive officers of thecountry’s biggest companies with a combined workforce of more than10 million employees.101 Thus, the extent to which its members be-lieved that cost shifting was primarily to blame for the nation’s rap-idly increasing private health-care costs reflects how widespread theexplanation had become throughout the entire U.S. business com-munity. According to Winters, “What happened in the late 1980sand in the early 1990s, was that health care costs became such a

RICK MAYES 163

significant part of corporate budgets that they attracted the very sig-nificant scrutiny of CEOs. . . . More and more CEO’s [were] saying,‘Goddammit, this has to stop!’ What was particularly attracting theirattention was costs, and they very quickly got animated by their rec-ognition of cost shifting.”102

Ultimately, America’s health-care system stumbled when double-digit increases in health insurance premiums coincided with the re-cession of the early 1990s. As Uwe Reinhardt notes, “Eventually,the increasingly desperate American employers began to reevaluatethe open-ended social contract they had written and supported forso long, and they looked around for an alternative deal. That dealwas known as ‘managed care.’”103

Final Domino: Employers’ Shift to Managed Care

There was a pronounced change in the health delivery system in theU.S. that began in the late 1980s. Cost control in the public sectorwith Medicare reform contributed significantly to medical inflationin the private sector, which triggered the private sector’s response: amassive switch to managed care (Table 5).104

“Once employers discovered the cost-saving potential of themanaged care system,” note Karen Titlow and Ezekiel Emanuel, “theyrapidly turned away from traditional indemnity plans. This trendwas spearheaded by Allied Signal Inc., which in 1988 moved all itsemployees from indemnity insurance into a Cigna health mainte-nance organization (HMO). By 1991, Allied Signal demonstratedthe cost-saving potential of managed care when it reported a 23 per-cent cut in health insurance expenditures.”105 Fein adds that “as thenumber of HMOs grew, employers discovered that just as they nego-tiated the price of steel, paper, or other ‘inputs’ (including labor),they could negotiate prices for health insurance. This became ad-vantageous when, as a result of the ability of HMOs and capitatedplans to control physician behavior, restrict expensive hospital uti-lization, and limit patient choice of providers, competing managed-care delivery/insurance organizations were often able to offeremployers premiums substantially lower than those available fromtraditional indemnity plans.”106

The initial shift to managed care in the late 1980s and early1990s had a self-reinforcing quality to it that fed back into the mo-mentum away from fee-for-service insurance. Managed-care organi-


zations initially attracted and enrolled low-risk individuals who wereleast likely to object to restrictions on utilization of services andphysician choice.107 These low-risk individuals also tended to behealthier than the general population, so they did not increase op-erating costs; on the contrary, they increased the profitability ofmanaged-care organizations. So although the rates of change inhealth insurance premiums generally moved in tandem, premiumsfor fee-for-service indemnity insurance grew substantially more thanmanaged-care premiums between 1986 and 1991.108 Moreover, ac-cording to Mark Pauly and Sean Nicholson, the private health in-surance market in the post–World War II era up to the mid-1980sessentially had been a “pooling equilibrium” in which the risk ofcovering an individual patient’s medical costs was spread out over alarge pool of individuals who all paid generally the same, commu-nity-rated insurance premiums: “Over 90 percent of employees hadindemnity insurance, mostly with Blue Cross and Blue Shield, whichexperience rated only reluctantly. This pooling equilibrium unrav-eled between 1984 and the early 1990s when [managed care] qua-drupled its share of the large employer market.”109

Employers’ shifting of their workers away from fee-for-servicehealth insurance was further facilitated by the maturation and im-proved infrastructure of the managed-care industry by the late1980s.110 Between 1987 and 1993, in particular, managed-care orga-nizations responded to employers’ demands for more cost control byconsolidating and applying extensive utilization review and guide-line development to their more traditional fee-for-service insuranceofferings.111 The traditional managed-care organizations, such as staff-or group-model HMOs (e.g., Kaiser Permanente), required signifi-cant expenditures in “bricks and mortar” when entering new mar-

Table 5. Enrollment in Indemnity Insurance and Managed Care, 1989�95

Type of Coverage 1989 1993 1995

Indemnity (fee-for-service) 71% 49% 30%Managed Care (HMO, PPO) 29% 51% 70%

Source: Employee Benefit Research Institute, Sources of Health Insurance(Washington, D.C., February 1995).

