The Future of Healthcare Reform in the United States 9780226255002

In the years since the passage of the Patient Protection and Affordable Care Act (PPACA, or, colloquially, Obamacare), m

358 25 3MB

English Pages 352 [354] Year 2015

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

The Future of Healthcare Reform in the United States
 9780226255002

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

The Future of Healthcare Reform in the United States

The Future of Healthcare Reform in the United States

edited by anup Malani & Michael H. Schill The University of Chicago Press

Chicago and London

Anup Malani is the Lee and Brena Freeman Professor at the University of Chicago Law School and professor at the Pritzker School of Medicine. Michael H. Schill is dean of and the Harry N. Wyatt Professor of Law at the University of Chicago Law School. Chapter 4, “Essential Health Benefits and the Affordable Care Act: Law and Process,” by Nicholas Bagley and Helen Levy originally appeared in Journal of Health Politics, Policy and Law, vol. 39, no. 2, pp. 441–465. Copyright, 2014, Duke University Press. All rights reserved. Republished by permission of the copyright holder, Duke University Press. www.dukeupress.edu. The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 2015 by The University of Chicago All rights reserved. Published 2015. Printed in the United States of America 24 23 22 21 20 19 18 17 16 15

1 2 3 4 5

ISBN-13: 978- 0-226-25495-1 (cloth) ISBN-13: 978- 0-226-25500-2 (e-book) DOI: 10.7208/chicago/9780226255002.001.0001 Library of Congress Cataloging-in-Publication Data The future of healthcare reform in the United States / [edited by] Anup Malani and Michael H. Schill. pages cm Includes bibliographical references and index. ISBN 978-0-226-25495-1 (cloth : alk. paper) — ISBN 978-0-226-25500-2 (ebook) 1. United States. Patient Protection and Affordable Care Act. 2. National health insurance—Law and legislation—United States. 3. Medical care—Law and legislation— United States. 4. Health care reform—United States. I. Malani, Anup, editor. II. Schill, Michael H., editor. KF3605.A328201A2 2015 362.1'0425—dc23 2015001692 This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

Contents Introduction 1 Michael H. Schill and Anup Malani PART

1.

ACA and the Law

Chapter 1. Postmortem on NFIB v. Sebelius: Early Reflections on the Decision That Kept the ACA Alive 13 Carter G. Phillips and Stephanie P. Hales Chapter 2.

Federalism, Liberty, and Risk in NFIB v. Sebelius Aziz Z. Huq

32

Chapter 3. The Future of Healthcare Reform Remains in Federal Court 50 Jonathan H. Adler Chapter 4. Essential Health Benefits and the Affordable Care Act: Law and Process 81 Nicholas Bagley and Helen Levy PART

2.

ACA and the Federal Budget

Chapter 5.

The Fiscal Consequences of the Affordable Care Act 109 Charles Blahous

Chapter 6.

Estimating the Impact of the Demand for ConsumerDriven Health Plans Following the 2012 Supreme Court Decision of the Constitutionality of the Patient Protection and Affordable Care Act 135 Stephen T. Parente

Contents

vi PART

3.

ACA and Healthcare Delivery

Chapter 7.

After the ACA: Freeing the Market for Healthcare 161 John H. Cochrane

Chapter 8.

Obamacare and the Theory of the Firm Einer Elhauge

202

Chapter 9. Can Federal Provider Payment Reform Produce Better, More Affordable Healthcare? 221 Meredith B. Rosenthal PART

4.

Healthcare Costs, Innovation, and ACA

Chapter 10. The Role of Technology in Expenditure Growth in Healthcare 249 Amitabh Chandra and Jonathan Holmes Chapter 11. Economic Issues Associated with Incorporating CostEffectiveness Analysis into Public Coverage Decisions in the United States 271 Anupam B. Jena and Tomas J. Philipson Chapter 12. The Complex Relationship between Healthcare Reform and Innovation 289 Darius Lakdawalla, Anup Malani, and Julian Reif PART

5.

ACA and Health Insurance Markets

Chapter 13. The Affordable Care Act and Commercial Health Insurance Markets: Fixing What’s Broken? 313 James B. Rebitzer Chapter 14. A Cautionary Warning on Healthcare Exchanges: A Plea for Deregulation 329 Richard A. Epstein List of Contributors 353 Index 357

Introduction

I

n 2010 President Barack Obama signed perhaps the most important piece of social welfare legislation since the New Deal—the Patient Protection and Affordable Care Act (henceforth Affordable Care Act, or ACA). This overhaul of the nation’s healthcare system aims to extend health insurance coverage to low-income groups by expanding Medicaid. In addition, individuals and families with modest incomes—up to 400 percent of the poverty level—who do not qualify for Medicaid or receive employer-provided insurance will be entitled to tax benefits that will subsidize their insurance premiums. The ACA raises funds to pay for this expansion through a combination of reductions in Medicare and Medicaid provider reimbursements and premiums and various taxes on medical providers and insurers. In order to maintain stable insurance markets without adverse selection the ACA most famously includes the “individual mandate,” which requires most people to purchase health insurance through their employers or on new exchanges. Although it was not the sole health reform of the new millennium—it came on the heels of the implementation of Medicare Part D, which provided drug coverage for seniors in Medicare, in 2006—it is much larger in scope. The ACA has engendered tremendous controversy, which was reflected in its enactment by a sharply divided Congress. Shortly after it was signed into law, it was challenged in court. In 2012, the United States Supreme Court took up the question of whether the individual mandate violated Article I of the Constitution and whether the Medicaid expansion impermissibly coerced the states. That summer, in National Federation of Independent Business (NFIB) v. Sebelius, it upheld the individual mandate while circumscribing the expansion of Medicaid.

2

Michael H. Schill and Anup Malani

The University of Chicago Law School and Medical School jointly convened a conference on healthcare reform months after NFIB was decided. Our goal was to bring together an interdisciplinary group of experts in Hyde Park—economists, lawyers, healthcare professionals—to discuss the meaning of the case, its impact on the implementation of the ACA, and more generally the question of healthcare reform. This volume is the product of that discussion.

Legal Challenges to the ACA The fi rst part of this volume focuses on legal challenges to the ACA. Although much of the statute was upheld, the Supreme Court’s decision shapes the contours of the law’s implementation. Our very fi rst chapter sets the stage for what follows. Carter Phillips, one of the nation’s most eminent Supreme Court litigators, and his colleague Stephanie Hales discuss the NFIB decision and what it means for the ACA. In upholding the individual mandate as an exercise of the tax power rather than the commerce power, the Court not only surprised most observers, but unsettled future jurisprudence on the relationship between federal power and the states. Similarly, its decision that Congress could not condition the receipt of all Medicaid subsidies on states agreeing to expand Medicaid wiped away years of precedent and contributes to a patchwork quilt of Medicaid coverage and rules throughout the nation. The authors speculate on how NFIB will complicate the implementation of healthcare reform over the next decade or two. The second chapter, by Aziz Huq, a constitutional law professor at the University of Chicago Law School, puts the NFIB decision in a jurisprudential context. Huq examines the relationship between the decision and notions of federalism. He concludes that despite the decision’s being viewed by many people as a decision about federalism (the appropriate relationship between the national government and state legislatures), federalism actually explains very little of the decision. Instead, Huq believes that the decision actually mediates the tension between risk spreading and individual liberty. He looks at each of the key aspects of the case and concludes that the decision reflects an early-twentieth-century ideology of managing social risk that combines “laissez-faire instincts and gendered understandings of welfare.”

Introduction

3

Jonathan Adler’s chapter looks to the future and the possibility of additional litigation on healthcare reform. Adler, an administrative and constitutional law professor at Case Western Reserve University, examines potential litigation over implementation of the ACA. An important theme of the chapter is that because of size of the statute, the speed with which it was passed, and the process that led to that passage (which did not include a conference committee), there are many parts that are either ambiguous or subject to legal challenge. Among these are the payment of federal tax subsidies to families in states that do not set up healthcare exchanges, the requirement that employer plans (except in certain limited circumstances) include coverage for contraception, and the unique administrative process for controlling the costs of Medicare. In addition, Adler speculates that if the charge assessed on individuals and families that do not purchase insurance is set too high, we might see another round of challenges to the individual mandate provision on the ground that it is not truly a “tax” as contemplated by Justice Roberts, but is actually a penalty. The fi nal chapter of our fi rst part on the law of healthcare reform focuses its attention not on constitutional challenges to the ACA but instead on potential claims that its implementation by the Department of Health and Human Services (HHS) violated principles of administrative law. While HHS followed formal notice-and- comment procedures for most provisions of the ACA, it adopted a more informal “subregulatory guidance” approach with respect to what constitutes “essential health benefits” of private insurance plans sold through the individual and smallgroup marketplace. Rather than publishing a formal rule in the Federal Register, HHS placed a bulletin on its website. According to Bagley and Levy, given the broad discretion contained in the law, this promulgation of rules is unlikely to be found to violate the law—nor should it. Indeed, according to the authors, the more “informal” method of administrative rule promulgation may have resulted in a process that was even more open and amenable to public scrutiny than the more typical, formal process used by federal agencies.

Fiscal Consequences of the ACA The second part of this volume examines the fi scal consequences of the ACA. In light of recurrent federal budget deficits and the seeming

4

Michael H. Schill and Anup Malani

inability of Congress and the president to achieve a long-lasting solution to the problem, a significant issue in the debate over the ACA is how much it will cost in terms of additional government spending and how much it will save in terms of reduced payments, particularly in the Medicare program. The fi rst chapter in this part, by Charles Blahous, a senior research fellow at George Mason University’s Mercatus Center and a public trustee for the Social Security and Medicare programs, paints a pessimistic picture of the costs of the program. According to Blahous, when the ACA was under consideration by members of Congress, the Congressional Budget Office (CBO) provided analyses that appeared to show that its passage would reduce future federal deficits. To obtain this result, the CBO assumed that if Congress did not pass the ACA it would go on funding the Medicare program at current levels, projected into the future. Blahous believes that this assumption is not appropriate and greatly understates the relative cost of the new law. Instead, he suggests that the ACA will add between $320 billion and $520 billion to federal deficits over the next decade. If the objective were to improve the nation’s fiscal balance, Congress would need to repeal two-thirds of the federal subsidies planned for participants in the new health insurance exchanges. The second chapter follows the thread on healthcare cost growth, but from an economic rather than an accounting perspective. Stephen Parente, a professor at the University of Minnesota’s Carlson School of Management, focuses his analysis on the demand for consumer- driven health plans (CDHP) (i.e., plans with high deductibles and savings accounts to help cover the cost of deductibles). Such plans have been promoted because, among other things, they may encourage beneficiaries to be more price conscious, and thus drive down healthcare expenditures or at least rationalize them from an economic perspective. Parente reports on a simulation that examines how the ACA, which seeks to reduce rather than increase consumer susceptibility or sensitivity—depending on your perspective—to healthcare prices, might affect consumer demand for CDHPs and thereby private and federal healthcare cost growth. The simulation also considers the effect of the NFIB decision, which gives states greater leeway to opt out of the ACA’s Medicaid expansion. He fi nds, unsurprisingly, that it will lead to fewer covered persons and contribute to lower federal cost.

Introduction

5

The Economic Merits of the ACA The third part of this volume steps back from the immediate legal and fiscal issues surrounding the ACA and examines the normative case for and against it. John Cochrane, a distinguished service professor of fi nance at the University of Chicago Booth School of Business, writes a highly critical chapter that argues that the ACA will make an already inefficient market for healthcare in the United States even worse. Cochrane posits that both the supply and the demand for healthcare are skewed by government programs and regulation. A more rational and economically efficient system would, among other things, increase supply (by reducing regulations that deter new entry) and reduce demand (by making consumers sensitive to the marginal cost of additional treatment). Despite his strong defense of free-market principles, Cochrane understands that a pure libertarian system of healthcare provisions would be politically infeasible in the United States. He suggests that to the extent we want people of modest incomes to consume more healthcare, we should provide them with vouchers and healthcare savings accounts. The chapter that follows Cochrane’s contribution takes a much more optimistic view of the ACA. Einer Elhauge, a professor of healthcare and antitrust law at Harvard University, examines the inefficiencies that result from the fragmented structure of healthcare production in the United States. Many of the reasons for this fragmentation are rooted in federal laws governing Medicare and antitrust, and state laws on hospital structure and torts. Elhauge suggests that elements of the ACA, especially its encouragement of vertical integration of medical service providers through so- called Accountable Care Organizations, may improve the quality and lower the cost of healthcare in the United States. He also points to ways in which the ACA institutionalizes innovation and provides incentives for cost saving as further drivers for integration. As Elhauge notes, the ACA incentivizes healthcare providers to create Accountable Care Organizations to improve production efficiency with the use of a global payment that allows such organizations to capture a share of the cost savings that they achieve for Medicare. In the next chapter Meredith Rosenthal, a professor at the Harvard School of Public Health, picks up the thread of payment reform and examines the Accountable Care Organization model and other payment reforms, some of which are also included in the ACA. Rosenthal uses the history of Medicare

6

Michael H. Schill and Anup Malani

reimbursement policies and the literature on provider payment policy to determine whether manipulation of the manner and conditions upon which physicians and hospitals are reimbursed for providing care, including pay-for-performance schemes, can improve the efficiency of medical care. Although the historical and empirical record on reimbursement incentives is mixed, Rosenthal concludes that the ACA makes some meaningful changes and is sanguine about its prospects for improving healthcare production.

Technology and the Future of Healthcare Reform The penultimate part of this volume turns its gaze away from the ACA and the present and focuses on healthcare technology, such as pharmaceuticals, and the future. Technology plays a central role in healthcare debates. On the one hand, it is considered an important contributor to longevity, which has enormous economic value (Murphy & Topel 2006). Yet technology is also considered an important driver of healthcare cost growth (Newhouse 1992; Chandra and Skinner 2012). Both supporters and critics of the ACA have to come to terms with such medical technology. Supporters must understand that the ACA, through its coverage expansion, has increased the number of people with access to medical technology, for good or ill. Critics, who have complained that the ACA does not do enough to address cost growth, must consider the role and value of technology in their criticisms. The fi rst chapter on technology, by Amitabh Chandra, a professor at Harvard’s Kennedy School of Government, and Jonathan Holmes, also from Harvard, examines the cost effectiveness of various technologies and divides them into three categories. They suggest that policy makers and healthcare providers may want to create market-based ways to reduce the use of interventions that have very little value in terms of adding to life expectancy and increase the use of those that do. A key insight is that some technologies might be quite cost effective for certain populations or stages of illness but wasteful for others. This makes it complicated to evaluate technologies in the abstract. Importantly, Chandra and Holmes offer policy options to help tame the inefficient use of technology, including allowing insurance companies to reimburse technologies based on how and when they are used, and to reward outcomes rather than treatment.

Introduction

7

One of the policy proposals that Chandra and Holmes discuss is costeffectiveness analysis (CEA), which monetizes the incremental health benefits of technology and compares them to the cost of those technologies. The chapter by Anupam Jenna and Tomas Philipson, professors at Harvard and the University of Chicago, respectively, discusses two pitfalls associated with CEA as it is currently performed. First, they argue that the cost of technology—the social resources required to create the technology and (re)produce it—is very difficult to estimate. The price that companies set for their technology is not the same as the social cost, because price is endogenous; companies set it with the aim of maximizing profits or at least reimbursement after the technology is created. Cost effectiveness should not, but often does, confuse price with cost. Second, cost-effectiveness thresholds set by government are in effect price controls, which can stifle innovation. When a government sets a threshold costeffectiveness amount (e.g., $100,000 per life year saved) above which it will not approve reimbursement for a technology, companies will respond by trying to price their technology as close to that threshold as possible, to ensure approval and maximize reimbursement given approval. As a result, the cost-effectiveness threshold functions as a price control, not as an objective determination of how much a technology costs society. The fi nal chapter in this part shifts readers’ attention from the pitfalls of valuing technology to the complex relationship between health insurance expansions, such as those in the ACA, and new medical technologies. Authored by Darius Lakdawalla, Anup Malani, and Julian Reif, professors at the University of Southern California, the University of Chicago, and the University of Illinois, respectively, this chapter makes three basic points. First, health insurance may both drive and hinder innovation—drive it because it increases demand for medical care and hinder it because it offers individuals an option to obtain care outside of medical trials, which reduces enrollment in such trials. Second, health insurance reduces the welfare costs of using patents to encourage medical innovation. Health insurance copayments lower the per-unit price of a drug so that it is closer to the marginal cost of the drug; at the same time, health insurance can offer the maker of a patented drug a price that is much higher than the marginal cost of the drug, so as to induce innovation. Third, medical innovation may actually be a substitute for health insurance from the perspective of a risk-averse person without access to full indemnity insurance. Empirical work by the authors in fact suggests that the welfare gains from risk reduction due to innovation may be large

8

Michael H. Schill and Anup Malani

relative to the welfare gains from risk reduction due to health insurance (Lakdawalla et al. 2013).

Healthcare Exchanges The fi nal part of this volume turns our attention back to the ACA, but focuses not on insurance and access to technology but on the marketplace for insurance, specifically the new health-insurance exchanges. The ACA provides for the creation of state or federal marketplaces for certain individuals to purchase health insurance. For individuals or households with incomes under 400 percent of the federal poverty level and without Medicaid coverage or employer-sponsored health insurance, the ACA also provides subsidies to reduce the price of health insurance premiums and out- of-pocket payments. The idea behind the marketplace is that it will simplify the purchase of health insurance, which can be a complicated product, and can be a platform used to implement certain health insurance regulations. The last two chapters debate the merits of the ACA’s exchanges. James Rebitzer, a professor of management economics and public policy at Boston University, observes that there are significant search frictions when individuals shop for health insurance contracts. These frictions can stem from the bounded rationality (or cognitive costs) of participants or from standard transaction costs that are intensified by the complexity of health insurance. Rebitzer fi nds evidence for search frictions from the fact that markets for employers who purchase insurance exhibit much greater turnover than markets for self-insurance. Rebitzer concludes that the ACA may reduce frictions and thereby promote efficiency by simplifying and standardizing products, and by reducing adverse selection. In the fi nal chapter of the book, Richard Epstein, a law professor at NYU and the University of Chicago, takes issue with Rebitzer’s position that search frictions might justify the ACA’s requirements for insurance exchanges. Epstein is particularly concerned that the ACA’s defi nition of “essential health benefits” will forestall innovation and force expensive plans on individuals who would prefer cheaper plans with less coverage. He also warns that the ACA’s cross-subsidy characteristics will exacerbate problems of adverse selection, and that its assurance of coverage will intensify moral hazard and lead to fewer precautions taken by at least part of the insured population.

Introduction

9

Any casual observer of the American political scene would be hard pressed not to recognize the incredibly charged environment in which healthcare reform exists. Party-line votes in Congress, challenges in court, and constant criticisms of the implementation of the ACA are commonplace. Perhaps this should not be surprising. Healthcare is bound up with life itself; the life of one’s loved ones and oneself. Even without killing, a catastrophic illness can wipe out a person’s financial resources and change that person’s life forever. Into this deeply significant and emotional environment, discussions of healthcare reform also directly implicate many of the most sensitive ideological cleavages in our society—ranging from the role of markets in distributing vital goods and services, to the liberty to make one’s choices free of government interference, to the relationship among different levels of government, on the one hand, and between government and the individual, on the other. In this book, we have attempted and hopefully succeeded in shedding some light on a set of key issues that inform current healthcare policy in the United States, the recent healthcare reform legislation, and future avenues of reform. We have no doubt excluded many important research topics and perspectives, the relevance of which may well become even more apparent as the ACA is implemented. It is our fervent belief that while the application of economic and legal analysis to these questions is not the only framework for analyzing this extraordinarily important subject, it is a fruitful one that will help pave the way to new and better interventions to promote social welfare.

References Chandra, Amitabh and Jonathan Skinner. 2012. “Technology Growth and Expenditure Growth in Health Care.” Journal of Economic Literature. 50(3):645– 80. Lakdawalla, Darius N., Anup Malani, and Julian Reif. 2013. “The Insurance Value of Medical Innovation.” University of Chicago Department of Economics Working Paper. Accessed October 30, 2014. https://economics.uchicago .edu/workshops/Malani%20-%20The%20Insurance%20Value%20of%20 Medical%20Innovation.CURRENT.pdf. Murphy, Kevin M., and Robert H. Topel. 2006. “The Value of Health and Longevity.” Journal of Political Economy. 114(5): 871– 904. Newhouse, Joseph P. 1992. “Medical Care Costs: How Much Welfare Loss?” Journal of Economic Perspectives, 6(3): 3– 21.

Chapter ONE

Postmortem on NFIB v. Sebelius Early Reflections on the Decision That Kept the ACA Alive Carter G. Phillips and Stephanie P. Hales

Introduction

B

ut for one vote by one Justice of the Supreme Court, there would be no reason for any discussion of healthcare under the Affordable Care Act (ACA). Chief Justice Roberts’s surprising opinion not only gives life to this symposium, but along with the other opinions by the various Justices serves as a useful context for beginning the process of looking to the future. This essay thus discusses the U.S. Supreme Court’s decision in National Federation of Independent Business v. Sebelius1 and its implications going forward. That the Supreme Court’s decision upholding the ACA bears on “The Future of Health Care Reform in the United States” is beyond doubt. Less clear at this time, however, are the precise form and contours of this impact—both in the particular space of healthcare reform and in the broader realm of Supreme Court jurisprudence addressing the scope of congressional power. But being lawyers, a lack of clarity begins our discussion rather than ends it. Indeed, this discussion was originally pitched to us as a “postmortem,” which is a somewhat strange way to think about an opinion that as of the date of this conference (October 12, 2012) had been public for less than four months, and as such was decidedly not dead. Even as of this

14

Carter G. Phillips and Stephanie P. Hales

writing, the yet-recent decision is very much alive. And thanks to it, so is the ACA. Thus, this essay is more a general analysis of the decision than a postmortem in the literal sense. We begin with an overview of the decision, featuring three main opinions with a host of unexpected and in many ways remarkable elements. We then discuss the three primary doctrinal areas of the Court’s decision: (1) its analysis of ACA’s “individual mandate” under the Commerce Clause, (2) its approval of the “individual mandate” under Congress’s taxing authority, and (3) its ruling on ACA’s Medicaid expansion provisions. Although a relatively short period of time has passed since the decision was announced, it paved the way for continued ACA implementation, which has proceeded along aggressive timelines (particularly following the reelection of President Obama in November). Thus, while appreciating that these implementation efforts are ongoing and in flux, and that additional challenges to the ACA continue to wind their way through the courts, we discuss not only what the Court held and our thoughts on why it did so, but also a few observations on how the Court’s decision appears to be affecting (or not) ACA implementation to date. The decision’s impact beyond the realm of healthcare reform likewise remains to be seen. Throughout the decision, the Court draws a number of distinctions that we imagine only lawyers can love, many of which center on—but do not clearly defi ne the boundaries of—the concept of “coercion”: At what point does an act of Congress become too coercive, either with respect to individuals or to the states, to withstand constitutional scrutiny? In many ways, the Justices’ opinions in NFIB v. Sebelius raise far more questions on this issue than they answer, planting seeds for continued constitutional litigation over where to draw lines.

Did Anyone Predict This? Many people expected that the ACA would survive the Court’s review (although we suspect that just as many felt certain it would not after the oral argument). But did anyone predict the constellation of elements in the decision that culminated in that result? Probably not. First, the Justices split into three main opinions, including an opinion written by Chief Justice Roberts only for himself. Second, the Chief Justice’s opinion was joined in certain places by the four liberals, in the opinion authored by Justice Ginsburg, 2 and not at all

Postmortem on

NFIB V. SEBELIUS

15

by the four dissenting conservatives;3 although they clearly agreed with some portions of the Chief Justice’s analysis, remarkably, the four conservatives never even mentioned his opinion in their dissent. Third, the dissent coauthored by the four conservatives has no individually named author. This is an extraordinary step that is exceedingly rare. Perhaps even more provocatively, and as noted, this dissent with no individually named author never mentions, let alone addresses, the analysis in the Chief Justice’s opinion. Even Justice Ginsburg, at various times— both when she joined with it and when she did not—at least acknowledged and did business with aspects of the Chief Justice’s opinion, as is customary in separate opinions written by individual members of the Court on their own behalf or for one or more other Justices. She also dealt with the four conservatives’ dissenting analysis (and notably, they dealt in some places with hers). Fourth, the Chief Justice was a “swing vote,” and further, was the lone swing vote; Justice Kennedy, the usual swing-vote suspect, obviously chose not to join the Chief on any point. Fifth, the individual mandate was upheld—but not on Commerce Clause grounds. Sixth, the Chief Justice included an in- depth opinion analyzing the individual mandate under the Commerce Clause, even though that issue was upheld under Congress’s taxing power, which made it completely unnecessary to address the Commerce Clause issue. Seventh, the Court held both that the individual mandate “penalty” is not a tax (jurisdictionally speaking) and is a tax (constitutionally speaking). This, of course, is a distinction only a lawyer can love. Eighth, the Medicaid expansion ruling was in many ways a “sleeper” issue that few people paid much attention to prior to the Court’s ruling; yet it produced an unprecedented result, in that the Court found (7 to 2, no less) unconstitutional “coercion” by Congress in connection with financial incentives provided to states under Spending Clause legislation designed to encourage the states to adopt certain programs. Although precious few people predicted this result, and perhaps no one predicted the nuanced outcome of rendering the Medicaid expansion “optional” for states while not striking down those provisions altogether—and, further, preserving every other aspect of the ACA—this ruling could have significant implications for cooperative federalism moving forward, both within and outside of the healthcare industry. Then again, perhaps it will not, as discussed further below. Time alone will tell.

16

Carter G. Phillips and Stephanie P. Hales

The balance of this essay addresses, briefly, the elements of the decision, how it came out, and the impact it may have moving forward for healthcare policy and in other contexts. Particularly for conversations about structural healthcare reform going forward, it is important to understand the metes and bounds of the Court’s decision, and in some ways to try to figure out how it will play out in practice and how long it will survive. To be sure, challenges to the ACA remain ongoing,4 but as its implementation plunges forward in the meantime, with NFIB v. Sebelius in place as the law of the land, the decision is a crucial one for policymakers and constitutional advocates to assess and understand.

The Commerce Clause: Constitutional Avoidance of “the Broccoli Horrible” A. Context for the Individual Mandate A central feature of the ACA, and certainly the headlining issue in NFIB v. Sebelius, is the individual mandate (or as the statute calls it, the “individual responsibility” requirement “to maintain minimum essential coverage”5). This provision requires that everyone—with certain exceptions,6 but virtually everyone—must buy healthcare insurance, or else pay what is called a “penalty,” although the assessment is codified in the Internal Revenue Code. The question is: can Congress impose that requirement in the context of an economy that is about one-fifth driven by healthcare costs7 and under circumstances in which we think everybody would concede that the system, if not broken, is certainly not functioning as efficiently as we might hope that it could? Faced with those circumstances, Congress took what many view as the extraordinary action to impose the individual mandate as part of a complex and intricate legislative package designed to fi x—or at least begin to address—an extremely inefficient and, many would argue, dysfunctional healthcare system. To be sure, the individual mandate is just one component of the ACA, a statute that consists of two public laws8 consuming more than 950 pages in addition to subsequent amendments that have enacted further technical and substantive changes to its provisions. Congress, in other words, was attempting in the ACA to deal with countless parts of the healthcare system, and the result is a complicated, intertwined web of provisions addressing more issues than most can imagine. To name only a small sample, the ACA includes provisions increasing

Postmortem on

NFIB V. SEBELIUS

17

reimbursement for primary care providers in Medicaid; addressing fraud and abuse in government healthcare programs; creating and funding preventative healthcare services and programs; closing the Medicare Part D “donut hole” to increase seniors’ access to affordable prescription drugs; providing grants for minority health workers; amending the Fair Labor Standards Act to require a reasonable break time for nursing mothers; establishing new annual fees and excise taxes for health insurers, pharmaceutical manufacturers, and medical device manufacturers; and authorizing dozens of demonstration projects and other initiatives pertaining to healthcare payment and delivery reforms. Within this context of trying to fi x countless different moving parts, Congress also addressed (again through many provisions, including but certainly not limited to the individual mandate) the current functioning of the health insurance market, which renders coverage unattainable for a substantial proportion of the population due to fi nancial constraints, preexisting conditions, or both. Congress did not seek a “single payer” solution, as some countries have implemented; instead, it sought to preserve the private market to the extent possible, including its feature that most Americans rely on employer-sponsored health insurance, particularly those under age sixty-five (i.e., those who do not qualify for Medicare, the country’s very popular single-payer healthcare system for the elderly and individuals with certain disabilities). In doing so, Congress considered the context and essentially said, “Look, we need to have people in the health insurance market in order to make the system work; health insurance is meaningless if those who are sick are shut out of the market; further, the market cannot be sustained if those who are not sick choose to ‘opt out’ of the market until the point at which they do get sick.” This was not a new idea, and indeed its origins came ideologically from the right, not the left.9 Congress clearly understood that its mandate decision would draw scrutiny, as it included comprehensive fi ndings within the legislation as to this issue’s effect on the national economy and interstate commerce.10 In these findings, Congress stated that “[t]he individual responsibility requirement provided for in [the ACA] is commercial and economic in nature, and substantially affects interstate commerce, as a result of the effects described in paragraph (2)”; paragraph (2) then sets forth, in eight subparagraphs, Congress’s explanation for why the individual mandate substantially affects interstate commerce.11 As a further point, the legislation adds that “[i]n United States v. South-Eastern Underwriters Association (322 U.S.

18

Carter G. Phillips and Stephanie P. Hales

533 (1944)), the Supreme Court of the United States ruled that insurance is interstate commerce subject to Federal regulation.”12 These provisions plainly are designed to justify the individual mandate as a valid exercise of the Commerce Clause under existing Supreme Court precedent.13 And, Congress added, the mandate would be enforced through what the legislation calls a penalty. In describing the consequence of noncompliance as a penalty and not a tax, Congress apparently was not anticipating the argument upon which the Court ultimately would uphold the individual mandate—Congress’s taxing authority. We address the tax issue further below. To stay with the Commerce Clause for the moment, however, the key question is: can Congress impose the individual mandate as a valid regulation of commercial and economic activity under Article I, Section 8, Clause 3 of the Constitution? B. A Brief Overview of Commerce Clause History As two students of constitutional law (or so we thought), we both learned from the Chief Justice’s decision that what we thought we knew about constitutional law perhaps is not as obvious as we had come to believe. In particular, we would have thought that under the Court’s decision in Wickard v. Filburn,14 decided more than sixty-five years before NFIB v. Sebelius and still on the books, there was no question that Congress had the authority to impose the individual mandate as codified—because, frankly, there are effectively no limits on what Congress can do so long as it is regulating economic activity. In Wickard, a farmer, Roscoe Filburn, was growing wheat on his farm for his own family’s consumption. The U.S. government had established limits on wheat production as part of a scheme for controlling wheat supply and therefore prices during the Great Depression. Filburn was growing more wheat than the laws permitted and was ordered to destroy his crops and pay a fi ne, even though he was growing the excess wheat for his own consumption and had no intention of selling it. In other words, Filburn argued, his activity was neither “interstate” nor “commerce.” (He was “inactive” in the commercial wheat market.) The Court did not agree, fi nding instead that Filburn’s wheat-growing activities affected the amount of wheat he otherwise would buy on the open market, and thus were affecting interstate commerce. The issue, the Court held, was not whether the activity was “local” but rather whether the activity “exerts a substantial economic effect on interstate commerce.”

Postmortem on

NFIB V. SEBELIUS

19

As an interesting sidelight to this, a scholar on Justice Robert H. Jackson—who wrote the opinion in Wickard for a unanimous Court— found a letter exchange between Justice Jackson and a Seventh Circuit judge.15 The judge had criticized Justice Jackson for his opinion in Wickard, stating that the decision provided no meaningful guidance to the lower courts or to litigants about how to proceed under the Commerce Clause. Justice Jackson reportedly wrote back and essentially said, “Well, that is because I can’t conceive of any way to articulate any limits, and so I leave it basically to the good judgment of Congress to decide ultimately.” And that is effectively what the Wickard opinion reflects—which at least through about the 1990s was the way that nearly everybody thought about Congress’s authority. You could (as many do) question the wisdom of what Congress had done, but so long as the action related to regulating economic activity in some way, it was not for the courts to decide whether it was constitutional or not, at least under the Commerce Clause. Now, there were exceptions to that general rule, most notably articulated in United States v. Lopez (1995) and United States v. Morrison (2000).16 In Lopez, Congress was told that there are limits to what it can do—and, for the fi rst time since the New Deal, a provision was struck down as exceeding those limits.17 In the context of that case, the Court held that Congress could not, under the Commerce Clause, prohibit a person from possessing a gun within a certain distance of a school. The theory was that no form of economic activity was involved, and that the relationship between the conduct and interstate commerce was so attenuated as to be insufficient; to reach the necessary link would require “pil[ing] inference upon inference in a manner that would . . . convert congressional authority under the Commerce Clause to a general police power of the sort retained by the States.”18 Therefore, the Court held (in a 5 to 4 decision authored by then– Chief Justice Rehnquist) that this was too far to go and struck down the relevant statute.19 Even in doing so, however, the Court upheld the general framework of Wickard—namely, that there are three broad categories of activity that Congress may regulate under the Commerce Clause: (1) the channels of interstate commerce; (2) the instrumentalities of interstate commerce, or persons or things in interstate commerce; and (3) activities that substantially affect or substantially relate to interstate commerce.20 Five years later, the Court struck down parts of the Violence Against Women Act of 1994—in particular, the law’s provision for a federal civil remedy to victims of gender-based violence, again on the basic theory

20

Carter G. Phillips and Stephanie P. Hales

that this was a noneconomic activity with at best an attenuated and indirect effect on interstate commerce, and therefore that was too far for Congress to go under the Commerce Clause. In the wake of Lopez and Morrison, the framework of Wickard and decades of its progeny remained in place, including the “substantial effect on commerce” standard. Even so, everyone began to accept the idea that “noneconomic” activity that is too attenuated to affect interstate commerce is outside the bounds of Congress’s authority (i.e., not “substantial” enough). That meant, however, that virtually everything else must be inside the bounds of that authority. The Court reinforced this idea in its recent decision dealing with marijuana, Gonzales v. Raich (2005).21 In Raich, the Court relied on Wickard in upholding the federal government’s power to prosecute individuals who grow their own medical marijuana pursuant to state law, without selling or distributing it in any way or placing it at any point in the channels of interstate commerce. The Court stated that Wickard “establishes that Congress can regulate purely intrastate activity that is not itself ‘commercial’, in that it is not produced for sale, if it concludes that failure to regulate that class of activity would undercut the regulation of the interstate market in that commodity.” [545 U.S. at 18.] In a concurring opinion in Raich, Justice Scalia stated his agreement that Congress could prohibit intrastate growing and use of marijuana, because Congress “could reasonably conclude that its objective of prohibiting marijuana from the interstate market ‘could be undercut’ if those activities were excepted from its general scheme of regulation.”22 Because the home-grown marijuana was usable in interstate commerce (even if not in fact used), Congress’s regulation of it was necessary and proper as an incident of its federal regulatory scheme under the Controlled Substances Act. In NFIB v. Sebelius, however, several Justices again demonstrated a desire to find a “limiting principle” for the Commerce Clause—apparently to ensure that Americans are not someday compelled to buy (or, even worse, to eat) broccoli. C. The Commerce Clause in NFIB v. Sebelius In analyzing the ACA’s individual mandate, the Chief Justice did not follow the framework of prior precedents and their repeated reliance on the “substantial effects” analysis. Instead, he reasoned that there is a fundamental difference between “activity” and “inactivity.” His basic analysis

Postmortem on

NFIB V. SEBELIUS

21

is that in order for Congress to regulate, there must be an activity that is ongoing. Through that lens, he focused on the issue of a specific individual being asked to buy health insurance and decided that there was no “activity” being regulated. Justice Ginsburg’s dissent, in contrast, would have analyzed the question through the lens of the entire market—as was done, for example, in Wickard and in Raich. The fear that the Chief Justice and the dissenters embraced in their opinions is the notion that without imposing this type of limiting principle on the Commerce Clause, there are essentially no limits on what Congress can do. That fear, in turn, gives rise to one of our favorite exchanges in the Court, both at the oral argument and otherwise, which is, of course, “the broccoli horrible” (to use Justice Ginsburg’s description of it). This is the notion that Congress will suddenly insist that everyone start buying broccoli and then, having done so, will probably go the extra mile and insist that we actually eat the broccoli that we are required to buy—and we all know that, once that happens, everything we hold dear in life is jeopardized. So that is the fear, and because of it the Justices must avoid the broccoli horrible. In order to do that, they must devise a limiting principle; here, they identify “inactivity” as that limiting principle. Justice Ginsburg, we think rightly, takes the Chief Justice and four dissenters to task on this on two counts. First, as noted, she takes an “entire market” perspective, as reflected in prior Court precedents, and sees no dearth at all of “substantial effects” on interstate commerce when individuals elect not to purchase insurance yet do, inevitably, wind up needing healthcare services down the road—which they often cannot afford in their uninsured state. In other words, she believes the Chief Justice and the other dissenters have chosen an odd way to look at the issue by focusing on particular individuals and not on the actual problem that Congress was regulating; and as a result of framing the issue in this odd way, they reach an odd answer. Second, she said, even following the notion of “activity” or “inactivity” as a framework, why isn’t the relevant “activity” the active use of self-insurance? Particularly striking was the dissenters’ attack on that argument, as their rebuttal is essentially to say that if “active in the self-insurance market” constitutes an activity, then that is mere “word play.” In all seriousness, that is what lawyers live for. On some level, all we do is wordplay, so that is hardly a criticism of any sort. At any rate, Justice Ginsburg’s view seems far more aligned with Supreme Court precedent, including

22

Carter G. Phillips and Stephanie P. Hales

Wickard and its holding that the farmer growing wheat for his own family instead of purchasing it from the commercial market (i.e., his activity of “self-feeding”) affected the entire market because of the market conditions at the time; in the same way the individual who does not purchase health insurance (i.e., who engages in “self-insuring”) affects the entire health insurance market because of how that market operates. (There are, in fact, a number of ways in which the health insurance market is nothing like, for example, the broccoli market.) In any event, it is rather stunning, given the history of Commerce Clause jurisprudence—including the long-standing Wickard, the more recent Raich, and a host of other cases—that the Court would have struck down the individual mandate on Commerce Clause grounds. Even more surprising in some ways is that the Chief Justice felt the need to write an opinion on the Commerce Clause even though he—and a majority of the Court—upheld the individual mandate based on the taxing authority of Congress. Essentially, the Chief Justice’s articulated position is that he had to decide the Commerce Clause issue, because if he had not, his reading of the statute would have been that the mandate “penalty” was not really a tax; thus, it was only in order to “save” the law under the Constitution that he was required to read the provision in a way that he would not otherwise have read it. Therefore, he had to decide the Commerce Clause issue. Any student of the Court, and likely many casual followers, knows that by and large the Court does not reach out to decide a difficult constitutional issue when there is an easier way to get to the result. Chief Justice Roberts’s detailed Commerce Clause opinion, therefore, strikes us as at least unpersuasive—as does his reasoning on why he had to engage in that analysis in this particular instance. D. Potential Implications With the decision on the books, including the Chief Justice’s Commerce Clause analysis, the question then is: What will the impact of the Commerce Clause be in the future? Will the Court continue to draw this particular line of “activity” vs. “inactivity”—and, if so, how much influence will it have on future reform? Potentially, this decision might not change much as a legal matter, because Wickard and Raich remain on the books. Moreover, as the different opinions in NFIB v. Sebelius show, the difference between “inac-

Postmortem on

NFIB V. SEBELIUS

23

tivity” and “activity” can be a matter of description in many ways. As a practical matter, too, this decision potentially might not change much, because it appears, as discussed below, that Congress can do this type of legislating whenever it wants, as long as it uses its taxing authority instead of the Commerce Clause. Ultimately, our guess is that the Commerce Clause aspect of this decision likely will have more impact outside the healthcare area than inside it, because other issues will raise Commerce Clause questions, and there is now more vitality to the Commerce Clause challenges, or so it seems.

The Taxing Authority: Distinctions That Only Lawyers Can Love While the Court’s decision in NFIB v. Sebelius seeks to impose a limiting principle on the Commerce Clause, it simultaneously affi rms Congress’s taxing authority rather strongly. The decision contains a clear 5 to 4 ruling that the individual mandate resides within the scope of Congress’s taxing authority and therefore is constitutional. Indeed, even the four dissenters never argue that the mandate would fall outside of Congress’s taxing power; instead, they argue that Congress did not, in this case, make it a tax (because Congress consciously called it a “penalty” instead). Congress did, after all, describe the consequences of noncompliance with the mandate in numerous places throughout the ACA—and every one of them describes it as a “penalty”; even so, the “penalty” is codified in the Internal Revenue Code and enforced by the Internal Revenue Service. Further, individuals with income below the federal income tax fi ling threshold are exempt from the penalty, and the mechanism for reporting compliance with the mandate (or for paying in the case of noncompliance) is, in fact, a person’s tax returns. In that sense, the “penalty” has, as Chief Justice Roberts reasoned, at least some indicia of a tax. The other twist, of course, is that Chief Justice Roberts decided that this is both a tax and not a tax. This conundrum arises because of the tax Anti-Injunction Act (AIA) issue raised in this case. Under the AIA, 23 it is not within a federal court’s authority to enjoin taxes before they are collected. As the ACA cases wound their way through the courts of appeals, the Fourth Circuit ruled that the individual mandate “penalty” was actually a tax, and further that the AIA was a jurisdictional provision

24

Carter G. Phillips and Stephanie P. Hales

that deprived courts of subject-matter jurisdiction to consider the constitutionality of the mandate until after the penalty had actually been assessed against a party. 24 In granting certiorari in NFIB v. Sebelius, the Court added a Question Presented specifically to address this AIA argument. Several amici supported the analysis of the Fourth Circuit, arguing to the Court that the AIA is jurisdictional and therefore would bar the Court from considering any challenge to the individual mandate until after collection of a penalty—which, these amici argued, would necessarily be a “tax.” Others contended either that the penalty was not a tax (such that the AIA did not apply at all) or that even if the penalty were a tax, the AIA is not jurisdictional (meaning that the Court could choose to address the constitutionality of the penalty even if the AIA applied to it). The Court, for its part, held that the penalty is a tax but that this was not a problem for AIA purposes, because although the “label” of a penalty cannot control whether the payment at issue is a tax for purposes of the Constitution, it can—and, the Court held, does—control the determination of whether the AIA applies. In other words, the Court reasoned that Congress did not want this payment to be a tax for purposes of the AIA’s statutory bar on courts’ consideration of lawsuits seeking to enjoin taxes, but it would be perfectly happy to have this payment be a tax in order to have its statute (i.e., the ACA and its individual mandate provision) upheld. One could argue in criticism, of course, that this is just mere wordplay again. We might respond here, though, that this is not much of a criticism, because so much of what lawyers do could be characterized, at the end of the day, as mere wordplay. We concede, though, that this likely is a distinction that only a lawyer can love. So for the doctors, nurses, teachers, and all other nonlawyers out there, we apologize for the legal profession’s overreaching in this context (although not for our loving it). In any event, the main point is that the Court seems convinced that the taxing authority under the Constitution extends well beyond the Commerce Clause authority, and further that the individual mandate “penalty” falls squarely within it. Going forward, then, it does seem that, to the extent that Congress wants to continue to be active in this area, it has a fairly easy pathway to get where it wants to go. There is one caveat, though; the Chief Justice did state that at some point taxing can become “coercive,” and in that context will be regarded again as a penalty. How far does that go? Well, that takes us back again to the wordplay and line drawing that lawyers embrace. We all know that the task of answering

Postmortem on

NFIB V. SEBELIUS

25

“How much coercion is too much coercion?” creates litigation, so we warmly thank the Court for that as they and we go forward from here.

The Medicaid Ruling: Knowing “Coercion” When We See It The Court’s “coercion” line- drawing exercise becomes even more interesting in the context of ACA’s Medicaid expansion provisions and, indeed, in the context of “cooperative federalism” more generally. This ruling was, to be frank, astonishing to us, all the more so given that seven Justices signed onto it. Never before had any majority of the Court, let alone a 7-to-2 majority, held that a statute enacted by Congress under the taxing and spending authority was unconstitutional because it was an overly “coercive” fi nancial inducement vis-à-vis the states. The Court has, of course, held that Congress has impermissibly “ ‘commandeered’ a state’s legislative or administrative apparatus for federal purposes,” 25 but those rulings were distinct from a holding under the Spending Clause that Congress has used federal “fi nancial inducements to exert a ‘power akin to undue influence.’ ”26 In other words, while the Court has long recognized that federal Spending Clause legislation potentially could offer incentives so coercive that “pressure turns to compulsion,”27 the Court had never, prior to NFIB v. Sebelius, held that this actually had happened. To take a step back, Medicaid is a program jointly administered by the federal government and the states, which provides healthcare coverage to certain low-income individuals. The federal government sets certain rules that apply to all Medicaid programs nationally, and the states can receive federal matching funds if they agree to abide by those federally established rules in setting up their own state Medicaid programs. These rules include certain eligibility criteria—i.e., standards for which individuals qualify for Medicaid coverage. Since its initial enactment in the mid-1960s (through the same legislation that created the Medicare program), Congress has from time to time amended the federal rules that apply to Medicaid, including those concerning eligibility.28 The Medicaid statute itself states that Congress has reserved the “right to alter, amend, or repeal” any provision of it. 29 Further, state participation in Medicaid is voluntary; a state can either accept the federal rules for the program, agree to follow them, and take the federal funding associated with such acceptance, or reject the Medicaid rules and in so doing reject the associated federal funding. Since 1982, when Arizona became the last state to

26

Carter G. Phillips and Stephanie P. Hales

sign up for this offer from the federal government, all states have participated in Medicaid. ACA’s Medicaid expansion provisions once again updated the federal eligibility standards applicable to the program, requiring that each state Medicaid program must, beginning in 2014, extend Medicaid eligibility to all individuals up to 133 percent of the Federal Poverty Level (FPL). To encourage states to accept this eligibility expansion, the law provides that the federal government will pay for 100 percent of the expansion costs through 2016, with that federal share then decreasing gradually yet slightly, ultimately to cover no less than 90 percent of those costs in 2019 and thereafter. The ACA further provides that if a state chooses not to get onboard with this expansion, then the law permits—although, notably, does not require—the Secretary of Health and Human Services to deprive that state of all Medicaid funding in the future (not just the “expansion” funding). So the question was, once again, is that unconstitutionally “coercive”? Uses of Congress’s taxing and spending authority occur regularly and frequently; the entire entitlement program concept is based on this idea, and all cooperative programs rest on the basic premise that “Congress may attach appropriate conditions to federal taxing and spending programs to preserve its control over the use of federal funds.”30 In a case decided about twenty-five years ago, South Dakota v. Dole,31 the Court considered a challenge to a provision in the Federal Highway Act in which Congress required states to increase their minimum drinking age limits as a condition of receiving some portion of federal highway funds. The Court upheld that provision as “encouragement” to states rather than “coercion,” stating in essence that while there may be some point at which a federal law’s fi nancial inducement under the Spending Clause becomes “coercive” beyond the bounds of constitutional permissibility, such was not the case there. Many years passed, and the Court never held that fi nancial conditions attached to Spending Clause legislation amounted to “coercion” of the states. Frankly, everybody seemed to assume that whatever meaning that dictum may have had, it had lost all practical meaning because there had been too many statutes in the past that seemed to cross the line into “coercion” but had never been subject to serious question. But here in NFIB v. Sebelius, the Court struck down the Medicaid expansion provisions of the ACA, 7 to 2, stating that this expansion “accomplishes a shift in kind, not merely degree”32 and that, even though the amount that the states are

Postmortem on

NFIB V. SEBELIUS

27

being asked to contribute is quite small—i.e., nothing at all in the short term and no more than 10 percent in the long term—the Chief Justice wrote a fairly colorful line equating the expansion provisions with “a gun to the head”33 and stated that “ ‘Your money or your life’ is a coercive proposition, whether you have a single dollar in your pocket or $500.”34 In other words, even if it’s only a dollar in your wallet, “your money or your life” is still your life, and so that command is coercive. That metaphor may prove too much, however. While many states did not like the idea of expanding their Medicaid rolls to this degree (or kind, in the Chief Justice’s view), we do not think that any state really felt as though there was a gun put to its head as a consequence of these provisions, particularly because the statute did not even require that states passing on the expansion funds would necessarily be in a position where they actually would lose all existing Medicaid funding—the law merely gave the Secretary discretion to terminate funding. Nonetheless, with Justices Breyer and Kagan joining the four conservatives and the Chief Justice, the Court struck down the expansion provisions on the ground of excessive coercion. On the other hand, with that decision made, a different majority of five Justices—the four liberals and the Chief Justice— agreed that the remedy for this violation would be a narrow one. Rather than striking down the entire statute on the ground of the constitutional infi rmity, they held that the expansion effectively would be “optional” for states. Each state could decide whether to expand its eligibility rules up to the ACA’s requirements; those that did so could receive the increased federal funding provided for under the law, whereas those that did not would lose out on that slab of new money, although the rest of their existing Medicaid entitlement would not be in jeopardy as a consequence. 35 Again, you have to love the Court. (Or at least we do.) “Coercion” is not a concept that is self- defi ning. So for people who make a living in this context, where wordplay really does count, this is a great decision. For purposes of healthcare policy moving forward, however, and how to reform Medicaid and virtually every other entitlement program, this could be a bit of a nightmare, even if only because the contours of the consequences remain uncertain. This is true not just for Medicaid and other healthcare programs, but also for any program based on cooperative federalism. In each context you will have the whole analysis of what the next steps will be and the question, at what stage do those steps become too coercive to pass constitutional muster? Policymakers and also Congress will have to be aware of that going forward.

28

Carter G. Phillips and Stephanie P. Hales

As a practical matter, the Department of Health and Human Services has strongly encouraged states to accept the ACA Medicaid expansion provisions and funding, and has told the states that there is no “deadline” by which they must declare whether they will do so or not. Thus, their status remains uncertain and in flux. Even so, several states have announced decisions or declared their intent regarding the expansion (and a number of Washington-based think tanks have compiled charts and interactive websites detailing the status of such decisions and declarations). 36 As of April 1, 2013, more than half of the fi fty states and the District of Columbia have announced that they are expanding or leaning toward doing so; notably, a number of these are states with Republican governors or states that vocally opposed the Medicaid expansion, including some that joined the multistate lawsuit against it. Perhaps the Court was right then, that this fi nancial inducement is “coercive”—if, that is, “coercive” means the offer is simply too good to refuse, even when it is optional. * * * The Court’s analysis in NFIB v. Sebelius has already inspired countless pages of commentary on its meaning both now and moving forward—for the ACA, for healthcare reform more generally, and for Supreme Court jurisprudence and even judicial behavior more broadly. The case and decision have been compared to Bush v. Gore, for example, in terms of a highly publicized case involving a high-profi le issue affecting the president, the Congress, the states, and the public at large. Many commentators have focused on Chief Justice Roberts: Did he view this decision as defi nitional to his legacy as Chief Justice? Did he switch his vote at the last minute? Was he strategic in setting out his constellation of views in this case, rather than sincere in his argument? For our part, we have difficulty psychoanalyzing the Chief Justice in terms of how he came out in this particular case. Part of the problem likely is as much a function of timing as anything else. This case was argued at the end of March, and the Court has an unbending rule that all cases for each Term conclude by the end of June. That rule compressed the decision-making process extraordinarily. To suggest a sort of stealth strategic approach on the Chief Justice’s part, an almost Machiavellian approach as some describe it, seems to require a lot of “plotting” in a fairly short period of time, as opposed to the Chief Justice’s simply

Postmortem on

NFIB V. SEBELIUS

29

concluding at the end of the day that he thought the statute really was constitutional,and should be upheld, then trying to develop the best rationale that he could to justify it and from there (if you accept the insider reports) trying to convince at least Justice Kennedy to go along with it. That seems to be a perfectly plausible way of thinking about the situation and the ultimate disposition by the Court, although of course we have no way of knowing. Ultimately, this case, like Bush v. Gore, focuses attention on the Court in a way that may be horribly misleading, because 99 percent of the Court’s work deals with complex ERISA questions, environmental issues, or other problems that nobody sees or pays much attention to, let alone understands. Those issues are probably where the Court is at its best, although the high-profi le cases, including this and, to take another example, the same-sex marriage cases argued nearly exactly one year after the oral arguments in NFIB v. Sebelius, shine a spotlight on the Court that inspires legalese-infused public debate and discussion. Whether this spotlight is healthy for the Court or for the public is itself subject to debate and differing views, which we will not attempt to address here. We will conclude only by saying that with its many remarkable elements and powerful impact on a landmark piece of healthcare reform legislation, NFIB v. Sebelius is a critical decision for health policy makers and for anyone concerned with or interested in the line- drawing exercises through which the Court continues to check and balance the contours of Congress’s authority. These may be distinctions that only lawyers can love, but they nevertheless affect us all.

Notes 1. 567 U.S. ____, 132 S. Ct. 2566 (2012). 2. Joined by Justices Breyer, Kagan, and Sotomayor. 3. Justices Scalia, Kennedy, Thomas, and Alito. 4. Numerous challenges are active in the courts, including an Origination Clause challenge, attacks on the so- called contraception mandate that relates to the law’s preventive care coverage provisions, and questions of whether the law’s federal premium subsidies are available to individuals who enroll in federally facilitated exchanges as well as those who enroll in state-run exchanges. Parts of the ACA are at issue in Congress, too, with legislation introduced to repeal various provisions, including the Independent Payment Advisory Board and the excise tax on medical devices.

30

Carter G. Phillips and Stephanie P. Hales

5. See Pub. L. No. 111-148 § 1501(b), 124 Stat. 119 (codified at Internal Revenue Code § 5000A). 6. See Pub. L. No. 111-148 §§ 1411 § 1501, 124 Stat. 119, (2010). For example, the law provides for exceptions to the individual mandate for religious objectors, individuals not lawfully present in the United States, and incarcerated individuals. It also provides for exemptions from the penalty for those who cannot afford coverage, taxpayers with income below the fi ling threshold, members of Indian tribes, those who have received a hardship waiver, and those who were not covered for a period of less than three months during the year. 7. See “Health Care Costs To Reach Nearly One-Fifth of GDP By 2021,” Kaiser Health News, last modified June 13, 2012, http://www.kaiserhealthnews.org / Daily-Reports/2012/June/13/health-care-costs.aspx. Healthcare costs consumed 17.9 percent of GDP in 2011. Id. 8. The Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119 and the Health Care and Education Reconciliation Act of 2010, Pub. L. No. 111-152, 124 Stat. 1029. 9. See Stuart M. Butler, “Assuring Affordable Health Care for All Americans” (lecture, Heritage Foundation, October 1, 1989) (including a provision to “mandate all households to obtain adequate insurance”). Multiple sources have reported on these conservative origins of the individual mandate, including Fox News. See “Individual Health Care Insurance Mandate Has Roots Two Decades Long,” Fox News Politics, last modified June 28, 2012, http://www.foxnews.com/politics/2012/06/28 /individual-health-care-insurance-mandate-has-long-checkered-past/. 10. Pub. L. No. 111-148 § 1501(a). 11. Id. § 1501(a)(1)– (2). 12. Id. § 1501(a)(3). 13. See United States v. Lopez, 514 U.S. 549 (1995) (specifically examining, as a key factor in Commerce Clause analysis, whether there had been congressional fi ndings of an economic link supporting the authority of Congress to enact the law at issue); United States v. Morrison, 529 U.S. 598 (2000) (same). 14. 317 U.S. 111 (1942). 15. The judge was Sherman Minton, who later went on to become a Supreme Court Justice himself. 16. 514 U.S. 549 (1995); 529 U.S. 598 (2000). 17. In full disclosure, Carter Phillips represented Lopez in that case. 18. 514 U.S. at 567. 19. The statute was 18 U.S.C. § 922(q), part of the Gun-Free School Zones Act of 1990. 20. See also Perez v. United States, 402 U.S. 146 (1971). 21. 545 U.S. 1 (2005). 22. 545 U.S. at 24 (quoting United States v. Lopez, 514 U.S. at 561). 23. 26 U.S.C. § 7421(a).

Postmortem on

NFIB V. SEBELIUS

31

24. Liberty Univ. v. Geithner, 671 F.3d 391 (4th Cir. 2011), remanded, 568 U.S. ____, 133 S. Ct. 679 (2012), on remand sub nom. Liberty Univ. v. Lew, 733 F.3d 73 (2013). 25. NFIB v. Sebelius, slip op. at 47 (citing Printz v. United States, 521 U.S. 898 (1997) and New York v. United States, 505 U.S. 144 (1992)). 26. Id. (citing Steward Machine Co. v. Davis, 301 U. S. 548, 590 (1937)). 27. Id. (citing Steward Machine, supra note 26). 28. See NFIB v. Sebelius, slip op. at 41–45 (Opinion of Ginsburg, J.). 29. 42 U.S.C. § 1304 (cited in NFIB v. Sebelius, slip op. at 39, 55) (Opinion of Ginsburg, J.). 30. NFIB v. Sebelius, slip op. at 49. 31. 483 U.S. 203 (1987). 32. Slip op. at 53. 33. Id. at 51 (“In this case, the fi nancial ‘inducement’ Congress has chosen is much more than relatively mild encouragement—it is a gun to the head.” (internal quotation marks omitted)). 34. Id. at 52 n.12. 35. You might wonder why the Court would reach out to decide this issue in this particular case, rather than waiting to see if any state actually would make a decision to take a pass on the expansion funding, and if so whether it would, in fact, be punished by the Secretary with a full withdrawal of all existing Medicaid funding. In other words, was this constitutional issue even ripe to decide in this case, absent a true concrete dispute? The Court did not deal with that issue, so we do not know the answer. 36. See, e.g., “Where Each State Stands on ACA’s Medicaid Expansion,” The Advisory Board Company, http://www.advisory.com/Daily-Briefi ng/2012/11/09 /MedicaidMap, last modified Mar. 4, 2013 (reporting where each state currently stands on the expansion “[b]ased on lawmakers’ statements, press releases, and media coverage”); “Status of the ACA Medicaid Expansion after Supreme Court Ruling,” Center on Budget and Policy Priorities, http://www.cbpp.org/files/status -of-the-ACA-medicaid-expansion-after-supreme-court-ruling.pdf, last modified March 13, 2013 (“Each state is categorized based on statements and other indications from key policy makers—including governors, Medicaid directors, or legislative leadership.”).

Chapter two

Federalism, Liberty, and Risk in NFIB v. Sebelius Aziz Z. Huq

Introduction

I

n the end, the market had it wrong. On June 28, 2012, Intrade estimated a 70 percent–plus probability that the individual mandate component of the Patient Protection and Affordable Care Act (PPACA or the Act)1 would be invalidated. Reporting the decision’s publication, CNN ran a banner proclaiming invalidation for several minutes before switching content. But if the decision in National Federation of Independent Business (NFIB) v. Sebelius 2 was a shock to mainstream media, it was no less surprising to constitutional scholars. Once its fragmented opinions had been assembled, the decision turned out not to flow simply from earlier precedent. To the contrary, the decision overruled no Supreme Court precedent and has ambiguous downstream consequences for legal doctrine. Its place in constitutional history is accordingly a puzzle. This essay offers an account of how NFIB fits into constitutional tradition. I argue that the NFIB Court’s reasoning takes root in profound (if poorly specified) fi rst- order normative principles about the appropriate role of the federal government, rather than arising from the lawyerly paraphernalia of precedent, rules, and syllogistic reasoning. To this end, I make two claims—one negative, the other positive. The fi rst concerns federalism. The NFIB decision superficially turns on the relationship of the federal government to the states, and so reflects judicial calibration of an appropriate “federal balance.”3 Federalism, of course, has been a

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

33

central term of American political contention since the 1790s.4 By the late twentieth century, “federalism” was invoked most often as a talisman against federal regulation, and as a way to favor state autonomy. 5 On a superficial reading, NFIB seems to advance that decentralizing vision of federalism. But I argue that federalism explains very little of its outcome. Examination of its central holdings suggests that the Court left the federal balance largely untouched. Second, in lieu of federalism, I suggest that a dyad of opposing values—liberty and risk—animates the Court’s ruling. The fi rst half of this pair is familiar; claims about liberty figured large in debates about the PPACA.6 It was less frequently observed in those debates that many species of risk management by the state limit some sort of “liberty” as a cost of spreading or mitigating risk. The ensuing trade- offs implicate questions about when individuals should bear the costs of diffusing shared risks. While framed as a case about constitutional law, the NFIB opinion is, in my view, better glossed as a function of the Justices’ normative judgments about the permissible domain of risk-liberty equilibriums the federal government can strike. It is, in other words, about which risk—and whose risks—the government may appropriately share, and conversely who must be cast to the vicissitudes of chance.

The PPACA In an obvious sense, the PPACA concerns risk and how we manage it. Unique among wealthy democracies, the pre-PPACA United States possessed no “guaranteed health coverage for all (or virtually all) citizens or measures to contain costs at a high level of aggregation.” 7 Consequently, nearly 87 million people were without insurance in the United States at some point between 2007 and 2009. Further, forty-five thousand workingage Americans died annually due to underinsurance.8 This is hardly a surprise. In a market- dominated system, competition predictably drives insurers toward actuarially fair pricing, by which individuals must pay in line with the risks they represent regardless of their ability to do so.9 Growing economic inequality since the 1970s has inevitably cashed out as more uneven distributions of healthcare coverage.10 Despite these high social-welfare costs, serial attempts at healthcare reform since the Truman presidency have foundered in a “policy trap”: “[A]n increasingly costly and complicated system . . . satisfied enough of

34

Aziz Z. Huq

the public and so enriche[d] the health care industry as to make change extraordinarily difficult.”11 A combination of tax subsidies for employerprovided insurance, “universal” benefits such as Medicare, and some measure of “welfare” such as Medicaid left uninsured “a mostly lowincome population with no coherence, organization, or political power.”12 Given this policy dynamic, it was somewhat remarkable reform occurred at all. Even if political constraints did not in the end foreclose reform via conventional legislative tools, they nonetheless shaped it. The PPACA is thus modest along two salient dimensions. First, the Act focused on underinsurance, not cost containment. Second, it made no attempt to recast the existing patchwork of private and public healthcare coverage. It instead accepted and broadened existing founts of coverage. As a result, many dysfunctional elements remain, such as the regressive tax expenditures on employer contributions for medical insurance. These cost the federal government about $177 billion in 2011.13 At its core, the PPACA expands or amends three longstanding elements of American healthcare provision.14 First, roughly 60 percent of working-age Americans have been until now covered by employersponsored health insurance.15 The Act largely exempts this large-group market from new regulations. It does require employers with more than 100 personnel to provide “minimum essential coverage” by 2014.16 The large-group market, though, is already heavily regulated by the national government in order to redistribute risk. The Health Insurance Portability and Accountability Act prohibits group health plans from discriminating on the basis of health factors respecting eligibility, benefits, or premiums.17 Interestingly, this federal nondiscrimination rule injects risk spreading into market provision of health insurance. At the same time, it is not (to date) politically controversial. Second, approximately a fi fth of the public is covered by federal statutory programs such as Medicare or Medicaid.18 The Act “made no fundamental changes to Medicare,”19 but its changes to Medicaid catalyzed public and judicial attention. Medicaid is a cooperative federalism program. Medicaid is jointly funded by the states and the federal government, although precise coverage defi nitions vary by state. 20 States submit a “State Plan” to the Secretary of Health and Human Services for authorization. 21 Until the PPACA, the plan had to cover a series of defi ned groups, including the elderly, disabled, blind, pregnant, and children. 22 The PPACA is the fi fth statutory expansion of Medicaid, 23 and adds a

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

35

new mandatory category to state plans as of 2014. This new category comprises all citizens and legal residents with incomes up to 133 percent of the federal poverty line. 24 It will add to Medicaid largely “non- elderly, non- disabled, low-income single adults or couples without children.”25 The federal government provides full funding for the first three years of this expansion, gradually reducing its contribution to a floor of 90 percent in 2020.26 Under long-standing language in Medicaid’s organic statute, the Secretary of Health and Human Services may cut some or all of a state’s Medicaid funding if it fails to comply with certain obligations.27 The fi nal and least functional element of the healthcare landscape is the small-group and individual market. It is here that the PPACA does most of its work. As many readers will know, it introduces a suite of measures designed to expand and render accessible the private market for health insurance. To this end, it imposes, inter alia, minimum essential coverage and community-rating rules on insurers; a mandate on states to create insurance exchanges to facilitate consumers’ purchases; and an “individual mandate” on lawful U.S. residents and citizens. With exceptions, the individual mandate provision requires any lawfully present person who lacks a defi ned quantum of minimum coverage as of 2014 to pay an escalating series of “penalt[ies]” (or “shared responsibility payments.”)28 Through these measures, the PPACA aims to nudge insurers to expand some coverage to the most needy, while preventing healthy individuals from exiting the insurance pool. The net intended result is a broad, diverse, and actuarially secure risk pool.

The Legal Challenges Within minutes of the PPACA’s enactment, states, individuals, and employers fi led suit challenging the Act.29 The challengers, their counsel, and supporters drew substantially from President Obama’s Republican Party and Tea Party opposition. 30 At the same time, the litigation marked a startling public volte-face for a Republican Party that in 2008 proclaimed (in a party platform no less) that “[j]udicial activism is a grave threat to the rule of law because unaccountable federal judges are usurping democracy.”31 In effect, elements of the Republican Party treated the federal courts as another stage in the federal legislative process on healthcare reform—one step beyond bicameralism and presentment. Whatever other long-term effects the case has, its fi rst consequence

36

Aziz Z. Huq

was thus bipartisan ratification of the federal bench as the Tocquevillian cockpit for resolution of divisive national policy questions. 32 The federal courts, however, are not mere extensions of Congress. Interest groups well positioned to influence elected officials are not necessarily capable of persuading judges. Law and politics are distinct mechanisms for the exercise of power in the United States. Different resources—fiscal and ideological—matter. In Congress, so- called big business, as represented by, say, the Chamber of Commerce, often prevails. In the courts, employers did not fare so well—their challenge to the PPACA’s employer mandate fi zzled. 33 After divergent results in various federal courts of appeals, 34 it was the challenges to the individual mandate and the Medicaid expansion that made it to the Supreme Court. 35 It was states and selfdeclared mavens of liberty who seemed to have caught the Court’s ear, not big business. The bottom line of the ensuing decision is this: the Court upheld both the individual mandate and the Medicaid expansion but invalidated the federal government’s power to punish states’ failure to expand Medicaid by withholding all Medicaid funds. 36 All parts of PPACA went into effect, but the federal government lost a stick to punish recalcitrant states. That simple result, however, obscures complex details. Across 193 pages of the U.S. Reports, nine Justices filed four interlocking, overlapping, and antinomic opinions—an opinion “for the Court” by Chief Justice Roberts (joined in places by Justices Breyer and Kagan); a partial concurrence by Justices Ginsburg and Sotomayor (also joined in places by Breyer and Kagan); a joint dissent from Justices Scalia, Kennedy, Thomas, and Alito; and a short solo dissent from Justice Thomas. Discerning the legally binding “holding” of the case requires some parsing to discern the narrowest ground of decision to which a majority of Justices subscribed. 37 But setting aside some important technicalities not relevant to the present argument, it can plausibly be said that the opinion rests on three central questions: (1) Is imposition of the individual mandate within Congress’s Article I power to regulate commerce? (2) Alternatively, is it within Congress’s power to tax for the general welfare? And (3) Does the Medicaid expansion violate the constitutional principle of federalism enshrined in the Tenth Amendment to the Constitution? Briefly, the answers are probably not, certainly yes, and somewhat. To begin with, the Chief Justice and the joint dissent reasoned that the Commerce Clause—long the most important repository of federal regulatory authority—could not underwrite the individual mandate. 38 Accord-

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

37

ing to the Chief Justice, the federal government has no Commerce Clause power to mandate insurance purchases by otherwise healthy individuals, even if they are likely to need coverage later. Previous cases, however, had permitted regulatory extrusions beyond the Commerce Clause’s scope in light of Congress’s supplemental power to enact “necessary and proper” additions to comprehensive federal regulatory schemes. 39 The mandate was fairly plainly a necessary element to the general scheme of healthcare regulation. Roberts, however, rejected this possibility by stating that Congress’s authority under the Necessary and Proper Clause did not encompass “any great substantive and independent power[s]” such as mandating the execution of a contract where none previously existed.40 A different coalition of five Justices, again with the Chief Justice penning the crucial opinion, then sustained the individual mandate as a valid exercise of congressional taxing authority.41 Roberts semaphored that such authority had limits: a tax tipped over into an impermissible “mere penalty” when it eviscerated the individual’s “lawful choice to do or not do a certain act.”42 Quite when this happens, Roberts declined to specify. The fi nal element of the decision concerned the Medicaid expansion. Again, Chief Justice Roberts’s opinion is useful taken as the dispositive guide. Until 2012, the Court had placed only modest restraints on the federal government’s power to impose conditions on grants to the states.43 These largely pertained to the clarity of conditions.44 This rule treated states as capable bargaining partners, but policed federal shading on statutory deals via a clear notice requirement for conditions that impinged on states’ autonomy over policy decisions or states’ budgets. The notice rule found additional justification as a means of fostering democratic accountability; it lowered the cost to voters of ascertaining when a given rule was mandated by state or federal law. Under these precedents, the Medicaid expansion seemingly presented no constitutional worry. States have been on notice since the program’s inception of Congress’s unfettered “right to alter, amend, or repeal any provision” of the Medicaid statute.45 Nor was there much doubt which level of government should be held accountable for “Obamacare.” But Chief Justice Roberts’s opinion ignored the notice rule. He fi rst said that the statutory reservation of authority to change Medicaid could not possibly mean what it said. Instead, Roberts asserted, the Medicaid expansion was “a shift in kind, not merely degree,” between what the Court described elsewhere in the opinion as “old” Medicaid and “new

38

Aziz Z. Huq

Medicaid.”46 The new program, he explained, “is no longer a program to care for the neediest among us, but rather an element of a comprehensive national plan to provide universal health insurance coverage.”47 Citing the magnitude of Medicaid spending as a proportion of state’s budgets (about a fi fth), Roberts proclaimed that the power to withdraw all Medicaid funding for failure to participate in the 2011 expansion was a “gun to the head” and “economic dragooning that leaves the States with no real option but to acquiesce.”48 Again without specifying any analytically crisp rule, he held that the power to withdraw all Medicaid funding was unconstitutionally coercive and therefore could not be exercised.49

Estimating the Footprint of NFIB v. Sebelius Not since the New Deal has the Court seriously grappled with the constitutionality of a major federal social program. 50 So anticipation of the decision ran high. But the decision’s immediate footprint, measured as a matter of constitutional doctrine or public policy, was faint. The Court overruled no earlier precedent, even though it had previously issued rulings upholding sweeping New Deal and post– New Deal regulatory programs. To the chagrin of commentators who hoped the Court would retrench federal power back to pre– New Deal contours, only Justice Thomas even hinted at a desire to go that far, and then in a conspicuously lonely dissent. 51 As to the cash value of NFIB’s new rules, the decision casts no existing federal program into constitutional peril. The leading legislative proposal that would face potential constitutional objections under the decision is the proposal to substitute mandatory individualized retirement accounts for Social Security— an idea typically offered by rightof- center politicians and think tanks. 52 Even with respect to Social Security privatization, the ironists should hold their guffaws: It is not hard to imagine that any now-unconstitutional mechanism could be replicated with well- designed tax incentives. Perhaps more consequential is the absence of any litmus test for ascertaining what constitutes an impermissibly coercive exercise of federal spending. This means there is no way to predict when any change to conditional spending efforts, including other elements of the PPACA’s amendments to the Medicaid statute, may risk invalidation. Judicial refusal to elaborate a clear rule leaves legislative drafters at sea. Because

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

39

legislators must labor under the burden of large ambiguity as to whether their compromises will hold, some otherwise viable deals will not be made. Some states will not benefit from sought-for federal spending because they fear a “heckler’s veto” by other litigious states. When legislation does emerge, it is likely be delayed and impeded by judicial challenges, frittering away federal dollars on lawyering in lieu of social goods. Ironically, the Chief Justice has elsewhere railed against rules that create open- ended uncertainty. 53 In sum, the NFIB presents something of a mystery. It does not break from, or even really extend, past precedent. It impacts no federal program other than the PPACA immediately—and then only at the margin. At best, its most short-term consequences flow from the absence of a clear rule at certain junctures of the decision, and not the prohibitory content of its announced rules. How can we account for such an opinion? In what follows, I offer an account of its immanent logic, that is, the fi rstorder normative principles that best fit the Court’s judgment. To begin with, though, we must examine and eliminate the most obvious explanation for the decision—federalism.

Federalism as Faux Ami The Court typically frames questions about the scope of federal power as a matter of federalism—that is, the balance of authority between several states and the national government. Vigilant policing of the outer boundaries to Congress’s power is underwritten by the Court’s perception that wayward federal lawmakers have only fragile incentives to honor states’ interests on Capitol Hill. 54 At fi rst blush, the Medicaid holding— framed initially by litigators in terms of states’ rights and the Tenth Amendment—seems squarely a function of judicial recalibration of the federal- state balance. Yet this moves too quickly. In my view, none of the main elements of NFIB are well explained by federalism concerns. To see this, we need to take each holding in sequence. As a threshold matter, neither the Commerce Clause nor the taxing power rulings have any clear effect on the scope of state regulatory authority. At best, the Court preserved a domain of state authority over mandatory purchases and capped one species of taxation. It is hard to see much practical significance in either holding. On the one hand, it is difficult to conjure up state mandatory-purpose

40

Aziz Z. Huq

laws that have free rein now in light of the Commerce Clause holding. But what good is fencing off a domain of independent state authority if it scarcely sees usage? In any case, it will often be feasible to reframe impermissible federal interference with inactivity as valid regulation of an activity. Consider, for example, the PPACA individual mandate. Had it been invalidated, Congress could—partisan politics permitting—have rewritten the law to mandate receipt of healthcare on the possession of continuing adequate insurance coverage. Close, if not perfect, substitutes are thus often available when the constitutional bar on federally mandated purchases kicks in. On the other hand, the Court’s constraint on the taxing power matters only so far as it protects states’ purses by reducing the “crowding out” effect of federal taxes. 55 But even if the Court limited federal taxes on activities, it did not address income or capital taxes. Any effect on federal crowd- out of state taxes from NFIB will consequently be de minimis in character. Additionally, there is less federalism “bite” in NFIB’s conditional spending holding than fi rst appears. Indeed, the Court’s holding may even diminish states’ latitude to strike beneficial bargains with the federal government. After NFIB, the federal government must be leery of entering into long-term cooperative federalism programs pursuant to which states develop implementation apparatuses. It must rationally anticipate being locked into such relationships on judicially specified terms. At least in some cases, Pareto optimal deals may not be struck because of this uncertainty. Some states will lose out because beneficial cooperative federal programs will not be created or expanded. The Court’s claim that states were “coerced” also conceals analytic confusion. In other constitutional cases, the Court tends to disparage individual claims of coercion. 56 As a result, it has developed no defi nition of individual coercion. In NFIB, an inchoate, haphazard concept of individual coercion is transposed mechanically to an anthropomorphized, subnational sovereign. 57 The result inevitably has an ad hoc flavor. So although the Court intimated that states’ settled expectations had been disrupted, 58 it is unclear why. Unlike individuals, states lack psychological dispositions such as anticipation. They also lack a constitutional entitlement to Medicaid funding. To the contrary, they can plainly operate without such federal subvention. As recently as 1982, Arizona opted wholly out of Medicaid. 59 A funding cutoff now would leave a state with two options: raise taxes or allow the poor to go without healthcare. Why should states be spared this fundamental public choice? Why in particu-

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

41

lar should deregulation-minded Republican governors who resisted the PPACA benefit from an entitlement to federal subsidies for welfare programs? Isn’t this to allow them to have their deregulatory cake while also “consuming” federal funding for regulation? Even if we assume that states could be coerced, another problem then comes into view. Judicial protection of states against coercion rests on the assumption that states are not full sovereign entities capable of contracting freely with the federal government. Superseding a notice regime with a coercion regime, the NFIB Court treated states as callow naïfs unable to navigate the political world. Rather than treating them with “dignity,”60 the NFIB Court took states as wards. In expressive and substantive terms, then, the conditional-spending holding in NFIB cannot be glossed in terms of federalism any more than the Commerce Clause and taxing power holdings can. Another analytic frame is needed to explain the decision.

The Liberty-Risk Dyad A central role of government is managing risk through prevention, risk shifting, and risk spreading.61 Since the Founding, governments in the United States have installed risk-management policies, from limited liability, early banking regulation, and bankruptcy laws to intellectual and real property rights.62 In recent decades, the federal government has experimented with new institutional forms to produce security against pandemic illnesses, terrorist attack, natural catastrophes, and economic shocks.63 Yet risk regulation has always been controversial.64 It often entails taxes, liabilities, or penalties on third parties better able to mitigate or absorb risk. The marginal social welfare cost of suffering different risks varies immensely by wealth and social circumstances. As a result of these dynamics, the choice of which risk to address (and how to do so) necessarily has redistributive consequences. Concomitantly, risk regulation is a focal point for interest-group activity and intensive ideological investments aimed at either legitimating or discrediting the federal regulatory state. Antistatist, laissez-faire intuitions, coupled with the nostalgic invocation of a prelapsarian smaller state, persist in American political culture. NFIB is profitably understood as an attempt to constitutionalize the fraught question of how to calibrate the federal role in managing risk

42

Aziz Z. Huq

in relation to lost liberties. Although the Constitution contains no express theory of risk regulation, each of the three elements of the NFIB decision can be interpreted as an attempt to erect boundaries around the state’s operation as risk manager—to calibrate, that is, the domain of permissible equilibriums between social risk and individual liberty. Rather than grounding this analysis in a simple or familiar social welfare function, however, the Court’s analysis employs categories drawn from American history to articulate a normative vision of what the federal government can and cannot do. Consider fi rst NFIB’s Commerce Clause and taxing power holdings. Both turn on a judgment about the appropriate quantum of social risk that can be assigned to an individual. The axial moment in Roberts’s Commerce Clause reasoning is his assertion that a federal power to mandate purchases falls outside both the Commerce Clause and the Necessary and Proper Clause powers because it is a “great substantive and independent power.” Set aside for a moment the peculiar genealogy of that last phrase, which has not figured in the U.S. Reports for almost two hundred years.65 Focus rather on the question why a power to mandate purchases should fall into the category. Only two years before NFIB, the Court had read the Necessary and Proper Clause to extend other enumerated powers to reach a confi nement power even in the absence of any criminal conviction.66 Since this earlier ruling was not overruled, NFIB’s conclusion implies that it is a “great substantive and independent power” to require a person to enter a contract, but not to lock that same person up without a criminal trial. To say the least, the intuition here is not obvious. Nothing in the Constitution’s text predicts it. Nor is it obviously justified in welfarist terms. The basic idea that decisions about private contracting relationships are beyond the reach of the state, however, is not unfamiliar to students of American legal thought. In the postbellum “classical” period of American legal thought, scholars and judges refi ned a notion of the “liberty of contract.” State and federal courts then pressed this into service to resist Progressive Era and New Deal regulation of a redistributive bent.67 Under the liberty- of- contract rubric, judges struck down minimum wage and maximum hour laws, labor laws, and union-protective measures. According to one of its theorists, William Graham Sumner, liberty of contract reflected and promoted a “society of free and independent men, who form ties without favor or obligation, and co- operate without cringing or intrigue . . . [that] gives the utmost room and chance for individual

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

43

development, and for all the self-reliance and dignity of a free man.”68 That is, it carved out a space of atomized, individual liberty against the emasculating risk-spreading depredations of the state. The analogy between the NFIB opinion and old-style liberty of contract is, to be sure, inexact. The Court in 2012 did not invalidate a legislative prohibition on contracting, as in earlier cases. It struck down a legislative mandate to contract. But both liberty to contract and liberty from contract are plausibly understood as rooted in the same ideal— Sumner’s “society of free and independent men.” Sumner implicitly contrasted a masculine norm of rugged, autonomous personhood with a vision of government risk-managers as the fons et origo of emasculating “ties [of] favor [and] obligation.” Although there is little chance that the Court will return soon to a full-blooded embrace of liberty of contract, the NFIB decision demonstrates that latter idea may still inform judicial approaches at the margin in cases where government risk regulation can be framed as constitutionally doubtful. The NFIB Court’s taxing power logic has a similar cast. The Court does not question the federal government’s power to impose taxes, but intimates that it cannot use this power to mandate the internalization of all spillover effects of risky individual behavior.69 The Court flags an outer boundary at which personal autonomy trumps collective social welfare. Each of us must then stand or fall on our own. To many, the Court’s insistence that the Constitution protects the wealthy and healthy from being required to contribute to mitigating the health perils of indigence will seem ungrounded in constitutional text and morally obtuse. But perhaps that is the point. It is harsh medicine indeed for a people to be “free and independent.” Just like the Commerce Clause and taxing power elements of the NFIB decision, the Court’s treatment of the Medicaid expansion rests on an implicit normative account of constitutionally permissible risk-liberty trade- offs. Again, this normative theory is not derived from the Constitution’s text. Again, it seems guided by unstated assumptions about the moral desert of different recipients of government largesse. Again, the Court’s normative boundaries have a somewhat gendered cast. There are two risk-related dynamics at work in NFIB’s conditional spending holding. First, recall that the NFIB Court took states to be incapable of freely contracting with the federal government, instead treating them as quasi wards of the Court. This holding shelters not only states’ purses (by allowing them to resist changes to Medicaid while maintaining

44

Aziz Z. Huq

historical levels of federal support), but also state politicians (who could avoid potentially hard choices between raising taxes and slashing benefits). In effect, the Court gave states and their elected leaders a species of constitutionally embedded insurance against future risk of federal policy change. Even if a statute plainly warns that a cooperative program may change over time, the Constitution as read in NFIB requires the lion’s share of change-related risk to fall on the federal government, not the states. Like any beneficiary of a compulsory insurance policy, state officials likely will fall prey to moral hazard—here, by shading on the new Medicaid program or scanting other federal requirements. This compulsory insurance term thus redistributes risk from state governments to individuals. It exposes the most impoverished and vulnerable slices of the population to greater expected health risk, while amplifying the political and fiscal capital of state elites. The conditional spending holding implies a second normative judgment about when and how government can legitimately step in as risk manager. A central move in Roberts’s analysis is the assertion that the PPACA effectuated “a shift in kind, not merely degree,” between “old” Medicaid and “new Medicaid.”70 The Chief Justice glossed this assertion by noting that Medicaid, as expanded, was “no longer a program to care for the neediest among us.”71 Of course, given the relatively low floor on eligibility, it is not obvious this is literally true. Further, federal law prior to the PPLACA mandated coverage up to 133 percent of the poverty line for some groups. Why then was extending this higher cutoff to other groups the crossing of a constitutional Rubicon?72 The reasons given in the opinion are unpersuasive. Roberts glossed his “different of kind, not quantity” argument by pointing out that the Medicaid expansion had been enacted as part of a comprehensive healthcare reform. But this is a non sequitur. There is no reason to conclude that the enactment context radically changed the effects of Medicaid reform. Moreover, it is not at all clear how the 2012 change is qualitatively distinct from five earlier changes beginning in 1967, which have included coverage expansions to reach young adults; redefinitions of mandatory benefit packages; and (most recently) major changes to the scope of prescription drug coverage.73 All these had significant fiscal and health consequences. Why then was the 2010 amendment distinct? To understand the Court’s reasoning, we might instead focus on Chief Justice Roberts’s italicized observation that the PPACA required states “to cover all individuals under the age of 65 with incomes below 133 percent of the federal poverty line.”74 The key word is the one that Roberts italicized:

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

45

“all.” The objection here is not that those at 133 percent of the poverty line are not in desperate straits—although that would be a controversial and hardly self- evident claim. Rather, Roberts’s point is that Medicaid had gone from covering largely women and children to covering the poor generally. That is, it was only when Medicaid ceased to concern women and children primarily—when it extended to poor men too—that the program changed in “kind, not merely degree.” The expected gender identity of Medicaid recipients hence seems to do important work in the Court’s analysis. We might gloss this line of argument as follows. States, the Court reasons, have a constitutionally protected expectation of what Medicaid covers. This expectation rests upon a “recognizably gendered vie[w] of what a welfare state should offer”—one that has been extensively analyzed by social historians.75 As summarized recently by one historian, in the twentieth century the United States developed a “discriminatory ‘two-track’ welfare state” that endows “white, male industrial workers and their dependents” with stable entitlement programs, while women and minorities are streamed into fiscally fragile and socially stigmatizing welfare programs.” 76 Notwithstanding its recent vintage, the NFIB Court treated this dichotomized model of welfare as constitutionally protected. It assumed, in other words, that states were entitled by the Constitution to rely on the existence of cooperative welfare programs that encompassed only those traditionally treated as deserving within the moralized, stereotyping lineaments of a gendered welfare state. It is worth noting that different coalitions of the Court have elsewhere warned that rules resting on the “social and economic inferiority of women” receive heightened equal protection scrutiny.77 The Court’s reasoning, in short, not only allocates risk between states and the federal government, but also entrenches hierarchical and subordinating modalities of managing social risk fi rst developed in the early twentieth century. In this fashion, the decision can be understood as enforcing plural normative limits on the federal government’s role as risk manager beyond the liberty of contract.

