National Federation of Independent Business v. Sebelius
Annotate this Case
567 U.S. ___ (2012)
- Syllabus |
- Opinion (Stephen G. Breyer) |
- Concurrence (Ruth Bader Ginsburg) |
- Dissent (Clarence Thomas) |
SUPREME COURT OF THE UNITED STATES
Nos. 11–393, 11–398 and 11–400
NATIONAL FEDERATION OF INDEPENDENT BUSINESS, et al., PETITIONERS
KATHLEEN SEBELIUS, SECRETARY OF HEALTH AND HUMAN SERVICES, et al.
DEPARTMENT OF HEALTH AND HUMAN SERVICES, et al., PETITIONERS
FLORIDA et al.
FLORIDA, et al., PETITIONERS
DEPARTMENT OF HEALTH AND HUMAN SERVICES et al.
on writs of certiorari to the united states court of appeals for the eleventh circuit
[June 28, 2012]
Justice Scalia, Justice Kennedy, Justice Thomas, and Justice Alito, dissenting.
Congress has set out to remedy the problem that the best health care is beyond the reach of many Americans who cannot afford it. It can assuredly do that, by exercising the powers accorded to it under the Constitution. The question in this case, however, is whether the complex structures and provisions of the Patient Protection and Affordable Care Act (Affordable Care Act or ACA) go be- yond those powers. We conclude that they do.
This case is in one respect difficult: it presents two questions of first impression. The first of those is whether failure to engage in economic activity (the purchase of health insurance) is subject to regulation under the Commerce Clause. Failure to act does result in an effect on commerce, and hence might be said to come under this Court’s “affecting commerce” criterion of Commerce Clause jurisprudence. But in none of its decisions has this Court extended the Clause that far. The second question is whether the congressional power to tax and spend, U. S. Const., Art. I, §8, cl. 1, permits the conditioning of a State’s continued receipt of all funds under a massive state-administered federal welfare program upon its acceptance of an expansion to that program. Several of our opinions have suggested that the power to tax and spend cannot be used to coerce state administration of a federal program, but we have never found a law enacted under the spending power to be coercive. Those questions are difficult.
The case is easy and straightforward, however, in another respect. What is absolutely clear, affirmed by the text of the 1789 Constitution, by the Tenth Amendment ratified in 1791, and by innumerable cases of ours in the 220 years since, is that there are structural limits upon federal power—upon what it can prescribe with respect to private conduct, and upon what it can impose upon the sovereign States. Whatever may be the conceptual limits upon the Commerce Clause and upon the power to tax and spend, they cannot be such as will enable the Federal Government to regulate all private conduct and to com- pel the States to function as administrators of federal programs.
That clear principle carries the day here. The striking case of Wickard v. Filburn, 317 U. S. 111 (1942) , which held that the economic activity of growing wheat, even for one’s own consumption, affected commerce sufficiently that it could be regulated, always has been regarded as the ne plus ultra of expansive Commerce Clause jurisprudence. To go beyond that, and to say the failure to grow wheat (which is not an economic activity, or any activity at all) nonetheless affects commerce and therefore can be federally regulated, is to make mere breathing in and out the basis for federal prescription and to extend federal power to virtually all human activity.
As for the constitutional power to tax and spend for the general welfare: The Court has long since expanded that beyond (what Madison thought it meant) taxing and spending for those aspects of the general welfare that were within the Federal Government’s enumerated powers, see United States v. Butler, 297 U. S. 1 –66 (1936). Thus, we now have sizable federal Departments devoted to subjects not mentioned among Congress’ enumerated powers, and only marginally related to commerce: the De- partment of Education, the Department of Health and Human Services, the Department of Housing and Urban Development. The principal practical obstacle that prevents Congress from using the tax-and-spend power to assume all the general-welfare responsibilities traditionally exercised by the States is the sheer impossibility of managing a Federal Government large enough to administer such a system. That obstacle can be overcome by granting funds to the States, allowing them to administer the program. That is fair and constitutional enough when the States freely agree to have their powers employed and their employees enlisted in the federal scheme. But it is a blatant violation of the constitutional structure when the States have no choice.
The Act before us here exceeds federal power both in mandating the purchase of health insurance and in denying nonconsenting States all Medicaid funding. These parts of the Act are central to its design and operation, and all the Act’s other provisions would not have been enacted without them. In our view it must follow that the entire statute is inoperative.
The Individual Mandate
Article I, §8, of the Constitution gives Congress the power to “regulate Commerce . . . among the several States.” The Individual Mandate in the Act commands that every “applicable individual shall for each month beginning after 2013 ensure that the individual, and any dependent of the individual who is an applicable individ- ual, is covered under minimum essential coverage.” 26 U. S. C. §5000A(a) (2006 ed., Supp. IV). If this provision “regulates” anything, it is the failure to maintain minimum essential coverage. One might argue that it regulates that failure by requiring it to be accompanied by payment of a penalty. But that failure—that abstention from commerce—is not “Commerce.” To be sure, purchasing insurance is ”Commerce”; but one does not regulate commerce that does not exist by compelling its existence.