RICK MAYES 165

kets. This served as a major barrier to entry because they were verti-cally integrated organizations that operated their own physical fa-cilities in different geographic locations, and whose physiciansworked solely for the managed-care organization.112 But beginningin the late 1980s, many new for-profit HMOs experienced rapidgrowth because they were “virtual organizations” or “organizationswithout walls,” built largely on contractual (paper) relationships withcommunity providers.113

By expanding their provider base and involving in their systemsmore physicians whose predominant practice was fee-for-service,managed-care organizations developed to the point that employerstook them more seriously and found them significantly more attrac-tive.114 Why? Because by increasing their number of affiliated medi-cal providers, managed-care organizations essentially became moreeffective “managed cost” plans, which could negotiate lower priceson behalf of larger numbers of patients and then pass the savings onto employers. Before this balance-of-power shifting to payers in theearly 1990s, providers had set prices and determined fees in mostmarkets.115

The changes so far described could have been predicates for thedevelopment of a single-payer health-care system similar to othercountries’ approaches to providing medical care (see Joseph White’sCompeting Solutions for elaboration on this point).116 The privatesector could have responded to increased cost shifting the way manyleaders from the employer community began to respond in the early1990s—in favor of national health insurance and a greater role forgovernment financing.117 President Bill Clinton’s proposal for com-prehensive health-care reform in the fall of 1993 represented a ma-jor political attempt to capitalize on just such sentiments. His electionin 1992, after twelve years of Republican control of the presidency,dramatically changed the political context for consideration of com-prehensive health-care reform. As Jacob Hacker notes, “For a briefmoment in the early 1990s, the strains on public programs and theerosion of private benefits shared the spotlight, as President Clintonsought to tackle the problems in American health insurance by put-ting in place the biggest missing piece of the American welfare state.The resounding failure of the Clinton health plan demonstrated notjust the fiscal barriers such efforts face, but also the powerful ongo-ing hold of antigovernment ideas and interests in American poli-tics—the last of the intertwined pressures that have placed thewelfare state under siege.”118


The death knell for the Clinton proposal, according to SallyannePayton (who was legal counsel to the Clinton White House for healthcare and a member of the Clinton Health Care Reform Task Force),came when top officials of the largest corporations—whose healthand benefit officers earlier had been supportive of Clinton’s efforts—did their own cost-benefit analysis of what would happen if Clinton’scomprehensive insurance reform plan went into effect. They con-cluded it would cost them more than they would gain.119 In effect,staying with managed care or shifting to it made more sense thanchanging to national health insurance.

The switch to managed care did succeed in significantly reduc-ing the annual increase in health insurance premiums during thefirst half of the 1990s. But overall cost control proved to be rela-tively short-lived as double-digit annual increases in health insur-ance premiums returned in the late 1990s (Fig. 3).120 What did notprove to be short-lived was the enormous public backlash againstmanaged care, which became one of the country’s most hated indus-tries.121

Managed-care executives essentially became the “flak catchers”for rationing medical care in much the same way that Clinton pro-posed government would do in his managed-competition plan. Op-ponents of Clinton’s proposal complained loudly at the time thathis plan would force government rationing of medical care. Howironic it was, then, as Uwe Reinhardt observes, that “only half adecade after embracing the idea of ‘managed competition with man-aged care,’ America’s ‘rugged individualists’ [began] to show theirmore tender side, as self-pitifully and pitiably they pleaded with theWhite house, with the Congress, with their state governments andwith the courts to jump right back onto their backs, to protect themfrom the forces of the private markets that, in their more ruggedmoments, they had professed to adore.”122

Conclusion

The reform of Medicare’s payment policy in 1983 appears to havebeen the initial catalyst that triggered a series of interconnectedevents resulting in an unintended consequence: the managed-carerevolution. Congress’s changing of Medicare to a prospective methodof hospital reimbursement in 1983 proved effective in slowing theprogram’s rapid rate of cost increase.123 But as an unintended conse-

RICK MAYES 167

02468

101214161820

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Yea

r

Pe

Rat

e o

f

CP

I

Incre

ase

Rat

e o

f

Ch

ang

e in

Pre

miu

ms

Fig

. 3. A

vera

ge a

nnua

l rat

e of

incr

ease

in a

ll he

alth

insu

ranc

e pr

emiu

ms,

198

9�20

00.