Conclusion Not federalism but a constitutionalized view of normatively permissible risk-liberty trade- off best explains NFIB. That normative vision is not

46

Aziz Z. Huq

grounded in constitutional text or tradition. It instead reflects laissez-faire instincts and gendered understandings of welfare familiar from the early twentieth century. Designers of future social insurance programs would do well to bear in mind this lingering historical hangover in the federal bench. Others may well reflect on whether it reflects a desirable approach to managing risk in our diverse, hazard-fi lled, and unequal republic.

Notes I am grateful to David Strauss and Alison LaCroix for comments and helpful conversation. The Frank Cicero, Jr. Faculty Fund supported work on this chapter. 1. Pub. L. No. 111-148, 124 Stat. 119 (2010). The PPACA was enacted alongside the Health Care and Education Reconciliation Act of 2010, Pub. L. No. 111152, 123 Stat. 1029. References to the PPACA or the Act encompass both. 2. 567 U.S. ____, 132 S. Ct. 2566 (2012). 3. United States v. Lopez, 514 U.S. 549, 578 (1995) (Kennedy, J., concurring). 4. Alison L. LaCroix, The Ideological Origins of American Federalism (2010). 5. It is hard to generalize here. Some Justices who have fervently pressed federalism values, understood as states’ rights, in respect to state sovereign immunity and congressional power also are exceedingly willing to fi nd that federal law preempts state regulation. 6. See Christopher Schmidt, The Tea Party and the Constitution, 39 Hastings Const. L. Q. 193, 237 (2011). 7. Jacob Hacker, Health Care Reform 2.0: Fulfi lling the Promise of the Affordable Care Act, in Shared Responsibility, Shared Risk: Government, Markets and Social Policy in the Twenty-First Century, 185, 186– 87 (Jacob Hacker & Ann O’Leary eds., 2012). 8. Id. at 189– 90. 9. See generally Jacob S. Hacker, The Great Risk Shift 140–43 (2006) (arguing that adverse selection leads to pervasive healthcare market failure in the United States). 10. Paul Starr, Remedy and Reaction: The Peculiar American Struggle Over Health Care Reform 5 (2011). 11. Id. at 2. 12. Id. at 7. 13. Suzanne Mettler, Reconstituting the Submerged State: The Challenges of Social Policy Reform in the Obama Era, 8 Persp. Pol. 803, 806 (2010). 14. The following relies on Tom Baker, Health Insurance, Risk, and Responsibility after the Patient Protection and Affordable Care Act, 159 U. Pa. L. Rev. 1577, 1580–1607 (2012).

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

47

15. Elise Gould, The Erosion of Employer- Sponsored Health Insurance: Declines Continue for the Seventh Year Running, 39 Int’l J. Health Servs. 669, 669 (2009) (“Employer-sponsored health insurance (ESI) remains the most prominent form of health coverage in the United States, at 62.9 percent of the under- 65 population; however, the rate of this coverage has fallen every year since 2000, when 68.3 percent had ESI.”). 16. See Baker, supra note 15, at 1592– 93 (summarizing PPACA regulation of the large-group market). 17. 29 U.S.C. § 1182 (2006). Healthcare plans are also covered by the Employee Retirement Income Security Act of 1974. 18. Kaiser Family Foundation, Fast Facts, Health Insurance Coverage in the U.S., 2011, available at http://facts.kff.org/chart.aspx?ch=2874. 19. Baker, supra note 15, at 1581– 83 (noting slightly increased progressivity of Medicare fi nancing under PPACA). 20. See Schweiker v. Gray Panthers, 453 U.S. 34, 37 (1981) (discussing scope of state discretion). 21. 42 U.S.C. § 1396. 22. 42 U.S.C. §1396a(a)(10)(A)(I)-(VII). 23. Nicole Huberfield, Elizabeth Weeks Leonard & Kevin Outterson, Plunging into Endless Difficulties: Medicaid and Coercion in the Healthcare Cases 15 (2012), abstract available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2128760. 24. 42 U.S.C. §1396a(a)(10)(A)(VIII). 25. Huberfield et al., supra note 24, at 18. 26. 42 U.S.C. §1397d(y)(1). 27. 42 U.S.C. §1396c. 28. 26 U.S.C. §5000A(b). 29. Complaint, Florida v. U.S. Dep’t. of Health & Human Servs., No. 3:10- cv- 91-RV-EMT (N.D. Fla. Mar. 23, 2010). 30. For enumeration of plaintiffs and counsel, see Ilya Shapiro, A Long Strange Trip: My First Year Challenging the Constitutionality of Obamacare, 6 FIU L. Rev. 29, 33– 34 (2010). 31. Republican National Committee, 2008 Republican Platform, Government Reform, available at http://www.gop.com/wp-content/uploads/2012/06 /2008platform.pdf. 32. Chief Justice Roberts’s averment of “respect for Congress’s policy judgments” at the opening of his opinion, 132 S. Ct. 2566, 2579– 80 (2012), assumes a different cast in this light. 33. See, e.g., Florida ex rel. McCollum v. U.S. Dept. of Health and Human Services 716 F.Supp.2d 1120 (N.D. Fla. 2010) (dismissing challenges to employer mandate). 34. NFIB, 132 S. Ct. at 2580– 81 (describing outcomes in D.C., Fourth, Sixth, and Eleventh Circuits).

48

Aziz Z. Huq

35. 132 S. Ct. 603 (2011) (granting writ of certiorari). 36. NFIB, 132 S. Ct. at 2608. 37. See Marks v. United States, 430 U.S. 188, 193 (1977). 38. NFIB, 132 S. Ct. at 2585– 93 (Roberts, C.J.); id. at 2644– 55. For a discussion of the Commerce Clause, see Aziz Z. Huq, Tiers of Scrutiny in Enumerated Powers Jurisprudence, 80 U. Chi. L. Rev.575 (2013). 39. The leading case is Gonzales v. Raich, 545 U.S. 1 (2005). 40. NFIB, 132 S. Ct. at 2591– 92. 41. Id. at 2593– 2601. 42. Id. at 2599– 2600. 43. South Dakota v. Dole, 483 U.S. 203 (1987). 44. Sossamon v. Texas, 131 S. Ct. 1651, 1658 (2011). 45. 42 U.S.C. §1304. 46. NFIB, 132 S. Ct. at 2605- 06. 47. Id. 48. Id. at 2604- 05. 49. Id. at 2641-42. 50. The Court, however, has grappled on several occasions with the constitutionality of major civil rights laws. See, e.g., Heart of Atlanta Motel v. United States, 379 U.S. 241 (1964). 51. NFIB, 132 S. Ct. at 2677 (Thomas, J., dissenting). 52. Hacker, The Great Risk Shift, supra note 10, at 37. 53. Caperton v. A.T. Massey Coal Co., Inc., 556 U.S. 868, 893– 99 (2009) (Roberts, J., dissenting), 54. See Aziz Z. Huq, Does the Logic of Collective Action Explain Federalism Doctrine? 66 Stan. L. Rev. 217, 233-37 (2014) (collecting authorizes). 55. See Brian Galle, A Republic of the Mind: Cognitive Biases, Fiscal Federalism, and Section 164 of the Tax Code, 82 Ind. L.J. 673, 702– 09 (2007). 56. See Janice Nadler, No Need to Shout: Bus Sweeps and the Psychology of Coercion, 2002 Sup. Ct. Rev. 153. 57. For a review of debates on coercion, see Aziz Z. Huq, Peonage and Contractual Liberty, 101 Colum. L. Rev. 351, 367- 70 (2001). 58. NFIB, 132 S. Ct. at 2604- 05. 59. Huberfield et al., supra note 23, at 11 n.71. 60. Judith Resnik & Julie Chi-hye Suk, Adding Insult to Injury: Questioning the Role of Dignity in Conceptions of Sovereignty, 55 Stan. L. Rev. 1921 (2003). 61. Tom Baker & David Moss, Government as Risk Manager, in New Perspectives on Regulation 87 (David Moss & John Cisternino eds., 2009). 62. David A. Moss, When All Else Fails: Government as the Ultimate Risk Manager 2- 9, 295-325 (2002). 63. See Mariano-Florentino Cuéllar, Governing Security: The Hidden Origins of American Security Agencies (2012).

Federalism, Liberty, and Risk in

NFIB V. SEBELIUS

49

64. Baker & Moss, supra note 62, at 296-302. 65. See Andrew Koppelman, ‘Necessary,’ ‘Proper,’ and Health Care Reform, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2010192. 66. United States v. Comstock, 130 S. Ct. 1949 (2010). 67. See Cass R. Sunstein, Lochner’s Legacy, 87 Colum. L. Rev. 873 (1987). 68. William Graham Sumner, What Social Classes Owe to Each Other 23 (Caxton Printers 1986) (1883). 69. Analogously, the Court has held that the Constitution bars tort judgments that aim to fully deter, and thus fully internalize social costs, in circumstances of imperfect enforcement. See, e.g., Exxon Shipping Co. v. Baker, 128 S. Ct. 2605 (2008). 70. NFIB, 132 S. Ct. at 2605- 06. 71. Id. 72. Huberfield et al., supra note 23, at 14. The federal poverty line is not an objective measure of need. Diana Karter Appelbaum. The Level of the Poverty Line: A Historical Survey, 51 Soc. Sci. Rev. 514 (1977). 73. Huberfield et al., supra note 23, at 15–16. 74. NFIB, 132 S. Ct. at 2601 (emphasis in original). 75. Linda Gordon, Social Insurance and Public Assistance: The Infl uence of Gender in Welfare Thought in the United States, 1879-1935, 97 a.m.. Hist. Rev. 19, 19 (1992). 76. Karen Tani, Welfare and Rights Before the Movement: Rights as a Language of the State, 122 Yale L.J. 314, 323 (2012); Linda Gordon, Pitied but Not Entitled: Single Mothers and the History of Welfare, 1890–1935 (1994). 77. United States v. Virginia, 518 U.S. 515, 538 (1996).

Chapter three

The Future of Healthcare Reform Remains in Federal Court Jonathan H. Adler

I

n National Federation of Independent Business v. Sebelius a closely divided Supreme Court upheld nearly all of the Patient Protection and Affordable Care Act (PPACA or the Act) against constitutional attack.1 Perhaps most signifi cantly, the Supreme Court upheld one of the Act’s central and most controversial provisions—a requirement that all Americans obtain “minimum essential” health coverage2—by recasting it as an exercise of the federal government’s taxing power. The only provision of the PPACA to fall was a requirement that states participate in a substantial expansion of Medicaid to continue to receive any Medicaid funding. NFIB will not be the judiciary’s last words on healthcare reform, however. PPACA litigation continues apace and could well increase in the years to come as federal agencies seek to implement this complex and contentious law. Having survived a frontal assault, the PPACA will continue to be the subject of legal attacks. This chapter provides a brief overview of how continuing litigation in federal court will affect the implementation—and perhaps even the ultimate viability—of Congress’s latest and most ambitious healthcare reform effort. First, this chapter surveys the factors that will contribute to a surge of litigation as the PPACA is implemented in the coming years. Implementation of the PPACA has already proven more difficult than most had anticipated, prompting numerous administrative fi xes and delays. Meanwhile, the surge of litigation has already begun.

The Future of Healthcare Reform Remains in Federal Court

51

One of the larger and more significant PPACA implementation challenges will be the establishment and operation of health insurance exchanges in all fi fty states. As the next part explains, this implementation will be complicated by the PPACA’s statutory language and consequent legal challenges to administrative fi xes. The PPACA’s authors hoped exchanges would play a key role in expanding access to affordable health insurance. Yet political miscalculation, drafting compromises, state resistance, and litigation could hinder the exchanges’ viability as a means to expand insurance coverage. Litigation and confl ict are inevitable for any policy reform that touches questions of reproductive healthcare and the sanctity of life, and the PPACA is no exception. The chapter next details the legal challenges to regulations adopted under the PPACA requiring group insurance plans to include coverage for all forms of contraception. These lawsuits are emblematic of the ideological and value- driven litigation that is likely to persist as federal agencies make policy choices about what sorts of healthcare services can or must be covered, under what conditions, and at whose expense. Challenges to PPACA provisions adopted to control healthcare costs could open another front in the legal battle over healthcare reform. Congress created a new federal agency—the Independent Payment Advisory Board (IPAB)—to constrain the growth of Medicare spending. To ensure the IPAB’s effectiveness, the PPACA insulates it from outside political pressure and entrenches its policy recommendations in unusual ways. These provisions, intended to strengthen the IPAB’s ability to constrain costs, could also be the source of legal vulnerabilities as the IPAB’s unique structure and authority raise constitutional questions that may need to be resolved by federal courts. Even the individual mandate could be the source of additional litigation. NFIB upheld the imposition of a tax penalty on individuals who fail to obtain qualifying health insurance coverage. Yet as the next part explains, it may also have constrained the federal government’s ability to use this penalty as a means of combating adverse selection in health insurance markets and exposed future reforms to the threat of further legal challenge. The bottom line throughout is that PPACA litigation is not over; it has scarcely begun.

52

Jonathan H. Adler

A Perfect Storm for Litigation NFIB presented the Court with a facial challenge to key provisions and the statute as a whole. Few other such facial challenges to the PPACA remain. The vast bulk of PPACA litigation going forward will concern the Act’s application and implementation. Many key provisions did not take effect until 2014, and many implementation details have yet to be worked out. The Congressional Research Service projects that federal agencies will be adopting new regulations to implement the PPACA for years, if not decades, to come. 3 With each new regulation and exercise of discretion by federal agencies will come another opportunity for litigation. Various interest groups will challenge agency actions under the PPACA as well as agency compliance with the Administrative Procedure Act. Commentators have raised serious questions about the legality of various administrative fi xes, including delaying implementation of the employer mandate and altering the rules governing health insurance plans for congressional staff.4 Such reforms have prompted legal challenges. 5 Additional constitutional challenges are also pending. Legal challenges against the implementation of large regulatory statutes are inevitable. Two decades after the 1990 Clean Air Act Amendments were adopted, legal challenges to implementing regulations continue to be heard in federal court.6 The PPACA is likely to spur even greater amounts of litigation. Healthcare represents nearly one-seventh of the domestic economy. Any effort to reform this sector necessarily creates winners and losers. With so much money on the table, litigation is inevitable as various interest groups seek to protect their gains, recapture losses, or seek out new rents within the PPACA’s healthcare regime. That the Department of Health and Human Services (HHS) has needed to adopt ad hoc fi xes to various provisions of the law only increases the likelihood of additional litigation. The economic incentives are substantial, but economic interests will not be the only driver of additional PPACA litigation. Even after the NFIB decision the law remains unpopular with a substantial portion of the public, and many Republican politicians are still clamoring for repeal. Ideological objections and partisan opposition to the law fuel litigation beyond that which might be economically justified. Republican state attorneys general along with conservative and libertarian public interest groups continue to seek opportunities to hamper full implementation of “Obamacare.” NFIB did little to quell the broader political debate over the PPACA.

The Future of Healthcare Reform Remains in Federal Court

53

Healthcare reform is inherently more controversial and divisive than many other sorts of large-scale administrative reform efforts. Healthcare reform inevitably encroaches on matters of deep ethical and personal concern for many Americans.7 Government decisions to pay for or subsidize some forms of healthcare and restrict others necessarily implicate contested questions of medical ethics and broader normative debates within society about the nature of life, the importance of individual autonomy, and the role of government in promoting public health and particular visions of individual freedom. This is most apparent in the context of reproductive healthcare and end-of-life decisions, but permeates much of healthcare policy. Even seemingly technical questions about the comparative cost effectiveness of various procedures necessarily implicates these broader ethical debates. As a consequence, healthcare reforms stir the passions and ignite ideological opposition in a way that policy initiatives in many other areas do not—and much of this passion will be channeled into the courts. An increasing array of public interest legal groups across the political spectrum stand ready to file legal challenges on behalf of various political, moral, and religious causes. The PPACA’s scope and complexity also make it particularly vulnerable to legal challenge. Such vulnerabilities were compounded by the unusual circumstances surrounding its passage and the need to resort to the budget reconciliation process—as opposed to a House- Senate conference—to iron out legislative language. The law was rushed to the president’s desk without benefit of the usual review and revision processes that can smooth a statute’s rough edges. Many members also voted on the bill without being fully aware of all that it contained. 8 Two different reform bills initially emerged from the legislature. After each house of Congress passed its reform proposal along party lines, House and Senate negotiators met to negotiate a conference bill. It was not to be, however. Republican Scott Brown won a special election in Massachusetts to replace Edward Kennedy in the Senate, thus depriving Democrats of a fi libuster-proof majority. This forced healthcare reform proponents to abandon their efforts to craft a conference bill. Enacting the PPACA required taking a less traveled path. Lacking a sixty-vote margin in the Senate, reform proponents’ options were limited. The only way to get a bill to the president’s desk was for the House to pass the bill that had already passed the Senate—the PPACA— and then amend it as much as would be allowed under the budget reconciliation process. Reconciliation only requires a majority vote to pass the

54

Jonathan H. Adler

Senate, but may be used only for budget-related measures. This limited the range of amendments that could be offered and constrained lastminute efforts to “fi x” the legislation. As reform advocates noted at the time, this presented a difficult choice: enact a flawed bill with many imperfections or risk enacting no bill at all.9 In this case passing a flawed bill meant enacting a PPACA that would be less effective at expanding health insurance coverage or controlling healthcare costs than its proponents had hoped. Yet that was the choice reform proponents ultimately embraced—even though, as one health law expert noted later, it meant enacting a law that no one had intended to become law.10 As a consequence, the PPACA would prove difficult to implement and particularly vulnerable to legal challenge.

Insurance Exchanges and the Consequence of Omission One of the central features of the PPACA is the creation of state-based health insurance exchanges, government-managed marketplaces in which consumers can shop for health insurance plans.11 Exchanges are a key element of the PPACA’s efforts to increase health insurance coverage. These marketplaces are intended to empower consumers to compare competing health plans by providing standardized comparative information about them. At the same time, exchanges facilitate government regulation of insurance markets. Exchanges also play a role in the provision of tax credits and subsidies for insurance coverage and enforcement of the requirement that all but the smallest employers provide health insurance for their employees. The creation of health insurance exchanges in every state is one of the greatest challenges of PPACA implementation. Section 1311 of the Act calls upon each state to create an “American Health Benefit Exchange” (Exchange).12 Section 1311’s requirement that states create Exchanges is not enforceable, however, as the federal government may not commandeer state governments to implement a federal regulatory scheme.13 Rather, the federal government must give states a choice of whether to cooperate. The federal government may offer various inducements for state cooperation, such as fi nancial support or regulatory consequences, but states must be left with a meaningful choice.14 Despite the obligatory language of Section 1311, the PPACA gives states a choice of whether to take responsibility for (and bear the cost of)

The Future of Healthcare Reform Remains in Federal Court

55

operating an Exchange.15 States that agree to set up an Exchange are eligible for start-up funds from the federal government and, as the PPACA is written, low-income residents of such states are eligible for tax credits and cost-sharing subsidies to aid in the purchase of insurance. Should a state refuse to create its own Exchange, Section 1321 provides that the federal government “shall . . . establish and operate” an Exchange in the state’s stead.16 In this respect, the PPACA embodies the sort of “cooperative federalism” common in many federal programs, from environmental regulation to Medicaid.17 As written, the PPACA provides generous tax credits and subsidies to low- and middle-income individuals and families for the purchase of qualifying health insurance plans in state-run Exchanges. Specifically, the Act offers “premium assistance” tax credits to households with incomes between 100 and 400 percent of the federal poverty level.18 These tax credits are refundable, which means that if the credit is larger than a taxpayer’s tax obligations, the taxpayer is eligible for a refund. The Act further offers “cost-sharing” subsidies to help low-income individuals and families obtain more than the minimum level of coverage at no additional cost. The plain text of the law limits these credits and subsidies to those who obtain health insurance through a state-run Exchange, however.19 Section 1401 of the PPACA creates a new section of the Internal Revenue Code—Section 36B—authorizing refundable premium assistance tax credits to aid in the purchase of health insurance in Exchanges. 20 Specifically, Section 1401 authorizes tax credits for each month in a given year in which a taxpayer has obtained qualifying health insurance. As defi ned by Section 1401, a “coverage month” is any month in which the taxpayer is “covered by a qualified health plan . . . that was enrolled in through an Exchange established by the State under section 1311.” 21 The amount of the tax credit is also calculated with reference to a qualifying health insurance plan “enrolled in through an Exchange established by the State under [Section] 1311 of the Patient Protection and Affordable Care Act.” Section 1311 further provides that an “Exchange” must be “a government agency or nonprofit entity that is established by a State.” 22 The cost- sharing subsidies provided under Section 1402 are similarly limited, as this section expressly provides that cost- sharing reductions are only allowed for “coverage months” for which Section 1401’s tax credits are allowed. 23 Section 1321 requires the Department of Health and Human Services to “establish and operate” an exchange in any state that does not choose

56

Jonathan H. Adler

to establish one of its own. 24 A federal exchange is intended to perform the same functions as a state Exchange. While a federal exchange may operate like a state Exchange, nothing in the PPACA authorizes the provision of tax credits or cost-sharing subsidies in federal exchanges. To the contrary, the relevant provisions of Section 1401 only provide for tax credits for the purchase of health insurance “established by a state under section 1311.” 25 Nothing else in the PPACA provides that exchanges established by the federal government under Section 1321 can be treated as Exchanges “established by a state” under Section 1311. Indeed, the PPACA expressly defines “State” as “each of the 50 states and the District of Columbia.”26 The textual limitation of tax credits to state- established Exchanges has implications beyond the affordability of health insurance. Under Section 1513 of the PPACA, employers with more than fi fty full-time employees are required to offer “minimum essential coverage” to their employees. 27 Failure to offer such insurance can subject employers to a $2,000 fi ne for every full-time employee beyond the fi rst thirty employees. 28 Significantly, this penalty is triggered when an employee becomes eligible for tax credits or cost-sharing subsidies by obtaining a qualifying health insurance plan through a state-run Exchange. In effect, the penalty is designed to help offset the federal government’s cost of providing tax credits and cost-sharing subsidies and prevent employers from dropping employee health insurance coverage due to the availability of subsidized insurance in exchanges. Yet if tax credits are unavailable in a given state due to the lack of a state-run Exchange, employers in that state will not face penalties for failing to offer qualifying health insurance. When the PPACA was enacted, it was generally assumed that most if not all states would willingly create Exchanges. 29 As President Obama explained shortly after signing the landmark legislation into law, “by 2014, each state will set up what we’re calling a health insurance exchange.”30 Allowing states to create their own Exchanges, in lieu of a federal exchange or a federally sponsored “public option” for insurance coverage, was intended to ameliorate concerns about a federal “takeover” of healthcare. 31 It would also enable exchanges to take advantage of state experience with health insurance regulation. 32 Few expected that many (if any) states would refuse. States have turned out to be far less cooperative than anticipated. The PPACA provides that the Secretary of Health and Human Services was to determine by January 1, 2013 whether or not states would have a quali-

The Future of Healthcare Reform Remains in Federal Court

57

fying Exchange up and running by 2014. 33 Accordingly, HHS initially set a November 16, 2012 deadline for states to declare their intentions. Yet by that date only seventeen states had indicated they would establish Exchanges under the law. 34 Even given more time, few additional states came forward. 35 Despite the Administration’s best efforts to encourage state cooperation, including an offer to create “partnership” exchanges with state governments, 36 over thirty states refused or otherwise failed to establish their own Exchanges as called for by the Act. Among the reasons offered by noncooperating states are expected operating costs, uncertainty about the legal and technical requirements HHS will impose, and skepticism that state officials would really be in control of state Exchange operations. “State authority to run a health insurance exchange is illusory,” Pennsylvania Governor Tom Corbett explained, warning that cooperating states “would end up shouldering all of the costs by 2015, but have no authority to govern the program.”37 In some states, political opposition to the PPACA also remains substantial, precluding state officials from cooperating with the Act’s implementation. 38 Faced with the prospect that widespread state refusal to establish Exchanges under the PPACA would make tax credits and cost-sharing subsidies unavailable in much of the country, the Internal Revenue Service sought to fi x the problem by reinterpreting (some would say disregarding) the relevant statutory language. In May 2012, the IRS adopted regulations concerning the availability of health insurance premium tax credits under the PPACA. 39 Under the IRS rule, taxpayers would be eligible for tax credits (and, as a consequence, cost-sharing subsidies) upon purchase of a qualifying health insurance plan without regard to whether the plan was obtained through a state-based Exchange under Section 1311 or a federal exchange under Section 1321. In response to concerns that such a rule would extend eligibility for tax credits beyond what was authorized by the PPACA, the IRS commented: The statutory language of section 36B and other provisions of the Affordable Care Act support [sic] the interpretation that credits are available to taxpayers who obtain coverage through a State Exchange, regional Exchange, subsidiary Exchange, and the Federally-facilitated Exchange. Moreover, the relevant legislative history does not demonstrate that Congress intended to limit the premium tax credit to State Exchanges. Accordingly, the final regulations maintain the rule in the proposed regulations because it is consistent with the language, purpose, and structure of section 36B and the Affordable Care Act as a whole.40

58

Jonathan H. Adler

No other explanation was offered in the Federal Register. Although commentators had argued that the express language of the PPACA limits the availability of the premium tax credits to those who enroll in qualifying health insurance plans through an Exchange established by a state under section 1311, the IRS did not identify any statutory language or legislative history to the contrary when it fi nalized the rule. Pressed by members of Congress to offer a more complete justification for its rule authorizing tax credits and cost-sharing subsidies outside of state- created Exchanges, the Department of the Treasury offered a fuller explanation some months later, embracing arguments put forward by some healthcare reform advocates.41 Specifically, the Treasury Department suggested that the language of Section 1321 could be interpreted to make a federally established exchange “the equivalent of a state exchange in all functional respects,” including an Exchange for purposes of determining eligibility for tax credits.42 The basis for this interpretation is that Section 1321 provides that if the HHS Secretary determines that a state will not have a required Exchange—that is, the Exchange required by Section 1311—operational by January 1, 2014, the Secretary is required to “establish and operate such Exchange within the State.”43 “Such Exchange,” according to the Treasury Department, is a Section 1311 Exchange and should be treated as such for the purposes of authorizing tax credits and cost-sharing subsidies. Further, as an Exchange established by the federal government under Section 1321 would be subject to the same requirements as an Exchange established by a state under Section 1311, there would be no reason to limit tax credits to the purchase of qualifying health insurance plans in state-run Exchanges. This would be a plausible interpretation of the relevant statutory text were it not for repeated references to the state role in establishing those Exchanges through which tax credits may be offered. As noted above, Section 1311 expressly requires that an authorized Exchange must be “established by a State.” Section 1304(d) also expressly defi nes “state” as “each of the 50 States and the District of Columbia.” Yet even if one were to set this language aside, as the Treasury Department suggests, and conclude that a Section 1321 Exchange is the equivalent of a Section 1311 Exchange, this is not enough to establish that tax credits are available to offset the costs of qualifying health insurance plans in either type of Exchange. The eligibility requirements for the tax credits are not found in either Section 1311 or Section 1321, but in Section 1401. This section repeatedly

The Future of Healthcare Reform Remains in Federal Court

59

defi nes qualifying health insurance plans eligible for tax credits as those purchased “through an Exchange established by the State under section 1311.” So even if one reads Section 1321 to provide that an Exchange established by the federal government is, for all intents and purposes, a Section 1311 Exchange, a federal Exchange is still not an Exchange “established by the State” as required by Section 1401. The repeated reference to the state role in creating the relevant exchanges is significant.44 Not all references to exchanges in the PPACA reference the state role as Section 1401 does. Section 1421, for example, provides tax credits to small businesses that make nonelective contributions to employee plans offered through an Exchange. Yet whereas Section 1401 repeatedly references Exchanges “established by a State,” Section 1421 only references “Exchanges.” Under the Treasury Department’s interpretation, the additional language in Section 1401 is reduced to surplusage.45 Despite months of prodding, neither the Department of the Treasury nor the Department of Health and Human Services has been able to identify any legislative history that expresses legislative intent to provide tax credits and cost-sharing subsidies in federal exchanges. The only legislative history identified by the federal government in support of its interpretation is the addition of information-reporting requirements when the PPACA was amended during the reconciliation process by the Health Care and Education Reconciliation Act of 2010 (HCERA). These requirements, which expressly apply to Exchanges established under both Section 1311 and Section 1321, include information relevant to the administration of the tax credits, such as information relating to taxpayer eligibility and the receipt of advance payments.46 According to the Treasury Department, the addition of this language “strongly suggests that all taxpayers who enroll in qualified health plans, either through the federally-facilitated exchange or a state exchange, should qualify for the premium tax credit.”47 The problem with this interpretation is that there is still no language in the PPACA that can be plausibly interpreted as authorizing the granting of tax credits and premium assistance in federal exchanges. That Congress chose to adopt a single set of information-reporting requirements for both state and federal exchanges does not suggest, let alone establish, that such exchanges are equivalent in all respects, particularly in the absence of any other language that would establish such equivalence. Given the various functions exchanges are required to perform, including determining Medicaid eligibility and monitoring insurance company

60

Jonathan H. Adler

compliance with applicable regulations, there were ample reasons to enact only one set of reporting requirements for both state and federal exchanges. Further, if the reference to “such Exchange” in Section 1321 truly made federal exchanges established under Section 1321 the full equivalent of state Exchanges established under Section 1311, there would have been no need to reference both sections in the HCERA’s reporting requirement. Had Congress sought to make federal exchanges created under Section 1321 identical to state Exchanges established under Section 1311, it could have done so. Indeed, when Congress amended the PPACA with the HCERA it adopted such equivalence language with regard to territorial Exchanges—expressly providing that Exchanges established by territories would be treated as the equivalent of Exchanges established by states and that tax credits would be available in such Exchanges as well.48 Had Congress meant to ensure that tax credits could be available in federal exchanges, one would have expected it either to adopt similar language to this effect or to remove the “established by a State” language in Section 1401. It did neither, despite making numerous changes to that section through the HCERA. Some commentators have suggested that the failure to authorize tax credits and cost-sharing subsidies in federal exchanges must have been a “drafting error,” as “[t]here is no coherent policy reason why Congress would have refused premium tax credits to the citizens of states that ended up with a federal exchange.”49 After all, to prevent the issuance of tax credits and cost-sharing subsidies in states that refuse to create their own Exchanges is to risk compromising the PPACA’s central goal of expanding healthcare coverage. Yet there are plenty of reasons why some in Congress may have believed a conditional offer of tax credits made sense, even if in hindsight it looks somewhat foolish or even absurd. The most plausible reason for conditioning the availability of tax credits and cost-sharing subsidies on state cooperation would be to provide an additional impetus for states to create Exchanges of their own accord. The authors of the Senate bill in particular wanted states to create Exchanges. Yet as noted above Congress cannot simply tell states what to do. If the federal government wants states to cooperate, particularly at their own expense, the federal government needs to provide some inducement. Financial support of related programs is the most obvious and commonly used incentive (see, e.g., Medicaid), but there is only so much money to go around. The PPACA authorized start-up funding to

The Future of Healthcare Reform Remains in Federal Court

61

help states get Exchanges off the ground, but left states responsible for funding their continued operation. Section 1321 also committed HHS to creating federal exchanges as a fallback if states were late to come around, but the threat of federal action of this sort is not the most powerful incentive for states to act. Nor did the PPACA provide for sustained federal funding of state- operated Exchanges. Another way to encourage state participation, identified as the Senate’s health reform legislation was fi rst taking shape, would be to condition the availability of tax credits or other subsidies on state cooperation. As one prominent health law scholar proposed in 2009, Congress could encourage states to create their own insurance exchanges “by offering tax subsidies for insurance only in states that complied with federal requirements.”50 While less common than threatening to withhold funds (as was done with Medicaid) this approach was not unprecedented. 51 Other draft healthcare reform bills introduced in the Senate contained similar provisions explicitly designed to encourage state cooperation. Moreover, on multiple occasions Congress has offered or withheld tax benefits based upon state cooperation with or resistance to federal policies. 52 Threatening to deprive needy individuals of greater access to health insurance because of state refusal to cooperate may seem like an “absurd” tactic for Congress to use, but it is hardly unprecedented. It can actually be found in other parts of the PPACA, as with the Medicaid expansion. As originally enacted, the PPACA provided that if a state were to refuse to participate in the Medicaid expansion, it would forfeit all federal funding for the expansion as well as all federal support for the preexisting Medicaid program. In other words, Congress threatened to withhold federal support for medical care for some of the most vulnerable populations in a state, were that state to refuse to implement the federally preferred policy. The result of such a sanction would have been to greatly reduce access to healthcare in an uncooperative state, particularly for highly vulnerable populations, thereby compromising efforts to maintain (let alone expand) health insurance coverage under the PPACA. 53 Yet there is no question this is what Congress intended (even if, as a majority of the Supreme Court ultimately concluded, such a threat was unconstitutional on other grounds). Congress decided to pursue the PPACA’s goal of expanding coverage by enlisting states in the cause, and it sought to encourage state participation with incentives, including a threat to withhold funding for benefits to needy populations. Congress did not think any state would refuse the

62

Jonathan H. Adler

Medicaid expansion, just as few considered that states might not be willing to create their own Exchanges. Whether due to the use of conditional tax subsidies or not, most commentators simply assumed that states would willingly create their own insurance exchanges, particularly when the most likely alternative would be a federal exchange—ominously characterized by some as a “federal takeover” of the healthcare system. 54 As it happened, some members of Congress were concerned that states might fail to implement Exchanges or otherwise cooperate with federal healthcare reform. For this reason, some members of the House of Representatives urged the House- Senate conference committee to reject the state-based Exchanges contained in the Senate PPACA in favor of a federally run model that had been included in the House bill. 55 Had Republican Scott Brown not been elected to the Senate, thereby depriving Senate Democrats of a fi libuster-proof majority, the conference negotiators may well have followed this advice. In the end, however, Massachusetts voters took this option off the table. The only way to enact comprehensive healthcare reform was to stick with the Senate bill—and this meant sticking with state-based Exchanges and a conditional offer of tax credits and cost-sharing subsidies. The fate of tax credits and cost-sharing subsidies in states without state-run Exchanges will ultimately be decided in federal court. In September 2012, the state of Oklahoma fi led suit challenging the IRS rule on both substantive statutory and procedural grounds. The suit alleges the IRS rule confl icts with the plain language of the PPACA and that the IRS failed to comply with the Administrative Procedure Act when promulgating the rule. 56 Suits raising similar claims were subsequently fi led in Virginia, Indiana, and the District of Columbia. On July 22, 2014, the U.S. Courts of Appeals for the D.C. and Fourth Circuits issued confl icting decisions on the IRS rule. A divided D.C. Circuit panel concluded the decision to provide tax credits in federal exchanges contradicted the text of the ACA 57 while the Fourth Circuit found that the ACA was sufficiently ambiguous to allow for the IRS’ interpretation. 58 As of this writing, the Indiana suit remains in district court. 59 Some initially thought challenges to the IRS rule were unlikely to proceed due to various jurisdictional concerns.60 As a general rule, taxpayers lack standing to challenge the misuse of federal funds or preferential tax treatment given to others. Were tax credits and premium assistance the only consequence of the IRS rule, there would be no viable litigation.

The Future of Healthcare Reform Remains in Federal Court

63

Yet because the availability of tax credits and cost-sharing subsidies triggers the imposition of penalties to enforce the employer mandate, employers in applicable states should have standing to sue provided they are threatened by these penalties. Some individuals in states with federal exchanges may be able to challenge the IRS rule as well, alleging injury due to the effect the authorization of tax credits and cost- sharing would have on whether given individuals are required to pay the tax penalty for failing to maintain qualifying health insurance under the minimum coverage provision. Some individuals could have standing because the IRS rule deprives them of an exemption from the individual mandate penalty for which they would otherwise qualify. This “affordability” exemption is based upon the out-of-pocket cost an individual would have to pay for qualifying health insurance in relation to that individual’s income. Specifically, if an individual’s “required contribution” exceeds 8 percent of household income, that individual is exempt from the penalty. By providing tax credits in federal exchanges, the IRS rule reduces the out- of-pocket cost of purchasing a qualifying health insurance plan for some individuals from above 8 percent of household income (where the taxpayer would be exempt from the penalty) to below 8 percent, thereby exposing some individuals who do not wish to purchase health insurance to the tax penalty. Therefore, individuals who live in a state that will not establish an Exchange by 2014 and who would otherwise qualify for the affordability exemption in the absence of tax credits would have standing to challenge the rule, provided that they earn between 100 and 400 percent of the federal poverty level, do not receive health insurance from their employer, and would be exposed to the tax penalty due to the availability of tax credits under the IRS rule. Several million Americans satisfy these criteria. Many taxpayers will also suffer injury because the IRS rule will deprive them of the ability to purchase a low- cost “catastrophic” plan, which the law makes available to those over age thirty who qualify for the affordability exemption. The creation of health insurance exchanges is one of the central features of the PPACA. Yet given the way the statute is written and the manner in which many states have responded, it could be difficult for these exchanges to operate in the way that many had hoped. No less significant, the operation of health insurance exchanges and the availability of tax credits and cost-sharing subsidies in states that refuse to cooperate

64

Jonathan H. Adler

with the PPACA are questions that will ultimately be decided by the federal courts.

Confl ict Over the Contraception Mandate Challenges to the IRS rule purporting to authorize tax credits and costsharing subsidies in federal exchanges may be among the most consequential for the ultimate operation of the PPACA, but they may not be the legal challenges that evoke the most popular concern. People care deeply about their healthcare. And some people care even more deeply about healthcare policy when it touches upon questions of sexual morality and reproductive health. Thus of all the decisions implementing the PPACA HHS has made thus far, none has been as controversial as the decision that employer health insurance plans must cover all forms of federally approved contraception, including sterilization and medications that some believe may act as abortifacients. None has been more litigated either. As of April 2014, over ninety separate lawsuits had been fi led challenging the so- called contraception mandate.61 Under Section 1001 of the PPACA, nongrandfathered group health plans are required to cover certain preventative healthcare services, in particular preventative healthcare services for women, without any copayments or other cost- sharing by the insured.62 As implemented by HHS, this requirement was interpreted to apply to all contraception methods that have been approved by the Food and Drug Administration.63 Somewhat controversially, such approved contraception methods include sterilization and some forms of contraception that may prevent the implantation of a fertilized egg or otherwise act as an abortifacient (such as intrauterine devices and the so- called morning-after pill). Such forms of contraception are opposed by some religious groups. The official doctrine of the Catholic Church, for example, prohibits the use of all such forms of contraception. Many Evangelical churches also oppose the use of contraceptive methods that they believe will terminate unborn human life. Failure to comply with the requirement subjects religious employers to substantial liability, however. Specifically, nonexempt employers are subject to a fi ne of $100 per employee, for each day they fail to provide the required coverage.64 In response to religious objections, HHS created a narrow exemption for religious institutions. As promulgated by the department, churches

The Future of Healthcare Reform Remains in Federal Court

65

and other religious entities would be exempt should they meet the following four criteria: 1. The organization’s purpose is the inculcation of religious values; 2. The organization primarily employs individuals who subscribe to the religious tenets of the organization; 3. The organization primarily serves individuals who subscribe to the religious tenets of the organization; and 4. The organization is a nonprofit.65

This exemption did not quiet the controversy over the contraception mandate, however. Under these criteria, many religious institutions, including religiously affi liated schools, universities, hospitals, and social service organizations, would be required to provide insurance that covers contraception methods that are contrary to their religious beliefs. As the head of Catholic Charities USA quipped, “the ministry of Jesus Christ himself” would not qualify under HHS’s criteria.66 Some private, for-profit corporations owned by religious individuals objected as well. HHS tried again to placate religious objections by announcing a one-year enforcement safe harbor and promising to adopt yet another accommodation for religious institutions.67 One accommodation suggested by HHS was to relieve religiously affi liated nonprofit employers from the obligation to provide insurance that covers all FDA-approved contraception methods and instead place the obligation to provide contraception coverage directly on insurers. So, for example, if a Catholic hospital objected to contraception coverage, it would no longer have to pay for insurance coverage that covered such contraception, provided it signed a certification to this effect. The insurer with which the hospital contracted, however, would be required to provide contraception at no charge to either the employer or the insured. Even assuming that HHS has the legal authority under the PPACA to impose such a requirement, it would not solve the problem because many religiously affiliated employers self-insure.68 In such cases, the employer and the insurer are one and the same, so the suggested accommodation would not, in fact, ameliorate the religious employers’ concerns. In addition, some religious employers objected to signing a form that would trigger contraception coverage. HHS faces two additional problems in that (a) this accommodation would not purport to do anything for privately owned, for-profit employers with owners who object to providing such contraception coverage on

66

Jonathan H. Adler

religious grounds, and (b) should HHS back off its initial commitment to ensuring that group insurance plans cover contraception, it would face a new set of objections from reproductive health advocates. Facing the prospect of having to pay for or otherwise provide health insurance coverage to which they object on religious grounds, numerous religious institutions and employers fi led suit against the contraception mandate on both constitutional and statutory grounds. Specifically, they alleged that the contraception mandate violates the Religious Freedom Restoration Act (RFRA) and, more ambitiously, the religion clauses of the First Amendment. Others challenged the sufficiency of the accommodations offered by HHS. The strongest legal argument against the contraception mandate is statutory, not constitutional. Under current doctrine, the First Amendment does not pose much of an obstacle to a general law of neutral application, even if it requires some individuals to engage in actions prohibited by, or refrain from actions compelled by, their religious faith.69 RFRA, however, presents a larger hurdle for the federal government. Under RFRA, the federal government may not adopt a policy that imposes a substantial burden on a person’s religious faith unless that policy represents the least restrictive means of achieving a compelling governmental interest.70 As the Supreme Court has recognized, this means that policies that are fully constitutional under the First Amendment are nonetheless barred under federal law. Moreover, by its express terms, RFRA applies to subsequently enacted legislation, so it governs the implementation of the PPACA. In defense of the contraception mandate, the federal government has argued that requiring group insurance plans to cover FDA-approved contraception methods does not represent a substantial burden on the practice of anyone’s religion because the connection between the religious employer and the use of the contraception methods to which they object is sufficiently attenuated. Setting aside those religious employers that self-insure, the employer is not required to pay for or arrange for contraception because such contraception is only obtained and used as a result of independent choices made by the insured, the insured’s doctor, and the insurance company. Religious employers counter that their objection is to the requirement that they cover such contraception, not that it may be later purchased or used. Lurking in the background of this debate is the question whether the government (and the courts) must

The Future of Healthcare Reform Remains in Federal Court

67

defer to a religious institution’s own conception of what does or does not impose a substantial burden on religious practice. Assuming that the contraception mandate does impose a substantial burden on religious exercise, defenders of the mandate contend it is nonetheless permissible because requiring coverage of FDA-approved contraception advances the compelling state interest in advancing gender equality and protecting women’s health.71 These government interests may well qualify as compelling under current doctrine. The government has also challenged the ability of private, for-profit employers to avail themselves of RFRA’s protections at all. The biggest problem for the federal government’s position is RFRA’s requirement that any burden on religious practice be narrowly tailored and no more restrictive than necessary to advance its asserted interest.72 Those challenging the contraception mandate note that many group health plans—those grandfathered under the law or those offered by smaller employers—are not subject to the mandate. A consequence of this exception is that even with the mandate, millions of individuals will have health insurance that does not cover contraception. Further, opponents argue that the federal government has other ways of expanding access to contraception that would be more effective and impose less of a burden on religious employers. As noted above, the contraception coverage requirement prompted several dozen legal challenges from a range of religious institutions and other objectors to the policy. As these challenges worked their way through the courts, HHS proposed additional measures to address the concerns of religious institutions, specifically by expanding the range of exempt nonprofit religious institutions and formally proposing to accommodate the objections of other religious institutions, such as universities and hospitals, by requiring their insurance companies to provide for separate policies covering contraception. HHS proposed to offset the costs of such plans by reducing the fees for participating in health insurance exchanges, and would address the concerns of self-insuring employers by placing the responsibility for arranging coverage on plan administrators. As before, the HHS proposal made no effort to address the concerns of for-profit employers who have religious objections to providing contraception coverage to their employees. Further, some nonprofit religious employers continue to maintain that the accommodations are insufficient.73

68

Jonathan H. Adler

Several of the legal cases challenging the contraception mandate proceeded quickly, with two challenges brought by for-profit employers reaching the Supreme Court during the 2013–14 term.74 In both of these cases—Burwell v. Hobby Lobby Stores and Conestoga Wood Specialties v. Burwell—the owners of closely held private corporations claimed that being forced to provide insurance that covers all forms of FDA-approved contraception would violate their deeply held religious beliefs and compromise their efforts to run their businesses in accord with their religious beliefs. In a 5-4 decision, the Supreme Court agreed with challengers, concluding both that private, for-profit corporations could raise RFRA claims and that the contraception coverage requirement was not the least restrictive means of advancing the government’s stated interest in expanding cost-free access to contraception coverage. Given that HHS had identified potential accommodations for nonprofit employers, the Court found no reason why similar accommodations could not be made for other religiously motivated employers, including for-profit corporations. The Court did not decide that the accommodation itself complied with RFRA, however, and at the time of this writing several challenges to the accommodation brought by religious nonprofits are pending.75 The heated legal battle over the contraception mandate may also foreshadow legal fights yet to come over the scope of health insurance coverage, the obligations of employers, and access to controversial healthcare technologies and services. Four decades after Roe v. Wade the nation remains fiercely divided on the question of abortion, and many have deep moral convictions as to the appropriateness or even acceptability of modern reproductive technologies and approaches to end-of-life questions. Insofar as the PPACA places the federal government in the position of deciding what sorts of treatments or care can or must be covered by various insurance plans, partisans of these battles will rush to court. Even where little money is at stake, the preferences of those involved in such debates are sufficiently intense to make additional litigation a certainty.

Constitutional Constraints on Controlling Costs Much of the PPACA seeks to expand health insurance coverage. The other major goal of healthcare reform was to tamp down rising healthcare costs. Medical inflation has exceeded overall inflation for most of the past few

The Future of Healthcare Reform Remains in Federal Court

69

decades and the cost of Medicare, in particular, has become a major budgetary concern, deemed “unsustainable” by the General Accounting Office.76 Though medical inflation has slowed in recent years, it is unclear whether this trend will continue. One of the PPACA’s primary cost- control measures is the Independent Payment Advisory Board (IPAB), a new independent agency tasked to “reduce the per capita rate of growth in Medicare expenditures.” 77 The Board consists of fi fteen members appointed by the president and subject to Senate confi rmation. Because Congress has shown itself incapable of enacting (or even allowing) limits on Medicare’s growth, the PPACA shifts this responsibility to the IPAB.78 Barring a sustained slowdown in health inflation, the IPAB must intervene and litigation is likely to follow.79 In any year in which Medicare spending is anticipated to grow by more than 1 percent above GDP, the IPAB is required to develop cost-reduction proposals that will bring Medicare growth rates back into line. 80 The primary means for achieving this goal is proposing measures to constrain outpatient reimbursement rates.81 Not all such measures are on the table, however. Under the PPACA, the IPAB is prohibited from proposing any measure to raise revenues, increase premiums or cost-sharing, limit benefits, or “ration healthcare,”82 yet the law does not defi ne what would constitute prohibited rationing and, as noted below, provides no basis for enforcing such limitations.83 Although the IPAB’s authority is limited to Medicare, supporters hope that its reform proposals will reverberate throughout the healthcare sector. 84 Whatever the IPAB puts forth is fast-tracked into law. Under the PPACA, the IPAB’s proposals are automatically introduced into Congress and take effect if Congress fails to act. Specifically, the HHS Secretary is required to implement the IPAB’s proposals unless Congress promptly enacts an alternative. Should Congress disapprove of the IPAB’s proposals, the PPACA provides that Congress must enact alternative measures by August of the same year that will generate equivalent cost savings. There is nothing at all unusual about delegating the authority to adopt policy measures with the force of law to an executive or independent agency. Yet such authority is typically subject to various procedural requirements, such as those provided under the Administrative Procedure Act for notice-and- comment rulemaking, and is subject to judicial review. Neither is the case with the IPAB. The PPACA imposes no meaningful

70

Jonathan H. Adler

administrative procedures on the Board and expressly precludes judicial review of IPAB actions and subsequent HHS implementation.85 The result is the lack of any meaningful checks should the IPAB exceed the scope of its delegated authority.86 Where Congress disapproves of specific agency action, regular legislation satisfying the constitutional requirements of bicameralism and presentment is sufficient to undo the agency’s work. Again, the IPAB is different, as the PPACA requires a three-fi fths vote in the Senate to waive the requirement that cost- control measures be adopted each year Medicare cost increases exceed the stated target.87 The PPACA also purports to hamstring Congress’s ability to revise the law to alter or eliminate the IPAB. Specifically, the PPACA provides that a joint resolution to repeal the IPAB provisions can only be introduced in January 2017, is subject to special rules governing floor consideration and debate, and can only be enacted by a three-fi fths supermajority in both houses.88 Although such provisions are designed to entrench the IPAB against subsequent political majorities, it is not clear these limitations are enforceable. Article I, Section 5 of the Constitution provides that “[e]ach House may determine the Rules of its Proceedings.” As traditionally understood, this provision prevents one Congress from entrenching procedural rules governing the consideration of legislation and preventing a subsequent Congress from determining its own rules. Should a future Congress exercise its prerogative to consider legislation modifying or eliminating the IPAB, litigation by groups seeking to enforce the law’s limitations may well ensue. The fi rst lawsuit against the IPAB was fi led in 2010, alleging among other things that Congress delegated excessive authority to the IPAB without providing a sufficiently constraining “intelligible principle” to guide the Board’s efforts. Unsurprisingly, this claim was dismissed. 89 It is particularly difficult to challenge an agency’s authority before it has taken any action. Once the IPAB swings into gear, however, additional litigation is likely, particularly once it puts forward specific proposals constraining provider reimbursements or otherwise compromising the economic interests of various providers. Once such suits are fi led, courts will have to confront the lack of administrative procedures governing the IPAB’s activities and the PPACA’s limitations on judicial review. Whether the IPAB will survive such challenges is an open question. Even supporters of the PPACA recognize that the IPAB, as enacted by Congress, represents “a troubling challenge to our constitutional order.”90

The Future of Healthcare Reform Remains in Federal Court

71

Relitigating the Individual Mandate Litigation over implementation of various portions of the PPACA will continue for some time. A wide range of legal challenges is pending in federal court and more have been threatened. Even the individual mandate could end up back in court. Although NFIB upheld the constitutionality of imposing a fi nancial penalty on individuals who fail to purchase and maintain qualifying health insurance, this may not be the last time a federal court has to consider the constitutionality of fi nancial assessments imposed for failing to acquire health insurance. Chief Justice Roberts upheld the assessment in the PPACA as a permissible exercise of the federal taxing power. Yet his decision did not immunize the assessment from all future challenge. “Even if the taxing power enables Congress to impose a tax on not obtaining health insurance,” he explained, “any tax must still comply with other requirements in the Constitution.” 91 Some commentators see this language as an invitation for future challenges to the tax penalty portion of the mandate, perhaps under the Constitution’s Uniformity Clause.92 Others think the Court’s resolution left open the question of whether such a tax could violate liberty interests, equal protection, or the Due Process Clause. Such challenges are speculative, but they will be litigated nonetheless.93 Two other pending suits allege that in enacting the PPACA, Congress violated the Origination Clause, which requires that bills for raising revenue originate in the House of Representatives.94 Neither is likely to prevail. Still more litigation is likely to result should Congress or HHS revisit the operation of the individual mandate to prevent adverse selection and further upward pressure on health insurance premiums. Chief Justice Roberts’s reinterpretation of the PPACA’s requirement that individuals obtain health insurance or pay a tax may have saved the Act’s constitutionality, but it also constrained the minimum coverage requirement’s effectiveness at preventing adverse selection and consequent increases in health premiums. The point of the mandate was to counteract the potential effect of imposing community rating and guaranteed issue requirements on health insurance providers. Under the PPACA, insurers may not deny coverage or charge higher premiums to individuals for health insurance based upon their health status or any preexisting medical conditions. Such a requirement makes insurance more affordable for high-risk individuals while increasing premiums for low-risk individuals. A potential result of such a requirement is a vicious cycle of adverse selection if

72

Jonathan H. Adler

some low-risk individuals react to increased premiums by dropping their insurance—causing further premium increases that could in turn cause more low-risk individuals to drop coverage, and so on.95 The mandate was intended to prevent adverse selection by requiring all individuals to obtain and maintain qualifying health insurance. Such a mandate can only be effective, however, if the penalty for noncompliance is large enough to discourage low-risk individuals from dropping their coverage. That is not the case with the penalty adopted in the PPACA. For many Americans, the penalty for failing to purchase health insurance will be substantially below the cost of purchasing a federally approved health insurance policy, and many of these people are likely to forego obtaining insurance as a result.96 After the NFIB decision, the Congressional Budget Office estimated that approximately six million Americans (of an estimated thirty million who would remain uninsured) will be required to pay the penalty in 2016.97 The most obvious way to address this concern and prevent adverse selection would be to increase the penalty until it was comparable with the out-of-pocket cost of a qualifying health insurance plan. At this point low-risk individuals would have little incentive to forego health insurance. The problem here is that were Congress to increase the penalty substantially, it might no longer qualify as a tax. Under NFIB, the relatively small amount uninsured individuals would be required to pay the government as a consequence of being uninsured was one of the primary factors that led Chief Justice Roberts to conclude the payment constituted a tax.98 A payment that equaled or exceeded the cost of obtaining insurance, on the other hand, could resemble the sort of “ ‘prohibitory’ fi nancial punishment” that would exceed the scope of the taxing power.99 While it is permissible to use a tax to “influence behavior,” Chief Justice Roberts explained, for an assessment to be a “tax” and not an unconstitutional penalty or mandate, it must leave an individual with a meaningful choice and not become “so punitive” that it begins to resemble a punishment or a mandate.100 Chief Justice Roberts’s opinion would seem to prohibit Congress from increasing the tax penalty by any sizable degree, virtually assuring that the minimum coverage requirement will not fulfi ll its intended purpose.101 It also ensures that any increase in the mandate tax would be met with a fresh legal challenge. Absent the ability to increase the assessment on those without health insurance, Congress or the administration may seek out other ways of discouraging adverse selection. According to some

The Future of Healthcare Reform Remains in Federal Court

73

reports, health insurers are encouraging HHS to implement additional measures, such as late-enrollment fees or other requirements, so as to discourage adverse selection, particularly as the mandate is fi rst phased in.102 Should the department seek to implement any such measures, however, legal challenges to its regulatory authority are equally likely.

Conclusion Some expected and many hoped that the Supreme Court’s resolution of the NFIB litigation would put an end to the legal challenges to the PPACA. Yet it was never to be this way. The PPACA is too expansive and significant a statute, affecting too many economic interests and implicating too many political, moral, and ideological divisions within the country for the litigators to stay their hand. As the federal government implements the law in the years to come, its choices will be scrutinized and challenged at every turn. As a consequence, the ultimate shape of healthcare reform will still be decided in federal court.

Notes The author would like to thank Ye Han for his research assistance. 1. 567 U.S. ____, 132 S. Ct. 2566 (2012). 2. 26 U.S.C. § 5000A. The Act exempts some groups from this requirement, including prisoners, undocumented aliens, and those with valid religious objections. Id. at § 5000A(d). 3. Curtis W. Copeland, Cong. Research Serv., R41180, Regulations Pursuant to the Patient Protection and Affordable Care Act (2010) (“[I]t seems likely that there will be a great deal of regulatory activity relating to the many provisions in PPACA for years, or even decades to come.”), available at: http://www.ncsl.org /documents/health/Regulations.pdf. 4. See, e.g., Nicholas Bagley, The Legality of Delaying Key Elements of the ACA, 370 New Eng. J. Med. 1967 (2014); Ron Johnson, I’m Suing Over ObamaCare Exemptions for Congress, Wall St. J., Jan. 5, 2014, http://online.wsj.com /news/articles/SB10001424052702304325004579296140856419808. 5. See Kawa Orthodontics, LLP v. Lew, No. 9:13- cv- 80990 (S.D. Fla. 2013); Johnson v. United States Office of Pers. Mgmt., No. 14- C- 0009, 2014 WL 1681691 (E.D. Wis. Apr. 28, 2014). 6. In 2011 and 2012 alone the U.S. circuit courts of appeal decided over two dozen cases concerning the implementation of the Clean Air Act.

74

Jonathan H. Adler

7. See B. Jessie Hill, What Is the Meaning of Health? Constitutional Implications of Defi ning ‘Medical Necessity’ and ‘Essential Health Benefi ts’ Under the Affordable Care Act, 38 Amer. J. L. & Med. 445, 336 (2012) (noting “the intensely fraught nature of any attempt to defi ne the essence of ‘health,’ ‘healthcare,’ or ‘medical necessity’ ”). 8. As the PPACA was being debated, the president and many of the bill’s supporters promised that those who liked their preexisting health insurance plans would be able to keep them. After passage, it became clear that this would not be true for several million people. In December 2013 PolitiFact.com declared that “If you like your health care plan, you can keep it” was the “Lie of the Year.” See Angie Drobnic Holan, Lie of the Year: “If You Like Your Health Care Plan, You Can Keep It,” PolitiFact.com (Dec. 12, 2013, 4;44 pm), http://www.politifact.com/truth-o-meter/ article/2013/dec/12/lie-year-if-you-like-your-health-care-plan-keep-it/. 9. Harold Pollack, 47 (Now 51) Health Policy Experts (Including Me) Say “Sign the Senate Bill.” The New Republic, The Treatment, (Jan. 22, 2010), http://www.tnr.com /blog/the-treatment/47-health-policy-experts-including-me-say-sign-the-senate-bill. 10. See Timothy Jost, Tax Credits in Federally Facilitated Exchanges Are Consistent with the Affordable Care Act’s Language and History, Health Affairs Blog, July 18, 2012 (noting “the Senate Bill was not supposed to be the fi nal law.”), http://healthaffairs.org/blog/2012/07/18/tax-credits-in-federally-facilitatedexchanges-are-consistent-with-the-affordable-care-acts-language-and-history/. 11. Some describe exchanges as the “centerpiece” of the PPACA reforms. See, e.g., Sandy Praeger, A View from the Insurance Commissioner on Health Care Reform, 20 Kansas J. L. & Pub. Pol’y 186, 189 (2011) (“The centerpiece of the reform is the new health insurance exchanges that will operate in every state.”); see also Robert Pear, Health Care Overhaul Depends on States’ Insurance Exchanges, N.Y. Times, Oct. 23, 2010, http://www.nytimes.com/2010/10/24/health /policy/24exchange.html. 12. See 42 U.S.C. §18031(b)(I). 13. See Printz v. United States, 521 U.S. 898, 925 (1997) (“the Federal Government may not compel the states to implement, by legislation or executive action, federal regulatory programs.”); New York v. United States, 505 U.S. 144, 162 (1992) (“the Constitution has never been understood to confer upon Congress the ability to require States to govern according to Congress’s instructions”). 14. See NFIB, 132 S. Ct. at 2602 (“Congress may use its spending power to create incentives for States to act in accordance with federal polices. But when ‘pressure turns into compulsion,’ the legislation runs contrary to our system of federalism.” (citation omitted)). 15. As the Supreme Court directed in New York v. United States, statutory commands that states enact federal programs should, if possible, be construed as incentives for state cooperation. 505 U.S. 144, 170 (1992); see also NFIB, 132 S. Ct. at 2597.

The Future of Healthcare Reform Remains in Federal Court

75

16. See 42 U.S.C. §18041(c)(I). 17. New York v. United States, 505 U.S. at 167 (“where Congress has the authority to regulate private activity under the Commerce Clause, we have recognized Congress’ power to offer States the choice of regulating that activity according to federal standards or having state law pre-empted by federal regulation. . . . This arrangement . . . has been termed “a program of cooperative federalism.”). 18. See 26 U.S.C. § 36B. 19. For a more extensive discussion of this issue and the implications for PPACA implementation, see Jonathan H. Adler & Michael Cannon, Taxation without Representation: The Illegal IRS Rule to Expand Tax Credits under the PPACA, 23 Health Matrix: Journal of Law-Medicine 119 (2013). 20. See 26 U.S.C. § 36B. 21. See id. § 36B(c)(2)(A)(i). 22. See id. § 36B(b)(2)(A), (b)(3)(B)(i). 23. See 42 U.S.C. § 18031(c)(6)(D). 24. See id. § 18041(c)(I). 25. See id. § 36B(c)(2)(A)(i). 26. See id. § 18024(d) 27. See 26 U.S.C. § 4980H. 28. The PPACA provides, in the alternative, that if an employer provides “minimum value” insurance coverage that is not “affordable,” the employer is fi ned $3,000 per employee that receives tax credits or cost-sharing subsidies or $2,000 per employee after the fi rst thirty employees, whichever is less. 29. Robert Pear, U.S. Offi cials Brace for Huge Task of Operating Health Exchanges, N.Y. Times, Aug. 5, 2012, available at: http://www.nytimes.com /2012/08/05/us/us-offi cials-brace-for-huge-task-of-running-health-exchanges. html. (“When Congress passed legislation to expand coverage two years ago, Mr. Obama and lawmakers assumed that every state would set up its own exchange”). Shortly before passage, Health and Human Services Secretary Kathleen Sebelius reported states were “very eager” to create Exchanges. Departments of Labor, Health and Human Services, Education, and Related Agencies Appropriations for 2011: Hearing Before the House Comm. on Appropriations, 111th Cong. 170–171 (Mar. 10, 2010) (Statement of Kathleen Sebelius, Secretary, Department of Health and Human Services), available at: http://www.gpo.gov/fdsys /pkg/CHRG-111hhrg58233/pdf/CHRG-111hhrg58233.pdf. 30. Barack Obama, U.S. President, Remarks by the President on Health Insurance Reform in Portland, Maine (Apr. 1, 2010), http://www.whitehouse.gov /the-press-office/remarks-president-health-insurance-reform-portland-maine. 31. See, e.g., Sam Stein, Latest Public Option Concession Comes Into Focus, Huffington Post (Mar. 18, 2010), http://www.huffi ngtonpost.com/2009/12/30 /latest-public-option-conc_n_407369.html; see also Carrie Brown, Nelson: National

76

Jonathan H. Adler

Exchange a Dealbreaker, Politico (Jan. 25, 2010), available at: http://www.po litico.com/livepulse/0110/Nelson_National_exchange_a_dealbreaker.html. 32. See Timothy S. Jost, Health Insurance Exchanges: Legal Issues, O’Neill Institute, Georgetown University Legal Center, no. 23, April 27, 2009, at 7, http://schol arship.law.georgetown.edu/cgi/viewcontent.cgi?article=1022&context=ois_papers. 33. See 42 U.S.C. § 18041(c). 34. See Abby Goodnuogh & Michael Cooper, Health Law Has States Feeling Tense Over Deadline, N.Y Times, Nov. 14, 2012, http://www.nytimes.com/2012/11/15 /health/health-law-has-states-feeling-tense-over-deadline.html. 35. See Robert Pear, Most Governors Refuse to Set Up Health Exchanges, N.Y. Times, Dec. 14, 2012, http://www.nytimes.com/2012/12/15/us/most-states -miss-deadline-to-set-up-health-exchanges.html. 36. Interestingly enough, nothing in the PPACA would appear to authorize the creation of a “partnership” exchange. 37. Pear, supra note 36; see also Robert Pear, States Will Be Given Extra Time to Set Up Health Insurance Exchanges, N.Y. Times, Jan. 14, 2013, http://www .nytimes.com/2013/01/15/us/states-will-be-given-extra-time-to-set-up-health -insurance-exchanges.html. 38. Some states also enacted legislation, passed ballot referenda, or adopted constitutional amendments that would appear to preclude those states from establishing their own exchanges. 39. Department of the Treasury, Internal Revenue Service, Health Insurance Premium Tax Credit, 77 Fed. Reg. 30377 (May 23, 2012), available at: http://www .gpo.gov/fdsys/pkg/FR-2012-05-23/pdf/2012-12421.pdf. 40. Id. at 30378. 41. See, e.g., Jost, supra note 11; but see Michael Cannon & Jonathan H. Adler, The Illegal IRS Rule to Expand Tax Credits Under the PPACA: A Response to Timothy Jost, Health Affairs Blog (Aug. 1, 2012), http://healthaffairs.org /blog/2012/08/01/the-illegal-irs-rule-to-expand-tax-credits-under-the-ppaca-a -response-to-timothy-jost/. 42. See Letter from Mark J. Mazur, Assistant Secretary for Tax Policy, U.S. Treasury Department, to the Honorable Darrell Issa, Chairman, Committee on Oversight and Government Reform, U.S. House of Representatives, (Oct. 12, 2012), http://www.cato.org/sites/cato.org/fi les/documents/20121012_mazur_irs_ letter_to_issa.pdf. 43. See id. (citing 42 U.S.C. §18041(c)(1). 44. See James F. Blumstein, testimony, House Ways & Means Committee Health Subcommittee, Implementation of Health Insurance Exchanges and Related Provisions, Wednesday, September 12, 2012, http://waysandmeans.house .gov/uploadedfi les/house_ways_and_means_testimony92112.pdf. 45. As a general rule, courts are not to treat any statutory provisions as mere surplusage. See, e.g., Duncan v. Walker, 533 U.S. 167, 174 (2001) (“We are  .  .  .

The Future of Healthcare Reform Remains in Federal Court

77

‘reluctan[t] to treat statutory terms as surplusage’ in any setting” (citation omitted)); Jones v. U.S., 529 U.S. 848, 857 (2000) (“Judges should hesitate . . . to treat statutory terms in any setting as surplusage” (citation and internal quotation omitted)); see also Russello v. United States, 464 U.S. 16, 23 (1983) (“Where Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.”); NFIB, 132 S. Ct. at 2583 (same). 46. See 26 U.S.C. § 36B(f)(3). 47. See Letter from Mark J. Mazur, supra note 43. 48. See 42 U.S.C. § 18043. 49. See Timothy S. Jost, Yes, the Federal Exchange Can Offer Premium Tax Credits, Health Reform Watch (Sept. 11, 2011), http://www.healthreformwatch .com/2011/09/11/yes-the-federal-exchange-can-offer-premium-tax-credits/; see also John D. Kraemer & Lawrence O. Gostin, The Power to Block the Affordable Care Act: What Are the Limits?, 308 J. Amer. Med. Assn. 1975 (2012) (characterizing relevant statutory language as “an apparent oversight”). 50. Jost, supra note 33, at 7. 51. See, e.g., 42 U.S.C. §§ 1397aa–1397mm (conditioning federal grants to states on their implementation of a health insurance program for children with low-to-moderate incomes); 26 U.S.C. § 35(e)(2) (eligibility for “health coverage tax credits” conditioned on states enacting specified laws); 26 U.S.C. § 223(c) (2) (individual’s eligibility to make tax-free contributions to health savings accounts conditioned on state providing the regulatory environment required by federal law). 52. The Supreme Court has also upheld the constitutionality of imposing differential tax burdens as a consequence of state cooperation with or resistance to federal policy priorities. As the Supreme Court noted in NFIB, the Court had previously upheld federal legislation “predicating tax abatement on a State’s adoption of a particular type of unemployment policy” in Steward Machine Co. v. Davis, 301 U.S. 548 (1937). See NFIB, 132 S. Ct. at 2603. See also New York v. United States, 505 U.S. 144 (1992) (upholding a provision of the Low-Level Radioactive Waste Policy Act Amendments that authorized surcharges on importation of low-level radioactive waste from noncompliant states). 53. As noted above, even in states where premium assistance tax credits are available, they are not provided for those earning less than 100 percent of the poverty line. See 26 U.S.C. § 36B(c)(1). 54. The Senate Democratic Policy Committee, for example, responded to claims that healthcare reform would result in a federal “takeover” with a “fact check” claiming that “All the health insurance exchanges  .  .  . are run by states.” See Senate Democratic Policy Committee, Responsible Reform for the Middle Class: Fact Check; Responding to Opponents of Health Insurance

78

Jonathan H. Adler

Reform, Sept. 21, 2009, available at: http://dpc.senate.gov/reform/reform-fact check-092109.pdf. 55. See Letter from Douglas H. Shulman, Comm’r, Internal Revenue Serv., to Rep. David Phil Roe, U.S. House of Representatives (Nov. 29, 2011), available at: http://roe.house.gov/uploadedfi les/irs_response_to_letter_on_ppaca_exchange .pdf. 56. See Pruitt v. Sebelius, 2013 WL 4052610 (E.D. Okla. 2013). 57. See Halbig v. Burwell, 758 F.3d 390 (D.C. Cir. 2014). 58. See King v. Burwell, 759 F.3d 358 (4th Cir. 2014). 59. See Indiana v. Internal Revenue Service, 2014 WL 3928455 (Aug. 12, 2014). 60. See, e.g., Timothy S. Jost, supra note 50. 61. See The Becket Fund, HHS Mandate Information Central, http://www .becketfund.org/hhsinformationcentral. 62. See 42 U.S.C. § 300gg-13(a) 63. See Health Resources and Services Administration, Women’s Preventative Services: Required Health Plan Coverage Guidelines (Aug. 1, 2011), http:// www.hrsa.gov/womensguidelines/. 64. 26 U.S.C. § 4980D(b). 65. 76 Fed. Reg. 46621, 46626 (Aug. 3, 2011). 66. Nancy Frazier O’Brien, HHS delays, but does not change, rule on contraceptive coverage, Catholic News Service (Jan. 20, 2012), http://www.catholic news.com/data/stories/cns/1200263.htm. 67. See 77 Fed. Reg. 8725 (Feb. 15, 2012); id. 16503 (Mar. 21, 2012); Department of Health and Human Services, Guidance on the Temporary Enforcement Safe Harbor (Feb. 10, 2012). 68. See, e.g., Katie Thomas, Self- Insured Complicate Health Deal, N.Y. Times, Feb. 16, 2012. 69. See Employment Div. v. Smith, 494 U.S. 872 (1990). 70. See 42 U.S.C. § 2000bb § 2000bb-4. 71. See, e.g., Caroline Mala Corbin, The Contraception Mandate, 107 Nw. U. L. Rev. Colloquy 151 (2012). 72. See, e.g., Edward Whelan, The HHS Contraception Mandate vs. the Religious Freedom Restoration Act, 87 Notre Dame L. Rev. 2179 (2012). 73. See, e.g., Little Sisters of the Poor v. Sebelius, 134 S. Ct. 1033 (2014). 74. See Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014). Note that the cases brought by Hobby Lobby Stores and Conestoga Wood Specialties were consolidated before the Supreme Court. 75. See Wheaton College v. Burwell, 134 S. Ct. 2806 (2014) (granting injunction against enforcement of contraception coverage mandate, even with accommodation, pending appeal). 76. See, e.g., U.S. General Accounting Office, Health Care: Unsustainable Trends, GAO- 04- 793SP, May 2004.

The Future of Healthcare Reform Remains in Federal Court

79

77. 42 U.S.C. § 1395kkk(b). 78. See Henry J. Aaron, The Independent Payment Advisory Board— Congress’s Good Deed, 364 N. Engl. J. Med. 2377 (2011); Timothy Stoltzfus Jost, The Real Constitutional Problem with the Affordable Care Act, 36 J. Health Pol., Pol’y & L. 501, 502 (2011); see also Peter R. Orszag, Too Much of a Good Thing: Why We Need Less Democracy, New Republic, Sept. 14, 2011. 79. Some recent reports suggest that healthcare cost increases could slow enough that the IPAB will not be required to adopt cost- control measures, at least in the immediate future. See, e.g., Sarah Kliff, The $2.7 Trillion Question: Are Health- Care Costs Really Slowing?, Wash. Post “WonkBlog,” Jan. 7, 2013, http:// www.washingtonpost.com/blogs/wonkblog/wp/2013/01/07/the-2-7-trillion -question-are-health-care-costs-really-dropping/. 80. Should the IPAB fail to act, for whatever reason, this responsibility falls to the HHS Secretary. 81. See Ann Marie Marciarille & J. Bradford DeLong, Bending the Health Cost Curve: The Promise and Peril of the Independent Payment Advisory Board, 22 Health Matrix 74, 78- 79 (2012). 82. See 42 U.S.C. § 1395kkk(c)(2)(A). 83. As Professor Jost notes, this means that the IPAB could adopt measures that arguably contravene the statutory limitations. See Jost, supra note 76, at 505. 84. See, e.g., Aaron, supra note 76, at 2379; Marciarile & Delong, supra note 79, at 79; Timothy Stoltzfus Jost, The Independent Payment Advisory Board, 363 N.Engl. J. Med. 103, 105 (2010). 85. See 42 U.S.C. § 1395kkk(e)(5). 86. As one commentator notes: “What is happening here is truly remarkable. Congress is delegating to HHS authority to waive the provisions of existing law, freeing it from judicial oversight and, in the case of the IPAB, even limiting Congress’s own authority to override the decisions of an executive agency.” Jost, supra note 76, at 503. 87. See 42 U.S.C. § 1395kkk(d)(3). 88. In litigation challenging the constitutionality of the IPAB, the Obama Administration attested that these provisions merely provide for “one way for Congress to repeal the Board” and that “Nothing prevents Congress from repealing the Board via ordinary legislation.” See Coons v. Geithner, Motion to Dismiss, CV-10-1714-PHX- GMS (D. Ariz. May 31, 2011). 89. See Coons v. Lew, Coons v. Lew, 2014 WL 3866475 13-15324 (9th Cir. 2014). 90. Jost, supra note 76, at 506. 91. See NFIB, 132 S. Ct. at 2598 (opinion of Roberts, C.J.). 92. See, e.g., David B. Rivkin Jr. & Lee A. Casey, The Opening for a Fresh ObamaCare Challenge, Wall St. J., Dec. 5, 2012. 93. See, e.g., Association of American Physicians & Surgeons, Inc. v. Sebelius, 901 F.Supp.2d 19 (D.D.C. 2012), aff’d 746 F.3d 468 (D.C. Cir. 2014).

80

Jonathan H. Adler

94. See Sissel v. U.S. Dep’t of Health & Human Servs., 760 F.3d 1 (D.C. Cir. 2014); Hotze v. Sebelius, 2014 WL 109407 (S.D. Tex. Jan. 10, 2014). 95. See, e.g., Thomas C. Buchmueller, Consumer Demand for Health Insurance, NBER Reporter (2006), http://www.nber.org/reporter/summer06/buchm ueller.html. 96. See Thomas A. Lambert, How the Supreme Court Doomed the ACA to Failure, Regulation (Winter 2012-13), pp. 32– 38. 97. Congressional Budget Office, Payments of Penalties for Being Uninsured Under the Affordable Care Act (Sept. 19, 2012), http://www.cbo.gov/publication/43628. 98. See NFIB, 132 S. Ct. at 2590 (opinion of Roberts, C.J.). 99. Id. (quoting Bailey v. Drexel Furniture, 259 U.S. 20, 37 (1922)); see also Randy E. Barnett, No Small Feat: Who Won the Health Care Case (and Why Did So Many Law Professors Miss the Boat)?, 65 Fla. L. Rev. 1331, 1340 (2013), http://www.floridalawreview.com/wp-content/uploads/9-Barnett.pdf. 100. See NFIB, 132 S. Ct. at 2600 (opinion of Roberts, C.J.). 101. See Thomas A. Lambert, supra note 93. The Roberts “tax” ruling undermines the new healthcare law. 102. See David Nather, Obamacare Mandate May Be ‘Mandate Plus,” Politico Pro, Jan. 13, 2013, http://www.politico.com/story/2013/01/obamacare-man date-may-be-mandate-plus-86115.html.

chapter four

Essential Health Benefits and the Affordable Care Act Law and Process Nicholas Bagley and Helen Levy

T

he Affordable Care Act (ACA) creates dozens of new programs that require some kind of implementation at the agency level. By and large, the regulations governing these new programs have been promulgated through relatively formal notice-and- comment procedures and subjected to review coordinated by the Office of Information and Regulatory Affairs (OIRA). But the federal agencies implementing the ACA have at times eschewed notice-and- comment rulemaking even where it might have seemed appropriate. They have instead announced a number of critically important policies through guidance documents—a broad category that encompasses bulletins, memoranda, and letters to state officials. These guidance documents are typically published not in Federal Register notices, but on agency websites. This implementation strategy raises questions—perennials in administrative law—about the virtues and vices of substituting guidance for notice-and-comment rulemaking. Does the use of guidance reflect the zealous pursuit of good policy by government officials reluctant to get bogged down in ritualistic bureaucratic exercises? Or does it represent an effort to avoid the rough-and-tumble of public deliberation over the merits of particular rules? We consider this question in the context of a case study. Beginning in 2014, the ACA will require private insurance plans sold in the individual

82

Nicholas Bagley and Helen Levy

and small-group markets to cover a roster of “essential health benefits.” Precisely which benefits should count as essential, however, was left to the discretion of the Department of Health and Human Services (HHS). The matter was delicate. An expansive bundle of mandatory services would assure comprehensive coverage, but would also raise the cost of insurance and could impede efforts to achieve near-universal coverage. Whatever HHS eventually decided, its choice would “influence the nature of coverage available to millions of people in the United States” (IOM 2011b: 17). In December 2011, HHS released its first official communication on essential health benefits: a thirteen-page bulletin posted on the agency’s website. The bulletin announced HHS’s intention of allowing each state to define essential benefits for itself by choosing a “benchmark” plan modeled on existing plans in the state. The benefits included in that benchmark plan (subject to some adjustments) would be considered essential within the state. On both substance and procedure, the move was surprising. The benchmark approach departed from the uniform, federal standard that the ACA appears to anticipate and that many informed observers expected HHS to adopt. And announcing the policy through an Internet bulletin came under immediate criticism because it allowed the agency to sidestep conventional administrative procedures—including notice and comment, immediate review in the courts, and OIRA oversight— notwithstanding the ACA’s command that HHS “provide notice and an opportunity for public comment” on the defi nition of essential health benefits (section 1302). By the time HHS issued a notice of proposed rulemaking (NPRM) on essential health benefits in November 2012, the deadline for states to submit their proposed benchmark plans to the agency was already two months in the past. What are we to make of this? The story of essential health benefits offers insight into the merits of using guidance documents; it is also interesting in its own right, both because of the importance of the policy question and because of HHS’s unexpected decision. In this essay, we explore two questions. First, is the benchmark approach a lawful exercise of HHS’s authority under the ACA? Although HHS has brushed up against the limits of its discretionary authority, we conclude that the approach likely will (and, in our view, should) be upheld in the event of a challenge. Second, did HHS’s announcement of the benchmark approach through an Internet bulletin allow the agency to sidestep the ad-

Essential Health Benefits and the Affordable Care Act

83

ministrative procedures that are meant to shape the exercise of its discretion? The answer, we believe, is no. In fact, the agency’s unconventional process was more open to public scrutiny and external oversight than conventional rulemaking would have been.