In Gibbons v. Ogden, 9 Wheat. 1, 196 (1824), Chief Justice Marshall wrote that the power to regulate commerce is the power “to prescribe the rule by which commerce is to be governed.” That understanding is con- sistent with the original meaning of “regulate” at the time of the Constitution’s ratification, when “to regulate” meant “[t]o adjust by rule, method or established mode,” 2 N. Webster, An American Dictionary of the English Language (1828); “[t]o adjust by rule or method,” 2 S. Johnson, A Dictionary of the English Language (7th ed. 1785); “[t]o adjust, to direct according to rule,” 2 J. Ash, New and Complete Dictionary of the English Language (1775); “to put in order, set to rights, govern or keep in order,” T. Dyche & W. Pardon, A New General English Dictionary
(16th ed. 1777). [ 1 ] It can mean to direct the manner of something but not to direct that something come into being. There is no instance in which this Court or Congress (or anyone else, to our knowledge) has used “regulate” in that peculiar fashion. If the word bore that meaning, Congress’ authority “[t]o make Rules for the Govern- ment and Regulation of the land and naval Forces,” U. S. Const., Art. I, §8, cl. 14, would have made superfluous the later provision for authority “[t]o raise and support Armies,” id., §8, cl. 12, and “[t]o provide and maintain a Navy,” id., §8, cl. 13.
We do not doubt that the buying and selling of health insurance contracts is commerce generally subject to federal regulation. But when Congress provides that (nearly) all citizens must buy an insurance contract, it goes beyond “adjust[ing] by rule or method,” Johnson, supra, or “direct[ing] according to rule,” Ash, supra; it directs the creation of commerce.
In response, the Government offers two theories as to why the Individual Mandate is nevertheless constitutional. Neither theory suffices to sustain its validity.
First, the Government submits that §5000A is “integral to the Affordable Care Act’s insurance reforms” and “necessary to make effective the Act’s core reforms.” Brief for Petitioners in No. 11–398 (Minimum Coverage Provision) 24 (hereinafter Petitioners’ Minimum Coverage Brief). Congress included a “finding” to similar effect in the Act
itself. See 42 U. S. C. §18091(2)(H).
As discussed in more detail in Part V, infra, the Act contains numerous health insurance reforms, but most notable for present purposes are the “guaranteed issue” and “community rating” provisions, §§300gg to 300gg–4. The former provides that, with a few exceptions, “each health insurance issuer that offers health insurance coverage in the individual or group market in a State must accept every employer and individual in the State that applies for such coverage.” §300gg–1(a). That is, an insurer may not deny coverage on the basis of, among other things, any pre-existing medical condition that the applicant may have, and the resulting insurance must cover that condition. See §300gg–3.
Under ordinary circumstances, of course, insurers would respond by charging high premiums to individuals with pre-existing conditions. The Act seeks to prevent this through the community-rating provision. Simply put, the community-rating provision requires insurers to calculate an individual’s insurance premium based on only four factors: (i) whether the individual’s plan covers just the individual or his family also, (ii) the “rating area” in which the individual lives, (iii) the individual’s age, and (iv) whether the individual uses tobacco. §300gg(a)(1)(A). Aside from the rough proxies of age and tobacco use (and possibly rating area), the Act does not allow an insurer to factor the individual’s health characteristics into the price of his insurance premium. This creates a new incentive for young and healthy individuals without pre-existing conditions. The insurance premiums for those in this group will not reflect their own low actuarial risks but will subsidize insurance for others in the pool. Many of them may decide that purchasing health insurance is not an eco- nomically sound decision—especially since the guaranteed-issue provision will enable them to purchase it at the same cost in later years and even if they have developed a pre-existing condition. But without the contribution of above-risk premiums from the young and healthy, the community-rating provision will not enable insurers to take on high-risk individuals without a massive increase in premiums.
The Government presents the Individual Mandate as a unique feature of a complicated regulatory scheme governing many parties with countervailing incentives that must be carefully balanced. Congress has imposed an extensive set of regulations on the health insurance industry, and compliance with those regulations will likely cost the in- dustry a great deal. If the industry does not respond by increasing premiums, it is not likely to survive. And if the industry does increase premiums, then there is a seri- ous risk that its products—insurance plans—will become economically undesirable for many and prohibitively ex- pensive for the rest.
This is not a dilemma unique to regulation of the health-insurance industry. Government regulation typically imposes costs on the regulated industry—especially regulation that prohibits economic behavior in which most market participants are already engaging, such as “piecing out” the market by selling the product to different classes of people at different prices (in the present context, providing much lower insurance rates to young and healthy buyers). And many industries so regulated face the reality that, without an artificial increase in demand, they cannot continue on. When Congress is regulating these industries directly, it enjoys the broad power to enact “ ‘all appropriate legislation’ ” to “ ‘protec[t]’ ” and “ ‘advanc[e]’ ” commerce, NLRB v. Jones & Laughlin Steel Corp., 301 U. S. 1 –37 (1937) (quoting The Daniel Ball, 10 Wall. 557, 564 (1871)). Thus, Congress might protect the imperiled industry by prohibiting low-cost competition, or by according it preferential tax treatment, or even by granting it a direct subsidy.
Here, however, Congress has impressed into service third parties, healthy individuals who could be but are not customers of the relevant industry, to offset the undesirable consequences of the regulation. Congress’ desire to force these individuals to purchase insurance is motivated by the fact that they are further removed from the market than unhealthy individuals with pre-existing conditions, because they are less likely to need extensive care in the near future. If Congress can reach out and command even those furthest removed from an interstate market to participate in the market, then the Commerce Clause becomes a font of unlimited power, or in Hamilton’s words, “the hideous monster whose devouring jaws . . . spare neither sex nor age, nor high nor low, nor sacred nor profane.” The Federalist No. 33, p. 202 (C. Rossiter ed. 1961).