Sour

c es:

Ad a

p ted

and

mo d

ified

fro

m M

erce

r/Fo

ster

Hig

gins

; U

.S.

Bur

eau

o f L

a bo r

Sta

tistic

s, N

a tio

n’s

Bus

ine s

s (J

uly

1998

), 2

0;H

ealth

Ins

ura n

ce A

sso c

iatio

n o f

Am

eric

a /K

PMG

Pea

t Ma r

wic

k Su

rve y

, Wa s

hing

ton

Post

(16

Jun

e 19

98),

C3;

Ann

ual C

h ang

e in

CPI

fro

m t

h e U

.S.

Dep

t. o

f La

b or,

Bur

eau

o f L

a bo r

Sta

tistic

s; M

e rc e

r-H

igg i

ns S

urve

y of

Em

ploy

e rs

Spon

sore

d H

e alt

h Pl

a ns,

1999

;a n

d Jo

el M

iller

, “D

é jà

All

Ove

r A

g ain

,” N

a tio

nal C

o alit

ion

o n H

ealth

Ca r

e(W

a sh i

ngto

n, D

.C.,

Ma y

200

0),

ii, 1

.


quence, much of Medicare’s cost containment came at the expenseof hospitals’ increased cost shifting to private patients, which be-came a primary motivation for businesses to shift their workers intovarious managed-care plans in order to restrain their health-carespending. Ultimately, the rapid and revolutionary paradigm shift inthe United States from fee-for-service health insurance to managedcare is a striking example of how the radical adjustment of a publicprogram (Medicare) can inadvertently trigger the transformation ofan entire industry (the U.S. health-care system), as individuals, in-stitutions, and organizations strategically readjust their behavior inresponse to changing incentives and regulations.

University of Richmond

Glossary of Terms

AHA American Hospital AssociationAMA American Medical AssociationCBO Congressional Budget OfficeCMS Centers for Medicare & Medicaid ServicesDRG Diagnosis-Related GroupFAH Federation of American HospitalsHCFA Health Care Financing AdministrationHHS U.S. Department of Health & Human ServicesHI Medicare’s Hospital Insurance trust fundHMO Health Maintenance OrganizationMedPAC Medicare Payment Advisory CommissionOASI Old Age & Survivors Insurance programPPO Preferred Provider OrganizationPPRC Physician Payment Review CommissionPPS Prospective Payment SystemProPAC Prospective Payment Assessment CommissionRBRVS Medicare’s Resource-Based Relative Value ScaleSMI Medicare’s Supplementary Medical Insurance trust fundSSA Social Security Administration

Acknowledgments

I want to thank several individuals who critiqued previous versions of this articleor offered helpful comments and suggestions along the way, including: MarthaDerthick, Rashi Fein, Christopher Howard, David Karol, Jennifer Mayes, JamesMorone, Paul Pierson, Uwe Reinhardt, Deborah Stone, Margaret Weir, four excel-lent anonymous reviewers, and especially Mark Peterson and Joseph White.

RICK MAYES 169

Notes

1. For example, see David F. Drake, “Managed Care: A Product of MarketDynamics,” JAMA, The Journal of the American Medical Association 277, no. 7 (19February 1997): 560–64.

2. Leiyu Shi and D. Singh, Delivering Health Care in America: A Systems Ap-proach (Gaithersburg, Md., 1998), 299.

3. Jacob S. Hacker and Paul Pierson, “Business Power and Social Policy: Em-ployers and the Formation of the American Welfare State,” paper presented at the2000 Annual Meeting of the American Political Science Association, Washing-ton, D.C. (September 2000), 8.

4. See Lawrence Brown, Politics and Health Care Organization (Washington,D.C., 1983).

5. Ibid.6. David Abernethy interview with the author, 19 June 2002.7. Many thanks to the anonymous reviewer who pointed this out.8. Christopher Howard, The Hidden Welfare State: Tax Expenditures and Social

Policy in the United States (Princeton, 1997).9. Many thanks to the same anonymous who pointed this out.