Defi ning Essential Health Benefits Starting in 2014, the ACA requires new health insurance plans in the individual and small-group markets to cover a minimum set of services that the ACA terms “essential health benefits.”1 This requirement applies to plans sold on state health insurance exchanges and to individual and small-group plans sold outside the exchanges. The statute (section 1302) enumerates ten different categories of services that essential health benefits must, at a minimum, include 1. Ambulatory patient services 2. Emergency services 3. Hospitalization 4. Maternity and newborn care 5. Mental health and substance use disorder services, including behavioral health treatment 6. Prescription drugs 7. Rehabilitative and habilitative services and devices 8. Laboratory services 9. Preventive and wellness services and chronic disease management 10. Pediatric services, including oral and vision care

Many of these inclusions are significant—for example, a more barebones approach might have excluded prescription drugs and pediatric dental care—but the list, by design, leaves much detail to be specified by subsequent regulation. For example, does the “habilitative services” category include behavioral treatment for autism, an expensive therapy with mixed evidence of effectiveness (Reichow 2012)? What do “preventive and wellness services” encompass beyond the ones that another provision of the ACA requires all plans to cover without cost sharing (sections 1001, 2713)? Sensitive to the need for greater detail, the ACA instructs the secretary of HHS to flesh out the definition of essential health benefits. Specifically, the statute directs her to “ensure that the scope of the essential health

84

Nicholas Bagley and Helen Levy

benefits . . . is equal to the scope of benefits provided under a typical employer plan, as determined by the Secretary.” Congress also instructed the secretary of labor to survey insurance plans “to determine the benefits typically covered by employers” and report back to the secretary of HHS. “In defi ning the essential health benefits,” the statute further provides, “the Secretary shall provide notice and an opportunity for public comment” (section 1302). In the normal course of regulatory events, HHS might have been expected to launch an orderly rulemaking process not long after the ACA’s enactment. It is hard to say exactly what a reasonable time frame for this might have been, but the ACA required states to demonstrate to HHS by January 2013 that they would have health insurance exchanges up and running within a year. To make such a demonstration, states would have to know well in advance about the scope of benefits that plans on the exchanges would have to cover. Working backward, a notice of proposed rulemaking would probably have had to issue by the end of 2011, followed by a fi nal rule in mid-2012, to have any hope of giving states the certainty they needed to create their exchanges. That is not, however, what happened. Shortly after the ACA’s enactment, HHS turned to the Institute of Medicine (IOM) for “advice on a process and considerations the Department needs to take into account in its initial establishment of [essential health benefits] and in updating them over time” (IOM 2011a). In other words, HHS asked the IOM not to defi ne essential benefits, but to offer suggestions on how HHS might do so. The IOM report was expected to be complete in the fall of 2011. Enlisting help from the IOM bought HHS time during which it might reasonably do nothing. Assuming the agency was prepared to issue a notice of proposed rulemaking soon after the release of the IOM report, the formal rulemaking process could proceed on a time frame that would allow for meaningful interaction with states and other interested parties—including insurers, health care providers, and consumer-advocacy groups—before bumping into deadlines for health insurance exchanges. The IOM tackled its assignment with dispatch, quickly convening an expert panel to write a report recommending methods for determining and updating essential benefits. It also invited members of the public to submit comments online, and held two public conferences, one in Washington, DC, in January 2011 and another in Costa Mesa, California, in March 2011. Featuring presentations from an impressive range of experts

Essential Health Benefits and the Affordable Care Act

85

and stakeholders, these conferences were later summarized in a volume that the IOM (2011b) released to the public. Meanwhile, in April 2011, the Department of Labor delivered its report on employer-sponsored coverage to HHS (Department of Labor 2011). Unfortunately, the Department of Labor surveys on which the report was based relied on “summary plan descriptions” that employers provide to their workers—and those descriptions lack detailed information about the scope of coverage for specific services. The report therefore gave HHS little to guide its decision; it certainly provided no help on whether, say, a “typical” employer plan covered behavioral treatment for autism. The IOM went considerably further. On October 6, 2011, the expert panel released a 256-page report recommending a method for determining essential benefits (IOM 2011a). Somewhat controversially, the report proposed a “premium target” approach in which a single national package of essential benefits would be tied to the cost of a typical benefits package in the small-group market. That national package would then be updated over time to reflect innovation and public deliberation. Following the release of the IOM report, the agency announced that it would hold a series of “listening sessions” over the subsequent months. These sessions—two-hour meetings at which members of the public could share their opinions with HHS officials—were conducted in each of ten HHS- defi ned regions. Sessions took place in Chicago, Boston, Philadelphia, Dallas, New York, Kansas City, Atlanta, Seattle, Denver, and, fi nally, on the Monday before Thanksgiving, San Francisco. Three and a half weeks later, reports began to circulate that HHS intended to release a “prerule” on essential health benefits, although the term prerule created confusion. “Not even the most seasoned Washingtonians seem to know what it means,” reported Politico (Feder and Millman 2011). Finally, on December 16, 2011, the prerule was posted on HHS’s website with the title “Essential Health Benefits Bulletin” (CCIIO 2011). Although the medium may have created some confusion, the bulletin’s message was both concise and surprising. Rather than specify a uniform national benefits package, the bulletin proposed to allow states to choose a “benchmark plan” to define essential health benefits. This approach was modeled on how states choose benefits under the State Children’s Health Insurance Program, also known as SCHIP (KFF 2007). A modified version of the same policy was introduced into Medicaid by the

86

Nicholas Bagley and Helen Levy

Deficit Reduction Act of 2005. Under the Medicaid version of the benchmark approach, states were allowed to offer a modified set of benefits, linked to a state-selected benchmark, to some groups of adult Medicaid enrollees (KFF 2010). Adapting the benchmark approach for essential health benefits, the bulletin proposed allowing each state to choose a “benchmark plan” from a menu of options, including the three largest insurance plans in the state’s small-group market and the three largest plans available to state employees. The default benchmark, for states that failed to select one, would be the largest small-group plan in the state. Subject to adjustments to assure their conformity with the ACA’s list of coverage requirements, these benchmark plans would defi ne essential benefits within the states. At just thirteen pages long, the bulletin provided few details. Additional information, the bulletin suggested, would be forthcoming in the fulldress rulemaking proceeding.

Responses to the Bulletin The benchmark approach was front-page news, described by the New York Times as a “major surprise” (Pear 2011). As we explain in greater detail below, the ACA was drafted on the assumption that HHS would choose a single, national defi nition of essential health benefits. The Congressional Budget Office, for one important example, had scored it on that assumption (CBO 2012: 8). Most expert observers had not seriously considered a state-specific benchmark prior to the bulletin. And the IOM report never mentioned the benchmark approach that the bulletin ultimately proposed, although it did offer a limited endorsement of the idea that states might deviate from the national defi nition of essential benefits, subject to approval by HHS (IOM 2011a: 129). 2 Why did HHS take such an unanticipated approach to essential health benefits? Politics certainly played a role. The benchmark approach gave states flexibility at a time when many were complaining about the lack of it. In addition—and although we lack the space here to thoroughly examine the question—smart politics may have made for smart policy. Because most health insurance plans “do not differ significantly in the range of services they cover” and “generally cover health care services in virtually all of the 10 statutory categories,” no state could select a threadbare benchmark plan that would thwart the ACA’s effort to guarantee the availability of

Essential Health Benefits and the Affordable Care Act

87

comprehensive coverage (CCIIO 2011: 4). And tying local benefits to local market conditions would probably result in less distortion (i.e., greater efficiency) than if benefits were required to be uniform. The benchmark approach does create winners and losers. Because of the way premium subsidies and tax-sharing credits are calculated, recipients of subsidized coverage in states with generous benefits will receive modestly more federal support than those in less generous states. But many policies give rise to that kind of differential treatment; variation across states is simply a feature of our federal system. Even John Ball, the chair of the IOM panel that recommended tying essential health benefits to a premium target, offered only gentle criticism of the benchmark approach. “Given where the department is coming from, giving flexibility to the states is a good thing,” he told Politico. “But I do think they missed an opportunity to take a crack at getting costs under control” (Millman 2011b). Leading Republicans reacted much more harshly, not to the substance of the policy but to the manner in which it was announced. In a letter to HHS, five influential Republican senators and congressmen offered the following objections: By issuing a “bulletin” rather than a proposed rule, the administration has sidestepped the requirement to publish a cost-benefit analysis estimating the impact these mandates will have on health insurance premiums and the increased costs to the federal government. . . . The administration is not required to respond to comments received regarding this “bulletin.” Publishing a “bulletin” rather than a proposed rule is the antithesis of an “open and transparent” process. . . . The bulletin also does not have the force of law and cannot, therefore, be considered an indication of what the proposed or fi nal rule will decree. Thus, states still have many unanswered questions and no more certainty than they had before the “bulletin” was released. . . . It is unreasonable to expect states to be ready to implement such draconian changes by 2014, if the Administration is not even ready to issue a proposed rule on such an integral part of the functioning of the law. (Enzi et al. 2012)

These concerns notwithstanding, the bulletin prompted many states to launch their own administrative processes for selecting benchmark plans. And that was the point: “By releasing the bulletin now,” the secretary of HHS explained, “we’re giving families, employers and states

88

Nicholas Bagley and Helen Levy

plenty of time to take this information into account as they plan for the big improvements the health care law will make to the insurance market in 2014” (Sebelius 2011). With just over a year to go before the January 2013 deadline for demonstrating readiness to run an exchange, and other reform-related tasks to complete at the same time, any state interested in running its own exchange could not really afford to wait. To be sure, some states declined to participate in this process. State officials cited the lack of guidance from HHS on essential health benefits and other exchange-related issues as one of their reasons. In September 2012, Michael Consedine, the insurance commissioner for the state of Pennsylvania, stated in testimony before the Health Subcommittee of the House Ways and Means Committee that “the lack of detailed information from HHS has put Pennsylvania, and many other states, in a very difficult position. We are traveling down a road, directionless, while knowing the road will end very soon—January 2014 is right around the bend.” Robert Bentley, the Republican governor of Alabama, was even more pointed in a letter sent a month later to the secretary of HHS: Your office released essential health benefits guidance on December 16, 2011, with the promise of more to come. It has yet to arrive. It has become clear to me that the states have been left to decide the fate of their insurance marketplaces with no additional guidance or regulations on essential health benefits. This places governors and other leaders in the untenable position of making a critical decision based on little more than vague guidance and guesswork. . . . I decline to make a decision on the essential health benefits benchmark plan. There is simply not enough valid information available now to make an informed choice for such an important decision. (Bentley 2012)

Ultimately, twenty-seven states held a public comment period on the subject of the benchmark plan as part of the selection process, and twentyfour states submitted their proposed benchmark plans to HHS by the October 1, 2012, deadline (Avalere Health 2012; Schwartz 2012). Alabama and Pennsylvania were not among them. On November 26, 2012—twenty days after the reelection of President Obama and almost a year after the release of the bulletin—HHS published in the Federal Register a notice of proposed rulemaking on essential health benefits. The NPRM formally proposed the benchmark approach that the bulletin had previewed. In its discussion of regulatory alternatives, the NPRM said,

Essential Health Benefits and the Affordable Care Act

89

At the request of some commenters, HHS considered one national defi nition of [essential health benefits] that would have applicable issuers offer a uniform list of benefits. However, this approach would not allow for state flexibility and issuer innovation in benefit design, would require a burdensome overhaul for issuers, and would disrupt the market (NPRM 2012: 70,665).

These two sentences represent the entirety of the NPRM’s discussion of the policy wisdom of the benchmark approach. The fi nal rule, issued on February 25, 2013, deviated little, if at all, from the NPRM. On the possibility of a single national defi nition of essential health benefits, the fi nal rule repeated the two sentences from the NPRM that are quoted above, without further elaboration (Final Rule 2013: 12,861).

Legality of the Benchmark Approach Because the ACA does not explicitly contemplate a state-based benchmark approach to essential health benefits, the question arises whether the approach is consistent with statute. In other words, has the secretary exceeded the bounds of her discretionary authority? Notwithstanding its considerable importance, the question has received scant public attention. HHS has not yet offered a legal defense of the benchmark approach, but its argument will probably run something like this: where agencies interpret open- ended phrases through notice-and- comment rulemaking, courts typically give agencies a lot of leeway. This practice is known as the “Chevron deference,” after the landmark case establishing it (Chevron U.S.A. v. Natural Res. Def. Council, 467 U.S. 837 [1984]). Here Congress delegated to the secretary of HHS broad authority to flesh out the meaning of “essential health benefits.” Under Chevron, the secretary’s interpretation of the statutory phrase will be upheld so long as that interpretation offers a reasonable construction of the ACA. Nothing in the statute precludes the secretary either from linking those benefits to state health plans or from giving the states the flexibility to select benchmark plans. Given congressional silence on those points, the secretary’s exercise of her authority is fully consistent with the ACA. This argument is a powerful one. There are, however, two ways in which the benchmark approach is arguably difficult to square with the text of the ACA. The fi rst is obvious. In a statute that attends carefully to the division of regulatory labor between the federal government and the

90

Nicholas Bagley and Helen Levy

states, the ACA repeatedly confi rms that “the Secretary shall defi ne the essential health benefits” (section 1302). This is not casual language: in three separate places in the same statutory section, the act contemplates that the secretary would be the one “defi ning” and then “revising” what counts as essential health benefits. The ACA even instructs the secretary to “ensure that the scope of essential health benefits . . . is equal to the scope of benefits provided under a typical employer plan, as determined by the Secretary.” The phrase “as determined by the Secretary” would do no work unless it was the secretary—not the states—doing the determining. This objection, however, is not terribly persuasive. Although a federal agency cannot delegate its powers to the states, it “may turn to an outside entity for advice and policy recommendations, provided the agency makes the fi nal decisions itself” (U.S. Telecom Ass’n v. F.C.C., 359 F.3d 554 [D.C. Cir. 2004]). Here the secretary gave the states a constrained set of options (e.g., choose a benchmark plan from among the three largest small-group plans in the state) and retained the authority to select a benchmark for any state that either does not pick a benchmark or chooses an inappropriate one (NPRM 2012: 70,667). As such, the secretary remains fi rmly in control. Nothing in the ACA prevents her from deferring to states that select benchmark plans from among the few options she has provided. That choice to defer is itself an exercise of her delegated powers. The second potential objection to the benchmark approach is both less obvious and more substantial. Notwithstanding the secretary’s wide discretion to defi ne essential health benefits, there are limits to the deference that courts afford to agencies that interpret open- ended statutory language. As the D.C. Circuit has explained, the notion that an agency interpretation is permissible just because the statute in question “does not expressly negate the existence of a claimed administrative power (i.e. when the statute is not written in ‘thou shalt not’ terms), is both flatly unfaithful to the principles of administrative law  .  .  . and refuted by precedent” (Ry. Labor Executives’ Ass’n v. Nat’l Mediation Bd., 29 F.3d 655 [D.C. Cir. 1994]). “The question,” the Supreme Court has recently emphasized, “is always whether the agency has gone beyond what Congress has permitted it to do” (City of Arlington v. FCC, 133 S.Ct. 1863 [2013]). In other words, agencies may fi ll in statutory gaps—but only up to a point. Where an agency’s interpretation of an open- ended provision clashes with the statutory scheme as a whole, the courts will not presume

Essential Health Benefits and the Affordable Care Act

91

that Congress meant to authorize the agency to so interpret the statute (Brown v. Gardner, 513 U.S. 115 [1994]). As we have discussed, the ACA was enacted on the assumption that HHS would establish a nationally uniform slate of essential health benefits. Under the benchmark approach, however, there will now be dozens of state-specific sets of essential health benefits. Many provisions of the ACA are inscrutable, extraneous, or impossible to implement in the face of that kind of variation. Consider again, for example, the requirement that essential health benefits must be “equivalent to the scope of benefits provided under a typical employer plan” (section 1302). How can a variable roster of state-specific essential health benefits be equivalent to the scope of benefits provided under “a” (which is to say, one) employer plan? Nowhere is the problem more apparent than in provisions governing state coverage mandates. Some states require insurers to cover specific benefits—for example, applied behavior analysis for autism or in vitro fertilization services—that Congress anticipated might exceed what the secretary would deem essential. Congress, however, did not want to devote the tax credits and cost-sharing payments available on the exchanges to the coverage of these state-mandated benefits. The ACA therefore limits federal subsidies to defraying the costs of essential health benefits (section 1401). States must pick up the rest of the tab to assure that exchange plans that include extra state-mandated benefits remain affordable (section 1311). Under the benchmark approach, however, this cost-sharing arrangement becomes irrelevant. A state’s benchmark plan will inevitably cover the treatments or services that the state has mandated. As such, a state’s essential health benefits will by defi nition include all of the benefits for which the state mandates coverage. State coverage mandates can therefore never exceed essential health benefits, and states with extensive coverage mandates will never assume the additional costs that the ACA anticipates they will assume. Did Congress really intend its cost-sharing provisions to do no work at all? The benchmark approach also raises questions about certain specialized insurance plans that will be sold on the exchanges. For the most significant example, the ACA instructs the Office of Personnel Management (OPM) to enter into contracts with health insurers to sell at least two “multi- state plans” on each state exchange. Those insurers must “offe[r] a benefits package that is uniform in each State and consists of the essential [health] benefits” (section 10104). Where essential health

92

Nicholas Bagley and Helen Levy

benefits vary from state to state, however, a multistate plan cannot both be uniform and cover only the essential health benefits. In proposing regulations for multistate plans, the OPM recognized this problem. Its solution was to read the “uniform in each State” language to require that “the benefits for each [multistate plan] must be uniform within a State, but not necessarily uniform among States” (NPRM 2012: 72,589). The bare language of the uniformity provision is perhaps amenable to this interpretation. Other statutory clues, however, suggest that Congress meant uniformity among states. Congress specified, for example, that a state can require a multistate plan to cover statemandated benefits, but only if the state picks up the increased expense (section 10104). There would have been no need for Congress to bless that limited inroad on uniformity among the states if the ACA required uniformity only within each state. In short, the benchmark approach to essential health benefits fits poorly with a number of provisions of the ACA. That poor fit raises the prospect of a legal challenge: perhaps it suggests that HHS exceeded its delegated powers in adopting the benchmark approach. In this, the claim is reminiscent of FDA v. Brown & Williamson Tobacco Corp. (529 U.S. 120 [2000]), where the Supreme Court rejected the FDA’s effort to assert jurisdiction over tobacco. If the FDA could regulate tobacco products, the Court reasoned, the agency’s statutory mandate to assure that such products were “safe” would require the FDA to ban cigarettes outright. The Court thought it unimaginable that Congress meant to arm the FDA with that sweeping power. The agency countered by arguing that it could continue to allow cigarettes to be sold because banning them would cause a shift to dangerous black-market cigarettes, harming overall public health. The Court rejected this public health interpretation as “incompatible” with several other provisions of the statute that asked the agency to weigh the therapeutic benefits and potential harms of individual products—not to address general questions of public health. The benchmark approach is subject to similar criticism for its incompatibility with provisions of the ACA that anticipate a single, uniform standard for essential health benefits. In the fi nal estimation, however, demonstrating the unlawfulness of HHS’s approach requires more than showing that its interpretation aligns awkwardly with scattered provisions of the ACA. The question remains whether the fit is so poor that it justifies the inference that

Essential Health Benefits and the Affordable Care Act

93

Congress could not have meant to allow HHS to interpret the ACA in the manner that it did (California Indep. Sys. Operator Corp. v. F.E.R.C., 372 F.3d 395 [D.C. Cir. 2004]). We think not, although the question is close. Unlike Brown & Williamson, this is not a case where HHS has exploited statutory ambiguity in an effort to intrude into regulatory domains that Congress never intended it to enter (American Bar Ass’n v. F.T.C., 430 F.3d 457 [D.C. Cir. 2005]). The agency has just chosen to involve the states in defi ning a statutory term that Congress gave it wide latitude to defi ne. Congress may not have contemplated that HHS would adopt a benchmark approach, but so what? Agencies routinely discharge their statutory obligations in ways that Congress never anticipated, particularly in complex and fast- changing regulatory environments. While HHS’s interpretation does fit uneasily with several provisions of the ACA, the agency could always shift course, and provisions that do no work today may be crucial tomorrow. More importantly, the interpretation of a complex statute is a messy business. Some statutory provisions will inevitably prove to be less significant than they would have been under alternative readings. An agency’s choice is not usually deficient for that reason alone (Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U.S. 687 [1995]). Something more than the statutory tension that the secretary’s interpretation creates—something closer to the incompatibility found in Brown & Williamson—would be necessary to conclude that Congress meant to preclude HHS from establishing essential health benefits with reference to state benchmarks.

Process-Based Concerns Setting aside the legality of the benchmark approach, the fact remains that HHS used a guidance document—a thirteen-page bulletin posted on its website—to announce its new policy. As is typically the case with such a policy statement, the bulletin nowhere committed the agency to the approach that it outlined. As a formal legal matter, it had no more significance than an advance notice of proposed rulemaking or a press release. Why, then, was this “atypical approach” so “widely criticized” (Cassidy 2011)? Given the looming deadlines for the exchanges, states and insurers understood that the benchmark approach outlined in the bulletin was not

94

Nicholas Bagley and Helen Levy

just one possibility among many. It was almost certainly the rule. HHS said as much when, in a call with reporters on the day the bulletin was issued, agency officials categorically rejected any suggestion that they might change their mind and bluntly said, “This is our intended regulatory approach” (Glied 2011). More significantly, the agency began the traditional rulemaking process required under the Administrative Procedure Act (APA) after the states’ deadline for selecting benchmark plans had passed and after insurers started developing new insurance products for the exchanges. The federal courts have a practice of asking whether policy statements, although they nominally lack the force of law, are nonetheless binding as a practical matter (Community Nutrition Inst. v. Young, 818 F.2d 943 [D.C. Cir. 1987]). Where policy statements have such binding effect, they can be deemed “legislative rules” and, if challenged, invalidated for failure to go through notice-and- comment rulemaking. Although the line separating policy statements from legislative rules is not crisp (Gersen 2007: 1712), two factors are especially important in suggesting that the line has been crossed: fi rst, when the policy statement has “present effect” by imposing “obligations” on the regulated community; and second, when the policy statement does not “genuinely” afford the agency the opportunity to change its mind (American Bus Assoc. v. United States, 627 F.2d 525 [D.C. Cir. 1980]). Judged by this standard, the bulletin arguably became a legislative rule once states, insurers, and employers realized that they had no choice but to conform to it and HHS had no time left to rethink it. The bulletin, however, did not go through the formal rulemaking process. From this perspective, HHS’s procedural approach—a bulletin, followed by a long wait and then a hurried notice-and- comment session—seems to confirm the fears of those who worry that agencies will use guidance to avoid administrative procedures and, at low cost and with relative ease, dictate to regulated entities how they must order their affairs (Anthony 1992). But what, exactly, did the use of guidance allow HHS to avoid? We consider three procedural requirements that would have applied in a conventional rulemaking setting, but to which the bulletin, as a guidance document, was not subject: notice and comment, immediate judicial review, and OIRA review. What is most notable about HHS’s unorthodox process, however, is not that the agency avoided these procedures. It is

Essential Health Benefits and the Affordable Care Act

95

that the agency voluntarily subjected itself to most of the obligations that adherence to the procedures would have entailed. Notice and Comment The APA’s notice-and- comment process is meant to allow for public feedback on proposed rules, to supply agencies with information that can aid them in revising those rules, and to publicly legitimate the rules they fi nally adopt (Breyer et al. 2002: 658). Notice and comment, however, can serve those purposes only if agencies are open to revisiting their proposed rules when they receive feedback. In the case of essential benefits, HHS committed itself to the benchmark approach long before issuing its notice of proposed rulemaking. That transformed the comment period—at least with respect to the high-level policy choice of whether to adopt the state-based benchmark approach—into an empty formality. HHS’s curt dismissal in the fi nal rule of comments suggesting a single, nationwide standard may have signaled its unwillingness at that late stage even to entertain the possibility. Still, it does not appear that HHS used the bulletin to avoid public feedback on the benchmark approach. Well before issuing the bulletin, HHS held a number of well-attended “listening sessions” where it sought views from states, insurers, providers, and consumer representatives. Additionally, starting in April 2010, HHS held weekly calls with state officials about implementation of the ACA, calls that informed its thinking about essential health benefits (NPRM 2012: 70,667). Around the same time, HHS also made it known to outside groups that it was toying with the idea of delegating authority to the states to establish their own essential health benefits. That provoked a consortium of about six dozen public interest groups, led by the National Health Law Program (NHeLP), to submit a lengthy letter to HHS—four months before the bulletin was issued—objecting preemptively to any sort of state-based approach (NHeLP 2011). These public discussions and the responses they engendered came on top of both the conferences that the IOM had organized and the notice-and- comment processes that twenty-seven states used to select their benchmark plans. The agency had no obligation to do any of this public outreach. Without informing anyone of its thinking, HHS could simply have issued a notice of proposed rulemaking announcing the benchmark approach.

96

Nicholas Bagley and Helen Levy

After receiving comments and issuing a fi nal rule, HHS would then have complied with all of the APA’s notice-and- comment requirements. Paradoxically, however, formal compliance might have undermined HHS’s effort to seek and receive meaningful public input on its proposed approach. Because courts insist that agencies provide a fulsome explanation of the basis for their proposed rules, HHS would have had to elaborate its benchmark approach in a lengthy notice of considerable specificity (Elliott 1992: 1494). HHS could then have discarded that approach only if it went through the laborious process of issuing a new notice of proposed rulemaking (Int’l Union, United Mine Workers of Am. v. Mine Safety & Health Admin., 407 F.3d 1250 [D.C. Cir. 2005]). But with statutory deadlines looming, the agency could have ill afforded the delay associated with restarting the notice-and-comment process. Instead, HHS used the bulletin to secure public input during the brief window where it could have reconsidered the benchmark approach. Although agencies do not have to solicit feedback on guidance documents, the bulletin’s very first sentence explicitly invited comments from the public. In response, the agency received more than eleven thousand comments (NPRM 2012: 70,646). HHS did not respond publicly to those comments, as the APA would have required in connection with a rulemaking. But HHS surely understood that the comments it received would preview the concerns voiced during the official notice-and- comment process— and that it would have to address those comments in issuing a fi nal rule. Nor does it seem that the absence of a requirement to publish comments received in response to the bulletin silenced any criticism. Just as the Internet allowed HHS to make the bulletin immediately available to the public, the Internet allowed commenters to publicize their concerns. A number of advocacy groups and state governments—like NHeLP and the officials in Alabama and Pennsylvania—did just that. Posting the bulletin was an ingenious way to invite public comment without irrevocably committing the agency to the benchmark approach. The bulletin served as a trial balloon—an effort to see if the approach would provoke the sort of outcry or incisive criticism that called for a change in thinking. If there was cause for serious concern, the informality of the prenotice process—in contrast to the rigidity of notice-andcomment rulemaking—would have afforded the agency an opportunity to quickly shift course. When reports surfaced just a week after the bulletin issued that “there was no backlash” to speak of, HHS learned something valuable about the acceptability of its chosen approach (Millman 2011a).

Essential Health Benefits and the Affordable Care Act

97

From the perspective of meaningfully involving the public in agency decision making, then, the bulletin-followed-by-rulemaking approach was superior to a routine process of APA notice-and- comment rulemaking. Significantly, it cleared a route for the agency to receive public comment at the critical prenotice phase of agency rulemaking. Commentators regularly lament that well-organized groups with concentrated interests have better access than diffuse public interest groups to this prenotice process where most important choices are made (Elliott 1992: 1492). The bulletin addressed this imbalance by telling everyone that the agency was open to hearing from them. Judicial Review By using a guidance document to announce the benchmark approach, HHS also arguably circumvented judicial review. Per section 702 of the APA, fi nal rules issued through notice-and- comment rulemaking can usually be challenged in court by those whose conduct the rule affects. In contrast, challenging guidance is much more difficult. Courts often conclude either that guidance is not fi nal agency action or that it is not ripe for review (Mendelson 2007: 411). Perhaps the perceived diffi culty with challenging guidance—and the expectation that the agency would soon resort to rulemaking, mooting any challenge that might be brought to the guidance—explains why no one tried to invalidate the bulletin. One purpose of judicial review is to assure that agencies do not force regulated interests to comply with rules that agencies lack the authority to issue (Abbott Laboratories v. Gardner, 387 U.S. 136 [1967]). It is therefore arguably worrisome that states, employers, and insurers did not have an opportunity to challenge HHS’s benchmark approach before making costly efforts to conform to that approach. Had HHS issued a proposed rule instead of the bulletin, the agency could have fi nalized the rule sometime in 2012 or early 2013. A more regular process would therefore have given affected interests an opportunity to challenge the fi nal rule before the requirement to cover essential health benefits sprang into force in January 2014. There’s something to this—but not much. The benchmark approach embodied in the bulletin will not evade pre- enforcement review altogether. At most, HHS’s decision to outline its approach in a bulletin allowed the agency to delay the date on which it issued a final rule. Now that

98

Nicholas Bagley and Helen Levy

the fi nal rule has issued, someone will probably bring a pre-enforcement challenge. Perhaps the plaintiff will be the mother of an autistic child who can fi nd no insurance plan in her state that covers needed services but who believes that, had the agency gone through the process of establishing a uniform federal standard, HHS might have included such services in the package of essential health benefits. Or perhaps it will be a California insurance company complaining that it must cover acupuncture services. The important point is that HHS knew when it issued the bulletin that any fi nal rule adopting the benchmark approach would eventually be challenged. The agency’s choice was therefore disciplined by the near- certainty that the courts would one day scrutinize that choice. HHS’s reliance on the bulletin just delayed when judicial review would occur. Nor is it especially worrisome that some states and insurers had to take immediate steps to comply with the approach HHS outlined in the bulletin. Even in the absence of fi nal agency action, parties often structure their affairs in anticipation of governmental action. Earlier review might have avoided some sunk costs: if the benchmark approach is invalidated, the efforts of states to select benchmark plans and of insurers to fashion new insurance products will have been wasted. Yet states and insurers could always have challenged the bulletin on the ground that it was a legally binding legislative rule. More significantly, there is no guarantee that HHS would have fi nalized its rule any more promptly had it issued a notice of proposed rulemaking instead of a bulletin. The agency might still have waited until the eleventh hour, devoting scarce resources to other pressing problems associated with implementation of the ACA. Against this backdrop, the notion that HHS used the bulletin to avoid judicial scrutiny is something of a stretch. OIRA Review Pursuant to Executive Order 12866, most important agency regulations are subject to OIRA review. Guidance documents are not. Late in his administration, President George W. Bush did issue an order subjecting guidance documents to OIRA review because they might otherwise “not receive the benefit of careful consideration accorded under the procedures for regulatory development and review” (OMB 2007: 3432). President Obama, however, rescinded that order shortly after taking office (Obama 2009). Although a memorandum from his budget director

Essential Health Benefits and the Affordable Care Act

99

clarified that significant guidance documents remain subject to review (Orszag 2009), review of guidance is in practice unsystematic and spotty (Nou 2013). Perhaps, then, HHS used a guidance document to sidestep the centralized oversight that is supposed to enhance the rationality and democratic legitimacy of agency decision making. In particular, because OIRA enforces the requirement that agencies undertake a rigorous, transparent regulatory impact analysis prior to taking significant actions, resorting to guidance allowed the agency to delay for more than a year any public effort to assess the costs and benefits of the benchmark approach. But did the delay of the regulatory impact analysis matter in practice? Some scholars have questioned whether such analysis has much effect at all on regulation (Hahn and Tetlock 2008). In the case of essential benefits, the delay in releasing the analysis may have been particularly inconsequential. Neither the analysis accompanying the NPRM in November 2012 nor the one accompanying the fi nal rule in February 2013 even bothered to estimate how costs and benefits might have differed if HHS had chosen a national, uniform defi nition of essential health benefits. Had they done so, it would probably not have made much of a difference to the estimated impacts. Because the scope of benefits under employer plans varies little across states, giving states the authority to select benchmark plans does not greatly affect the scope of the obligation to cover essential health benefits. The delay in providing a regulatory impact analysis, then, seems of little consequence. More generally, considerable evidence suggests that HHS did not “go rogue” and use guidance to evade presidential oversight. For one thing, HHS did share its bulletin with the White House. OIRA’s website reports that it received the bulletin from HHS on December 14, 2011, and cleared it (with some revisions) two days later—the same day HHS released it (OIRA 2011). Sources inside the administration have confi rmed that White House involvement ran much deeper. Administration officials were consulted about the benchmark approach as early as the summer of 2011, more than five months before the bulletin was issued. This close involvement is unsurprising. Deciding what counts as essential health benefits is perhaps the single most consequential policy choice that HHS will make in connection with the implementation of President Obama’s signal legislative achievement. As a matter of practical politics, HHS had no choice but to vet the bulletin at the very highest levels.

100

Nicholas Bagley and Helen Levy

Discussion At first blush, HHS’s release of a terse bulletin to announce a major regulatory decision looks unusual, even improper. It seems to reinforce the concern that agencies routinely use guidance documents to establish binding rules while evading the procedural obstacles that might otherwise deter them from acting (House Committee on Government Reform 2000). And it appears to confirm the wisdom of the academic consensus that guidance documents should be tolerated only grudgingly. Banning all guidance would be imprudent—better that regulated entities have some inkling of how agencies will carry out their duties than that they have none—but too much guidance risks undermining the procedural regularity of the administrative state (Mendelson 2007: 413; Raso 2010: 787). This consensus, however, rests on an unflattering view of administrative motivation. On this view, agencies are staffed not by public officials anxious to assure that their choices are workable and publicly legitimate, but instead by bureaucrats trying to avoid procedural obstacles that stand in the way of doing what they think best. Guidance is tempting precisely because it allows those bureaucrats to avoid the sort of public input, judicial review, and executive oversight that, by fostering accountability to a broader public, could impede their efforts. To put it in terms more familiar to political scientists: to the extent that administrative procedures allow political principals to better control their agents (McCubbins et al. 1987), agencies will use guidance to evade those procedures and exploit the slack between them and their principals. Doubtless this accurately describes some agencies some of the time. But what then should we make of the fact that HHS voluntarily replicated the very procedures that guidance is supposed to let it avoid? The agency may not have been driven only by a desire to secure public input on the benchmark approach—other motivations surely shaped the unusual regulatory process—but it is hard to conclude that the turn to guidance was a ploy to avoid accountability. What is more, HHS is hardly alone in adopting more administrative procedures than strictly necessary. Croley documents a series of important rulemakings from the late 1990s and early 2000s, including the EPA’s imposition of stringent controls on ozone and particulate matter, the FDA’s attempt to regulate tobacco products, and the Forest Service’s efforts to curtail road building in national forests. Although not a representative sampling, Croley’s examples “would all unquestionably make a short list of some of the most

Essential Health Benefits and the Affordable Care Act

101

significant regulatory activity in more than a decade” (Croley 2008: 160). And in every case, the agency “provided more notice, data, and opportunities for participation tha[n] the APA (or any other legal authority) demanded” (Croley 2008: 160). Mendelson (2007: 425) similarly identifies a number of agencies that make a habit of soliciting public input on guidance documents in the absence of legal compulsion to do so. In other words, there is nothing especially unusual about what HHS has done here. Far from ducking procedural obligations wherever possible, agencies sometimes embrace them. Why? At least for salient policy questions of substantial importance—a small but critical slice of agency action—agencies have a number of incentives having little or nothing to do with formal legal requirements to secure public input and assure political oversight. Doing so provides the agencies with technical information that might otherwise be difficult to obtain about how to craft policies that are capable of implementation. It arms them with scientific data that can help them better calibrate their rules. It teaches them about the political acceptability—and hence long-term sustainability—of the regulatory initiative. It identifies wellsprings of potential political support for (and opposition to) the rulemaking. It lends legitimacy to the regulatory initiative by assuring that all interested parties have had the opportunity to be heard both at the agency and in the courts. And it eases the concerns of those concerned that the agency is regulating by fiat. In sum, procedures can improve the workability and legitimacy of agency rules while protecting them from judicial or political attack (Croley 2008: 259).

Conclusion What, if anything, does this essential health benefits case study teach us about the future process of ACA implementation? We close with three observations. First, the intensely politicized nature of health reform will color anything and everything connected to implementing health care reform. The essential health benefits rule, for example, was issued against a backdrop of fierce partisan tension. The Supreme Court had agreed just a month before HHS posted the bulletin that it would consider a constitutional challenge to the so- called individual mandate—a challenge that, if successful, could have brought down the rest of the law with it. At the same time, governors and other state officials hostile to health reform

102

Nicholas Bagley and Helen Levy

were bridling at what many perceived as heavy-handed federal implementation efforts. Although we have chosen not to dwell on this political confl ict, it would be naive to suggest that politics played no role in HHS’s decision making. It would be equally naive to think it will play no role going forward. Second, ACA implementation will continue to depend on guidance. Part of the reason is the sheer volume of regulation that implementing the ACA requires. The Department of Health and Human Services, the Treasury Department, the Labor Department, and the other implementing agencies face daunting challenges rolling out the various programs that the ACA establishes. It is unrealistic to expect that each regulation will chart a clean course through the APA rulemaking process. Equally important, the digital revolution may itself encourage a turn to guidance. Long gone are the days when proposed rules were only put “on display” at the Office of the Federal Register—when there was a single point of official interaction between the regulators and the regulated. Agencies can now communicate directly with the public in ways that the formal regulatory process never envisioned. A vast literature considers how technology is transforming the relationship between government and its citizens (Mendelson 2011). Using the Internet to publicize guidance is an excellent example of this phenomenon. Third, the case study illuminates the larger debate about agency guidance. If agencies sometimes have powerful incentives to adhere voluntarily to administrative procedures, the distrust of guidance that runs like a leitmotif through much of the literature on administrative law seems misplaced. This is not to deny that agencies use guidance to avoid the costs of adhering to burdensome procedural requirements. Of course they do. That’s why issuing guidance is attractive to begin with. But do agencies systematically use guidance to avoid scrutiny from the public, the courts, and the president? The story of essential health benefits suggests that such evasion is far from inevitable. Nor does the available empirical evidence—slim as it is—lend support to the story. A recent review of significant guidance documents issued by five agencies over the span of a decade fi nds no indication that agencies routinely use guidance to shirk public accountability (Raso 2010). Perhaps, then, administrative lawyers should temper the reflexive assumption that agencies turn to guidance to avoid answering for their actions. It may be worth exploring the possibility that agencies are often sincere about what they use guidance for: to give regulated entities insight into their thinking, to

Essential Health Benefits and the Affordable Care Act

103

shape how line officials carry out their duties, and even—as with essential health benefits—to spur a public debate about the wisdom of a regulatory approach.

Notes 1. Large-group plans, such as those provided by large employers, are not required to provide essential health benefits, although they are subject to a different set of requirements governing the actuarial value of coverage. 2. The IOM did, however, explore the possibility of allowing states to deviate on a piecemeal basis from the national defi nition of essential benefits (IOM 2011a: 129).

References Anthony, Robert A. 1992. “Interpretive Rules, Policy Statements, Guidances, Manuals, and the Like—Should Federal Agencies Use Them to Bind the Public?” Duke Law Journal 41, no. 6: 1311– 84. Avalere Health. 2012. “EHB Implentation Progress.” State Reform Insights, updated October 4. www.avalerehealth.net/news/spotlight/index.php?year=2012. Bentley, Robert, Governor of Alabama. 2012. Letter to Kathleen Sebelius, Secretary, Department of Health and Human Services. October 1. governor.ala bama.gov/. Breyer, Stephen G., et al. 2002. Administrative Law and Regulatory Policy. 5th ed. New York: Aspen. Cassidy, Amanda. 2011. “Health Policy Brief: Essential Health Benefits.” Health Affairs (online), April 25. www.healthaffairs.org. CCIIO (Center for Consumer Information and Insurance Oversight), Department of Health and Human Services. 2011. Essential Health Benefi ts Bulletin. Washington, DC: CCIIO. cciio.cms.gov/resources/fi les/Files2/12162011/essen tial_health_benefits_bulletin.pdf. CBO (Congressional Budget Office). 2012. Updated Estimates for the Insurance Coverage Provisions of the Affordable Care Act. Washington, DC: CBO. www.cbo.gov/sites/default/fi les/cbofi les/attachments/03-13-Coverage%20 Estimates.pdf. Croley, Steven P. 2008. Regulation and Public Interests: The Possibility of Good Regulatory Government. Princeton, NJ: Princeton University Press. Department of Labor. 2011. Selected Medical Benefits: A Report from the Department of Labor to the Department of Health and Human Services. April 15. Washington, DC: Department of Labor. www.bls.gov/ncs/ebs/sp/selmedbensreport .pdf. Elliott, E. Donald. 1992. “Re-Inventing Rulemaking.” Duke Law Journal 41, no. 6: 1490– 96.

104

Nicholas Bagley and Helen Levy

Enzi, Michael B., et al., US Senators and Representatives. 2012. Letter to Kathleen Sebelius, Secretary, Health and Human Services. January 13. edworkforce .house.gov/uploadedfi les/1-13-12_letter_to_sebelius.pdf. Feder, J. Lester, and Jason Millman. 2011. “Guidance by Any Other Name.” Politico Pro (blog), December 15. www.politicopro.com. Final Rule. 2013. Patient Protection and Affordable Care Act: Standards Related to Essential Health Benefits, Actuarial Value, and Accreditation. 78 Fed. Reg. 12,834. Gersen, Jacob. 2007. “Legislative Rules Revisited.” University of Chicago Law Review 74: 1705– 22. Glied, Sherry. 2011. Conference Call Briefi ng with Secretary of Health and Human Services Kathleen Sebelius; Sherry Glied, Assistant Secretary for Planning and Evaluation; and Steve Larsen, Deputy Administrator and Director of the Center for Consumer Information and Insurance Oversight (CCIIO), Centers for Medicare and Medicaid Services. Transcript. Federal News Service, December 16. Hahn, Robert W., and Paul C. Tetlock. 2008. “Has Economic Analysis Improved Regulatory Decisions?” Journal of Economic Perspectives 22, no. 1: 67– 84 House Committee on Government Reform. 2000. Non- Binding Legal Effect of Agency Guidance Documents. H.R. Rep. 106–1009. Implementation of Health Insurance Exchanges and Related Provisions: Hearing before the Subcommittee on Health of the Health Committee on Ways and Means. 2012. 112th Cong., 2d sess. (statement of Michael Consedine, Commissioner, Pennsylvannia Office of the Commissioner, Department of Insurance). IOM (Institute of Medicine). 2011a. Essential Health Benefi ts: Balancing Coverage and Cost. Washington, DC: IOM. www.iom.edu/Reports/2011/Essential -HealthBenefits-Balancing-Coverage-and-Cost.aspx. IOM (Institute of Medicine). 2011b. Perspectives on Essential Health Benefi ts: Workshop Report. Washington, DC: IOM. www.iom.edu/Reports/2011/Per spectives-on-Essential-Health-Benefits-Workshop-Report.aspx. KFF (Kaiser Family Foundation). 2007. State Children’s Health Insurance Program (SCHIP) at a Glance. Menlo Park, CA: Kaiser Family Foundation. www.kff.org/medicaid/upload/7610.pdf. KFF (Kaiser Family Foundation). 2010. Focus on Health Reform: Explaining Health Reform: Benefi ts and Cost- Sharing for Adult Medicaid Benefi ciaries. Menlo Park, CA: Kaiser Family Foundation. www.kff.org/healthreform/up load/8092.pdf. McCubbins, Mathew D., et al. 1987. “Administrative Procedures as Instruments of Political Control.” Journal of Law, Economics, and Organization 3, no. 2: 243– 77. Mendelson, Nina. 2007. “Regulatory Beneficiaries and Informal Agency Policymaking.” Cornell Law Review 92, no. 3: 397–452.

Essential Health Benefits and the Affordable Care Act

105

Mendelson, Nina. 2011. “Rulemaking, Democracy, and Torrents of E-Mail.” George Washington Law Review 79: 1343– 80. Millman, Jason. 2011a. “First Crack at Essential Health Benefits Draws No Backlash.” Politico (blog), December 18. www.politico.com. Millman, Jason. 2011b. “IOM Authors Fault HHS on Benefit Safeguards.” Politico Pro (blog), December 21. www.politico.com. NHeLP (National Health Law Program). 2011. Letter to Kathleen Sebelius, Secretary, Department of Health and Human Services. December 2. Nou, Jennifer. 2013. “Agency Self-Insulation.” Harvard Law Review 126, no. 7: 1755. NPRM (Notice of Proposed Rulemaking). 2012. Patient Protection and Affordable Care Act: Standards Related to Essential Health Benefits, Actuarial Value, and Accreditation. 77 Fed. Reg. 70,644. Obama, Barack. 2009. Executive Order 13497. 74 Fed. Reg. 6113. OIRA (Office of Information and Regulatory Affairs). 2011. “Historical Reports.” Last modified January 2013. www.reginfo.gov/public/do/eoHistoricReport. OMB (Office of Management and Budget). 2007. Final Bulletin for Agency Good Guidance Practices. 72 Fed. Reg. 3432. Orszag, Peter R., Director, Office of Management and Budget. 2009. “Memorandum to Heads and Acting Heads of Executive Departments and Agencies.” March 4. www.whitehouse.gov/sites/default/fi les/omb/assets/memoranda_ fy2009/m09-13.pdf. Pear, Robert. 2011. “Health Care Law Will Let States Tailor Benefits.” New York Times, December 17. Raso, Conor N. 2010. “Strategic or Sincere? Analyzing Agency Use of Guidance Documents.” Yale Law Journal 119, no. 4: 782– 824. Reichow, Brian. 2012. “Overview of Meta-Analyses on Early Intensive Behavioral Intervention for Young Children with Autism Spectrum Disorders.” Journal of Autism and Developmental Disorders 42, no. 4: 512– 20. Schwartz, Sonya. 2012. “States Take a First Step on the Path to Essential Health Benefits.” Health Affairs (blog), October 3. healthaffairs.org/blog/2012/10/03 /states-take-a-fi rst-step-on-the-path-to-essential-health-benefits/. Sebelius, Kathleen. 2011. Conference Call Briefi ng with Secretary of Health and Human Services Kathleen Sebelius; Sherry Glied, Assistant Secretary for Planning and Evaluation; and Steve Larsen, Deputy Administrator and Director of the Center for Consumer Information and Insurance Oversight (CCIIO), Centers for Medicare and Medicaid Services. Transcript. Federal News Service, December 16.

Chapter five

The Fiscal Consequences of the Affordable Care Act Charles Blahous

Fiscal Effects of the ACA’s Use of Medicare Savings to Finance a New Health Entitlement

P

rior to the passage of the 2010 Affordable Care Act (ACA), there existed a fairly broad consensus that comprehensive reforms to federal health entitlement programs were necessary to ameliorate the federal government’s long-term fi scal imbalance. Granted, there were substantial disagreements between analysts as to the extent to which health care reform could by itself correct the fi scal outlook, as there were also disagreements over what shape those reforms should take. But there was general agreement that long-term projected federal spending growth was driven substantially by growth in federal health programs such as Medicare and Medicaid, and that corrective legislative action was required to limit these programs’ cost growth to levels that the underlying economy and tax base could sustain. This was by no means the only policy objective associated with health care reform, but it was regarded as an imperative by supporters and opponents of the ACA alike. Unfortunately the ACA as enacted in 2010 substantially worsened the federal fiscal outlook relative to prior law and previous budgetary practice. Even under an optimistic scenario in which its most controversial cost-savings provisions are all maintained and successfully implemented, the legislation would be expected to add over $340 billion to federal deficits and over $1.15 trillion to net federal spending over the fi rst ten years.

110

Charles Blahous

This fi nding flows directly from the projections and assumptions of the Congressional Budget Office (CBO) in scoring the legislation.1 Under a more pessimistic scenario in which some of the legislation’s most controversial provisions are reindexed to grow at rates more consistent with prior legislative practice, the legislation would add over $520 billion to deficits in the fi rst ten years, and over $1.24 trillion to net spending. This fi nding that the ACA worsens the fiscal outlook relative to prior law is unambiguous and results under a wide variety of plausible future legislative scenarios. This fiscal result is not everywhere understood, in large part because of incomplete appreciation of the scorekeeping methods under which the CBO operates dating back to the Deficit Control Act of 1985. 2 When the ACA was under legislative consideration, as well as in the years immediately following, the CBO published analyses that superficially appeared to show that the ACA would reduce future federal deficits. 3 CBO analysis, however (per existing congressional scorekeeping conventions), did not explore the fiscal consequences of the actual change of law embodied in the ACA. To fully understand this requires some familiarity with the basics of Medicare fi nancing. The ACA enacted substantial changes to the federal Medicare program. Medicare, like its sister program Social Security, is fi nanced in certain distinguishing ways. Like Social Security, the Medicare Hospital Insurance (HI) program is only permitted to make benefit payments to the extent there is a positive balance in its trust fund.4 If the trust fund were to be depleted, benefit payments would in effect be reduced through the mechanism of delaying such payments until sufficient revenues had come in to finance them. In deference to this requirement of law, legislators have periodically adjusted benefit and tax schedules in both Social Security and Medicare to prevent an interruption of benefit payments. CBO is instructed to ignore these aspects of Medicare fi nancing when it scores legislation. CBO instead assumes that both Medicare and Social Security will make full scheduled benefit payments in perpetuity without regard for whether these programs command adequate financial resources and statutory authority to do so. CBO routinely discloses in its reports that this assumption does not reflect actual law. 5 It should go without saying that overriding all restrictions on trust fund fi nancing to make full scheduled Medicare benefit payments in perpetuity would be an extremely, indeed unsustainably, expensive scenario

The Fiscal Consequences of the Affordable Care Act

111

relative to existing law. Yet this is the hypothetical scenario with which CBO is obliged to compare the effects of legislation. The ACA was projected to reduce federal deficits only relative to this untenable and extralegal scenario. It bears emphasis that this scorekeeping approach for Medicare is unusual among those under which CBO is generally obliged to operate. With many other areas of ongoing law, CBO models the literal effects of existing statutory text even where the law’s application is widely regarded as extremely unlikely. For example, for years lawmakers acted periodically to readjust the income thresholds of the Alternative Minimum Tax (AMT) for inflation before fi nally enacting automatic indexation as part of the resolution of the 2012–13 “fiscal cliff.” Despite the longstanding expectation of continuing legislative action to index the AMT thresholds, the scorekeeping conventions had CBO factor the assumption that the AMT thresholds would cease to be periodically raised into its baseline, because that is what statutory text specified.6 CBO thus scored each subsequent legislated increase in the AMT thresholds as adding to federal deficits. Similarly, for many years lawmakers have overridden pending Medicare physician payment reductions such as the 31 percent reduction that had been scheduled to occur in January 2013. Even though it was widely expected that Congress would act again to override this sudden payment reduction—which it eventually did—CBO was required to score legislation against the assumption that Congress would not. CBO thus scored the long- expected override as an addition to federal deficits.7 Thus even if it had been widely expected (which it was not) that Congress would later override statutory restrictions on Medicare spending beyond the resources in its trust funds, for CBO to be directed to assume such an override is quite atypical of general federal scorekeeping practices. CBO is usually directed to score the law as it is, whether its execution is regarded as plausible or not. In any event, the apparent CBO fi nding that the ACA would reduce federal deficits did not reflect the actual changes of law—particularly of Medicare law—embodied in the ACA. The ACA is merely expected to produce a better fiscal outcome than the overrides of existing Medicare law CBO is currently directed to use as its hypothetical scorekeeping baseline. By contrast, the ACA is expected to produce a much worse fiscal outcome than would have transpired under pre-ACA law.

112

Charles Blahous

Two summary explanations may help to condense this complex budgetary picture. The first involves understanding that prior to the passage of the ACA, there was a preexisting statutory obligation to keep the Medicare HI Trust Fund’s books balanced. Under a literal no-action scenario this would be enforced by the interruption of benefit payments at the point of trust fund depletion; past congressional behavior suggests that this fi nancing limitation would more typically be enforced instead by legislating specific increases in revenue or reductions in the growth of payments sufficient to prevent the trust fund from being depleted. But no matter which path was to be taken, this fi nancing constraint existed in law prior to the ACA. Thus the Medicare cost constraints contained within the ACA were enacted pursuant to a preexisting requirement; had they not been enacted, alternative savings measures would have been necessitated to avert trust fund insolvency. But while this statutory fi nancing discipline was in effect prior to the ACA, the legislation’s enormous new spending initiatives—an ambitious expansion of Medicaid and new federally subsidized health exchanges—were in no way required under prior law. The ACA thus coupled previously required savings provisions with wholly new spending. This combination worsened the fiscal outlook. Another way of understanding the fiscal worsening is by reference to the self-financing construct of the Medicare HI program. As with Social Security, spending in Medicare HI is constrained by the level of its incoming revenues. If revenue is insufficient to fund scheduled benefits, only a lower level of benefit payments can be made. If the program brings in more tax revenue, it is permitted to spend more on benefits. Because additional savings in either Social Security or Medicare increase the amount of other spending these programs are allowed to do, such savings cannot simultaneously be employed to fi nance new spending elsewhere in the federal budget (such as the new health entitlement in the ACA). The savings provisions of the ACA are insufficient both to extend the solvency of Medicare and to fund a massive health coverage expansion. Overall the new spending authorized by the ACA well exceeds the savings generated by its other provisions, so on balance it adds substantially to federal deficits. Well before the ACA was drafted, federal lawmakers on both sides of the aisle appreciated this general phenomenon. This is why congressional pay-go rules typically prohibit the use of savings within Social Secu-

The Fiscal Consequences of the Affordable Care Act

113

Figure 5.1 Net Annual Budgetary Effect of the ACA. Source: Figure 7 in Charles Blahous, “The Fiscal Consequences of the Affordable Care Act,” Mercatus Center at George Mason University, April 10, 2012. Author’s calculations based on CBO and CMS data and projections. The figures above are positive if they improve the budget outlook and negative if they worsen deficits.

rity, for example, to fi nance other federal spending. There was a loophole in the scoring rules for Medicare, however, which permitted the ACA to be scored as reducing the deficit in a way that it would not have if such restrictions had been universally applied. This can readily be seen by comparing the ACA’s fiscal effects relative to prior law with those that result from the scorekeeping conventions that bind CBO. The upper two lines on the graph show the apparently positive long-term fiscal effects of the legislation relative to a scenario in which Medicare trust fund fi nancing restrictions are overridden. These upper two lines differ because one includes the effects of the ACA’s Community Living Assistance Services and Supports (CLASS) provision, originally part of in the enacted law but suspended the following year due to its fi nancial unworkability. The bottom three lines on the graph show that under a wide range of possible implementation scenarios, the legislation will worsen the fiscal outlook relative to prior law, and that this worsening will grow more severe over time. After this study was published the Supreme Court issued a ruling that could bear significant ramifications for the ultimate fiscal effects of the

114

Charles Blahous

Figure 5.2 Updated: Annual Budget Effect of the ACA. (Reflects CBO 2012 re- score after SCOTUS Medicaid ruling) “Author’s calculations based on CBO and CMS data and projections.” Elmendorf, Douglas W. “Direct Spending and Revenue Effects of H.R. 6079.” Letter to John Boehner. 24 July 2012. MS. Congressional Budget Office, Washington, D.C.

ACA. In June 2012, the Court upheld the constitutionality of most of the ACA but struck down the federal government’s power to compel the states to participate in the law’s ambitious Medicaid expansion. CBO’s updated analysis pursuant to the ruling shows the projected long-term effects of the ACA qualitatively unchanged. CBO now expects that some of those who would have been enrolled under Medicaid will remain uninsured (reducing federal costs) while some will be covered under the law’s new health exchanges (increasing federal costs). On balance CBO slightly reduced its cost estimates for the ACA’s coverage provisions, while its estimates of other provisions grew slightly more pessimistic. The new figures still suggest that the ACA will add over $340 billion to federal deficits over the next ten years (see Figure 5.2) and over $1.2 trillion to net federal spending. As earlier noted, the largest reason for the difference between the evaluation relative to prior law and the evaluation relative to the CBO scoring convention pertains to the treatment of the ACA’s Medicare savings. These Medicare savings provisions are substantial and projected to grow dramatically over time under CBO scorekeeping assumptions. The Medicare savings are much smaller, however, relative to the cost constraints already operative under pre-ACA law (see Figure 5.3). The ACA does contain some significant net reductions in future Medi-

The Fiscal Consequences of the Affordable Care Act

115

Figure 5.3 Medicare Savings under the ACA. Source: Author’s calculations based on CBO and CMS Medicare actuary projections, updated for 2011 data.

care spending, including reductions in the growth of Medicare Parts B and D that there was no prior obligation to enact. But on balance these net savings are insufficient to fi nance the law’s other increases in federal spending.

Practical Consequences for Federal Budgeting It is an understandable temptation to believe that the ACA’s practical effect on federal fi nances might be more positive than is captured by measuring the effect of the literal change in law as quantified in the preceding section of this paper. Given that the scorekeeping baseline CBO is directed to use differs in a fundamental way from actual law, the question arises whether this baseline is used because it more accurately reflects past and likely future budgetary practice than actual law does. This is not, however, the case. The ACA’s worsening of the fi scal outlook becomes even more evident when considering decades of historical budgetary practice. The scorekeeping baseline used to suggest a favorable budgetary effect for the ACA is in fact without legislative precedent. It implicitly assumes that both Medicare HI and Social Security will be given an open tap on general revenues and permitted to spend on scheduled benefits without restraint. Such a scenario would require

116

Charles Blahous

not only a change in law but a marked departure from past legislative behavior. Congress’s historical management of Social Security and Medicare fi nances may well be imperfect and is justly subject to critical scrutiny, but lawmakers have nevertheless generally respected the limitations of trust fund fi nancing. The evidence of this is extensive, but perhaps the best anecdotal example is embodied in the 1983 Social Security reforms. In 1983 Social Security was rescued from the brink of insolvency by a comprehensive reform package containing many controversial measures, including a six-month delay in the program’s cost- of-living adjustments, the exposure of benefits to income taxation, a retirement age increase, an accelerated payroll tax increase, and bringing new federal employees and their payroll tax contributions into the system.8 These measures exposed supporting lawmakers to great political risks, including the spirited opposition of the powerful seniors’ lobby, the American Association of Retired Persons. The legislative effort was undertaken, however, pursuant to lawmakers’ collective responsibility to maintain Social Security solvency. None of these measures would have been necessary if it had been deemed acceptable simply to allow Social Security to spend in excess of the resources in its trust fund, as is assumed in the CBO baseline for the ACA.9 The conventional scorekeeping methodology thus not only compares the ACA to a scenario that represents a substantial change in law, but to a scenario fundamentally at odds with historical practice. Historically, both with Social Security and Medicare HI, lawmakers have been more likely to enact cost-savings provisions when insolvency was approaching than when it was more distant. This behavioral pattern provides further evidence that the ACA has worsened the fiscal outlook. Having just enacted substantial Medicare HI cost constraints and postponed the projected date of Medicare’s insolvency from 2016 to 2024, lawmakers are under lessened pressure to enact additional Medicare cost constraints. The ACA could only represent a net improvement in federal finances if lawmakers stood just as ready after its passage to enact further Medicare cost reductions as they were in 2009. This is clearly, however, not the case. The ACA’s reduction of the projected Medicare HI seventy-five-year actuarial imbalance from $13.4 trillion (in present value) to $2.4 trillion and its postponement of the HI insolvency date from 2016 to 2024 permit lawmakers to enact far fewer future cost constraints henceforth than

The Fiscal Consequences of the Affordable Care Act

117

would be the case if HI insolvency were still projected to be 2016 and its actuarial shortfall $13.4 trillion.10 Indeed, some of the ACA’s most vocal advocates argue that one of its positive achievements was to postpone the insolvency of Medicare HI and thereby spare seniors the threat of a benefit interruption in 2016.11 If that is true—that is, if the Medicare savings in the ACA permit an extension of secure and uninterrupted Medicare benefit payments—then those same Medicare savings cannot be used to fi nance a vast health coverage expansion without increasing the federal deficit. The assertion that the ACA extends Medicare HI solvency from 2016 to 2024, and that this is an economically meaningful effect, is implicitly an acknowledgment that the ACA has worsened the overall federal fiscal imbalance. These adverse fi nancial effects of the ACA cannot be wished away, as is sometimes more persuasively done with other changes to federal law. There are indeed other areas of law where one could make a strong argument that statutory text is not the best guide to future fiscal practice: examples include the aforementioned AMT and physician payments under Medicare Part B. In both of those issue areas there is an ample historical record of legislative overrides of statutory fiscal restraints. This is not the case, however, with respect to the fi nancing restrictions of the Medicare HI Trust Fund. Historically Medicare law has mattered; lawmakers have not yielded to the temptation to simply do away with its restrictions and give the program an open tap on general revenues, as is assumed in the scorekeeping baseline. Arguing against this point, others have correctly noted that it is customary for CBO’s scorekeeping methods to ignore certain statutory restraints upon federal spending, such as with the statutory debt ceiling and when projecting future appropriations levels.12 For example, CBO does not assume that all future annual appropriations will vanish, even though the actual appropriations will not be provided for until the required legislation is enacted. Nor does CBO assume that future spending will be limited by the existing statutory debt ceiling. However, CBO’s scorekeeping rules ignore these limitations for the obvious practical reason that historically they have not been useful predictors of subsequent spending levels; basing CBO scorekeeping solely on such phenomena as the statutory debt ceiling would indeed destroy its ability to provide useful information about the wide array of legislation affecting federal finances. Thus it should not be inferred that because CBO scorekeeping rules ignore certain elements of law, they should ignore all

118

Charles Blahous

elements of law. To do so would be to destroy any utility of CBO scorekeeping. The practical reality remains that as a general historical rule, Medicare payment levels have been best predicted by actual Medicare law and not by such tangential phenomena as the statutory federal debt ceiling. Thus, approaching the federal budget from a variety of practical perspectives only reinforces the conclusion that the provisions of the ACA, taken as a whole, have substantially worsened the federal fiscal outlook.

Alternative Fiscal Scenarios The first and most important basis for a fiscal evaluation of the ACA is to compare its projected fiscal effects with those of prior law, without regard for the political plausibility of either scenario. But as discussed earlier in this paper, federal scorekeepers are often confronted with situations in which the literal path of current law is regarded as extremely unlikely to be implemented as written. This raises the question whether a literal current-law projection is always the only relevant or even the most meaningful one. Referring to one aforementioned example, it was never regarded as likely that Medicare physician payments would suddenly be cut by 31 percent in January 2013, even though this is what the text of law specified for some time. Such discrepancies between what is written in statute and what is deemed likely to occur create challenges for federal scorekeepers. On the one hand, the scorekeeper’s primary job is to model the effects of current law, not to construct and evaluate alternative policies. On the other hand, failing to account for near- certain changes to future law can create a distorted picture of the likely state of federal fi nances. The government’s nonpartisan scorekeepers, including both CBO and the Medicare/Social Security trustees, have sometimes approached this problem by presenting multiple baseline scenarios. One scenario presents literal law as written, without regard for political plausibility. Other illustrative baselines may involve extensions of current government policies that require future changes in law. The construction of these alternative baselines necessarily involves subjective judgments by scorekeepers concerning which current policies are likely to be extended, as well as what specific legislative changes would be most consistent with recent policy practice. The scorekeeping bodies typically take care to emphasize that these alternative scenarios are illustrative and do not represent

The Fiscal Consequences of the Affordable Care Act

119

either a fi rm political prediction or a policy recommendation. The information is provided to lawmakers purely so that the fiscal consequences of certain likely deviations from current law are quantified and understood. For example, the 2012 Medicare trustees’ report includes both a projection of literal law as then written (assuming among other things a 31 percent cut in physician payments in 2013) and another projection showing what costs would be if lawmakers continued to override pending physician payment cuts roughly as has been done on average over the last decade. The trustees describe these projected overrides as a “virtual certainty” and thus opine that “future Medicare costs could be substantially larger than in the trustees’ current-law projections.”13 The trustees’ report illustrates another alternative scenario in which various provider payment reductions called for under the ACA are phased out over the period from 2020 to 2035. In a footnote to the report, the trustees state that “readers should not infer any endorsement of the policies represented by the illustrated alternatives.” The alternatives are simply provided to acquaint lawmakers and the public with the fiscal consequences of reasonably likely legislative overrides. CBO engages in a similar exercise with many provisions of existing law.14 The ACA presents a particularly salient version of this scorekeeping dilemma because the implementation of many of its provisions would require lawmakers to behave in the future in ways different than they have in the past. It is of course theoretically possible that all of the ACA’s provisions will be fully implemented as written, even those that require a sharp break with prior legislative practice. Under this scenario the law would add roughly $340 billion to federal deficits over the next ten years and roughly $1.2 trillion to total federal spending. But it is also important to know what the fiscal consequences will be if the ACA’s provisions are implemented in ways more consistent with historical practice. The April 2012 study “The Fiscal Consequences of the Affordable Care Act” did not explore all potential modifications to the cost-savings provisions of the ACA, even though many of these provisions have already proved controversial. Many analysts have opined, for example, that the aggressive provider payment reductions in the ACA will prove unworkable over time because they will eventually result in payment rates under Medicare that are but a small fraction of those projected for the private health care sector.15 If these analyses are correct, then unless lawmakers repeal the ACA’s Medicare cost constraints beneficiaries will lose access to care, as providers either withdraw from the program

120

Charles Blahous

or are unable to remain profitable. On the other hand, there are experts who believe that this payment-rate wedge will not materialize because the ACA will facilitate greater efficiencies in both Medicare and private sector health care. I take no position on this controversy other than to note that we are heading into uncharted waters with respect to whether such aggressive Medicare payment restraints can be sustained over long periods of time, rendering predictions of future legislative behavior extremely uncertain. Since the April 2012 study was published, other provisions of the ACA have been also subject to calls for repeal. In 2013 a new 2.3 percent tax established by the ACA on the sale of certain medical devices became effective.16 Although the original ACA was passed in 2010 by congressional Democrats essentially on a party-line basis, in December 2012 18 Democratic senators and senators- elect called for repeal of the new tax, which was intended to help defray the cost of the ACA’s subsidized coverage expansion.17 While it must be assumed that this tax will remain in force unless repealing legislation is enacted, there is clearly now a nontrivial risk that this particular fi nancing source may not produce the full amount of revenue now projected for it. While the fate of such provisions as the ACA’s Medicare provider reductions and medical device tax will be settled by yet-unpredictable legislative events, history suggests that other ACA fi nancing sources are still less likely to be sustained as currently written. The ACA imposes a new 3.8 percent tax on investment income (the so- called Unearned Income Medicare Contribution or UIMC, although it is not provided to the Medicare trust funds) for those with incomes above $200,000 for singles and $250,000 for married couples. Though set to initially capture only a small minority of taxpayers, the tax’s income thresholds are not indexed for inflation, meaning that over time more and more taxpayers will be subject to it. The 2012 Medicare trustees’ report projected that by the end of the trustees’ long-term valuation window fully 80 percent of taxpayers would be affected by the tax.18 To implement the UIMC as written would require lawmakers to behave quite differently over the long-term future than they have with other nonindexed tax provisions in the past. The AMT, for example, was also originally written into law so that its income thresholds would not rise over time, thereby capturing rising numbers of taxpayers. Congress nevertheless legislated frequent increases of the AMT income thresholds before fi nally indexing them to inflation in early 2013. If lawmakers

The Fiscal Consequences of the Affordable Care Act

121

treat the UIMC in the future as they have the AMT, the UIMC will produce far less in revenues than the rapidly escalating amounts now projected for it. The ACA also contains a so- called Cadillac plan tax provision expressly designed to capture more health insurance plans over time, by virtue of being indexed to grow with general U.S. price inflation rather than the higher rate of cost growth historically associated with health insurance. There is already considerable reason to believe that the rapidly rising revenues now projected for this tax will not fully materialize. Indeed the Cadillac plan tax as initially passed did not even survive its fi rst clash with political realities, being immediately amended in nearly simultaneous budget reconciliation legislation to postpone its effective date to 2018 and reduce the number of plans to which it would apply. When 2018 approaches more closely it is virtually certain that there will be energetic political opposition to applying the tax as currently written. For the tax to be fully implemented would require future legislators to be more completely committed to it than were the original authors of the ACA. These are by no means the only fi nancing provisions of the ACA that remain highly controversial and subject to substantial risk of weakening or of outright repeal. The ACA’s subsidies for low-income purchasers of exchange-based health insurance, for example, are designed so that over time these vulnerable individuals must shoulder a steadily rising percentage of their own health care costs.19 Also intensely controversial is the ACA’s Independent Payment Advisory Board (IPAB), empowered to make further unspecified reductions in Medicare benefit payments sufficient to hold program cost growth to the rate of per capita GDP plus 1 percent. Should the IPAB be weakened or eliminated, as many lawmakers have announced their intentions to do, post-ACA Medicare costs may well be higher than currently projected. 20 It is not necessary for all of these various controversial provisions to be fully repealed for the ACA to have a considerably more adverse effect on the federal budget than now projected. Even if lawmakers simply reindex the 3.8 percent UIMC and Cadillac plan tax so that these revenues grow in proportion to GDP and also reindex the health exchange subsidies to allow them to keep pace with national health cost growth while repealing only the IPAB, the ACA would add more than $520 billion to federal deficits and more than $1.24 trillion to total federal spending over its fi rst ten years. 21

122

Charles Blahous

Again, this does not represent a prediction or a recommendation that these provisions be reindexed in this manner. Policy makers and the public should be kept fully aware, however, that the fiscal damage of the ACA will be less only if lawmakers allow specific tax and healthcare cost burdens to mount much more in the future than elected officials have historically been willing to countenance.

Fiscal Corrections The foregoing analysis prompts the question what modifications of the ACA are required to prevent it from having a damaging budgetary impact. The answer depends on the fi scal standard applied. If the fi scal goal is as originally stated, to scale back the untenable projected growth in federal healthcare spending, the vast majority of the ACA’s coverage expansion subsidies would need to be eliminated or alternative fi nancing sources found. Without alternative fi nancing sources, meeting this standard would require eliminating roughly $1.2 trillion in new spending under the legislation over the next ten years. This could be accomplished by eliminating the entirety of the law’s subsidies for new exchanges in addition to roughly two-thirds of its Medicaid expansion. 22 Many of the law’s advocates would regard this measure as defeating another essential purpose of the legislation, that being a substantial expansion of national health insurance coverage. Another possible and more relaxed fiscal standard is the one embraced by the Obama administration closer to the time that the legislation was being fi nalized and voted upon—that is, to guarantee merely that it not add to federal deficits in its fi rst ten years. 23 This would still permit the ACA to add further costs to the fastest-growing portion of the federal budget, and should thus not be regarded as embodying genuine healthcare reform. It would, however, at least be somewhat consistent with the goal of fiscal consolidation, albeit achieved solely through tax increases. To provide for a reasonable degree of protection against fiscal damage, this standard should be met not only under the assumption that all of the ACA’s cost-savings provisions will be fully enforced as originally written, but also assuming that they evolve in the future according to historical precedent. This more relaxed fi scal standard could be met while leaving the law’s Medicaid expansion fully in place, but roughly

The Fiscal Consequences of the Affordable Care Act

123

two-thirds of its exchange subsidies would still need to be eliminated or alternative fi nancing sources found. Regardless of which fiscal standard is pursued, if the ACA is not to worsen the budgetary outlook corrections need to be enacted before its coverage expansion is fully implemented. History shows the tremendous political difficulty of slowing cost growth in federal programs once beneficiaries become dependent on them, even after the programs become as expensive and as fi nancially troubled as Social Security and Medicare currently are. Thus the window of opportunity for enacting fiscal corrections to the ACA was open widest prior to 2014 and is in the process of closing as individuals enroll in the exchanges and states implement their Medicaid expansion plans. If the ACA’s substantial coverage expansion is not considerably scaled back before these processes are completed, it thereafter becomes quite certain to have substantial adverse budgetary effects, with subsequent legislation likely confi ned to rearguard actions to contain the damage.

Alternative Viewpoints The aforementioned analysis showing that the ACA has worsened the fiscal outlook inevitably provokes some concerns among ACA’s advocates as well as others. This section answers some of the questions that have been raised. Consistency with CBO. The fi rst question that naturally arises is whether these results contradict CBO’s published analysis, which appears to show that the ACA would reduce projected federal deficits. There is no contradiction; indeed the analysis underlying this study borrows heavily from CBO’s participation assumptions as well as from its evaluation of the ACA’s fiscal effects. Any apparent contradiction is simply the result of the respective studies analyzing two different questions. Whereas this study evaluates the fiscal effects of the ACA in comparison with prior law, per congressional scorekeeping conventions CBO’s evaluates in comparison with an alternative hypothetical scenario in which statutory Medicare and Social Security fi nancing restrictions are overridden in future legislation. The two studies anticipate similar results but compare with two different alternatives. Unified budget perspective. A second question that has been raised is whether this analysis depends unduly on trust fund accounting conventions

124

Charles Blahous

instead of heeding more meaningful realities of the unified federal budget. Specifically, the question pertains to whether this analysis treats exchanges of intragovernmental debt (such as debt issued to the Medicare trust funds) as being more significant economically than they actually are.24 Many economists believe that the most appropriate frame of reference for analyzing federal deficits and debt is with respect to the unifi ed federal budget, referencing only debt borrowed from the public. In this perspective debt issued to federal trust funds is less meaningful economically because the government need not enter into credit markets to issue intragovernmental debt; its ability to issue such debt is limited only by restrictions the federal government places on itself. Because this study references the bonds held by the Medicare trust fund when quantifying the benefits payable under current law, some have suggested that the fi ndings might be different if trust fund debt were ignored and only unified budget operations considered. As one example, the Committee for a Responsible Federal Budget stated that “if you break away from trust fund accounting and use a unified budget perspective in which benefit levels are fully funded, this . . . doesn’t really change the estimate of the Affordable Care Act (relative to CBO estimates).” 25 However, the fi nding that the ACA worsens the federal fiscal outlook is fully consistent with a unified budget perspective. The key distinction in the quote above is not the budget perspective adopted but rather the key specification that scheduled benefit levels will be “fully funded.” It is this specification that is determinative; it invokes a particular legislative scenario in which current law is changed in important ways to permit the perpetual payment of now- scheduled Medicare benefit levels. The ACA, by contrast, worsens deficits relative to how prior law would have determined Medicare benefit levels; it mitigates deficits only relative to the payment of higher benefit levels requiring overrides of existing law. Both of these fi ndings reflect a unified budget perspective. Unlikelihood of Medicare insolvency. This paper has already discussed how, to appreciate that the ACA has worsened the fiscal outlook, it is not necessary to believe that otherwise the Medicare trust fund would have been depleted and benefit payments suddenly interrupted as projected under a literal reading of law. All that is needed is an appreciation of the historical norm in which lawmakers have taken seriously their obligations to keep Medicare fi nances in balance. Lawmakers have not historically permitted the wholesale overrides of trust fund fi nancing restrictions assumed in the CBO baseline.

The Fiscal Consequences of the Affordable Care Act

125

The implausibility of the conventional scorekeeping baseline. It is also worth noting that while it is highly unlikely that Congress would permit sudden benefit payment interruptions as a result of trust fund insolvency, the baseline used to suggest a favorable fiscal effect for the ACA is equally implausible. Such a baseline assumes that full scheduled Medicare benefit payments are made in perpetuity regardless of fi nancial resources, statutory authority, or pressures elsewhere in the federal budget. Applying such a “current policy” approach to projecting all components of the budget would produce projections of unsustainable deficits, yet these baseline projections assume limited effects on such phenomena as interest rates and economic growth. 26 Accordingly, an analytical approach based on judgments of plausibility does not favor the existing scorekeeping convention. The seriousness of the long- term fi scal problem. Others have correctly noted that the long-term federal fi scal imbalance does not look nearly so daunting if it is assumed that, as a literal current-law interpretation requires, Social Security and Medicare are kept in fi nancial balance. Since most observers do not regard the federal government’s long-term fiscal picture as looking so benign, some are tempted toward the conclusion that a fi nding based on this picture that the ACA worsens deficits must therefore be disbelieved. 27 This disbelief reflects an apparent misunderstanding of the nature of the projected long-term fiscal imbalance. The fiscal situation facing the U.S. government is such that much of the projected shortfall consists of an imbalance between scheduled Medicare/Social Security revenues and benefit payments. 28 If these program imbalances were to be addressed in accordance with longstanding statutory requirements, then indeed much of the projected federal fiscal imbalance would be eliminated. It is only in comparison with a scenario in which these statutory balancing obligations are ignored or overridden that the ACA appears to improve the long-term fiscal outlook. A hypothetical example may render it clearer how the ACA’s fiscal policy approach worsens federal fi nances. Suppose that legislation is enacted to bring Social Security and Medicare benefit and revenue schedules closer to balance by $1 trillion. Under the ACA’s approach to budgeting, no fiscal harm would be seen if that $1 trillion were spent on an entirely new program in the same legislation. But following such a policy self- evidently worsens federal fi nances. Applying the policy consistently would ultimately replace projected federal fi scal imbalances, which would be prevented by continued adherence to current Social Security

126

Charles Blahous

and Medicare law, with large projected fiscal imbalances that nothing in those programs’ statutory constrictions would prevent. Ignoring other aspects of current law. As earlier mentioned, some have also argued that Medicare fi nancing restrictions should be ignored in fiscal analysis because certain other aspects of current law, such as the debt ceiling, are also ignored under Congress’s conventions. 29 This suggestion, however, leads to a reductio ad absurdum: if the law is everywhere ignored, then all analyses of statutory changes to fiscal policy become meaningless. Historically, Medicare spending levels have been much better predicted by Medicare law than by other tangential phenomena such as the statutory debt ceiling. Current law versus current policy. Some objections voiced to these fi ndings reflect confusion about congressional scorekeeping rules. For example, New York Times columnist Paul Krugman asserted “in general, you almost always want to assess legislation against ‘current policy,’ not ‘current law’; there are lots of things that legally are supposed to happen, but that everyone knows won’t, because new legislation will be passed to maintain popular tax cuts, sustain popular programs, and so on.”30 While it is true that analysts frequently debate whether “current policy” or “current law” is a more reliable predictor of future federal fi nances, the congressional scorekeeping treatment of Medicare is actually an anomaly among current practices; income tax legislation, as has been mentioned, is officially scored against “current law” even when current law contains pending expirations widely regarded as improbable and when periodic overrides have been the historical norm. It is atypical of general scorekeeping practices that the existing conventions direct CBO to assume overrides of Medicare fi nancing restrictions although such overrides—it should be noted—have not been the historical norm. 31 (In any case, legislation that adds to the deficit should be scored as such, even if there is a precedent for it.) Assumptions re Medicare provisions. There exists an ongoing debate over whether the ACA’s provider payment restraints will prove workable over time. 32 The 2012 study did not take sides in this controversy or consider possible overrides of these provisions. Perhaps because of awareness of this ongoing debate, some of the ACA’s advocates incorrectly assumed that this study posited that the ACA’s Medicare cost- containment measures would eventually be overridden. 33 This is not the case; the fi ndings in this study assume that the ACA’s Medicare savings provisions

The Fiscal Consequences of the Affordable Care Act

127

are maintained and enforced, with the minor exception of the IPAB in only the pessimistic and mixed-outcome scenarios. Follow-up analysis has confirmed that the ACA will add to federal deficits if it is assumed that all aspects of law are enforced without regard to political plausibility, as well as under scenarios in which judgments of political plausibility leaning heavily in the ACA’s favor are applied.34 In sum, the unwelcome conclusion that the ACA will add substantially to federal deficits is upheld even after considering various possible objections and alternative approaches to the analysis. There is no significant debate about whether Medicare spending is limited under law by its trust funds; nor is there significant debate over the fact that CBO scorekeeping methods do not reflect this critical aspect of law. Though the policy case for and against the ACA is complex and involves many important and subjective value judgments, it is unambiguous that at least in comparison to actual prior law it has worsened the federal fiscal outlook.

Recent Developments Since the original version of this analysis was published a number of events have transpired that bear upon an analysis of the ACA’s fiscal effects. Among the most important of these is the aforementioned Supreme Court ruling fi nding the ACA’s compulsory expansion of Medicaid to be unconstitutional. Subsequent to the ruling CBO changed its assumptions, from universal state participation in the full Medicaid expansion to a wide range of state participation decisions. This lowered CBO’s projected increase in Medicaid coverage by roughly one-third relative to previous estimates. This change in CBO assumptions did not qualitatively change the expected fiscal effects of the ACA; CBO projected that federal costs would be reduced to the extent potential Medicaideligibles would remain uninsured, while federal costs would be increased to the extent these individuals would move into the ACA’s new health exchanges. CBO also estimated that other noncoverage portions of the ACA would have less favorable fiscal effects than previously projected. 35 Taken as a whole the updated analysis results in expectations that the ACA will add roughly $340 billion to federal deficits over the next ten years, and about $1.2 trillion to total federal spending. 36

Charles Blahous

128

Around the same time CBO took steps to explain its scorekeeping conventions more transparently, pursuant to many of the points made in this study. In its Long-Term Budget Outlook of June 2012, CBO further clarified that its projections under scorekeeping conventions necessarily deviate from what would transpire under law: Once the HI trust fund was exhausted, it appears that total payments to health plans and providers for services covered under Part A of Medicare would be limited to the amount of revenues subsequently credited to the trust fund. If that occurred, beneficiaries’ access to health care services would almost certainly be reduced. However, projections in this report are consistent with a statutory requirement that CBO, in its baseline projections, assume that benefit payments will continue to be made after trust funds have been exhausted, even if there is no legal authority to make such payments. 37

CBO’s Long-Term Outlook also contained an illustrative alternative scenario similar to this study’s exploration of “mixed outcome” and “pessimistic” fiscal scenarios. In this scenario the ACA’s health exchanges are permitted to grow so that eligibility and per ,capita federal support will not decline over time as would happen if the provisions of the ACA remained unchanged. 38 This policy alternative is labeled CBO’s “extended alternative fiscal scenario.” The point of CBO’s illustration is the same as for the one contained within this study: while the primary analysis must be of the ACA as now written, attention should also be paid to fiscal outcomes that could transpire if lawmakers behave more in accordance with historical precedent. The original publication of this analysis included statements by the author to the effect that existing scorekeeping rules are generally appropriate but have simply been misunderstood in the particular case of the ACA. Since that time the author’s view of these issues has evolved, as reflected in a more recent article suggesting that Congress consider changing these scorekeeping rules. 39 Problems with the existing rules include the facts that they do not fully reflect current law, that they are inconsistent across federal programs in their treatment of current law and current policy, and that they create incentives for fiscally irresponsible behavior. Events continue to unfold that could change the ultimate fiscal effects of the ACA. After the Supreme Court ruling a number of state governors

The Fiscal Consequences of the Affordable Care Act

129

announced that they would not participate in the Medicaid expansion that federal lawmakers had attempted to make compulsory through the ACA.40 Other governors inquired of HHS Secretary Kathleen Sebelius whether states must expand coverage to the ACA’s specification of 138 percent of the Federal Poverty Level (FPL) in order to receive the ACA’s enhanced federal match rate.41 This is important because the combination of the Supreme Court decision and the ACA’s statutory text gives states powerful incentives to limit Medicaid coverage to those below 100 percent of FPL, allowing childless adults at higher income levels to receive more generous health insurance subsidies fi nanced entirely by the federal government.42 In December 2012 Secretary Sebelius replied that states would not receive the ACA’s generous federal match rate for a partial expansion.43 Though clearly intended to coax states to expand Medicaid more fully per the terms of the ACA, HHS’s decision may instead result in fewer states expanding Medicaid at all, potentially dampening the ACA’s costs relative to prior estimates. The ACA’s CLASS program was suspended in October 2011 when an actuarial analysis required by a provision inserted by then- Senator Judd Gregg (R-NH) determined that the program was fi nancially unworkable. More recently the CLASS program was repealed outright as part of the so- called fi scal cliff legislation enacted at the start of 2013. These actions mitigate the ACA’s potential long-term fiscal damage, though they worsen the scoring of the legislation in its fi rst ten years, during which time CLASS had been projected to collect premiums in excess of benefit payments.44 Though CLASS’s repeal was not legislated until early 2013, this analysis had already accounted for the demise of CLASS resulting from HHS’s prior suspension of the program. As of this writing, a number of decisions that will affect the ultimate fi nancial effects of the ACA remain unmade and in the hands of both state and federal officials. These decisions could yet mitigate the fi nancial damage of the ACA or they could further worsen it.

Conclusion The enactment of the Affordable Care Act in 2010 unambiguously worsened the federal fi scal outlook relative to previous law. Public understanding of this has been incomplete, largely due to misunderstanding

130

Charles Blahous

of federal publications comparing the effects of legislation not to prior law but to a hypothetical baseline alternative in which certain statutory Medicare financing restrictions are legislatively overridden. Specifically, the ACA would only reduce federal deficits relative to a scenario in which Medicare makes full scheduled benefit payments irrespective of limitations upon its fi nancial resources and statutory authority; the ACA would substantially worsen federal deficits relative to the law as it stood prior to its enactment. Assuming that the provisions of the ACA remain unchanged, it would be expected to add roughly $340 billion to federal deficits over the next ten years and approximately $1.2 trillion to net federal spending. If key provisions of the ACA were reindexed to grow more in line with historical precedent, it would add more than $520 billion to federal deficits over the next decade. It is critical to understand that considerations of practical budgeting behavior accentuate rather than undercut this conclusion. There is no historical precedent for the legislative behavior to which the ACA is compared in producing a favorable fiscal projection. Historically Congress has tended to legislate corrections to Medicare’s payment and revenue schedules whenever insolvency has approached, and to relax its financial corrective efforts when insolvency is more distant. Relative to a continuation of this historical pattern, the ACA has severely worsened the federal fiscal outlook. To correct the law’s fiscal damage, much of the ACA’s subsidized coverage would need to be repealed. Roughly two-thirds of federal subsidies for participation in the health exchanges would need to be eliminated to meet a minimum standard for fiscal improvement. Since the initial publication of this analysis, a number of the ACA’s advocates have raised questions about its fi ndings. A full examination of all of these questions returns one to the conclusion that the ACA has worsened the federal fiscal outlook. There is no significant debate over whether Medicare payments are statutorily limited to the balances of its trust funds, nor is there significant debate over the reality that CBO scorekeeping conventions ignore these limitations. It is thus unambiguous that the ACA has worsened federal fi nances relative to prior law in ways that the existing scorekeeping conventions fail to capture. Since the publication of the original version of this analysis, the CBO has refi ned the language of its published reports to accentuate critical analytical points underlying the conclusions reached in this study.