At the outer edge of the commerce power, this Court has insisted on careful scrutiny of regulations that do not act directly on an interstate market or its participants. In New York v. United States, 505 U. S. 144 (1992) , we held that Congress could not, in an effort to regulate the disposal of radioactive waste produced in several different industries, order the States to take title to that waste. Id., at 174–177. In Printz v. United States, 521 U. S. 898 (1997) , we held that Congress could not, in an effort to regulate the distribution of firearms in the interstate mar- ket, compel state law-enforcement officials to perform background checks. Id., at 933–935. In United States v. Lopez, 514 U. S. 549 (1995) , we held that Congress could not, as a means of fostering an educated interstate labor market through the protection of schools, ban the possession of a firearm within a school zone. Id., at 559–563. And in United States v. Morrison, 529 U. S. 598 (2000) , we held that Congress could not, in an effort to ensure the full participation of women in the interstate economy, subject private individuals and companies to suit for gender-motivated violent torts. Id., at 609–619. The lesson of these cases is that the Commerce Clause, even when sup- plemented by the Necessary and Proper Clause, is not carte blanche for doing whatever will help achieve the ends Congress seeks by the regulation of commerce. And the last two of these cases show that the scope of the Necessary and Proper Clause is exceeded not only when the congressional action directly violates the sovereignty of the States but also when it violates the background principle of enumerated (and hence limited) federal power.
The case upon which the Government principally relies to sustain the Individual Mandate under the Necessary and Proper Clause is Gonzales v. Raich, 545 U. S. 1 (2005) . That case held that Congress could, in an effort to restrain the interstate market in marijuana, ban the local cultivation and possession of that drug. Id., at 15–22. Raich is no precedent for what Congress has done here. That case’s prohibition of growing (cf. Wickard, 317 U. S. 111 ), and of possession (cf. innumerable federal statutes) did not represent the expansion of the federal power to direct into a broad new field. The mandating of economic activity does, and since it is a field so limitless that it converts the Commerce Clause into a general authority to direct the economy, that mandating is not “consist[ent] with the letter and spirit of the constitution.” McCulloch v. Maryland, 4 Wheat. 316, 421 (1819).
Moreover, Raich is far different from the Individual Mandate in another respect. The Court’s opinion in Raich pointed out that the growing and possession prohibitions were the only practicable way of enabling the prohibition of interstate traffic in marijuana to be effectively enforced. 545 U. S., at 22. See also Shreveport Rate Cases, 234 U. S. 342 (1914) (Necessary and Proper Clause allows regulations of intrastate transactions if necessary to the regulation of an interstate market). Intrastate marijuana could no more be distinguished from interstate marijuana than, for example, endangered-species trophies obtained before the species was federally protected can be distinguished from trophies obtained afterwards—which made it necessary and proper to prohibit the sale of all such trophies, see Andrus v. Allard, 444 U. S. 51 (1979) .
With the present statute, by contrast, there are many ways other than this unprecedented Individual Mandate by which the regulatory scheme’s goals of reducing insurance premiums and ensuring the profitability of insurers could be achieved. For instance, those who did not purchase insurance could be subjected to a surcharge when they do enter the health insurance system. Or they could be denied a full income tax credit given to those who do purchase the insurance.
The Government was invited, at oral argument, to suggest what federal controls over private conduct (other than those explicitly prohibited by the Bill of Rights or other constitutional controls) could not be justified as necessary and proper for the carrying out of a general regulatory scheme. See Tr. of Oral Arg. 27–30, 43–45 (Mar. 27, 2012). It was unable to name any. As we said at the outset, whereas the precise scope of the Commerce Clause and the Necessary and Proper Clause is uncertain, the proposition that the Federal Government cannot do everything is a fundamental precept. See Lopez, 514 U. S., at 564 (“[I]f we were to accept the Government’s arguments, we are hard pressed to posit any activity by an in- dividual that Congress is without power to regulate”). Section 5000A is defeated by that proposition.
The Government’s second theory in support of the In- dividual Mandate is that §5000A is valid because it is actually a “regulat[ion of] activities having a substantial relation to interstate commerce, . . . i.e., . . . activities that substantially affect interstate commerce.” Id., at 558–559. See also Shreveport Rate Cases, supra. This argument takes a few different forms, but the basic idea is that §5000A regulates “the way in which individuals finance their participation in the health-care market.” Petitioners’ Minimum Coverage Brief 33 (emphasis added). That is, the provision directs the manner in which individuals purchase health care services and related goods (directing that they be purchased through insurance) and is therefore a straightforward exercise of the commerce power.