10. See Paul Pierson, “Not Just What, but When: Timing and Sequence inPolitical Processes,” Studies in American Political Development 14 (Spring 2000): 72–92.

11. Ibid., 72.12. Ibid.13. Ibid., 84.14. Paul Pierson, “Big Slow-Moving, and . . . Invisible: Macro-Social Processes

in the Study of Comparative Politics,” paper presented at the 2000 American Po-litical Science Association Meeting (Washington, D.C.), 5–6.

15. See Paul Pierson, “Increasing Returns, Path Dependence, and the Study ofPolitics,” American Political Science Review 94 (June 2000); and Jacob Hacker, “TheHistorical Logic of National Health Insurance,” Studies in American Political Devel-opment 12 (Spring 1998).

16. See D. Rueschemeyer, E. H. Stephens, and J. Stephens, Capitalist Develop-ment and Democracy (Chicago, 1992), 387. See also Paul Pierson, “When EffectBecomes Cause: Policy Feedback and Political Change,” World Politics 45 (July 1993):595–628.

17. Paul Pierson, Dismantling the Welfare State? Reagan, Thatcher, and the Poli-tics of Retrenchment (Cambridge, 1994), 40.

18. Jacob S. Hacker, The Divided Welfare State: The Battle Over Public and Pri-vate Social Benefits in the United States (Cambridge, 2002), 26.

19. Pierson, Dismantling the Welfare State? 40.20. See Theodore Marmor, The Politics of Medicare, 2d ed. (New York, 2000),

108–19; Jonathan Oberlander, The Political Life of Medicare (Chicago, 2003), chap.4; Eric M. Patashnik, Putting Trust in the U.S. Budget: Federal Trust Funds and thePolitics of Commitment (Cambridge, 2000), chap. 5; Jacob S. Hacker, The DividedAmerican Welfare State: Public and Private Welfare Benefits (Cambridge, 2002).

21. Marmor, The Politics of Medicare, 2d ed., 118.22. Ibid., 94, 137.23. Eric Patashnik and Julian Zelizer, “Paying for Medicare: Benefits, Budgets,

and Wilbur Mills’s Policy Legacy,” Journal of Health Politics, Policy, and Law 26, no.1 (February 2001): 11.

24. As quoted in Linda Demkovich, “Devising New Medicare Payment PlanMay Prove Easier Than Selling It,” National Journal 14, no. 47 (20 November 1982):1981.

25. Robert Ball, “Medicare’s Roots: What Medicare’s Architects Had in Mind,”Generations 20 (Summer 1996): 7.


26. John Rapport, Robert Robertson, and B. Stuart, Understanding Health Eco-nomics (Rockville, Md., 1982), 330.

27. Rosemary Stevens, In Sickness and in Wealth: American Hospitals in the Twen-tieth Century (New York, 1989), 284.

28. E-mail exchange with Clif Gaus, former associate administrator of Policy,Planning & Research, HCFA (11 February 2003).

29. Jonathan Oberlander, “Medicare: The End of Consensus,” paper presentedat the Annual Meeting of the American Political Science Association, Boston (3–6 September 1998), 4.

30. The Board of Trustees, Federal Hospital Insurance Trust Fund, 1983 An-nual Report, Federal Hospital Insurance Trust Fund, House of Representatives, 98thCong., 1st sess., Doc. No. 98–75 (Washington, D.C., 1983), 19–20.

31. See: http://www.ssa.gov/history/interfund.html.32. See Board of Trustees, 1996 Annual Report, Federal Hospital Insurance Trust

Fund (Washington, D.C., 1996).33. See Oberlander, The Political Life of Medicare, 122–24.34. Carolyne Davis, “The Federal Role in Changing Health Care Financing,

Part II,” Nursing Economics (September–October 1983): 98–104.35. For more on the origins of DRGs and how New Jersey’s plan became the

national model, see Robert Fetter, David Brand, and Dianne Gamache, DRGs: TheirDesign and Development (Ann Arbor, 1991); Health Care Finance Administration,A Prospective Reimbursement System Based on Patient Case-Mix for New Jersey Hospi-tals, 1976–1981 (Washington, D.C., 1981); Health Care Finance Administration,Diagnosis-Related Groups: The Effect in New Jersey—The Potential for the Nation(Washington, D.C., 1984); Leah Curtin and Carolina Zurlage, eds., DRGs: TheReorganization of Health (Chicago, 1984); Mohan L. Garg and Barbara M. Barzansky,eds., The Medicare System of Prospective Payment (New York, 1986); and HowardSmith and Myron Fottler, Prospective Payment (Rockville, Md., 1985).