The Fiscal Consequences of the Affordable Care Act

131

Notes Thanks to Jakina Debnam for her assistance with “The Fiscal Consequences of the Affordable Care Act,” Mercatus Center, April 2012, http://mercatus.org /sites/default/files/publication/The-Fiscal-Consequences-of-the-Affordable -Care-Act_1.pdf, for which some of the research underlying this paper was originally performed. 1. Charles Blahous, “The Fiscal Consequences of the Affordable Care Act,” Mercatus Center, April 2012; Congressional Budget Office, H.R. 4872, Reconciliation Act of 2010 (March 20, 2010), and H.R. 2, Repealing the Job-Killing Health Care Act (February 18, 2011). 2. Congressional Budget Office, The 2012 Long-Term Budget Outlook, 62 (June 2012), http://www.cbo.gov/sites/default/files/cbofiles/attachments/06-05-Long-Term_ Budget_Outlook_2.pdf. 3. Congressional Budget Office, CBO Publications Related to Health Care Legislation (December 2010). 4. Boards of Trustees, Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, 2011 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds (Washington, DC: U.S. Government Printing Office, 2011), 26. 5. Congressional Budget Office, The 2012 Long-Term Budget Outlook, 62. 6. See for example Congressional Budget Office, An Analysis of the President’s 2013 Budget (March 2012), 7– 8, http://www.cbo.gov/sites/default/fi les/cbo fi les/attachments/03-16-APB1.pdf, which scores President Obama’s proposal to index the AMT income thresholds for inflation as adding to the federal deficit. 7. See Congressional Budget Office, Detail on Estimated Budgetary Effects of Title VI of H.R. 8 (revised January 9, 2013), http://www.cbo.gov/sites/default/fi les /cbofi les/attachments/SenateHR8-TitleVI_0.pdf. 8. Charles Blahous, Social Security: The Unfi nished Work (Stanford, CA: Hoover Institution Press), 2010. 9. Indeed, the same could well be said of the ambitious Medicare Hospital Insurance provider payment constraints in the ACA itself. These constraints are examples of the seriousness with which lawmakers have complied with the restraints of trust fund fi nancing. 10. Board of Medicare Trustees, 2010 Annual Report of the Medicare Trustees, 81. 11. David Cutler, “Memorandum to Interested Parties,” March 22, 2012, http://images.politico.com/global/2012/03/memo_on_gop_plans_for_ss_and_ medicare.pdf. Cutler writes that repealing the ACA would mean that “the trust fund would become insolvent in 2016.” This would not be meaningful if it were assumed that Medicare payments were never in jeopardy regardless of the status of its trust funds, as is assumed in the CBO scorekeeping baseline.

132

Charles Blahous

12. Peter Orszag, “Washington Stuck Fighting Wrong Health- Care Battle,” April 17, 2012, http://www.bloomberg.com/news/2012-04-17/washington-stuck -fighting-wrong-health-care-battle.html. 13. Board of Medicare Trustees, 2012 Annual Report of the Medicare Trustees, 2– 3. http://www.cms.gov/Research-Statistics-Data-and-Systems/StatisticsTrends-and-Reports/ReportsTrustFunds/Downloads/TR2012.pdf. 14. Congressional Budget Office, The 2012 Long-Term Budget Outlook. 15. See, for example, Richard Foster’s actuarial opinion attached to the 2010 Annual Report of the Medicare Trustees, http://www.cms.gov/Research-Statistics -Data-and-Systems/Statistics-Trends-and-Reports/ReportsTrustFunds /Down loads/TR2010.pdf: “The fi nancial projections shown in this report for Medicare do not represent a reasonable expectation for actual program operations in either the short range (as a result of the unsustainable reductions in physician payment rates) or the long range (because of the strong likelihood that the statutory reductions in price updates for most categories of Medicare provider services will not be viable).” 16. Internal Revenue Service, “Medical Device Excise Tax: Frequently Asked Questions,” http://www.irs.gov/uac/Medical-Device-Excise-Tax:-Frequently-Asked -Questions. 17. Chris Camire, “Senators, Firms: Repeal Medical Device Tax,” December 23, 2012, http://www.sentinelandenterprise.com/local/ci_22249852/senators -fi rms-repeal-medical-device-tax. 18. Board of Medicare Trustees, 2012 Annual Report. 19. See Congressional Budget Office, The 2012 Long-Term Budget Outlook, 57: “A smaller percentage of people will be eligible for subsidies over time because incomes are projected to increase more quickly than the eligibility thresholds, and federal subsidies will cover a declining share of the premiums over time because of the additional indexing factor described above.” 20. Kaiser Health News, “House Votes to Repeal Medicare’s IPAB,” March 22, 2012, http://www.kaiserhealthnews.org/daily-reports/2012/march/22/ipab-repeal .aspx. 21. Blahous, “ Fiscal Consequences.” 22. Ibid. 23. Peter Orszag, “Health Care Reform and Fiscal Discipline,” Offi ce of Management and Budget Blog, May 29, 2009, accessed March 5, 2012, http://www .whitehouse.gov/omb/blog/09/05/29/HealthCareReformandFiscalDiscipline. 24. See, for example, Josh Barro, “Why You Should Ignore the Medicare Trust Fund,” Forbes.com, April 11, 2012, http://www.forbes.com/sites/josh barro/2012/04/11/why-you-should-ignore-the-medicare-trust-fund/. Len Nichols in a post on the Health Affairs blog, also suggested (incorrectly) that this author’s analysis implicitly holds a unified budget view to be “illegitimate.” “Is

The Fiscal Consequences of the Affordable Care Act

133

Health Reform Fiscally Responsible?” April 20, 2012, http://healthaffairs.org /blog/2012/04/20/is-health-reform-fi scally-responsible/. 25. Committee for a Responsible Federal Budget, “The Affordable Care Act and the HI Trust Fund,” April 10, 2012, http://crfb.org/blogs/affor dable-care-act-and-hi-trust-fund. 26. Congressional Budget Office, The 2012 Long-Term Budget Outlook, 30. 27. Some who have made this leap include Ezra Klein, “The Bizarre Baseline that Obamacare’s Opponents Are Using—in One Graph,” Washington Post, April 11, 2012, http://www.washingtonpost.com/blogs/wonkblog/post/the-bizarre -baseline-that-obamacares-opponents-are-using—in-one-graph/2012/04/11 /gIQAj4cUAT_blog.html and Peter Orszag, “Washington Stuck Fighting Wrong Health- Care Battle,” Bloomberg View, April 17, 2012, http://www.bloomberg.com /news/2012-04-17/washington-stuck-fighting-wrong-health-care-battle.html. 28. Committee for a Responsible Federal Budget, “The Affordable Care Act.” 29. Orszag, “Washington Fighting Wrong Battle.” 30. Paul Krugman, “Another Bogus Attack on Health Care Reform,” New York Times, April 10, 2012, http://krugman.blogs.nytimes.com/2012/04/10/another -bogus-attack-on-health-reform/. 31. Charles Blahous, “Should Congress Change CBO’s Scorekeeping Rules?” e21, May 29, 2012, http://www.economics21.org/commentary/should- congress -change-cbos-scorekeeping-rules. 32. See Foster, Statement of Actuarial Opinion, 2012 Medicare Trustees Report. 33. Nichols, “Is Health Reform Fiscally Responsible?” 34. Blahous, Charles, “Yes, the Health Law Worsens the Deficit,” e21, April 18, 2012, http://www.economics21.org/commentary/yes-health-law-worsens-deficit. 35. Congressional Budget Office, Estimates for the Insurance Coverage Provisions of the Affordable Care Act Updated for the Recent Supreme Court Decision (July 2012), http://www.cbo.gov/sites/default/fi les/cbofi les/attachments/43472-07 -24-2012-CoverageEstimates.pdf. 36. Blahous, Charles, “Yet Another Fiscal Turn for the Worse: Understanding the CBO Re- Score of the 2010 Health Care Law,” e 21, July 25, 2012, http://www .economics21.org/commentary/yet-another-fi scal-turn-worse-understanding -cbo-re-score-2010-health-care-law. 37. Congressional Budget Office, The 2012 Long-Term Budget Outlook, 62. 38. Ibid., 8. 39. Blahous, Charles, “Should Congress Change CBO’s Scorekeeping Rules?” 40. Advisory Board Company, “Where Each State Stands on ACA’s Medicaid Expansion,” June 14, 2013, http://www.advisory.com/Daily-Briefi ng/2012/07/05 /Where-each-state-stands-of-the-Medicaid-expansion.

134

Charles Blahous

41. Dan Crippen, National Governors’ Association, to HHS Secretary Kathleen Sebelius, July 2, 2012, http://www.nga.org/cms/home/federal-relations/nga -letters/executive-committee-letters/col2-content/main-content-list/july-2-2012 -letter—affordable.html. 42. Congressional Budget Office, Estimates Updated. 43. Commonwealth Fund, “Washington Policy Week in Review,” December 17, 2012, http://www.commonwealthfund.org/Newsletters/Washington-Health -Policy-in-Review/2012/Dec/December-17-2012/Sebelius-Tells-Governors-They -Cannot-Partially-Expand-Medicaid.aspx. 44. Blahous, “ Fiscal Consequences.”

Chapter six

Estimating the Impact of the Demand for Consumer-Driven Health Plans Following the 2012 Supreme Court Decision of the Constitutionality of the Patient Protection and Affordable Care Act Stephen T. Parente

Introduction

C

onsumer- directed health plans (CDHPs) are a recent development in the U.S. health insurance market. Their principal economic aim is to make the consumer more engaged in healthcare purchasing decisions through the provision of an account from which medical services can be purchased. This system is tied to traditional indemnity insurance, usually in the form of a high- deductible health plan (HDHP) with a deductible gap of roughly the size of the account—though there are many variations. As of 2012, CDHPs have reached parity in size (18 percent of the market) with HMOs (Kaiser Family Foundation 2011). When the Patient Protection and Affordable Care Act (PPACA) was passed in 2010, CDHPs’ growth was considered uncertain. Nearly two years on, these plans continue to grow. With the implementation of much of PPACA coming in 2014, gauging the impact of the law, as well as potential political and judicial challenges to the legislation that offer viable

Stephen T. Parente

136

alternatives, will be valuable for health policy development. This chapter seeks to determine which of these proposals may have the most impact, in terms of cost per capita, on covering the uninsured. Earlier microsimulations (Feldman et al. 2005 and Parente et al. 2005) estimated the effect of the Medicare Modernization Act of 2003 (MMA) on take-up of HDHPs in the individual health-insurance market. HDHPs feature a large deductible coupled with a health savings account (HSA) owned by the individual that can be used to pay for eligible medical expenses. The MMA made it possible for contributions to the HSA to be made on a tax-preferred basis. That is, contributions less than the size of the deductible are exempt from federal income taxes. If the contribution is made by an employer, it is exempt from Social Security taxes as well. In this paper, policy simulations are completed with three key attributes: 1. Simulations are created to identify the amount of money that individuals and families contribute to their HSAs. Specifically, HSA contributions are allowed to vary by age and income of the policyholder, while constraining the overall average contribution to be the same. The effect of prior health status is also incorporated into health plan choice—a necessary step if one wants to predict enrollment taking risk adjustment into consideration. 2. The simulation approach uses an “iterative” model for determining premiums and health plan choices. Using the results of a baseline choice model, healthcare costs are predicted for the people who enrolled in each plan. Costs are turned into premiums by adding a loading fee and then choices are predicted again with the new premiums. The model is iterated until the predicted premiums do not change from one round to the next. The new outcomes in terms of premiums and enrollment represent a true equilibrium for the simulation model. 3. Simulations are made policy relevant by simulating both the impact of PPACA and policy alternatives following the 2012 Supreme Court decision that states can opt out of the proposed Medicaid expansion. In this step a method for estimating the tax cost of various subsidy proposals is introduced by including an offsetting reduction in tax subsidies for people who drop subsidized employer-sponsored health insurance (ESI) to purchase an individual HDHP policy.

Impact of the Demand for Consumer-Driven Health Plans

137

Simulation Development Approach There are three major components of this methodological approach: (1) model estimation, (2) choice set assignment and prediction, and (3) policy simulation. Often more than two databases were required to complete the task. Integral to this analysis was the use of consumer- directed health plan data from four large employers working with the study investigators. The model estimation had several steps. As a fi rst step, data from three employers offering CDHPs were used to estimate a conditional logistic plan-choice model, similar to earlier work by Parente, Feldman, and Christianson (2004). In the second step the estimated choice-model coefficients are used to predict health plan choices for individuals in the Medical Expenditure Panel Survey– Household Component (MEPS-HC). In order to complete this step, it was necessary fi rst to assign the number and types of health insurance choices that are available to each respondent in the MEPS-HC. For this purpose the smaller but more detailed MEPS Household Component–Insurance Component linked fi le, which contained the needed information, is used. The third step was to generate HSA premiums and benefit designs. The fi nal step was to apply planchoice model coefficients to the MEPS data with premium information and 2010 PPACA statute to get fi nal estimates of insurance take-up and subsidy costs. Each of these three steps is described below. Estimate Plan Choice Regression Plan- choice data were pooled from the three employers offering CDHPs to specify a conditional logistic regression model. Conceptually, a choice model based on utility maximization is used wherein utility is considered to be a function of personal attributes such as health status, health plan attributes such as the out- of-pocket premium, and the interaction of premium and health status, formally stated as: Uij = f(Zj,Yi,Xij) where i is the decision-making employee choosing among: •

j = health plan choices



Yi = employee personal attributes

Stephen T. Parente

138 •

Zj = health plan attributes and



Xij = interactions between alternative-specific constants and personal attributes.

A very important constraint in this modeling was that any plan attribute used in the model from the employer data also had to be available in the MEPS data to permit a simulation. As a result, the key variables used in the plan choice model were: •

SCALEDPREM

After-tax premium paid by the employee



CLB

Amount of money in the employee’s health reimburse-



CUB

ment account (HRA), if any Difference between the employee’s plan deductible and the HRA •

COIN

Coinsurance rate



CHRONIC

Employee or dependent’s chronic illness status: 1 if em-



AGE

Employee’s age (in years)



FEM

Employee’s gender (1 if female, 0 if male)



FAM

Employee’s contract status: 1 if employee has a two-



INC

Employee’s annual wage income.

ployee or dependent has a chronic illness, else 0

person or family contract, else 0

Also included in the regression were alternative-specific constants (intercepts) for each of the possible health plan choices. These intercepts are used to capture plan-specific features not represented by other identifiers of plan design. They are also included as interaction terms with age, gender, family status, and income. The intercept terms include: •

PPO_L

PPO (preferred provider organization) Low (e.g., restrictive network, high copay, 15 percent coinsurance)



PPO_M

PPO Medium (e.g., better network, lower copay and



PPO_H

PPO High (e.g., open network, lowest copay, no coinsurance)



HRA

Health Reimbursement Account CDHP



HSA_E

Employer-sponsored HSA, modeled on higher-premium-



HSA_S

coinsurance)

cost HRA Employee-paid HSA, no employer contribution, modeled on lower-premium- cost HRA •

HMO

Health Maintenance Organization

Impact of the Demand for Consumer-Driven Health Plans

139

Choice Set Assignment and Prediction Several data sources were used to develop two sets of plan-choice predictions for the simulation: one set of data for workers with insurance offers and a second set for individuals who do not have employer offers of coverage. This second set includes both uninsured individuals and those who take up nongroup policies. One group of individuals excluded from the simulation is nonoffered individuals who reported having employer group coverage through another household member. The analytic steps taken to develop the individuals’ choice sets for the simulations are outlined below. First, the original four choices predicted earlier, including three PPOs and an HMO, are used. Since a worker was assigned between one and four plans, assumptions were made for each plan design: •

Four choices: Low PPO, Medium PPO, High PPO, HMO



Three choices: Low PPO, High PPO, HMO



Two choices: Medium PPO, HMO



One choice: Medium PPO

Here “low,” “medium,” and “high” refer to the cost and quality of the plans (e.g., “low” implies low cost and lower quality). To these choices four additional options were added: •

Self-fi nanced (full cost) HSA—additional choice for all workers



No health coverage—additional choice for all workers to turn down coverage



Employer-sponsored HSA—available to all workers in establishments with



Employer-sponsored HRA—available to all workers in establishments with

more than 500 employees, not available to other workers more than 500 employees, not available to other workers

Individuals who did not have health insurance offered to them at work or who were not employed faced five health plan choices regardless of income, age, or gender: •

High PPO



Medium PPO



Low PPO



Self-fi nanced HSA



No insurance

140

Stephen T. Parente

With this total set of possible choices for workers with insurance offers and individuals without insurance offers, plan- choice regression results were used to predict plan- choice probabilities for each MEPS-HC sample respondent.1 However, before the predicted probabilities could be used, some specific assumptions about benefit plan design and premiums for individual plans were made. To get premium estimates, MEPSlinked insurance data were used to develop a hedonic price model to predict premiums for individual plans. The same hedonic plan regressions described above were used. An exception was for individuals without offers of coverage. In that case, the premium model for the smallestestablishment-size category was used. The rationale for this assumption was that this choice most closely represents an individual policy in terms of the loading charge for plan administrative costs. The plan characteristics used to defi ne the three PPOs (low, medium, and high) came from the eHealthinsurance.com survey of plans purchased in the individual market. Roughly speaking, the twenty-fi fth, fi ftieth, and seventy-fi fth percentiles of coinsurance and deductibles for assigning the plan characteristics are used. The likelihood that premiums in the individual market vary a lot by a person’s age was also recognized. The MEPS survey included a table of average premiums by age cohort. Originally, an index using the information on this table was created. The index was set equal to the age group corresponding to the median age of adults in this sample (thirty-five to thirty-nine). Older individuals, who had higher premiums, had index values that were greater than 1.0. Younger individuals, who had lower premiums, had index values less than 1.0. The index values ranged from .59 to 2.18 for single- coverage policies and .453 to 1.65 for family- coverage policies. Age, gender, family contract, and chronic illness were taken into account to predict premiums using the health plans claims data. Finally, all premiums were adjusted to current levels. One significant issue with the simulation is its inability to predict whether or not an individual would take up insurance in the employeroffered market or be uninsured in the individual market. This limitation occurs because the CDHP employer data only include information on offered workers who held coverage. To address this issue, the model needed to be calibrated to accurately reflect both the actual percentage of people who turn down employer offers and the actual percentage of people in the individual market who are uninsured. To obtain more accurate estimates, these calibrations were

Impact of the Demand for Consumer-Driven Health Plans

141

completed by four quartiles of income and then compared to previously published national nontake-up and uninsurance rates. National population weights were also applied to the calibrated model to represent the entire adult population, excluding full-time students, those with public insurance, and individuals with employer-based coverage through another household member. This fairly tedious process was performed for each reestimation or modification of the conditional logistic regression. Policy Simulation Two of the most substantial advances in this approach to microsimulation were inclusion of chronic illness in the plan choice model and generation of premiums through an iterative process using prior years’ claims data to create actuarially fair estimates of premium. These advances are described in greater detail below. In the simulations, health status was used as a variable that affects plan choice. This is important for two reasons. First, health status may be an important factor in predicting plan choice, so the addition of this variable should improve the fit of the choice model, other things being equal. Second, it may be that case that sick people prefer certain plans, which would drive the premiums up, and conversely for healthy people. Specifically, if sick people are attracted to traditional plans, it could lead to a death spiral of increasing premiums and falling enrollment for those plans. One of the goals of the new simulations is to determine whether the addition of an HDHP as a choice will tend to destabilize the market for health insurance. To account for health status, claims data were used for contract holders (employees) of the employers examined during prior plan choice analysis. For this analysis, claims data were obtained for the year prior to the contract holders’ possible enrollment in an HDHP. Diagnosis code information from these prior-year claims records was used to calculate a set of thirty-four adjusted diagnosis groups using a methodology developed by Johns Hopkins University researchers (Weiner et al. 1991). Several of these thirty-four groups identify a diagnosis indicating the presence of a chronic condition. With this information a dummy variable was constructed indicating the presence of chronic illness. Construction of this variable permitted us to develop a medical- care cost-regression model to predict future medical expenditure by the MEPS population enrolled in each plan type. Using an aggregate measure of health status

142

Stephen T. Parente

represented as a binary variable allowed the creation of a variable to be mapped from the MEPS database in order to predict health-plan choice. Besides the inclusion of health status, premium and cost-sharing information interacted with more demographic variables. These interactions were introduced to account for possible associations not accounted for by the own-price response to premium and cost-sharing variables. To capture the relationship between costs and health risk, a healthcare cost model was estimated for the individuals who chose each plan. The cost model was used to develop premium estimates that fed back into the choice model. The microsimulation was “iterated”—that is, applied repeatedly—between the choice model and the cost model until the market converged to a stable set of choices and premiums. This method is illustrated in Figure 6.1. Starting from a premium that is too high for equilibrium (point A), the premium falls and enrollment increases until the two lines converge on a single premium and enrollment (point B). There is no guarantee that the model will be stable, as shown here. The “choice depends on premium” line (i.e., the demand curve) slopes down, but the “premium depends on choice” line might slope up or down. The model will be stable if the demand curve is the steeper of the two lines. To implement this new iterative approach, premiums were constructed from expected healthcare costs in the individual and ESI markets. Premiums obviously depend on expected costs, but they also depend on how costs are aggregated across individuals. How many individuals are in the insurance pool? Does the premium for a particular person depend on his or her experience, or on the experience of the group? In other words, how are premiums “rated” in the individual and ESI markets? The two rating methods used were individual experience rating (IER) for the individual health insurance market and group experience rating (GER) for the ESI market. 2 The premium for each rating method was generated as follows: Individual Market: Given that the premium for each person in the individual market is based on that person’s own healthcare costs, how much each person would spend under each type of insurance plan (PPO, HSA, etc.) was estimated. A loading fee of 20 percent was used to arrive at the premium for each choice in the individual market. 3 Employer- sponsored Market: The fi rst step in GER is to defi ne the pool that determines the premium rates. Three pools based on estab-

Impact of the Demand for Consumer-Driven Health Plans

143

Figure 6.1 Model of Health Plan Demand and Cost.

lishment size—small establishments, midsize establishments, and large establishments were used.4 The costs of each person in each plan were predicted. Then the average cost across all people who work for employers in each of the three pools was calculated. For example, the average cost of the HMO for employees of small establishments may be $7,000 for a single policy and $15,000 for a family policy. The average cost of the HMO was different in midsize and large groups. Then, loading fees were added to get predicted premiums for each pool in the ESI market.

Patient Protection and Affordable Care Act Simulation Results The simulation methods described were used to estimate the impact of the Patient Protection and Affordable Care Act as well as variants of the law following the June, 2012 Supreme Court decision to allow states to opt out of Medicaid expansion. The two primary elements modeled from PPACA were the expansion of the private insurance market through state and federal insurance exchanges and the expansion of the Medicaid program as outlined in the law for those with incomes less than 133 percent of the Federal Poverty Level (FPL). The data used to complete

Stephen T. Parente

144

the state- specific estimates were an MEP- derived database of synthetic state assignment for survey participants, described in detail in (Parente et al. 2011). PPACA in 2012 and 2014 The results of these simulations are presented in Tables 6.1 through 6.3. In Table 6.1 estimates of 2012 and 2014 demand for insurance across the entire market of those less than sixty-five years old and affected by PPACA are estimated. In 2012, the number of uninsured was estimated to be 54.8 million 5. Due to PPACA this number is predicted to drop 38 percent to 34 million. The majority of the change will occur from an expansion of the private insurance market, with over 21 million newly covered enrollees. Over 17 million will gain Medicaid coverage due to an expansion of the program in all fi fty states. The iteration results at baseline focus on the group market. At baseline, the iterations did not change the results of the individual-market healthplan choices. With respect to the individual market, there will be 27 percent growth in bronze or catastrophic programs and a 55 percent increase in enrollment in silver or more generous plan designs. HSA/HDHP de-

Table 6.1. PPACA Impact on Insurance Demand from 2012 to 2014 Individual Market Silver or Higher Bronze or Catastrophic Medicaid Uninsured Group Market HMO HRA HSA-Funded ESI 2 Self-Pay Low PPO PPO High PPO Low PPO Medium ESI to Other Insurance/ Exchange Employee refuses coverage Total Uninsured Total Private Insurance Total Medicaid

2012

2014

Difference

% Change

13,077,268 19,539,213 34,855,438 48,854,416

20,285,299 24,721,721 52,012,641 28,542,663

7,208,030 5,182,508 17,157,203 (20,311,753)

55% 27% 49% −42%

2012

2014

Difference

% Change

5,896,887 18,060,845 2,045,882 102,199 22,395,612 2,821,926 81,931,286 19,584,084

5,870,908 17,665,282 3,909,279 3,537,014 18,486,371 3,391,734 69,754,544 30,358,810

(25,979) (395,563) 1,863,397 3,434,815 (3,909,241) 569,808 (12,176,743) 10,774,725

0% −2% 91% >400% −17% 20% −15% 55%

5,916,908 54,771,324 205,669,993 34,855,438

5,507,608 34,050,271 226,904,491 52,012,641

(409,300) (20,721,053) 21,234,498 17,157,203

−7% −38% 10% 49%

Impact of the Demand for Consumer-Driven Health Plans

145

signs that qualify as bronze plans may have too high a premium to maintain the growth they had prior to 2012. However, the design appears likely to continue to grow from these results. For the group market, the iterations lead to substantial migration out of the “turned down,” “PPO high,” and “PPO medium” choices with losses of 7 percent, 17 percent and 15 percent respectively. The plan designs forecast to have the largest in-migrations are HSAs in which the consumer self-funds the account, PPO medium, and an opt- out lowoption PPO in which the consumer takes a voucher and buys insurance on a qualified exchange, and the employer pays a penalty. There is also a large share of individuals (10.8 million) who fi nd new coverage as a consequence of a switch in spousal coverage or are dropped from an employers plan and directed towards a health insurance exchange. Ten-Year Estimates of Health Reform Following the 2012 PPACA Supreme Court Decision The impact of PPACA over a ten-year period from 2012 to 2021 was estimated as a baseline simulation. To test how this microsimulation could forecast policy changes, the effect of two policy changes based on the recent PPACA Supreme Court decision were examined. On June 28, 2012, the Supreme Court ruled that the PPACA could not force states to take a proposed Medicaid expansion. This decision could greatly impact both the insurance coverage and the cost of the law. Shortly after the decision, it was reported that fifteen states would not take or would be very unlikely to take the Medicaid expansion in 2014. Below, estimates of the coverage and federal cost impact of the PPACA, compared with the impact of the Supreme Court ruling on coverage and federal cost for Medicaid and insurance exchanges from 2012 to 2021, are generated. To complete this simulation, the following assumptions were made: •

Individuals who do not take Medicaid and are eligible for the exchange subsidies starting at 100 percent of the FPL will consider taking them, even with a minimal premium required for enrollment.



States that do not offer an exchange will have a federal exchange available to their citizens and the federal exchange will be able to route subsidies to citizens in these states.



States offering Medicaid will use the same health plan fi nancing arrangements present today through 2021.

Stephen T. Parente

146 •

Insurance exchange premiums will grow at a higher rate than Medicaid program costs.



Prior to the Supreme Court ruling, states would have autoenrolled anyone below 134 percent of the FPL in federally fi nanced Medicaid expansion plans and not allowed anyone below 134 percent of the FPL to enroll in a state or federal exchange for private insurance coverage.

Table 6.2A presents the coverage impact of the PPACA before and after the Supreme Court decision; Table 6.2B reports the federal cost impact. As of January 1, 2013, the governors of six states (Florida, Louisiana, Mississippi, Nebraska, South Carolina, and Texas), all with Republican governors, have openly declared that they do not plan to expand their Medicaid programs. It is estimated that the number of uninsured will increase by 7.9 million by 2021 over the number before the Supreme Court decision. Several million childless adults who are not disabled or aged earn sufficient income to qualify for an exchange policy in the six states choosing not to expand Medicaid. There will be a drop of 4.4 million covered lives under Medicaid in 2014 from the number before the Supreme Court ruling. There will be a 3.2 million increase in the number of people taking up private insurance coverage in 2014. The cost of a private insurance crowd- out of Medicaid expansion will be an additional $7.5 billion in 2014 and a total of $75.8 billion from 2014 to 2021. If citizens in the states refusing to expand their Medicaid programs can accept federal subsidies for private health insurance, the impact will be greater federal cost and greater loss of coverage than the baseline PPACA estimate before the Supreme Court decision. An alternative scenario, a comparison of pre– Supreme Court coverage and cost effects, was contrasted with one where only states with Medicaid benefits already at 100 percent of the FPL or above would take the expansion. The rationale for this policy alternative is that each of these states is already investing state revenues to support an expanded Medicaid program. As a result, PPACA offers these states fi nancial relief from self-funding their expanded programs. These states include the prototype state for PPACA, Massachusetts, with 133 percent FPL qualification, and Minnesota, the state with the highest FPL threshold 233 percent. The coverage and cost results of this simulation are presented in Tables 6.3A and 6.3B, respectively. Regarding coverage, there are 11.8 million fewer people in Medicaid under the PPACA in 2014. However, there are 7.5 million more people insured through private insurance. This

2012 32.6 31.7 48.9 2012 152.6 3.2 5.9 0.1 2012 185.4 34.9 54.8 275.0

INDIVIDUAL COVERAGE

Insured Medicaid Uninsured

GROUP COVERAGE

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL COVERAGE

Insured Medicaid Uninsured TOTAL

Pre– SCOTUS Ruling

183.7 35.9 58.1 277.7

2013

152.4 3.8 7.2 0.1

2013

31.3 32.1 50.9

2013

194.4 52.0 34.1 280.5

2014

144.9 10.1 5.5 4.5

2014

45.0 41.9 28.5

2014

196.2 54.1 33.0 283.3

2015

137.7 11.5 5.9 11.7

2015

46.9 42.6 27.1

2015

196.3 56.2 33.6 286.1

2016

131.8 12.7 6.3 17.6

2016

47.0 43.5 27.3

2016

194.7 58.7 35.6 289.0

2017

132.2 13.8 6.8 17.3

2017

45.3 44.9 28.8

2017

193.6 60.9 37.5 291.9

2018

132.8 14.7 7.2 17.0

2018

43.7 46.1 30.3

2018

Scenario: Six states decline to take expansion: FL, LA, MS, NE, SC, TX

Table 6.2A. Pre- and Post- SCOTUS PPACA Coverage Impact

192.6 63.0 39.3 294.8

2019

133.8 15.6 7.4 16.8

2019

42.0 47.4 31.9

2019

191.9 66.8 42.0 300.7

2021

136.8 16.9 7.1 16.2

2021

38.9 50.0 34.9

2021

(continued)

192.1 65.0 40.7 297.8

2020

135.1 16.2 7.3 16.5

2020

40.4 48.7 33.4

2020

2012 32.6 31.7 48.9 2012 152.6 3.2 5.9 0.1 2012 185.4 34.9 54.8 275.0 2012 0.0 0.0 0.0

INDIVIDUAL COVERAGE

Insured Medicaid Uninsured

GROUP COVERAGE

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL COVERAGE

Insured Medicaid Uninsured TOTAL

Net Impact

Insured Medicaid Uninsured

Post– SCOTUS Ruling

Table 6.2A. (continued)

0.0 0.0 0.0

2013

183.7 35.9 58.1 277.7

2013

152.4 3.8 7.2 0.1

2013

31.3 32.1 50.9

2013

3.2 −4.4 1.1

2014

197.7 47.7 35.2 280.5

2014

145.8 8.7 5.6 4.9

2014

46.9 38.9 29.6

2014

2.7 −4.6 1.9

2015

198.9 49.5 34.9 283.3

2015

138.6 9.9 6.0 12.2

2015

48.1 39.6 28.9

2015

1.9 −4.7 2.8

2016

198.2 51.5 36.4 286.1

2016

132.8 11.0 6.5 18.2

2016

47.3 40.5 29.9

2016

1.0 −4.9 3.9

2017

195.8 53.7 39.5 289.0

2017

133.2 12.0 7.0 17.8

2017

44.7 41.7 32.6

2017

0.2 − 5.2 5.0

2018

193.8 55.7 42.5 291.9

2018

133.9 12.9 7.3 17.6

2018

42.2 42.8 35.1

2018

2020 − 2.2 − 5.0 − 7.2

−1.0 −5.1 − 6.1

189.9 60.0 47.9 297.8

2020

136.4 14.4 7.5 17.0

2020

36.5 45.6 40.4

2020

2019

191.5 57.9 45.4 294.8

2019

135.0 13.7 7.5 17.2

2019

39.3 44.2 37.9

2019

− 3.2 −4.8 − 7.9

2021

188.7 62.1 50.0 300.7

2021

138.1 15.0 7.2 16.7

2021

33.9 47.1 42.7

2021

2012

$0.0 $105.4 —

2012

$0.0 $10.7 $0.0 $0.0

2012

$0.0 $116.1 $0.0 $116.1

INDIVIDUAL FEDERAL $$

Insured Medicaid Uninsured

GROUP FEDERAL $$

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL FEDERAL $$

Insured Medicaid Uninsured TOTAL

Pre– SCOTUS Ruling

$0.0 $122.0 $0.0 $122.0

2013

$0.0 $13.1 $0.0 $0.0

2013

$0.0 $108.9 —

2013

$246.6 $179.4 $0.0 $426.0

2014

$0.0 $35.6 $0.0 $26.9

2014

$219.7 $143.8 —

2014

$244.2 $190.5 $0.0 $434.7

2015

$0.0 $41.1 $0.0 $25.4

2015

$218.9 $149.4 —

2015

$245.3 $201.9 $0.0 $447.1

2016

$0.0 $46.3 $0.0 $25.2

2016

$220.1 $155.6 —

2016

$242.7 $214.5 $0.0 $457.3

2017

$0.0 $51.3 $0.0 $25.2

2017

$217.5 $163.3 —

2017

$237.3 $226.9 $0.0 $464.2

2018

$0.0 $55.8 $0.0 $25.5

2018

$211.8 $171.1 —

2018

Scenario: Six states decline to take expansion: FL, LA, MS, NE, SC, TX

Table 6.2B. Pre- and Post- SCOTUS PPACA Cost Impact (billions of US dollars)

$228.5 $239.4 $0.0 $467.9

2019

$0.0 $60.1 $0.0 $25.4

2019

$203.0 $179.3 —

2019

$220.6 $251.9 $0.0 $472.5

2020

$0.0 $64.0 $0.0 $26.0

2020

$194.6 $187.9 —

2020

(continued)

$213.1 $264.4 $0.0 $477.5

2021

$0.0 $67.9 $0.0 $26.1

2021

$187.0 $196.6 —

2021

2012

$0.0 $105.4 $0.0

2012

$0.0 $10.7 $0.0 $0.0

2012

$0.0 $116.1 $0.0 $116.1

2012 $0.0 $0.0 $0.0 $0.0

INDIVIDUAL FEDERAL $$

Insured Medicaid Uninsured

GROUP FEDERAL $$

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL FEDERAL $$

Insured Medicaid Uninsured TOTAL

Net Impact

Insured Medicaid Uninsured TOTAL

Post– SCOTUS Ruling

Table 6.2B. (continued)

$0.0 $0.0 $0.0 $0.0

2013

$0.0 $122.0 $0.0 $122.0

2013

$0.0 $13.1 $0.0 $0.0

2013

$0.0 $108.9 $0.0

2013

$20.1 −$12.5 $0.0 $7.5

2014

$266.7 $166.9 $0.0 $433.6

2014

$0.0 $31.0 $0.0 $30.6

2014

$236.0 $135.9 $0.0

2014

$22.4 −$13.6 $0.0 $8.8

2015

$266.6 $176.8 $0.0 $443.5

2015

$0.0 $35.7 $0.0 $30.6

2015

$236.0 $141.1 $0.0

2015

$23.2 −$14.3 $0.0 $8.9

2016

$268.5 $187.6 $0.0 $456.0

2016

$0.0 $40.4 $0.0 $31.2

2016

$237.3 $147.2 $0.0

2016

$25.8 −$15.2 $0.0 $10.6

2017

$268.5 $199.4 $0.0 $467.9

2017

$0.0 $45.1 $0.0 $31.1

2017

$237.4 $154.3 $0.0

2017

$31.4 −$16.3 $0.0 $15.1

2018

$268.7 $210.6 $0.0 $479.4

2018

$0.0 $49.3 $0.0 $31.0

2018

$237.7 $161.3 $0.0

2018

$29.3 −$16.3 $0.0 $13.0

2019

$257.8 $223.1 $0.0 $480.9

2019

$0.0 $53.5 $0.0 $30.2

2019

$227.6 $169.7 $0.0

2019

$24.0 −$16.0 $0.0 $8.0

2020

$244.6 $235.9 $0.0 $480.5

2020

$0.0 $57.2 $0.0 $30.2

2020

$214.4 $178.7 $0.0

2020

$19.2 −$15.4 $0.0 $3.8

2021

$232.3 $249.1 $0.0 $481.3

2021

$0.0 $60.8 $0.0 $30.1

2021

$202.1 $188.3 $0.0

2021

$75.8

2012 32.6 31.7 48.9 2012 152.6 3.2 5.9 0.1 2012 185.4 34.9 54.8 275.0

INDIVIDUAL COVERAGE

Insured Medicaid Uninsured

GROUP COVERAGE

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL COVERAGE

Insured Medicaid Uninsured TOTAL

Pre– SCOTUS Ruling

183.7 35.9 58.1 277.7

2013

152.4 3.8 7.2 0.1

2013

31.3 32.1 50.9

2013

194.4 52.0 34.1 280.5

2014

144.9 10.1 5.5 4.5

2014

45.0 41.9 28.5

2014

196.2 54.1 33.0 283.3

2015

137.7 11.5 5.9 11.7

2015

46.9 42.6 27.1

2015

196.3 56.2 33.6 286.1

2016

131.8 12.7 6.3 17.6

2016

47.0 43.5 27.3

2016

194.7 58.7 35.6 289.0

2017

132.2 13.8 6.8 17.3

2017

45.3 44.9 28.8

2017

193.6 60.9 37.5 291.9

2018

132.8 14.7 7.2 17.0

2018

43.7 46.1 30.3

2018

Scenario: Only states with Medicaid FPL ≥ 100% take the expansion

Table 6.3A. Pre- and Post- SCOTUS PPACA Coverage Impact (millions of US dollars)

192.6 63.0 39.3 294.8

2019

133.8 15.6 7.4 16.8

2019

42.0 47.4 31.9

2019

192.1 65.0 40.7 297.8

2020

135.1 16.2 7.3 16.5

2020

40.4 48.7 33.4

2020

(continued)

191.9 66.8 42.0 300.7

2021

136.8 16.9 7.1 16.2

2021

38.9 50.0 34.9

2021

32.6 31.7 48.9 2012 152.6 3.2 5.9 0.1 2012 185.4 34.9 54.8 275.0 2012 0.0 0.0 0.0

Insured Medicaid Uninsured

GROUP COVERAGE

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL COVERAGE

Insured Medicaid Uninsured TOTAL

Net Impact

Insured Medicaid Uninsured

0.0 0.0 0.0

2013

183.7 35.9 58.1 277.7

2013

152.4 3.8 7.2 0.1

2013

31.3 32.1 50.9

2013

7.5 −11.8 4.3

2014

202.0 40.2 38.4 280.5

2014

147.2 6.8 5.8 5.2

2014

49.5 33.4 32.6

2014

7.6 −12.8 5.1

2015

203.8 41.3 38.1 283.3

2015

140.8 6.9 6.3 12.6

2015

50.4 34.4 31.8

2015

7.5 -13.6 6.1

2016

203.8 42.6 39.7 286.1

2016

135.8 6.9 6.9 18.8

2016

49.2 35.8 32.8

2016

States taking expansion: AZ, CA, CO, CT, DE, DC ,HI, IL, ME, MD, MA, MN, NJ, NY, RI, TN, VT, WI.

2012

INDIVIDUAL COVERAGE

Post– SCOTUS Ruling

Table 6.3A. (continued)

7.1 −14.4 7.3

2017

201.9 44.2 42.9 289.0

2017

137.1 6.8 7.5 18.6

2017

46.2 37.4 35.4

2017

6.7 −15.1 8.4

2018

200.3 45.7 45.9 291.9

2018

138.6 6.7 8.0 18.5

2018

43.2 39.0 37.9

2018

5.7 −15.3 9.6

2019

198.2 47.7 48.9 294.8

2019

140.3 6.7 8.3 18.2

2019

39.7 41.0 40.6

2019

4.5 −15.2 10.6

2020

196.6 49.8 51.3 297.8

2020

142.3 6.7 8.3 18.0

2020

36.4 43.1 43.1

2020

3.7 −15.1 11.4

2021

195.6 51.8 53.4 300.7

2021

144.5 6.7 8.0 17.7

2021

33.3 45.1 45.4

2021

2012

$0.0 $105.4 $0.0

2012

$0.0 $10.7 $0.0 $0.0

2012

$0.0 $116.1 $0.0 $116.1

INDIVIDUAL FEDERAL $$

Insured Medicaid Uninsured

GROUP FEDERAL $$

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL FEDERAL $$

Insured Medicaid Uninsured TOTAL

Pre- SCOTUS Ruling

$0.0 $122.0 $0.0 $122.0

2013

$0.0 $13.1 $0.0 $0.0

2013

$0.0 $108.9 $0.0

2013

$246.6 $179.4 $0.0 $426.0

2014

$0.0 $35.6 $0.0 $26.9

2014

$219.7 $143.8 $0.0

2014

$244.2 $190.5 $0.0 $434.7

2015

$0.0 $41.1 $0.0 $25.4

2015

$218.9 $149.4 $0.0

2015

$245.3 $201.9 $0.0 $447.1

2016

$0.0 $46.3 $0.0 $25.2

2016

$220.1 $155.6 $0.0

2016

$242.7 $214.5 $0.0 $457.3

2017

$0.0 $51.3 $0.0 $25.2

2017

$217.5 $163.3 $0.0

2017

$237.3 $226.9 $0.0 $464.2

2018

$0.0 $55.8 $0.0 $25.5

2018

$211.8 $171.1 $0.0

2018

Scenario: Only states with Medicaid FPL ≥ 100% take the expansion

Table 6.3B. Pre- and Post- SCOTUS PPACA Cost Impact (billions of US dollars)

$228.5 $239.4 $0.0 $467.9

2019

$0.0 $60.1 $0.0 $25.4

2019

$203.0 $179.3 $0.0

2019

$220.6 $251.9 $0.0 $472.5

2020

$0.0 $64.0 $0.0 $26.0

2020

$194.6 $187.9 $0.0

2020

(continued)

$213.1 $264.4 $0.0 $477.5

2021

$0.0 $67.9 $0.0 $26.1

2021

$187.0 $196.6 $0.0

2021

$0.0 $105.4 $0.0

2012

$0.0 $10.7 $0.0 $0.0

2012

$0.0 $116.1 $0.0 $116.1

2012

$0.0 $0.0 $0.0 $0.0

Insured Medicaid Uninsured

GROUP FEDERAL $$

Insured Take Medicaid Coverage Refuse Coverage Take Individual Coverage

TOTAL FEDERAL $$

Insured Medicaid Uninsured TOTAL

Net Impact

Insured Medicaid Uninsured TOTAL

$0.0 $0.0 $0.0 $0.0

2013

$0.0 $122.0 $0.0 $122.0

2013

$0.0 $13.1 $0.0 $0.0

2013

$0.0 $108.9 $0.0

2013

$34.4 −$42.0 $0.0 −$7.6

2014

$281.0 $137.4 $0.0 $418.5

2014

$0.0 $24.0 $0.0 $33.3

2014

$247.7 $113.4 $0.0

2014

$35.0 −$46.3 $0.0 −$11.3

2015

$279.2 $144.2 $0.0 $423.4

2015

$0.0 $24.6 $0.0 $34.5

2015

$244.7 $119.6 $0.0

2015

$33.4 −$50.3 $0.0 −$16.8

2016

$278.7 $151.6 $0.0 $430.3

2016

$0.0 $24.9 $0.0 $36.5

2016

$242.2 $126.7 $0.0

2016

2017

$32.5 −$54.4 $0.0 −$21.9

2017

$275.2 $160.2 $0.0 $435.4

2017

$0.0 $25.1 $0.0 $37.8

2017

$237.5 $135.1 $0.0

States taking expansion: AZ, CA, CO, CT, DE, DC, HI, IL, ME, MD, MA, MN, NJ, NY, RI, TN, VT, WI.

2012

INDIVIDUAL FEDERAL $$

Post- SCOTUS Ruling

Table 6.3B. (continued)

$33.5 −$58.0 $0.0 −$24.6

2018

$270.8 $168.9 $0.0 $439.7

2018

$0.0 $25.2 $0.0 $38.9

2018

$231.9 $143.7 $0.0

2018

$24.8 −$59.9 $0.0 −$35.1

2019

$253.3 $179.5 $0.0 $432.9

2019

$0.0 $25.5 $0.0 $38.9

2019

$214.4 $154.1 $0.0

2019

$13.7 −$60.8 $0.0 −$47.1

2020

$234.4 $191.1 $0.0 $425.4

2020

$0.0 $25.9 $0.0 $39.8

2020

$194.6 $165.2 $0.0

2020

$3.9 −$61.5 $0.0 −$57.7

2021

$216.9 $202.9 $0.0 $419.9

2021

$0.0 $26.6 $0.0 $40.3

2021

$176.6 $176.3 $0.0

2021

−$221.9

Impact of the Demand for Consumer-Driven Health Plans

155

still leaves a net reduction in uninsured in 2014, but less potential Medicaid crowd- out of the private market. By 2021, the total net difference in uninsured has grown to 11.4 million as the private insurance market take-up erodes compared to 2014. With respect to cost, we estimate that the net federal cost will be $7.6 billion less in 2014 due to less Medicaid expansion and $57.7 billion less by 2021 under the PPACA. The sum of 2014 to 2021 net federal cost savings will be $221.9 billion. Under this scenario states that already made a commitment to their populations to expand Medicaid coverage would get federal fi nancing relief and states that chose not to expand to 100 percent would get fewer federal dollars for Medicaid and larger numbers of uninsured.

Caveats While an improved model was developed for this analysis, several caveats are critical. The fi rst is lack of observed uninsured or Medicaid takeup in plan- choice estimation. Thus intercept terms have to be added in the prediction equation to calibrate the level of uninsured to match that reported in the market for both the individual and ESI populations. The second caveat is that both the HDHP and MEPS data are several years old and need to be inflation adjusted for this analysis. However, there is some confidence in making these adjustments, because the plan designs in this analysis are largely the same as in previous analyses when the plan choices were observed, and the model’s premium estimates are based on claim expenditures with a medical- care inflation rate applied. The 2009 MEPS data used are the oldest component of the analysis, but the linked insurance component and household interview survey data have not been made available beyond 2001. The third caveat is that the estimated individuals enrolled in plans from the simulations are actually summed probabilities of a person’s enrollment in a plan. For example, 100 actual people are predicted to join a HDHP. Instead, it is predicted that 1,000 people have (on average) a 10 percent probability of joining an HDHP (which sums to 100). This is what a plan choice model enables one to do. It also provides a platform for changes in policy to be predicted, but not to the point of saying that a person will absolutely choose a particular health plan. A fourth caveat is the lack of actual HDHP plan choices with a Health Savings Account. Instead, the results of a low option Health

156

Stephen T. Parente

Reimbursement Account (HRA) design are used that later became the standard benefit design template for an HSA. Recent data on HSA choices within employers suggest that this approach still remains valid. The fifth caveat is that adjusted changes were made to the HDHP benefit design. The adjustment made was to increase the coinsurance rate of the HDHP from 5 percent to 15 percent. This is also consistent with more recent plan designs for HDHPs in 2010–12 than early HDHP designs in 2004–5. The fi nal caveat is that greater price elasticity was found from both premium and cost sharing responses than before. Once again this change may have merit appears to have merit. The most recent plan choice model produced results suggesting a greater elasticity response from cost sharing than found previously.6 As improvement a nested logit model to provide more accurate premium elasticity estimates to verify the increase in elastic response is being developed but not yet working well enough for simulation purposes. This change, combined with the cost regressions predicting higher premiums for low option PPOs and lower premiums for low option HDHPs, led to greater take-up of HDHPs than before.

Summary This simulation model provides a policy analysis tool to gauge the national impact of federal health reform. The application of the simulation was found to predict significant reductions in levels of uninsured following the full implementation of PPACA in 2014. This approach employed novel characteristics that reflect an evolving insurance market with greater demand for HDHPs and HSAs as well as taking health status into consideration as part of health insurance demand. When considering the impact of the recent Supreme Court decision, this simulation model predicted a nontrivial impact on federal cost and the level of uninsured, depending on whether a state decides to accept or decline Medicaid expansion. Generally, states’ refusing coverage will lead to more uninsured and lower federal cost. The one exception is the six-state model in which a small population of uninsured entering private exchanges, as opposed to a Medicaid expansion program, results in greater federal cost and more uninsured. However, when considering the assumption that all states providing more than 100 percent of the FPL

Impact of the Demand for Consumer-Driven Health Plans

157

for Medicaid do not accept expansion, thus defraying some of their own outlays, the result of the PPACA is a sizable reduction in the number of covered lives as well as a substantially lower federal cost. As alternatives and refi nements to the PPACA are proposed, either as Republican “Repeal and Replace” options or further legislative refi nements, the simulation model outlined above can inform policy to assess the cost per capita of covering the uninsured, as well as the federal cost, resulting from these proposals. At best, full implementation of the PPACA would reduce the number of uninsured by more than 20 million. If achieved in 2014, it will be the largest person-level coverage expansion in U.S. history.

Notes This paper was funded by the University of Chicago. Additional funding supporting the development of the microsimulation model described was provided by the Department of Health and Human Services (Contract: HHS-P23320054301ER) and the Agency for Health Care Research and Quality (Contract: HHS-N263200500063293). 1. Regression coefficients for this plan choice model are available in Parente et al. (2007). Note that HMO copays were converted to actuarially equivalent coinsurance rates for predicting the HMO enrollment probability. 2. Pauly and Herring (1999) have suggested that individual policies contain some degree of group experience rating and vice versa. According to Pauly and Herring, premiums in the individual market do not rise one-for- one with predictable expenses, and premiums in the ESI market have a positive association with predicted individual medical expenditures, contrary to the GER hypothesis. Notwithstanding these fi ndings, it was decided to use IER and GER as rating assumptions because these methods are more tractable and because it is not clear how to combine them to form “mixed” ratings systems as suggested by Pauly and Herring. 3. See Pauly, Percy, and Herring (1999) for data on loading fees in the individual health insurance market. 4. The Medical Expenditure Panel Survey (MEPS) uses “establishment size” rather than employer size. The three size classes are fewer than fi fty employees, fi fty to two hundred, and more than two hundred. The loading factors for these classes are assumed to be 20 percent, 15 percent, and 10 percent respectively. 5. The 54.8 million estimate comes from Center for Health and Economy 2013 report. 6. Estimates of previous premium elasticities are available from a 2005 NBER working paper presentation at the www.ehealthplan.org website.

158

Stephen T. Parente

References Kaiser Family Foundation. 2011. Employer Health Benefits Survey. Accessed September 24, 2012. http://ehbs.kff.org/pdf/2011/8225.pdf. Feldman, Roger, Stephen T. Parente, Jean Abraham, Jon B. Christianson, and Ruth Taylor. 2005. “Health Savings Accounts: Early Evidence of National Take-Up from the 2003 Medicare Modernization Act and Future Policy Proposals.” Health Affairs 24 (6):1582– 91. Parente, Stephen, Roger Feldman, Jean Abraham, and Jon Christianson. 2007 “Continuation of Research on Consumer Directed Health Plans: HSA Simulation Model Refi nement. Final Technical Report for DHHS Contract HHSP23320054301ER.” Parente, Stephen T., Roger Feldman, Jean M. Abraham, and Yi Xu. 2011. “Consumer Response to a National Marketplace for Individual Insurance.” Journal of Risk and Insurance 78 (2): 389–411. Parente, Stephen T., Roger Feldman, Jean Abraham, and Yi Xu. 2011. “Consumer Response to a National Marketplace for Individual Health Insurance.” Journal of Risk and Insurance 78 (2): 389–411. Parente, Stephen T., Roger Feldman, and Jon B. Christianson 2004. “Employee Choice of Consumer-Driven Health Insurance in a Multiplan, Multiproduct Setting.” Health Services Research 39(4, part 2): 1091–1111. Parente, Stephen T., Roger Feldman, Jon B. Christianson, and Jean Abraham. 2005. “Health Savings Accounts: Early Estimations on National Take-up from 2003 MMA and Future Policy Proposals.” Final Report on Contract HHSP233200400573P: Analytic Support in Assessing the Impact of Health Savings Accounts on Health Insurance Coverage and Costs. Pauly Mark, and Bradley Herring. 1999. Pooling Health Insurance Risks. Washington, DC: AEI Press. Pauly, Mark, Allison Percy, and Bradley Herring. 1999. “Individual Versus JobBased Health Insurance: Weighing the Pros and Cons,” Health Affairs 18 (6): 28–44. Weiner, Jonathan, Barbra Starfield, Donald Steinwachs, and Laura Mumford. 1991. “Development and Application of a Population Oriented Measure of Ambulatory Care Case-Mix.” Medical Care 29: 452– 72.

Chapter seven

After the ACA Freeing the Market for Healthcare John H. Cochrane

Introduction

M

ost health-economics policy discussion takes the bulk of our current legal and regulatory structure for granted, in particular that the government will have a heavy hand in providing, paying for, and directing the private provision of and payment for healthcare. Opponents of the optimistically named Affordable Care Act (ACA) delight in pointing out its unintended consequences, mangled incentives, and exploding budgets.1 Fans work to patch it up with new layers of regulation or “reforms.” Neither takes a ground-up, first-principles approach to understand why our current system is such a mess and how a better system might emerge. That is my goal. I survey the supply, demand, and market for healthcare and health insurance, to think about how those markets should work to provide quality care, low cost, and technical innovation. A market-based alternative does exist, and it is realistic. Healthy markets do not emerge because our current web of healthcare laws and regulations forbids them from doing so, not because of intractable market failures. But deregulation is not easy. The impediments to well-functioning healthcare and insurance markets go deep into federal, state, and local law, regulation, and practice. And the pieces are linked; greater competition, innovation, and entry by suppliers, greater control by consumers, and insurance innovations that cure the current mess each need the others in order to function.

162

John H. Cochrane

This analysis is obviously aimed at the long run. Thinking through how a freer and more competitive healthcare and insurance market can work and how most of the regulatory apparatus is doomed will not get anyone hired as a consultant, lobbyist, or adviser, nor will it generate bundles of government- or industry-provided research funding. It will not lead to immediate policy impact. But such long-run thinking is important nonetheless. Opponents of the ACA who would see it repealed need a detailed, coherent alternative, and even if the alternative is “leave it to the market,” they need to understand and explain how that alternative can realistically address the cost, “access,” and other evident pathologies of contemporary healthcare and insurance markets. The status quo was a mess, and the concerns that motivated the ACA were real. If the ACA remains, as is likely, but stumbles from one crisis to another and eventually falls apart of its own weight, it will be equally important to have that detailed coherent alternative in our back pockets. I focus on the supply and demand for health care, which gives this essay a bit of novelty. Curiously, most of the current policy debate, and most of our regulation, focuses on health insurance, the question of who will pay the bill, as if the market for health care were functioning normally. The market for health care, which is if anything even more dysfunctional, and which underlies any health insurance scheme, is relatively neglected.

Healthcare Supply We all agree what we would like to see: healthcare needs to become efficient and innovative and to provide high- quality care at reasonable cost. A. Cost Reduction and Innovation: Some Examples How will this happen? Well, we have before us many good examples. WalMart and Home Depot revolutionized retail. Airlines are dramatically cheaper than in the 1970s. Consumer electronics, telecommunications, computers, and even cars are much better and cheaper, for what you get, than ten or twenty years ago. These revolutions are not just about technology. In most of these cases, we see process innovation, reorganizing activities to deliver com-

After the ACA: Freeing the Market for Healthcare

163

plex services at lower cost and with better and more uniform quality. This process efficiency is most glaringly absent in healthcare. 2 Southwest Airlines turns a plane around in twenty minutes, and has fi nally figured out how to get people on it without the chaos at United and American. Wal-Mart’s and Home Depot’s success is as much about organizing and standardizing the motion of people and inventory as it is about adopting technology, outsourcing supply, or negotiating lower prices. Toyota assembles a car with thirty hours of labor. As Atul Gawande asked in the New Yorker, 3 the Cheesecake Factory delivers a complex service- oriented product with remarkable quality, efficiency, and cost—why can’t hospitals do the same? Beyond stories, Amitabh Chandra and Jonathan Skinner summarize the academic literature, writing that “there is increasing evidence of the potential for cost-saving technologies (with equivalent or better outcomes) in the management and organization of healthcare to yield substantial productivity gains. But these types of innovations are unlikely to diffuse widely through the healthcare system until there are much stronger incentives to do so.”4 But our hopes for healthcare go beyond the obvious need to streamline process and delivery and to adopt cost-saving technology. We do not want 1950s care at 1920s prices. Technical innovation is, fundamentally, why we can be so much healthier than our grandparents. Healthcare markets need to bring that innovation as fast as possible—and then diffuse it quickly down to the mass market. My example industries are also great at this sort of technology innovation and diffusion. Healthcare is a paradox in that innovation is widely reviled as a cause of increased costs. The standard economists’ answer is that we are mistaking “cost” for “price” and “introduction of new goods.” A new $500,000 treatment represents a reduction in cost—widening of the budget constraint—over a less effective but still available $50,000 older treatment. But though economically correct, this answer is unsatisfying—especially to those needing the care and those paying the bills—because we all see the monstrous inefficiencies in healthcare. That $500,000 could be $100,000. We know we could get more and faster technical innovation and lower prices. Why does Moore’s law not apply to medical devices? Why has the price of cell phones, GPS devices, and computers come down so fast relative to the prices of medical technology? Where is the home MRI? There is nothing deeply different about medical and other technology.

164

John H. Cochrane

The answer is that supply and demand—in the current highly regulated system—is not producing the Moore’s law incentives. In my example industries, innovation also does not always mean lower cost. I paid $1500 in 1982 for an IBM PC with 16K of RAM and one floppy disk drive. I paid about the same (nominal) for my most recent laptop, with vastly more power. Nissan plans to sell $3,000 cars in China and India5—with no airbags. We have chosen much better cars for higher prices. But my example industries did a good job of pushing the cost/innovation/quality frontier out to its limits, and then discovering where people really want to be. If we “spend more” today, we know we’re getting a good deal, and simply choosing a different point on a far better frontier than we faced twenty years ago. We know a better healthcare frontier is possible. My example industries do not cut costs by selling shoddy products or service. Instead, they provide consistent quality on the dimensions people turn out to really care about, and save on those that people do not really care about. Southwest gets you where you want to go at convenient times, with a good on-time record and admirable safety. And seats twenty- seven inches apart, while feeding you peanuts. People are not willing to pay the extra $20 that slightly more legroom would cost. The iPhone error rate is a lot lower than the medical error rate. Wal-Mart shirts use inexpensive materials, and they are sold in environments far less sexy than Michigan Avenue boutiques, but it is rare to fi nd one torn, or missing buttons. The theory that unregulated competitive suppliers will pawn off shoddy merchandise on consumers, so often expressed in medical contexts, is exactly false in every other industry. Restaurants and hotels tremble at a poor Yelp.com review. The corporatization and standardization of my example industries, which many people bemoan, is a good part of their ability to deliver consistent quality. If each airplane and pilot were a different practice, quality would vary a lot more. B. Competition and Entry How can healthcare emulate the quality improvement and cost reductions of these successful service- oriented industries? My examples share a common thread: intense competition—in particular, competition from new entrants, who put old companies out of business or force unwelcome and disruptive changes. Microsoft displaced IBM, and Google is displacing Microsoft. Wal-Mart displaced Sears, and Amazon may displace

After the ACA: Freeing the Market for Healthcare

165

Wal-Mart. Typewriter companies did not invent the word processor; word-processing companies did not invent the PC. The post office did not invent FedEx or email. Kodak is out of business, famously hobbling its digital cameras to protect a dying fi lm business. Toyota, not competition between Ford, GM, and Chrysler, brought us cheaper and better cars. When the older businesses survive, it is only the pressure from new entrants that forces them to adapt. I will not dwell on just how uncompetitive healthcare is, as I do not think the point needs belaboring. The simple fact that hospitals will not tell you a price ahead of time makes it blatantly obvious. No competitive industry would dream of getting away with this. As one good academic study of this phenomenon, Jaime Rosenthal, Xin Lu, and Peter Cram posed as an elderly patient seeking a hip replacement and wishing to pay cash6. Few hospitals could even quote a price, and the price quotes the authors fi nally received varied from $11,100 to an amazingly precise $125,798. My examples share another common thread. They remind us how painful the cost control, efficiency, and innovation processes are. When airlines were regulated, artificially high prices did not primarily go to stockholders. They went to unionized pilots, fl ight attendants, and mechanics. They produced an easy life more than fi nancial reward. Protection for domestic car makers supported generous union contracts and inefficient work rules more than outsize profits. “Bending down cost curves” in these examples required cleaning out these rents, through offshoring, elimination of union contracts and work rules, mechanization, pressure on suppliers, internal restructurings, and painful bankruptcies and mergers in which lots of people—both workers and well-paid managers—lost their jobs to others. The fact that so much cost reduction comes from new entrants, not reform at the old companies, is testament to how painful this process is and the ability of incumbents to protect the status quo. The Big Three still take forty hours to build a car as opposed to Toyota’s thirty. And two of them went bankrupt, while Toyota sits on a cash reserve. American and United are still struggling to match Southwest’s efficiencies, after thirty years. The parts of Kodak invested in fi lm simply could not let the company exploit its technical knowledge in optics and electronics. Chicago’s teacher unions are fighting charter schools tooth and nail. A quick look at a modern hospital and its suppliers reveals many similar ossified structures. It suggests just how wrenching the same transformations

166

John H. Cochrane

will be. And it suggests just how hard healthcare incumbents will fight to stop it, if they can. C. Competition and Regulation So, where are the Wal-Marts and Southwest Airlines of healthcare? They are missing, and for a rather obvious reason: regulation and legal impediments. A small example: in Illinois as in 35 other states,7 every new hospital or even major purchase requires a “certificate of need.” This certificate is issued by our “hospital equalization board,” appointed by the governor and, like much of Illinois politics, regularly in the newspapers for various scandals. The board has an explicit mandate to defend the profitability of existing hospitals. It holds hearings at which they can complain that a new entrant would hurt their bottom line. Specialized practices that deliver single kinds of service or targeted groups of customers cheaply face additional hurdles, because they undermine the cross-subsidization provided by “full service” hospitals. For example, the Institute for Justice is bringing a major suit on behalf of a specialty colonoscopy practice in Virginia, which local full service hospitals managed to ban.8 This is exactly the form of regulation put in place by the Civil Aeronautics Board until the late 1970s, which produced airline prices much higher than they are today. Airlines had to show need for a new route, and incumbents defended monopoly rents on the grounds that they crosssubsidized service to small airports. This one deregulation is pretty much what brought us cheap airline fl ights now. Revealingly, certificate of need laws were part of an earlier round of “cost containment” and were federally mandated for a while. The theory sounds sensible enough, and you can easily imagine it echoing through academic conferences to gentle approval. In a fee-for-service system, there can be an incentive to buy too many MRI machines and then prescribe “needless” scans, which insurance companies and the government would be forced to pay for. “Well,” said an earlier round of health-policy experts, “we’ll patch that up by having a regulatory board review the ‘need’ for major investments or hospital expansion to avoid ‘overinvestment.’ ” It is a story worth remembering; how a regulatory cost- containment patch to one broken system (poor incentives in fee-for-service reimbursement)

After the ACA: Freeing the Market for Healthcare

167

turned swiftly into a well- captured barrier to competition and wound up increasing costs. How occupational licensing is captured to restrict supply and push up prices should be obvious by now—Milton Friedman wrote his PhD dissertation on it, and a chapter in Capitalism and Freedom in 1962. Little has changed. For example, Uwe Reinhart recently covered the AMA’s opposition to a California measure that would have allowed nurse practitioners to perform some simple primary care services.9 He particularly savaged the usual argument that consumers have a right to the quality of a licensed doctor, noting that half of California’s physicians do not take new Medicare patients. If you are a parent, you have been there. It is 2:00 a.m. in a strange city. The kid has an ear infection. She needs amoxicillin, now. Getting it is going to be a three-hour trip to an emergency room, hundreds of dollars, so a “real doctor” can peer in her ear, then off to the pharmacy to fi ll the prescription. A nurse practitioner at the Wal- clinic could handle this in five minutes for $15. I am not arguing that we have to get rid of licensing. But licensing for quality does not have to mean restriction of supply to keep wages up—including state-by-state licensing, restriction of residency slots, restrictions on the number of new medical schools, and restrictions that encourage overuse of doctors where they are not needed. Restrictions on immigration of doctors and nurses keep prices up here, as they keep out high-skilled workers in many fields. Here our immigration law dovetails with occupational licensing restrictions. Immigration law is explicitly designed to keep American wages up. We forget that we pay those wages, or kid ourselves that we can drive wages up and costs down. Einer Elhauge examines “fragmentation” of medical care in detail,10 that is, the fact that care is bought essentially from different doctors and specialists, even in hospital settings, rather than in an integrated manner—as, say airline travel is, where you do not separately purchase pilot, fl ight attendant, fuel, and baggage services. My examples suggest a consolidation, integration, and corporatization of overall health service provision, in the same way restaurant chains displace individual stores. What stops this defragmentation? Elhauge surveys research concluding that nothing in the nature of healthcare seems to require this fragmented structure, since hospitals in other countries have salaried

John H. Cochrane

168

doctors. He concludes instead: “The dominant cause of fragmentation instead appears to be the law, which dictates many of the fragmented features described above and thus precludes alterative organizational structures.”11 He lists a long string of legal impediments, including Medicare reimbursement rules, laws against corporate practice of medicine, and tort doctrines. Referring to private insurance: State laws generally make it illegal for physicians to split their fees with anyone other than physicians with which a physician is in a partnership. More important, alternative payment systems, such as paying a hospital (or other fi rm) to produce some health outcome or set of treatments, would make sense only if it has some control over the physicians and other contributors to that outcome and treatments. And other laws preclude such control. . . . The corporate practice of medicine doctrine provides that fi rms—whether hospitals or HMOs—cannot direct how physicians practice medicine because the fi rms do not have medical licenses, only the physicians do. Although some states allow hospitals to hire physicians as employees, that change in formal status does not help much if the employer cannot tell the employee what to do. Even if the law did not prohibit such interference, tort law generally penalizes fi rm decisions to interfere with the medical judgments of individual physicians, making it unprofitable to try. . . . Further, hospital bylaws usually require leaving the medical staff in charge of medical decisions, and those bylaws are in turn required by hospital accreditation standards and often by licensing laws. . . . Private insurer efforts to directly manage care have likewise been curbed by the ban on corporate practices of medicine and the threat of tort liability. In addition, states have adopted laws requiring insurers to pay for any care (within covered categories) that a physician deemed medically necessary, banning insurers from selectively contracting with particular providers, and restricting the fi nancial incentives that insurers can offer providers.12

Laws against the “corporate practice of medicine” are another example of restrictions that end up limiting competition and innovation. The American Health Lawyers Association explains:13 The CPM doctrine generally prohibits a business corporation from practicing medicine or employing a physician to provide professional medical services. Corporate employment of a licensed professional has been prohibited on the grounds that such a relationship “tends to the commercialization and debasement of those professions.”

After the ACA: Freeing the Market for Healthcare

169

Commercialization is what competition is all about. My cost- cutting examples are all for-profit companies. About 70 percent of hospitals and 85 percent of healthcare employment is in nonprofits,14 whose legal and regulatory treatment protects much inefficiency from competition. If United did not have to pay taxes, Southwest’s job would have been that much harder. Maybe for-profit companies pay too much attention to stock prices. But nonprofits can go on inefficiently forever, with no stockholders to complain. The whole point of a nonprofit is to pursue goals other than economic efficiency. More importantly, if a for-profit company is inefficiently run, another company or a private-equity firm can buy up the stock cheaply, replace management, and force reorganization. Nonprofits (especially their management) are protected from this “market for corporate control.”15 Many nonprofit hospitals are too small, cannot merge, and, unable to issue stock by defi nition of “nonprofit,” are undercapitalized. Recognizing some of these pathologies, there is a wave of mergers and transfers between for-profit and not-for-profit status. But there is lots of gum in the works. When a nonprofit is sold or converts to for-profit, the state attorney general and courts can weigh in on the sale; legally to ensure that the proceeds benefit a charitable cause related to the nonprofit’s original mission. This is a great opportunity for competitors to block the change.16 The FTC is ramping up antitrust action against hospital mergers.17 Hospitals need economies of scale for expensive, specialized modern medicine and to comply with the avalanche of regulation and insurance paperwork. The FTC worries about local monopolies able to raise prices, especially given the inelastic demand by insurers and government reimbursement. So here we have the government forcing small size in order, it hopes, to boost competition with one hand; stopping entry, explicitly to protect hospitals from competition, with another; trying to force larger “networks” through “affordable care organizations” to obtain the needed economies of scale with the third; and preserving doctor independence from competitive pressure by law with the fourth. The schizophrenic attitude of our regulatory regime to size and competition comes partly down to its desire to enforce cross-subsidies and mandates. For example, the Emergency Medical Treatment and Active Labor Act (EMTALA) requires pretty much all acute- care hospitals to provide care

170

John H. Cochrane

for emergencies and active labor patients, without any provisions for reimbursement. And Medicare reimbursement rates are notoriously lower than costs. So hospitals have to make up the difference by overcharging other patients, both those with insurance and the few cash customers. But you cannot have cross-subsidies with competition—those being overcharged will quickly leave. Thus, the hospitals providing EMTALA service and the insurance companies cross-subsidizing Medicare have to be protected from competition, or they will not be able to stay in business. That is fi ne for a while, but businesses protected from competition and able to cross-subsidize money-losing operations soon become complacent and sclerotic and fi nd other ways to lose money. They also fi nd ways to lobby regulators for even more protection from competition, so as to continue to provide the regulators’ desired cross-subsidy. Regulation, mandated cross-subsidies, and protection from competition also help to hide the size of the government’s interventions from a skeptical electorate. If the government taxed corporations and used the revenue to provide health-insurance subsidies, that action would count on the budget as “taxing and spending.” The government instead mandates that employers shall provide health insurance, and then neither the tax nor the spending show up in the government’s budget. The economic effect is exactly the same, and the distortions are exactly the same. (Witness the number of jobs suddenly cut down to 29.5 hours a week once the ACA required health insurance for jobs over 30 hours a week.) We are just kidding ourselves in many ways. It is amazing that computerizing medical records was part of the ACA and stimulus bills. Why in the world do we need a subsidy for this? My bank computerized records twenty years ago. So did my car repair shop. Why, in fact, do doctors not answer emails? And do they still send you letters by the post office, probably the last business to do so? Or maybe grudgingly by fax, only 20 years obsolete? Why, when you go to the doctor, do you answer the same twenty questions over and over again, and what the heck are they doing trusting your memory to know what your medical history and list of medications are? Part of the answer: they’re afraid of being sued. Confidentiality regulations, apparently more stringent than those for your money in the bank. They can’t bill email time. So medical records offer a good parable; rather than look at an obvious pathology, rather than ask what features of current law and regulation are causing hospitals to avoid the computer revolution that swept

After the ACA: Freeing the Market for Healthcare

171

banks and airlines twenty years ago, and rather than remove those roadblocks, the government adds a new layer of subsidies and contradictory legal pressure. One regulation says move right, the other says move left. The impediments to supply- side competition go far beyond formal legal restrictions. Our regulatory system has now evolved past laws, past simple, explicit, and legally challengeable regulations, to hand vast discretionary power to officials and their administrative bureaucracy—either directly (“the Secretary shall determine” is the chorus of the ACA) or through regulations so lengthy, vague, and contradictory that administrative discretion is the effect. Witness the wave of waivers to ACA that HHS handed out to friendly companies. Those administrators can easily be persuaded to take actions that block a disruptive new entrant, with little recourse for the potential entrant. And criticizing a regulator with such power is a dangerous business. (Lobbying government to adopt rules or take actions to block entrants is legal, even if those actions taken directly would violate antitrust laws, under the Noerr-Pennington doctrine.) Forget about Wal- clinics; Chicago and New York have kept Wal-Mart from selling food and clothes to their residents for years, at the behest of unions and competitors, by denying Wal-Mart all the necessary permits and approvals. So many citizens, especially our poor and vulnerable, continue to live in employment and retail deserts. The increasing spread of medical tourism to cash- only offshore hospitals is a revealing trend. Why does this have to occur offshore? What’s different about the hospital location? Answer: the regulatory regime. So, what’s the biggest thing we could do to “bend the cost curve,” as well as fi nally tackle the ridiculous inefficiency and consequent low quality of healthcare delivery? Look for every limit on supply of healthcare services, especially entry by new companies, and get rid of it. D. The Reregulation Path Now, this is of course not the way of current policy. The ACA and the health-policy industry are betting that additional layers of new regulation, price controls, effectiveness panels, “accountable care” organizations, and so on will force efficiency from the top down. And they plan to do this while maintaining the current regulatory structure and its protection for incumbent businesses, management, and employees. Well, let us look at the historical record of this approach, the great examples in which industries, especially ones combining mass-market

172

John H. Cochrane

personal service and technology, have been led to dramatic cost reductions, painful reorganizations towards efficiency, improvements in quality, and quick dissemination of technical innovation by regulatory pressure. In other words, let’s have a moment of silence. No, we did not get cheap and amazing cell phones by government ramping up the pressure on the 1960s AT&T. Southwest Airlines did not come about from effectiveness panels or an advisory board telling United and American (or TWA and Pan Am) how to reorganize operations. The mass of auto regulation did nothing to lower costs or induce efficient production by the Big Three. When has this approach ever worked? The Postal Service? Amtrak? The Department of Motor Vehicles? Road construction? Military procurement? The TSA? Regulated utilities? European state-run industries? The last twenty or so medical cost control ideas? The best example and worst performer of all . . . wait for it . . . public schools? It simply has not happened. Government-imposed efficiency is, to put it charitably, a hope without historical precedent. And for good reasons. Regulators are notoriously captured by industries, especially when those industries feature large and politically powerful businesses with large and politically powerful constituencies, as in health insurance or as in most cities’ hospitals. In turn, regulated industries quickly become dominated by large and politically powerful businesses. See “Banks, comma, too big to fail.” (Several insurance companies were also bailed out in the fi nancial crisis, on the theory that failure of their retirement contracts was somehow a “systemic” danger. Many states now have only a few health insurers left. Too-big-to-fail protection for health insurers is not an abstract and distant worry.) This is not to say that regulators are not well-meaning and do not put great pressure on many industries. But the deal, “You do what we want, we’ll protect you from competition” is too good for both sides to resist. The addendum “And support us and our administration politically or else” is emerging fast. Needless to say, price controls have been a disaster in every case they have been tried. Long lines for gas in the 1970s are only the most salient reminder. Price controls’ predictable result is vanishing supply, abundant demand, and low quality. Try finding a doctor who will take new Medicare or Medicaid patients. The over-the-counter additional payments many providers now require will predictably become the under-the-counter payments or personal connections you need to get treated in many countries.

After the ACA: Freeing the Market for Healthcare

173

The current regulatory approach is not really well described as simple price controls—for example “Thou shalt not charge more than $3 per gallon of gas”—but rather as fiddling with a payment system of mindnumbing complexity and endlessly discovered unintended consequences. The past record of cost control and “incentive” efforts should warn us of how likely adding more complex rules is to work. There are already conferences for doctors to teach them how to maximize Medicare billing codes (there are sixty- eight thousand) for each visit,18 and there are 2.2 people doing medical billing for every doctor who actually sees patients, costing $360 billion.19 This regulatory failure seems instead to be a challenge to the next generation of planners. 20 But capture and the failure of price controls are only the beginning. Real cost reduction is a painful process, as my examples remind us, and our political system is allergic to pain. Can a regulator, appointee, or politician in a democracy really become a union buster, forcing painful concessions on workers, managers, suppliers, and other “stakeholder” beneficiaries of rents? Can a regulator realistically demand that jobs be outsourced or replaced by software? Can a regulator really preside over a wave of bankruptcies, mergers, and mergers, in which new businesses send old ones to the dustbin? They can stand back and let the market do it, but can they possibly take direct responsibility for these events? Consider a small example now in the news. Hospitals are starting to outsource the reading of X-rays, even to India. This activity is still heavily regulated—the radiologists must still be US-trained and certified, and also state certified. But already it is a cause célèbre for its potential to cost jobs. When the obvious happens—“Hmm, we have some good Indian doctors who can read the X-rays just as well”—you can imagine the scandal. And doesn’t every American deserve the best—a US radiologist on staff and present twenty-four hours a day, ready to consult with the doctor? Personal-injury law fi rms are already lining up to sue based on the “inferior quality” of outsourced readings, with requisite horror stories. 21 How could a regulator not just allow but demand outsourcing radiology and using Indian doctors? A big stated point of regulation is to ensure quality. It is interesting how bad a job it does. Regulators can impose minimum standards, requiring degrees, certification, inspections, and the like, and keep out really dangerous quacks. But beyond that they are terrible at pushing for higher quality, especially when quality means so much in the experience of a cus-

174

John H. Cochrane

tomer in a service- oriented business. Restaurant regulation keeps restaurants reasonably safe, but there is no regulatory pressure for Joe’s Tacos to use better cuts of beef, let alone to adopt molecular gastronomy, seat you quickly, or be polite. Yelp.com ratings do that in a way no regulator can hope to. Yet mind-numbing and competition- destroying regulation is routinely instituted on the argument that quality must be forced on businesses that are for some reason unwilling to provide it. Well, of course they are unwilling to provide it if they are not competitive. And they are not competitive when the regulator protects them from competition. My examples also do a remarkable job of getting rich people to pay through the nose voluntarily, covering fi xed costs for medium-income consumers. Two words: business class. But a politician who proposed taxing people this way to provide air travel would be hanged as a socialist. And a regulator who consigned middle-income patients to seat 25D while wealthier patrons got business class would be hanged as a fascist. E. Realism Now by being concrete and therefore realistic, I invite obvious complaints. What, I like airlines and Wal-Mart? Have I flown Southwest or shopped at Wal-Mart? (Yes to both, incidentally.) But I think the examples are good to remind us what efficiency looks like in the real world and how it is achieved, and to keep us from fantasies about what healthcare can look like and what outcomes regulators are likely to be able to achieve. We love to complain about airlines. But aside from the TSA’s security theater and air traffic control—both run by the government—what we really want is 1970s service at 2010 prices. Sorry, we can’t afford privatejet medicine for everyone. Southwest medicine has to be the goal—safe, effective, and just as comfortable as people are willing to pay for. Shop at Wal-Mart? Wal-Mart is putting all those cute mom-and-pop stores out of business. It is putting pressure on union jobs, the main reason Chicago kept it out all these years. It pushes suppliers relentlessly. It buys from China. Am I not being heartless? No. I am being realistic. The lesson from all our experience with other industries is that cost control and innovation are hard and brutal processes. Not just the businesses, but their suppliers and employees clamor for protection. Many of you are probably still squirming in your seats. You want some other way. You want to keep unionized jobs, “living wages,” “worker

After the ACA: Freeing the Market for Healthcare

175

protections,” or to “keep our community hospitals going.” Perhaps you mourn the bank tellers replaced by ATM machines and jobs sent to China. More deeply, you are probably squirming in your seats at my observation that quality varies enormously in efficient industries; some fly economy middle seat, and some fly in private jets. Some get shirts from Wal-Mart and some get shirts from Macy’s. Surely, does not every American deserve the best when it comes to healthcare? If so, you are not serious about reducing costs—that is, finding the efficient point on the quality-cost curve. This is simply a fact: you are adding other goals to the mix, so you are accepting rising costs to fund those other goals. Or you are fantasizing that you can have it both ways. And if you are having trouble putting those other considerations aside and accepting a consumer-focused Wal-Mart / Southwest Airlines model for healthcare, imagine how unlikely it is that the Department of Health and Human Services will force that model to emerge through its regulatory power.

Healthcare Demand The demand side of the healthcare market is just as severely distorted. A. Payment Plans and “Need” Most basically, with either government provision or private insurance, healthcare is bought in “payment plan” form. You pay a tax or a premium, then your expenses are “covered.” We all understand that when somebody else is paying, people do not economize on expensive services, shop for better deals, or accept less convenient but cheaper alternatives. More importantly, I think, demand affects supply, and demand distortions inhibit needed supply competition; it is a lot harder for new entrants to attract business when people are paying with someone else’s money. Is there something about the nature of healthcare, as an economic good, that necessitates payment-plan provision? Thinking about it, I think the opposite is true; healthcare, as an economic good, is a particularly poor candidate for payment-plan provision. I think people have in mind examples such as a simple wound or a broken arm. Even if it is free, nobody is going to overuse broken-arm

176

John H. Cochrane

treatment. Nobody will have a good arm put in a cast or have stitches just for fun. Pretty much any qualified doctor can handle such procedures; you do not need to fi nd one who is “really good at setting bones” (or so people think) but charges a higher price. So the “good” is well defi ned, it is a pretty generic commodity, the demand curve is steep, and what you “need” is clearly observable. But these are misleading examples. The actual demand curve for healthcare is incredibly elastic. When healthcare is provided at low cost, people consume prodigious amounts of healthcare services. Every cost estimate for government provision or subsidy, from the U.K. NHS to Medicare, Medicaid, and beyond has missed its mark by orders of magnitude. Furthermore, though it is common to disparage “overuse” in health policy circles, the elastic demand curve is real. These are real people, with painful and debilitating illnesses, and the “extra” test or visit to the specialist, the one more last-ditch treatment, might just be the one to finally help them. Conversely, when asked to pay more, consumers economize rapidly, refusing “too much” care in the judgment of the medical community. So we have attempted payment plans with limits—insurance rules, managed care, HMOs, effectiveness panels, affordable care organizations, and so on—to cut off the flat demand curve. Ezekiel Emanuel, Neera Tanden, and Donald Berwick, writing in the Wall Street Journal,22 explained the idea behind affordable care organizations: “Instead of paying a fee for each service, providers should receive a fi xed amount for a bundle of services or for all the care a patient needs.” Hmm. “Need.” “From each according to his ability, to each according to his need.” It has a nice ring to it. Why do I feel a certain foreboding? “Need” is not an economic concept. Would this setup work for clothes? Your employer gives you “access” to clothes by including a “clothes plan” in your benefits. Then your appointed “primary style consultant” will determine how many shirts you need, which you can pick from the preferred shirt-provider network (Kmart). (And if you show up at Kmart saying “I’d rather pay cash,” they charge $1,000 for a shirt.) Home repair? The home-repair effectiveness board will conduct peer-reviewed research on appropriate materials for kitchen counters. Sorry, granite is off the approved list, you don’t need it. Healthcare? For many patients, just getting through the diagnosis to decide what treatment they might try is an expensive and inconclusive nightmare, with trip after trip to various specialists. How much diagnosis do you really need in these circumstances?