The primary problem with this argument is that §5000A does not apply only to persons who purchase all, or most, or even any, of the health care services or goods that the mandated insurance covers. Indeed, the main objection many have to the Mandate is that they have no intention of purchasing most or even any of such goods or services and thus no need to buy insurance for those purchases. The Government responds that the health-care market involves “essentially universal participation,” id., at 35. The principal difficulty with this response is that it is, in the only relevant sense, not true. It is true enough that everyone consumes “health care,” if the term is taken to include the purchase of a bottle of aspirin. But the health care “market” that is the object of the Individual Mandate not only includes but principally consists of goods and services that the young people primarily affected by the Mandate do not purchase. They are quite simply not participants in that market, and cannot be made so (and thereby subjected to regulation) by the simple device of defining participants to include all those who will, later in their lifetime, probably purchase the goods or services covered by the mandated insurance. [ 2 ] Such a definition of market participants is unprecedented, and were it to be a premise for the exercise of national power, it would have no principled limits.
In a variation on this attempted exercise of federal power, the Government points out that Congress in this Act has purported to regulate “economic and financial decision[s] to forego [sic] health insurance coverage and [to] attempt to self-insure,” 42 U. S. C. §18091(2)(A), since those decisions have “a substantial and deleterious effect on interstate commerce,” Petitioners’ Minimum Coverage Brief 34. But as the discussion above makes clear, the decision to forgo participation in an interstate market is not itself commercial activity (or indeed any activity at all) within Congress’ power to regulate. It is true that, at the end of the day, it is inevitable that each American will affect commerce and become a part of it, even if not by choice. But if every person comes within the Commerce Clause power of Congress to regulate by the simple reason that he will one day engage in commerce, the idea of a limited Government power is at an end.
Wickard v. Filburn has been regarded as the most expansive assertion of the commerce power in our history. A close second is Perez v. United States, 402 U. S. 146 (1971) , which upheld a statute criminalizing the eminently local activity of loan-sharking. Both of those cases, however,
involved commercial activity. To go beyond that, and to say that the failure to grow wheat or the refusal to make loans affects commerce, so that growing and lending can be federally compelled, is to extend federal power to virtually everything. All of us consume food, and when we do so the Federal Government can prescribe what its quality must be and even how much we must pay. But the mere fact that we all consume food and are thus, sooner or later, participants in the “market” for food, does not empower the Government to say when and what we will buy. That is essentially what this Act seeks to do with respect to the purchase of health care. It exceeds federal power.
A few respectful responses to Justice Ginsburg’s dissent on the issue of the Mandate are in order. That dissent duly recites the test of Commerce Clause power that our opinions have applied, but disregards the premise the test contains. It is true enough that Congress needs only a “ ‘rational basis’ for concluding that the regulated activity substantially affects interstate commerce,” ante, at 15 (em- phasis added). But it must be activity affecting com- merce that is regulated, and not merely the failure to engage in commerce. And one is not now purchasing the health care covered by the insurance mandate simply because one is likely to be purchasing it in the future. Our test’s premise of regulated activity is not invented out of whole cloth, but rests upon the Constitution’s requirement that it be commerce which is regulated. If all inactivity affecting commerce is commerce, commerce is everything. Ultimately the dissent is driven to saying that there is really no difference between action and inaction, ante, at 26, a proposition that has never recommended itself, neither to the law nor to common sense. To say, for example, that the inaction here consists of activity in “the self-insurance market,” ibid., seems to us wordplay. By parity of reasoning the failure to buy a car can be called participation in the non-private-car-transportation market. Com- merce becomes everything.
The dissent claims that we “fai[l] to explain why the individual mandate threatens our constitutional order.” Ante, at 35. But we have done so. It threatens that order because it gives such an expansive meaning to the Commerce Clause that all private conduct (including failure to act) becomes subject to federal control, effectively destroying the Constitution’s division of governmental powers. Thus the dissent, on the theories proposed for the validity of the Mandate, would alter the accepted constitutional relation between the individual and the National Government. The dissent protests that the Necessary and Proper Clause has been held to include “the power to enact criminal laws, . . . the power to imprison, . . . and the power to create a national bank,” ante, at 34–35. Is not the power to compel purchase of health insurance much lesser? No, not if (unlike those other dispositions) its application rests upon a theory that everything is within federal control simply because it exists.
The dissent’s exposition of the wonderful things the Fed- eral Government has achieved through exercise of its assigned powers, such as “the provision of old-age and survivors’ benefits” in the Social Security Act, ante, at 2, is quite beside the point. The issue here is whether the federal government can impose the Individual Mandate through the Commerce Clause. And the relevant history is not that Congress has achieved wide and wonderful results through the proper exercise of its assigned powers in the past, but that it has never before used the Commerce Clause to compel entry into commerce. [ 3 ] The dissent treats the Constitution as though it is an enumeration of those problems that the Federal Government can address—among which, it finds, is “the Nation’s course in the economic and social welfare realm,” ibid., and more specifically “the problem of the uninsured,” ante, at 7. The Constitution is not that. It enumerates not federally soluble problems, but federally available powers. The Federal Government can address whatever problems it wants but can bring to their solution only those powers that the Constitution confers, among which is the power to regulate commerce. None of our cases say anything else. Article I contains no whatever-it-takes-to-solve-a-national-problem power.