36. Office of Technology Assessment, “Diagnosis Related Groups (DRGs) andthe Medicare Program Working Paper: Using Diagnosis Related Groups in HospitalPayment—The New Jersey Experience,” Congress of the United States (Washing-ton, D.C., December 1983).

37. John K. Iglehart, “Health Policy Report: The New Era of Prospective Pay-ment for Hospitals,” New England Journal of Medicine 307 (11 November 1982):1291.

38. Robert Rubin interview with Edward Berkowitz, 16 August 1995.39. Ibid.40. Jack Owen interview with the author, 17 February 2003.41. Bruce Vladeck interview with the author, 14 August 2002.42. Louise Russell, Medicare’s New Hospital Payment System: Is It Working?

(Washington, D.C., 1989), 7–9.43. Ibid.44. See Stevens, In Sickness and in Wealth, 322–27.45. Sheila Burke interview with the author, 2 October 2002.46. John K. Iglehart, “Health Policy Report,” New England Journal of Medicine

23 (9 June 1983): 1429.47. For more on the Social Security crisis, see Joseph White and Aaron

Wildavsky, The Deficit and the Public Interest (Berkeley and Los Angeles, 1989),chap. 14, “A Triumph of Governance: Social Security”; and Paul Light, Artful Work:The Politics of Social Security Reform (New York, 1985).

48. For a comprehensive analysis of how Congress devised, passed, and imple-mented the new reimbursement system, see David G. Smith, Paying for Medicare(New York, 1992), 23–120.

RICK MAYES 171

49. Timothy Clark, “Congress Avoiding Political Abyss by Approving SocialSecurity Changes,” National Journal 15 (19 March 1983): 611–15. See also Light,Artful Work, 3.

50. ProPAC, Medicare and the American Health Care System, Report to the Con-gress (Washington, D.C., June 1991), 90.

51. Robert F. Coulam and Gary L. Gaumer, “Medicare’s Prospective PaymentSystem: A Critical Appraisal,” Health Care Financing Review, 1991 Annual Supple-ment, U.S Department of Health and Human Services, Health Care Financing Ad-ministration (Baltimore, 1991), 46.

52. David Burda, “What We Have Learned from DRG’s,” Modern Healthcare (4October 1993): 44.

53. Traditionally, there have been two hospital sectors—whose missions oftenoverlap—that policymakers have explicitly used Medicare to subsidize: (a) teach-ing and (b) indigent safety-nets. In the first sector, teaching, Medicare providestwo types of extra payments to hospitals with graduate medical education programsto compensate them for their higher institutional costs. The indirect medical edu-cation (IME) adjustment, which accounted for $3.7 billion in 1999, pays the costsof treating sicker patients and additional tests needed for training purposes. Teach-ing hospitals also receive a direct graduate medical education (DGME) adjustment,which accounted for $2.2 billion in 1999, for training medical residents. In thesecond hospital sector, indigent safety-nets, Medicare provides what are known as“disproportionate share” payments to hospitals that treat a large number of Medic-aid and uninsured patients. Initiated by policymakers in 1986, the Medicare Dis-proportionate Share (DSH) program increases payment rates to hospitals that pro-vide a disproportionately large share of health care to the poor whose conditionsare often more severe than average patients and, yet, are less able to pay. Thisexplicit adjustment costs the government (Medicare) approximately $5 billion peryear.

54. See Smith, Paying for Medicare: The Politics of Reform.55. “Health Care Cost Containment Strategies,” Hearing Before the Committee

on Labor and Human Resources, United States Senate, 98th Cong., 2d sess. (21 June1984), 192.

56. “Issues Relating to Medicare Hospital Payments,” Hearing Before the Sub-committee on Health of the Committee on Ways and Means, House of Representatives,99th Cong., 1st sess. (14 May 1985), 158.