After the ACA: Freeing the Market for Healthcare

177

Many diseases are chronic, requiring widely varying and individualspecific treatment plans. Nothing really works, and we are trading off different options with different bad side effects, and needing different levels of commitment from the patient. End- of-life care, care for the elderly, infi rm, handicapped, and mentally ill are very expensive, and all lie on a long string of quality versus quantity choices. Does grandma really need a five-star nursing home, a helper (a highly personal service!—Could insurance or government “provide” needed housecleaning services successfully?), or just support from family? Does need without considering cost, that is, willingness to pay, really even begin to describe the economics of this decision? Should a family that decides to provide care, saving the nation hundreds of thousands of dollars, receive no benefit? I had a back pain episode recently. (Somehow health policy always ends up with here’s-where-it-hurts anecdotes!) Did I need an MRI to really see the structural problem? Cortisone shots? Surgery? Physical therapy, or just a photocopy of recommended exercise? Physical therapy at the University of Chicago hospital, or at the specialty sports-rehab clinic that patches up the Bears? Or just a handful of ibuprofen and time to let it heal? Did my planned trip to Europe matter in this medical need? And why not speak the dirty little secrets? For most patients, “Stop smoking, exercise and lose some weight” is the best advice they could take. Patients’ awful compliance is an open secret. How much drugs and treatment do patients need who won’t stop smoking, lose weight, exercise, do the physical therapy, or comply with drug regimes? Another dirty little secret: quality, both actual and perceived, varies enormously. Rates of medical errors, infection rates, rates of success in difficult procedures, just getting basic diagnoses right, or even washing hands often enough vary widely. The quality of service provided, including everything from waiting times to convenience of making an appointment to whether the doctor answers emails, varies as well. Do you need an MRI this afternoon at 5:00 p.m. near your work, or on the other side of town two weeks from now? Conversely on supply: Yelp.com ratings have a huge effect on spurring this sort of attention to detail in restaurant services. Can bureaucratic rules really substitute in medical services? And medicine is not perfect. For a range of conditions, we have imperfect treatments, with varying side effects, and scientific knowledge of what works or does not is changing fast. What does need mean then?

178

John H. Cochrane

If only it were so simple to determine need. If only people like me went away quickly when told we do not need an MRI to fi nd out why our backs hurt. Or if people with hard-to- diagnose but debilitating illnesses like food allergies quietly went away rather than hold out hope that the next specialist will figure out the problem. B. Need and Willingness to Pay or Forego So what does “need” really mean? The only sensible economic defi nition I can think of is that “need” is the bundle of services you would choose if you were paying with your own money at the margin. You need that MRI to make sure your back pain will not just heal after six weeks of ibuprofen if you would be willing to shell out $1,000 of your own money to get it. And you need it delivered at a convenient hour, tomorrow, rather than next week, across town, if you are willing to pay that extra cost. “At the margin” is an important qualifier, because intuitive thinking soon mixes up “what you’d rather spend money on” with “what you can ‘afford.’ ” As economists, we are expected to avoid that confusion. A good way to do so is to pose the question in the positive rather than the negative: suppose we offered each patient the choice, “Your doctor prescribed this MRI. You can have the MRI or you can have $1,000 in cash.” The patient “needs” the MRI if he or she foregoes the cash and goes through with the MRI. This is an important and unsettling conceptual experiment. If the patient chooses to forego treatment, or to fi nd a cheaper alternative and keep half the cash, you can not argue the patient “cannot afford” treatment. It’s unsettling, because I think we suspect lots and lots of people would take the cash, especially at current inflated prices. So there is a lot of paternalism in healthcare policy, which we might be more upfront about. In any case, once defined, it’s pretty clear that this need is essentially impossible to measure externally for a personal service with so much variety and imperfection as healthcare. Moreover, many more people would need MRIs, by any definition of “need,” if competition and innovation drove the price down to $50. So we are just arguing about who makes the cost/benefit decision. What you “want” is where you make the cost/benefit decision. What you need is what I—or some panel of bureaucrats—think you should get.

After the ACA: Freeing the Market for Healthcare

179

I think the word “need” also has a moral tone, implying “what society owes you.” This seems even harder to defi ne or measure. How much back treatment did society owe me? Now economists might quibble with my defi nition of “need” as willingness to pay or forego because I left out income effects. My patient taking the cash instead of an MRI might have wanted to pay for food and rent. Perhaps “need” should mean “what you would be willing to pay or forego if you earned $1 million a year.” Alas, we don’t have the resources to pay for that defi nition of “need.” We simply cannot all fly on private jets at public expense. So while private jet stories are fun, given the social budget constraint the relevant question is whether someone earning $50,000 a year would give a much different answer than someone earning $80,000 per year. Care for the very poor and indigent is a separate question, which I discuss below. Now it is not so obvious that income is a large source of variation in “willingness to pay,” in this relevant range. For every other good and complex service, variation of demand across people within income categories is far greater than variation of demand with income by the average person. At Denny’s and at Alinea, some eat steak and some eat chicken. This pattern is likely to hold for healthcare as well. So, while it is a relevant quibble, in the end I think an argument based on income effects in the defi nition of “need” is distraction. C. Healthcare Demand, Bottom Line In sum, healthcare is a complex, highly varied personal service, not a simple well-defined commodity. The demand curve is as elastic as any in economics. When, where, how, how much, by whom are vital components of that service. Objective and subjective quality and corresponding cost vary tremendously, and many dimensions of that quality are not easily measurable. The distinction between “need” and “want” is at best unmeasurable and at worst simply meaningless. The broken arm is a horrendously misleading example. But healthcare is an economic good. Healthcare is not that different from the services provided by lawyers, auto mechanics, home remodelers, tax accountants, fi nancial planners, restaurants, airlines, or college professors. Payment-plan provision, with rationing by some bureaucratic determination of need, is based on the opposite and false assumptions and

180

John H. Cochrane

thus is pretty hopeless for healthcare. No planner can mimic the market outcome in which what you need is what you are willing to pay for at the margin. To some extent, private insurers offer high- quality versus generic plans to sort patients ex ante by quality versus willingness to pay. But regulation makes that sorting much harder; once we force guaranteed issue at the same price, it is next to impossible for insurers to maintain bare-bones versus Cadillac plans. The minute a bare-bones customer gets sick, that customer will demand to be issued a Cadillac plan at the same cost as everyone else. And health insurers will respond by tailoring plans to attract healthy consumers—for example, by offering free health club benefits—and discourage sick ones. The whole guaranteed issue plus mandate arrangement assumes that health insurance is a generic good, not one with good-better-best quality and price points. If not generic already, health insurance will soon be forced to be generic by this regulation. And regulatory rationing cannot say that anyone must shop at Wal-Mart. I conclude that at the margin, the consumer needs to be paying a lot closer to full marginal cost of healthcare, or, equivalently, receiving the full fi nancial benefits of any economies that consumer is willing to accept.

The Healthcare Market—Supply and Demand The obvious problem with my demand analysis is that the cash market is dead. Making people pay, and shop, is unrealistic. If you walk in to the University of Chicago Hospitals and say, “I don’t have insurance. I have a bank account. I’ll be paying cash,” their eyes will light up (after they figure out you actually have the cash). “We’ll pay for a hundred Medicare patients with this guy.” That is like walking up to United Airlines and saying “I want to go to Paris, fi rst class. Sell me a ticket.” Actually, it is worse—at least United will quote you a price up front and on its website, and let you compare with American. It will not usher you into a back room for a one- on- one negotiation over what you will pay. Nobody reading this essay really needs health insurance—income protection—for anything less than catastrophes. We pay for transmis-

After the ACA: Freeing the Market for Healthcare

181

sion repairs, leaking roofs, and vet bills out of pocket. We could easily “afford” most of our routine medical expenses, and even pretty big unplanned expenses, especially if we were paying commensurately lower health-insurance premiums. But we all have health insurance, and we deal with the paperwork nightmare. Why? Because we know we cannot simply pay for healthcare when we need it. Insurance companies now function as purchasing agents, negotiating complex deals on our behalf. But why, again? You do not need an insurance company to negotiate your cell phone contract, home repair and rehab, mortgage, airline fare, legal bills, or clothes, as we do for health. Moreover, why do we mix this negotiation with “insurance,” and a payment plan? Dr. Jones is in Humana’s network, Dr. Smith is in Blue Cross’s. What economic principle means I should not see Dr. Jones, just because some arcane negotiation took place behind the scenes? And what about the new low- cost specialty clinic that Dr. Thomas is setting up, which cannot get into either network? Part of the answer is the tax- deductibility of employer-provided group insurance. The 10 percent who really do not need health insurance pay high marginal income-tax rates, so a great deal of inefficiency is worth a tax dodge. But the bigger answer is that the market is missing robust supply-side competition. Hospitals would never get away with obscure pricing, hidden rebates, or massive cross-subsidies if they were facing serious competition from new entrants who could peel you away—and peel you away from your expensive “price negotiator” as well. The cash market is also dead because of the demand-side distortion: too many people have insurance, that is, highly regulated “payment plans.” Competing for cash customers just does not make enough money to keep a hospital going, and the pool of cash customers is a lot sicker. A hospital must choose, basically, to be all insurance or all cash. If it offers clear transparent prices to cash consumers, it cannot also play the game with insurance companies. (The spread of “concierge medicine,” the equivalent of private schools for people so fed up they just throw away health insurance, is an interesting phenomenon. But it is still too small to affect the overall market. There are no concierge, cash- only hospitals. That business seems to have to move offshore.)

182

John H. Cochrane

In a vicious circle, the absence of a functional cash market lies at the heart of many insurance pathologies and government cost- control problems. Insurance functions best when it is a small part of a market, in which prices are set by marginal consumers paying cash and competitive businesses supplying them. With little price discovery left in healthcare, health insurers have to do all the price negotiation in a vacuum. Airlines, restaurants, and car repair work reasonably well even though in each case a large fraction of consumers are not paying with their own money—instead using expense accounts in the fi rst two cases, insurance in the third. Each has competitive supply and a remaining fraction of consumers who feel marginal decisions, enough to allow price discovery and competitive pressure for efficiency. Healthcare is so far gone that it is missing the price discoverers. Many pricing decisions are based on Medicare reimbursement rates. But from where does Medicare get its rates? With no supply- equalsdemand pricing going on anywhere, how does price discovery happen? Ed Lazear reported that in the Soviet Union, which had no price discovery mechanism, central planners used the Sears Catalog to set relative prices. 23 But what happens when there is no Sears catalog left? I suspect this is the reason that we cannot even separate negotiator and insurance functions. It has long puzzled me why insurance companies do not offer the very sickest patients or rich people who do not want “insurance” the following deal: “You need our negotiating power. But we do not want to take you on as a risk. So you get access to all our negotiating power, but you have to pay all your bills.” Alas, hospitals and insurance companies have negotiated contracts with lump-sum rebates, so the cost of a particular patient is not really measured. The phony-baloney bills you see really are phony-baloney bills. Perhaps the companies also fear that insurance regulators would quickly put a stop to the practice and force the company to pay for the sick person’s care rather than just pass on huge bills. Part of the reason for phony pricing is that hospitals know most “cash” customers will not end up paying, so they will end up negotiating charity care. Then they can report the discount as a contribution to charity care. Nicholas Kristof’s story in the New York Times of the travails of an uninsured friend who got cancer unwittingly but beautifully illustrates my point. 24 Kristof cites completely ridiculous prices, then explains how his friend applied for charity care and had a $550,000 bill knocked down to

After the ACA: Freeing the Market for Healthcare

183

$1,339. But just to reiterate how ridiculous the cash pricing is, the hospital still wanted to charge $1,400 for an ambulance ride. Freeing up either supply or demand without freeing up the other will do little good. Increasing co-pays can help to ration expensive or overpriced services, but co-pays do not stimulate supply or efficiency as long as new entrants cannot come in and compete for business. Allowing new entrants to compete for business does not do any good as long as few consumers are able to vote with their money.

Health Insurance I have written a lot about how to fi x health insurance, so I will not repeat all that here. 25 To summarize briefly, health insurance should be individual, portable, lifelong, guaranteed renewable, transferrable, competitive, and lightly regulated, mostly to ensure that companies keep their contractual promises. “Guaranteed renewable” means that your premiums do not increase and you cannot be dropped if you get sick. “Transferable” gives you the right to change insurance companies, increasing competition. Insurance should be insurance, not a negotiator and payment plan for routine expenses. It should protect overall wealth from large shocks, leaving as many marginal decisions unaltered as possible. “Access” should mean a checkbook and a willing supplier, not a federally regulated payment plan. Such insurance would, of course, be a lot cheaper. And insurance can be all these things, in a free or lightly regulated market. Preexisting conditions, failure by the young and healthy to obtain insurance, and spiraling insurance costs—the main problems motivating the ACA—are neatly addressed by this alternative, as I and others have explained at length elsewhere. Why do we not have a system? First, because law and regulation prevent it from emerging. Before the ACA, the tax deduction and regulatory pressure for employer-based group plans were the elephants in the room. This distortion killed the long-term individual insurance market, and thus directly caused the preexisting conditions mess. Anyone who might get a job in the future will not buy long-term individual insurance. Mandated coverage, tax deductibility of regular expenses if cloaked as

184

John H. Cochrane

“insurance,” prohibition of full rating, barriers to insurance across state lines—why buy long-term insurance if you might move and are forbidden to take it with you?—and a string of other regulations did the rest. Now, the ACA is the whale in the room; the kind of private health insurance I described is simply and explicitly illegal. The second reason we do not have a system is that functional socalled insurance requires a functioning underlying market, which law and regulation have also prevented from emerging. We cannot reasonably write contracts about who pays the bill when the bill itself is so meaningless. If there were functional cash markets, health savings accounts could also substitute for much of the necessarily cumbersome functions of insurance. Health borrowing accounts, that is, health savings accounts with a preapproved line of credit that you can tap for unexpected expenses, are not insurance in the sense of transferring overall wealth but would help even more. But without functional (competitive) cash markets, health savings accounts are not that helpful either. Unfortunately, individual long-term policies were one of the fi rst casualties of Obamacare. In fall 2013, a large number of insurers canceled individual policies, most of which were guaranteed renewable under ACA requirements. Many customers faced large premium increases and more restrictive new policies under the exchanges, and may choose to go without insurance instead. Here was a population that did the right thing and bought insurance, even if badly overpriced, precisely for the right to keep it if they should get sick in later years. And the fi rst act of the ACA, just before the disastrous healthcare.gov rollout, was to cancel that insurance. The only silver lining is the number of voters who began to fi nd out what is really in the system, epitomized by a young woman writing a letter to Pam Kehaly, president of Anthem Blue Cross in California, on receiving a 50 percent rate hike: “I was all for Obamacare until I found out I was paying for it.” 26

Objections The idea that healthcare and insurance can and should be provided by deregulated markets and that existing regulations are the main source of our problems is fairly radical within the current policy debate. Let me deal with a few of the standard objections.

After the ACA: Freeing the Market for Healthcare

185

A. The Poor “What about the homeless guy with a heart attack?” Let us not confuse the issue with charity. The goal here is to fi x health insurance for the vast majority of Americans—people who have jobs, people who buy houses, cars, and cell phones, people who buy insurance for their houses and life insurance for their families. Yes, we will also need charity care for those who fall through the cracks, the victims of awful disasters, the very poor, and the mentally ill. This will be provided by government and by private charity. It has to be good enough to fulfi ll the responsibilities of a compassionate society, and just bad enough that few will choose it if they are capable of making choices. I wish it could be better, but that is the best that is possible. For people who are simply poor, but competent, vouchers to buy health insurance or to refi ll health savings accounts make plenty of sense. But supplying decent charity care does not require a vast “middle class” entitlement and regulation of health insurance and healthcare for everyone in the country, any more than providing decent homeless shelters (which we are pretty scandalously bad at) or Section 8 housing subsidies for the poor requires that we apply ACA-style payment and regulation to your house and mine, to Holiday Inn or to the Four Seasons. To take care of homeless people with heart attacks, where does it follow that your health insurance and mine must cover fi rst- dollar payment for wellness visits and acupuncture? The ACA is hardly a regulation minimally crafted to solve the problems of homeless people with heart attacks.

B. The Straw Man There is a more general point here, which will appear time and again as I answer criticisms. Critics adduce a hypothetical situation in which one person might be ill served by a straw-man completely unregulated market, with no charity or other care (which we have had for over eight hundred years, 27 long before any government involvement at all), which nobody is advocating. They conclude that the hypothetical justifies the thousands of pages of the ACA, tens of thousands of pages of subsidiary regulation, and the mass of additional federal, state, and local regulation applying to every single person in the country. How is it that we accept this deeply illogical argument, or that anyone making it expects it to be taken seriously?

186

John H. Cochrane

Will not one person fall through the cracks or be ill served by the highly regulated system? If I find one Canadian grandma denied a hip replacement or one elderly person who cannot get a doctor to take her as a Medicare patient, why do I not get to conclude that all regulation is hopeless and that only an absolutely free market can function? Both straw men are ludicrous, but somehow smart people make the fi rst one, in print, and everyone nods wisely. C. Adverse Selection We all took that economics course in which the professor shows how asymmetric information makes insurance markets impossible due to adverse selection. Sick people sign up in greater numbers, so premiums rise and the healthy go without. George Akerlof’s justly famous “The Market for Lemons” 28 proved that used cars cannot be sold because sellers know more than buyers. Yet CarMax thrives. Life, property, and auto insurance markets at least exist, and function reasonably well despite the similar theoretical possibility of asymmetric information. Life insurance is also “guaranteed renewable,” meaning you are not dropped if you get sick. Is the story even true? Do most people, with knowledge of aches and pains, really know so much more about likely cost than an insurance company armed with a full set of computerized health records, actuaries, health economists, and whatever battery of tests it wants to run? Or is asymmetric information market failure in health insurance just a myth passed from generation to generation, despite functioning markets in front of our eyes? Now the real world does see a big “adverse selection” phenomenon. Sick people are more likely to buy insurance, and healthy people forego it. But the insurance company does not charge people the same rate because it can’t tell who is sick or likely to cost more—the fundamental, technological, and intractable information asymmetry posited in your economics class. The insurance company charges the same rate because law and regulation force it to do so. The insurance company is barred from using all the information it has. Regulation seems to feel that we have the opposite information problem; insurers know too much. The centerpiece of the ACA, after all, is banning the use of information, that is, preexisting conditions, not a

After the ACA: Freeing the Market for Healthcare

187

great regret that insurers cannot tell who has preexisting conditions in order to charge them more. This source of adverse selection is the legal and regulatory problem, not the information problem of economic theory, and easily solved. If insurance were freely rated, nobody would be denied. Sick people would pay more, but “health status” insurance or guaranteed renewability solve that problem and eliminate the preexisting conditions problem. (See note 21 for references.) Adverse selection due to fundamental information asymmetry in an unregulated market is, as far as I can tell, a cocktail-party market failure. It is a nice story, but does not quantitatively account for the real world. Furthermore, the ACA is not a minimally crafted regulation to solve the problem that people know more than their insurance companies can know about their health. Once again we are subject to the logical fallacy of accepting the entire regulatory structure because of one alleged failure of a hypothetical free market. D. Shopping Paternalism Defenders of regulation reiterate the view that markets can’t possibly work for health decisions:29 “A guy on his way to the hospital with a heart attack is in no position to negotiate the bill.” “One point I cannot agree with is that competition can work in healthcare, at least as it does in other markets. I cannot fathom how people faced with serious illness will ever make cost-based decisions.” “What about those who currently don’t have the background and/or the economic circumstances to consume healthcare, (e.g. take anti-hypertensive medicine instead of [buying] an iPhone)?”

Ezra Klein trying to understand why healthcare prices are so high and so obscure, writes:30 Health care is an unusual product in that it is difficult, and sometimes impossible, for the customer to say “no.” In certain cases, the customer is passed

John H. Cochrane

188

out, or otherwise incapable of making decisions about her care, and the decisions are made by providers whose mandate is, correctly, to save lives rather than money. In other cases, there is more time for loved ones to consider costs, but little emotional space to do so—no one wants to think there was something more they could have done to save their parent or child. It is not like buying a television, where you can easily comparison shop and walk out of the store, and even forgo the purchase if it’s too expensive. And imagine what you would pay for a television if the salesmen at Best Buy knew that you couldn’t leave without making a purchase.

New York Times columnist Bill Keller put it clearly, in “Five Obamacare Myths:”31 [Myth:] The unfettered marketplace is a better solution. To the extent there is a profound difference of principle anywhere in this debate, it lies here. Conservatives contend that if you give consumers a voucher or a tax credit and set them loose in the marketplace they will do a better job than government at fi nding the services—schools, retirement portfolios, or in this case health insurance policies—that fit their needs. I’m a pretty devout capitalist, and I see that in some cases individual responsibility helps contain wasteful spending on health care. If you have to share the cost of that extra M.R.I. or elective surgery, you’ll think hard about whether you really need it. But I’m deeply suspicious of the claim that a health care system dominated by powerful vested interests and mystifying in its complexity can be tamed by consumers who are strapped for time, often poor, sometimes uneducated, confused and afraid. “Ten percent of the population accounts for 60 percent of the health outlays,” said Davis [Karen Davis, president of the Commonwealth Fund]. “They are the very sick, and they are not really in a position to make cost- conscious choices.”

Now, “dominated by powerful vested interests and mystifying in its complexity” is a good point, which I also just made. But why is it so? Answer: because law and regulation have created that complexity and protected

After the ACA: Freeing the Market for Healthcare

189

powerful interests from competition. And is the ACA really creating a simple clear system that will not be “dominated by powerful vested interests?” Or is it creating an absurdly complex system that will be, completely and intentionally, dominated by powerful vested interests? But the core issue is these consumers who are “passed out, or otherwise incapable of making decisions about [their] care,” “strapped for time, often poor, sometimes uneducated, confused and afraid,” and “not really in a position to make cost-conscious choices.” Yes, a guy in the ambulance on his way to the hospital with a heart attack is not in a good position to negotiate. But what fraction of healthcare and its expense is caused by people with sudden, unexpected, debilitating conditions requiring immediate treatment? How many patients are literally passed out? Answer: next to none. What does this story mean about treatment for, say, an obese person with diabetes and multiple complications, needing decades of treatment? For a cancer patient, facing years of choices over multiple experimental treatments? For a family, choosing long-term care options for a grandmother with dementia? Most of the expense and problem in our healthcare system involves treatment of chronic conditions or (what turns out to be) end- of-life care, and involve many difficult decisions involving course of treatment, extent of treatment, method of delivery, and so on. These people can shop. Our healthcare system actually does a pretty decent job with heart attacks. And even then . . . have they no families? If I’m on the way to the hospital, I call my wife. She is a heck of a negotiator. Moreover, healthcare is not a spot market, which people think about once, at fi fty-five, when they get a heart attack. It is a long-term relationship. When your car breaks down at the side of the road, you’re in a poor position to negotiate with the tow-truck driver. That is why you join AAA. If you, by virtue of being human, might someday need treatment for a heart attack, might you not purchase health insurance, or at least shop ahead of time for a long-term relationship to your doctor, who will help to arrange hospital care? And what choices really need to be made here? Why are we even talking about “negotiation?” Look at any functional, competitive business. As a matter of fact, roadside car repair and gas stations on interstates are remarkably honest, even though most of their customers meet them once. In a competitive, transparent market, a hospital that routinely overcharged

190

John H. Cochrane

cash customers with heart attacks would be creamed by Yelp.com reviews, to say nothing of lawsuits from angry patients. Life is not a oneshot game. Competition leads to clear posted prices, and businesses anxious to give a reputation for honest and efficient service. So this is not even a realistic situation. To be sure, some conditions really are unexpected and incapacitating. Not everyone has a family. There will be people who are so obtuse they would not get around to thinking about these things even if we were a society that let people die in the gutter, which we are not, and maybe some hospital somewhere would pad someone’s bill a bit. (As if they do not now!) But now we are back to the straw man fallacy. Once again, the idea that ACA is a thoughtful, minimally designed intervention to solve the remaining problem of poor negotiating ability by people with sudden unexpected and debilitating health crises is ludicrous. As is the argument that we should accept the entire ACA because of this issue. Take a closer look at Keller and Davis’s statement: “strapped for time, often poor, sometimes uneducated, confused and afraid,” and “not really in a position to make cost- conscious choices.” We are talking about average Joe and Jane here, sorting through the forms on the insurance offerings to see which one offers better treatment for their multiple sclerosis or diabetes-related complications. If Joe and Jane cannot be trusted to sort through this, how in the world can they be trusted to figure out whether they want a fi xed or variable mortgage? Which cell phone or cable plan to buy? To deal with auto mechanics, contractors, lawyers, and fi nancial planners? How can they be trusted to sign marriage or divorce documents, drive, or . . . vote? We have a name for this state of mind: legal incompetence. Keller, Davis, and company are saying that the majority of Americans, together with their families, are legally incompetent to manage the purchase of health insurance or healthcare. And, by implication, much of anything else. Yes, there are some people who are legally incompetent. But—straw man again—Keller and Davis are not advocating social services for the incompetent. They are defending the ACA, which applies to all of us. So they must think the vast majority of us are incompetent. If not blatant illogic, this is a breathtaking aristocratic paternalism. Noblesse oblige. The poor little peasants cannot possibly be trusted to take care of themselves. We, the bien-pensants who administer the state, must make these decisions for them.

After the ACA: Freeing the Market for Healthcare

191

Let me ask any of you who still agree, does this mean you? When you are faced with cancer, do you really want to place your trust in the government health panel, because they will make better decisions than you, with your doctor and family? Or is this just for the benighted lower classes, and you and I, of course, know how to find a good doctor and work the system? Choice is always between alternatives. Sure, some people make awful decisions. The question is, can the ACA bureaucracy and insurance companies really do better? Yet you would not trust them to buy your shirts? And once again does the entire gargantuan bureaucratic apparatus of the ACA follow, not from the proposition that there is some fundamental economic market failure, but because . . . Americans are no good at shopping? No. Health is not too important to be left to the market. Health is so important—and so varied, so personal, and so subjective—that it must be left to the market. If you do not trust the vast majority of people to make the most important decisions of their lives, and a government bureaucracy can make better decisions on their behalf, you are a devout patrician, not a devout capitalist. E. Theory and Experience I’m often told, “Well, fi ne, but this is just theory. Free-market healthcare has not been tried in a modern economy. All countries regulate healthcare or governments provide it.” That is the point of my extensive examples of other industries. As an economic good, there really is not much difference between healthcare and other complex personal services such as auto repair, legal services, home repair and remodeling, or college education. Yet these markets do not require payment plans styled as “insurance” for “access,” nor must bureaucracies decide what every American “needs.” In all these other industries, the providers also have considerably more expertise than the customer. One hundred fi fty years ago, the United States looked across the ocean and nearly all the governments of the time were monarchies. That observation did not prove that monarchy was a better system. Over and over again, from the guilds against which Adam Smith railed to the telecommunications, trucking, and airlines industries

192

John H. Cochrane

deregulated in our time, people have told us that industries cannot possibly be left to market forces. And time and time again they have been wrong. No country in the world let private markets operate telephones and television stations when we deregulated them, either. The pockets of healthcare that are allowed relatively competitive free entry operate reasonably well. Plastic surgery and dentistry are not disasters. Radial keratotomy (corrective eye surgery) is a good example, as specialization and competition have led to both lower costs and increased quality. I am not the fi rst dog owner to notice how easy and relatively inexpensive cash-and- carry veterinary medicine is compared to the same treatment for humans. Concierge medicine is taking off. So is cash-and- carry medical tourism. If anyone is guilty of theorizing in the face of experience, it would seem to be those who hold the faith that the next round of brilliant ideas for layering on ACA-style regulation will lead fi nally to successful cost control that is not simply rationing, or will induce the radical quality improvement and innovation that we need, whereas the past ones have all failed, over and over again.

Realistic Freedom, Help, and Vouchers I do not require that you follow me to some unrealistic libertarian nirvana. “The unfettered free market” where the improvident die in the gutter is another ridiculous straw man. Southwest’s pilots have FAA licenses. Wal- Mart’s products meet the standards set by the Consumer Product Safety Commission. True-blue libertarians argue about this last 5 percent of deregulation, but we do not have to. A little freedom will go a long way. A. Better Regulation through Transfers and Vouchers In addition to the need for genuine charity care, there can stlll be lots of government help in various places. But a central principle of economics is “don’t transfer income by distorting prices, mandating transfers, or via government-provided services.” The vast majority of any help and transition smoothing can and should be given in the form of on-budget, lump-sum subsidies or vouchers, leaving marginal incentives intact, and avoiding programs, protections, and incentives that last forever.

After the ACA: Freeing the Market for Healthcare

193

When we transition to freely rated lifelong individual insurance, individuals who are already sick face high premiums. That problem is easily solved with a voucher, or a lump-sum payment to their health savings accounts. The same principle applies to genetic diseases. Economics has long recognized the principle that insurance cannot insure events that have already happened, so lump-sum transfers are appropriate. But one-time lump-sum transfers based on clearly defined events over which no one has control, such as a DNA marker, are much less distorting, or subject to abuse, than perpetual regulation and intervention in a market, to “provide care” as “needed.” If we want to subsidize healthcare or insurance for old people, poor people, or veterans, give them a voucher. There is no reason the government should try to run an insurance company or run hospitals in order to provide fi nancial assistance to people it wants to help. Insurance is about money, after all, period. There is no reason for government to pass an implicit tax by mandating that businesses “provide” insurance. If we want to subsidize emergency rooms, let us just do it, on budget. That will be much more efficient than forcing a big cross-subsidy scheme and blocking competition to keep that scheme afloat. Subsidies do not require the competition-smothering protective regulation needed to prop up mandated cross-subsidies. Letting Wal-Mart set up more clinics would be a lot cheaper too. If you think people do not get enough checkups when paying with their own money, give them a voucher. That is much easier than passing a mandate that every company must provide fi rst- dollar health payments with a long range of mandated benefits. More generally, there is an income-based paternalism at work in healthcare policy, somewhat more reasonable than the “they cannot shop” paternalism I decried above, worth making explicit. Most people, when spending their own money at the margin, are likely to choose less healthcare than we, the self-appointed advisers to “policy makers,” would like. Already, they exhibit trade- offs that imply less health than we would like— they drink sugared sodas, eat fast foods, and do not exercise enough. In my example in which patients were offered an MRI or $1,000 in cash, I think we suspect that a lot of patients would choose the cash, and we would prefer that they did not. A true libertarian would say, “Well, let people choose more iPhones and less health if that is what they want.” But we do not have to have

194

John H. Cochrane

this argument. If you think people will spend too little on health overall, give them vouchers in a health-savings account that can only be used for healthcare expenses or insurance. This system maintains the efficiency of patient- driven choice. It distorts the overall health versus nonhealth price so they will choose health-related expenses, but without distorting relative healthcare prices, destroying healthcare competition, or writing ten thousand pages of regulations and supply-side restrictions that gum up the entire system. Now you might object that all these subsidies and vouchers will raise “costs” on the budget. But this happens simply because of phony accounting. If the government mandates that cardiac patients cross-subsidize emergency rooms, this is exactly the same as a tax on cardiac services and an expenditure on emergency rooms. Actually, it is a lot worse because the distortion of the current system is much greater. So any economically relevant accounting would recognize that we save money overall. Fixing the accounting is a much better and cheaper project than keeping our ridiculously inefficient healthcare system. B. “Politically Feasible” “Well,” my typical critic concludes, “maybe you’re right about all this as a matter of economics, but it’s not ‘politically feasible.’ ” No, not now. But the alternative is not economically feasible, and economics is a sterner taskmaster. The ACA is becoming less and less politically feasible by the day as well, and inevitable scandals will not help it. What was not feasible today can quickly become feasible tomorrow if it is correct—once people understand it, once people see the alternative fall apart, and once people realize there is no option. Our job as economists is to figure out what works and explain it, not to bend reality to some notion of what today’s politicians are willing to say in public, or to hire us as advisers to defend for them. The “politically feasible” conversation is truly lunatic. It is taken for granted in policy discussion that no American can be asked to “pay” (directly, rather than through taxes or cross-subsidies) for one cent of health cost risk, while they routinely pay for broken and crashed cars and destroyed houses, suffer huge risks in the job market, and shoulder housing, transport, and other expenses much greater than the cost of healthcare. Yet while we are pretending nobody should pay for things, unfortunates

After the ACA: Freeing the Market for Healthcare

195

who fall through the cracks can be handed ridiculous $550,000 bills for cancer treatment. We can start by saying, out loud, that healthcare is an economic good like any other. It is okay to ask Americans to pay for it, and to allow American companies to competitively supply it, just like all the other goods and services we routinely purchase. It is okay for insurance to retreat to its proper role, that of protecting people from large shocks to wealth, rather than that of a hugely inefficient payment plan. Car insurance does not pay your oil changes after you fax in the forms in quintuplicate, obtain permission from your mechanic, go to the in-network mechanic, wait six weeks, and answer a twenty-page questionnaire about your repair history and driving habits. It is okay for Americans to bear small risks of expenditure in healthcare as they do in everything else.

Conclusion Healthcare is a complex personal service, with wide variation in quality, both along measures of health outcomes and along more subjective dimensions of satisfaction. Its demand curve is very elastic—people will consume a lot at subsidized prices. The distinction between “want” and “need” is conceptually fuzzy, and nearly impossible to measure. The big improvements in healthcare come from better technology. But big improvements in healthcare delivery, average quality, and cost are also attainable. The latter come from much better human organization, as has happened recently in many other industries that have witnessed revolutionary supply competition. Achieving those improvements requires that newcomers can sell products at a profit and enter new markets, while displacing lots of entrenched interests before facing competition themselves. From these observations, simple conclusions follow. Health care markets need a big supply-side revolution, in which the healthcare equivalents of Southwest Airlines, Wal-Mart, and Apple enter, improving business practices, increasing quality and transparency, and spurring innovation. And disrupting the many entrenched interests and cross-subsidies of the current system. I outlined a long string of restrictions on competition that must be repealed or modified to allow competition. At a minimum, every new

196

John H. Cochrane

regulation should be evaluated by its effect on competition by new entrants or protection of incumbents, a consideration not even spoken of in policy discussion today. (Even when regulatory cost-benefit calculations are made, they do not consider the effects of regulation on competition, capture, and cronyism.) Healthcare is singularly ill suited to payment-plan provision, either by government directly or by heavily regulated insurance issued by a few large well-protected businesses. A functional cash market must exist in which patients can realistically feel the marginal dollar cost of their treatment or (equivalently) enjoy the full fi nancial benefits of any economies of treatment they are willing to accept, and are not patsies for huge cross-subsidization and rent seeking by an obscure system negotiated behind the scenes between big insurance companies, hospitals, and government. Both supply and demand must be freed, along with insurance. Without supply competition, asking consumers to pay more will do little to spur efficiency. Without demand competition, new suppliers will not be able to succeed. So the alternative to the current healthcare and health insurance mess (both pre- and post-ACA) is clear. Getting there will be a long hard road. It is not a simple matter of “deregulation,” given how deep and widespread the offending restrictions are, and the many legitimate purposes they purport to serve, and sometimes do. We need to construct a different but wiser legal and regulatory regime. I know an interest group when I see one: do not worry, there will be lots of jobs for health economists, policy analysts, and lawyers. The alternative, doubling down regulations on an already highly regulated system, full of protected and politically connected incumbents and rent seekers, has little chance of achieving these goals. Whether in the post- office model (government provision) or the 1950s-style regulated airline, utility, or bank model (the ACA), this effort will just produce less efficiency, more costs, and another generation of bright ideas dashed. Reformers, remember that the last twenty bright ideas did not fail simply because the people in charge were not as smart as you are, or as well meaning. There are some bright spots. As Uber has undermined taxi regulation by swiftly implementing a better system and creating a large enough interest group of consumers unwilling to be taken advantage of by taxi

After the ACA: Freeing the Market for Healthcare

197

regulations, so Internet-based startups are undermining many aspects of healthcare, including obscure hospital and pharmacy pricing and obscure quality. Many ACA exchange policies have large co-payments, and as more workers are thrown on to exchanges, a critical mass of pricesensitive consumers who are also voters may demand change as it is being supplied.

Notes I am grateful to Anup Malani and W. E. Heasley for extensive and very helpful comments and gratefully acknowledge research support from the Center for Research in Securities Prices and from the Guggenheim Foundation. 1. An excellent example: Richard A.Epstein and David A. Hyman, “Fixing Obamacare: The Virtues of Choice, Competition and Deregulation,” Annual Survey of American Law 68 (2013): 493, http://ssrn.com/abstract=1158547. Epstein and Hyman end up calling for a much more free-market system, as I do, but build their case from the failures of the ACA and current regulation. 2. If personal experience is not enough to remind you how inefficient the current system is, I recommend Jonathan Rauch’s YouTube video, “If Air Travel Worked Like Health Care,” YouTube video, 7:00, uploaded by TheNewAltons, January 4, 2010, http://www.youtube.com/watch?v=5J67xJKpB6c. Hat tip to Einer Elhauge, who showed it at the conference. 3. Atul Gawande, “Big Med: Restaurant Chains Have Managed to Combine Quality Control, Cost Control, and Innovation. Can Health Care?” New Yorker, August 13, 2012, http://www.newyorker.com/reporting/2012/08/13/120813fa_fact_gawande. 4. Amitabh Chandra and Jonathan Skinner, “Technology Growth and Expenditure Growth in Health Care,” Journal of Economic Literature 50 (September 2012): 643 – 80. 5. Chester Dawson, “For Datsun revival, Nissan Gambles on $3,000 Model, Wall Street Journal, October 1, 2012, http://online.wsj.com/article/SB1000087239 6390443890304578009284279919750.html. 6. Jaime A. Rosenthal, Xin Lu, and Peter Cram, “Availability of Consumer Prices From US Hospitals for a Common Surgical Procedure,” JAMA Internal Medicine 173, no. 6 (2013): 427– 32 doi:10.1001/jamainternmed.2013.460. In a hilarious follow-up, Jillian Bernstein and Joseph Bernstein called twenty Philadelphia hospitals to inquire about the cost of a simple ECG and the cost to park at the hospital to obtain the ECG. Only three hospitals were able to provide the cost of an ECG, while nineteen were able to provide the cost of parking. Of these, ten offered free or discounted parking. “This demonstrates not only that hospitals are able to provide cost information by telephone but, we infer, that

198

John H. Cochrane

they can respond to consumers’ concern about cost.” And “hospitals seem able to provide prices when they want to.” Jillian R. H. Bernstein and Joseph Bernstein, “Availability of Consumer Prices From Philadelphia Area Hospitals for Common Services: Electrocardiograms vs Parking,” 292 JAMA Internal Medicine 174 (2) (2014): 292. http://jamanetwork.com/data/Journals/INTEMED/929736 /ild130153.pdf. 7. In a literature review, the Civitas Institute writes: “One hospital industry respondent to a National Institute for Healthcare Reform Study reported “member hospitals initially had mixed views about the benefits of CON but banded together to support the process after realizing it was a valuable tool to block new physician-owned facilities.” “Certificate of Need: Does It Actually Control Healthcare Costs?” http://www.nccivitas.org/2011/certificate-of-need-does-it-actually -control-healthcare-costs/. Innovation and competition are thus stifled in order to continue the profitability of existing healthcare providers. Physicians and multiphysician groups fi nd it harder to open and operate ambulatory surgery centers, freestanding radiology practices, and other facilities that would allow consumers to enjoy healthcare that is potentially both lower cost and higher quality. The Washington State Certificate of Need website, http://www.doh.wa.gov/Li censesPermitsandCertificates/FacilitiesNewReneworUpdate/CertificateofNeed. aspx, makes fun browsing. The “methodology” sets out numerical targets for facilities in “planning areas.” Thus, the idea of building an “unneeded” facility simply because you can do it better and cheaper than an incumbent is explicitly prohibited. In North Carolina, “All new hospitals, psychiatric facilities, chemical dependency treatment facilities, nursing home facilities, adult care homes, kidney disease treatment centers, intermediate care facilities for mentally retarded, rehabilitation facilities, home health agencies, hospices, diagnostic centers, and ambulatory surgical facilities must fi rst obtain a CON before initiating development. In addition, a CON is required before any upgrading or expansion of existing health service facilities or services, which involves a capital expenditure above specified minimums.” “Certifi cate of Need,” North Carolina Division of Health Service Regulation, http://www.ncdhhs.gov/dhsr /coneed/index.html, 8. Institute for Justice, “CON JOB: How A Virginia Law Enriches Established Businesses by Limiting Your Medical Options, and How IJ Is Going to Stop It” 2012, http://www.ij.org/vacon; Michael Tennant, “The Big Health Care Con,” June 28, 2102, http://fff.org/explore-freedom/article/the-big-health-care-con /gives a short CON review and covers the Virginia case. 9. Uwe E. Reinhardt, “The Dubious Case for Professional Licensing,” Economix: Explaining the Science of Everyday Life, October 11, 2013, http://economix .blogs.nytimes.com/2013/10/11/the-dubious-case-for-professional-licensing.

After the ACA: Freeing the Market for Healthcare

199

10. Einer Elhauge, ed., The Fragmentation of U.S. Health Care— Causes and Solutions (Oxford: Oxford University Press, 2010). These quotes from the introductory chapter are available at http://www.law.harvard.edu/faculty/elhauge /pd f / El hauge%2 0T he%2 0Frag mentation%2 0 of%2 0US%2 0Hea lth%2 0 Care%20--%20Introductory%20Chpt.pdf. 11. Einer Elhauge, ed., The Fragmentation of U.S. Health Care—Causes and Solutions (Oxford: Oxford University Press, 2010), 11. 12. Ibid., 12. 13. American Health Lawyers Association, “Corporate Practice of Medicine,” www.healthlawyers.org/hlresources/Health%20Law%20Wiki/Corporate%20 Practice%20of%20Medicine.aspx. 14. Darius Lakdawalla and Tomas Philipson, “Non- Profit Production and Industry Performance,” Journal of Public Economics 90, no. 9 (2006), 1681– 98. 15. Eugene F. Fama and Michael Jensen, “Agency Problems and Residual Claims,” Journal of Law and Economics 26 (1983), 327– 49. Fama and Jensen also view the presence of donors on boards of directors as an imperfect substitute for knowledgeable insiders and corporate- control market discipline. 16. For a description of the process, with a view, however, that it needs more not less regulation, see Jill R. Horwitz, “State Oversight of Hospital Conversions: Preserving Trust or Protecting Health?” (working paper, Hauser Center for Nonprofit Organizations, Kennedy School of Government, Harvard University, 2012), http://www.hks.harvard.edu/hauser/PDF_XLS/workingpapers/work ingpaper_10.pdf. 17. For an example of recent news coverage see “Regulators Seek to Cool Hospital-Deal Fever,” Wall Street Journal, March 18, 2012. 18. Gregory Warner, “The Battle over Billing Codes,” Marketplace, April 10, 2012, http://www.marketplace.org/topics/life/health-care/battle-over-billing-codes. 19. David M. Cutler and Dan P. Ly, “The (Paper)Work of Medicine: Understanding International Medical Costs,” Journal of Economic Perspectives 25, no. 2 (2011): 8, http://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.25.2.3. 20. At the conference, Meredith Rosenthal gave a wonderful presentation highlighting a wide range of complex payment schemes and how they didn’t work out, a wider range of bright new ideas, and how little we know about how they work. Her conclusion was that lots more research will lead to something workable to patch up each leak. Mine was that jiggering health payment systems is the best modern example of the hopelessness of central planning. You can get some idea from Meredith B. Rosenthal, “What Works in Market- Oriented Health Policy?” New England Journal of Medicine 360 (May 21, 2009):,2157– 60. http://www.nejm.org/doi /full/10.1056/NEJMp0903166. See especially the table in Meredith B. Rosenthal, “Beyond Pay for Performance—Emerging Models of Provider-Payment Reform,”

200

John H. Cochrane

New England Journal of Medicine 359 (September 18, 2008): 1197–1200, http:// www.nejm.org/doi/full/10.1056/NEJMp0804658. 21. For example, http://www.personalinjurylawupdate.com/damorelaw/2012/04 /what-is-outsourced-radiology.html complete with a link, “Read the tragic story of a now brain-damaged young woman who had 2 sets of x-rays - yet no one diagnosed her brain abscess,” and “Outsourcing radiology abdicates 3 of 4 of the core responsibilities of radiologists.” 22. Ezekiel J. Emanuel, Neera Tanden, and Donald Berwick, “The Democrats’ Market-Friendly Health- Care Alternative,” Wall Street Journal, September 25, 2102, http://online.wsj.com/article/SB1000087239639044401750457764519 3107383610.html. 23. Edward P. Lazear, Prices and Wages in Transition Economies” (Stanford: Hoover Institution Press, 1992), 16. 24. Nicholas D. Krist of, “A Possibly Fatal Mistake,” New York Times, October 12, 2012, http://www.nytimes.com/2012/10/14/opinion/sunday/kristof-a-possi bly-fatal-mistake.html?ref=healthcare. 25. “Health- Status Insurance,” Cato Institute, Policy Analysis No 633 (2009); “Time- Consistent Health Insurance,” Journal of Political Economy 103, no. 3 (1995): 445– 73; “What to Do about Pre-Existing Conditions,” Wall Street Journal, August 14, 2009; “Forget about the Mandate,” Bloomberg View, July 12, 2012; “What to Do on the Day after Obamacare,” Wall Street Journal, April 2, 2012; “The Real Trouble with the Birth- Control Mandate,” Wall Street Journal, February 9, 2012; all available at http://faculty.chicagobooth.edu/john .cochrane/research/news.htm#health, except for “Time- Consistent Health Insurance,” available at http://faculty.chicagobooth.edu/john.cochrane/research /index.htm#health. 26. Terhune, Chad, “Some Health Insurance Gets Pricier as Obamacare Rolls Out,” Los Angeles Times, October 26, 2013, http://www.latimes.com /business/la-fi-health-sticker-shock-20131027-story.html. The article also states, “All these cancellations were prompted by a requirement from Covered California, the state’s new insurance exchange. The state didn’t want to give insurance companies the opportunity to hold on to the healthiest patients for up to a year, keeping them out of the larger risk pool that will influence future rates.” 27. The Misericordia charitable hospital was founded in Florence, Italy, in 1244, http://www.misericordia.fi renze.it/Home/ChiSiamo. 28. Akerlof, George A., “The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism,” Quarterly Journal of Economics 84, no. 3 (1970): 488- 500. 29. These quotes are from commenters on my blog, not a very authoritative source, but they put the view so clearly I couldn’t resist. Grumpy Economist, The. http://johnhcochrane.blogspot.com/search/label/Health%20economics.

After the ACA: Freeing the Market for Healthcare

201

30. Ezra Klein, “Why an MRI Costs $1,080 in American and $280 in France,” Washington Post, March 3, 2012, http://www.washingtonpost.com/blogs/ezra-klein /post/why-an-mri-costs-1080-in-america-and-280-in-france/2011/08/25/gIQAVHz toR_blog.html. 31. Bill Keller, “Five Obamacare Myths,” New York Times, July 15 2012, http://www.nytimes.com/2012/07/16/opinion/keller-five-obamacare-myths.html.

Chapter eight

Obamacare and the Theory of the Firm Einer Elhauge

T

he fragmented nature of the U.S. healthcare system is quite anomalous. A hilarious video on YouTube makes the point well, asking what it would be like if air travel worked like healthcare.1 In this alternative world, we see an unfortunate consumer trying to book a crosscountry fl ight to Oregon. He discovers he needs to book separately with an East Coast specialist to get him to Chicago and a West Coast specialist to get him to Oregon, then book a separate baggage specialist and fuelist for each leg, all of whom bill separately and require their own paperwork, none of whom publicly post their rates, with the whole uncoordinated mess resulting in an astronomical cost to travel on a day different than when he wants to fly. You cannot watch the video without thinking “thank heavens we do not live in that world.” But for U.S. healthcare, we do live in that world. Even the physicians who practice in a hospital are typically independent from each other and from the hospital and its nurses. If you are lucky, the hospital will have a case manager to try to coordinate all these actors, but the case manager will have a hard time getting the physicians to pay attention because they are paid separately and the hospital depends on the admitting physician for business. Outside of hospitals, the situation is even worse. The average Medicare patient sees seven doctors a year, ten if the patient has a chronic condition, and no one is paid to coordinate them. The evidence indicates, as I show in Part I, that this fragmentation raises costs and worsens health outcomes. Further, as Part II explains, the eco-

Obamacare and the Theory of the Firm

203

nomic theory of the firm suggests that allowing greater integration should improve these results. The problem, Part III demonstrates, is that current law stands in the way. Regulatory laws restrict how hospitals and physicians can work together, while payment laws require disaggregated payments for specific services. This legal framework today prevents healthcare markets from reaching the optimal level of healthcare integration, instead favoring fragmentation of healthcare provision and payment.2 The good, and perhaps surprising, news is that Obamacare might well provide the solution to this problem. More formally known as the Patient Protection and Affordable Care Act, 3 Obamacare contains a number of provisions that could, Part IV shows, lift current legal obstacles to efficient healthcare integration. All we need is appropriate implementing regulations. This approach is probably the least painful way to lower healthcare costs because it actually increases quality. It should also have bipartisan appeal, because it would use provisions of Obamacare to adopt the sort of deregulatory reforms that generally appeal to Republicans but should also appeal to Democrats because they will likely be necessary to make Obamacare a success. Best of all, it can be done through executive action, thus sparing us the agony of trying to pass another healthcare statute.

How Fragmentation Raises Costs and Hurts Patients Fragmentation within hospitals. Fragmentation at the level of a single illness can occur when there is a failure to coordinate among the various providers with whom a patient interacts during a single hospital visit.4 Attending physicians in a hospital are typically independent of each other and of the hospital. Further, hospital administrators have no direct control over physician decisions. Nor do they have much fi nancial leverage because attending physicians typically bill separately. Indeed, the fi nancial incentives tend to run the other way because physicians are usually the primary source of the hospital’s business. While a dedicated case manager (when provided by a hospital) can help prevent some failures of communication, case management does not give the hospital actual control, nor does it change the fi nancial incentives of either the doctors or the hospital. To illustrate how fragmentation can impair healthcare even in a worldleading hospital, consider a recent article’s account of the organization of

Einer Elhauge

204

surgical instruments at Stanford University Hospital. 5 Surgeons at the hospital, each of whom was an independent contractor who gets a fee for each surgery, indicated the instruments they required by submitting preference cards. Technicians, who are hospital employees, were responsible for loading the requested supplies onto a cart to follow the varying physician specifications. Under this system, errors could occur in fi lling out the cards, loading the supplies, mislabeling bins, or a host of other possibilities. Physicians had no direct contact with these technicians, so they tended to blame the nurses when errors occurred even though the nurses had nothing to do with loading the instruments. These failures led to potentially unsafe practices, such as nurses keeping instruments in lockers, doctors taking instruments home with them, and flash sterilizations of instruments rather than the preferred six-hour sterilization process. Neither this situation nor this article’s assessment of the problem is idiosyncratic. They are perfectly in line with reports by the Institute of Medicine, the highly influential medical branch of the National Academy of Sciences that offers independent evidence-based advice on health policy. The Institute has concluded that similar problems exist throughout the system because it is fragmented and focuses on “professional prerogatives and separate roles” rather than on “cooperation and teamwork.”6 As a result, [p]atients and families commonly report that caregivers appear not to coordinate their work, or even to know what others are doing. Suboptimization is seen, for example, in operating rooms that must maintain multiple different surgical tray setups for different doctors performing the same procedure. Each doctor gets what he or she wants, but at the cost of introducing enormous complexity and possible error into the system.” 7

All this would be less worrisome if medical errors in hospitals were not a serious problem, but they are. According to the Institute of Medicine, preventable medical errors in hospitals result annually in forty-four thousand to ninety-eight thousand deaths and cost between $17 billion and $29 billion.8 The Institute concludes that “[t]he decentralized and fragmented nature of the healthcare delivery system . . . contributes to unsafe conditions for patients, and serves as an impediment to efforts to improve safety.”9 In short, medical errors in hospitals annually cause tens of thousands of deaths and billions of dollars in costs, and healthcare fragmentation causes many of these medical errors. The Institute adds:

Obamacare and the Theory of the Firm

205

A highly fragmented delivery system that largely lacks even rudimentary clinical information capabilities results in poorly designed care processes characterized by unnecessary duplication of services and long waiting times and delays. And there is substantial evidence documenting overuse of many services—services for which the potential risk of harm outweighs the potential benefits.  .  .  . [P]atients tell stories of fragmented care in which relevant information is lost, overlooked, or ignored; of wasted resources; of frustrated efforts to obtain timely access to services; and of lost opportunities. When clinicians and their families and those steeped in health management become patients, they, too, fi nd that there appears to be no one who can make the systems function safely and effectively.10

Fragmentation across providers. Fragmentation in the care provided to a single patient can also occur when there is a failure to coordinate between different providers treating different conditions or even different aspects of the same condition. The typical Medicare beneficiary sees two primary- care and five specialist physicians a year; those with a chronic disease such as coronary artery disease see ten physicians annually, on average.11 Worse, as Professor David Hyman notes, each physician is “focused on the discrete symptoms and/or body parts within their jurisdiction.”12 Medical histories from other providers are often unavailable or distrusted, leading to imaging studies or laboratory tests being unnecessarily repeated.13 Payment structures contribute to this disjointedness. Neither Medicare nor other insurers pay physicians to coordinate care.14 To the contrary, because providers are paid separately for the amount of care they provide, coordination that solves medical problems more effectively would reduce the need for provider services and thus reduce their revenue.15 The perverse result is that “providers can actually do better if their patients do worse.”16 These conclusions again comport with the assessment of the renowned Institute of Medicine, which concludes that “physician groups, hospitals, and other healthcare organizations operate as silos, often providing care without the benefit of complete information about the patient’s condition, medical history, services provided in other settings, or medications prescribed by other clinicians.”17 More generally, the Institute states: “Today’s health care system is not well designed to meet the needs of patients with common chronic conditions. . . . For too many . . . care for even a single condition is fragmented across many clinicians and settings

206

Einer Elhauge

with little coordination or communication, and some needs remain undetected and/or unmet.”18 One striking empirical study directly addressed whether having more physicians worsens care. It studied the outcome differences for Medicare patients after a heart attack depending on whether they saw a relatively low number of physicians (4.8 physicians per patient) or a relatively high number (9.2 physicians per patient).19 It found that having more physicians involved not only increased patient costs by $3,331 (with a 99.9 percent statistical level of confidence), but also reduced their odds of surviving by 2.5 percent (with a 94.0 percent statistical level of confidence).20 So seeing more physicians not only increases costs, but contrary to common intuition worsens medical outcomes, here increasing the odds of death by 2.5 percent. This sort of fragmentation also helps explain why, for a nation that spends so much money on healthcare, our overall metrics are so unimpressive. Of course, it is well known that measures of U.S. health are worse than developed nations that spend much less on healthcare. But because those sorts of statistics are subject to the objection that this difference may reflect our diet or lifestyle, consider a simpler metric: What percentage of us get medically recommended levels of preventive care, or when we have a chronic illness, receive the recommended treatments for it? It turns out that the answers are only 55 percent on the fi rst question and only 56 percent on the second. 21 That is remarkably low, and it seems likely that part of the explanation is that ensuring that recommended care is provided often falls through the cracks in our fragmented system, where lines of responsibility are unclear. U.S. healthcare fragmentation also produces outsize administrative costs. As of 1999, healthcare administration cost $1,059 per person in the United States, which was not only a remarkable 31 percent of total U.S. healthcare costs but also more than triple the $307 per person administrative cost in Canada.22 Part of this reflects the fact that U.S. health insurers are far more fragmented than Canada’s nationalized health insurance system; insurance overhead costs $259 per person in the United States versus $47 per person in Canada.23 But much of the difference reflects the sort of fragmentation in healthcare provision at issue in this article. For hospitals, administrative costs per person are $315 in the United States versus $103 in Canada. For practitioners, administrative costs per person are $324 in the United States versus $107 in Canada. 24 In short, the average administrative costs per U.S. patient, hospital, and doctor are each triple those of their Canadian counterparts.

Obamacare and the Theory of the Firm

207

The Theory of the Firm Although our healthcare system clearly seems excessively fragmented, that does not mean all integration is good. Well-functioning markets always feature some mixture of integration and disintegration. After all, buying airplane tickets may allow travelers to avoid choosing and coordinating pilots, planes, fl ight attendants, ticket agents, baggage handlers, and mechanics, but airlines do not also provide our taxicab to the airport or our hotel when we arrive. Moreover, the optimal level of integration often changes over time with changing technologies or economics. At one time, people bought cars without wipers or bumpers and selected those separately. 25 Airlines themselves now often charge separately for food, leading many passengers to buy their food before boarding from other fi rms, thus suggesting that air travel is becoming disaggregated from airplane food. Of course, airplane food is so bad that it became a comedic cliché, but other airline services also might become disaggregated. For example, airlines now frequently charge for baggage, and some separate fi rms are beginning to offer the service of taking your baggage directly from your home to your destination. While the latter service currently seems priced more at luxury levels, one could easily imagine the economics changing so that the transport of humans and their baggage became efficiently disintegrated in the future. Indeed, even now, some beneficial changes in healthcare organization may involve disintegration. For example, retail health clinics have separated some routine healthcare from other healthcare. But getting this routine care while shopping could be beneficial, especially if the lower costs, greater convenience, and decreased delay result in patients getting medically beneficial healthcare more regularly and on time. 26 The economic theory of the fi rm explains how markets efficiently determine what activities to integrate into fi rms rather than leave outside them. As Professor Ronald Coase fi rst pointed out, a defi ning characteristic of business fi rms is that they use centralized control rather than internal markets to allocate and coordinate resources. 27 They will fi nd this profitable only when centralized control provides an efficiency advantage over decentralized market transactions. Professors Armen Alchian and Harold Demsetz then showed that the major efficiency advantage firms have is that centralized control can mitigate the incentives to shirk that characterize a market system when it is hard for the market

208

Einer Elhauge

to measure and reward each individual’s contribution to team production that requires joint effort. 28 Firms solve these inefficiencies by having a single owner that can both “(1) select, direct, monitor, and reward or punish team members based on their contributions to the joint product and (2) ha[ve] a residual claim to any profits on the sale of the joint product that are left after all the team members are paid.” 29 A residual claim to profits, coupled with the ability to monitor and control the inputs, is important in giving the owner both the ability and incentives to coordinate most efficiently—minimizing shirking by individual team members and ensuring that the joint product is maximally profitable. Team production may not have been as important in healthcare decades ago, but has become vital to modern healthcare. Physicians, hospitals, nurses, drugs, devices, tests, technicians, and other inputs must be combined to produce the joint result of health. Yet it is difficult to determine the contribution of each participant to the joint result without close observation. In healthcare, “shirking” generally does not take the form of failing to work—people in healthcare tend to work remarkably hard. Instead, shirking usually consists of failing to coordinate with other providers who also affect the same patients’ health on strategy, timing, and information in a way that maximizes health benefits and minimize costs. Thus, where providers are able to shirk in this way, greater coordination would be beneficial. However, the right level of coordination will vary across different areas of healthcare, depending on where direct observation and shirking are more or less likely, and will likely change with changing technology. Unfortunately, U.S. healthcare refuses to adopt centralized ownership structures to deal with this team-production problem. A single hospital stay requires treatment by multiple physicians, each of whom is independent of the others and the hospital. The hospital cannot direct or monitor the medical decisions of the physicians, and because the hospital does not pay the physicians, it cannot leverage payments to influence them. In any case, the hospital has insufficient incentive to coordinate because it is not a residual claimant that stands to gain profit by coordinating physicians. The medical staff, which can review the decisions of other physicians, similarly lacks the incentive that a residual claimant would have to control physician decision making. Beyond a single hospital stay, the problems multiply: each physician bills for her own services and no one receives payment to coordinate among the providers. The most obvious

Obamacare and the Theory of the Firm

209

choices to coordinate care, either the primary- care physician or the insurer, are not residual claimants and have little incentive to serve in such a coordinating function. In any case, both the primary- care physician and the insurer lack the power to direct the decisions of other providers, even if doing so would lower costs and improve care. True, medical providers sometimes coordinate in heroic ways to ameliorate this problem. But even then the system fails because the payment system rewards each participant for the amount of care they provide. The system does not pay a residual claimant for the value of the care, which would create the normal fi rm profit incentive to increase value and minimize the costs of providing that value. For example, Duke University Hospital once adopted an integrated program to treat congestive heart failure. The program reduced costs by approximately 40 percent by improving outcomes and lowering hospital admissions. But while the program was a resounding medical success, it was a business failure. By reducing the health problems it could bill to treat, Duke actually lost money. 30 It is admirable how often we see medical institutions attempt these sorts of Herculean efforts to coordinate in ways that improve medical outcomes, but we are unlikely to see widespread adoption of such efforts if our payment system continues to penalize them fi nancially.

How Current Law Mandates Fragmentation So we have strong evidence that fragmentation worsens medical outcomes and costs, as well as sound economic theory that greater integration could alleviate those problems. Why, then, have we not seen healthcare institutions actually integrate in way that solve these fragmentation problems? After all, calls to address healthcare fragmentation are not new, and efforts to institute organizational change have been made. They have just not been successful. The Institute of Medicine observes: What is perhaps most disturbing is the absence of real progress toward restructuring health care systems to address both quality and cost concerns, or toward applying advances in information technology to improve administrative and clinical processes.  .  .  . Mergers, acquisitions, and affi liations have been commonplace within the health plan, hospital, and physician practice sectors. Yet all this organizational turmoil has resulted in little change in the way health care is delivered. 31

Einer Elhauge

210

Why has organizational consolidation produced so little improvement? Part of the problem is that organizations adopt the permissible forms of integration that are the most profitable, and our payment system rewards fragmentation rather than medical efficiency. As the Institute put it: The current payment system often reinforces fragmentation by paying separately according to the setting of care and provider type, and by not giving providers the flexibility needed to customize care for individual patients. . . . Furthermore, the fragmentation of payment by service can make it difficult for care to be coordinated efficiently across multiple settings. There is a misalignment among what the patient needs, the services provided, and how needed services are paid for. 32

The Institute thus recommends that “purchasers and health plans  .  .  . should eliminate or modify payment practices that fragment the care system.”33 But that shifts the question to a new level: Why haven’t institutions changed payment practices to encourage more efficient medical organization? The answer is simple. Current law gets in the way of private efforts to reform both organization and payment structure. This has stymied adoption of the Institute’s recommendations on both fronts. On organization, the law inhibits the development of fi rms that control the provision of care and have the profit motive of a residual claimant. The law does so through various legal obstacles to prevent corporations from controlling physicians or charging for medical services. The “corporate practice of medicine” doctrine prohibits the employment of physicians by corporations (including hospitals) on the ground that corporations cannot hold a medical license and thus cannot practice medicine or charge for medical services. 34 Even in states that create exceptions to this doctrine to allow hospitals to hire physicians as employees, the doctrine limits the ability of the hospital to control the decisions of the physician.35 Tort law provides a further disincentive by imposing liability on firms that interfere with the medical judgments of individual physicians. 36 Accreditation standards and sometimes licensing laws mandate that hospitals adopt bylaws that leave the medical staff in charge of medical decisions. 37 Medicare also reinforces physician autonomy by requiring physicians to certify the medical need for the services that they render, 38

Obamacare and the Theory of the Firm

211

and by prohibiting federal officials from supervising the practice of medicine or selecting some providers over others. 39 On payment structure, the law requires a separation of payments that effectively prohibits integrated payments to fi rms that can serve as a residual claimant that would orchestrate all the providers necessary to jointly produce some health outcome. The law does so by generally requiring separate payments for hospitalization, physician services, drugs, and outpatient services that must go directly to each provider. Medicare explicitly separates payments for hospitals (Part A) from those for physicians (Part B) and those for pharmaceuticals (Part D).40 Each of these programs is separate, focused only on services performed, often with Medicare paying different amounts for the same thing depending on who performed the service.41 Medicare thus bars a fi rm from charging for everything necessary to treat a specific illness. Further, Medicare does not reimburse for the coordination of care or case management.42 Because Medicare is the biggest source of hospital revenue, typically providing 35– 55 percent of the money hospitals receive, hospitals cannot afford to organize themselves in a way that does not comply with Medicare. And because hospital care is generally an important part of integrated care, other organizations cannot afford to do so either. Even if they wanted to do so, separate payments are generally required by fee-splitting statutes that prohibit splitting fees either to induce treatment (referral fees) or deter treatment (antireferral fees).43 Not only does Medicare have a general bar on both types of fees,44 but many states directly criminalize referral fees or specify that referral fees are grounds for the suspension or revocation of a physician’s license.45 Such bans on fee splitting not only constrain efforts to substitute incentive payments for direct control, but also effectively prevent fi rms from being the residual claimant on profits earned from medical services. Thus, although healthcare has seen many mergers and organizational changes, these laws have constrained vertical mergers or consolidations that integrate complementary inputs into team production in a way that produces the kind of efficient integration we see in industries like airlines. Instead, the mergers and consolidations we have seen in healthcare have tended to involve horizontal combinations of competing services, because combining hospitals or physician practices does not violate the fragmenting role divisions required by these laws. To be sure, antitrust scrutiny remains available to check horizontal combinations, but the enforcement agencies have generally lost cases challenging hospital mergers,

212

Einer Elhauge

in part because of the intuition that some integration would be helpful. The perverse upshot is that in healthcare the laws have posed a much greater barrier to vertical combinations that could efficiently reduce fragmentation than to horizontal combinations that increase market power in a way that worsens efficiency. This is precisely the opposite of what prevails in other industries, where antitrust is generally the operative constraint and imposes much tighter limits on horizontal combinations than on vertical ones. Together, these regulatory and payment laws limit organizational and payment innovation to protect a form of individual physician autonomy that once made a great deal of sense, when medical care was largely provided by a single physician to the patient with minimal equipment. But protecting individual physician autonomy makes little sense in the modern world where many medical treatments require intricate teamwork and expensive equipment. To be sure, sometimes laws allow a specific form of integration, like HMOs. But such laws do not allow corporations to pick whatever level of integration is most efficient to achieve a valuable result. Moreover, the forms of integration that are allowed have problems. HMOs, for example, are not paid for the value of the care they provide but rather receive a fi xed annual fee per insured. This means that HMOs are not a residual claimant that receives payment for achieving a particular valued result and has incentives to pay team members to achieve that result with maximum efficiency. Instead, HMO profits are the difference between their flat fee and the cost of the care they provide, which provides an incentive to undercare, even if that worsens health outcomes.46 This incentive to undercare has limited their market appeal. Further, to offset this incentive to undercare, HMOs are subject to laws that restrict their ability to control their physicians. The result may be better than fee-for-service medicine, but the combination of flawed incentives with imperfect control is far from optimal, and it means that HMOs do not combine the control and incentive structure required by the theory of the fi rm. Regardless of whether HMOs are better than the traditional fee-forservice model, they provide an important lesson. Instead of dictating a particular form of integration, the law should be neutral as to the appropriate level, without restricting forms of integration that may be efficient. The experience with HMOs demonstrates how mandating a specified type of integration may provide inappropriate incentives, which must then be tinkered with on an ad hoc basis. The law should instead allow

Obamacare and the Theory of the Firm

213

market forces to determine optimal levels of integration (much as the market does for air travel), and focus on fashioning payment and liability systems to give fi rms incentives to choose whatever method optimizes team production by medical professionals. Only in this way can the law encourage those in the market to innovate in a way that develops efficient systems that balance high- quality delivery with low- cost provision.

How Obamacare Can Help Can Obamacare help address the fragmentation problem? Surprisingly, yes. Although public attention has focused on other controversial aspects of Obamacare,47 it also contains a number of provisions that create regulatory authority to address fragmentation in healthcare. For policy insiders, the most well-known of these are the provisions that allow for the creation of Accountable Care Organizations (ACOs), which can coordinate care, and, if they meet quality performance standards, receive a share of savings that they can in turn distribute among providers. These provisions allow groups of physicians and hospitals with “shared governance” to participate as ACOs.48 Such groups must “be willing to become accountable for the quality, cost, and overall care” of the patients that join the ACO.49 Further requirements specify, among other things, that ACOs must join the program for at least three years, must have a minimum size in terms of patients assigned to the ACO, and must meet quality and reporting thresholds. 50 Obamacare then provides for a “shared savings program.”51 Should the ACO’s average per capita costs (including hospital and physician payments under Medicare Parts A and B) fall below a benchmark set by regulation, the ACO will be eligible to receive payments that equal a share of those savings, which they can distribute among the providers belonging to the ACO. 52 Eligibility also depends on meeting quality benchmarks and not taking affi rmative steps to avoid higher-risk patients. 53 Unfortunately, the regulations implementing the ACO provisions so far continue to separate Medicare payments to each hospital and provider. 54 For reasons discussed above, such separate payments provide an incentive for hospitals and physicians to increase care. This incentive to overcare (and receive the full price for any services provided) can easily override the counterincentive created by shared savings payments, which

Einer Elhauge

214

give ACOs only a fraction of the savings from cutting this care that they then have to split among the participating hospitals and physicians. To be sure, we are not limited to these initial implementing regulations. The statutory provisions allow the future adoption of regulations that could change the separate payment model itself in a way that eliminates this obstacle to efficient integration. 55 But it seems less likely that the ACO provisions would allow regulations that remove legal obstacles to firm control over physicians. Thus, the ACO provisions are unlikely to provide a complete solution to the fragmentation problem. Still, one has to walk before one can run, and the creation of ACOs seems likely to be an important fi rst step in the evolution toward less fragmentation of healthcare. The ACOs will not be fully integrated fi rms like airlines, but will link providers in ways that could more easily morph into such integrated fi rms in the future. In any event, although the ACO provisions have received the most attention from policy insiders, other provisions offer the promise of a more complete solution to our fragmentation problem. In particular, consider the provisions on the Center for Medicare and Medicaid Innovation (CMI) and the Independent Payment Advisory Board (IPAB). The CMI provides a way to test new healthcare payment and delivery systems, including those that decrease fragmentation. Introducing the idea, the provisions state: The purpose of the CMI is to test innovative payment and service delivery models to reduce program expenditures  .  .  . while preserving or enhancing the quality of care furnished to individuals. . . . In selecting such models, the Secretary shall give preference to models that also improve the coordination, quality, and effi ciency of health care services. . . . 56

The statute then sets out criteria to guide the selection of models to be tested, as well as particular payment and delivery models that might merit examination. Models tested should be those that “address[] a defi ned population for which there are deficits in care leading to poor clinical outcomes or potentially avoidable expenditures.”57 This language certainly encompasses models that address fragmentation, given the abundant evidence that the lack of coordination reduces quality and increases cost.58 Further, several of the specific models mentioned in the CMI provisions address concerns that lead to fragmentation. One potential model

Obamacare and the Theory of the Firm

215

that can be tested is “Contracting directly with groups of providers of services and suppliers to promote innovative care delivery models, such as through risk-based comprehensive payment or salary-based payment.”59 Another model would be “Establishing comprehensive payments to Healthcare Innovation Zones, consisting of groups of providers that include a teaching hospital, physicians, and other clinical entities, that, through their structure, operations, and joint activity deliver a full spectrum of integrated and comprehensive health care services.”60 These provisions would allow the CMI to adopt regulations that override the legal obstacles to integrated payments and control that cause undue fragmentation. The CMI provisions tend to be overlooked because they seem to provide merely for experimentation rather than to authorize national regulation. But the provisions actually allow any successful experiment to be implemented on a national basis and thus to become national policy. The statute provides that [T]he Secretary may, through rulemaking, expand (including implementation on a nationwide basis) the duration and the scope of a model that is being tested  .  .  . if  .  .  . the Secretary determines that such expansion is expected to—(A) reduce spending under applicable title without reducing the quality of care; or (B) improve the quality of care and reduce spending.61

The IPAB provisions create a new independent agency, the Independent Payment Advisory Board, that is required to produce proposals to lower Medicare spending in years in which payments are expected to exceed targets.62 An interesting feature of the IPAB is that it can make proposals that become law unless Congress enacts legislation to override the specific proposal.63 In meeting its duty to reduce Medicare costs, the IPAB cannot ration care, increase premiums or cost sharing, or restrict benefits or eligibility.64 What distinguishes the IPAB from the CMI is that the IPAB must act if the statutory triggers are met: “The [IPAB] shall develop detailed and specific proposals related to the Medicare program. . . .”65 Moreover, the IPAB must make proposals that improve health or efficiency through greater integration or coordination if it can. “In developing and submitting each proposal . . . the [IPAB] shall, to the extent feasible . . . include recommendations that . . . improve the health care delivery system and health outcomes, including by promoting integrated care, care coordination, prevention and wellness, and quality and efficiency improvement.”66

216

Einer Elhauge

Thus, the IPAB provisions provide a strong mechanism to defragment U.S. healthcare. Whenever Medicare spending is projected to exceed targets, which seems sadly inevitable, the IPAB must make proposals that include efforts to integrate care and improve care coordination if they would improve medical quality and efficiency. Given the evidence noted above, this arguably creates an affi rmative duty for IPAB to adopt regulations that allow fi rms to defragment healthcare. Under either the CMI or the IPAB, the federal government could and should promulgate several regulations that cut to the core of the fragmentation that exists in U.S. healthcare. Possible regulations could preempt state laws that prevent fi rms from controlling physician behavior, such as the corporate practice of medicine doctrine and various tort doctrines. Regulations could limit the scope of fee-splitting laws to allow fi nancial coordination between hospitals and physicians. Payment systems could be changed to make integrated payments to a fi rm that orchestrates the providers necessary to achieve a health outcome. Paying for the value of health outcomes may often be too difficult because it requires putting a fi nancial value on those outcomes and determining the extent to which fi rms improved them. But an alternative payment system that could optimize the incentives of fi rms would be to give each fi rm both (1) an amount per enrollee attracted, which they could keep as profits; and (2) a separate risk-adjusted payment that must be spent on care for the group of enrollees, and thus cannot go to fi rm profits.67 Such a system would eliminate incentives to undercare because the payments for care cannot be kept if unspent, while creating a powerful incentive to spend those care payments efficiently to maximize health benefits because that will attract the most enrollees. It would thus give integrated fi rms the ability and incentives to optimize team production by the medical professionals within their control. This proposed approach of separating profit payments from care payments has some similarity to the medical loss ratio requirements of Obamacare, which require that insurers spend 80– 85 percent of their premiums on healthcare.68 But the proposal differs in various key ways. First, it would extend beyond insurers to integrated providers, thus providing a solution to the integration problem rather than merely an effort to reduce insurer profiteering. Second, it would give a profit payment per enrollee that is not set as a percentage of spending, thus eliminating the incentive to spend more on care to get more. Third, it would separate profits from spending rather than try to micromanage the allo-

Obamacare and the Theory of the Firm

217

cation of money between administration and medical care, which may be counterproductive when better administration would lead to more efficient care. Nonetheless, the proposal has enough of a family resemblance to the medical loss ratio rules that it would not require a great leap in regulatory strategy, which might smooth the transition to such a system.

Conclusion The fragmentation of U.S. healthcare increases costs and decreases quality. The main reason such fragmentation persists is a combination of regulatory and payment laws that entrench physician autonomy and prevent the development of integrated firms that have the incentives and control necessary to achieve the team coordination needed in modern medicine. In an era when the biggest question facing the country may be the longterm trend in healthcare costs,69 the Obamacare provisions that enable the federal government to remove legal barriers to efficient healthcare integration offer a critical and useful tool. Used effectively, regulations under these provisions could improve healthcare, potentially saving tens of thousands of lives, avoiding hundreds of thousands of injuries, and saving hundreds of billions of dollars in medical costs. Given the persisting objections to the costs of Obamacare, adopting regulations like this that can save huge sums of money while improving quality may indeed be necessary to make it a success and fend off efforts to undermine it. These regulatory tools also have the clear benefit of allowing progress to be made without requiring another round of politically volatile federal healthcare lawmaking.

Notes I am grateful for summer research support from Harvard Law School and the Petrie-Flom Center, for research assistance from Jordan Wish, and for comments from Scott Altman, Nicholas Bagley, Amitabh Chandra, Nancy-Ann DeParle, Alain Enthoven, Julia Feldman, Allison Hoffman, Mark Hall, Joe Newhouse, Frank Pasquale, James Rebitzer, Barak Richman, Chris Robertson, Bill Rubenstein, Jonathan Schenker, Alan Stone, and Patrick Taylor, and participants at workshops at the University of Chicago and Harvard Law Schools.

218

Einer Elhauge

1. “If Air Travel Worked Like Healthcare,” YouTube video, 7:00, uploaded by TheNewAltons, January 4, 2010, http://www.youtube.com/watch?v=5J67xJKpB6c. 2. For a comprehensive set of essays that address causes, effects, and remedies for fragmentation in the U.S. healthcare system, see generally The Fragmentation of U.S. Healthcare (Einer R. Elhauge ed., 2010). 3. Pub. L. No. 111-148, 124 Stat. 119 (codified as amended in scattered sections of 26 and 42 U.S.C.). Although the term “Obamacare” was originally a mocking Republican characterization designed for political effect, even President Obama now embraces it on the grounds that saying that Obama cares is not exactly an insult. I thus use the term because it is certainly shorter and more memorable, and I think it has lost the political spin it once had. 4. See Einer Elhauge, Why We Should Care About Health Care Fragmentation and How to Fix It, in The Fragmentation of U.S. Healthcare, supra note 3, at 1, 1– 6. 5. See Randal Cebul, James B. Rebitzer, Lowell L. Taylor, and Mark Votruba, Organizational Fragmentation and Care Quality in the U.S. Health Care System, in The Fragmentation of U.S. Healthcare, supra note 3, 37, at 47–48. 6. Comm. on Quality of Health Care in Am., Inst. of Med., Crossing the Quality Chasm 83 (2001) [hereinafter “IOM, Chasm”.] 7. Id. 8. Comm. on Quality of Health Care in Am., Inst. of Med., To Err Is Human 1– 2 (2001). The Institute of Medicine estimates that healthcare costs represent approximately half of the $17– 29 billion cost; the other half represents lost income and production. Id. 9. Id. at 3. 10. IOM, Chasm, supra note 7, at 3, 43. 11. Hoangmai Phan, Deborah Schrag, Ann S. O’Malley, Beny Wu, and Peter B. Bach Care Patterns in Medicare and Their Implications for Pay for Performance, 356 New Eng. J. Med. 1130, 1134 tbl.1 (2007). 12. David A. Hyman, Health Care Fragmentation: We Get What We Pay for, in The Fragmentation of U.S. Health Care, supra note 3, 21, at 23. 13. Id. 14. Id. at 26. 15. Id. at 26– 27. 16. Id. at 27. 17. IOM, Chasm, supra note 7, at 4. 18. IOM, Chasm, supra note 7, at 90. 19. Jonathan S. Skinner, Douglas O. Staiger, and Elliott S. Fisher, Is Technological Change in Medicine Always Worth It? The Case of Acute Myocardial Infarction, 25 Health Affairs W34, W42–43 (2006). 20. Id.

Obamacare and the Theory of the Firm

219

21. Elizabeth A. McGlynn, Steven M. Asch, John Adams, Joan Keesey, Jennifer Hicks, Alison DeCristofaro, Eve A. Kerr, The Quality of Health Care Delivered to Adults in the United States, 348 New Eng. J. Med. 2635, 2641, 2642 tbl.3 (2003). 22. Id. at 772. Steffie Woolhandler, Terry Campbell, David U. Himmelstein, Costs of Health Care Administration in the United States and Canada, 349 New Eng. J. Med. 768, 772 (2003). 23. Id. at 771 tbl.1. 24. Id. 25. Einer Elhauge, U.S. Antitrust Law & Economics 369 (2d. ed. 2011). 26. See Elhauge, supra note 5, at 2; Hyman, supra note 13, at 34. 27. R. H. Coase, The Nature of the Firm, 4 Economica (n.s.) 386 (1937). 28. See Armen A. Alchian & Harold Demsetz, Production, Information Costs, and Economic Organization, 62 Am. Econ. Rev. 777 (1972) 29. Elhauge, supra note 5, at 6. 30. This example is laid out in Regina E. Herzlinger, Why Innovation in Health Care is So Hard, 84 Harv. Bus. Rev. 58, 64 (2006). 31. IOM, Chasm, supra note 7, at 3. 32. Id. at 101, 202. 33. Id. at 13. 34. See Mark A. Hall, Institutional Control of Physician Behavior: Legal Barriers to Health Care Cost Containment, 137 U. Pa. L. Rev. 431, 509–18 (1988). Some states have codified the doctrine. See, e.g., Cal. Bus. & Prof. Code § 2400 (West 2013). 35. Elhauge, supra note 5, at 12. 36. Id. See generally James F. Blumstein, Of Doctors and Hospitals: Setting the Analytical Framework for Managing and Regulating the Relationship, in The Fragmentation of U.S. Healthcare, supra note 3, 135. 37. Elhauge, supra note 5, at 12. 38. 42 U.S.C. § 1395n (2006 & Supp. V 2011); Elhauge, supra note 5, at 11. 39. 42 U.S.C. § 1395 (2006); Elhauge, supra note 5, at 11. 40. Elhauge, supra note 5, at 11. These names for the various Medicare programs come from the codification of Medicare in 42 U.S.C., chapter 7, subchapter XVIII. 41. See Hyman, supra note 13, at 26. 42. Elhauge, supra note 5, at 11. 43. Hall, supra note 35, at 488. 44. See 42 U.S.C. § 1320a- 7a(b) (2006 & Supp. V 2011); id. § 1395nn(a). 45. See, e.g., Cal. Bus. & Prof. Code §§ 650– 652; Fla. Stat. § 395.0185 (2011); id. § 458.331(1)(i). 46. See Elhauge, supra note 5, at 9. 47. See Einer Elhauge, Obamacare on Trial (2012).

220

Einer Elhauge

48. 42 U.S.C. § 1395jjj(b)(1). 49. Id. § 1395jjj(b)(2)(A). 50. Id. § 1395jjj(b)(2)(B); id. § 1395jjj(b)(2)(D); id. § 1395jjj(b)(3). 51. 42 U.S.C. § 1395jjj(a)(1). 52. Id. § 1395jjj(d)(1)(B)(i). 53. Id. § 1395jjj(d)(3)– (4). 54. Medicare Program; Medicare Shared Savings Program: Accountable Care Organizations, 72 Fed. Reg. 67,802, 67,802 (Nov. 2, 2011) (to be codified at 42 C.F.R. pt. 425). 55. See 42 U.S.C. § 1395jjj(i). 56. Id. § 1315a(a)(1) (emphasis added). 57. Id. § 1315a(b)(2)(A). 58. See Part I, supra. 59. 42 U.S.C. § 1315a(b)(2)(B)(ii) (emphasis added). 60. Id. § 1315a(b)(2)(B)(xviii) (emphasis added). 61. Id. § 1315a(c) (emphasis added). 62. Id. § 1395kkk(b). See generally Timothy Stoltzfus Jost, The Independent Medicare Advisory Board, 11 Yale J. Health Pol’y L. & Ethics 21 (2011). 63. 42 U.S.C. § 1395kkk(b)(3). The ACA also contains expedited procedures for Congress to consider proposals by the IPAB. See id. § 1395kkk(d). 64. 42 U.S.C. § 1395kkk(c)(2)(A)(ii); see also id. § 1395kkk(c)(2)(A)(iii) (prohibiting IPAB proposals before 2020 that would lower provider reimbursements). 65. Id. § 1395kkk(c)(1)(A). 66. Id. § 1395kkk(c)(2)(B)(ii)(I). 67. See Elhauge, supra note 5, at 19– 20. 68. 42 U.S.C. § 300gg-18. 69. See, e.g., Cong. Budget Office, The Long-Term Budget Outlook 7, tbl.1-2 (June 2010, revised August 2010) (projecting federal healthcare spending to rise from 5.5 percent of gross domestic product in 2010 to 9.7 percent in 2035); Louise Radnofsky, Steep Rise in Health Costs Projected, Wall St. J., June 12, 2012, at http://online.wsj.com/articles/SB1000142405270230376810457746273171 9000346 (US healthcare spending projected to rise from 17.9 percent in 2010 to 19.6 percent in 2021).