The dissent dismisses the conclusion that the power to compel entry into the health-insurance market would include the power to compel entry into the new-car or broccoli markets. The latter purchasers, it says, “will be obliged to pay at the counter before receiving the vehicle or nourishment,” whereas those refusing to purchase health-insurance will ultimately get treated anyway, at others’ expense. Ante, at 21. “[T]he unique attributes of the health-care market . . . give rise to a significant free-riding problem that does not occur in other markets.” Ante, at 28. And “a vegetable-purchase mandate” (or a car-purchase mandate) is not “likely to have a substantial effect on the health-care costs” borne by other Americans. Ante, at 29. Those differences make a very good argument by the dissent’s own lights, since they show that the failure to purchase health insurance, unlike the failure to purchase cars or broccoli, creates a national, social-welfare problem that is (in the dissent’s view) included among the unenumerated “problems” that the Constitution authorizes the Federal Government to solve. But those differences do not show that the failure to enter the health-insurance market, unlike the failure to buy cars and broccoli, is an activity that Congress can “regulate.” (Of course one day the failure of some of the public to purchase Amer- ican cars may endanger the existence of domestic automobile manufacturers; or the failure of some to eat broccoli may be found to deprive them of a newly discovered cancer- fighting chemical which only that food contains, producing health-care costs that are a burden on the rest of us—in which case, under the theory of Justice Ginsburg’s dissent, moving against those inactivities will also come within the Federal Government’s unenumerated problem-solving powers.)
The Taxing Power
As far as §5000A is concerned, we would stop there. Congress has attempted to regulate beyond the scope of its Commerce Clause authority, [ 4 ] and §5000A is therefore invalid. The Government contends, however, as expressed in the caption to Part II of its brief, that “the minimum coverage provision is independently authorized by congress’s taxing power.” Petitioners’ Minimum Coverage Brief 52. The phrase “independently authorized” suggests the existence of a creature never hitherto seen in the United States Reports: A penalty for constitutional purposes that is also a tax for constitutional purposes. In all our cases the two are mutually exclusive. The provision challenged under the Constitution is either a penalty or else a tax. Of course in many cases what was a regu- latory mandate enforced by a penalty could have been imposed as a tax upon permissible action; or what was im- posed as a tax upon permissible action could have been a regulatory mandate enforced by a penalty. But we know of no case, and the Government cites none, in which the imposition was, for constitutional purposes, both. [ 5 ] The two are mutually exclusive. Thus, what the Government’s caption should have read was “alternatively, the minimum coverage provision is not a mandate-with-penalty but a tax.” It is important to bear this in mind in evaluating the tax argument of the Government and of those who support it: The issue is not whether Congress had the power to frame the minimum-coverage provision as a tax, but whether it did so.
In answering that question we must, if “fairly possible,” Crowell v. Benson, 285 U. S. 22, 62 (1932) , construe the provision to be a tax rather than a mandate-with-penalty, since that would render it constitutional rather than un- constitutional (ut res magis valeat quam pereat). But we cannot rewrite the statute to be what it is not. “ ‘ “[A]l- though this Court will often strain to construe legis- lation so as to save it against constitutional attack, it must not and will not carry this to the point of perverting the purpose of a statute . . .” or judicially rewriting it.’ ” Commodity Futures Trading Comm’n v. Schor, 478 U. S. 833, 841 (1986) (quoting Aptheker v. Secretary of State, 378 U. S. 500, 515 (1964) , in turn quoting Scales v. United States, 367 U. S. 203, 211 (1961) ). In this case, there is simply no way, “without doing violence to the fair meaning of the words used,” Grenada County Supervisors v. Brogden, 112 U. S. 261, 269 (1884) , to escape what Congress enacted: a mandate that individuals maintain minimum essential coverage, enforced by a penalty.
Our cases establish a clear line between a tax and a penalty: “ ‘[A] tax is an enforced contribution to provide for the support of government; a penalty . . . is an exaction imposed by statute as punishment for an unlawful act.’ ” United States v. Reorganized CF&I Fabricators of Utah, Inc., 518 U. S. 213, 224 (1996) (quoting United States v. La Franca, 282 U. S. 568, 572 (1931) ). In a few cases, this Court has held that a “tax” imposed upon private conduct was so onerous as to be in effect a penalty. But we have never held—never—that a penalty imposed for violation of the law was so trivial as to be in effect a tax. We have never held that any exaction imposed for violation of the law is an exercise of Congress’ taxing power—even when the statute calls it a tax, much less when (as here) the statute repeatedly calls it a penalty. When an act “adopt[s] the criteria of wrongdoing” and then imposes a monetary penalty as the “principal consequence on those who transgress its standard,” it creates a regulatory penalty, not a tax. Child Labor Tax Case, 259 U. S. 20, 38 (1922) .
So the question is, quite simply, whether the exaction here is imposed for violation of the law. It unquestion- ably is. The minimum-coverage provision is found in 26 U. S. C. §5000A, entitled “Requirement to maintain minimum essential coverage.” (Emphasis added.) It commands that every “applicable individual shall . . . ensure that the individual . . . is covered under minimum essential coverage.” Ibid. (emphasis added). And the immediately following provision states that, “[i]f . . . an applicable individual . . . fails to meet the requirement of subsection (a) . . . there is hereby imposed . . . a penalty.” §5000A(b) (emphasis added). And several of Congress’ legislative “findings” with regard to §5000A confirm that it sets forth a legal requirement and constitutes the assertion of regulatory power, not mere taxing power. See 42 U. S. C. §18091(2)(A) (“The requirement regulates activity . . .”); §18091(2)(C) (“The requirement . . . will add millions of new consumers to the health insurance market . . .”); §18091(2)(D) (“The requirement achieves near-universal coverage”); §18091(2)(H) (“The requirement is an essential part of this larger regulation of economic activity, and the absence of the requirement would undercut Federal regulation of the health insurance market”); §18091(3) (“[T]he Supreme Court of the United States ruled that insurance is interstate commerce subject to Federal regulation”).