57. Allen Dobson, Nancy Bray, et al., “An Evaluation of Winners and LosersUnder Medicare’s Prospective Payment System,” Lewin-ICF Report to the Prospec-tive Payment Assessment Commission (11 May 1992).

58. George W. Whetsell, “The History and Evolution of Hospital Payment Sys-tems: How Did We Get Here?” Nursing Administration Quarterly 23 (Summer 1999):1–10.

59. Ibid.60. Jack Owen interview with the author, 17 February 2003.61. See Dennis S. Ippolito, Uncertain Legacies: Federal Budget Policy from

Roosevelt Through Reagan (Charlottesville, 1990), chaps. 6 and 7; White andWildavsky, The Deficit and the Public Interest, chaps. 19 and 21.

62. “1987 Medicare Budget Issues,” Hearing Before the Subcommittee on Healthof the Committee on Ways and Means, House of Representatives, 99th Cong., 2dsess. (6 March 1986), 338–33.

63. “Status of the Medicare Hospital Prospective Payment System,” HearingBefore the Subcommittee on Health of the Committee on Ways and Means, 100th Cong.,2d sess. (1 March 1988), 83–84.

64. Bill Gradison interview with the author, 12 June 2002.65. Leon Panetta interview with the author, 13 August 2002.


66. Rick Pollack interview with the author, 27 December 2002.67. Lisa Potetz interview with the author, 24 July 2002.68. Congressional Budget Office, The Economic and Budget Outlook: Fiscal Years

1993–1997 (Washington, D.C., January 1992), 58.69. Patashnik, Putting Trust in the U.S. Budget, 104.70. Robert Reischauer interview with the author, 16 August 2002.71. Oberlander, “Medicare: The End of Consensus,” 6.72. Stuart Guterman, Jack Ashby, and Timothy Greene, “Hospital Cost Growth

Down: Unprecedented Cost Constraint by Hospitals Has Maintained Their Bot-tom Line. But Can It Continue?” Health Affairs 15 (Fall 1996): 136.

73. Stuart Altman interview with the author, 22 July 2002.74. See Stuart Guterman, Stuart Altman, and D. Young, “Hospitals’ Financial

Performance in the First Five Years of PPS,” Health Affairs 9 (1990): 125–34; andStuart Altman and Jack Ashby, “The Trend in Hospital Output and Labor Produc-tivity, 1980–89,” Inquiry 29 (Spring 1992): 80–92.

75. MedPAC, Report to the Congress: Medicare Payment Policy, 52.76. See Jack Needleman, “Cost Shifting or Cost-Cutting: Hospital Responses

to High Uncompensated Care,” (Cambridge, Mass., 1994); and Michael Morrisey,Cost Shifting in Health Care (Washington, D.C., 1994).

77. “Health Care Cost Containment Strategies,” Hearing Before the Committeeon Labor & Human Resources, U.S. Senate, 98th Cong., 2d sess. (21 June 1984),148.

78. Ibid.; see also D. Coddington, D. Keen, K. Moore, and R. Clarke, The Cri-sis in Health Care: Costs, Choices, and Strategies (San Francisco, 1990), chap. 6,103–13; and D. Banks, S. Foreman, and T. Keeler, “Cross-Subsidization in HospitalCare,” Health Matrix Journal of Law-Medicine 9 (Winter 1999): 1–35.

79. See K. Levit, H. Lazenby, L. Sivarajan, et al., “National Health Expendi-tures, 1994,” Health Care Financing Review 17 (Fall 1996): 205–42.

80. See Karl Pozer, “Strategies to Contain Health Care Costs,” Business & Health(September 1990): 35.

81. See Drake, “Managed Care: A Product of Market Dynamics” (as in note 1).82. Tom Scully interview with the author, 24 October 2002.83. Marilyn Moon interview with the author, 2 August 2002.84. “Options for Health Insurance,” 187.85. James Mongan interview with the author, 3 October 2002.86. Michael Bromberg interview with the author, 23 July 2002.87. Rick Pollack interview with the author, 29 October 2002.88. For more on the role of cost shifting, see George W. Whetsell, “The His-

tory and Evolution of Hospital Payment Systems: How Did We Get Here?” NursingAdministration Quarterly 23 (Summer 1999): 9–15.