Chapter nine

Can Federal Provider Payment Reform Produce Better, More Affordable Healthcare? Meredith B. Rosenthal

W

ith the ongoing implementation of comprehensive federal insurance coverage reform, there is intense pressure on the federal government to address long-standing problems of uneven quality, waste, and seemingly unsustainable spending growth in the U.S. healthcare delivery system. Federal policy initiatives to improve affordability and quality have traditionally avoided direct regulation of the delivery system and instead set up the conditions for a form of “managed competition” in which market forces are channeled or redirected in ways that achieve policy aims. These types of policies typically involve manipulating consumer or supplier incentives and choice architecture to change behavior and thus delivery system outcomes. At this juncture, federal policy—directed by the Affordable Care Act and other legislation, as well as ongoing programmatic developments—is addressing the need for improved healthcare value primarily through provider payment reform. While some of these activities include engaging state Medicaid agencies in payment restructuring, Medicare is the focus of reform and this chapter, in part because unlike Medicaid it is fully controlled by the Centers for Medicare and Medicaid Services and in part because of its leading role in shaping both healthcare delivery and the policies of commercial insurers. Payment reform is nothing new to Medicare. Indeed, since the Prospective Payment Assessment Commission (the predecessor to the Medicare

222

Meredith B. Rosenthal

Payment Assessment Commission, MedPAC) was formed in the 1980s, Medicare’s leaders and advisers have engaged in a nearly continuous series of incremental and occasionally significant payment experiments and reforms. It is important to ask whether payment reform is an effective lever for policy, given how long we have been at it. Are past and present payment reforms improvements along a path to better and more affordable healthcare or just the technocratic flavor of the week? To answer whether and how payment policy can be an effective tool for improving healthcare delivery, I begin by sketching out the theoretical framework that health economists use to examine provider payment and summarizing the empirical tests of this theory at a high level. I then summarize the evolution of Medicare from cost-based to prospective reimbursement and its entry into value-based purchasing, noting some of the major economic and policy lessons along the way. In this section I also assess whether there is any reason to believe payment policy is making progress rather than moving in a hapless circle of reform and counterreform. Finally, I consider what federal policy makers might do to improve the chances that provider payment reforms will succeed. In my concluding remarks I also endeavor to address a reasonable question that critics of payment reform might ask: Is it impossible to get payment “right,” either because the economic realities are more complicated than we allow or because politics intervenes?

Why Not Rely on Demand to Achieve Efficiency in Healthcare? While this chapter is about the design and implementation of administrative pricing, it is appropriate to begin with at least a nod to the rationale for a procurement model for healthcare as opposed to a consumer- driven market. After all, it is only fair to judge the performance of payment reform against some reasonable alternative, the leading example of which is competition. Policy proposals in this category generally start with the notion that for competition to work in healthcare, consumers fi rst need more “skin in the game” than traditional health insurance provides.1 Greater sensitivity to cost in this manner might come about through defi ned- contribution benefit plans including vouchers, or highdeductible health plans. Better information about cost and quality, deci-

Federal Provider Payment Reform

223

sion support, and a broader range of choices (facilitated by freer entry and less restrictive licensing) typically complete the package. Would such a consumer- driven system lead to lower cost and better quality care?2 Even before we consider the empirical evidence, theoretical concerns arise out of the poor fit between the market paradigm and the reality of healthcare delivery. Most work in this area has emphasized the presence of conditions—such as consumers’ lack of independent information about appropriate diagnosis and treatment—that cause markets to be inefficient (Hurley 2000). Seminal work by Nobel laureate Kenneth Arrow further described the proliferation of “nonmarket” institutions, such as the trust that is critical to the physician-patient relationship, that fi ll the breach left by market failure (Arrow 1963). A wide range of empirical evidence supports the theory that information problems cause significant distortions in healthcare markets. First, while increasing patient cost sharing has been demonstrated to reduce utilization of healthcare, there are limitations to the usefulness of cost sharing as a mechanism for improving efficiency. Most importantly, patients do not prioritize healthcare in accordance with clinical notions of relative value; patients with high cost sharing appear equally likely to cut back on essential healthcare services as on services of low or no value (Newhouse and the Health Insurance Experiment Group 1993). While some patient behavior that goes against expert opinion (e.g., care guidelines for people with hypertension) might be individually rational due to patient preferences regarding side effects or discounting of future benefits, in other cases failures of information or decision making (e.g., myopia) mean that cost sharing does not effectively lead consumers to optimize healthcare choices. Many studies have demonstrated worrisome responses to increasing cost sharing, and the landmark RAND Health Insurance Experiment showed that patients reduced utilization without regard to value, leading to health harms in lower income, sicker populations (Newhouse and the Health Insurance Experiment Group 1993; Greene 2008; Roblin et al. 2005; Tamblyn et al. 2001; Blustein 1995). Concerns about unintended consequences that might ensue as a result of poor consumer information or understanding could be addressed directly. Such efforts include offering patients third-party support for healthcare decisions—when to seek a doctor’s care, how to manage a chronic condition, how to participate in decision making around major treatment choices. For example, many health insurance plans offer toll-free

224

Meredith B. Rosenthal

telephone access to nurses who can help patients navigate medical concerns. Another more targeted approach, shared decision making, makes use of video and other learning aids to help patients understand differences in the experience and outcomes associated with alternative therapeutic choices. Studies of shared-decision-making aids for surgical and other major interventions have shown them to produce benefits in terms of more patient-centered care (patients being more actively engaged in decisions and getting care that is consistent with their stated preferences) and more realistic expectations about treatment effects, although the impact on cost is less clear (O’Connor et al. 2008). While this approach may hold promise as a means of reducing waste, shared-decision-making practices are not widespread, in part due to lack of provider acceptance (Frosch et al. 2012). On the other hand, substantial effort—and federal dollars—has gone into arming consumers with information to select physicians and hospitals that are high quality, low cost, or both. Informing consumers in this way has two intended effects: improving average quality or efficiency by helping patients fi nd the best providers, and causing providers to compete and therefore improve care along reported dimensions. Such provider report cards have been available since the 1980s and have grown in scope and sophistication. Research published over the past several decades casts doubt on the value produced by provider quality report cards (Werner and Asch 2005; Sinaiko, Eastman, and Rosenthal 2012). In particular, studies show that patients frequently do not understand, trust, or rely on report cards to select a provider. One of the earliest studies of patient responses to publicly reported information on the quality of coronary bypass surgery found that only 12 percent of patients who underwent the procedure were aware of a widely publicized report card prior to surgery and only 2 percent reported using the ratings in selecting a hospital or surgeon (Schneider and Epstein 1998). Two decades later, only about 20 percent of U.S. consumers and even fewer physicians say they use healthcare quality reports, and observed behavior suggests only minimal effects on choice of provider (Kaiser Family Foundation and Agency for Healthcare Research and Quality 2004). While consumer “shopping” behavior does not appear to be stimulated by quality report cards, providers may nonetheless respond to publicized reports. In a controlled experiment in Wisconsin, hospitals whose quality data were reported publicly were more likely to participate in quality improvement activities than hospitals receiving the same quality

Federal Provider Payment Reform

225

information privately (Hibbard, Stockard, and Tusler 2003). Public (as opposed to private) reports may cause this effect because providers anticipate eventual response by the market, or simply due to shame or competitive instincts. On the negative side of the ledger, public reporting has been observed to yield other supply responses, including market exit and avoidance of patients perceived to be high risk (Werner and Asch 2005). On the whole, efforts to transform healthcare by mobilizing or redirecting competition through consumers have proven to be problematic, due to the limits of patient expertise about health and healthcare and to hidden information and actions by physicians and other providers. The fact that building a better health system through consumerism requires that consumers take on more fi nancial risk should not be minimized. Insurance serves a valuable economic purpose given risk aversion and the uncertainty surrounding health events and medical costs. Cost sharing not only diminishes risk protection but also may act as a barrier to needed care for those with low income. All of this theory and evidence taken together is sufficient to suggest that a system of incentives delivered through a well- designed procurement model may be able to outperform a consumer- directed one. Thus, it is worth examining that procurement model and the prospects for improving it.

Some Economics of Payment Incentives in Healthcare There is an extensive theoretical and empirical literature on provider payment incentives in healthcare (McGuire 2000; Pauly 2000). Health economists take the view that providers act as agents for payers (or patients, depending upon the framing of the problem) and value both health benefits—through professionalism or altruism—and their own bottom line. This model has a number of implications for payment contracts. First, a change in the profitability of delivering a service at the margin will change volume. Second, while the usual assumption is that physicians will “do no harm,” the marginal benefit of care delivered under a fee-for- service system will be lower than that delivered when payment is prospective (i.e., not a function of volume). Theoretically, it is impossible for fee for service alone (i.e., without some other rationing mechanism) to generate efficient levels of care. Whether care decisions are optimal under prospective payment depends in theory on the way physicians trade off benefits and costs—often parameterized as the degree of

226

Meredith B. Rosenthal

agency. If physicians care too little about patients, pure prospective payment may lead to underprovision of services (where the marginal benefit to the patient at the level of services provided exceeds the marginal cost). Another concern that arises with prospective payment is that providers will have an incentive to avoid patients with higher-than-average expected costs that are not recognized by the payment system—for example, through risk adjustment. Both of these unintended consequences of prospective payment can be addressed by using a mix of fee for service and prospective payment rather than either alone. While the focus of agency theory in health economics has largely been on cost and the intensity of service delivery, some work has also addressed the quality of care. In theory, demand might enforce efficient provision of quality, as long as quality (however defined) is observable to patients (Ma 1994). Alternatively, pay for performance—where fi nancial incentives are explicitly tied to quality measures—may be an effective way to mimic the role that demand would play in other markets. The theoretical desirability of pay for performance depends upon a number of factors, including the precision of quality measurement, the extent to which payers capture patient preferences in selection of quality targets, and spillover effects between measured and unmeasured quality (Hart and Holmstrom 1987). Concerns about appropriate accounting for patient differences in quality measurement or the setting of performance targets also arise in the theory and practice of pay for performance (Rosenthal et al. 2005). In terms of the empirical evidence, data strongly support the theory of provider payment incentives sketched above, both in terms of the relative importance of payment as a driver of care patterns and the effects of changes in the level and structure of payment (McGuire 2000). The most indisputable fi nding is that the marginal payment for quantity matters; when provider payment switches from volume- or cost-based to prospective the quantity of care declines measurably. Specifically, studies have documented differences in visits, testing, and hospitalization between physicians who earn revenues through fee-for-service payment compared to physicians who are salaried or paid by capitation (Hellinger 1996; Jennison and Ellis 1987; Epstein, Begg, and McNeil 1986; Hemenway et al. 1990; Stearns, Wolfe, and Kindig 1992; Rosenthal 2000). While paying for volume increases costs, it is less clear whether the unintended adverse consequences of either fee for service or prospective payment are realized. There is a small literature that examines qual-

Federal Provider Payment Reform

227

ity differences between fee for service and capitation or salary, but most studies are confounded by differences in benefit design (e.g., HMO versus indemnity insurance) and delivery system features, or rely on relatively weak measures of quality. With these important caveats, what the literature shows overall is that quality may be higher or lower in systems that favor fee-for-service payment (Miller and Luft 1997; Rosenthal 1999). Even less is known about the extent to which healthcare providers differentially avoid high-risk patients when they are paid prospectively. Hospital prospective payment has been associated with increased admission of lower-risk patients compared to cost-based reimbursement, but similar evidence does not exist for physician practices (Feinglass and Holloway 1991). In contrast to the strong evidence linking volume-based payment and service intensity, the literature on pay for performance in the United States (and internationally) is mostly negative. Systematic reviews of pay for performance in a wide variety of healthcare settings have concluded that such incentives often fail to produce a measurable benefit and that even the best-performing programs may not be cost- effective (Eijkenaar et al. 2013; Peterson et al. 2006; Van Herck et al. 2010). In terms of physician payment, the largest pay-for-performance program to date in the United States is that organized through the Integrated Health Care Association (IHA) in California. Founded in 2001, the IHA pay-forperformance program includes eight health plans representing approximately 10 million insured persons cared for by thirty-five thousand physicians in over two hundred physician groups. While the IHA initiative has important strengths, including multistakeholder engagement around a shared quality measure set and the involvement of virtually all commercial insurers in California, its beneficial impact on the quality of care has only been demonstrated for a single preventive care measure: cervical cancer screening (Rosenthal et al. 2005; Mullen, Frank, and Rosenthal 2010; Damberg et al. 2009). The weak evidence base to support pay for performance in the United States includes evaluations of two major Medicare demonstration projects. First, the Premier Demonstration targeted a set of quality measures focused on three types of hospital discharges and provided bonuses to the top two deciles of participating hospitals. During the initial two-year period of the demonstration, average performance increased by roughly two to four percentage points for participating hospitals compared to nonparticipating hospitals (Lindenauer et al. 2007). However, a more

228

Meredith B. Rosenthal

recent longer-run study demonstrates that these modest early improvements in hospital quality attenuated over time (Jha et al., 2012). The second initiative, the Prepaid Group Practice Demonstration, included bonus potential for both quality and savings relative to the projected total cost of care for patients assigned to a participating institution. This program involved substantially more complicated incentives because both quality and cost considerations factored into the payment formula. Moreover, savings were determined after the fact by comparison to the performance of similar providers in the same geographic market. Bonuses were ultimately paid to several of the participating organizations for meeting savings or quality- of- care targets. However, subsequent research suggests that some of the observed change in performance may have been caused by coding changes that made the covered populations appear to increase in severity of illness, a factor used in the savings calculations (Colla et al. 2012). The lackluster results of pay-for-performance efforts to date might be seen as an indication of the failure of the principal-agent model to predict behavior. More likely, however, they prove the theory that providers understand and optimize over a complete range of behaviors. In some cases, designers of pay-for-performance programs have been guilty of design flaws called out by the classic “On the Folly of Rewarding A, While Hoping for B” (Kerr 1975)—they pay for proxy measures and get proxy performance as a result. One example is targeting testing measures for chronic diseases like diabetes and coronary artery disease. The idea behind the testing measures is that physicians need test results to manage blood sugar or cholesterol. Testing is thus necessary but not sufficient for improved control of many common chronic illnesses and the prevention of long-term complications. Testing measures tend to be used in pay for performance because they can be measured using billing data, everyone can agree they should be done, they are reasonably in the control of providers (physicians and their clinical staffs), and they are not as sensitive to patient differences as the associated control or health- outcome measures are. Nonetheless, testing per se is not what payers are “hoping for,” and there is no guarantee that improved outcomes will follow. Paying directly for outcomes, on the other hand, might be desirable in theory but there are two major barriers to this alternative. First, most patient outcomes vary widely as a function of variables outside the control of providers, and this variation creates risk in a pay-for-performance program. To the extent that fi xed and observable factors such as patient socioeconomic

Federal Provider Payment Reform

229

status affect the provider’s ability to attain pay-for-performance targets (whether outcome levels or improvement), this will also encourage avoidance of some patients as one dimension of the optimizing response to the incentive program. Two other design flaws associated with the nature of the marginal payment in pay for performance in healthcare have led, in my opinion, to most of the nonresponse observed in the literature. First, many programs simply do not offer enough money relative to the costs of improving quality. Incentives may be too small because they are offered by only one payer of many or because payers are reluctant to commit substantial funds without assurance of a return, thus creating a self-fulfi lling disappointment. The second way in which the marginal payment in pay for performance may be inadequate to yield the hoped-for quality improvement is that the reward does not vary with the work required to achieve it. If bonuses are all- or-none and providers vary in terms of baseline performance, then a pay-for-performance program generates uneven incentives where the marginal benefit of achieving a quality target (i.e., the bonus payment) may exceed the marginal cost for some (indeed, the marginal cost of achieving the target may be zero for the best providers) and fall short for others.

Healthcare Procurement under Medicare Past and Present The history of Medicare payment is a book-length topic on its own (Newhouse 2002; Mayes and Berenson 2008). In this brief section I try to give some perspective on both the changing ends and means of Medicare provider payment policies. If payment reform is to be judged according to its effectiveness it is appropriate to measure effectiveness based on Medicare’s goals. These have shifted over time, in part due to the increasing burden of healthcare spending and in part due to growing awareness of healthcare quality and patient-safety problems. Cost-based reimbursement was the central tenet of Medicare payment at its inception. The Social Security Act Amendments of 1965 explicitly link reimbursement to cost as follows: “The amount paid to any provider of services with respect to services for which payment may be made . . . shall . . . be the reasonable cost of such services.”3 The original intent of Congress was to pay a reasonable amount to providers for the care of Medicare patients. In a 1995 article Robert

Meredith B. Rosenthal

230

Ball, who served as commissioner of Social Security under Presidents Kennedy, Johnson, and Nixon, provides insider’s insights about the intentions of Medicare’s sponsors (Ball 1995). In connection with hospital payment, he recollects: By and large, our posture at the beginning was one of paying full costs and not intervening very much in how hospitals, at least the better ones, conduct their business. . . . We believed in paying fully. We opposed shifting costs to other payers, and we avoided discounts beyond what our contractors might have secured for their own insured persons.

During this formative period, the Health Care Financing Administration and Medicare’s congressional sponsors were concerned primarily with access to care for the elderly, and cost-based reimbursement was an effective instrument. Healthcare utilization and access grew rapidly for the elderly with Medicare’s implementation. By the early 1970s, however, the escalating cost of the Medicare program—particularly in the form of hospital (Part A) payments—alongside a crisis of economic stagnation coupled with inflation, forced a change in priorities and a closer look at the incentive effects of cost-based reimbursement. The result was the Prospective Payment System (PPS) for hospitals, introduced in 1983. The diagnosis-related group payment approach under PPS appears to be a middle ground that yielded important efficiencies and moderated cost increases without dramatically altering the nature of Medicare’s relationship to the hospitals or access to care. PPS accomplished cost savings not just by paying a fi xed amount per admission (or discharge) but also through the use of yardstick pricing, which is a regulatory mechanism that mimics competition (Shleifer 1985). Studies of PPS and its effects have generally demonstrated decreases in cost (largely through length of stay), with few if any adverse effects on patient care including readmissions and thirty- day mortality. Moreover, hospitals were remarkably swift and predictable in the contours of their response to the new incentives, both of which support the application of the underlying theory of provider behavior to hospitals (Feinglass and Holloway 1991; Chulis 1991). This is not to say that the inpatient PPS is without problems. Some diagnosis-related groups (DRGs) include procedures in their defi nition, leading payments to be less than fully “prospective” (McClellan 1997). And while readmission rates did not increase when PPS was implemented,

Federal Provider Payment Reform

231

readmissions have grown steadily over the past decade (Jencks, Williams, and Coleman 2009). While there is a great deal of controversy about the meaning and cause of readmissions, at a minimum the discharge-based payment system does nothing to discourage these patterns. In the decades since the inpatient PPS was introduced the concept has been applied to postacute and outpatient hospital care, with similar results (Cotterill and Gage 2002). While these individual prospective budgeting systems have achieved cost savings within their individual silos, their proliferation has also brought the need for coordination and integration of care across the continuum, an outcome that eludes the current system, into relief. It is this notion—that payment should encourage holistic, proactive management of patient care—that animates the current wave of Medicare payment reform. While the history of Medicare institutional payment system reform can be viewed largely as a success story in changing healthcare delivery through procurement reform, physician payment has proven more difficult. Because of physicians’ largely independent status relative to hospitals, their services are not bundled with the institutional payment for inpatient care and a parallel prospective system is both conceptually and practically more difficult to construct. Instead, physician payment has remained fee for service, with the only major reform being the introduction of yardstick pricing in the form of the Medicare Fee Schedule—the Resource Based Relative Value Scale (RBRVS). The RBRVS was designed to set payments to physicians fairly in recognition of the average costs—both fi xed and variable—of delivering different types of cognitive and procedural services (Hsiao et al. 1992). Relative to the “customary, prevailing, and reasonable” method of pricing services that preceded it, the RBRVS reduced the payment gap between cognitive services and technical services such as imaging and surgery. While the repricing of services using the resource estimates gathered by auditors may have redistributed payments along these lines initially, the fi xing of a single price for procedures with relatively infrequent updates caused technologically advanced services to become overpriced with time relative to the true cost of delivering them. When the RBRVS was adopted, it was accompanied by a total expenditure target for physicians’ services—the Volume Performance Standard. In the event that total spending exceeded the target, physician fees would grow more slowly or be cut. The Volume Performance Standard was applied to all physician spending without regard to specialty or location.

232

Meredith B. Rosenthal

Thus overuse by cardiothoracic surgeons in Florida would negatively affect the fees of general practitioners in Minnesota. Overall, the effect of the RBRVS plus expenditure targets was to substantially slow the growth of Medicare Part B spending in the years immediately following its implementation (Mayes and Berenson 2008). But over time the collective budget and the inherent long-run differences in pricing between cognitive (low-technology) and procedural (high-technology) services, coupled with the fact that it is much easier for procedure-oriented specialists to increase volume, have led to distortions and renewed spending growth. After all, the existence of a national spending target does nothing to discourage individual physicians from increasing volume: the sustainable growth rate (SGR) concept is a classic problem of the commons. Moreover, Congress ultimately proved unwilling to enforce drastic across-the-board fee cuts as the SGR requires. At present, many stakeholder groups have called for the repeal of the SGR, although there is a great deal of debate about what Medicare should do instead to control physician spending. The history of physician payment under Medicare highlights both the responsiveness of physicians to fi nancial incentives and the challenge of anticipating fully the consequences of any complex payment system. It does not support the notion that payment reform is fruitless or unmanageable. While it may have been difficult to predict the erosion of cognitive service margins (with its disastrous consequences for primary care) and the responsiveness of imaging and other types of diagnostic and therapeutic procedures, it should have been expected that a single national expenditure target would not be an effective means of restraining the inevitable volume response to price control. No doubt economists and many policy makers understood this when the policy was designed but political or practical constraints prevented them from doing more. In summary, there are several key lessons from the history of Medicare physician payment that are pertinent to the present exercise of putting Medicare on a sustainable path. First, price control is insufficient to control spending; tweaking the RBRVS may deliver some short-run savings but it is not a solution. Moreover, a price control that is applied uniformly to physician services (like the SGR) will differentially affect procedures and evaluation and management services, which are much less elastic. Finally, global budgets or spending targets need to be assigned to some accountable unit—there is no national physician organization that has the authority to enforce the SGR—and be enforceable.

Federal Provider Payment Reform

233

Those lessons, along with a growing sense that fragmented payment systems have prevented better integration of care around the needs of patients, have motivated a series of new efforts to improve healthcare delivery through payment reform. These policies build upon prospective payment and the fi rst tentative steps that Medicare began taking towards value-based purchasing during the past decade. The most significant change in the current era of reform, however, is not in the method of payment but in the recognition for the fi rst time of health systems, including medical groups and integrated delivery systems, as contracting units. There is a limited amount that Medicare can accomplish as long as it pays only individual physicians and institutions; contracting with larger, more sophisticated entities that control a broader scope of services opens up a wide range of new possibilities.

Payment Reform and the Affordable Care Act The central accomplishments of the ACA were twofold: to establish private health insurance market conditions that, coupled with mandates and subsidies, would maximize uptake of private insurance; and to expand public health insurance to all those in poverty, regardless of other criteria. Payment reform, while recognized by policy makers as essential to the sustainability of universal coverage, was secondary to coverage expansion. Payment reform policies in the ACA (see Table 9.1) include some modest changes to Medicare’s payment systems for hospitals and physicians, to restrain price increases and incorporate new incentives to address specific problems such as healthcare acquired conditions and the use of evidence-based protocols. More fundamental changes in the structure of reimbursement under Medicare and Medicaid, such as the Accountable Care Organization Shared Savings Program and bundled payments, are also established in opt-in programs, through pilots and demonstration programs. Finally, the payment reform provisions of the ACA include the establishment of two new policy institutions designed to overcome some of the political obstacles to meaningful payment reform: the Centers for Medicare and Medicaid Innovation (CMI) and the Independent Payment Advisory Board (IPAB). The CMI has the ability to conduct pilot initiatives. If successful, these pilots can be disseminated throughout the program without new legislation. The IPAB, which has not been formed as of November 2014, is intended to help the Centers for Medicare

Table 9.1. Payment Reform Provisions of the Affordable Care Act 2702

Prohibition of Medicaid Payment for Healthcare- Acquired Conditions. Requires the Secretary to promulgate regulations specifying healthcare-acquired conditions for which Medicaid payment is prohibited.

Regulations to take effect by July 1, 2011

2704

Medicaid Bundled Payment Demonstration. Establishes a five-year demonstration project (Jan. 1, 2012 through Dec. 31, 2016) in up to eight states to study the use of bundled payments for hospital and physician services under Medicaid.

Report to Congress due one year after conclusion of demonstration project

2706

Medicaid Pediatric Accountable Care Organization Demonstration Program. Requires the Secretary to conduct a five-year demonstration project (Jan. 1, 2012 through Dec. 31, 2016), under which a participating state is allowed to recognize pediatric providers as an accountable care organization for the purpose of receiving incentive payments. Authorizes the appropriation of such sums as may be necessary for the project.

No specified deadlines

3001

Hospital Value-Based Purchasing Program. Establishes a value-based purchasing (VBP) program for hospitals starting in FY2013, under which value-based incentive payments will be made based on hospital performance on quality measures related to common and high- cost conditions such as cardiac, surgical, and pneumonia care. The hospital VBP program does not include measures of hospital readmissions.

Value-based incentive payments are for hospital discharges on or after Oct. 1, 2012

3007

Physician Value- Based Payment Modifi er. Requires the secretary to develop and implement a budget-neutral payment system that will adjust Medicare physician payments based on the quality and cost of the care they deliver. Quality and cost measures will be risk-adjusted and geographically standardized. The new payment system is to be phased in over a two-year period beginning in 2015.

Payment adjustments apply to certain physicians beginning on Jan. 1, 2015, and apply to all physicians by Jan. 1, 2017

3008

Hospital-Acquired Condition Payment Adjustment. Starting in FY2015, hospitals in the top 25th percentile of rates of hospital-acquired conditions for certain highcost and common conditions will be subject to a payment penalty under Medicare. Requires the secretary to report to Congress on the appropriateness of establishing a healthcare-acquired- condition policy related to other providers participating in Medicare, including nursing homes, inpatient rehabilitation facilities, long-term- care hospitals, outpatient hospital departments, ambulatory surgical centers, and health clinics.

Report to Congress due by Jan. 1, 2012; payment adjustments are for hospital discharges on or after Oct. 1, 2014

3021

Center for Medicare and Medicaid Innovation (CMI). Requires the secretary, no later than Jan. 1, 2011, to establish the CMI within CMS. The purpose of CMI is to test and evaluate innovative payment and service delivery models to reduce program expenditures under Medicare, Medicaid, and CHIP while preserving or enhancing the quality of care furnished under these programs. Appropriates $5 million for FY2010 for the selection, testing, and evaluation of new payment and service delivery models, and $10 billion for the period FY2011 through FY2019 plus $10 billion for each subsequent 10-fi scal-year period, to continue such activities and for the expansion and nationwide implementation of successful models.

Effective Jan. 1, 2011

Table 9.1. (continued) 3022

Medicare Shared Savings Program. Directs the secretary to implement an integrated care delivery model using Accountable Care Organizations (ACOs), modeled on integrated delivery systems. While ACOs can be designed with varying features, most models put primary care physicians at the core, along with other providers, and emphasize simultaneously reducing costs and improving quality. Under the Medicare Shared Savings Program, CMS will contract for ACOs to assume responsibility for improving quality of care provided, coordinating care across providers, and reducing the cost of care Medicare benefi ciaries receive. If cost and quality targets are met, ACOs will receive a share of any savings realized by CMS.

Effective Jan. 1, 2012

3023

National Medicare Payment Bundling Pilot Program. Requires the secretary to establish a five-year national, voluntary pilot program encouraging hospitals, doctors, and postacute care providers to improve patient care and achieve savings for the Medicare program through bundled payment models. Before Jan. 1, 2016, the secretary is also required to submit a plan to Congress to expand the pilot program if doing so will improve patient care and reduce spending. Authorizes the secretary to expand the pilot program if it is found to improve quality and reduce costs. Also, directs the secretary to test bundled payment arrangements involving continuing care hospitals within the bundling pilot program.

Pilot to be established by Jan. 1, 2013

3024

Medicare Independence at Home Demonstration Program. Requires the secretary to conduct a three-year Medicare demonstration program to test a payment incentive and service delivery model aimed at reducing expenditures and improving health outcomes that uses physician- and nurse practitioner– directed primary care teams to provide homebased services to chronically ill patients. The secretary must submit a plan no later than Jan. 1, 2016 for expanding the program if it is determined that such expansion would improve the quality of care and reduce spending. Transfers $5 million from the Medicare trust funds for each of FY2010 through FY2015 for the demonstration (i.e., $30 million in total).

Demonstration to begin by Jan. 1, 2012

3025

Hospital Readmissions Reduction Program. Beginning in FY2013, adjusts payments for hospitals paid under the inpatient prospective payment system, based on the dollar value of each hospital’s percentage of potentially preventable Medicare readmissions for the three conditions with riskadjusted readmission measures that are currently endorsed by the NQF (i.e., acute myocardial infarction (heart attack), heart failure, and pneumonia). Authorizes the secretary to expand the policy to additional conditions in future years and directs the secretary to calculate and make publicly available information on all hospital patient readmission rates for certain conditions.

Payment reductions are for hospital discharges on or after Oct. 1, 2012

(continued)

Meredith B. Rosenthal

236 Table 9.1. (continued) 3403

Independent Payment Advisory Board (IPAB). Creates an independent, 15-member Payment Advisory Board tasked with presenting Congress with comprehensive proposals to reduce excess cost growth and improve quality of care for Medicare beneficiaries. In years when Medicare costs are projected to exceed a target growth rate, the board’s proposals will take effect unless Congress passes an alternative measure that achieves the same level of savings. Congress would be allowed to consider an alternative provision on a fast-track basis. Appropriates from the Medicare trust funds $15 million for FY2012 and, for each subsequent fi scal year, an amount equal to the previous fi scal year’s appropriation adjusted for inflation.

Advisory reports may be submitted to Congress beginning Jan. 15, 2014

Source: Sheldon Whitehouse. “Health Care Delivery System Reform and the Patient Protection and Affordable Care Act.” excerpted from Appendix B: Affordable Care Act Delivery System Reform Status Chart. Washington, DC: United States Senate, 2012.

and Medicaid Services (CMS) achieve a greater degree of discipline over healthcare spending. Its mandate is to propose changes to Medicare payment that will result in cost reductions if Medicare spending exceeds a target growth rate. By law, when the target growth rate is exceeded, the Secretary of HHS must accept IPAB’s recommendations or adopt an actuarially equivalent alternative. This mechanism was clearly intended to make it more difficult for interest-group politics to undermine the government’s ability to adhere to a budget constraint. Global Payment and Accountable Care The flagship payment reform of the current era is the legislative establishment by the ACA of the concept of the Accountable Care Organization (ACO), which paves the way for population-based payment and accountability. Without the ability to contract with an entity that could manage and take fi nancial risk for the continuum of care, Medicare would be (and is, outside of ACO arrangements) significantly constrained in its ability to encourage higher-value care. Take the notion of providing incentives for improving quality and reducing readmissions to hospitals, which the ACA does through §§ 2701, 3001, and 3007. Given the role of physicians—both in the hospital and in the community—in the care of patients implicated by these provisions, incentives that apply only to hospitals and not to those who write the orders cannot be optimal. Moreover, hospitals are not in a position to share fi nancial gains from improving quality with physicians, due to rules intended to prevent kick-

Federal Provider Payment Reform

237

backs for referrals. ACOs with incentives tied to population costs and quality, however, can make appropriate trade- offs across settings of care, and implement internal mechanisms (fi nancial or otherwise) to share accountability with individual physicians as appropriate. Based in part on the Physician Group Practice (PGP) demonstration, which had only modest success, the Medicare ACO Shared Savings Program uses a retrospective method of patient assignment and savings calculations. While ACOs can choose upside- only arrangements initially, risk sharing both above and below the target is required in out years of the three-year contract. The payment formula for ACOs is somewhat different from the PGP demonstration as well: quality of performance interacts with cost in determining bonus payments, and both the width of the risk corridor and rate of sharing have been altered. Nonetheless, the primary ACO Shared Savings Program, which any qualifying provider organization can enter (more than 250 are participating to date), resembles the PGP demonstration closely with its generally modest incentives. An alternative ACO program, the Advanced Payment ACO Model, uses this same shared savings framework but offers the opportunity for eligible providers to receive a portion of potential shared savings payments up front to permit investments that may be required to transform care. The Advanced Payment ACO program, a CMI pilot, was intended to make it possible for smaller, less-well- capitalized providers to participate in payment reform. A third ACO program—dubbed Pioneer—was launched as a pilot in 2012 by the CMI. Pioneer ACOs are prospectively assigned patients for whom they are accountable, and they bear full risk for the costs of caring for those patients. As the name suggests, the Pioneer program is intended for organizations at the cutting edge of care management with greater capacity to manage costs and fi nancial risks than typical provider organizations. As of November 2014, there are nineteen Pioneer ACOs. The Pioneer ACO model test is is scheduled to last until 2015 (Centers for Medicare and Medicaid Services 2015). Episode-Based Payment As early as the 1990s, CMS experimented with broader episode-based payment through the Medicare Heart Bypass Center Demonstration. This initiative, which tested the feasibility and impact of paying an allinclusive fee for facility and physician services associated with bypass

238

Meredith B. Rosenthal

graft surgery, was deemed successful in both cost and quality terms but was never adopted by Medicare. Prior to the ACA, CMS initiated a new episode-based payment demonstration—the Acute Care Episode Demonstration, which paid providers all-inclusive fees for cardiac and orthopedic surgery. A pilot program launched by CMI tests the boundaries of bundled payment for a wider array of acute conditions (all episodes are triggered by an index hospitalization). Providers can choose alternative models of payment covering either acute or postacute care with fully prospective and retrospective payment approaches. Episode payment has some advantages and disadvantages relative to global payment. The advantages are that patient needs and services for which a provider is accountable are well defi ned and typically relatively limited. There is less fi nancial risk involved and a lower- order care management and coordination problem. Many more organizations could successfully manage an episode of care around a hip replacement or heart surgery than can manage all the healthcare needs of a population. On the negative side, there is less opportunity for savings because episodebased payment (as currently designed) does nothing to reduce the number of inappropriate or preventable episodes. Moreover, while episodes may be easier to manage than global payment, they are considerably more difficult to defi ne and implement: When does an episode begin and end? What services count as part of the episode? (Hussey, Ridgely, and Rosenthal 2011.) Dividing accountability into episodes may also result in efforts to shift costs between bundled and unbundled payment systems (Struijis and Baan 2011). Expanded Use of Pay for Performance The ACA includes provisions that institutionalize Medicare’s use of pay for performance across the program, particularly for hospitals. Physician pay for performance (Physician Value Modifiers) will be implemented in a phased approach beginning in 2015 and will include measures of both quality and resource use. While the fi nal rules implementing the Physician Value Modifier Program are not yet written, the hospital Value Based Purchasing Program incorporates at least one important feature that improves upon the majority of its predecessors. In particular, the formula for bonus payments rewards both high levels of performance and improvement (through a failure-rate-reduction measure, which increases

Federal Provider Payment Reform

239

the marginal reward as the hospital approaches “optimal” care). The total at stake for performance remains modest under Medicare’s payfor-performance programs, including the Value Based Purchasing Program bonuses and penalties associated with high rates of readmissions and hospital-acquired infections. Perhaps this is inevitable given the limits of quality measurement; both multitasking concerns and risk aversion associated with imprecise performance measures point to the need to limit the amount of money at risk for the current set of measures. Can Medicare Do Better with the Current Policy Direction? Medicare provider payments systems have come a long way since 1965, when Medicare’s primary concern was access and physicians and hospitals were reimbursed with little constraint on either prices or quantities. In the intervening decades, Medicare has made significant strides in addressing the inefficiencies engendered by its payment systems, first by making hospital payment prospective, then following suit with other non-physician services—with many refinements along the way to mitigate potential unintended consequences including gaming, overtreatment of low-need patients, and dumping of high-risk ones. What are the prospects for the current wave of reforms? Given the current priority around cost control and better integration of care for those who need it most, a push towards prospective payment (more bundling, moving towards global budgets) makes sense on a number of levels. First, the diffuse system of accountability that exists now, wherein each provider is responsible only for the care he or she delivers, is an obstacle to holistic, integrated care. Second, the broader the scope of prospective payment, the greater the cost-savings opportunities from the prevention of complications and elimination of waste. Third, global budgets leave the least room for cost-increasing distortions such as demand inducement at either the service or episode level. The design of global payment arrangements in Medicare in the various ACO programs remains a work in progress. Both theory and empirical evidence suggest that a narrow band of shared savings will not deliver the results that policy makers are looking for, but this may be a necessary glide path. Eventually, risk sharing should include a broader corridor of spending with significant shared risk both above and below the targeted

240

Meredith B. Rosenthal

budget. Some organizations, such as those in the Pioneer ACO Program, may be fully capable of accepting and managing fully prospective global budgets. All global payment arrangements should include robust risk adjustment, outlier provisions, and monitoring for patient access and quality concerns to protect beneficiaries and ensure fairness to participating providers. One concern with the direction of Medicare policy reform is the transition from the current payment system to global payment. There is a great deal of capacity building required for most healthcare providers to be able to accept global payment. Needless to say, it is even more difficult to build that capacity with one foot in a fee-for-service world and one foot in an ACO world. In some healthcare markets, commercial insurers are adopting similar approaches, a trend that the federal government should encourage. One way it could do that is to require Medicare Advantage plans to transition to ACO arrangements for their provider contracts. From a provider perspective, global payment is now an opt-in system under Medicare policy, which is probably the only practical shortrun approach. But if ACOs and global payment are to take hold, the policy will need to ramp up incentives for nonparticipants to join. The best way to do this is to remove the status quo as an option by imposing fee limits for providers that do not participate in an ACO arrangement, with some exceptions for underserved areas and providers that do not have a reasonable opportunity to enter a global payment arrangement. The political feasibility of imposing and enforcing a Medicare fee schedule cliff at the end of some transition period remains in question, but the more Medicare can do to make ACO participation attractive in advance of that option, the greater the likelihood that such a policy would succeed. Even if global payment is projected to cover most of the patient care that Medicare pays for, there are two reasons for the continued development of episode payments. First, Medicare could begin paying some providers for broader and longer episodes built around a DRG relatively quickly, and the evidence suggests these efforts would save money. Most hospitals could manage an enhanced DRG that included postacute care and readmissions within some defi ned period of time. Such episode payments might also be a transitional step away from fee for service towards participation in an ACO for hospital-based physicians if the episode combined facility and physician costs. A second reason to pursue refinements to episode-based payment is that there may be an argument for carving out some kinds of care from global budgets to preserve access

Federal Provider Payment Reform

241

for very-high- cost patients who might otherwise be seen as an enormous liability. Medicare’s expanded use of pay for performance is the fi nal piece of the reform puzzle. With the adoption of higher-powered incentive systems to encourage cost control, it is critical to provide some balance in terms of quality incentives. But the limits of pay for performance as a counterbalance to global payment have to be acknowledged. These limits appear particularly relevant to the Physician Value Modifier Program that will be implemented in 2015. Measurement of resource use and quality at the individual physician level is crude at best and nonexistent for some specialties. Medicare should use a light touch there because there will be more noise than signal in many of the measures. The hospital pay-for-performance programs (rewards and penalties), on the other hand, hold more promise. The magnitude of pay-for-performance payments to hospitals may need to be titrated upward in order to produce the desired effect. There continues to be concern that the targeted measures (especially readmissions) are the wrong ones, so continued study of the outcome effects of existing programs and development of new measures should be undertaken. Pay for performance should not be viewed as the only tool for addressing quality, however. As underwhelming as the response has been in the past, Medicare should continue to invest in efforts to engage patients in evidence-based care and patient safety initiatives through transparency and population management. Healthcare’s economic heft makes provider payment reform both critical and highly politically charged. From Medicare’s inception its payment decisions have been highly politicized. Nonetheless, the program has reformulated and reduced payments on many fronts despite sometimes painful consequences to powerful interest groups. But costs continue to rise unsustainably, and it is clear there are opportunities for savings from better integration and proactive management of highrisk patients. As long as payment for physician services is based on a fee schedule and each provider is in its own silo, cost control will remain an extremely difficult policy and political challenge. While ACOs and risk sharing between Medicare and providers are works in progress, these policies are a step in the right direction. Should they prove successful, Medicare itself will have to transform from a bill-paying entity into one that provides timely information to providers to enable population management, monitors the quality of care, and enables beneficiaries to navigate their healthcare choices in a transformed system.

242

Meredith B. Rosenthal

Notes 1. See, for example, Antos (2012). 2. In my framing of the problem these policies are trying to solve I assume that the goal is to improve welfare by reducing waste (and possibly improving quality) in the healthcare system. Some proposals in favor of using vouchers take the goal to be reducing the federal budget impact of healthcare, which is not equivalent. 3. 42 U.S.C. §§ 1395f and 1395x(v).

References Antos, Joseph R., Mark V. Pauly, and Gail R. Wilensky. 2012.“Bending the Cost Curve through Market-Based Incentives.” New England Journal of Medicine 367 (10): 954– 58. Arrow, Kenneth J. 1963. “Uncertainty and the Welfare Economics of Medical Care.” American Economic Review 53 (5): 941– 73. Ball, Robert M. 1995. “Perspectives On Medicare.” Health Affairs 14 (4): 62– 72. Blustein, Jan. 1995. “Medicare Coverage, Supplemental Insurance, and the Use of Mammography by Older Women.” New England Journal of Medicine 332 (17): 1138–43. Centers for Medicare and Medicaid Services. “Pioneer Accountable Care Organization Model: General Fact Sheet”. Accessed November 2014. http://in novation.cms.gov/initiatives/Pioneer-ACO-Model/. Chulis, George S. 1991. “Assessing Medicare’s Prospective Payment System for Hospitals.” Medical Care Research and Review 48 (2): 167– 206. Colla, Carrie H., David E. Wennberg, Ellen Meara, Jonathan S. Skinner, Daniel Gottlieb, Valerie A. Lewis, Christopher M. Snyder, and Elliott S. Fisher. 2012. “Spending Differences Associated With the Medicare Physician Group Practice Demonstration.” Journal of the American Medical Association 308 (10): 1015– 23. Cotterill, Philip G., and Barbara J. Gage. 2002. “Overview: Medicare Post-Acute Care since the Balanced Budget Act of 1997.” Health Care Financing Review 24 (2): 1– 6. Damberg, Cheryl L., Kristiana Raube, Stephanie S. Teleki, and Erin dela Cruz. 2009. “Taking Stock of Pay-for-Performance: A Candid Assessment from the Front Lines.” Health Affairs 28 (2): 517– 25. Eijkenaar, Frank, M. Emmert, M. Scheppach, and O. Schöffski. 2013. “Effects of pay for performance in health care: A systematic review of systematic reviews.” Health Policy 110 (2): 115– 30. Epstein, Arnold M., Colin B. Begg, and Barbara J McNeil. 1986. “The Use of Ambulatory Testing in Prepaid and Fee-for- Service Group Practices.” New England Journal of Medicine 314 (17): 1089– 94.

Federal Provider Payment Reform

243

Feinglass, Joe, and James J. Holloway. 1991. “The Initial Impact of the Medicare Prospective Payment System on US Health Care: A Review of the Literature.” Medical Care Research and Review 48 (1): 91–115. Frosch, Dominick L., Suepattra G. May, Katharine A. S. Rendle, Caroline Tietbohl, and Glyn Elwyn. 2012. “Authoritarian Physicians And Patients’ Fear Of Being Labeled ‘Difficult’ Among Key Obstacles To Shared Decision Making.” Health Affairs 31 (5): 1030– 38. Greene, Jessica, Judith Hibbard, James F. Murray, Steven M. Teutsch, and Marc L. Berger. 2008. “The Impact Of Consumer-Directed Health Plans On Prescription Drug Use.” Health Affairs 27 (4): 1111–19. Greenfield, Sheldon, Eugene C. Nelson, Michael Zubkoff, Willard Manning, William Rogers, Richard L. Kravitz, Adam Keller, Alvin R. Tarlov, and John E. Ware. 1992. “Variations in Resource Utilization Among Medical Specialties and Systems of Care: Results From the Medical Outcomes Study.” Journal of the American Medical Association 267 (12): 1624– 30. Hart, Oliver and Bengt Holmstrom. 1987. The Theory of Contracts. Edited by T. F. Bewley. Vol. 12, Advances in Economic Theory: Fifth World Congress. Cambridge: Cambridge University Press. Hellinger, Frederick J. 1996. “The Impact of Financial Incentives on Physician Behavior in Managed Care Plans: A Review of the Evidence.” Medical Care Research and Review 53: 294– 314. Hemenway, David, Alice Killen, Suzanne B. Cashman, Cindy Lou Parks, and William J. Bicknell. 1990. “Physicians’ Responses to Financial Incentives— Evidence from a for-Profit Ambulatory Care Center.” New England Journal of Medicine 322 (15): 1059– 63. Hibbard, J. H., J. Stockard, and M. Tusler. 2003. Does Publicizing Hospital Performance Stimulate Quality Improvement Efforts? Health Affairs 22: 84– 94. Hickson, G. B., W. A. Altemeier, and J. M. Perrin. 1987. “Physician Reimbursement by Salary or Fee-for- Service: Effect on Physician Practice Behavior in a Randomized Prospective Study.” Pediatrics 80 (3): 344– 50. Hsiao, William C., Peter Braun, Daniel L. Dunn, Edmund R. Becker, Douwe Yntema, Diana K. Verrilli, Eva Stamenovic, and Shiao-Ping Chen. 1992. “An Overview of the Development and Refi nement of the Resource-Based Relative Value Scale: The Foundation for Reform of US Physician Payment.” Medical Care 30 (11): NS1–NS12. Hurley, Jeremiah. 2000. “An Overview of the Normative Economics of the Health Sector.” In Handbook of Health Economics, vol. 1A, edited by Anthony J. Culyer and Joseph P. Newhouse, 56–188. Oxford: North-Holland. Hussey, Peter S., M. Susan Ridgely, and Meredith B. Rosenthal. (2011). “The PROMETHEUS Bundled Payment Experiment: Slow Start Shows Problems in Implementing New Payment Models.” Health Affairs 30 (11): 2116– 24.

244

Meredith B. Rosenthal

Jencks, Stephen F., Mark V. Williams, and Eric A. Coleman. 2009. “Rehospitalizations among Patients in the Medicare Fee-for- Service Program.” New England Journal of Medicine 360 (14): 1418– 28. Jennison. K., and R. P. Ellis. 1987. “Comparison of Psychiatric Service Utilization in a Single Group Practice.” The Economics of Mental Health Services: Advances in Health Economics and Health Services Research, vol. 8, edited by Thomas G. McGuire and Richard Scheffler, 175– 94. Greenwich, CT: JAI Press. Jha, Ashish K., Karen E. Joynt, E. John Orav, and Arnold M. Epstein. 2012. “The Long-Term Effect of Premier Pay for Performance on Patient Outcomes.” New England Journal of Medicine 366 (17): 1606–15. Kaiser Family Foundation and Agency for Health Care Research and Quality. 2004. National Survey on Consumers’ Experiences with Patient Safety and Quality Information. Washington, DC: Kaiser Family Foundation. Kerr, S. 1975. “On the Folly of Rewarding A, While Hoping for B.” Academy of Management Journal 18: 769– 83. Lindenauer, Peter K., Denise Remus, Sheila Roman, Michael B. Rothberg, Evan M. Benjamin, Allen Ma, and Dale W. Bratzler. 2007. “Public Reporting and Pay for Performance in Hospital Quality Improvement.” New England Journal of Medicine 356 (5): 486 – 96. Ma, C. Albert. 1994. “Health Care Payment Systems: Cost and Quality Incentives.” Journal of Economics and Management Strategy 3 (1): 93–112. Mayes, Rick, and Robert A. Berenson. 2008. Medicare Prospective Payment and the Shaping of U.S. Health Care. Baltimore: Johns Hopkins University Press. McClellan M. 1997. “Hospital Reimbursement Incentives: An Empirical Analysis.” Journal of Economics and Management Strategy 6 (1), 91–128. McGuire, Thomas G. 2000. “Physician Agency.” In Handbook of Health Economics vol. 1A, edited by Anthony J. Culyer and Joseph P. Newhouse. Oxford: North- Holland. Miller, R. H., and H. S. Luft. 1997. “Does Managed Care Lead to Better or Worse Quality of Care?” Health Affairs 16 (5): 7– 25. Mullen, Kathleen, Richard G. Frank, and Meredith B. Rosenthal. 2010. “Can You Get What You Pay For? Pay for Performance and the Quality of Health Care Providers.” RAND Journal of Economics 41 (1): 64– 91. Newhouse Joseph P., and the Health Insurance Experiment Group. 1993. Free for All? Lessons from the RAND Health Insurance Experiment. Cambridge, MA: Harvard University Press. Newhouse, Joseph P. 2002. Pricing the Priceless: A Health Care Conundrum. Cambridge, MA: MIT Press. O’Connor, A. M., C. L. Bennett, D. Stacey, M. Barry, N. F. Col, K. B. Eden, and V. A. Entwistle. 2008. “Decision Aids for People Facing Health Treatment or Screening Decisions.” The Cochrane Collaboration published in the

Federal Provider Payment Reform

245

Cochrane Library 3. http://summaries.cochrane.org/CD001431/COMMUN_ decision-aids-to-help-people-who-are-facing-health-treatment-or-screening -decisions. Pauly, Mark V. 2000. “Insurance Reimbursement.” In Handbook of Health Economics vol. 1A, edited by Anthony J. Culyer and Joseph P. Newhouse. 537– 60. Oxford: North-Holland. Petersen, Laura A., LeChauncy D. Woodard, Tracy Urech, Christina Daw, and Supicha Sookanan. 2006. “Does Pay-for-Performance Improve the Quality of Health Care?” Annals of Internal Medicine 145 (4): 265– 72. Roblin, Douglas W., Richard Platt, Michael J. Goodman, John Hsu, Winnie W. Nelson, David H. Smith, Susan E. Andrade, and Stephen B. Soumerai. 2005. “Effect of Increased Cost-Sharing on Oral Hypoglycemic Use in Five Managed Care Organizations: How Much is Too Much?” Medical Care 43 (10): 951–59. Rosenthal, Meredith B. 1999. “Risk Sharing and Delegation in Managed Behavioral Health Care.” Health Affairs 18 (5): 204–13. ———. 2000. “Risk Sharing and the Supply of Mental Health Services.” Journal of Health Economics 19 (6): 1047– 65. Rosenthal, Meredith B., Richard G. Frank, Li Zhonghe, and Arnold M. Epstein. 2005. “From Concept to Practice: Early Experience with Pay-forPerformance.” Journal of the American Medical Association 294 (14): 1788– 93. Schneider, E. C., and A. M. Epstein. 1998. “Use of Public Performance Reports: a Survey of Patients Undergoing Cardiac Surgery.” Journal of the American Medical Association 279 (20) :1638–42. Scott, Anthony, Peter Sivey, Driss Ait Ouakrim, Lisa Willenberg, Lucio Naccarella, John Furler, and Doris Young. 2011. “The Effect of Financial Incentives on the Quality of Health Care Provided by Primary Care Physicians.” Cochrane Database Systematic Reviews 9. Shleifer, Andrei. 1985. “A Theory of Yardstick Competition.” RAND Journal of Economics 16 (3): 319– 27. Sinaiko, Anna D., Diana Eastman, and Meredith B. Rosenthal. 2012. “How Report Cards on Physicians, Physician Groups, and Hospitals Can Have Greater Impact on Consumer Choices.” Health Affairs 31 (3): 602–11. Stearns, Sally C., Barbara L. Wolfe, and David A. Kindig. 1992. “Physician Responses to Fee-for- Service and Capitation Payment.” Inquiry 29 (4): 416–425. Struijs, Jeroen N., and Caroline A. Baan. 2011. “Integrating Care through Bundled Payments—Lessons from the Netherlands.” New England Journal of Medicine 364 (11): 990– 991. Tamblyn, Robyn, Rejean Laprise, James A. Hanley, Michael Abrahamowicz, Susan Scott, Nancy Mayo, and Jerry Hurley. 2001. “Adverse Events Associated with Prescription Drug Cost- Sharing among Poor and Elderly Persons.” Journal of the American Medical Association 285 (4): 421–429.

246

Meredith B. Rosenthal

Van Herck, Pieter, Delphine De Smedt, Lieven Annemans, Roy Remmen, Meredith B. Rosenthal, and Walter Sermeus. 2010. “Systematic Review: Effects, Design Choices, and Context of Pay-for-Performance in Health Care.” BMC Health Services Research 10 (1): 247. http://www.biomedcentral.com/content /pdf/1472-6963-10-247.pdf. Werner, Rachel M. and David A. Asch. 2005. “The Unintended Consequences of Publicly Reporting.” Journal of the American Medical Association 293(10): 1239–1244.

Chapter ten

The Role of Technology in Expenditure Growth in Healthcare Amitabh Chandra and Jonathan Holmes

Introduction

T

he costs of healthcare are becoming an acute issue for governments around the world. This is particularly the case in the United States, which boasts both the highest level of health spending as a proportion of GDP in 2012 and the highest rate of growth in the share of GDP devoted to health care in the rich world since 1980 (Chandra & Skinner 2012). The United States has maintained this high growth rate in spending despite having slower improvement in health outcomes than other rich countries, even after adjusting for smoking and obesity trends (Muenning & Glied 2010). While health spending is lower in other countries, the trend towards higher expenditures is a global one, and expenditures are projected to rise in most countries because of an aging population and technological change. There are many causes of spending growth in the healthcare sector in the United States. Costs are increasing partly due to an aging population, because elderly patients are more expensive to treat on average than the young. Rising obesity rates also contribute to poorer health outcomes and higher costs, although this is offset by falling smoking rates. Finally, trends in the cost of fraud, medical malpractice lawsuits, and administration fees may also explain some of the growth. Nevertheless, even after controlling for all of these trends, most of the growth in U.S. health expenditures remains unaccounted for. This remaining growth is

250

Amitabh Chandra and Jonathan Holmes

Figure 10.1 A simple model of the effect of the adoption of medical technology on health. The slope of the curve represents the “amount” of health (which can be measured in Quality Adjusted Life Years) gained from a $1 increase in spending.

due to changes in technology, from the introduction and diffusion of new techniques and procedures into the market (Newhouse 1992). Spending an ever greater fraction of our income on medical technology may be warranted if this spending results in better health outcomes and increased longevity. If new medical drugs and procedures have increased benefits with fewer side effects, the benefits to increased spending on these new technologies may outweigh their costs. Higher medical expenditures may also be justified as incomes grow, because yearly income has a declining marginal value. As per capita incomes grow, more people may choose to put a rising share of their income into increasing their lifespan rather than buying expensive luxuries. If these factors are the primary sources of increasing spending on new medical technology, then the spending growth in the United States is warranted. In order for increased spending to be worthwhile, though, we must fi rst ask whether it has a positive impact on patient health. The traditional way to think of this question is to imagine a health production function such as the one in Figure 10.1. This graph shows the maximum possible level of health outcomes (measured in Quality Adjusted Life Years or some other index of health quality) that can be achieved from a given level of spending at a particular point in time. Because the technologies that have the greatest benefit (the “low-hanging fruit”) are the fi rst to be used, the marginal productivity of health technology declines

The Role of Technology in Expenditure Growth in Healthcare

251

as spending rises. Below point A, extra spending is used on high-net-gain technologies that do not cost a lot of money, such as aspirin for heart attack patients. Between point A and point B the benefits of new technologies decline, but additional spending still yields some health benefits. Once spending has passed point B, however, additional technologies cost money but no longer make a difference for patient health. For example, after a certain point additional diagnostic tests no longer yield any useful information to doctors, even though they do not hurt patient health. Finally, technologies introduced between point C and point D are actually damaging to patient health. An example of this type of treatment is aggressive end- of-life care for non-small- cell lung cancer in elderly patients. In a recent randomized trial, patients undergoing the intensive procedure actually lived less long than patients who received cheaper early palliative care (Temel et al. 2010). Interpreting recent studies on health spending by applying the simplistic model from Figure 10.1 yields confl icting results. Some studies have found that increasing health costs in the aggregate led to better patient outcomes (Cutler 2004; Lichtenberg 2001). These studies would suggest that the health economy is at point A, and that greater expenditures may be warranted. On the other hand, when U.S. health regions are compared, the relationship between per-person regional health costs and health outcomes is either zero or negative, even after risk adjusting for other factors (Chandra, Gruber, and McKnight 2010). This would suggest that our economy is somewhere between points B and C. In the highly studied case of heart attack patients, one study found that increasing expenditures between 1984 and 1998 led to a large decrease in mortality (Cutler and McClellan 2001). However, a more recent study found that expenditure growth between 1996 and 2002 was no longer associated with increased survival rates (Skinner, Staiger, and Fisher 2006). To reconcile these claims, we need a better model. The problem with Figure 10.1 is that it does not take the huge amount of heterogeneity in the medical marketplace into account. Per person health spending in the United States varies from $6,432 in Anchorage, Alaska to $15,568 in Miami, Florida.1 As a consequence, some regions may be at point A in the graph, whereas other regions may be at point D. Moreover, there is also a huge amount of variation from region to region in the practices and technologies that are used. By adopting different technologies and practices, and by having access to different resources, some health regions may be able to achieve greater health outcomes simply and for less money.

252

Amitabh Chandra and Jonathan Holmes

Figure 10.2 A disaggregated model of the health economy. In these graphs, points A, B, C, and D each represent different hospitals, and are located on separate health production functions determined by fi xed factors at each hospital. In both panel A and panel B, health spending is uncorrelated with health outcomes when compared across hospitals. In panel A, each hospital has the ability to improve patient outcomes through higher average spending, but in panel B, the adoption of more expensive technologies will actually hurt patients’ health in all regions.

Because of this large regional heterogeneity, thinking about the health economy in the aggregate as in Figure 10.1 is misleading. To illustrate why, consider the modification made in Figure 10.2, which simply assumes that four hospitals A, B, C, and D each have a different health production function. Variation in hospital production functions may be driven by differences in factors beyond the control of hospitals, such as the availability of local high- quality universities, or may be due to past investments in physical infrastructure and organizational structure. Consistent with current U.S. data, there is no relationship between average hospital spending and average health, but each panel implies a different policy recommendation. In panel A, an increase in health spending would improve outcomes, because hospitals still have access to productive technology. In contrast, an increase in health spending would decrease health outcomes in panel B, because all productive technology has been used. Because both of these panels are consistent with observed trends, it is not possible to know whether we should increase health spending using only aggregate statistics. The determinants of high medical productivity are also missed in the simplistic aggregate model. In Figure 10.2, hospital B is very produc-

The Role of Technology in Expenditure Growth in Healthcare

253

tive, because it is able to achieve excellent health outcomes at a low cost. While research on what distinguishes a high-productivity hospital like B from a low-productivity hospital like D is still ongoing, initial evidence is that technology use plays a big role. In the case of heart attacks, highproductivity regions tend to be the ones that use proven cheap and effective medical technologies like aspirin and beta blockers (Skinner and Staiger 2009). In contrast, low-productivity hospital D may have invested heavily in expensive but unproven technologies such as proton beam cancer therapy. The key difference between panel A and panel B in Figure 10.2, and between hospital D and hospital B, is the medical technologies used in each institution. Under the current institutional structure of the medical sector, the incentives to use only the best and most cost- effective technologies available are weak. In a fee-for-service environment, 2 where patients do not directly pay for the care they receive, physicians have an incentive to overprescribe care. As a consequence, many procedures with negligible or unknown value continue to be used, even when these technologies have a large price tag. At the same time, capacity constraints and skewed reimbursement schemes may cause other technologies with a high net benefit to be underused. Hospitals may balk at shifting patients from high- cost surgery to low- cost rehabilitation when this would also eliminate revenue streams due to lost reimbursements. Thus, there is nothing that guarantees that we are getting the most value per dollar of spending given current technologies. In this environment, it is plausible that we can reduce costs while maintaining health outcomes simply by shifting medical expenditures from low-productivity technologies to high-productivity technologies.

A Typology of Medical Technologies In considering which medical technologies have the greatest average cost effectiveness, the fi rst thing to understand is that a given technology does not have the same effect on all patients. For example, many surgeries can be of greater benefit to younger patients, and drugs have different effects on patients depending on their comorbid conditions. Patients who receive a given treatment can thus fall into one of the following groups:

Amitabh Chandra and Jonathan Holmes

254

Group A comprises patients who receive the treatment and for whom it is cost effective. For these patients, the benefits clearly exceed the costs of administering the treatment. Group B comprises patients who receive the treatment but for whom it is not cost effective. These patients have benefits that are either small or negligible compared to the cost of administering the treatment. Group C comprises patients for whom the treatment is of unknown value.

A given medical technology may have patients who fall into only one group, or it may have patients in all three groups. An example is angioplasty, known as PCI (percutaneous coronary intervention), which is an invasive procedure to remove blockages of arteries related to heart attacks or coronary heart failure. Group A patients for PCI are young patients who receive treatment in the fi rst twelve to twenty-four hours following a heart attack, for whom the procedure is highly cost effective (Hartwell et al. 2005; Hochman et al. 2006). In contrast, the benefits to survival and functioning for patients with stable coronary disease are small to nonexistent relative to the best available alternative (Boden et al. 2007; Weintraub et al. 2008). There is also suggestive evidence that older patients have either small or negligible increases in survival rates (Chandra and Staiger 2007). Given its high cost—typically above $15,000—elderly patients and those who have stable coronary disease would fall into group B. The average productivity of a given medical technology is thus dependent on the proportion of patients treated with it who fall into each of groups A, B, and C respectively. Treatments for which most patients fall into group A are highly productive, whereas treatments for which most patients fall into group B have a low average productivity. Treatments with many patients in group C may benefit patients, but they may also hurt patients or be ineffective. While a certain amount of experimentation with such technologies with uncertain benefits may be warranted where no proven alternative exists or where additional evidence is necessary, heavy reliance on unproven technologies is unlikely to be cost effective. Chandra and Skinner (2012) classify these different medical procedures and technologies into three categories based on how the benefits of a procedure are distributed. Category 1 technologies are those in which most patients fall into group A. Category 2 procedures have some patients in group A, but also have many patients in group B. Finally,

The Role of Technology in Expenditure Growth in Healthcare

255

category 3 procedures have most of their patients in either group B or group C. Category 1—“Home Run” Technologies Category 1 technologies are those for which there is a straightforward delineation between patients who clearly benefit from treatment and those who clearly do not. There exist very few gray areas in treatment decisions where it is unclear whether a patient will benefit. As a consequence, there are very few group B or group C patients for a category 1 technology, and cost effectiveness is high. Many category 1 technologies are classified as such because their costs are very low but their benefits are large. The canonical example of this is routine hand washing and use of gloves in hospitals, which costs almost nothing but which provides a high value to patients by preventing the spread of diseases (Lister 1867). Another example is the use of aspirin and beta blockers, both of which are highly effective and cost very little for doctors to administer, for the treatment of heart attack patients (Yusuf et al. 1985). In both of these cases, costs are so low and average benefits are so high that overuse of the technologies has a negligible impact on average cost effectiveness. However, category 1 drugs, for example antiretroviral drugs, are not necessarily cheap. For those who have HIV/AIDS, antiretroviral drugs have a very high value relative to their cost; one estimate puts the benefits in terms of life years gained at twenty times the drugs’ cost of development (Philipson and Jena 2006). But for those who do not have HIV/ AIDS, antiretroviral drugs provide negligible benefits with substantial side effects. As a consequence, the potential for overuse is small. The key feature of a category 1 technology is a strongly negative slope of the value function, as shown in Figure 10.3. For this technology, the value per patient is very high relative to cost for some patients, but then quickly drops off for patients not in the target group. In a fee-for-service environment where patients do not pay for a procedure, a well-meaning physician will administer any treatments that benefit the patient. On the diagram, this includes all patients for whom the value function has a value greater than zero. Of these patients, group A, which is the group of patients for whom the benefits of the treatment are greater than the costs, is much larger than group B. As a consequence, the average cost effectiveness of a category 1 technology is high.

256

Amitabh Chandra and Jonathan Holmes

Figure 10.3 Costs and benefits of a category 1 technology. The value of a procedure per patient is equal to the number of Quality Adjusted Life Years gained times the value of a Quality Adjusted Life Year. Source: Adapted from Chandra and Skinner (2012).

In our current policy framework, category 1 technologies are those that are most likely to be used at close to their optimal level, because there are very few areas where either overuse or underuse is likely. However, there are still some cases in which policy could encourage better uptake of category 1 technologies. A puzzling example is the case of beta blockers, a highly effective category 1 treatment for heart attack patients. While the effectiveness of beta blockers was established by 1985 (Yusuf et al. 1985), the median state-level use for patients who would benefit was only 68 percent (Jencks, Huff, and Cuerdon 2003). Clearly, something is hampering the diffusion of certain types of highly effective treatments. More importantly, in a system where there is a great fi nancial incentive to develop technologies that can be billed, there are only weak incentives to develop many category 1 technologies, like hand washing or innovative management structures, that are not strictly a billable category of goods. As a consequence, hospitals may hesitate to make use of new technologies that cannot be reimbursed, or that reduce the usage of other procedures that can be reimbursed at a higher rate. To take one example, a back pain clinic elected to steer patients toward low- cost rehabilitation rather than referring them to a hospital for back surgery. While health outcomes improved and costs decreased, the hospital lost a substantial revenue stream and was forced to petition the insurance company for lost profits (Fuhrmans 2007).

The Role of Technology in Expenditure Growth in Healthcare

257

Category 2 Technologies Category 2 technologies are those with varying results, which yield positive net benefits for some patients, but only minimal benefits to a large number of other patients. Because category 2 technologies have many patients in group B, it is possible that the average cost effectiveness of the technology across all patients is negative, even when some patients benefit a great deal. An example of a category 2 treatment is angioplasty. While angioplasty can provide substantial benefits to younger patients who have had a heart attack within the previous day, it provides almost no benefit to older patients and those with stable coronary disease. Since about onethird of patients who receive angioplasties have stable coronary disease, the average cost effectiveness of the procedure across all patients is low. Many forms of medical imaging are also examples of category 2 technologies. Medical imaging technologies such as computed tomography, magnetic resonance imaging, and positron- emission tomography clearly have substantial benefits for some patients. In recent years, though, there has been substantial growth in their use, and some studies suggest that these new uses are of low marginal effectiveness (Medicare Payment Advisory Committee 2003; M. Miller 2005). Group B patients of medical imaging are those for whom the chance that a medically useful piece of information will be gleaned from the procedure is so small that medical imaging is not worth the cost, both in terms of price and in terms of side effects due to radiation exposure. The main characteristic of a category 2 technology is a value function that has a weakly negative slope for those patients for whom benefits are less than the cost of the treatment. A typical example is found in Figure 10.4. Just as with category 1 treatments, there are still patients in group A who clearly benefit from the treatment. However, there also exists a large set of group B patients who do not benefit substantially. In a fee-for-service environment, these group B patients will still receive the treatment from a well-meaning physician because it does benefit the patient slightly, even though the benefits do not warrant their price tag. As a result, the average cost effectiveness of a category 2 technology across all patients will be lower than a similar category 1 technology, and may actually be negative. In a fee-for-service framework, category 2 technologies clearly have a tendency to be overused, because physicians will prescribe any technology

Figure 10.4 Costs and benefits of a category 2 technology. The value of a procedure per patient is equal to the number of Quality Adjusted Life Years gained times the value of a Quality Adjusted Life Year. Source: Adapted from Chandra and Skinner (2012).

Figure 10.5 The effect of different physician beliefs on utilization of a category 2 technology. The dotted line represents slightly more favorable beliefs about a technology than the technology merits. Small changes in beliefs make a big difference in utilization when the slope of the value per patient curve is small in absolute value. Source: Adapted from Chandra and Skinner (2012).

The Role of Technology in Expenditure Growth in Healthcare

259

they believe to have a nonzero benefit. Moreover, the usage of category 2 technologies can be much more easily altered by slight changes in beliefs about their effectiveness. Evidence suggests that there exist wide variations in the perceived effectiveness of many medical technologies, particularly for category 2 and category 3 treatments that have not been extensively studied. Figure 10.5 shows a case in which physicians have a slightly higher opinion of a category 2 technology than it actually merits. In this case, overuse increases substantially, because there is a large number of patients in the gray area who are now perceived to benefit from the treatment. This characteristic makes category 2 treatments particularly susceptible to marketing campaigns, which need only be effective at convincing patients and physicians that procedures are slightly more effective than they actually are. This also means that increased information sharing about ineffective treatments may be able to reduce utilization of ineffective category 2 technologies on Group B patients. Potentially the biggest problem with category 2 technologies, however, is that there is a strong financial incentive for companies to develop, patent, and market these new techniques. Unlike many category 3 technologies, category 2 technologies have a clear benefit for a small number of patients. This makes it easier for these techniques to be approved for use and covered by insurance plans. Once they are covered by insurance, there is very little to prevent their use in marginal patients who may not benefit much (if at all). As a consequence, companies that develop and sell category 2 technologies can expect to have access to a much larger market than category 1 technologies that have a similar number of group A core users. Category 3 Technologies Category 3 technologies have dubious or unknown benefits for most or all patients. Despite being shown to be ineffective, or despite our lack of knowledge about their actual benefits, many category 3 procedures are regularly performed and reimbursed by Medicare and private insurance. Some category 3 procedures are known to be of minimal benefit. A typical example is arthroscopic surgery to treat osteoarthritis of the knee. Arthroscopic surgery works by making an incision in the knee to allow a doctor to clean out any debris that may be promoting inflammation and worsening arthritis symptoms. While this procedure was widely used in the past, a randomized trial has shown that the surgery has no discernible benefits when compared with a sham surgery (Moseley et al. 2002).

260

Amitabh Chandra and Jonathan Holmes

In consequence, the alleged benefits of the surgery could only be attributed to the placebo effect and any rehabilitation that was performed following the procedure. Many routine and mundane procedures are also category 3. For example, many types of preoperative screening have been shown to be of limited value. Archer, Levy, and McGregor (1993) find that preoperative Xrays only found abnormalities which were enough to change physicians’ disease management strategy in 0.1 percent of cases. Schein et al. (2000) also fi nd that for cataract surgery, patients who underwent preoperative tests had results that were no better than those who did not. Faced with evidence like this, some have argued that preoperative testing is of limited value for almost all cases (Smetana and Macpherson 2003). Given that patients do not have to pay for these tests directly, medical providers have incentives to substantially overprescribe these techniques. Many other technologies are also considered to be category 3 simply because we currently do not know whether or not they provide any benefit at all. An example of this type of treatment is proton beam therapy for prostate cancer. While facilities to perform this technique can cost up to $100 million, there have been no randomized trials on the therapy as of 2012 and we still do not know whether it actually performs better than conventional radiation in practice (Terasawa et al. 2009). Another less expensive example is PSA testing for prostate cancer. While PSA testing may increase the detection rate of cancer, the substantial false positive rate and existence of known side effects are concerns. For young, healthy patients, it is unknown whether the early detection of cancers provided by regular PSA tests is greater than the negative side effects over the long term (Harris and Lohr 2002). 3 As a consequence, PSA testing rates vary substantially across the United States and the developed world. Finally, many medical technologies, like postacute care in skilled nursing facilities, are considered category 3 treatments because they are impossible to evaluate using randomized trials. For these types of care, the question is not whether they are necessary (because they clearly are in many cases), but rather how much intervention is necessary. In the case of skilled nursing facilities, there is little consensus on the ideal length of stay following surgery. As a consequence, standards vary widely across regions, and postdischarge care is responsible for a substantial amount of the variation in costs between U.S. health regions for Medicare patients (D. Miller et al. 2011). However, those regions that perform high amounts of postdischarge care do not have better health outcomes on average than

The Role of Technology in Expenditure Growth in Healthcare

261

regions that perform less care. Since postdischarge care represents a substantial part of spending growth over the past twenty years (Chandra, Holmes, and Dalton 2013), it is very important to use what data we have to set reasonable standards on usage of these types of technologies.

Why Are Costs Increasing in the United States? Armed with this categorization of medical technology, it is natural to ask what portion of the dramatic increases in both costs and longevity can be attributed to each type of medical technology. If the drivers of spending growth and improved health are different, it could be possible to decrease costs while still maintaining or improving health outcomes. To answer this question, we first look to a study by Ford et al. (2007) that explained the decline in mortality from coronary disease from 1980 to 2000 by its specific causal factors. As shown in Table 10.1, 43.8 percent of the decline in death rates can be attributed to changing health-risk

Table 10.1. Number of Deaths Prevented or Postponed from Coronary Disease 1980– 2000 Number of deaths prevented/postponed

Percent of total mortality decline

Type of medical/surgical treatment or risk factor change

209,000

61.2%

− 59,370

−17.4%

149,630

43.8%

Subtotal: Deaths prevented or postponed because of health risk factors

83,285

21.9%

Category 1: Aspirin, heparin, warfarin, antihypertensives, beta-blockers, diuretics

45,225

13.2%

Category 1+: Statins, ACE inhibitors, IIb/IIIa antagonists, thrombolytics

30,830

11.5%

Category II: Angioplasty/stents, bypass surgery (CABG), cardiopulmonary resuscitation, cardiac rehabilitation

159,340

46.6%

Subtotal: Deaths prevented or postponed by medical/surgical treatments

32,775 341,745

9.6% 100.0%

Health risk reduction: Declines in prevalence of smoking, hypertension, cholesterol, physical inactivity Health risk increase: Rise in prevalence of bodymass index (BMI) and diabetes

Unexplained by model Total deaths prevented or postponed

Source: Data are taken from a study by Ford et al. (2007). Table is reprinted from Chandra and Skinner (2012).

262

Amitabh Chandra and Jonathan Holmes

factors. Whereas BMI and diabetes increased risk, this was more than offset by reductions in other risk factors such as smoking and cholesterol. Category 1 treatments, such as aspirin and beta blockers, are responsible for 21.9 percent of the decline; 13.2 percent is attributable to treatments like ACE inhibitors that we consider to be category 1+ because of their high value-to- cost ratio, even though they may be overused in some circumstances. Category 2 procedures such as bypass surgery are responsible for 11.5 percent of the decline in mortality; the last 9.6 percent of mortality reduction cannot be explained by the study, and may thus be attributable to category 3 technologies with unknown benefits. So for the example of coronary disease, category 1 technologies are the key drivers of increased longevity and category 2 and 3 technologies are responsible for only up to 20 percent of longevity gains. Finding which types of technologies drive spending is difficult. Clearly, some of the most effective technologies, such as aspirin and handwashing, constitute a very small portion of spending. There have also been many anecdotal cases of extremely expensive but unproven techniques diffusing widely in the U.S. health system from other countries. For example, the United States has a much higher adoption of robotic surgical technology; over 75 percent of the most common robotic technology systems are located in the United States, even though it is unknown whether these systems improve overall longevity (Barbash and Glied 2010). U.S. spending on category 3 technologies at the end of life is also high, as is evidenced by the fact that 11.3 percent of people over 85 die in an intensive care unit in the United States, compared to only 1.3 percent in the United Kingdom (Wunsch et al. 2009). Because elderly patients are more likely to suffer from a range of health problems, the potential benefits to treating one particular condition are lower (Welch et al. 1996). Finally, the United States also leads the world in spending on category 3 drugs such as Avastin, which was approved for use by the FDA to treat blioblastoma (brain tumors) based on a single uncontrolled study but was denied approval by the European Medicines Agency.4 Aggregate estimates also support the hypothesis that category 3 procedures are the main causes of spending growth. In a recent paper, we looked at spending growth for patients in Medicare who are diagnosed with a hip fracture, heart attack, or congestive heart failure. We found that a large fraction of the recent cost-growth trends among the U.S. Medicare population is driven by spending on long-term postacute care (Chandra, Holmes, and Dalton 2013), which is category 3 spending due

The Role of Technology in Expenditure Growth in Healthcare

263

to our lack of knowledge about what level of spending is appropriate. If this trend towards increased postacute care arose because hospitals are reducing the number of Group B patients who are treated with high- cost acute care, then it may have reduced cost growth. Much of this spending may also be high value because it improves patient quality of life. But the enormous variation in utilization rates for postacute care across the country suggests that there is a huge gray area where benefits are uncertain at best. So although our understanding of the drivers of both costs and health outcomes is still evolving, current studies suggest that category 1 procedures drive mortality gains and category 3 procedures drive cost growth. This difference explains why some studies fi nd that increased expenditures improve health outcomes, but others do not. Health outcomes rise substantially when use of category 1 technologies is growing, but do not rise when spending is driven by category 3 technologies or category 2 technologies performed on group B patients. In the aggregate, the fact that outcomes and costs are uncorrelated suggests that there is little correlation between uptake of category 1 technologies and overall spending among U.S. health regions.

Policy Options A good analysis of healthcare policy should take into account both the effect of the policy on current spending and the compounding effect of the policy on the incentive to invest in new technologies. This chapter has listed many questionable spending decisions, from unnecessary end- oflife care to unproven proton beam therapy. But only reducing these types of spending in the present will not affect spending growth unless the incentives to innovate are also changed. Practices and technologies used by medical practitioners develop slowly over time, and are a product of the incentive structures created by the way new medical technologies are reimbursed. So, for example, the introduction of Medicare in 1965 led to a much higher growth rate in spending than expected, which is probably because it induced changes in market structure (Finkelstein 2007). As we have shown above, there are strong fi nancial incentives to develop new category 2 or category 3 technologies, and the incentive to create a money-saving category 1 innovation in a delivery system, let alone a cheaper version of an existing technology, is much smaller. The following

264

Amitabh Chandra and Jonathan Holmes

are a few different policy options to reduce costs while still maintaining or improving health outcomes. Allow Insurance Companies to Treat Category 1, 2, and 3 Spending Differently One theoretical way to control costs is for insurance companies to reduce their reimbursements for ineffective procedures. In practice, though, insurance companies are constrained from doing this by the legal system. Legal action has forced insurance companies to reimburse for procedures that have been found to be ineffective in the past (Ferguson, Dubinsky, and Kirsch 1993). In addition, Medicare is not allowed to take costs into account in its decision making; this has a tendency to increase Medicare spending on procedures that are not cost effective and also makes it harder for private insurance plans to justify denying coverage for these same procedures. Finally, state laws require insurance companies to cover any procedure deemed medically necessary by a physician, which prevents insurance companies from setting out a list of predefi ned conditions to reimburse. As a result, insurance companies are not able to control costs by offering slimmed- down insurance plans to frugal consumers. One way around this legal issue would be for insurance companies to charge different patient copayments based on whether a treatment has been shown to be cost effective. This approach has been used to promote the use of certain category 1 procedures (see, e.g., Chernew et al., 2008), although it has not been used yet to decrease the use of treatments with small or uncertain benefits. For elective treatments in which the treatment options do not differ much in their survival benefits, letting patients bear a larger portion of the costs of the expensive treatment option may empower them to make decisions that better reflect their personal beliefs about the underlying trade- offs. Take robotic surgery, for example, which has been proven to reduce postoperative pain and hospital length of stay, but has not yet been proven to have any long-term longevity gains compared with conventional surgery (Barbash & Glied, 2010). Charging a higher copayment to patients who elect to take robotic surgery would allow those patients who care more about the reduced pain and improved recovery times to elect to use it, while those who do not care for these factors would still receive the effective conventional option. Money saved could then be passed back to consumers through lower insurance premiums.

The Role of Technology in Expenditure Growth in Healthcare

265

Reward Health Outcomes, Not Quantity of Procedures Existing reimbursement schemes calculate reimbursements based on the number of procedures that are performed or the amount of medical equipment that is bought. If hospitals were instead reimbursed for health outcomes, this could encourage them to increase their use of high-value treatments and reduce their use of low-value treatments. While quality-based payments are beginning to be used, reimbursement for quality is still a very small part of Medicare reimbursements and private insurance programs. Part of this is because rewarding quality can be difficult; indicators of health outcomes are difficult to determine and few markers of quality (such as survival rates) are both objective and readily available. Moreover, it is hard to construct measures of quality that control for a patient’s initial health, which is essential in order to dissuade hospitals from inflating their quality statistics by treating only relatively healthy patients with few comorbid conditions and other health risks. Another form of reimbursement that changes provider incentives is payment bundling. A bundled health payment occurs when an institution gets paid a fi xed amount of money to treat a patient over the entire life cycle of the illness. Under bundled payments, hospitals have an incentive to reduce costs, because hospitals get to keep the proceeds from reducing the number of procedures they perform. Bundling also encourages high- quality care, because hospitals are required to pay for readmissions due to faulty care. Finally, an integrated approach to healthcare may reduce the tendency for institutions to pass the buck to other institutions, which has led to expensive hospital readmissions following a stay at a skilled nursing facility (Mor et al. 2010). Some initial experiments have shown promising results in reducing both costs and readmission rates (Casale et al. 2007; Cromwell, Dayhoff, and Thoumaian 1997).

Comparative Effectiveness Analysis Finally, one of the most powerful methods of reducing hospital expenditures is to study them to understand what technologies are effective and which are not. Many procedures are categorized as category 3 only because it is not known whether they are cost effective or not. An increased use of comparative effectiveness analysis would give both

266

Amitabh Chandra and Jonathan Holmes

doctors and policy makers a better idea of which procedures are effective and for whom. Particularly for category 2 and category 3 technologies, these trials are expensive, because treatments must be tested on a variety of groups that differ by factors such as age and genetic background. But the benefits to this research also have the potential to be great by allowing all parties to make better decisions (Chandra, Jena, and Skinner, 2011).

Notes Amitabh Chandra acknowledges support from the National Institute on Aging (Grant Nos. P01-AG005842-24 and P0-AG19783). The research on which this article is based was funded by the National Institute of Aging NIA P01-AG19783 and the Robert Wood Johnson Foundation. 1. Data are taken from (Trustees of Dartmouth College 2009). 2. Technically, many Medicare reimbursements are not pure fee-for-service, and are instead subject to a prospective payment system, according to which reimbursements are calculated based on the disease code of an admitted patient. But in practice, the disease codes are often fairly closely linked to their associated procedures, and hospitals that treat patients more intensively are permitted to use disease codes which have higher reimbursements. So, for example, a hospital that performs minimal surgery on elderly patients with terminal conditions would be reimbursed less than a hospital which performs intensive surgeries on patients with similar conditions on admission. 3. For elderly patients, it is clear that routine PSA testing is not worth the costs (Walter et al. 2006) 4. See Wick et al. (2010) and Friedman et al. (2010) for a discussion of the case.