The Government and those who support its view on the tax point rely on New York v. United States, 505 U. S. 144 , to justify reading “shall” to mean “may.” The “shall” in that case was contained in an introductory provision—a recital that provided for no legal consequences—which said that “[e]ach State shall be responsible for providing . . . for the disposal of . . . low-level radioactive waste.” 42 U. S. C. §2021c(a)(1)(A). The Court did not hold that “shall” could be construed to mean “may,” but rather that this preliminary provision could not impose upon the oper- ative provisions of the Act a mandate that they did not contain: “We . . . decline petitioners’ invitation to con- strue §2021c(a)(1)(A), alone and in isolation, as a command to the States independent of the remainder of the Act.” New York, 505 U. S., at 170. Our opinion then proceeded to “consider each [of the three operative provisions] in turn.” Ibid. Here the mandate—the “shall”—is contained not in an inoperative preliminary recital, but in the dispositive operative provision itself. New York provides no support for reading it to be permissive.
Quite separately, the fact that Congress (in its own words) “imposed . . . a penalty,” 26 U. S. C. §5000A(b)(1), for failure to buy insurance is alone sufficient to render that failure unlawful. It is one of the canons of interpretation that a statute that penalizes an act makes it unlawful: “[W]here the statute inflicts a penalty for doing an act, although the act itself is not expressly prohibited, yet to do the act is unlawful, because it cannot be supposed that the Legislature intended that a penalty should be inflicted for a lawful act.” Powhatan Steamboat Co. v. Appomattox R. Co., 24 How. 247, 252 (1861). Or in the words of Chancellor Kent: “If a statute inflicts a penalty for doing an act, the penalty implies a prohibition, and the thing is unlawful, though there be no prohibitory words in the statute.” 1 J. Kent, Commentaries on American Law 436 (1826).
We never have classified as a tax an exaction imposed for violation of the law, and so too, we never have classified as a tax an exaction described in the legislation itself as a penalty. To be sure, we have sometimes treated as a tax a statutory exaction (imposed for something other than a violation of law) which bore an agnostic label that does not entail the significant constitutional consequences of a penalty—such as “license” (License Tax Cases, 5 Wall. 462 (1867)) or “surcharge” (New York v. United States, supra.). But we have never—never—treated as a tax an exaction which faces up to the critical difference between a tax and a penalty, and explicitly denominates the exaction a “penalty.” Eighteen times in §5000A itself and else- where throughout the Act, Congress called the exaction in §5000A(b) a “penalty.”
That §5000A imposes not a simple tax but a mandate to which a penalty is attached is demonstrated by the fact that some are exempt from the tax who are not ex- empt from the mandate—a distinction that would make no sense if the mandate were not a mandate. Section 5000A(d) exempts three classes of people from the definition of “applicable individual” subject to the minimum coverage requirement: Those with religious objections or who participate in a “health care sharing ministry,” §5000A(d)(2); those who are “not lawfully present” in the United States, §5000A(d)(3); and those who are incarcerated, §5000A(d)(4). Section 5000A(e) then creates a separate set of exemptions, excusing from liability for the penalty certain individuals who are subject to the minimum coverage requirement: Those who cannot afford coverage, §5000A(e)(1); who earn too little income to require filing a tax return, §5000A(e)(2); who are members of an Indian tribe, §5000A(e)(3); who experience only short gaps in coverage, §5000A(e)(4); and who, in the judgment of the Secretary of Health and Human Services, “have suffered a hardship with respect to the capability to obtain coverage,” §5000A(e)(5). If §5000A were a tax, these two classes of exemption would make no sense; there being no requirement, all the exemptions would attach to the penalty (renamed tax) alone.
In the face of all these indications of a regulatory requirement accompanied by a penalty, the Solicitor General assures us that “neither the Treasury Department nor the Department of Health and Human Services interprets Section 5000A as imposing a legal obligation,” Petitioners’ Minimum Coverage Brief 61, and that “[i]f [those subject to the Act] pay the tax penalty, they’re in compliance with the law,” Tr. of Oral Arg. 50 (Mar. 26, 2012). These self-serving litigating positions are entitled to no weight. What counts is what the statute says, and that is entirely clear. It is worth noting, moreover, that these assurances contradict the Government’s position in related litigation. Shortly before the Affordable Care Act was passed, the Commonwealth of Virginia enacted Va. Code Ann. §38.2–3430.1:1 (Lexis Supp. 2011), which states, “No resident of [the] Commonwealth . . . shall be required to obtain or maintain a policy of individual insurance coverage except as required by a court or the Department of Social Services . . . .” In opposing Virginia’s assertion of standing to challenge §5000A based on this statute, the Government said that “if the minimum coverage provision is unconstitutional, the [Virginia] statute is unnecessary, and if the minimum coverage provision is upheld, the state statute is void under the Supremacy Clause.” Brief for Appellant in No. 11–1057 etc. (CA4), p. 29. But it would be void under the Supremacy Clause only if it was contradicted by a federal “require[ment] to obtain or maintain a policy of individual insurance coverage.”