89. David Burda and Cathy Tokarski, “Hospitals Are Under Pressure to JustifyCost Shifting: But Some Payers Are Rejecting Hospitals’ Excuses and Are Demand-ing Data,” Modern Healthcare (12 November 1990), 28–36.

90. Uwe Reinhardt, “The Predictable Managed Care Kvetch on the Rocky Roadfrom Adolescence to Adulthood,” Journal of Health Politics, Policy, and Law 24 (Oc-tober 1994): 902–3; “Payroll Taxes, Health Insurance, and SBA Budget Proposals,”Hearing Before the Committee on Small Business, House of Representatives, 101stCong., 2d sess. (29 March 1990), 4–9.

91. See Coddington, Keen, Moore, and Clarke, The Crisis in Health Care, 103;and G. Ruffenach, “Health Insurance Premiums to Soar in ‘89,” Wall Street Journal(23 October 1989), B12.

92. See M. McGarvey, “The Challenge of Containing Health Care Costs,” Fi-nancial Executive 8 (January–February 1992): 34–40; Jan P. Clement, “Dynamic Cost

RICK MAYES 173

Shifting in Hospitals: Evidence from the 1980s and 1990s,” Inquiry 34 (Winter 1997–98): 340–50.

93. Rashi Fein, Medical Care, Medical Costs: The Search for a Health InsurancePolicy, 2d ed. (Cambridge, Mass., 1999), 95.

94. See Michael Morrissey, “Hospital Cost Shifting: Revisited,” EBRI Issue Brief,No. 180 (Washington, D.C.: Employee Benefits Research Institute, EBRI, Decem-ber 1996); Jack Hadley, Stephen Zuckerman, and Lisa Iezzoni, “Financial Pressureand Competition: Changes in Hospital Efficiency and Cost-Shifting Behavior,” Medi-cal Care 34 (March 1996): 205–19.

95. See J. Geisel, “Solution Proposed for N.J. Cost Shifting,” Business Insur-ance 26 (2 November 1992): 3–5; Allen Dobson and Richard Clarke, “Shifting NoSolution to Problem of Increasing Costs,” Healthcare Financial Management 46 (July1992): 24–30; D. Shalowitz, “Hospitals Shift Costs to Cover Costs,” Business Insur-ance 22 (15 February 1988): 14; McGarvey, “The Challenge of Containing HealthCare Costs”; R. Clarke, “Cost Shifting Merits an Explanation,” Healthcare FinancialManagement 46 (July 1992): 12.

96. Z. Schiller, “The Humana Flap Could Make All Hospitals Feel Sick,” Busi-ness Week (4 November 1991): 34.

97. D. McFadden, “The Legacy of the $7 Aspirin,” Management Accounting 71(April 1990): 38–41.

98. Malcolm Gladwell, “Insurance System Squeezes Some Hospitals; PaymentsDrop as Governments Cut Reimbursements, Private Insurers Fight Cost-Shifting,”Washington Post, 29 March 1992, A9.

99. Dobson and Clarke, “Shifting No Solution to Problem of Increasing Costs.”100. Ibid.101. See http://www.brtable.org/newsroom.htm.102. Robert Winters interview with the author, 28 August 2002.103. Reinhardt, “The Predictable Managed Care Kvetch,” 903.104. See Cathie Jo Martin, “Nature or Nurture? Sources of Firm Preference for

National Health Reform,” American Political Science Review 89, no. 4 (December1995): 898–914. For further evidence of and explanation for businesses respondingto cost shifting by moving to various forms of managed care, see Coddington, Keen,and Moore, “Cost Shifting Overshadows Employers’ Cost-Containment Efforts”;Clarke, Coddington, Keen, and Moore, The Crisis in Health Care; Leslie Berkman,“Employers Treat Health Plan Costs with Managed Care,” Los Angeles Times, 24January 1990, 7; Paul Kenkel, “Employers More Willing to Trade Workers’ Choiceof Providers for Genuine Cost Containment,” Modern Healthcare (27 July 1992):70; Patrice Duggan, “Power Shift: Desperate to Control Health Insurance Costs,Employers Are Using Their Market Power to Force Down Doctors’ Bills,” Forbes(18 February 1991), 80; Louise Kertesz, “Health Cost Hikes Checked: Large FirmsReport Managed Care Success,” Business Insurance (30 March 1992): 2; Paul Kenkel,“HMOs’ Savings Have ‘Spillover Effect,’” Modern Healthcare (13 January 1992), 30.