References Archer, Chantai, Adrian R. Levy, and Maurice McGregor. 1993. “Value of Routine Preoperative Chest X-Rays: A Meta-Analysis.” Canadian Journal of Anaesthesia 40 (11): 1022– 27. Barbash, Gabriel I., and Sherry A. Glied. 2010. “New Technology and Health Care Costs—The Case of Robot-Assisted Surgery.” New England Journal of Medicine 363 (8): 701–4. Boden, William E., et al. 2007. “Optimal Medical Therapy With or Without PCI for Stable Coronary Disease.” New England Journal of Medicine 356 (15): 1503–16. Casale, Alfred S., Albert Bothe Jr., Ronald Paulus, Mark Selna, Karen McKinley, Michael C. Doll, Scott A. Berry, and Glenn D. Steele. 2007. “ProvenCareSM:

The Role of Technology in Expenditure Growth in Healthcare

267

a Provider-Driven Pay-for-Performance Program for Acute Episodic Cardiac Surgical Care.” Annals of Surgery 246 (4): 613–23. Chandra, Amitabh, Jonathan Gruber, and Robin McKnight. 2010. “Patient CostSharing and Hospitalization Offsets in the Elderly.” American Economic Review 100 (1): 193– 213. Chandra, Amitabh, Jonathan Holmes, and Maurice Dalton. 2013. “Large Increases in Spending on Postacute Care in Medicare Point to the Potential for Cost Savings in These Settings.” Health Affairs 32 (5): 864– 72. Chandra, Amitabh, Anupam B. Jena, and Jonathan Skinner. 2011. “The Pragmatist’s Guide to Comparative Effectiveness Research.” Journal of Economic Perspectives 25 (2): 27–46. Chandra, Amitabh, and Jonathan Skinner. 2012. “Technology Growth and Expenditure Growth in Health Care.” Journal of Economic Literature 50 (3): 645– 80. Chandra, Amitabh, and Douglas O. Staiger. 2007. “Productivity Spillovers in Health Care: Evidence from the Treatment of Heart Attacks.” Journal of Political Economy 115 (1): 103–40. Chernew, Michael E., Mayur R. Shah, Arnold Wegh, Stephen N. Rosenberg, Iver A. Juster, Allison B. Rosen, Michael C. Sokol, Kristina Yu-Isenberg, and A. Mark Fendrick. 2008. “Impact of Decreasing Copayments on Medication Adherence within a Disease Management Environment.” Health Affairs 27 (1): 103–12. Cromwell, Jerry L., Debra A. Dayhoff, and Armen H. Thoumaian. 1997. “Cost Savings and Physician Responses to Global Bundled Payments for Medicare Heart Bypass Surgery.” Health Care Financing Review 19 (1): 41– 57. Cutler, David M. 2004. Your Money or Your Life: Strong Medicine for America’s Health Care System. Oxford and New York: Oxford University Press. Cutler, David M., and Mark McClellan. 2001. “Is Technological Change in Medicine Worth It?” Health Affairs 20 (5): 11– 29. Ferguson, John H., Michael Dubinsky, and Peter J. Kirsch. 1993. “Court-Ordered Reimbursement for Unproven Medical Technology: Circumventing Technology Assessment.” Journal of the American Medical Association 269 (16): 2116–21. Finkelstein, Amy. 2007. “The Aggregate Effects of Health Insurance: Evidence from the Introduction of Medicare.” Quarterly Journal of Economics 122 (1): 1– 37. Ford, Earl S., Umed A. Ajani, Janet B. Croft, Julia A. Critchley, Darwin R. Labarthe, Thomas E. Kottke, Wayne H. Giles, and Simon Capewell. 2007. “Explaining the Decrease in U.S. Deaths from Coronary Disease, 1980– 2000.” New England Journal of Medicine 356 (23): 2388– 98. Friedman, Henry S., et al. 2010. “Bevacizumab and Recurrent Malignant Gliomas: A European Perspective: Reply.” Journal of Clinical Oncology 28 (12): E190– 92.

268

Amitabh Chandra and Jonathan Holmes

Fuhrmans, Vanessa. 2007. “A Novel Plan Helps Hospital Wean Itself Off Pricey Tests.” Wall Street Journal, January 12. Harris, Russell, and Kathleen N. Lohr. 2002. “Screening for Prostate Cancer: An Update on the Evidence from the U.S. Preventative Services Task Force.” Annals of Internal Medicine 137 (11): 917– 29. Hartwell, Debbie, Jill L. Colquitt, Emma Loveman, Andrew J. Clegg, H. Brodin, Norman Waugh, Pam Royle, Peter Davidson, Luke Vale, and Lynda MacKenzie. 2005. “Clinical Effectiveness and Cost-Effectiveness of Immediate Angioplasty for Acute Myocardial Infarction: Systematic Review and Economic Evaluation.” Health Technology Assessment 9 (17): 1– 99. Hochman, Judith S., et al. 2006. “Coronary Intervention for Persistent Occlusion after Myocardial Infarction.” New England Journal of Medicine 355 (23): 2395– 2407. Jencks, Stephen F., Edwin D. Huff, and Timothy Cuerdon. 2003. “Change in the Quality of Care Delivered to Medicare Beneficiaries, 1998–99 to 2000– 2001.” Journal of the American Medical Association 289 (3): 305– 312. Lichtenberg, Frank R. 2001. “Are the Benefits of Newer Drugs Worth Their Cost? Evidence from the 1996 MEPS.” Health Affairs 20 (5): 241– 51. Lister, Joseph. 1867. “On the Antiseptic Principle in the Practice of Surgery.” Lancet 90 (2299): 353– 56. Medicare Payment Advisory Committee. 2003. “Report to the Congress: Variation and Innovation in Medicare.” Washington, DC: Medicare Payment Advisory Committee. Miller, Mark E. 2005. “MedPAC Recommendations on Imaging Services.” Testimony Before the Subcommittee on Health, Committee on Ways and Means, U.S. House of Representatives. Miller, David C., Cathryn Gust, Justin B. Dimick, Nancy Birkmeyer, Jonathan Skinner, and John D. Birkmeyer. 2011. “Large Variations in Medicare Payments for Surgery Highlight Savings Potential from Bundled Payment Programs.” Health Affairs 30 (11): 2107–15. Mor, Vincent, Orna Intrator, Zhanlian Feng, and David C. Grabowski. 2010. “The Revolving Door of Rehospitalization from Skilled Nursing Facilities.” Health Affairs 29 (1): 57– 64. Moseley, J. Bruce, Kimberly O’Malley, Nancy J. Petersen, Terri J. Menke, Baruch A. Brody, David H. Kuykendall, John C. Hollingsworth, Carol M. Ashton, and Nelda P. Wray. 2002. “A Controlled Trial of Arthroscopic Surgery for Osteoarthritis of the Knee.” New England Journal of Medicine 347 (2): 81– 88. Muennig, Peter A., and Sherry A. Glied. 2010. “What Changes in Survival Rates Tell Us About US Health Care.” Health Affairs 29 (11): 2105–13. Newhouse, Joseph P. 1992. “Medical Care Costs: How Much Welfare Loss?” Journal of Economic Perspectives 6 (3): 3– 21.

The Role of Technology in Expenditure Growth in Healthcare

269

Philipson, Tomas J, and Anupam B. Jena. 2006. “Who Benefits from New Medical Technologies? Estimates of Consumer and Producer Surpluses for HIV/ AIDS Drugs.” Forum for Health Economics & Policy 9 (2): 1005. Schein, Oliver D., Joanne Katz, Eric B. Bass, James M. Tielsch, Lisa H. Lubomski, Mark A. Feldman, Brent G. Petty, and Earl P. Steinberg. 2000. “The Value of Routine Preoperative Medical Testing before Cataract Surgery.” New England Journal of Medicine 342 (3): 168– 75. Skinner, Jonathan, and Douglas O. Staiger. 2009. “Technology Diffusion and Productivity Growth in Health Care.” National Bureau of Economic Research Working Paper 14865. Skinner, Jonathan, Douglas O. Staiger, and Elliott S. Fisher. 2006. “Is Technological Change in Medicine Always Worth It? The Case of Acute Myocardial Infarction.” Health Affairs 25 (2): W34–47. Smetana, Gerald W., and David S. Macpherson. 2003. “The Case against Routine Preoperative Laboratory Testing.” Medical Clinics of North America 87 (1): 7–40. Temel, Jennifer S., Joseph A. Greer, Alona Muzikansky, Emily R. Gallagher, Sonal Admane, Vicki A. Jackson, Constance M. Dahlin, Craig D. Blinderman, Juliet Jacobsen, William F. Pirl, J. Andrew Billings, and Thomas J. Lynch. 2010. “Early Palliative Care for Patients with Metastatic Non-Small- Cell Lung Cancer.” New England Journal of Medicine 363 (8): 733–42. Terasawa, Teruhiko, Tomas Dvorak, Stanley Ip, Gowri Raman, Joseph Lau, and Thomas A. Trikalinos. 2009. “Systematic Review: Charged-Particle Radiation Therapy for Cancer.” Annals of Internal Medicine 151 (8): 556– 65. Trustees of Dartmouth College. 2009. “Dartmouth Atlas of Health Care.” http:// www.dartmouthatlas.org. Accessed October 29, 2014. Walter, Louise C., Daniel Bertenthal, Karla Lindquist, and Badrinath R. Konety. 2006. “PSA Screening Among Elderly Men With Limited Life Expectancies.” Journal of the American Medical Association 296 (19): 2336–42. Weintraub, William S., et al. 2008. “Effect of PCI on Quality of Life in Patients with Stable Coronary Disease.” New England Journal of Medicine 359 (7): 677– 87. Welch, Gilbert H., Peter C. Albertsen, Robert F. Nease, Thomas A. Bubolz, and John H. Wasson. 1996. “Estimating Treatment Benefits for the Elderly: The Effect of Competing Risks.” Annals of Internal Medicine 124 (6): 577– 84. Wick, Wolfgang, Michael Weller, Martin van den Bent, and Roger Stupp. 2010. “Bevacizumab and Recurrent Malignant Gliomas: A European Perspective.” Journal of Clinical Oncology 28 (12): E188– 89. Wunsch, Hannah, Walter T. Linde-Zwirble, David A. Harrison, Amber E. Barnato, Kathryn M. Rowan, and Derek C. Angus. 2009. “Use of Intensive

270

Amitabh Chandra and Jonathan Holmes

Care Services during Terminal Hospitalizations in England and the United States.” American Journal of Respiratory and Critical Care Medicine 180 (9): 875– 80. Yusuf, S., R. Peto, J. Lewis, R. Collins, and P. Sleight. 1985. “Beta Blockade during and after Myocardial Infarction: An Overview of the Randomized Trials.” Progress in Cardiovascular Diseases 27 (5): 335– 71.

Chapter eleven

Economic Issues Associated with Incorporating Cost-Effectiveness Analysis into Public Coverage Decisions in the United States Anupam B. Jena and Tomas J. Philipson

Introduction

H

ealthcare continues to account for a growing share of national income in the United States. In 1970, national healthcare expenditures accounted for 7 percent of the country’s gross domestic product, compared with 18 percent in 2011. In addition to the largely fee-for-service payment structure that characterizes the U.S. healthcare delivery system and that encourages high levels of healthcare spending per capita, a leading explanation for growth in U.S. healthcare spending is technological innovation.1 New medications, medical devices, procedures, and diagnostic studies command higher prices than their predecessors, forcing healthcare spending upward in an environment where new medical therapies are sometimes indiscriminately adopted. It should come as no surprise, then, that the governments of most Westernized nations, including the United States, are grappling with how best to assess the value of these medical innovations, in an effort to stem rising public healthcare expenditures without compromising health. In the United States, both public and private payers are increasingly demanding more information on the costs and benefits of new medical technologies. Expensive biologic treatments for cancer and other diseases,

272

Anupam B. Jena and Tomas J. Philipson

relatively unproven therapies such as proton beam therapy for prostate cancer, and costly radiographic imaging studies such as magnetic resonance imaging have come under increased scrutiny by policy makers and third-party payers seeking to curtail use of expensive technologies whose benefits are potentially outweighed by their costs. The major scientific approach that has been utilized to assess the costs and benefits of new medical technologies is cost- effectiveness analysis (CEA), also referred to as cost-benefit and cost-utility analysis. 2 As the name suggests, CEA offers payers a way to allocate scarce healthcare budgets based on the costs and benefits of existing medical treatments. The main method through which CEA has been implemented is the use of cost- effectiveness thresholds, which state that a given treatment will only be reimbursed by a payer if the incremental cost to the payer (above a comparison treatment) is less than a prespecified threshold. For instance, in many high-income countries such as the United Kingdom and Australia, treatments that cost $50,000 more than a baseline treatment and that lengthen life by less than one Quality Adjusted Life Year (QALY) relative to that treatment are looked at skeptically by reimbursement authorities. In Australia, for example, a study of coverage submissions to the country’s Pharmaceutical Benefits Advisory Committee—the organization responsible for determining whether a treatment will be covered by public funds—found that only two out of twenty-six submissions whose incremental cost per QALY exceeded $57,000 were accepted for coverage. 3 In a review of similar coverage submissions to the United Kingdom’s National Institute for Clinical Excellence (NICE), one study found that all treatments recommended for coverage by NICE had costs per QALY below $30,000.4 Although explicit cost- effectiveness thresholds have not been used in coverage decisions by the Centers for Medicare and Medicaid Services (CMS) in the United States, their use has been discussed extensively and there appears to be a growing movement towards incorporating CEA in coverage decisions in the future. Moreover, although CEA is less well incorporated in the publicly fi nanced portion of the U.S. market, it is often used by private payers to justify coverage decisions; the majority of published cost- effectiveness studies are conducted for the U.S. market and funded by U.S. manufacturers. For example, one study found that 51 percent of 1,393 cost- effectiveness studies published from 1976 to 1996 were performed from a U.S. perspective. Similarly, during this same pe-

Cost-Effectiveness Analysis and Public Coverage

273

riod, seven of the top ten companies in terms of industry-sponsored costeffectiveness studies were based in the United States5 In this chapter, we briefly review the past and current role of CEA in treatment coverage determinations in the United States and discuss two important shortcomings of CEA implied by economic analysis. The intent of our discussion is to analyze the potential shortcomings of formally implementing CEA into U.S. public coverage decisions, which we view as possible in the near future. The discussion is based on our own work on this important issue; the interested reader is referred to those studies for a more in- depth analysis.6 First, we demonstrate why CEA, as it has ordinarily been applied to guide treatment reimbursement decisions, can fail to allocate scarce resources in a way that is economically efficient. This occurs because treatments that are equally valuable to society—in the sense that they lead to equivalent improvements in health—may have similar prices and yet cost society very different amounts to produce. From the perspective of insurers and government payers, treatments of equal effectiveness and price will appear equally cost effective, even if from the perspective of society they are not. Costs of production include manpower, human capital, facilities, and machinery, all of which may differ dramatically between treatments (e.g., pharmaceuticals that cost pennies to produce per pill versus costly medical devices). Cost- effectiveness-based policies that focus only on the prices that payers face will lead to the coverage of some treatments with higher resource costs of production than other treatments that are equally effective but command higher prices (and are therefore not reimbursed). We discuss how coverage policies based on CEA can exacerbate this problem by impacting how companies set prices in response to these policies. Second, we highlight how a country’s decision to base coverage decisions only on a treatment’s cost effectiveness can stifle innovation. Coverage policies based on cost effectiveness, such as cost- effectiveness thresholds, closely resemble price controls. Price controls, in turn, reduce incentives for innovation as do other mechanisms to contain costs. Consider, for instance, a coverage policy that requires all treatments covered by a payer to cost less than $50,000 per QALY. This policy acts as a price control by fi xing the price of treatments to be less than this dollar amount per QALY. If this price is too low, it may limit the incentive for companies to innovate by reducing the price they can expect to charge

274

Anupam B. Jena and Tomas J. Philipson

for their treatment. In the same way that the patent system exists to encourage innovation through inefficiently high prices, cost- effectiveness thresholds that lower the prices companies receive can limit innovation. This idea is particularly important when considering the role of costeffectiveness-based policies in the United States and their impact on global incentives for innovation. Because U.S. spending on pharmaceuticals is close to half of global spending, reductions in U.S. sales arising from price reductions could substantially lower incentives for pharmaceutical innovation.

Cost-Effectiveness Analysis in the United States Unlike in many other Westernized nations, in which CEA is formally incorporated into public coverage decisions of medical therapies, the provision of healthcare in the United States is not organized as a singlepayer system. As a result, a national set of coverage guidelines based on CEA does not exist in the United States, although the role of CEA in guiding public coverage decisions has been extensively discussed. For instance, as the largest single supplier of health insurance in the United States, the Medicare program has grappled with whether or not to include CEA in coverage decisions. To date, Medicare is required by law not to consider costs of treatment in coverage decisions, but some evidence suggests that cost effectiveness has played an informal role in such decisions in the past. For example, although the initial Medicare statute did not cover preventive services, the cost effectiveness of vaccines for streptococcal pneumonia, hepatitis B, and influenza, and that of screening for breast and cervical cancer, led to their later coverage by Medicare.7 Similarly, Medicare’s decision to cover podiatric services for diabetic patients was based on the cost effectiveness of this service. The extent to which Medicare has considered incorporating cost effectiveness of treatments to guide coverage decisions in the past is not fully known. In a notice of intent issued by the Health Care Financing Administration in 2000, it was suggested that Medicare coverage decisions be based on the “medical benefits” and “added value” of the treatment being evaluated.8 This statement was opposed by industry lobby groups on the basis that Medicare is legally precluded from incorporating a treatment’s cost effectiveness into its coverage determination.9 Since that time, Medicare has attempted on several occasions to incorporate

Cost-Effectiveness Analysis and Public Coverage

275

some form of CEA into coverage considerations, but has been unable to do so because of political opposition.10 As evidenced by a statement of the Director of the Office of Clinical Standards and Quality in a 2004 editorial in the New England Journal of Medicine, Medicare’s inability to formally consider cost effectiveness in coverage determinations has not precluded its use in guiding the level of scrutiny of investigation: The Medicare statute is silent on the role of costs, and Medicare has not explicitly considered costs in making coverage decisions. Health care services are generally covered when there is adequate evidence that they improve health outcomes, irrespective of the unit or aggregate cost. However, technology that is associated with very high costs is also very likely to have substantial clinical ramifications for the Medicare population and, therefore, these forms of technology receive comparatively greater scrutiny than other devices, procedures, and services.11

The recurring theme of whether cost effectiveness should be considered in Medicare coverage decisions, along with the rising burden of publicly fi nanced healthcare in national income, suggests a growing movement towards formally incorporating CEA into Medicare coverage decisions in the future. Outside of Medicare, private, federal, and state public insurers have increasingly used CEA to guide treatment coverage decisions.12 For example, pharmacy benefit design in the Veterans Administration has routinely relied on cost- effectiveness assessments of medications being considered for inclusion in the administration’s formulary.13 The medication formulary and clinical practice guidelines of the Department of Defense’s military health system have similarly relied heavily on CEA.14 Finally, a discussion of the role of CEA in public coverage decisions in the United States would not be complete without a discussion of the Oregon experiment of the 1980s. Beginning in 1989, Oregon embarked on an effort to expand its state Medicaid program. In order to fi nance this expansion, it sought to offer its Medicaid beneficiaries a more limited set of healthcare services that were initially chosen and ranked based on cost effectiveness.15 Controversy and the ultimate termination of the program began when that list was published and was found to rank lifesaving surgical procedures for ectopic pregnancy and appendicitis as less cost effective than procedures such as dental caps for pulp exposure.

276

Anupam B. Jena and Tomas J. Philipson

The evolving trajectory of CEA in public and private coverage decisions is unknown. Private payers have already incorporated CEA into coverage determinations and the role of CEA is only likely to increase as delivery systems become accountable for the quality and costs of care of their patients through Accountable Care Organizations and the like. Rising costs of Medicare and the transition of baby boomers into Medicare eligibility may lead policy makers to ultimately reverse Medicare’s statutory injunction to not formally consider cost effectiveness in coverage decisions. Although incorporating CEA into public coverage decisions in the United States may seem attractive to those seeking to reduce public healthcare spending, there are several important limitations to consider, two of which we discuss in detail next.

Cost-Effectiveness Policies, Economic Efficiency, and Healthcare Spending In this section we demonstrate how cost- effectiveness policies that could be implemented by the United States to guide public coverage decisions could paradoxically lead to preferential coverage of treatments that are more costly to society to produce than comparison treatments that are similarly effective, and to greater healthcare spending. In practice, CEA uses the price charged to payers by companies as the “cost,” rather than the actual resource costs of production that ordinarily determine the best use of scarce resources. Prices, rather than costs, therefore determine the cost effectiveness of treatments that would be considered for coverage by public payers such as Medicare. This has several important implications. First, among a set of treatments of equal effectiveness, prices may not necessarily correlate well with costs, in that certain high- cost treatments may exhibit relatively low markups, while certain low- cost treatments exhibit substantial markups. Pharmaceuticals and medical devices provide an excellent illustration. Patented pharmaceuticals for diseases like HIV/AIDS cost upward of $10,000 per person annually and yet, like other pharmaceuticals, cost pennies to produce. Meanwhile, medical devices such as coronary stents for patients with coronary artery disease may cost more to produce but exhibit substantially lower markups in price. In general, the true cost effectiveness of treatments may not correlate well with cost effectiveness based on price, simply because many factors contribute to price including patient

Cost-Effectiveness Analysis and Public Coverage

277

demand, patent protection, and the availability of competing treatments. Two treatments that are equally effective in terms of improving longevity (i.e., QALYs) may have different prices because one is covered by patent and the other is not. Comparing two treatments, if one is inexpensive to produce (e.g., a patented pharmaceutical medication) whereas the other is more costly (e.g., a medical device), society would be better served by covering the pricier but less resource-intensive treatment—in this instance, the medication. Of course, from the perspective of a public payer with a fi xed budget who does not weigh these costs, the less expensive but more resource-intensive option would make more sense for public coverage. This issue of choosing the appropriate cost perspective in CEA (i.e., viewing cost effectiveness from the perspective of society or the payer) is not the only concern with basing treatment coverage decisions on it. In fact, it should come as no surprise that public payers with a limited budget such as Medicare would make coverage decisions based on the costs they face rather than the production costs to society. An additional and perhaps more important concern about using CEA to determine public coverage of treatments is that companies will respond in predictable ways to these policies by changing the prices that they charge. To illustrate how cost- effectiveness policies themselves impact pricing decisions, consider how companies might respond to cost- effectiveness thresholds. In the United Kingdom, for example, NICE is responsible for making recommendations to the British National Health Service (NHS) on the coverage of selected medical treatments. These decisions are based on a treatment’s clinical effectiveness, cost per QALY gained for patients using the treatment, and the overall impact on NHS costs.16 Companies that would like their products covered by NHS submit costeffectiveness information based on the costs that would be incurred by the NHS in covering treatment (i.e., the price charged to the NHS). Based on this information as well as similar cost- effectiveness estimates produced by independent groups, academic groups, and patient advocacy groups, NICE makes a determination on whether a treatment will be covered by the NHS. In the past, NICE has employed an implicit cost- effectiveness threshold that recommends coverage of treatments whose cost per QALY gained falls below a threshold of about $50,000. Companies facing such a threshold have an incentive to raise the price of their treatment up to the threshold when there are no other competitive factors (e.g., other

278

Anupam B. Jena and Tomas J. Philipson

competitors) that drive the submitted price downwards. For instance, a company that produces a new biologic treatment that can be used to treat metastatic colon cancer may face no competition from other products, and as a result would charge a price to the NHS that is consistent with a cost effectiveness of $50,000 per QALY. More clearly, if the treatment extends life by 0.5 QALYs, the company would have an incentive to charge the NHS $25,000 for the treatment, implying a cost effectiveness of $50,000 per QALY gained. Companies therefore respond to the cost- effectiveness policies they face by charging prices that maximize their profits. This has several important implications for what we might expect if the United States were to implement cost- effectiveness-based coverage decisions. First, the intent of such policies to maximize health under a fi xed budget would be satisfied, but at potentially greater-than- expected resource costs to society. To illustrate how this occurs, note that with a cost- effectiveness threshold policy, companies producing new, patent-protected treatments would have an incentive to raise the prices of their treatments to meet the threshold. From the perspective of Medicare, all of these treatments would appear to have the same cost effectiveness, since all companies price to the cost- effectiveness threshold. The cost to Medicare of generating an additional QALY would be identical across treatments and equal to the threshold (e.g., a cost of $50,000 per QALY). As a result, Medicare would be indifferent as to which treatments it should cover, even though the resource costs used to produce each of these treatments may be vastly different. Patent-protected pharmaceuticals may have the same cost effectiveness from the perspective of Medicare as medical devices, although the latter are substantially more costly to produce. While the overall health of Medicare’s beneficiaries would be maximized under its fi xed budget by utilizing cost effectiveness to make coverage decisions, overall social welfare (which considers the costs of resources used to produce various treatments) could still be improved. The reason is that the fi xed budget could instead be allocated towards treatments of identical cost effectiveness (from the perspective of Medicare) that are less costly to society to produce. In addition to the economic efficiency that may be created by using CEA in public coverage decisions, a second implication of costeffectiveness-based coverage is that the intended goal of these policies to reduce healthcare spending may fail. The reason is that with public insurance (such as Medicare or the NHS), the prices that patients face

Cost-Effectiveness Analysis and Public Coverage

279

(termed by economists as “demand” prices and also commonly referred to as copayments) are different and substantially lower than the prices that the public payer faces (termed “supply” prices). Public insurers set demand prices for patients that are often considerably lower than the prices that would occur if insurance did not exist and companies were to charge prices according to the market. When the prices that patients and public insurers pay are separated, it is possible that cost- effectiveness policies will raise healthcare spending, because the prices charged to insurers by companies may exceed prices that would exist in the private market. For instance, cost- effectiveness thresholds may lead companies to raise the price of their treatments up to the threshold, a price that would not be possible in a private market in which patients paid the full price. By facilitating prices that are higher than would exist in a private market or in a market without cost- effectiveness thresholds, costeffectiveness thresholds may, paradoxically, raise healthcare spending rather than lower it.

Cost-Effectiveness Policies and Incentives for Innovation This section demonstrates how cost- effectiveness policies have the potential to impact incentives for innovation adversely. We begin with a basic economic description of how cost effectiveness is related to the economic concepts of consumer and producer surplus, the latter of which determines the incentives for companies to innovate. We then briefly illustrate why the link between cost effectiveness and innovation is empirically relevant by using case studies from HIV/AIDS and cancer. We conclude by revisiting the impact of implementing cost- effectiveness-based policies in the United States on innovation, emphasizing the important role that the United States plays in determining global innovation incentives. Cost Effectiveness, Consumer Surplus, and Company Profi ts At its core, CEA encourages the public coverage of treatments for which the benefits to patients today outweigh the costs to public payers. The most cost- effective treatments are those for which the incremental costs to the healthcare system are far outweighed by the value to patients. In economics, the difference between the monetized value patients place on a treatment and what is actually paid is termed “consumer surplus.” For

280

Anupam B. Jena and Tomas J. Philipson

life-saving medications that have low copayments to patients, the consumer surplus is very high. For similarly life-saving treatments that have high copayments to patients (e.g., surgery), the consumer surplus is lower. In a standard model of supply and demand for treatment, the value individuals place on a treatment is the area under the demand curve—that is, the total willingness to pay for treatment. That total willingness to pay is divided between patients in the form of consumer surplus and companies in the form of profits. In this framework, the cost effectiveness of a treatment is closely related to its consumer surplus. A treatment for which the value to a patient far exceeds the costs the patient faces is both highly cost effective and has a large consumer surplus. That treatment, in turn, generates lower profits for the company that produced the treatment, since most of the treatment’s social value is appropriated by patients. Cost Effectiveness and Innovation The link between a treatment’s cost effectiveness, consumer surplus, and the profits it generates motivates the following question: If a treatment’s cost effectiveness were further improved by lowering price, would patients be better or worse off? The answer to this question depends on whether one takes a static or dynamic view of the world. When a treatment’s cost effectiveness is improved through lowered prices, patients utilizing the treatment today are clearly better off, since they enjoy the same improvement in health at lower cost. Future patients may be adversely affected, however, if the shift of profits towards consumers (in the form of greater consumer surplus) reduces the incentive for companies to innovate. Indeed, after a treatment has been discovered, society is clearly best off when the price of the treatment equals its marginal cost of production. In that scenario, the quantity of the treatment utilized is at its highest, as are cost effectiveness and consumer surplus. In reality, however, the research and development required to produce new medical treatments is costly and uncertain, to the point that we as a society allow companies to charge prices that exceed marginal costs in order to ensure that they have adequate incentives to undertake the costs of innovation. This is the rationale of the patent system, which trades off the well-being of patients today (who face higher prices) to ensure the well-being of patients in the future (who have access to new medical therapies they might not otherwise have been able to utilize).

Cost-Effectiveness Analysis and Public Coverage

281

Cost effectiveness is therefore intrinsically linked to the incentives companies have to innovate. This can be further illustrated with an extreme example of price discrimination, in which the producer of a treatment is able to charge each individual patient what that patient is willing to pay for the treatment. Doing so would completely eliminate consumer surplus—since the price each patient pays would be equal to, not less than, the amount that patient is willing to pay—and would maximize producer surplus or profits. This complete transfer of consumer surplus to producer profits would make the treatment not cost effective from the perspective of current patients, but would ensure that the producers have the maximum incentive to innovate as measured by profits. In this case, cost effectiveness would be minimized from a static perspective on the world but potentially maximized from a dynamic perspective that incorporates the well-being of future patients who benefit from innovation. Importantly, in this example, the health of both current and future patients is actually maximized. Those utilizing the treatment today pay what they are wiling (or able), and this maximizes health because even those patients with low willingness (or ability) to pay are able to utilize treatment at a lower price. The health of future patients is maximized as well, since they have access to innovative treatments that they would not if consumer surplus were higher and producer surplus lower. Of note, an important consideration in advocating a policy that lowers consumer surplus is the distributional consequence of shifting surplus from consumers toward producers. While economists often do not distinguish between the two parties when calculating the welfare created by policies, the same is not true for policy makers. It is therefore important to note that when patients themselves are shareholders or employees of healthcare companies, or owners of pension plans that directly invest in those companies, the distinction between consumers and producers becomes less clear. Cost-Effectiveness Thresholds as a Price Control The link between a treatment’s cost effectiveness and the implied division of the treatment’s value between patients and producers raises important questions about the impact of cost- effectiveness-based coverage policies on incentives for innovation. Cost- effectiveness thresholds very closely resemble price controls in that companies seeking public

282

Anupam B. Jena and Tomas J. Philipson

coverage of their treatments must price then no higher than the threshold if they expect coverage. As these thresholds are set lower, treatments that would ordinarily command prices well above marginal costs because of patent protection will instead have prices pressured downward by the cost- effectiveness threshold. Stricter cost- effectiveness thresholds will therefore unarguably lower profits by lowering the supply price companies receive for their product. It is important to note, however, that while stricter cost-effectiveness thresholds will unarguably lower profits, those profits may be still higher than in a private market in which demand is not subsidized by low demand prices (i.e., copayments).17 For example, when utilization is heavily subsidized through low demand prices (i.e., low copayments), lower supply prices induced by cost-effectiveness thresholds may still lead to higher profits than an unsubsidized private market in which demand and supply prices would be identical to each other but higher than in subsidized markets, and in which utilization would be lower than in subsidized markets. Stricter cost-effectiveness thresholds may therefore have varying effects on profits compared to private markets without insurance, but will always lower profits when there is public insurance. The reduction in profits that occurs when public payers lower costeffectiveness thresholds will reduce incentives to innovate, which will be socially harmful or helpful depending on whether there are already insufficient or ample incentives to innovate. For example, when private incentives to innovate are inefficiently low from an economic perspective—implying that the marginal benefits of additional research and development are outweighed by their costs—reductions in profits through stricter costeffectiveness thresholds will reduce needed innovation. Alternatively, if incentives to innovate are already abundant, encouraging more generous cost- effectiveness thresholds which raise profits will induce a gold rush of research and development efforts that may be economically inefficient in that those same research and development efforts would be best allocated toward other areas (e.g., environmental research or computing). While it is impossible to definitely know whether additional private healthcare research and development is needed and should be incentivized, the fact that nearly 50 percent of healthcare research and development arises from public sources such as the National Institutes of Health suggests that more and not less private research and development would be economically efficient. By serving as an implicit price control, the implementation of costeffectiveness-based coverage policies by the United States may threaten

Cost-Effectiveness Analysis and Public Coverage

283

incentives for innovation. The patent system exists to reward innovation by allowing companies to charge prices that exceed marginal costs; the use of cost- effectiveness thresholds in U.S. coverage decisions would thus appear to contradict the intended goal of the patent system. An illustrative case of this may be vaccines, which are very cost effective, in part due to government monopsony power, yet lack substantive investments in research and development.18 An additional concern of implementing cost- effectiveness thresholds in U.S. coverage decisions relates to the United States’ place in the global healthcare marketplace. Because of the large role played by the United States in the global demand for healthcare, the implementation of cost- effectiveness thresholds in the United States would have large consequences not only for U.S. citizens but around the world. A country like Belgium, which accounts for a minor share of the world market for healthcare and yet has a similar income per capita to the United States, could implement cost- control policies without expecting to have an impact on global healthcare profits. The same would not be true for the United States, which accounts for nearly a third of global pharmaceutical spending. A cost- effectiveness threshold policy that reduces prices paid by U.S. consumers could dramatically impact global incentives for research and development, an important issue that has received little recognition. Case Studies of HIV/AIDS and Cancer An important empirical question raised by our analysis of the link between cost effectiveness and innovation is whether producers of medical treatments currently capture little or much of the value their treatments generate. If most of the value of innovative therapies already accrues to producers in the form of profits, advocacy for cost- effectiveness thresholds that somewhat lower these profits may be warranted. Alternatively, if producer profits reflect only a small fraction of a treatment’s social value, stricter cost-effectiveness thresholds and lower profits may be particularly deleterious. Therapies to treat HIV/AIDS and cancer serve as two useful illustrations of how the social value of treatments has been divided between patients and producers in practice. Both diseases have exhibited substantial gains in innovation and longevity in the last several decades—HIV with the introduction of highly active antiretroviral therapy (HAART)

284

Anupam B. Jena and Tomas J. Philipson

in 1996, and cancer with the introduction of new chemotherapeutic and targeted treatment regimens since the War on Cancer was initiated by President Nixon in 1971. In prior work, we have shown the longevity gains in each of these diseases to be large and extremely valuable from a social perspective.19 Yet industry profits for treatments of each of these diseases have been substantially lower than the social value of this greater life expectancy to patients. We briefly describe each of these case studies below. Since the onset of the HIV/AIDS epidemic nearly three decades ago, survival of individuals infected with HIV has improved dramatically, largely due to medical advances to treat the disease and its complications. Within ten years of the fi rst reported case of HIV, AZT (azidothymidine) was introduced, and less than ten years later the medications comprising HAART became available after expedited review by the Food and Drug Administration. Averaged across cohorts infected with HIV since the start of the epidemic, life expectancy has increased by at least five years; those diagnosed with AIDS early in the epidemic died quickly of the disease, while those diagnosed with HIV or AIDS today live substantially longer, upward of fi fteen years on average after diagnosis. Assuming a value of a statistical life year of $100,000, the gain in longevity has been worth more than $500,000 per person infected with HIV and more than $750 billion for the nearly 1.5 million Americans infected to date. Adding the value of improved survival for those who will be infected with HIV in the future raises this aggregate social value to over $1 trillion. 20 This value can be compared to the profits of companies producing HIV/AIDS treatments that we have estimated in other work. Using annual data on national revenues from HIV-related medications obtained from other studies, we estimate profits of companies producing HIV/AIDS treatments to be approximately $60 billion to date. This implies that of the nearly $750 billion worth of value generated by medications to treat HIV/AIDS, producers of those therapies captured less than 10 percent in the form of profits. While HIV may seem an extreme example of a particular disease, these results extend more broadly to a much larger class of diseases, cancer. In 1971, President Nixon initiated the War on Cancer with the passage of the National Cancer Act. 21 While cancer continues to be the second leading cause of death in the United States, rates of death from cancer have fallen dramatically through improvements in rates of screening and survival after diagnosis. Many of the genes responsible for cancer have been identified, and chemotherapeutic and targeted cancer therapies

Cost-Effectiveness Analysis and Public Coverage

285

have been refi ned. In our own work, we have estimated that the gains in cancer survival from 1988 to 2000 were large and were mainly attributable to improvements in treatment rather than screening.22 For example, we estimated that survival for all cancers combined increased nearly 3.9 years between 1988 and 2000, while survival for patients with nonHodgkin’s lymphoma and breast cancer (two areas of substantial medical innovation) increased 3.6 and 3.5 years, respectively. Approximately 80 percent of these survival gains were estimated to be due to improvements in treatment; changes in the likelihood of cancer detection were much less important for improving survival than improvements in survival at any given stage of disease. We have estimated that compared to an individual diagnosed with cancer in 1988, an individual diagnosed in 2000 would be willing to pay approximately $31,000 annually for the improved survival prospects of being diagnosed in 2000 rather than 1988. Multiplying this value by the nearly 23 million additional life-years generated by improved survival in 2000 versus 1988 suggests a social value of improvements in cancer survival worth approximately $1.9 trillion. When compared to the profits earned by companies and individuals responsible for improving cancer care (drug companies, hospitals, physicians, and other health professionals), the surplus to patients of improved survival from cancer is vastly greater. For instance, we estimated profits ranging from approximately $100 to $400 billion between 1988 and 2000, implying that those responsible for improving cancer survival during this period appropriated between 5 percent and 20 percent of the value of improved longevity.

Conclusions As public expenditures on healthcare continue to account for an evergrowing share of national income in the United States and other Westernized nations, governments are facing difficult decisions about how best to evaluate and fi nance new medical therapies, which are a large contributor to rising healthcare costs. In the United States, the publicly fi nanced expansion of insurance mandated by the Affordable Care Act will place greater fi nancial pressure on public payers such as Medicare and Medicaid, which must evaluate new and expensive medical therapies. The public coverage of treatments on the basis of their cost effectiveness has been debated widely in the United States and that discussion

286

Anupam B. Jena and Tomas J. Philipson

is only likely to intensify. The majority of published cost- effectiveness studies are already conducted by U.S. companies for the United States market. While it is clear why public payers would attempt to maximize the health benefits they obtain from the limited resources available for healthcare spending, the use of cost effectiveness to guide reimbursement should be viewed with at least some caution. Companies may respond to costeffectiveness policies in ways that are unanticipated by policy makers but may be predicted by economic theory. Those responses may in fact lower economic inefficiency and increase healthcare spending if not carefully considered. Similarly, cost-effectiveness policies may have adverse impacts on innovation, an issue that is particularly salient in the U.S. marketplace given its role in driving global incentives for innovation. Each of these issues deserves serious policy consideration. It is imperative to understand the consequences and implications of cost- effectiveness-based reimbursement policies before sailing further into this uncharted sea.

Notes Anupam Jena acknowledges funding from the Office of the Director, National Institutes of Health (NIH Early Independence Award, Grant 1DP5OD017897- 01). 1. Joseph P. Newhouse, “Medical Care Costs: How Much Welfare Loss?” Journal of Economic Perspectives 6, no. 2 (1992): 3– 21; Amitabh Chandra and Jonathan Skinner, “Technology Growth and Expenditure Growth in Health Care,” Journal of Economic Literature 50, no. 3 (2012): 645– 80. 2. Milton C. Weinstein and William B. Stason, “Foundations of CostEffectiveness Analysis for Health and Medical Practices,” New England Journal of Medicine 296, no. 13 (1997): 716–21; Charles E. Phelps and Stephen T. Parente, “Priority Setting in Medical Technology and Medical Practice Assessment,” Medical Care 28, no. 8 (1990): 703–23; Milton C. Weinstein and William G. Manning Jr., “Theoretical Issues in Cost-Effectiveness Analysis,” Journal of Health Economics 16, no. 1 (1997): 121–28; Magnus Johannesson and Milton C. Weinstein, “On the Decision Rules of Cost-Effectiveness Analysis,” Journal of Health Economics 12, no. 4 (1993): 459– 67; Alan M. Garber and Charles E. Phelps, “Economic Foundations of Cost-Effectiveness Analysis,” Journal of Health Economics 16, no. 1 (1997): 1– 31. 3. Bethan George, Anthony Harris, and Andrew Mitchell, “Cost-Effectiveness Analysis and the Consistency of Decision Making: Evidence from Pharmaceutical Reimbursement in Australia (1991 to 1996),” Pharmacoeconomics 19, no. 11 (2001): 1103– 09.

Cost-Effectiveness Analysis and Public Coverage

287

4. James Raftery, “NICE: Faster Access to Modern Treatments? Analysis of Guidance on Health Technologies” BMJ 323(7324) (2001): 1300–1303. 5. Peter J. Neumann, Chi-Hui Fang, and Joshua T. Cohen, “30 Years of Pharmaceutical Cost-Utility Analyses: Growth, Diversity and Methodological Improvement,” Pharmacoeconomics 27, no. 10 (2009): 861– 72. 6. Anupam B. Jena and Tomas Philipson, “Cost-Effectiveness as a Price Control,” Health Affairs 26, no. 3 (2007): 696– 703; Anupam B. Jena and Tomas J. Philipson, “Cost-Effectiveness Analysis and Innovation,” Journal of Health Economics 27, no. 5 (2008): 1224– 36: Amitabh Chandra, Anupam B. Jena, and Jonathan S. Skinner, “The Pragmatist’s Guide to Comparative Effectiveness Research,” Journal of Economic Perspectives 25, no. 2 (2011): 27–46; Darius N. Lakdawalla, Eric C. Sun, Anupam B. Jena, Carolina M. Reyes, Dana P. Goldman, and Tomas J. Philipson, “An Economic Evaluation of the War on Cancer,” Journal of Health Economics 29, no. 3 (2010): 333–46; Tomas J. Philipson and Anupam B. Jena, “Who Benefits from New Medical Technologies? Estimates of Consumer and Producer Surpluses for HIV/AIDS Drugs,” Forum for Health Economics & Policy (2006): 1005–1005. Anupam B. Jena and Tomas J. Philipson, Innovation and Technology Adoption in Health Care Markets (Washington, DC: AEI Press, 2008). 7. U.S. Department of Health and Human Services Health Care Financing Administration. Program Memorandum to Carriers (Woodlawn, MD 1993); Elaine J. Power, Sean R. Tunis, and Judith L.Wagner, “Technology Assessment and Public Health,” Annual Review of Public Health 15 (1994): 561– 79. 8. U.S. Department of Health and Human Services Health Care Financing Administration. “Criteria for Making Coverage Decisions, Notice of Intent to Publish Federal Rule.” Federal Register 65 (95) (May 16, 2000). 9. Robin Strongin, “Medicare Coverage: Lessons from the Past, Questions for the Future.” National Health Policy Forum Background Paper, Washington, D.C. August 2001. 10. Susan Bartlett Foote and Peter J. Neumann, “The Impact of Medicare Modernization on Coverage Policy: Recommendations for Reform,” American Journal of Managed Care 11, no. 3 (2005): 140–42. 11. Sean R. Tunis, “Why Medicare Has Not Established Criteria for Coverage Decisions,” New England Journal of Medicine 350, no. 21 (2001): 2196– 98. 12. Power, Tunis, and Wagner, “Technology Assessment and Public Health,” 561– 79; Michael Dickson, Jeremy Hurst, and Stéphane Jacobzone, “Survey of Pharmacoeconomic Assessment in Eleven Countries,” OECD Health Working Paper No. 4 (2003). 13. Tunis, “Medicare,” 2196– 98. 14. Ibid. 15. David C. Hadorn, “Setting Health Care Priorities in Oregon: CostEffectiveness Meets the Rule of Rescue,” Journal of the American Medical Association 265, no. 17 (1991): 2218– 25.

Anupam B. Jena and Tomas J. Philipson

288

16. Raftery, “NICE,” 1300–1303. 17. Department of Health and Human Services, Program Memorandum. 18. Power, Tunis, and Wagner, “Technology Assessment ,” 561– 79. 19. Lakdawalla et al., “War on Cancer,” 333–46; Philipson and Jena, “HIV/ AIDS Drugs,” 1005. 20. Philipson and Jena, “HIV/AIDS Drugs,” 1005. 21. Chandra, Amitabh and Jonathan Skinner, “Technology Growth and Expenditure Growth in Health Care,” Journal of Economic Literature 50 no.3 (2012): 645– 80. 22. Tunis, “Medicare,” 2196– 98.

Chapter twelve

The Complex Relationship between Healthcare Reform and Innovation Darius Lakdawalla, Anup Malani, and Julian Reif

Introduction

L

ife expectancy has increased dramatically over the past century for individuals of all ages. For example, life expectancy at birth was 47.3 years in 1900. Today it is 78.1 years (Arias 2012). Although more difficult to measure, there is substantial evidence that the quality of life has increased also. Childhood is no longer plagued by dangerous infectious diseases such as measles, mumps, and rubella. In 1980, only 5 percent of marathon runners were over the age of fi fty. Today this figure has more than doubled (Cutler 2005). The value of this health increase is enormous. Murphy and Topel (2006) estimate that the cumulative gains in life expectancy since 1900 are worth $1.2 million (in 2004 dollars) to an individual alive today, with increases in longevity over the past forty years alone adding about $3.2 trillion per year to national wealth. They also offer suggestive evidence that the increase in quality of life may be as valuable as the increase in life expectancy. Although some of this increase is attributable to changes in lifestyle (e.g., a decline in smoking), much is due to medical innovation (Cutler, Kadiyala et al. 2003; Cutler 2007; Lichtenberg 2007). Vaccines have rid us of measles, mumps, and rubella. In 1950, patients who suffered a heart attack were treated with bed rest. Today, they are treated with an assortment of drugs and, if necessary, bypass surgery. Death rates for

290

Darius Lakdawalla, Anup Malani, and Julian Reif

cardiovascular disease have decreased by more than 60 percent over the past fi fty years (National Heart Lung and Blood Institute 2012). None of this is free, of course. Healthcare expenditure in the United States has grown rapidly, exceeding the annual growth in GDP by 2.5 percent since 1960. In 2011, the United States spent about 17.9 percent of its GDP on healthcare, or about $8,700 per person (Center for Medicare and Medicaid Services 2013). This rise has strained private and public budgets and is of central importance in the current national debate over health reform. Conventional wisdom among health economists is that the primary driver of cost growth is medical innovation.1 Newhouse (1992) argues that innovation is the only factor that can explain the vast increase in medical costs since 1950. Similarly, Cutler (2005) argues that medical innovation, though it has generated a large increase in life expectancy, is responsible for the recent rapid growth in costs. There is substantial heterogeneity in both the costs and benefits of different innovations, however, so one must be careful not to overgeneralize. Some innovations, such as using aspirin to treat acute myocardial infarction, cost almost nothing, while others, such as the invention of antiretrovirals for HIV/AIDS, cost $15,000 per year. Innovations that are both cheap and effective—like aspirin treatment for myocardial infarction—are clearly valuable. By contrast, expensive innovations that contribute to rising costs but offer little benefit may not increase social welfare. Vertebroplasty, for example, is an expensive procedure that injects cement into the spine to stabilize vertebrae but has been shown to be ineffective (Chandra and Skinner 2012). Many innovations, of course, fall between these two extremes. Antiretrovirals are costly but highly effective. Linkages between treatments complicate valuation. A cheap and effective diagnostic test for a disease that is treated with a costineffective treatment may actually increase costs with little to no health benefit. The ideal healthcare system would accomplish multiple objectives. It would provide consumers with quality healthcare at reasonable prices. It would provide incentives for innovation that improve the quality of care and thus health over time. Finally, an ideal system would insulate riskaverse individuals from health shocks and their fi nancial consequences. The task of healthcare reform is to inch our current system towards the ideal system. However, the complexity of markets for medical innovation, medical care, and fi nancing—and the connections between them—

Healthcare Reform and Innovation

291

makes it challenging to achieve all the ends of an ideal system. Some worthy objectives may counteract others, often in surprising ways. In this chapter we examine some of the interesting and important ways in which healthcare reform, particularly expansion of health insurance, affects innovation. We focus on three particular linkages between insurance markets and medical innovation. First, we examine how expansion of health insurance directly and indirectly affects the demand and supply of medical innovation. The expansion of insurance lowers the cost of fi nancing medical care, increasing demand for that care. This demand spills over into demand for future medical innovation (Acemoglu and Linn 2004; Finkelstein 2004; BlumeKohout and Sood 2008; de Mouzon et al. 2011; Clemens 2013). At the same time, the expansion of insurance weakens the incentives for patients to enroll in clinical trials, making it more costly for drug and device companies to develop innovations and reducing the profits they earn from those innovations (Malani and Philipson 2013). The net effects are uncertain. Yet because medical research in one country has value for all countries, the demand and supply effects from U.S. healthcare reform can have spillover impacts on the rest of the world (Lakdawalla, Goldman et al. 2009). Second, we examine how health insurance can cure an unfortunate side effect of using patents to encourage medical innovation: high prices that make medical care unaffordable for many. Health insurance has this benefit not because it bargains down prices from medical providers but because of the way in which it separates the price it pays providers for their care and the price it charges consumers for care (Lakdawalla and Sood 2013). Providers get a high (perhaps monopoly) price consistent with their negotiating power in the market, but consumers are charged a low copayment that may be close to or sometimes even below marginal cost. The result is both strong incentives to innovate and little deadweight loss due to underuse of medical care. Third, we examine how innovation both promotes health insurance and reduces demand for such insurance (Lakdawalla, Malani, and Reif 2013). It is healthcare, not health insurance, that reduces the negative consequences of sickness on health. Health insurance merely pays for healthcare, which would not exist but for innovation. Thus, innovation makes health insurance possible. Yet, if the price of healthcare is lowered, individuals face less fi nancial uncertainty for sickness, reducing the demand for health insurance. Thus medical innovation, depending on

292

Darius Lakdawalla, Anup Malani, and Julian Reif

how it is priced, can be a substitute for health insurance in the eyes of a risk-averse consumer. In the conclusion we argue that the complexity and uncertainty of the relationship between insurance and innovation puts a premium on ability to experiment with alternative forms of healthcare reform.

Insurance Expansion Affects Both the Quantity and Nature of Research and Development A. Direct Demand- Side Effects Profits drive innovation. Firms are more likely to invest in research and development if the rewards are large. In other words, the larger the potential market, the larger the investment. Clear evidence in favor of this hypothesis comes from a counterexample: Africa. Malaria and tuberculosis kill millions of people every year in Africa, yet companies invest little in fi nding cures for these diseases. The reason is simple. Africa is a poor continent and cannot afford to pay much for treatment. Investing large sums of money into research for these diseases is unlikely to generate profits for fi rms. The link between innovation and market size is the subject of numerous papers. Acemoglu and Linn (2004) show that a significant number of new drugs for the elderly were invented in response to the aging of the baby boomers in the United States. They estimate that a 1 percent increase in the size of the potential market leads to a 4 percent increase in the entry of new drugs. De Mouzon et al. (2011) examine the effect of the aging population on drug innovation in a global context. They also fi nd that potential market size has a significant effect on drug entry, although their estimate is an order of magnitude smaller than the one from Acemoglu and Linn. Whereas demographic changes like the aging of the population affect market size—in this instance, for drugs for the elderly—by mechanically increasing the number of potential customers, insurance expansions, which decrease the marginal price of medical care, affect market size by increasing the quantity demanded per capita. Blume-Kohout and Sood (2008) show that the enactment of Medicare Part D is associated with a significant increase in investment in drugs for the elderly. Similarly, Finkelstein (2004) fi nds that public policies that encourage vaccinations induce a two-and-a-half-fold increase in clinical trials for potential new vaccines.

Healthcare Reform and Innovation

293

These fi ndings imply that reducing insurance payment rates, although a politically expedient way to cut costs, may have adverse long-term consequences by depressing the incentives to innovate. Medicare currently reimburses providers below the levels paid by the typical private insurer, and Medicaid reimburses at even lower rates. The Independent Payment Advisory Board (IPAB), created by the Affordable Care Act, is expected to reduce Medicare spending by $2.4 billion over the next ten years (Congressional Budget Office 2011). The ACA prohibits the IPAB from rationing care, restricting benefits, raising premiums, or changing eligibility criteria, which means that reducing payment rates is the only feasible method for reducing spending. All of these reductions are expected to reduce fi rm profits and investment. Changes in market size have long-run effects not just on the United States but also on the world because innovation is a public good. Drugs invented for the U.S. market are eventually sold in foreign markets as well. It is well known that pharmaceutical firms earn substantially more revenue per capita in the United States than in Europe, where price controls lower the price paid to manufacturers (Danzon and Furukawa 2006). Increases in spending benefit both the United States and Europe; conversely, American spending cuts may have adverse international effects. Lakdawalla and Sood (2009) estimate that U.S. price controls that lower fi rm revenue by 20 percent would reduce life expectancy by 2.8 percent for both the United States and Europe in the long run. Although the cuts generate some fi nancial savings, the authors estimate that the costs of decreased innovation significantly outweigh the fiscal benefits. Although the effect of insurance on market size is important, Lakdawalla et al. (2013) note that this is not the only way insurance encourages innovation. They show that if consumers are risk averse then insurance also raises the value of innovation by allowing them to transfer fi nancial resources across different health states. This generates an increase in demand that reinforces the other effects previously discussed. One way of framing this issue is to note that insurance makes consumers better off and thus serves as an increase in wealth. This encourages further use of medical technology, and consequently raises the returns earned by innovators on their discoveries. For example, consider an individual who exists in one of two possible states, “healthy” or “sick.” A sick individual purchases a treatment to return to the healthy state. For simplicity, suppose the good is priced so high that it generates no consumer surplus in the conventional sense. Even so,

294

Darius Lakdawalla, Anup Malani, and Julian Reif

it is valuable to risk-averse consumers because it allows them to purchase access to the treatment prior to knowing whether they will be sick or not. The technology makes the individual relatively better off in the “sick” state by improving health. Since fi nancial payment for the technology is equal across states, this increases the value of the sick state overall. By the usual logic of insurance, compressing the difference between a “bad” state and a “good” state is valuable to a risk-averse consumer. Medical insurance affects not only the amount of innovation but also the types of innovation. Policies that include catastrophic coverage (i.e., coverage for severe illnesses that require prolonged hospitalization) encourage expensive, intensive procedures like ICU technology because consumers pay little or nothing for them. Similarly, policies with generous “fi rst dollar” coverage encourage investment in routine- care treatments like vaccines. Fee for service is the dominant payment schedule used by insurers in the United States. Understanding how insurance has affected innovation requires understanding the incentive effects of fee for service, which reimburses providers for each service they perform. It is widely blamed for increasing healthcare costs by providing incentives for physicians to provide more treatments regardless of the fi nal value to the patient. Cutler (2005) argues that it also encourages the development of intensive procedures like heart bypasses that create generous reimbursements for providers and manufacturers. Some of these procedures are valuable, but others are ineffective for at least some patients. Physicians will generally use the procedures irrespective of their cost so long as they provide at least a marginal health improvement for some patients. B. Indirect Effects through Subject Markets The previous section described how insurance expansions increase innovation by increasing market size or increasing the value of innovation. In other words, insurance expansions increase the demand for innovation. Health insurance also has a second, subtler effect on innovation through the market for human subjects. Drug regulators require that drug and device manufacturers conduct clinical trials to demonstrate that their products are safe and effective before allowing these manufacturers to sell their products. The medicalproduct market is unique in that potential participants in these clinical trials are also potential consumers of already-approved competing

Healthcare Reform and Innovation

295

products. Moreover, patients who participate in clinical trials cannot also consume a competing product. Malani and Philipson’s (2013) insight is that when deciding whether to enroll in a clinical trial, patients compare the benefits and costs of enrolling to the benefits and costs of “conventional care”—consuming an already-approved product. The merits of the trial are the expected medical benefits from the experimental treatment (and perhaps free medical care). The merits of the approved products are the health benefits of conventional care net of the costs of purchasing it. Anything that improves the merits of conventional care lowers the incentive of patients to enroll in trials. Malani and Philipson call this the “subject market effect.” The authors provide evidence for this effect by examining clinical trial recruitment before and after the regulatory approval of highly active antiretroviral therapy (HAART), a treatment for HIV/AIDS. As soon as it was approved HAART became the conventional treatment for HIV/AIDS because it was so effective at reversing that disease. 2 Because HAART constituted a dramatic improvement in the quality of conventional care, it reduced HIV-positive patients’ incentives to enroll in clinical trials. 3 Figure 12.1 shows that this innovative treatment, which became available in 1996, is associated with a steep decline in the fraction of HIV-positive patients who enrolled in clinical trials for new HIV/ AIDS treatments. The authors estimate that HAART reduced participation by 50– 75 percent and that this reduction was primarily driven by exits from existing trials rather than a reduction in the demand for new participants. Health insurance expansions can also reduce enrollment rates in clinical trials. Health insurance reduces the cost of purchasing the already-approved product because it charges patients a copay for that product rather than its full price. In this manner, health insurance makes already-approved products relatively more attractive than enrolling in a trial. Therefore, health insurance expansions reduce enrollment in clinical trials.4 This reduction in enrollment rates due to health insurance increases the time required to complete a clinical trial. Statistical considerations such as power and confidence require trials to enroll a given number of patients. In the United States, the Food and Drug Administration imposes such requirements when companies fi le for required approval to conduct clinical trials, during the so- called Investigational New Drug application process. The lower the enrollment rate per year, the more years are required to enroll a given number of patients.

296

Darius Lakdawalla, Anup Malani, and Julian Reif

Figure 12.1 Introduction of HAART reduces the rate of trial participation, by category of HIV patient. Notes: Data are from Multicenter Aids Cohort Study as described in Malani and Philipson (2013). Subjects with HIV are subjects in MACS with HIV. Subjects on HIV/AIDS medication are MACS subjects with HIV and on some medication to control HIV/AIDS.

The resulting delay in trials has two effects on innovation. First, it increases the opportunity cost of clinical trials. The delay ties up money that could be used for alternative investments or increases the repayment required for money borrowed to conduct trials. Because drug companies face a high cost of capital (Scherer 2010), this increases the cost of clinical trials. DiMasi, Hansen, and Grabowski (2003) estimate that opportunity costs are a substantial driver of the costs of clinical trials. In this manner, delays in clinical trials increase the cost of innovation, reducing its supply. Second, delays in completing a trial reduce the potential profits a company can earn from a medical product even if it is approved. Patents on medical products have a fi xed duration of twenty years. Because drug companies typically obtain patents prior to conducting clinical trials, and because they cannot sell drugs until trials are completed and the drugs are approved, the time required to conduct clinical trials reduces the amount of time the companies have to sell drugs under the patent.

Healthcare Reform and Innovation

297

This reduction in the so- called effective patent life reduces the expected profits from drugs. This demand-side effect reduces companies’ incentives to conduct medical research and development. To summarize, due to the subject market effect, health insurance expansions can reduce medical innovation in two ways. First, they increase the cost of trials of new treatments, which operates like an inward shift in the supply curve for innovation. Second, they reduce the expected profits from any innovation and operate like an inward shift in the demand for innovation. These two changes both tend to reduce the amount of medical innovation in the economy. 5 These two changes do not imply that health insurance reduces innovation, on net. The net effect depends on the combined effect of these two negative changes and the positive change discussed in the previous section. That positive effect was an outward shift in demand due to the prospect of additional quantity sales after a medical product is tested and approved for sale. It is possible that innovation could fall or rise. The important point is that the conventional wisdom that health insurance increases demand for innovation paints an incomplete picture of the effect of insurance on the amount of innovation because it neglects the subject market effects of insurance.

Insurance Expansions Reduce the Cost of Medical Innovation A primary method by which our legal system encourages innovation is by granting patents to innovators. A company that invents a new product is given a patent that bars any other company from producing that new product. The patent enables its holder to sell the new product at supracompetitive prices, earning correspondingly high profits. The prospect of these profits is what encouraged the company to invest in the research and development that generated the product. Without this research and development, no consumers would have enjoyed the product. Economists call this the dynamic efficiency from the patent. However, the high prices enabled by a patent also preclude some consumers from purchasing the new product after it is invented. Among those excluded by the patent are consumers who were willing to pay the marginal cost of the new product but not the supracompetitive price that patent holder charges. The loss borne by these consumers is what economists call the static inefficiency. For most products, the duration

298

Darius Lakdawalla, Anup Malani, and Julian Reif

of a patent is thought to trade off the dynamic efficiency from longer duration with the static inefficiency from that duration (Nordhaus 1969).6 In the healthcare context, however, Lakdawalla and Sood (2009) show that health insurance can provide a way out of this trade- off by decoupling the price that consumers pay from the price that innovators receive. Insurance does not charge beneficiaries the full cost of a drug, but rather a copay.7 This copay is not substantially different from the marginal cost of the drug. Yet insurance may reimburse innovators at a rate close to the price that the latter would like to charge under their patent. Thus, health insurance is able to increase welfare by increasing drug consumption without reducing incentives for innovation. This fi nding contrasts with the common intuition that insurance leads to inefficiency due to overconsumption of healthcare, that is, ex post moral hazard. According to this intuition, health insurance does not charge consumers the full price of care, but rather a (low) copay. This price reduction leads beneficiaries to consume more care than they would at full price. Because beneficiaries consume even when their willingness to pay is less than the full price of care, ordinarily we would say that the extra consumption is inefficient. But in the case of medical products covered by patent, we would say that the full price is “too high” relative to the social optimum as a result of market power due to the patent. However, unlike in traditional goods markets, health insurance provides a way out of this dilemma in the context of healthcare consumption. Specifically, it is efficient for health insurers to charge ex post copayments to consumers that approximate the efficient price for the treatment, even when monopolistic healthcare providers are receiving much higher prices upstream. 8 The effect of insurance on deadweight loss can be tested empirically by examining what happens to quantity sold when a drug’s patent expires. Theoretically, drugs that are well insured ought to exhibit lower deadweight loss due to patent monopolies. Therefore, patent expirations should have more modest effects on such drugs, compared to their lesswell-insured counterparts. Table 12.1 uses data from Lakdawalla and Sood (2013) to provide support for this hypothesis. It reveals that quantity sold increases for drugs that are not well insured, but changes negligibly for drugs covered by insurance. (Drugs not covered by insurance tend to be expensive, with prices far above marginal cost.) The results are consistent with the hypothesis that insurance limits deadweight loss from monopoly.

Healthcare Reform and Innovation

299

Table 12.1 Effect of Patent Expiration on Drug Quantity and Branded Drug Revenue by Insurance Penetration Effect of Patent Expiration On: Insurance Penetration Percentile 25th percentile Median 75th percentile

Insurer Share of Spending

Quantity

Branded Revenue

41% 57% 78%

+11% +5.3% −1.9%

− 22% − 30% −40%

Source: Data from Lakdawalla and Sood (2013).

A health insurance contract that charges an upfront premium and a (low) ex post out- of-pocket cost resembles a two-part tariff that is often used by monopolists to eliminate deadweight loss. An example illustrates the analogy. Consider a monopolist that produces healthcare and provides health insurance. By charging its customers a copayment equal to social marginal cost, this monopolist can ensure that consumers use care efficiently and derive the maximum possible gross consumer surplus. The monopolist then profits from this strategy by charging a premium equal to this gross consumer surplus. Here, consumers are willing to participate in the health insurance market, utilization occurs at the efficient level, and the fi rm earns profits equal to gross consumer surplus. While this simple example features a single fi rm producing both healthcare and insurance, its logic can be shown to extend to the real-world case in which separate fi rms provide healthcare and insurance (Lakdawalla and Sood 2013). This is the usual logic through which two-part pricing generates maximum profits and fi rst-best utilization (Oi 1971). The analogy to two-part pricing suggests that health insurance premiums must rise, as a by-product of the way in which health insurance solves the problem of static inefficiency from medical product patents. If insurance companies are charging beneficiaries a low price but paying providers a high price, the difference must be extracted via the premium charged to beneficiaries. This suggests that health insurance may just transfer static inefficiency from the medical product market to the health insurance market. There are three reasons to suspect that the static inefficiency in the insurance market is small. First, patented (or branded) drugs are a minor component of health expenditures and thus health insurance premiums. Total drug expenditures, including both branded and generic drugs, are approximately 10 percent of overall national health expenditures (U.S.

300

Darius Lakdawalla, Anup Malani, and Julian Reif

Census Bureau 2012 Table 136). Moreover, roughly 70 percent of all drug purchases are generic drugs (Hemphill and Sampat 2011). Therefore, branded- drug costs are a small driver of health insurance premiums. Second, the tax code subsidizes the purchase of employer-sponsored health insurance. It allows employers to expense their contributions to premiums and employees to deduct their contributions to premiums from taxable income. This proportional subsidy reduces the extent to which consumers bear the cost of higher insurance premiums (Gruber and Lettau 2004). Thus, higher premiums due to medical product patents are unlikely to reduce the consumption of insurance much. Not surprisingly, a survey of the empirical literature reveals that “demand for health insurance is, in general, price inelastic” (Liu and Chollet 2006). Third, the recently enacted Patient Protection and Affordable Care Act (ACA) virtually eliminates static inefficiency in the health insurance market. The ACA includes a so- called individual mandate that requires nearly all persons to buy health insurance.9 To ease the burden of this mandate, the ACA provides premium tax credits to individuals below 400 percent of the federal poverty level and expands Medicaid coverage for individuals too poor to afford health insurance.10

The Effect of Innovation on the Demand for Health Insurance Thus far we have focused on how health insurance affects the amount of innovation and the cost of using patents to spur innovation. In this section we turn the tables and focus on how innovation affects health insurance. Ostensibly this discussion is less relevant to healthcare reform, which in the United States typically focuses on expansion of health insurance. However, healthcare policy could focus on promoting innovation rather than insurance. In this section, we consider the benefits of this alternative policy strategy. Conventional analysis of new medical technologies often treats these technologies—or more precisely the need to pay for them and their high prices—as sources of fi nancial risk. This view leads naturally to the conclusion that healthcare reform should focus on providing insurance to reduce financial risk. In a recent working paper, Lakdawalla, Malani, and Reif (2013) explain that this view misses the forest for the trees. They argue that therapeutic medical innovation in particular plays a fundamental role in reducing health risks, as important as health insurance if

Healthcare Reform and Innovation

301

not more.11 As a result, innovation may be as important a policy goal as health insurance. To make their point, the authors explain how therapeutic medical innovations are more valuable to risk-averse individuals than they are to risk-neutral individuals, that is, more valuable to individuals who value insurance than to those who do not. To be sure, medical treatments have value to risk-neutral individuals, value that also accrues to risk-averse individuals. After they fall ill (or ex post), all individuals have some willingness to pay for medical treatment. If that treatment is offered at a price below that willingness to pay, economists say the individual has earned a consumer surplus. This is a benefit enjoyed regardless of risk preference. The value that a risk-neutral individual obtains from a medical treatment prior to falling ill (or ex ante) is the same as the value that individual obtains after falling ill, albeit discounted by the probability of falling ill. However, the value that the risk-averse individual obtains from that treatment ex ante is greater than the value she obtains from it ex post discounted by the probability of falling ill. The reason is that the riskaverse individual dislikes risk more than a risk-neutral individual does. Therapeutic medical treatments have particular value to such a person because they address that risk in two ways. First, risks to life and health are fundamentally physical, not fi nancial, and it is medical technology, not health insurance, that directly reduces these physical risks. Health insurance merely pays for medical treatment. Indeed, in the absence of medical technology, health insurance can do nothing to reduce risks to life and health. Once a medical treatment is developed, however, individuals face only the risk of having to pay for treatment, a risk for which they can buy market insurance. Another way to put this is that medical innovation converts uninsurable health risks into insurable fi nancial risks. Lakdawalla (2013) call this the marketinsurability value of innovation to risk-averse individuals. Second, if medical treatment is sold at a value below willingness to pay, the magnitude of risk faced by individuals is reduced. Instead of facing a high health risk, individuals face a smaller fi nancial risk. In the terminology of Ehrlich and Becker (1972), innovation therefore functions as a form of self-insurance. The authors call this the self-insurance value of innovation. A simple example can explain these two additional sources of value from innovation. Imagine the plight of an HIV-negative individual prior to the introduction of HAART. To make our calculations easy, suppose

302

Darius Lakdawalla, Anup Malani, and Julian Reif

that without HAART, an individual who gets HIV progresses to AIDS in five years and dies; with HAART, that individual lives 25 years at full health with HIV and then dies; and individuals do not discount the future. If the value of a life year is $100,000, then the HIV-negative individual confronts a health shock of $2 million (twenty years $100,000). Before HAART, the individual has no value for health-insurance coverage for HIV treatment even if she is risk averse: there is no treatment for her to purchase with health insurance. Once HAART comes along, however, health insurance has value. Even if HAART is priced at $100,000 per year,12 the risk-averse individual can use insurance to spread that cost out over HIV and non-HIV states. This is the market-insurability value of HAART to the HIV-negative risk-averse individual. Moreover, if HAART is sold—as it actually is—for $15,000 per year, then HAART reduces a $100,000 annual fi nancial shock by $85,000 (Steinbrook 2001). This is the self-insurance value of HAART. This reduces the amount of insurance a risk-averse person has to buy; if the insurance load were 10 percent, then the innovation would provide $8,500 of additional value to the risk-averse consumer. At this price, we can also compare the value of HAART to health insurance. A risk-averse individual obtains value from HAART because it reduces the cost of the HIV shock by $85,000. By contrast, insurance coverage for HAART only smooths out a $15,000 shock. The ratio—85/15 or 566 percent— yields the relative value of risk reduction from innovation and smoothing from health insurance in this (cooked-up) example. Lakdawalla et al. (2013) estimate the market-insurability and selfinsurance value of medical innovation using data on 1,257 medical treatments from Tufts University’s Cost-Effectiveness Analysis Registry. They report results for a hypothetical individual with a constant-relative risk-aversion utility function: u(w + h) =

1 (w + h)1−γ 1−γ

where w is wealth used for nonhealth consumption and h is health. The parameter ≥ 0 measures the degree of risk aversion. A value of = 0 implies risk neutrality; as  approaches infi nity, so does risk aversion. As Table 12.2 illustrates, Lakdawalla et al. (2013).fi nd that for a plausible range of risk aversion parameters, the median market-insurability value of insurance is 65–1061 percent of the ex post consumer surplus from innovation and the median self-insurance value of insurance is 9–169

Healthcare Reform and Innovation

303

Table 12.2. Estimates of the Market- Insurability and Self- Insurance Value of Innovation, by Level of Risk Aversion Surplus Gamma

P10

0.5 1.1 1.7 3 5

$0.11 $0.11 $0.11 $0.11 $0.11

Median $2.42 $2.42 $2.42 $2.42 $2.42

Self-insurance value P90

P10

$44.39 $44.39 $44.39 $44.39 $44.39

$0.00 $0.01 $0.01 $0.02 $0.04

Median $0.21 $0.51 $0.85 $1.81 $4.09

Market-insurance value

P90

P10

$6.19 $16.15 $29.64 $69.21 $194.13

$0.00 $0.00 $0.00 $0.00 $0.01

Median $1.58 $3.67 $6.11 $11.99 $25.67

P90 $24.57 $62.04 $105.96 $235.34 $602.90

Notes: Table is from Lakdawalla, Malani, and Reif (2013). P10 and P90 indicate 10th and 90th percentile, respectively.

percent of that consumer surplus. This implies that medical innovation may have significantly more value to a risk-averse person than a risk-neutral one. The additional value that new medical treatments offer to risk-averse individuals (relative to risk-neutral individuals) also illustrates the complicated relationship between medical innovation and health insurance. On the one hand, the market-insurability value of medical innovation suggests that innovation and health insurance are complements. Medical innovation makes health insurance more valuable. Policies that neglect or reduce the amount of medical innovation may thereby reduce the potential value of health insurance, even public health insurance. On the other hand, the self-insurance value of medical innovation suggests that innovation and insurance are substitutes. The lower the price of medical innovation, the more innovation provides self-insurance and the less need there is for market health insurance. Thus, lower prices for medical innovations are a substitute for health insurance. One might be tempted to suggest that this logic suggests that price controls are a (cheap) substitute for health insurance expansions. To some extent this is true. Given the innovations currently on the market, price controls increase the self-insurance from those innovations. However, price controls also reduce the potential profits to innovators and thereby discourage future innovation. In the long run, price controls may just shift self-insurance from the future to the present, and perhaps even reduce total self-insurance value from a present-value perspective. Moreover, price controls forego future market-insurability value from the future innovation that they deter. Given the relative magnitude of the estimates of market-insurability and self-insurance value in Table 12.2, this second cost could dwarf the short-run self-insurance value from price controls.

304

Darius Lakdawalla, Anup Malani, and Julian Reif

Conclusion We have argued that the relationship between innovation and health reform involves a wide variety of causal relationships. The complexity of this relationship makes it hard to predict with great certainty what the effects of new policies, like the ACA, are likely to be. Local experimentation, whereby different areas implement different reforms, could provide valuable data on the effectiveness of different reforms. Successful reforms can then be adopted gradually by different localities or eventually implemented by a centralized authority. The creation of Accountable Care Organizations by the ACA is a partial step in this direction. These organizations are allowed to experiment with different methods of caring for patients, although they are subject to certain quality requirements, such as reducing hospital readmissions. The ACA encourages companies to reduce costs by allowing them to keep some of the savings they generate. Success could create a system wherein providers are rewarded for quality rather than quantity. To be sure, there are significant risks with a decentralized approach. States that experiment with more generous public programs, for instance, risk attracting migrants in search of these subsidies. The result can be strategic interaction across states, which end up competing to keep unattractive migrants out rather than to construct efficient health policy. Evidence in the context of welfare recipients suggests that this is an important dynamic (Gelbach 2004). This problem becomes even more acute at the level of small areas or localities. For instance, if a city began to impose community-rating requirements, healthy people might flee to the suburbs, while only the sickest consumers would remain. Finally, decentralized health policy that involves cost sharing between federal and state governments creates additional problems. States have an incentive to foist expenditures onto the federal government and vice versa. A good example is the interaction between federally funded Medicare and Medicaid programs that involve partial state funding. States have little incentive to prevent acute illness in elderly residents of longterm care institutions. Acutely ill long-term care patients get admitted to hospitals, where they become Medicare’s fi nancial liability. While they remain in long-term care facilities, on the other hand, they are at least in part the responsibility of the state.

Healthcare Reform and Innovation

305

On balance, there are good reasons to promote decentralization in healthcare policy, but the practical aspects of decentralizing policy frequently lead to wasteful confl icts between states and between state and federal governments. The latter dynamic would disappear in a system in which state governments had complete control and fully internalized the costs of public health-insurance schemes. The former effect could be solved by state-level policies that promoted efficiency and tended to draw productive fi rms and citizens into successful states.

Notes Darius Lakdawalla is grateful to the National Institute on Aging for fi nancial support (grant numbers RC4 AG039036 and P01 AG033559). 1. Of course, there are many other contributors to cost growth. Healthcare is a normal good, so one expects demand to increase as a country becomes wealthier. The U.S. population is aging rapidly, which means a larger portion of the population is elderly and in need of expensive medical care. The expansion of health insurance coverage to a majority of the population increases the accessibility of medical care, but may also encourage inefficient spending if consumers do not face the full marginal price of their healthcare. Physicians may recommend treatments based on profitability rather than patient value, or may overprovide (practice “defensive medicine”) in order to combat medical malpractice lawsuits. Newhouse (1992) argues that the fi rst three factors cannot explain much of the growth in healthcare costs between 1950 and 1990. Mello et al. (2010) suggests medical malpractice premiums are a small percentage of total healthcare costs. 2. HAART did not cure HIV. It reduced the replication of the HIV virus so that patients did not experience a full collapse of their immune system, a fatal condition known as AIDS. 3. Improvements in the quality of conventional care do not always reduce enrollment in all clinical trials. Medical innovation in one market may affect innovation in another market positively. For example, innovation in the treatment of heart disease will result in a larger population facing comorbid risks such as Alzheimer’s disease. This increases both the demand for Alzheimer’s research and the supply of available clinical trial participants for that research. As a result, while innovation may be self-limiting within a disease, it may increase the returns to innovation in other diseases. 4. Using the same logic, price controls also reduce enrollment rates in clinical trials. 5. Malani and Philipson (2013) note that the subject market effect, by itself, does not demonstrate inefficiency in the market for medical innovation. Just as

306

Darius Lakdawalla, Anup Malani, and Julian Reif

an increase in demand due to aging may efficiently improve the quantity of innovation, reductions in the supply of innovation due to health insurance or improvements in the quality of conventional care may efficiently reduce the quantity of innovation. Better health insurance or conventional care may represent increases in the real opportunity costs for patients to enroll in clinical trials. The only feature of the subject market that suggests adjustments in that market may yield inefficient levels of innovation is the fact that ethical rules governing clinical trials generally cap the amount that companies can pay patients to enroll in their clinical trials. Just as rent control can lead to inefficiencies in the allocation of housing, these ethical wage controls can lead to inefficiencies in the allocation of patients to trials. 6. The dynamic efficiency and static inefficiency are not the only benefits and cost of the patent system. For example, there is some cost associated with multiple companies investing in research and development to develop a given product. Since only one company obtains the patent, the investments of other companies are, in hindsight, a waste (Loury 1979). Another example is that because a patent does not allow the holder to expropriate all the social surplus from the new product, it gives suboptimal incentives for innovation (Lakdawalla, Goldman et al. 2009). 7. The beneficiary will typically have to pay a deductible, but that is not specific to a given treatment. After paying the deductible, the beneficiary is only responsible for the copay, up until the annual or lifetime limits kick in. The recently enacted Patient Protection and Affordable Care Act bars such limits in the health insurance plans to which it applies (Healthcare.gov 2013). 8. In theory, if the copay is less than social marginal cost even for patented products, then beneficiaries may overconsume care relative to the social optimum. In the context of the conventional pharmaceutical market, this is unlikely to be the case, because marginal cost is so low (Caves, Whinston, and Hurwitz 1991). As a result, health insurance is almost certainly welfare improving in the pharmaceutical context. 9. (26 U.S.C. § 5000A). 10. Of course the pre-ACA tax subsidy for the purchase of health insurance and the ACA’s new premium tax credits and Medicaid expansion may increase government expenditures on healthcare. But for the same reasons that branded drug expenditures are likely to have only nominal effects on insurance premiums, they are likely to have negligible effects on government expenditures. Moreover, the effects on government expenditures are smaller since the government shares only a portion of the cost of health insurance premiums. Finally, the welfare effects depend on how the government fi nances those additional expenditures. If the government cuts other less efficient expenditures, the causal effect of higher insurance premiums on welfare may be zero or even negative. 11. The authors do not extend their analysis to preventative medical innovations or to diagnostic medical innovations, though they intend to explore these other innovations in future work.

Healthcare Reform and Innovation

307

12. We assume the HIV-positive individual begins HAART five years after diagnosis with HIV, when the viral load rises such that the individual is at risk for AIDS.

References Acemoglu, Daron, and Joshua Linn. 2004. “Market Size in Innovation: Theory and Evidence from the Pharmaceutical Industry.” Quarterly Journal of Economics 119 (3): 1049– 90. Arias, Elizabeth. 2012. “United States Life Tables, 2008.” National Vital Statistics Reports 61 (3): 1– 64. Blume-Kohout, Margaret E., and Neeraj Sood. 2008. The Impact of Medicare Part D on Pharmaceutical R&D. National Bureau of Economic Research. Caves, Richard E., Michael D. Whinston, and Mark Hurwitz. 1991. “Patent Expiration, Entry, and Competition in the US Pharmaceutical Industry.” Brookings Papers on Economic Activity: Microeconomics 1991: 1– 66. Center for Medicare and Medicaid Services. 2013. “National Health Expenditure Data: Historical.” Accessed October 29, 2014. http://www.cms.gov/Re search-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/National HealthExpendData/NationalHealthAccountsHistorical.html. Chandra, Amitabh B., and Jonathan Skinner. 2012. “Technology Growth and Expenditure Growth in Health Care.” Journal of Economic Literature 50 (3): 645– 80. Clemens, Jeffrey. 2013. “The Effect of US Health Insurance Expansions on Medical Innovation.” NBER Working Papers 19761. National Bureau for Economic Research, Inc. Congressional Budget Office. 2011. Scoring of Proposed Changes to the Independent Payment Advisory Board Mechanism. Washington, DC.: Congressional Budget Office. Cutler, David M. 2005. Your Money or Your Life: Strong Medicine for America’s Health Care System. New York. Oxford University Press. ———. 2007. “The Lifetime Costs and Benefits of Medical Technology.” Journal of Health Economics 26 (6): 1081–1100. Cutler, David M., S. Kadiyala, et al. 2003. “The return to biomedical research: Treatment and behavioral effects.” In Measuring the Gains from Medical Research, edited by Jevin M. Murphy and Robert H. Topel. Chicago: University of Chicago Press. Danzon, Patricia M., and Michael F. Furukawa 2006. “Prices and Availability of Biopharmaceuticals: An International Comparison.” Health Affairs 25 (5): 1353– 62. de Mouzon, Oliver, Pierre Dubois, Fiona M. Scott Morton, and Paul Seabright. 2011. Market Size and Pharmaceutical Innovation, CEPR Discussion Papers.

308

Darius Lakdawalla, Anup Malani, and Julian Reif

DiMasi, Joseph A., Ronald W. Hansen, and Henry G. Grabowski. 2003. “The Price of Innovation: New Estimates of Drug Development Costs.” Journal of Health Economics 22 (2): 151– 85. Finkelstein, Amy. 2004. “Static and Dynamic Effects of Health Policy: Evidence from the Vaccine Industry.” The Quarterly Journal of Economics 119 (2): 527– 64. Gelbach, Jonah B. 2004. “Migration, the Life Cycle, and State Benefits: How Low Is the Bottom?” Journal of Political Economy 112 (5): 1091–1130. Gruber, Jonathan, and Michael Lettau. 2004. “How Elastic is the Firm’s Demand for Health Insurance?” Journal of Public Economics 88 (7): 1273– 93. Healthcare.gov. 2013. “Lifetime & Annual Limits.” Accessed March 25. http:// www.healthcare.gov/law/features/costs/limits/index.html. Hemphill, C. Scott, and Bhaven N. Sampat. 2011. “When Do Generics Challenge Drug Patents?” Journal of Empirical Legal Studies 8 (4): 613–49. Lakdawalla, Darius N., Dana P. Goldman, Pierre- Carl Michaud, Neeraj Sood, Robert Lempert, Ze Cong, Han de Vries and Italo Gutierrez. 2009. “U.S. Pharmaceutical Policy in a Global Marketplace.” Health Affairs 28 (1): w138–50. Lakdawalla, Darius N., Anup Malani, and Julian Reif. 2013. “The Insurance Value of Medical Innovation.” The Petrie Flom Center for Health Law Policy, Biotechnology, and Bioethics at Harvard Law School Working Papers. Accessed October 29, 2014. http://petrieflom.law.harvard.edu/assets/publica tions/Malani.pdf. Lakdawalla, Darius N., and Neeraj Sood 2009. “Innovation and the Welfare Effects of Public Drug Insurance.” Journal of Public Economics 93 (3–4): 541– 548. ———. 2013. “Health Insurance as a Two-Part Pricing Contract.” Journal of Public Economics. Lichtenberg, Frank R. 2007. “The Impact of New Drugs on US Longevity and Medical Expenditure, 1990-2003: Evidence from Longitudinal, DiseaseLevel Data.” American Economic Review 97 (2): 438–43. Liu, Su, and Deborah Chollet 2006 “Price and Income Elasticity of the Demand for Health Insurance and Health Care Services: A Critical Review of the Literature.” Ref. No. 6203- 042. Washington, DC: Mathematica Policy Research Loury, Glenn C. 1979. “Market Structure and Innovation.” Quarterly Journal of Economics 93 (3): 395–410. Malani, Anup, and Tomas J. Philipson. 2013. The Economics of Medical R&D: The Subject Market Effect. Mello, Michelle, Amitabh Chandra, Atul A. Gawande, and David Studdert. 2010. “National Costs of the Medical Liability System.” Health Affairs 29(9): 1569– 77. Murphy, Kevin M., and Robert H. Topel. 2006. “The Value of Health and Longevity.” Journal of Political Economy 114 (5): 871– 904.