Against the mountain of evidence that the minimum coverage requirement is what the statute calls it—a requirement—and that the penalty for its violation is what the statute calls it—a penalty—the Government brings forward the flimsiest of indications to the contrary. It notes that “[t]he minimum coverage provision amends the Internal Revenue Code to provide that a non-exempted individual . . . will owe a monetary penalty, in addition to the income tax itself,” and that “[t]he [Internal Revenue Service (IRS)] will assess and collect the penalty in the same manner as assessable penalties under the Internal Revenue Code.” Petitioners’ Minimum Coverage Brief 53. The manner of collection could perhaps suggest a tax if IRS penalty-collection were unheard-of or rare. It is not. See, e.g., 26 U. S. C. §527(j) (2006 ed.) (IRS-collectible pen- alty for failure to make campaign-finance disclosures); §5761(c) (IRS-collectible penalty for domestic sales of to- bacco products labeled for export); §9707 (IRS-collectible penalty for failure to make required health-insurance premium payments on behalf of mining employees). In Reorganized CF&I Fabricators of Utah, Inc., 518 U. S. 213 , we held that an exaction not only enforced by the Commissioner of Internal Revenue but even called a “tax” was in fact a penalty. “[I]f the concept of penalty means anything,” we said, “it means punishment for an unlawful act or omission.” Id., at 224. See also Lipke v. Lederer, 259 U. S. 557 (1922) (same). Moreover, while the penalty is assessed and collected by the IRS, §5000A is administered both by that agency and by the Department of Health and Human Services (and also the Secretary of Veteran Affairs), see §5000A(e)(1)(D), (e)(5), (f)(1)(A)(v), (f)(1)(E) (2006 ed., Supp. IV), which is responsible for defining its substantive scope—a feature that would be quite extraordinary for taxes.
The Government points out that “[t]he amount of the penalty will be calculated as a percentage of household income for federal income tax purposes, subject to a floor and [a] ca[p],” and that individuals who earn so little money that they “are not required to file income tax returns for the taxable year are not subject to the penalty” (though they are, as we discussed earlier, subject to the mandate). Petitioners’ Minimum Coverage Brief 12, 53. But varying a penalty according to ability to pay is an utterly familiar practice. See, e.g., 33 U. S. C. §1319(d) (2006 ed., Supp. IV) (“In determining the amount of a civil penalty the court shall consider . . . the economic impact of the penalty on the violator”); see also 6 U. S. C. §488e(c); 7 U. S. C. §§7734(b)(2), 8313(b)(2); 12 U. S. C. §§1701q–1(d)(3), 1723i(c)(3), 1735f–14(c)(3), 1735f–15(d)(3), 4585(c)(2); 15 U. S. C. §§45(m)(1)(C), 77h–1(g)(3), 78u–2(d), 80a–9(d)(4), 80b–3(i)(4), 1681s(a)(2)(B), 1717a(b)(3), 1825(b)(1), 2615(a) (2)(B), 5408(b)(2); 33 U. S. C. §2716a(a).
The last of the feeble arguments in favor of petition- ers that we will address is the contention that what this statute repeatedly calls a penalty is in fact a tax because it contains no scienter requirement. The presence of such a requirement suggests a penalty—though one can imagine a tax imposed only on willful action; but the absence of such a requirement does not suggest a tax. Penalties for absolute-liability offenses are commonplace. And where a statute is silent as to scienter, we traditionally presume a mens rea requirement if the statute imposes a “severe penalty.” Staples v. United States, 511 U. S. 600, 618 (1994) . Since we have an entire jurisprudence addressing when it is that a scienter requirement should be inferred from a penalty, it is quite illogical to suggest that a penalty is not a penalty for want of an express scienter requirement.
And the nail in the coffin is that the mandate and penalty are located in Title I of the Act, its operative core, rather than where a tax would be found—in Title IX, containing the Act’s “Revenue Provisions.” In sum, “the terms of [the] act rende[r] it unavoidable,” Parsons v. Bedford, 3 Pet. 433, 448 (1830), that Congress imposed a regulatory penalty, not a tax.
For all these reasons, to say that the Individual Mandate merely imposes a tax is not to interpret the statute but to rewrite it. Judicial tax-writing is particularly troubl- ing. Taxes have never been popular, see, e.g., Stamp Act of 1765, and in part for that reason, the Constitution requires tax increases to originate in the House of Representatives. See Art. I, §7, cl. 1. That is to say, they must originate in the legislative body most accountable to the people, where legislators must weigh the need for the tax against the terrible price they might pay at their next election, which is never more than two years off. The Federalist No. 58 “defend[ed] the decision to give the origination power to the House on the ground that the Chamber that is more accountable to the people should have the primary role in raising revenue.” United States v. Munoz-Flores, 495 U. S. 385, 395 (1990) . We have no doubt that Congress knew precisely what it was doing when it rejected an earlier version of this legislation that imposed a tax instead of a requirement-with-penalty. See Affordable Health Care for America Act, H. R. 3962, 111th Cong., 1st Sess., §501 (2009); America’s Healthy Future Act of 2009, S. 1796, 111th Cong., 1st Sess., §1301. Imposing a tax through judicial legislation inverts the constitutional scheme, and places the power to tax in the branch of government least accountable to the citizenry.