105. Karen Titlow and Ezekiel Emanuel, “Employer Decisions and the Seeds ofBacklash,” Journal of Health Politics, Policy, and Law 24, no. 5 (October 1999): 943–44.

106. Fein, “The HMO Revolution: How It Happened, What It Means,” 33–34.107. Mark Pauly and Sean Nicholson, “Adverse Consequences of Adverse Se-

lection,” Journal of Health Politics, Policy, and Law 24 (October 1999): 925.108. Ibid.109. Paul and Nicholson, “Adverse Consequences of Adverse Selection,” 924.110. Douglas Wholey, Jon Christianson, John Engberg, and Cindy Bryce, “HMO

Market Structure and Performance: 1985–1995,” Health Affairs 16 (November–De-cember 1997): 75–84.

111. Marsha Gold, “DataWatch: HMOs and Managed Care,” Health Affairs 10(Winter 1991): 189–206.


112. Jon Gabel, “Ten Ways HMOs Have Changed During the 1990s,” HealthAffairs 16 (May–June 1997): 134–45.

113. Ibid.114. Gold, “DataWatch: HMOs and Managed Care,” 192–93.115. G. Halvorson, “Health Plans’ Strategic Responses to a Changing Market-

place,” Health Affairs 18 (March–April 1999): 28–29.116. Joseph White, Competing Solutions: American Health Care Proposals and In-

ternational Experience (Washington, D.C., 1995).117. Thanks to Mark Peterson for showing me this line of argument.118. Hacker, The Divided Welfare State, 318.119. See Sallyanne Payton, “The Politics of Comprehensive National Health

Reform: Watching the 103rd and 104th Congress,” in Health Policy, ed. Max Heirichand Marilynn M. Rosenthal (Boulder, 1998).

120. “Foster Higgins Survey: Managed Care Slows Hike in Employer HealthCosts,” Best’s Review: Life-Health Insurance Edition (April 1993), 7.

121. For example, see Titlow and Emanuel, “Employer Decisions and the Seedsof Backlash,” 941–47.

122. Reinhardt, “The Predictable Managed Care Kvetch,” 908.123. Oberlander, The Political Life of Medicare, 199.124. Personal communication with Jack Ashby, MedPAC Hospital Research Di-

rector (7 August 2003): “The 14% PPS margins [in Fig. 1] come from ProPAC pub-lications and are based on Medicare cost report data. The 0.98 to 1.01 payment-to-cost ratios [in Table 4] are, of course, from the AHA annual survey. The first andperhaps primary difference between the two measurements is that the cost reportfigure is an inpatient margin, while the AHA numbers cover all services hospitalsprovide for Medicare beneficiaries. Medicare inpatient margins have always been,and still are, much higher than Medicare outpatient margins. Besides that, though,the two data sources are fundamentally different in two ways. First, the cost reportmeasure is based on Medicare-allowable costs while the AHA measure captures allcosts per the hospitals’ books. This difference also leads to a higher margin valuefor the cost report data. Second, the cost report measure reflects a complex methodfor allocating costs among payers, while the AHA data reflect a simple applicationof an RCC to charges by payer to produce costs by payer. While the proof has beenillusive to date, we have anecdotal evidence that hospitals over the years have settheir charges so as to maximize the allocation of costs to Medicare, which thenbiases the AHA payment to cost ratio downward. Charges are used in the costreport allocation also, but to a lesser degree than in the AHA data. This factor alsoleads to a higher value for the cost report data, and this manipulation of chargeswas at its zenith in the first few years of the PPS. The net result of all this in ourminds [at MedPAC] is that the AHA data are quite useful for monitoring trends(which includes providing evidence that there has been cost shifting), but are muchless useful in establishing the level of margins or payment/cost ratios.”

Causal Chains and Cost Shifting: How Medicare’s Rescue …bmayes/pdf/16.2mayes.pdf · Causal...

Documents

Transcript of Causal Chains and Cost Shifting: How Medicare’s Rescue …bmayes/pdf/16.2mayes.pdf · Causal...