Healthcare Reform and Innovation

309

National Heart Lung and Blood Institute. 2012. Morbidity & Mortality: 2012 Chart Book on Cardiovascular, Lung, and Blood Diseases. Washington, DC: National Institutes of Health. Newhouse, Joseph P. 1992. “Medical Care Costs: How Much Welfare Loss?” Journal of Economic Perspectives 6 (3): 3– 21. Nordhaus, William D.1969. Invention, Growth and Welfare: A theoretical Treatment of Technological Change. Cambridge, MA: MIT Press. Oi, Walter Y. 1971. “A Disneyland Dilemma: Two-Part Tariffs for a Mickey Mouse Monopoly.” Quarterly Journal of Economics 85: 77– 96. Scherer, F. M. 2010. Chapter 12—“Pharmaceutical Innovation.” In Handbook of the Economics of Innovation, vol. 1, edited by H. H. Bronwyn and R. Nathan, 539– 74.Amsterdam: North-Holland. Steinbrook, Robert. 2001. “Providing Antiretroviral Therapy for HIV Infection.” New England Journal of Medicine 344 (11): 844–46. U.S. Census Bureau. 2012. Statistical Abstract of the United States 2012. Washington, DC: U.S. Department of Commerce.

Chapter thirteen

The Affordable Care Act and Commercial Health Insurance Markets Fixing What’s Broken? James B. Rebitzer

Introduction

T

he Affordable Care Act (ACA or the Act) is a famously complex piece of legislation designed to address many problems in the U.S. health care fi nance and delivery system. Some parts of the Act are concerned with ensuring access to health insurance through a combination of subsidies, mandates, and expansions of Medicaid. Other parts of the ACA focus on policies to reduce the rate of growth of healthcare costs. Still other parts of the Act seek to improve the operation of commercial health insurance markets. This chapter concerns the impact of the ACA on the market for commercial health insurance. The analysis is conventional in that it considers whether and how the ACA can address failures in the market for commercial health insurance. The analysis is unconventional, however, in two respects. First, it emphasizes important but often overlooked heterogeneity in commercial health insurance markets. Secondly, the focal insurance market failure is neither moral hazard nor adverse selection. Rather, attention is focused on search frictions. “Search frictions” refers to any factor that inhibits the flow of information through a market. Frictions may emerge from the confusion or

314

James B. Rebitzer

bounded rationality of market participants—especially when the transaction involves complex contingencies or hard-to-observe product attributes. Alternatively frictions can result from more mundane factors such as the time and effort costs of shopping and negotiation (Rebitzer and Votruba 2011). Whatever their cause, search frictions give one party to the transaction some degree of market power over counterparties. In the case of labor markets, where models of search frictions have had great influence and yielded a recent Nobel Prize, frictions give employers some degree of monopsony power over employees. In many product markets—and as we shall see, in insurance markets—frictions enable sellers to exercise some monopoly power over buyers. As in frictionless markets, this market power has the effect of raising prices and transferring consumer surplus to sellers. A bit of institutional background is required to understand the role of search frictions in health insurance markets. The market for commercial group health insurance has two large segments. The fi rst is the fully insured segment, in which relatively small and unsophisticated employers buy insurance from commercial health insurance companies on behalf of their employees. The second market segment, often referred to as the self-insured segment, features larger and more sophisticated employers who bear the risk of their employees’ healthcare costs and use insurance companies to help design and administer their plans. A striking difference between these two segments is that rates of turnover are much higher in the fully insured market than the self-insured market. This high turnover appears not to be the result of normal labor-market mobility in which individual members leave an insurer when they change jobs. Rather high mobility in the fully insured market is the result of entire employer groups changing their insurers with great alacrity. Cebul et al. (2011) present evidence from a private insurer indicating that this type of turnover may be 18 percent per year. From the perspective of microeconomics, employer group turnover at this rate is puzzling. In well-functioning markets the law of one price prevails, meaning that similar products sell for similar prices to similar consumers. Given the substantial transaction costs to switching insurers, the prices offered by outside insurance companies to induce movement away from incumbent insurers must have been substantially less than the prices offered by incumbents. What is it about competition in health insurance that inhibits the workings of the law of one price? Why are these issues manifest in the fully insured rather than the self-insured part of the market?

Affordable Care Act and Commercial Health Insurance

315

Models of search frictions offer a parsimonious answer to both these questions. In frictional markets the law of one price does not hold. In place of a single price, an equilibrium distribution of prices for identical products emerges in these markets. Given this distribution it is no surprise that frictional insurance markets exhibit high rates of turnover— there are deals to be had for employers who switch insurers. The fully insured commercial health insurance market is more affl icted by health frictions than the self-insured market because the purchasers are less sophisticated and because they are buying a more complex product from insurers. Insurance brokers, who are quite prevalent in the fully insured market, compound the problem of frictions when they fail to recommend the low-price equivalent of high-price policies to their clients. This chapter examines how the various features of the ACA may mitigate friction-related distortions. The discussion proceeds as follows. The next section develops the microeconomic logic of search frictions in insurance markets more fully and discusses the distortions they create in some detail. The subsequent three sections discuss policy questions that emerge from our analysis: Can market efficiency be improved by policies that alter the equilibrium distribution of insurance prices? Can efficiency be enhanced by simplifying the search process within insurance markets? Can efficiency be increased by promoting more efficient investments in future health? The ACA is relevant to each of these questions because it focuses new exchanges and regulations on the individual and small group markets where search frictions are most severe.

The Logic of Search Frictions in Insurance Markets Equilibrium Price Distributions and Turnover At the most basic level, search frictions make it hard for buyers to comparison shop. If outside offers do not reach potential purchasers or if these offers cannot be easily compared to current policies, search frictions enable sellers to charge prices above marginal cost. A second, less intuitive, feature of frictional markets is that they do not have a single equilibrium price; rather, equilibrium is characterized by a distribution of prices. Consider a market for a set of actuarially equivalent insurance policies that are sold to identical customers. Suppose as well that the average cost of insurance is equal to its marginal cost, so that insurers who set

316

James B. Rebitzer

premiums at marginal cost will earn zero economic profit while those who set premiums above marginal cost will earn positive economic profits on each policy sold. If all consumers are well informed about all price offerings, the equilibrium in this market must be a single price with premiums set equal to the marginal costs of insurance. Any premium above marginal cost will attract no customers and so insurers will not post these high prices. Now suppose that information about prices flows imperfectly through the market. In this case, setting insurance premiums equal to marginal cost cannot be an equilibrium. The policies with premiums equal to marginal cost earn zero economic profit, but policies with a price above marginal cost would earn positive economic profits. Why? Because imperfect information flows create the possibility that the high-priced policy is the best price that at least some consumers encounter. So long as there is a positive probability that consumers purchase a policy at a price above marginal cost, the expected economic profit for this pricing strategy will be greater than zero. The same logic we used to establish that the equilibrium premium must exceed marginal cost can be extended to make the stronger point that equilibrium is characterized by a distribution of prices rather than a single price. The key to this result is the commonsense observation that even in frictional markets, customers will be disproportionately drawn towards the lowest-priced products. Thus if all insurers decided to coordinate on a single premium greater than marginal cost, a policy with a slightly lower price would succeed in a attracting a large number of customers. Indeed a single low-price policy would be more profitable than any of the other policies offered at the single higher price because it would attract so many customers that the increase in volume would more than offset the smaller markup. If we stipulated that two policies were offered at the low price rather than one, these two policies would have to divide the flow of customers between them, and this reduces the profits of each individual policy offered at the low price. The movement of a second policy to the low price point allows the high-price market share to be divided among a smaller number of fi rms and so increases profits at the high price point. Following this logic, as the number of policies offered at the low price grows, each policy’s profits fall until the low price and high-price policies generate the same economic profit. In equilibrium, a fraction of policies will be offered at the low price point (which nevertheless exceeds marginal cost) and the remaining policies will be high

Affordable Care Act and Commercial Health Insurance

317

price. Individual policies will have no incentive to switch from low price to high-price strategies because they are equally profitable.1 It is easy to show that the equilibrium distribution of prices will be right-skewed in the sense that there will be more low-price policies than high-price policies and that the mean insurance premium exceeds the median (Cebul et. al. 2011). The fi nding that the equilibrium distribution of prices is everywhere above marginal cost reflects friction-induced insurer market power. The greater the frictions, the larger the average price in the market and the larger the share of consumer surplus captured by insurers. Persistent price differentials for identical products also create incentives for employers and individuals to switch insurers in search of lower- cost insurance. The extent of this turnover also depends on the magnitude of frictions, but the relationship is not linear. If frictions are very small, the distribution of prices degenerates to something quite close to marginal- cost pricing and the incentive to switch policies becomes quite small. Similarly at the other end of the spectrum, if frictions are so great that policyholders have almost no information about outside offers, the distribution of prices degenerates to something close to consumers’ maximum willingness to pay. In this case too, the gain from switching policies is also small. High rates of turnover emerge in frictional markets when frictions are moderate so that the spread of prices is large. Cebul et. al. (2011) fit a structural search model to information on price distributions. They report that insurers set prices high enough to capture 13.2 percent of consumer surplus (about $34 billion in 1997); turnover at the average policy was about 64 percent higher than it would have been in the absence of search frictions. These results are what one would expect with modest search frictions. Others have also found evidence consistent with insurer market power in the fully insured health insurance market. Dafny (2010) reports that the premiums of a sample of large fully insured fi rms increase with positive profit shocks. This sort of “direct price discrimination” is feasible only in imperfectly competitive settings. Brokers and Failures in the Market for Expert Advice: If search is impeded by frictions it is natural to expect a market for expert search advisers to evolve, and that is what happens. Most employers in the fully insured market rely on insurance brokers to assist them in

318

James B. Rebitzer

their search for insurance. For search frictions to persist in the presence of such expert advisers, it must be the case that a nontrivial number of brokers fail to act in the best interest of their clients. Widespread broker malfeasance of this sort is feasible in principle because it is difficult for employers to directly assess the advice they get from their brokers. In addition, brokers receive commissions from insurance companies and can profit by steering clients towards products paying the highest commissions rather than products that serve client needs most cost efficiently. This odd arrangement in which expert advisers are paid by the other side of the transaction is common to many fi nancial products and services (Jackson 2008). Many times these side payments are not clearly disclosed to consumers, and sometimes they are made in secret. Recent studies fi nd that these confl icts of interest lead to consumer losses in these other fi nancial services markets as well. Woodward and Hall (2011), for example, study the market for mortgage origination. At the time of the study, mortgage brokers received side payments for persuading borrowers to agree to loans with higher interest rates. This confl ict of interest would seem to make comparison shopping among brokers prudent, yet the data suggest that borrowers considered bids from no more than two brokers. Woodward and Hall estimate that a borrower seeking a mortgage with a $200,000 principal amount could save $1866 by talking with a third broker. Passing up such large gains does not seem sensible, and the authors conclude that confusion about how the market works causes borrowers to trust brokers too much and therefore to shop too little. From an economic perspective, these failures in the market for expert advice raise a number of important questions. If brokers offer biased advice, why do employers not learn to adjust for their broker’s biases? Failing this, why does a market for independent brokers who take no commissions not emerge? Sadly there are no settled answers to these important questions. A number of recent small- scale experiments tackle this first question (Cain, Loewenstein, and Moore 2005, 2011; Cain and Detsky, 2008) They fi nd that subjects generally do not discount advice from biased advisers as much as they should, even when confl icts of interest are disclosed. In addition, they fi nd that disclosure actually increases the bias in adviser advice and can worsen the quality of advice in this way. Gabbaix and Laibson’s (2006) model of hidden attributes offers some potential insight into why markets for independent brokers may not

Affordable Care Act and Commercial Health Insurance

319

spontaneously appear in response to confl icts of interest resulting from commission payments by insurers. In this model, fi rms exploit myopic consumers by shrouding high-priced add- ons. Shrewd consumers cannot be fooled by such tactics. Indeed they exploit fi rms by avoiding the highpriced add- ons, and as a result receive basic services at low prices that are subsidized by the exploitation of myopic consumers. Although nobody has yet done this, it seems straightforward to extend Gabbaix and Laibson’s model to include brokers. Brokers, who know a great deal about their clients, may be very good at identifying who is a myope and who is a shrewd customer, steering the former towards exploitative policies paying high commissions and the latter towards low-price/low- commission policies. Independent brokers would have a hard time entering such a market because they could not profitably attract shrewd shoppers. Similarly, they could not attract myopes without turning them into shrewd shoppers. If this extension of Gabbaix and Laibson’s model is correct, market competition cannot be relied upon to produce efficient, independent brokers. Frictions and Adverse Selection In frictional insurance markets, insurers gain market power because their customers are at an informational disadvantage—they do not know the entire distribution of outside prices. In insurance markets with adverse selection, it is the insurers who are at an informational disadvantage because buyers know more about their health status than insurers. On this basis one may well wonder how search frictions and adverse selection interact. This question has not yet received much attention from microeconomists, but there is good reason to believe that adverse selection reinforces frictions. 2 To see this, consider an insurer attempting to bid an employer group away from an incumbent insurer. The outside insurer must consider the possibility that they might offer too low a price to the employer group, especially if the better-informed incumbent insurer chooses not to match the outsider’s price. Thus adverse selection can magnify the effect of search frictions by dampening incentives to bid new business away from incumbent insurers. It follows that features of the ACA designed to reduce adverse selection in the individual and small-group markets can have the effect of reducing the importance of search frictions.

320

James B. Rebitzer

Improving Efficiency by Altering the Distribution of Insurance Prices? Market frictions lead to an equilibrium distribution of prices such that the entire distribution sits above marginal costs. An important implication of this result is that insurers can always profitably sell an additional insurance policy. In frictional markets, the most natural way to increase sales is to spend money on marketing and there are good theoretical reasons to expect an “arms race” in marketing expenditures. This arms race, in turn, suggests the possibility of efficiency gains from using public policy to alter the equilibrium distribution of prices. Suppose an insurer tries to attract more clients by raising the commission it offers independent brokers. 3 The higher commission will no doubt attract the broker’s attention and reduce efforts to sell competing insurance policies, but this incentive effect will be eliminated when other insurers match the fi rst movers’ new commission. As a result of this escalation in premiums, the original insurance company causes the other insurance companies to incur higher marketing costs without any company gaining an enduring competitive advantage. The amount of money tied up in this wasteful marketing arms race is determined by the insurer with the most to gain from making additional sales. In frictional markets this will, of course, be the insurer whose premiums are most heavily marked up over marginal cost. In other words, the high-price part of the equilibrium distribution of prices drives a marketing arms race that causes inefficient marketing expenditures to cascade throughout the entire insurance market. It follows from this that any policy that thins out the high-priced part of the price distribution can improve the efficiency of the entire insurance market. Cebul et. al. (2011) make this point by considering the effect of a hypothetical public insurance option. Employers can either buy insurance from the government or accept a better offer from the private sector if they fi nd one. In this model public insurance is, by assumption, less desirable than private insurance and so it will not completely crowd out private insurance. The primary economic function of the public option in this setting is to serve as a widely publicized backstop policy to limit the highest price offered in the private sector. By depressing the highest prices in the distribution of prices, a public insurance option can reduce the private sector’s burden from an inefficient marketing arms race. Indeed in the stylized Cebul et. al. (2011) framework these efficiency gains

Affordable Care Act and Commercial Health Insurance

321

are sufficiently great that the socially optimal government policy would be to price the public insurance backstop option below marginal cost. A public option is not the only way to thin out the high-price part of the distribution of prices. The approach adopted currently in the ACA is to specify minimum medical loss ratios: at least 80 percent of small- and medium-sized group premiums must go to pay medical claims or services. For large groups this minimum is set at 85 percent. If premiums exceed the minimum medical loss ratio, insurers are required to return money to their customers. Rebates from 2011 experience totaled $1.1 billion with an average rebate per household of $151. (Harrington 2012) The message from this section is that in frictional markets, competition does not get prices right, and there is a possibility of enhancing efficiency by thinning out the right tail of the distribution of prices. It is worth noting, however, that clumsy government policy can make things worse rather than better. For example, the regulations classify some “cost reduction” expenses as administrative expenses and others as expenditures on “quality improvement,” which counts towards meeting the minimum medical loss ratio (ibid.). With potentially costly rebates at stake, one can expect insurers and regulators to fight over what constitutes a cost reduction or quality improvement expenditure. In addition, it is important to note that insurers do not have full control over their medical costs in any given year (ibid.). If medical costs end up being lower than projected insurers face the risk of paying rebates to customers to achieve mandated minimum medical loss ratios, but if they end up being higher than projected they lose money as well. Since the latter loss is uncapped and the former is capped by the medical loss ratio, insurers might reasonably see higher premiums than they would have in the absence of the minimum medical loss requirement. These sorts of administrative complications are not limited to minimum medical loss ratio rules. Publicly subsidized public insurance options are similarly not immune from the costs of clumsy policy. It is, for example, not hard to imagine that publicly subsidized public insurance options might be priced to crowd out too much of the private insurance sector. If price caps are hard to administer, what about policies that place a cap on the distribution of prices? Ericson and Starc (2012) have studied the effect of such limitations in the Massachusetts healthcare reform. Under the law, insurers can charge their older customers more than younger ones for actuarially equivalent plans, but this differential cannot exceed 2:1. Older customers are less price sensitive than younger

322

James B. Rebitzer

ones. As a result, insurers raise the prices of the youngest customers 8 percent so that they can increase prices on the less inelastic, older market segment.4 Given the many policy pitfalls of directly influencing prices, policy makers might want to consider alternative strategies for reducing inefficiencies resulting from search frictions. Of these, perhaps the most direct is to try simplifying the search process using modern information technology. This is a topic we take up in the next section.

Improving Efficiency by Simplifying Search? The ACA aims to reduce search frictions by simplifying search. The new insurance exchanges it mandates will enable web-based shopping, require standardized products, and cap price differentials across plans. It is not yet clear, however, how effective these strategies will be in eliminating search frictions. We can learn something about the effectiveness of web-based shopping in eliminating search frictions from recent experience with the market for Medicare Part D pharmaceutical benefits. Under Medicare Part D, seniors and their families could select from a wide range of plans. Information on formularies, copays, and deductibles was widely disseminated on the web. In addition, Medicare made available a decision tool, Plan Finder, that recommends plans based on drugs consumed in the current year. A number of recent studies have examined how efficiently individuals selected their Medicare Part D plans. The results are mixed. Abaluck and Gruber (2011) found that plan choice fell considerably short of the efficient benchmark. Elders, according to this study, put more weight on plan premiums than on expected out-of-pocket costs, and placed almost no value on variance-reducing aspects of plans. The authors estimate that welfare would have been 27 percent higher if patients had chosen plans rationally. Heiss et. al. (2012) estimate in a separate analysis that enrollees lost about $300 per year on average by not optimizing effectively. In contrast, Ketcham et. al. (2012) fi nd that overspending declines markedly over time, especially among those who initially overspent the most. We can also learn something about the effects of product and price standardization in insurance exchanges from the experience of Massachusetts’s Connector, a precursor to the exchanges mandated under

Affordable Care Act and Commercial Health Insurance

323

the ACA. Plans in the Connector are placed into bands (bronze, silver, gold) based on their actuarial values. In addition insurers could offer a young-adult plan with an actuarial value lower than bronze. A substantial fraction of buyers chose the least generous tier plan (bronze), which is less generous than employer plans; these tend to be in the silver and gold range. Prior to 2010, insurers had wide latitude to introduce plans and there was substantial variation in premiums among actuarially similar plans. Beginning in 2010, however, the Connector introduced more standardization. Insurers were limited to only seven defi ned plans (two in each actuarial band and one young adult). Within each “metal” band, plans had to offer the same deductibles, copays, and coinsurance. Even in this more standardized setting, prices varied—presumably because the plans were not perfect substitutes. Plans differed, for example, in their physician networks and reputations. Standardization seemed to have the effect of shifting consumers into more generous products, but it had no effect on the price offered by insurers. More generally, there are theoretical reasons to wonder whether reductions in consumer search costs necessarily make markets more or less frictional. Ellison and Wolitzky (2009) model a market in which it is rational for fi rms to obfuscate, that is, to take actions that make it more time- consuming for customers to inspect a product and learn its price. In this setting they fi nd that reductions in consumer search costs are at least partially offset by changes in the equilibrium level of obfuscation undertaken by fi rms. These concerns are not simply a matter of theoretical interest. Ellison and Ellison (2009) examine a group of Internet retailers operating in an environment where a price search engine plays an important role in shopping. They fi nd that easy price search in this environment makes demand very price sensitive and that retailers engage in obfuscation that effectively reduces price sensitivity on some products.

Promoting Efficient Investments in Future Health? Efficient treatment for important and costly chronic diseases often takes the form of investments in future health. Costly efforts to control a diabetic’s blood sugar and hemoglobin A1c levels today, for example, can prevent or delay even more costly complications in the future that result from strokes, blindness, and amputations. Search frictions depress the returns from investments in future health by increasing insurance

324

James B. Rebitzer

turnover and by giving insurers heightened market power. In frictional markets then, mandates and other policies encouraging investments in future health may be efficiency enhancing. Broadly speaking there are two types of investments in future health: relationship-specific and general. To fi x ideas about relationship-specific investments, imagine an insurer that has a disease management program in which nurses establish a relationship with an employer group’s diabetics to help them manage their disease. Suppose further that after a lot of money and hard work, all the diabetics learn to become good at managing their blood sugar and therefore they all postpone heart attacks for ten years. If the employer should switch insurance companies after five years, the original insurer will have lost half of the return from investing in the relationship with the employer and its diabetic employees. In this way, friction-induced insurance turnover discourages relationship-specific investments in future health. The logic of general investments in future health is subtly different from relationship-specific investments. Suppose, for example, that it is the employer that invests in teaching its diabetics to control their blood sugar. One would expect the employer to enjoy returns on this investment through lower insurance premiums. In contrast to the example above, these returns should be general in the sense that every insurer should be willing to offer lower premiums to the employer because of the newly well- controlled diabetics. In frictional markets these returns to general investments are less than they would be in the absence of frictions. The source of the difficulty is not turnover but rather insurer market power. If insurers mark up 30 percent over marginal costs, employers gain only 70 percent of the savings from reducing their employees’ future healthcare costs. As an empirical matter, very little is known about the effects of insurance turnover and market power on investments in future health. Fang and Gavazza (2011) examine the effect of employee turnover on investments in future health, and their results suggest there may be significant inefficiencies. Similarly, Herring (2010) fi nds that markets with higher rates of employee turnover have a significantly negative effect on the utilization of preventive services but no effect on the utilization of acute services. To the extent that the ACA-mandated insurance exchanges reduce friction-induced insurer market power in the individual and small group market, they will increase incentives to investments in future health. It is

Affordable Care Act and Commercial Health Insurance

325

less clear if these exchanges will have a similar effect on reducing insurance turnover. Ericson and Starc (2012) report that tenure on the Massachusetts Connector was quite short, inasmuch as individuals appeared to use the Connector as a way to acquire insurance when they were between jobs.

Conclusions: Euthanasia of Commercial Health Insurance? The fi rst rule of public policy regarding the economy is to fi x only what is broken. In the United States, the self-insured market segment appears to function well for both employers and employees, and so it is a good thing that the ACA leaves the self-insured segment of the commercial health insurance market almost entirely untouched. The search-frictioninduced inefficiencies highlighted in this paper are primarily found in the small-group and individual market segments and it is on these sectors that the ACA will have its greatest effect—for good or for ill. The presence of moderate search frictions in the small-group market (and by extension the individual market) significantly increases insurer market power as well as insurance turnover. From this perspective, various features of the ACA may be efficiency enhancing. This chapter has focused on potential gains from policies that thin out the right tail of the distribution of prices, simplify search, and encourage investments in future health. The study of search frictions in insurance markets is a relatively new endeavor and there are many important gaps in our understanding. Of these, perhaps the most important is the absence of an empirically grounded understanding of the source of search frictions. If frictions result, for example, from the cost of search and the complexity of health insurance contracts, then simplifying and standardizing products may go a long way towards making insurance markets more efficient. If frictions are significantly magnified by adverse selection, then the insurance mandates of the ACA may reduce friction-induced distortions. Other interventions may be called for, however, if frictions result primarily from failures in the market for brokers and other trusted expert advisers. In one of the concluding chapters of his General Theory of Employment, Interest and Money, John Maynard Keynes impishly speculated that the “euthanasia of the rentier” might be imminent because in the 1930s

326

James B. Rebitzer

environment of insufficient aggregate demand and with diminished prospects for profitable private investments, thrift no longer served a positive social function. In a similar spirit one may speculate about the future of commercial health insurance brokers and carriers in the small-group and individual market. If these private entities cannot fi nd a way to create value for themselves and their clients in the new ACA environment that works to strip away the distortions resulting from search frictions, what positive social function do they serve?

Notes Much of the analysis of search frictions and insurance markets described in this chapter was the result of research conducted jointly with Randall Cebul, Lowell Taylor, and Mark Votruba and published in Cebul et. al. (2011). Interested readers are referred to that paper for technical and other details. My coauthors have not participated in the writing of this chapter and I alone bear responsibility for omissions or errors. 1. These low-price policies can themselves be undercut by other low-price policies. The result is a distribution of prices ranging from some point above marginal cost to consumers’ maximum willingness to pay. 2. For an important exception see J. Li (2010). 3. I could fi nd no comprehensive data on the prevalence of independent health insurance brokers, but industry informants suggest that many health insurers rely heavily on independent brokers to sell their policies. Brown and Minor (2012) use data from the Texas Department of Insurance and fi nd that in 2010, 253,604 agents were licensed to sell all types of insurance in the state and roughly 75 percent of them were independent. 4. Of course, the overall effect of the Massachusetts healthcare reform law is to substantially reduce prices for young people relative to what they would have been in the absence of the insurance exchanges and the mandate requiring everyone to purchase insurance (Ericson and Starc 2012).

References Abaluck, J., and J. Gruber. 2011. “Choice Inconsistencies among the Elderly: Evidence from Plan Choice in the Medicare Part D Program.” American Economic Review 101(4): 1180–1210. Brown, J., and D. Minor. 2012. “Misconduct in Credence Goods Markets.” Working Paper presented at Allied Social Science Association Meetings, San Diego, California. January 2013. Cain, D. M., G. Loewenstein, and D.A. Moore. 2011. “When Sunlight Fails to Disinfect: Understanding the Perverse Effects of Disclosing Conflicts of Interest.” Journal of Consumer Research 37(5): 836– 57.

Affordable Care Act and Commercial Health Insurance

327

———. 2005. “The Dirt on Coming Clean: Perverse Effects of Disclosing Confl icts of Interest.” Journal of Legal Studies 34(1): 1– 25. Cain, D. M., and A. S. Detsky 2008. “Everyone’s a Little Bit Biased (Even Physicians).” Journal of the American Medical Association 299(24): 2893– 95. Cebul, Randall, James B. Rebitzer, Lowell Taylor, and Mark Votruba. 2011. “Unhealthy Insurance Markets: Search Frictions and the Cost and Quality of Insurance.” American Economic Review, 101(5): 1842– 71. Dafny, L. S. 2010. “Are Health Insurance Markets Competitive?” American Economic Review 100(4): 1399–1431. Ellison, G., and S. F. Ellison 2009. “Search, Obfuscation, and Price Elasticities on the Internet.” Econometrica 77(2): 427– 52. Ellison, G., and A. Wolitzky 2009. “A Search Cost Model of Obfuscation” NBER Working Papers 15237, National Bureau of Economic Research, Inc. Ericson, K. M. M., and A. Starc. 2012. “Designing and Regulating Health Insurance Exchanges: Lessons from Massachusetts.” Inquiry 49(4): 327– 28. Fang, H., and A. Gavazza. 2011. “Dynamic Inefficiencies in an Employment-Based Health Insurance System: Theory and Evidence.” American Economic Review 101(7): 3047– 77. Gabbaix, X., and D. Laibson. 2006. “Shrouded Attributes, Consumer Myopia, and Information Suppression in Competitive Markets.” Quarterly Journal of Economics 121(2): 505–40. Harrington, S. E. 2012. “Medical Loss Ratio Regulation Under the Affordable Care Act.” Abstract. Accessed October 29, 2014. http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=2180965. Heiss, F., A. Leive, D. McFaden, and J. Winter. 2012. “Plan Selection in Medicare Part D: Evidence from Administrative Data.” NBER Working Paper.18166. National Bureau of Economic Research, Inc. Cambridge, Massachusetts. Herring, B. 2010. “Suboptimal Provision of Preventative Healthcare Due to Expected Enrollee Turnover Among Private Insurers.” Health Economics 19(4): 438– 48. Jackson, H. E. 2008. “The Trilateral Dilemma in Financial Regulation: How to Improve the Effectiveness of Financial Education and Savings Programs.” In Overcoming the Savings Slump, edited by A. M. Lusardi, 82–118. Chicago: University of Chicago Press. Ketcham, J. D., C. Lucarelli, et al. 2012. “Sinking, Swimming, or Learning to Swim in Medicare Part D.” American Economic Review 102(6): 2639– 73. Li, J. 2010. “Job Mobility, Wage Dispersion, and Asymmetric Information.” Kellogg School of Management Working Paper. Evanston, Illinois. Rebitzer J., and M. E. Votruba. 2011. “Organizational Economics and Physician Practices.” NBER Working Paper No. 17535. National Bureau for Economics Research, Inc. Cambridge, Massachusetts.

James B. Rebitzer

328

Woodward, S. E., and R. E. Hall. 2011. “Diagnosing Consumer Confusion and Sub- Optimal Shopping Effort: Theory and Mortgage-Market Evidence.” American Economic Review 102(7): 3249– 76. Woolhandler, S., T. Campbell, and D. U. Himmelstein. 2003. “Costs of Health Care Administration in the United States and Canada.” New England Journal of Medicine 349(8): 768– 75.

Chapter fourteen

A Cautionary Warning on Healthcare Exchanges A Plea for Deregulation Richard A. Epstein

Introduction: Grand Hopes Meet Harsh Realities

R

ight now the United States is in the midst of one of the most ambitious and comprehensive experiments in mass marketing that the world has ever seen. The passage of the Patient Protection and Affordable Care Act has brought with it a determined effort to create a massive set of healthcare exchanges,1 now rebranded on the official government site as “Marketplaces,” 2 whose purpose is to aid individuals who are otherwise unable to do so to acquire healthcare in some form of voluntary market. The basic objective of the Act is to provide a single convenient place in each state where individuals in need of healthcare coverage can acquire it from qualified fi rms that wish to sell it. “Whether you’re uninsured, or just want to explore new options, the Marketplace will give you more choice and control over your health insurance options.”3 It is a highly debatable proposition that individuals will have “more choice and control” over their healthcare options. The term “marketplace” evokes an image of a world in which people can come or go as they see fit, which is surely not the case with the institutional arrangements under the ACA. The term “exchange” is more accurate because it highlights the simple fact that the operator of the exchanges has a huge say in who participates on it. Just that emphasis is found in the Wikipedia

330

Richard A. Epstein

defi nition—important because of its easy public availability—which begins its description of healthcare exchanges as follows: “A health insurance exchange is a set of government-regulated and standardized health care plans in the United States, from which individuals may purchase health insurance eligible for federal subsidies.”4 The government’s new marketplace metaphor downplays both the massive regulations and the subsidies built in at the ground level in the ACA exchanges. Subject to these critical caveats, these healthcare exchanges are designed as places where individuals on both sides of the market can congregate to facilitate trade. In standard economic analysis, the exchange establishes the rules for parties to join an open market in which consumers can move quickly from booth to booth in order to compare the quality and prices of the products in question, in order to match each buyer with the right seller, and the reverse. Its key functions are often to ensure that buyers have sufficient assets to meet any obligations that they incur through their activities on the exchange and that vendors allow their customers to gain clear title to merchantable goods. The ease of head-to-head comparisons reduces information search costs and thus increases the likelihood of efficient pairing in ways that will reduce the risk of monopoly profits, which is magnified whenever search is inefficient. There is little doubt that the basic appeal of the ACA lies in its outward embrace of this image of voluntary exchanges that has been deployed so successfully in countless markets from stocks to diamonds to pork bellies to collectibles. It is important to understand the similarities between these voluntary exchanges and the complicated exchanges fi rst created and then extensively regulated by the ACA. But it is equally important to understand the profound differences between the two types of exchanges—most notably the exhaustive requirements that any fi rm must accept in order to seek customers on the ACA exchanges. In my view, these extensive requirements are so ill- considered that they are likely to negate many, if not all, of the possible benefits that ordinarily derive from organizing voluntary exchanges. The net effect is that the planned healthcare exchanges contemplated by the ACA are likely to break down in many ways. Studies being published as this essay is written confidently predict serious dislocations within the overall exchange system. The most recent report of the Congressional Budget Office estimates that about 7 million people will lose their employer healthcare coverage because of the complex interaction of the tax incentives and penalties created under the new exchanges. 5 The problem is compounded by major delays in set-

A Cautionary Warning on Healthcare Exchanges

331

ting up these exchanges both at the federal and state level. A recent news story reports that the National Association of Insurance Commissioners (whose new CEO is Ben Nelson, formerly a Democratic senator from Nebraska) has urged the Obama administration to push back the October 1, 2013 start date for the operation of the healthcare exchanges so as not “to create chaos.”6 These difficulties are inherent in any effort to regulate the exchanges within a federal system that requires sign- off and cooperation between state and federal officials before they can be made operational. Nelson was hired to explicitly make sure that the lines of communication were open to the White House and that officials at HHS would return his phone calls. The task that Nelson faces is indeed formidable, because some insurance commissioners work in states that have decided to establish their own exchanges while an even larger number work in states that have refused to establish these exchanges—which requires the federal government to assume front-line operational responsibilities with which it is ill equipped to deal. It is therefore no wonder that the Obama administration has extended the deadlines for states to opt into the program on multiple occasions,7 indeed until after the October 2013 date when these are supposed to go into operation.8 At this point, the current tally is that of the fi fty- one entities eligible to participate in the program nineteen have opted in, seven are still on the fence, and twenty-five jurisdictions have “defaulted” to the federal option.9 In November 2012, it was estimated that some 25 million individuals would be covered by the exchanges.10 Thus far federal grants to support the creation of these state exchanges are in excess of $2 billion.11 That figure does not include the expenses that individual states will have to bear in setting up their own exchanges or those of the federal government in setting up its own exchanges. Awareness of the delay has crept through the system, so that HHS pushed back the March 1, 2013 start date by which employers are under a duty to supply employees information about the public health insurance exchange options indefi nitely.12 All the evidence to date indicates that few if any exchanges will go live as of their stated day. But even if they were to surmount these formidable administrative hurdles, there are good reasons to believe that the design conditions imposed on the exchanges will make it very difficult for their participants to realize gains from trade. Just because the exchanges are open does not mean that either consumers or sellers will come, at least in sufficient numbers for the exchanges to work. Most notably, the ACA

332

Richard A. Epstein

does not let the fi rms on the exchanges set the terms and conditions on which they may offer various products. Instead they are required to meet an extensive set of minimum conditions that are likely to prove so onerous that the packages offered will not be priced at a level that individuals can afford, even after taking into account the major government subsidy on the consumer side of the transaction. The overall rigidity of the ACA system rests on a defi nition of “competition” that assumes that people can compete only on dimensions that are predetermined by the state, which in this case are limited to price and quality for the goods stipulated by regulation. But true competition works with innovations that leapfrog current practices by offering, or not offering, bundles of goods that consumers prefer to those that are already available to them. That sense of discovery is wholly blocked by the requirements that are imposed under the ACA. The limitations will make it difficult for these exchanges to thrive, perhaps even survive, going forward. The larger and oft-told lesson is simple: aspirations are easier cherish than to implement. In order to defend this grim prediction, I shall proceed as follows. The initial section offers a critical analysis of the well-regarded paper, “Unhealthy Insurance Markets: Search Frictions and the Cost and Quality of Health Care Insurance.”13 The title of the paper reveals its basic thesis, which is the corrosive effect of friction, or transaction costs, on the operation of the healthcare markets. In the second section of this paper I explain why frictions are not unique to healthcare, and that they are by no means dispositive on the question of regulation with respect to other markets. High transaction costs are a fact of life, but they are neither a necessary nor a sufficient condition for government regulation having a desirable role in the market. The fi nal section of this paper offers an overview of some specific flawed design features of healthcare exchanges found in Title I of the ACA, which account for many of the difficulties that the government has experienced to date in implementing the law. Ironically, fi nally getting the exchanges up and running will only aggravate the dangers that Rebitzer highlights: the heavy frictions that the ACA and regulation impose on the operation of the healthcare market. The supposed cure is worse than the underlying disease. To be sure, the Rebitzer paper was not written as a defense of the ACA. But the paper’s framework shows why the high hopes for this healthcare experiment are likely to be dashed. In principle, someone might be able to devise regulations that control the inevitable frictions in the operation of voluntary markets. But

A Cautionary Warning on Healthcare Exchanges

333

the odds of doing that are long indeed, for the familiar obstacles to comprehensive government regulation remain. Public administrators have at best imperfect knowledge of the fi rms and individuals they regulate. But even with perfect information, powerful political forces and regulatory zeal are almost certain to undermine the cause of reform. In general, the best reforms are modest ones that do as little as possible to upset the delicate balance of voluntary markets.

The Asserted Difficulties of Healthcare Insurance Markets Frictions in Healthcare Markets The potential inefficiencies in the healthcare market arise from the use of multidimensional healthcare contracts with many moving parts. They arise both in individual healthcare insurance contracts and in group insurance policies, where the composition of the voluntary group shapes the coverages, exclusions, deductibles, and rates found in any private plan. Within this environment, the alleged inefficiency in healthcare markets stems from the high search costs needed to match buyers with sellers. It is a commonplace since Ronald Coase’s pioneering work that positive transaction costs are always an obstacle to efficient contracting.14 Indeed it counts as a general theorem that the best way to maximize social welfare is to minimize transaction costs that stand in contractors’ way. That low- cost approach includes simple requirements: land contracts are not enforceable unless in writing, and all claims on real property must be recorded on some computerized public registry with robust search functionality that allows all private persons to fi nd out the state of the title. These simple formalities may look like restrictions on freedom of contract, because they reduce or eliminate the benefits that people derive from agreements that do not observe them. But that ostensible detriment is largely illusory. The additional costs imposed ex ante are more than offset by the great savings in dispute resolution ex post, while preserving contractual freedom on key issues of price and terms of service. The availability of these simple devices should put us on guard against using major government interventions as the initial line of regulatory strategy. All major markets have high transaction costs of one kind or another. If public regulation is needed in the healthcare market, is it also needed in real estate markets for leases, sales, and mortgages? This

334

Richard A. Epstein

generalized argument presents an open invitation for the comprehensive regulation of all markets outside the healthcare area. Consumer protection is an iffy rationale for government regulation, even without accounting for the risk of failure. The Consumer Financial Protection Bureau has proposed massive restrictions on the operation of mortgage brokers that could easily drive large numbers of them from the business.15 The Securities and Exchange Commission uses investor protection as its rationale for tightly controlling disclosures of new stock issues to a world fi lled with naïve consumers. But this entire system could be dispensed with by instructing ordinary consumers that they are well advised to pay a small commission to one of a thousand fi nancial investment houses or advisers to manage their money for them. The decision to hire an agent, of course, substitutes an agency- cost risk for a personal incompetence risk. These agents have a level of expertise from repeat play that ordinary consumers cannot hope to match. They can represent multiple consumers who are similarly situated at far lower cost than the consumers can represent themselves. And they can develop a reputational bond that limits bad behavior without the necessity of a costly and lengthy lawsuit. In light of these systematic and pervasive advantages, the frequency of fi nancial intermediation powerfully suggests that the difficulties of consumers in transacting for themselves present the far greater risk; consumers who seek help gain far more in professional advice than they lose from a combination of service fees and the dangers of abuse. Intermediates, whom a given person can pick and monitor, in general offer a better cure for shortfalls in information than do intrusive forms of government regulation imposed from afar, often by government officials who have little local knowledge of the markets they are regulating. Information Shortfalls—Which Way? Rebitzer’s account of the information shortfalls in these insurance markets cuts against the grain of the standard insurance model, which looks at asymmetrical information in the exact opposite way. The Rebitzer model assumes that informational advantages in the health insurance markets lie with the insurer and not the insured. By that account, the only real point of contention is whether these companies gain only a noncontroversial competitive profit or whether they can extract some monopoly surplus from their unsuspecting commercial customers.

A Cautionary Warning on Healthcare Exchanges

335

Rebitzer’s proposition goes very much against the grain of conventional insurance market practice, which holds that the differential advantage is possessed by the insured and not by the insurer. For example, in the history of marine insurance, in which the major perils are associated with ships lost at sea, all the duties of disclosure were imposed upon the insured for the benefit of the insurer—both as to whether or not to take the risk and, if so, how to set the premium. That basic rule was set out in London Assurance v. Mansel, which held that disclosures must be made of all information known to an insured even if the insured does not appreciate its relevance to the risk. “Whether it is a life, or fi re, or marine assurance, I take it good faith is required in all cases.”16 That rule was in turn incorporated into statutory law.17 There is no reason why this principle of good-faith disclosure by an insured should not apply to markets in health or disability insurance. To be sure, modern healthcare markets are more complex, and healthcare insurers are not passive vehicles ready to take on uncompensated risk. But by the same token, the Rebitzer argument is not, at least at this juncture, addressed to ordinary individuals who may be (but often are not) innocent in the ways of the world. The insureds here are sophisticated commercial parties that have access to the third-party expertise needed to negotiate with the insurance companies, as well as the informational advantage that comes from having employee applications and medical records on hand. In the face of these institutional realities, it is hard to credit the Rebitzer position when the opposite conclusion is more likely true. If there is any informational asymmetry in these markets, it may well rest with employees who get or retain jobs precisely because they are able to conceal information from both their employer and the third-party insurer.18 Indeed it is a feature of the modern disability discrimination rules that the employer cannot ask for information that would have been allowed as a matter of course in all standard markets under the goodfaith principle. The point of that regulation is, of course, not to foster the efficient operation of the health insurance market but to induce a set of cross-subsidies so that disabled individuals can get insurance at a lower pooled rate rather than their higher individual rate.19 That cross-subsidy risk is one reason these markets for individual health insurance can break down when employers or healthy individuals fi nd ways to detect high- cost individuals with concealed disabilities, as they often do. These institutional risks also undermine the insurance market, but in ways that are orthogonal to the problems of group insurance raised by Rebitzer.

336

Richard A. Epstein

Once we take all these other complications into account, many private employers will end up providing insurance to certain individuals at a loss. Since these fi rms know their actuarial position, it seems unlikely that any insurance companies will be able to extract profits from their unsuspecting insureds. At this point both sides may know something the other does not, which makes it all the more difficult to estimate the direction of any deviation from the ideal actuarial rates; it could run in either direction. Why then assume any systematic advantage to the insurance company? In practice, the composition of employees in any particular fi rm will often prove critical to the kinds of coverage it is prepared to buy in a voluntary market for workers who differ by age, by sex, by education, by hobbies and outside activities, by their willingness to smoke or drink alcohol. The use of fi nancial penalties has expanded rapidly in the past few years and is likely to do so in the future, wholly without reference to anything in the ACA. It often turns out to be more efficient to use a mixture of sticks and carrots than to rely solely on the carrots. 20 This recent trend confi rms the sensible proposition that as the costs of healthcare increase, both insurers and employers will turn away from standardized prices when they may well face nonstandardized risks from their individual insureds. Looking at these negotiations, they seem to satisfy the usual condition that the outcome works for the parties’ mutual advantage. It is very hard, therefore, to think that there is some built-in informational bias that operates in favor of the insurer. Turnover Rates Rebitzer also argues that the high turnover rate in portions of the healthcare insurance market offers some evidence of latent market imperfection. But the significance of overall turnover rates is questionable. A high turnover rate is a function of decisions by insured companies to look elsewhere or by insurers not to continue policies. One inference from this is that the insured has acquired information that allows it to get a better deal elsewhere in the marketplace. But the switch could just as likely take place because the insurer wants to raise rates for what it fi nds to be a more expensive pool. Either way, it is likely that the relevant information can be acquired with relative ease on both sides of the market. Turnover, therefore, does not in and of itself suggest any exploitation on either side. Nor, for that matter, does the stability of some long-term relationships, which may well survive precisely because the two sides were

A Cautionary Warning on Healthcare Exchanges

337

able to adjust coverage or rates that preserved their mutual gains from the transaction. Nor do I attach the same weight as Rebitzer to the fact that more midsize fi rms go into the voluntary market for third-party insurance while larger fi rms prefer to self-insure—by hiring insurance companies chiefly in a management rather than a risk-bearing capacity. One benign explanation for this result is that the loss exposure over time is likely to prove more stable in a larger company with many employees in diverse locations than it is for a smaller fi rm with fewer employees, often concentrated in fewer locations, and thus subject to some common-mode risk. The higher level of variation in potential exposure for smaller fi rms may make them prize insurance more than larger fi rms, so that they go to the market in larger numbers. A second benign reason derives from the cost curves associated with setting up an internal management plan. The fi xed costs are likely to be relatively insensitive to the size of the fi rm. The larger fi rm can therefore spread these costs out over more employees at a lower per-unit cost. The smaller fi rm may fi nd it easier to go to a third-party provider in order to share those front- end costs with others. At this point, both types of fi rm are on average making the right decisions, and will continue to do so if the choice between different strategies of risk are left unregulated. The corrective force of new entry should go a long way to dissipate any supposed monopoly profits on insurance, unless, as seems most unlikely, incumbent insurers can keep their favorable returns hidden from the world. It is unlikely that midsize insureds are perpetually duped by the their insurers. The existence of some but not complete turnover is not any evidence of market failure. Indeed, if there is anything that is likely to roil these markets, it is regulation imposed through the disability law that makes it hard for either insurers or employers to get information about employees. It is not just regulations in the ACA that impact insurance markets. It is the full range of controls. The Perils of Public Marketing of Insurance The expertise required to do health insurance work makes it unlikely that any public agency has the ability to operate successfully in this market niche. In an ordinary private setting a fi rm has to estimate its costs of production for various quantities of its services, and will only enter the market if it thinks that its revenues are sufficient to cover its costs at various levels of production. Clearly this is a chancy process at best, given the

338

Richard A. Epstein

difficulties of estimation. But owing to the enormity of the stakes it is a task that is well worth undertaking, for even if most new businesses fail, the ones that do succeed often produce huge social gains that are more than sufficient to cover the losses. But the same cannot be said about government programs, which often operate free of the same scarcity constraints. Programs seem destined to succeed by government fiat, even when they are unable to generate the revenues needed to cover costs. Indeed, often their stated rationale is to make sure that the individual program participants do not have to cover their full costs, which necessarily requires the government to set aside funds, raised either by taxation or borrowing, to cover those losses. At this point, the government fi rm has many more options, especially in the American context where we do not normally commit ourselves to honoring the entitlement fi rst, only worrying about the funding afterward. So long as the government program has access to public revenues, it can continue in operation even though the subsidies required for it to do so far exceed original estimates, which is surely the case with both Medicare and Medicaid. In addition, these government programs will face exactly the same difficulties as the private programs once they discover that it is not easy to integrate any new healthcare program with those that are already in existence. The major delays in getting the healthcare exchanges off the ground are just one additional piece of evidence that the government as entrepreneur has no advantage over the fi rms with which it is in competition. All too often it gains illicit advantages through constant access to cheap public funds that allow it to undersell efficient private competitors. This kind of distortion has taken place with the provision of electrical power, where for example the Tennessee Valley Authority receives certain advantages over private competitors, in the form of cash transfers or exemption from various regulations. Too Many Products I also think that it is a mistake for Rebitzer to attack the private market for the proliferation of alternatives that it offers to consumers. It is of course the case that no individual, and indeed few if any fi rms can intelligently evaluate the terms and conditions of several hundred plans at reasonable cost. But by the same token it is equally clear that no individual or fi rm can evaluate the thousand types of vacation homes, breakfast ce-

A Cautionary Warning on Healthcare Exchanges

339

reals, or telephones offered for sale in a competitive market. Fortunately, that observation is largely beside the point, because no savvy consumer ever tries to cover the entire market before making selections. The common strategy in all these variegated settings is to break the market down into segments and to search only those that promise the greatest return. So one person may have no interest in looking at sugared cereals, and thus lops off a huge portion of the available options with a single cut. Make a few other global judgments and the relevant choices will be down to a manageable dozen or so. All of that winnowing does not mean, however, that the rest of the options are mere waste on the marketplace. Quite the contrary. Each of us relies on different search parameters, so that while we all end up with a few serious options, the options differ dramatically from person to person and group to group. Firms of course understand that an “original” anything cannot blanket the market with a single product. But they also know that the “original” something is likely to be the most popular brand in a new market niche, so it pays to launch that product fi rst. If it fails, then the experiment comes to an end. But if it succeeds then new varieties are introduced to pick up the smaller but still substantial components of the market that do not choose to buy the original. Some people will only buy Honey Nut Cheerios and not the simpler variety. This process of product differentiation goes on until the cost of coming up with a new variation is low relative to the anticipated gain. Yet it is precisely because innovative fi rms also have sound incentives about when to quit that the system works well in market segment after market segment. If this analysis is correct, then it is a mistake to think that the proliferation of consumer healthcare options poses an impediment to successful market operation. Any decision such as that made in healthcare markets—to require given fi rms to offer only a particular type of contract with predetermined coverage—does not facilitate competition but thwarts it by restricting the dimensions over which innovative fi rms can compete. To be sure, it is unlikely that either midsize fi rms or ordinary consumers can canvass the entire market. But they can make a series of initial cuts to focus on the market segment they care about most. At this point, one of the key drivers of good competition is the ability to offer a particular configuration of goods and services that make sense to some segment of the overall market. The standardization of service packages thus prevents innovation along certain key dimensions, which hardly improves the overall competitive market. Put otherwise, product

340

Richard A. Epstein

differentiation is the great and beneficent spoiler because it allows rapid and discontinuous changes in the market, such as the rise and fall of the BlackBerry and the now possible decline of Apple in the face of potentially disruptive technological developments from a host of competitors. In my view, these large gains dominate any negative effects. Indeed the constant use of product differentiation, both large and small, in market after market, suggests that healthcare regulators engage in a dangerous gambit by limiting product choice to a few set choices in order to reduce the buyer’s costs of search. People can truncate searches using sensible strategies. They do not have similar ways to expand market options. Expanding Market Alternatives My fi nal objection to the Rebitzer argument is that it moves matters in the wrong direction. It takes no learned citations to remind us that healthcare markets were heavily regulated at the state and national level long before the arrival of the ACA. In earlier work, soon to be published, David Hyman and I have argued that the current drive to increase regulation takes matters in exactly the wrong direction. 21 The best thing that we could do to open up healthcare is to remove many of the mandates and restrictions that have been imposed at either the state or federal level. The list of those mandates and restrictions is long. But it is useful to mention just a few high points of what we think the source of the difficulty is. First, one clear difficulty with healthcare markets is that they are now burdened with heavy restrictions on entry that prop up the monopoly institutions that Rebitzer addresses. In this regard, the three most conspicuous devices are the state licensure requirements that hamper the movement of physicians from state to state, thereby restricting entry in ways that create artificial shortages that could lead to supracompetitive prices. Next, most states have restrictions on the ability of out- of-state insurance carriers to sell policies in all states, which again reduces competition. Third, there are the constant restrictions on the corporate practice of medicine that make it hard for large operations like Wal-Mart and CVS to enter these markets en masse, by offering low-priced services with short wait times that could induce many uninsured people to avoid emergency rooms when faced with the need for routine care. It is critical to note that a reform of these three obstacles displays the following twin virtues: administrative costs go down while access to care increases.

A Cautionary Warning on Healthcare Exchanges

341

Lower prices and improved quality become possible with expanded market participation. The ACA makes no sensible moves in this direction. Second, it is commonplace today for states to impose extensive mandates on employers who wish to supply insurance coverage to their employees. The logic of these mandates is simple enough. No employer is required to offer health insurance to its employees, but if it does it must maintain certain minimum standards. It is possible to condemn this tying arrangement on per se grounds, wholly apart from the desirability of particular mandates. Clearly, employee insurance that supplies no coverage is worth nothing, so that employers have a strong incentive to supply those benefits that cost less than their worth to employees on average, so that both sides gain. There is no reason to mandate coverages that are supplied anyhow. But it is all too easy for these mandates to impose coverage that costs more to employers than it is worth to employees. That form of mandate operates as a tax on the relationship between the two parties. In some instances, it will lower the joint surplus between the parties (itself a bad) but not lead to a decision to jettison employer coverage, so long as the overall value of the policy exceeds its total costs. But that game cannot go on forever, and given the literally dozens if not hundreds of mandates, in some instances a given mandate will be the straw that breaks the camel’s back and lead to a discontinuation of coverage that is too expensive for either side to want. At that point, the costs of the mandate program become explicit. And these costs are large. The percentage of employees with employer insurance has dropped from about 60 percent in 1980 to about 50 percent today, which translates to a loss of employer coverage for about 15 million workers. Scarcity presents this stark choice. The state can have ubiquitous coverage for all persons or comprehensive coverage for some persons. It cannot afford to pay for both. Yet the ACA with its strict minimum essential benefits packages continues what is likely to prove a ruinous policy. Third, it should be possible to relax the extensive and cumbersome privacy regulations that are introduced under the Health Insurance Portability and Accountability Act. 22 The statute imposes an extensive preclearance system for the use of key medical data in every transaction in the United States. Yet prior to the passage of the statute, cases of documented abuse were few, and those that did occur exposed the wrongdoers to suits for invasion of privacy, breach of fiduciary duty, medical malpractice, intentional infl iction of emotional distress, and defamation, as well as a large number of regulatory sanctions. There was no need to

342

Richard A. Epstein

invoke a hugely expensive ex ante system of regulation that needlessly covers billions of transactions in the absence of any proof that the older set of judicial, regulatory, and social sanctions had failed. The compliance costs of this statute are in the billions. 23 The loss of operational flexibility is greater. Repeal would yield lower administrative costs and lower prices for services, which would again increase both the quality of and access to healthcare. Fourth, medical malpractice liability is by no means the dominant driver of the current healthcare crisis, but there is little doubt that the inability of private institutions to contract out of the standard-issue terms for medical malpractice does create some dislocation in these markets, which increases costs and reduces access to care, often from rural clinics that face high liability costs. Allowing parties to contract away from these rules will not in my view lead to absurd contracts that would face serious consumer resistance, but would allow, if the first point is accepted, greater ease of entry to these markets by lowering their anticipated liability costs. The trade-off that is involved is this. On the positive side, tort liability may supply compensation to parties who have been grievously wronged by bad medical treatment. But at the same time, the potential price barrier will delay, diminish, or deny access to other individuals who are far better off taking the risk of the low probability of negligence than going without medical care at all. Yet those cases of people who suffer because they cannot gain access to the system are not made visible in a liability system that looks at only one kind of failure and not the other. Reform in this regard could lower costs, increase access, and increase the average level of care for those people who now lie outside the system. There are, in effect, ways to deal with the many imperfections of healthcare markets that rely first on deregulation. The Rebitzer approach, to my mind, overstates the gains from its form of regulation but overlooks its defects. The next section addresses the systematic risks of excessive regulation that should weigh heavily in the overall social calculus of costs and benefits.

Defects of the ACA: A Brief Catalogue Legal reform often suffers from this difficulty: it is easy to identify imperfections in some existing market, but far more difficult to cure any one imperfection without creating a second. Even—perhaps espe-

A Cautionary Warning on Healthcare Exchanges

343

cially—in healthcare, the cure can be worse than the disease. In this section, I confine myself to three major ways in which Title I of the ACA, which regulates the existing insurance market and sets up the healthcare exchanges, deviates from standard insurance principles. These involve the definition of essential minimum benefits; the various rules dealing with eligibility for coverage on the healthcare exchanges; and the computation of the medical loss ratio used to limit the fraction of plan revenues that can be spent on administrative expenses. Essential Minimum Benefi ts The centerpiece of the ACA is that, as the HHS website proclaims, “the law ensures that health plans offered in the individual and small group markets, both inside and outside of Affordable Insurance Exchanges (Exchanges), offer a core package of items and services, known as ‘essential health benefits.’ ”24 Excluded from this program are persons on Medicare and Medicaid as well as individual members of large employer healthcare groups, at least for the present. The ACA identifies ten key coverage areas: ambulatory patient services, emergency services, hospitalization, maternity and newborn care, mental health and substance abuse disorder services including behavioral health treatment, prescription drugs, rehabilitative and habilitative services and devices, laboratory services, preventive and wellness services and chronic disease management, and pediatric services including oral and vision care. The statute does not specify the procedures and treatments covered under the law. Nor does it require HHS to start from scratch in all cases in setting out the appropriate parameters, but builds off current practice, as defi ned by “typical employer plans,” a phraseology that deliberately understates the level of heterogeneity found in the private healthcare market. Accordingly, HHS picks multiple “benchmarks” that mirror the business choices of the dominant fi rms within any market, without controlling for differences in plan membership in smaller fi rms located within any given state. Defi ned in this fashion, the benchmarks could squelch desirable market heterogeneity by forcing highly expensive plans on individuals who previously opted for cheaper plans with less expensive coverage. This simplification of coverage is in line with Rebitzer’s recommendation. But the metallic plans (bronze, silver, platinum) all contain generous fi xed- design features that are so demanding that no existing plan supplies them. At this juncture HHS must invent various solutions

344

Richard A. Epstein

out of whole cloth. Notwithstanding this lack of any market information, HHS is alert to pounce on any allegedly discriminatory practice against any actual or potential recipient. Charges of discrimination are easy to make but difficult to refute, thereby raising the possibility of ratcheting up overall coverage levels. For this system to work, the government must pay hefty subsidies for eligible persons, which are sufficient to meet the costs of the coverage when added to the relatively small individual contributions. But there is no reason to be confident that the subsidies will be sufficient to cover the extra costs, raising the risk of serious market breakdown, especially if employers strategically choose to dump all or some of their employees into the public plan even before it is up and running. Just that risk was highlighted when Chicago mayor Rahm Emanuel announced his intention to transfer many retired city workers into the federal plan, 25 thereby hoping to convert the city’s liabilities into national ones. If Chicago is imitated by other cash-strapped local governments, the financial stress on the ACA could prove intolerable. One deep intellectual mistake drives the ACA’s move toward uniform high benefit standards. Some overeager regulators assume that if any one insurer or employer in any particular market offers a particular coverage, it becomes efficient for all insurers whose insureds have different population demographics to do so as well. The same forces that drove the creation of excess state mandates now operate with ever greater power at the federal level. But the ACA regulates groups of individuals with enormous variation in risk profi les. A modern tech fi rm with young workers may be able to offer insurance against alcohol or drug abuse or psychiatric disorders at low cost. But a larger fi rm with less educated and more risk-prone workers could fi nd the risks prohibitive. In such settings, the required reduction in wages needed to fund the insurance obligation could prove too large for a given group of workers. What the ACA does is to force a trade that is ex ante in the interest of neither party to that insurance contract. It is not my job to explain how key coverages should be tailored by individual fi rms. Quite the opposite. The basic point here is that it is easier to know what regulators do not know than to pretend I have the right answers for all private plans, as HHS is forced to do. These informational complexities augur ill for the ACA, because decisions must be made in a knowledge void on such mundane but vital issues as defi ning the set of compensable events, resolving disputes of proof, adverse selection, and moral hazard. At this point RomneyCare in

A Cautionary Warning on Healthcare Exchanges

345

Massachusetts, with its own major cost overruns, 26 is the dead canary at the bottom of the mine, having driven the cost of healthcare in the state from 25 percent of state expenditures in 2006 to 41 percent for 2013. Yet even this comparison is misleading because it is far easier to run a comprehensive health system at the state level, which is rid of the complex layers of state/federal cooperation that have already proved a fertile source of delays in federal plan implementation. Insurance Risks The second element of note in the federal plan is how it consciously ignores all the standard risks associated with private insurance. 27 It is well known that all insurance plans must take steps to guard against some combination of moral hazard and adverse selection. The former refers to the risk that the mere fact of insurance coverage will increase the likelihood of occurrence of the compensable event. The latter refers to the risk that potential insureds will use their private information to time their purchase of insurance to maximize their net gains from insurance coverage. They buy when their known medical risks are high and ditch coverage when they are not. Virtually all private insurance plans include explicit safeguards against both forms of opportunistic behavior. They exclude coverage for certain activities that are likely the result of willful neglect. They impose waiting periods before covering any preexisting conditions. They avoid community rating systems that impose uniform rates on individuals of different ages. They shy away from covering conditions like mental illness that are hard to verify. They require minimum waiting periods before coverage goes into effect, and they require policies to have a certain minimum duration to prevent individuals from signing up when medical expenses are hig, and deserting the plan once their condition has been treated. One major purpose of private insurance conditions is to avoid cross-subsidies, so that people know that they will be more likely to get coverage that exceeds the costs of premiums in value, even allowing for administrative expense. The ACA specializes in the creation of just these cross-subsidies, which requires coercive measures to offset the adverse selection risks against which private plans erect institutional safeguards. The ACA, however, regards insurance coverage as an individual right that necessarily eliminates all underwriting discretion. Accordingly, the rules require insurance companies to take customers on a fi rst- come,

346

Richard A. Epstein

fi rst-served basis. 28 That job is not as easy as it seems, because it requires some regulatory determination of the number of customers that a fi rm can sensibly cover. Once the number of applications exceeds the number of supposed places, the ACA requires insurers to follow certain randomization protocols in selecting the lucky winners, a process that does nothing to stem the risk of adverse selection. To counter the risk of getting a bad draw, the ACA introduces complex procedures to even out burdens among insurers after the fact. But these devices are slow and balky and require a complex pooling system among insurers without curtailing any adverse selection risk. At best they only spread it around to other fi rms, which may also be weakened by the same statutory procedures. The assured availability of insurance should, at the margin, lead to a systemwide reduction in the level of private precautions for health, which could in turn lead to sicker populations in need of larger quantities of healthcare. All of these dangers are, in my view, far more pressing than any concern with the asserted failures of midsize fi rms to accurately price their insurance coverages. Medical Loss Ratios The third major pressure point of the ACA is its extensive use of the medical loss ratio, which essentially limits the percentage of total premium that can be expended on an undefi ned class of administrative expenses. 29 The implicit assumption behind this prohibition is that direct regulation is the appropriate way to bleed out costly administrative expenses that supply no direct benefits to the companies’ insureds. But the decision to impose a 15 percent administrative cap on small group plans and 20 percent on individual plans ignores several key points of insurance management. First, there is no reason to think that any fi rm in a competitive market will want to waste money on unnecessary administrative costs. Rather, the rational coverage program seeks to make sure that the last dollar that it spends on administrative expense produces as much benefit as the last dollar spent on direct patient care. Collectively, we know for certain that the optimal level of administrative expenses is not zero. But beyond that it is never clear where the equimarginal point is for any particular plan, let alone for all plans. The list of risk considerations set out in the previous section offers some insight into the range of management issues that require direct expenditure, and it is nothing more than

A Cautionary Warning on Healthcare Exchanges

347

a wide- eyed guess whether the 15 and 20 percent figures represent an intelligent estimate of the correct number. What is wholly lacking from the ACA procedures is any sense as to how those particular limits are chosen or what expenditures—wellness mailings to customers, calls for follow-up visits, and so on—are subject to the statutory cap. It is doubly ironic that the compliance costs generated by the ACA will count as administrative costs subject to the cap. Yet there is little flexibility in these numbers, save for the possibility that states can unilaterally lower these figures, which only adds uncertainty to the situation. What makes matters worse is that these rules are applied on a onesize-fits-all basis, even though administrative costs will vary widely as a function of the benefit levels under the plan and of the composition of the covered population. In general, we should expect that plans that offer lower benefits to relatively transient low-income employee pools will have the highest medical loss ratios. The smaller the benefits, the smaller the allowable administrative costs. The higher the turnover rate, the greater the administrative costs. Faced with the choice of being out of compliance or dropping the coverage, many fi rms would choose the latter, which would only increase the level of pressure on the ACA healthcare exchanges and Medicaid programs, both of which are already under severe fiscal pressure. As Charles Blahous explains: “CBO is instructed to disregard this restriction on Social Security and Medicare fi nancing whenever it scores legislation. Instead, CBO is directed to assume that Social Security and Medicare will make full scheduled benefit payments in perpetuity, even though the programs lack the fi nancial resources and statutory authority to do so.”30 There is no principled release from these fi nancial pressures, which have resulted in what I have elsewhere called “government by waiver.”31 There is no official repeal of the program, but HHS issues individual short-term waivers to firms certifying that they need not comply with the medical loss ratio to stay in business. The program serves all local interests in ways that can only undermine the rule of law. The public administrators gain an additional lever that they can use in some cases but not in others, thereby shifting the parity between various plans. The fi rms that receive the waivers can survive for another year, when they may reapply. Other plans can aspire to the same preferred status. Selective application of the waiver principle can be made in favor of those institutions—unions—that support the administration, and in certain states that lean Democratic.32

348

Richard A. Epstein

The entire makeshift program can continue to go forward, since no outsider has either the incentive—or under current law, the standing—to challenge the preferential treatment. The simple truth in all these cases is that delay in implementation offers both fiscal relief and political cover for plans that are not fi nancially sustainable in the long run. The only question is how long this particular game can be played when all evidence points to the fact that the underlying program is unsustainable, for once the benefit structure goes into effect, the political costs will make it difficult to undo the entitlements thereby created.

Conclusion: What Direction? There is one sobering conclusion that follows from this review of the pros and cons of regulation in the healthcare market. The Rebitzer paper shows in a sense what all of us know to be true. No market institution works without error and scars. The more difficult it is to understand the good or service to be provided, the greater the error rate, and the higher the associated administrative costs. But the relevant issue is not whether we can identify some market imperfection, for we always can. The harder question is what direction the law should move once imperfections are found. There is much evidence that indicates some increase in premature deaths—by one estimate up to 18,000 people—divided among such conditions as breast cancer, high blood pressure, and HIV/AIDS. 33 But these numbers have to be put in perspective, given that this total is less than 0.2 percent of all cancer and heart disease deaths alone, and is 15 percent of the number of accidental deaths per year according to the Centers for Disease Control. 34 Nor do those static numbers take into account the ability of overall prosperity to reduce the number of deaths by keeping individuals out of harm’s way prior to the time of a particular illness. Higher incomes mean better foods, safer cars, and less stress, which surely feed back into the system in major, if undocumented ways. These factors point to the central moral in this area. Once we remember that the healthcare markets have been highly regulated prior to the introduction of the ACA, the key question is whether it is better to move toward more regulation or less to respond to asserted claims of inefficiency. The ACA took the position that heavy intervention in healthcare and insurance markets is a necessary prerequisite to untangling the mess. David Hyman and I have urged that this nation move in the op-

A Cautionary Warning on Healthcare Exchanges

349

posite direction. Start with deregulation. What we need are more market forces at work, not fewer. What has happened is that political forces have worked heavily to undermine the resilience and durability of the private insurance markets. It is a bad bargain that we should hasten to undo.

Notes My thanks to Benjamin Margo and Joshua Stanton, NYU Law School Class of 2014, for their usual excellent research assistance on this paper. Editors’ note: This paper was originally given at a conference at the University of Chicago in October 2012. Consequently, some of the references to impending federal actions under PPACA have already transpired. We believe Professor Epstein’s arguments and predictions are still relevant, both as a demonstration of the best deregulatory arguments of the period prior to PPACA implementation and as a response to James Rebitzer’s article in this volume. We present Professor Epstein’s paper in its original form without changes. 1. Patient Protection and Affordable Care Act, Pub. L. No. 111-148, § 1341 (2010), 124 Stat. 119, 208–11 (codified as amended at 42 U.S.C.A § 18061 (West 2012)) [hereinafter ACA or the Act]. 2. See the official government site, Prepare for the Health Insurance Marketplace, Healthcare.gov, http://www.healthcare.gov/marketplace/index.html. 3. Id. 4. Health Care Exchanges, Wikipedia (Feb. 14 2013, 10:25 p.m.) http:// en.wikipedia.org/wiki/Health_insurance_exchange. 5. Stephen Dinan, Seven million will lose insurance under Obama health law, Wash. Times, (Feb. 7, 2013), http://www.washingtontimes.com/blog/inside -politics/2013/feb/5/obama-health-law-will-cost-7-million/. 6. Alex Nussbaum & Zachary Tracer, Obama Health Law Needs Delay, State Insurance Head Says, Bloomberg (Jan. 22, 2013) http://www.bloomberg.com /news/2013-01-22/obama-health-law-needs-delay-state-insurance-head-says.html. 7. Robert Pear, U.S. Extends a Deadline for States on Coverage, N.Y. Times (Nov.  9, 2012) http://www.nytimes.com/2012/11/10/us/us-extends-deadline-on -health-coverage-for-states.html. 8. Robert Pear, States Will Be Given Extra Time to Set Up Health Insurance Exchanges, N.Y. Times (Jan. 14, 2013) http://www.nytimes.com/2013/01/15/us /states-will-be-given-extra-time-to-set-up-health-insurance-exchanges.html (reporting that Secretary of Health and Human Services Kathleen Sebelius stated “she would waive or extend the deadline [beyond October] for any states that expressed interest in creating their own exchanges or regulating insurance sold through a federal exchange”).

350

Richard A. Epstein

9. For exhaustive data on the progress of the exchanges, see Kaiser Family Foundation, StateHealthFacts.org, http://statehealthfacts.kff.org/healthre formsource.jsp. One state still on the fence is Florida, which led the challenge against the Medicaid extension found in Title II of the ACA. Jim Saunders, Lawmakers Consider to Put in Place Health Insurance Exchanges Under Obamacare, Miami Herald (Jan. 1, 2013), http://www.miamiherald.com/2013/01/23/3195446 /lawmakers-consider-to-put-in-place.html. 10. See Robert Pear, “Health Insurers Will Be Charged to Use the New Exchanges.” The New York Times. (November 9, 2012) http://www.nytimes.com / 2012/12/01/health/health-insurers-will-be-charged-to-use-new-exchanges .html. 11. See Kaiser Family Foundation, supra note 9, (Facts at a Glance). 12. See, e.g., Jerry Geisel, Health Insurance Exchange Reporting Delayed, Crain’s N.Y. Business (Jan. 25, 2013), http://www.crainsnewyork.com/article /20130125/HEALTH_CARE/130129924. 13. Randall D. Cebul, James B. Rebitzer, Lowell J. Taylor, & Mark E. Votruba, Unhealthy Insurance Markets: Search Frictions and the Cost and Quality of Health Insurance, 101 American Econ. Rev. 1842 (2011). This article was presented at the University of Chicago Conference on Health Care Reform on October 12, 2012 by one of its authors, James Rebitzer. I shall therefore refer to Rebitzer in speaking of the collective effort of these authors. 14. Ronald H. Coase, The Problem of Social Cost, 3 J. L. & Econ. 1 (1960). 15. See Anna Maria Andriotis, Mortgage Brokers Face New Hurdles, Wall St. J., Jan. 26– 27, 2013, at B8. 16. 11 Ch. D. 363, 367 (1879); see also Lindenau v. Desborough, 8 Born. & Cress. 586, 108 Eng. Rep. 1160 (1828). 17. Marine Insurance Act, 1906, § 18(1). 18. See 42 U.S.C. § 12112(d)(4)(A) (2010) (requiring that fi rms “shall not conduct a medical examination or make inquiries of a job applicant as to whether such applicant is an individual with a disability or as to the nature or severity of such disability”); id. § 12102 (defi ning “disability” under the statute broadly to cover most major medical conditions). 19. For the same difficulty on a related issue, see Richard A. Epstein, The Legal Regulation of Genetic Discrimination: Old Responses to New Technology, 74 Bost. U. L. Rev. 1 (1994). 20. See Reed Abelson, “The Smoker’s Surcharge” The New York Times. November 16, 2011. http://www.nytimes.com/2011/11/17/health/policy/smokers -penalized-with-health-insurance-premiums.html?pagewanted=all&_r=0. 21. Richard A. Epstein & David A. Hyman, Fixing Obamacare: The Virtues of Choice, Competition and Deregulation, 68 Ann. Surv. Am. Law 493 (2013), available at http://ssrn.com/abstract=1158547.

A Cautionary Warning on Healthcare Exchanges

351

22. See Richard A. Epstein, HIPAA on Privacy: Its Unintended and Intended Consequences, 22 Cato J. (2002), abstract available at http://ssrn.com /abstract=305326. 23. Arthur R. Williams, David C. Herman, James P Moriarty, Timothy J Beebe, Sandra K Bruggeman, Eric W Klavetter, Paul H Steger, & Janet K Bartz. HIPAA Costs and Patient Perceptions of Privacy Safeguards at Mayo Clinic, 34 Joint Comm’n J. on Quality & Patient Safety 27 (2008) (fi nding HIPAA startup costs alone to be greater than $4.6 billion). 24. Essential Health Benefi ts, Actuarial Value, and Accreditation Standards: Ensuring Meaningful, Affordable Coverage, HealthCare.gov, http://www.health care.gov/news/factsheets/2012/11/ehb11202012a.html. 25. Rahm’s ObamaCare Brainstorm: Chicago May Dump Its Retiree Health Costs on Federal Taxpayers, Wall St. J., Jan. 26, 2013, http://online.wsj.com/ar ticle/SB10001424127887323968304578245702495382788.html. 26. The RomneyCare Bill Comes Due, Wall St. J., Jan. 23, 2013, http://online .wsj.com/article/SB10001424127887323968304578249694095831444.html. 27. For a more detailed discussion of these points, see Richard A. Epstein & Paula Stannard, Constitutional Ratemaking and the Affordable Care Act: A New Source of Vulnerability, 38 Am. J. of Law & Med., 243, 245– 61 (2012). 28. ACA §§ 1341-1342, discussed in Epstein & Stannard, supra note 27, at 254– 56. 29. For discussion, see id. at 259– 61. 30. For a recent account of the fiscal straits and the interdependence between the two plans, see Charles Blahous, Obamacare’s Fiscal Nightmare. Feb. 7, 2013, Accessed October 29, 2014. http://www.hoover.org/research/obamacares-fiscal-nightmare. 31. Richard A. Epstein, Government by Waiver, National Affairs, Spring 2011, http://www.nationalaffairs.com/publications/detail/government-by-waiver. 32. See Richard A. Epstein & David A. Hyman, Why Obamacare Will End Health Insurance as We Know It, Manhattan Inst. Issues (March 2012), http:// www.manhattan-institute.org/html/ir_7.htm. 33. Charles Marwick, For the Uninsured, Health Problems are More Serious, 94 J. Nat’l Cancer Institute, 967– 68. 34. Centers for Disease Control, Deaths and Mortality, http://www.cdc.gov /nchs/fastats/deaths.htm.

Contributors JONAT H A N H . A DLER

A M I TA BH CH A N DR A

Johan Verheij Memorial Professor of

Professor of Public Policy

Law and Director of the Center for

Harvard University

Business Law & Regulation

John F. Kennedy School of

Case Western University School of Law Cleveland, OH 44106

Government Cambridge, MA 02138 USA

USA JOH N H . COCH R A N E N ICHOLAS BAGLEY

Assistant Professor of Law University of Michigan Law School Ann Arbor, MI 48109 USA

AQR Capital Management Distinguished Service Professor of Finance, University of Chicago Booth School of Business; Senior Fellow, Hoover Institution; Research Associate, National

CH AR LES BLA HOUS

Senior Research Fellow and Direc-

Bureau of Economic Research; Adjunct Scholar, Cato Institute

tor of the Spending and Budget

University of Chicago

Initiative at the Mercatus Center;

Booth School of Business

Research Fellow at the Hoover

Chicago, IL 60637

Institution

USA

Mercatus Center George Mason University

EI N ER ELH AUGE

Arlington, VA 22201

Petrie Professor of Law

USA

Harvard University Law School Cambridge, MA 02138 USA

Contributors

354 R ICH AR D A. EPST EI N

Harvard Medical School

Laurence A. Tisch Professor of Law,

Department of Health Care Policy

New York University School of

Boston, MA 02115

Law; Peter and Kirsten Bedford

USA

Senior Fellow, Hoover Institution; James Parker Hall Distinguished Service Professor of Law, University of Chicago University of Chicago Law School Chicago, IL 60637 USA ST EPH A N I E P. H A LES

Associate

DA R I US LA K DAWA LLA

Quintiles Professor of Pharmaceutical Development and Regulatory Innovation, Schaeffer Center for Health Policy and Economics University of Southern California Sol Price School of Public Policy Los Angeles, CA 90089 USA

Sidley Austin LLP

H ELEN LEV Y

Washington, D.C. 20005

Research Associate Professor

USA

University of Michigan School of

JONAT H A N HOLM ES

Institute for Social Research

Research Fellow

Ann Arbor, MI 48104

Harvard University

USA

Public Health

John F. Kennedy School of Government Cambridge, MA 02138 USA

A N U P M A LA N I

Lee and Brena Freeman Professor of Law University of Chicago Law School

AZIZ Z . H UQ

Chicago, IL 60637

Professor of Law

USA

University of Chicago Law School Chicago, IL 60637 USA

ST EPH EN T. PA R EN T E

Minnesota Insurance Chair of Health and Finance University of Minnesota

A N UPA M B . JENA

Assistant Professor of Health Care Policy and Medicine at Harvard Medical School and Massachu-

Carlson School of Management Minneapolis, MN 55455 USA

setts General Hospital; Faculty

TOM AS J. PH I LI PSON

Research Fellow, National Bureau

Daniel Levin Professor of Public

of Economic Research

Policy Studies

Contributors University of Chicago Harris School for Public Policy

355

University of Illinois at Urbana- Champaign

Chicago, IL 60637

College of Business and IGPA

USA

Champaign, IL 61820 USA

CA RT ER G. PH I LLI PS

Partner Sidley Austin LLP Washington, D.C. 20005 USA

M ER EDI T H B . ROSEN T H A L

Professor of Health Economics and Policy and Associate Dean for Diversity Harvard University

JA M ES B. R EBI TZER

Professor of Management, Economics, and Public Policy; Lord Distinguished Faculty Scholar Boston University

Department of Health Policy and Management Boston, MA 02115 USA

School of Management Boston, MA 02115

M ICH A EL H . SCH I LL

USA

Dean and Harry N. Wyatt Professor

J U LI A N R EI F

University of Chicago Law School

Assistant Professor of Finance and

Chicago, IL 60637

of Law

Economics

USA

Index

Commerce Clause, 14–16, 18– 24, 30n13, 36– 37, 40–43, 75n17 Congressional Budget Office (CBO), 4, 86, 110–19, 123– 24, 126– 28, 130, 347

incentives, 5– 6, 15, 25, 38, 39, 52, 60, 61, 72, 74n14, 101– 2, 128– 29, 161, 163– 64, 166, 168, 173, 192, 203–13, 216–17, 221, 225– 30, 232– 37, 241–43, 253, 259– 60, 263, 265, 273– 74, 277– 83, 286, 290– 91, 293– 95, 298, 304, 306n6, 317, 319– 20, 324, 330, 339, 341, 348 Independent Payment Advisory Board (IPAB), 29n4, 51, 69, 79n88, 121, 214–15, 233, 236, 293 Institute of Medicine (IOM), 84, 85, 86, 87, 95, 103n2, 204, 205, 209, 218n8

Department of Health and Human Services (HHS), 3, 26, 28, 34, 35, 52, 55– 59, 61, 64– 71, 73, 75n29, 82–102, 129,171, 236, 331, 343–44, 347, 349n8

Kagan, Justice Elena, 27, 36 Kennedy, Justice Anthony, 15, 29, 36, 53 Kennedy School of Government, 8, 199

FDA v. Brown & Williamson Tobacco Corp., 92, 93 Federal Poverty Level (FPL), 8, 26, 55, 129,143, 145–46, 151, 153, 156, 300 Federal Register, 58, 88, 102 fi rst- dollar payments, 185, 193; fi rst- dollar coverage, 294

Medicare Hospital Insurance (HI), 110, 112, 115, 116, 117, 131n9 Medicare Part B, 115, 117, 213, 232 Medicare Part D, 1, 17, 115, 292, 322 Medicare Payment Assessment Commission (MedPac), 222 metallic plans, 323, 343; bronze, 144–45, 323; gold, 282, 323; silver, 144, 184, 323, 343

Accountable Care Organizations (ACO), 5, 171, 213, 214, 233– 37, 239–41, 276, 304 Alito, Justice Samuel, 36 antiretroviral, 255, 283, 290, 295 Breyer, Justice Stephen, 27, 36, 95 Brown, Scott, 53, 62

Ginsburg, Justice Ruth Bader, 36 healthcare savings accounts (HSA), 5, 184, 194; savings accounts, 4, 5, 155 high- deductible health plan (HDHP), 135– 36, 141, 144, 155, 156, 222 highly active antiretroviral therapy (HAART), 283– 84, 295– 96, 301– 2, 305n2, 307n12

National Health Law Program (NHeLP), 95– 96 New Deal, 1, 19, 38, 42 Obama, President Barack, 1, 14, 35, 56, 88, 98, 99, 131n6, 218n3; Obama Administration, 122, 331

Index

358 Office of Information and Regulatory Affairs (OIRA), 81– 82, 94, 98– 99 Patient Protection and Affordable Care Act (PPACA): Section 1311, 54– 56, 58, 59– 60, 91; Section 1401, 55– 56, 58– 60, 91; Section 1421, 59 pay for performance, 226, 227, 229, 238, 241 preexisting conditions, 17, 71, 183, 186– 87, 345 provider report cards, 224 Quality Adjusted Life Year (QALY), 272– 73, 277– 78

readmission, 230– 31, 234, 235– 36, 239–41, 265, 304 Religious Freedom Restoration Act (RFRA), 66– 68 religious objections, 64– 65, 67 Roberts, Justice John, 3, 13–14, 22– 23, 28, 36, 37– 38, 42, 44–45, 71– 72 Scalia, Justice Antonin, 20, 36 Sotomayor, Justice Sonia, 36 Thomas, Justice Clarence, 36, 38 Unearned Income Medicare Contribution (UIMC), 120– 21