Finally, we must observe that rewriting §5000A as a tax in order to sustain its constitutionality would force us to confront a difficult constitutional question: whether this is a direct tax that must be apportioned among the States according to their population. Art. I, §9, cl. 4. Perhaps it is not (we have no need to address the point); but the meaning of the Direct Tax Clause is famously unclear, and its application here is a question of first impression that deserves more thoughtful consideration than the lick-and-a-promise accorded by the Government and its supporters. The Government’s opening brief did not even address the question—perhaps because, until today, no federal court has accepted the implausible argument that §5000A is an exercise of the tax power. And once respondents raised the issue, the Government devoted a mere 21 lines of its reply brief to the issue. Petitioners’ Minimum Coverage Reply Brief 25. At oral argument, the most prolonged statement about the issue was just over 50 words. Tr. of Oral Arg. 79 (Mar. 27, 2012). One would expect this Court to demand more than fly-by-night briefing and argument before deciding a difficult constitutional question of first impression.
The Anti-Injunction Act
There is another point related to the Individual Mandate that we must discuss—a point that logically should have been discussed first: Whether jurisdiction over the challenges to the minimum-coverage provision is precluded by the Anti-Injunction Act, which provides that “no suit for the purpose of restraining the assessment or collection of any tax shall be maintained in any court by any person,” 26 U. S. C. §7421(a) (2006 ed.).
We have left the question to this point because it seemed to us that the dispositive question whether the minimum-coverage provision is a tax is more appropriately addressed in the significant constitutional context of whether it is an exercise of Congress’ taxing power. Having found that it is not, we have no difficulty in deciding that these suits do not have “the purpose of restraining the assessment or collection of any tax.” [ 6 ]
The Government and those who support its position on this point make the remarkable argument that §5000A is not a tax for purposes of the Anti-Injunction Act, see Brief for Petitioners in No. 11–398 (Anti-Injunction Act), but is a tax for constitutional purposes, see Petitioners’ Minimum Coverage Brief 52–62. The rhetorical device that tries to cloak this argument in superficial plausibility is the same device employed in arguing that for constitutional purposes the minimum-coverage provision is a tax: confusing the question of what Congress did with the question of what Congress could have done. What qualifies as a tax for purposes of the Anti-Injunction Act, unlike what qualifies as a tax for purposes of the Constitution, is entirely within the control of Congress. Compare Bailey v. George, 259 U. S. 16, 20 (1922) (Anti-Injunction Act barred suit to restrain collections under the Child Labor Tax Law), with Child Labor Tax Case, 259 U. S., at 36–41 (holding the same law unconstitutional as exceeding Congress’ taxing power). Congress could have defined “tax” for purposes of that statute in such fashion as to exclude some exactions that in fact are “taxes.” It might have prescribed, for example, that a particular exercise of the taxing power “shall not be regarded as a tax for purposes of the Anti-Injunction Act.” But there is no such prescription here. What the Government would have us believe in these cases is that the very same textual indications that show this is not a tax under the Anti-Injunction Act show that it is a tax under the Constitution. That carries ver- bal wizardry too far, deep into the forbidden land of the sophists.
The Medicaid Expansion
We now consider respondents’ second challenge to the constitutionality of the ACA, namely, that the Act’s dramatic expansion of the Medicaid program exceeds Congress’ power to attach conditions to federal grants to the States.
The ACA does not legally compel the States to participate in the expanded Medicaid program, but the Act authorizes a severe sanction for any State that refuses to go along: termination of all the State’s Medicaid funding. For the average State, the annual federal Medicaid subsidy is equal to more than one-fifth of the State’s expenditures. [ 7 ] A State forced out of the program would not only lose this huge sum but would almost certainly find it necessary to increase its own health-care expenditures substantially, requiring either a drastic reduction in funding for other programs or a large increase in state taxes. And these new taxes would come on top of the federal taxes already paid by the State’s citizens to fund the Medicaid program in other States.
The States challenging the constitutionality of the ACA’s Medicaid Expansion contend that, for these practical reasons, the Act really does not give them any choice at all. As proof of this, they point to the goal and the struc- ture of the ACA. The goal of the Act is to provide near-universal medical coverage, 42 U. S. C. §18091(2)(D), and without 100% State participation in the Medicaid program, attainment of this goal would be thwarted. Even if States could elect to remain in the old Medicaid program, while declining to participate in the Expansion, there would be a gaping hole in coverage. And if a substantial number of States were entirely expelled from the program, the number of persons without coverage would be even higher.
In light of the ACA’s goal of near-universal coverage, petitioners argue, if Congress had thought that anything less than 100% state participation was a realistic possibility, Congress would have provided a backup scheme. But no such scheme is to be found anywhere in the more than 900 pages of the Act. This shows, they maintain, that Congress was certain that the ACA’s Medicaid offer was one that no State could refuse.
In response to this argument, the Government contends that any congressional assumption about uniform state participation was based on the simple fact that the offer of federal funds associated with the expanded coverage is such a generous gift that no State would want to turn it down.
To evaluate these arguments, we consider the extent of the Federal Government’s power to spend money and to attach conditions to money granted to the States.
No one has ever doubted that the Constitution authorizes the Federal Government to spend money, but for many years the scope of this power was unsettled. The Constitution grants Congress the power to collect taxes “to . . . provide for the . . . general Welfare of the United States,” Art. I, §8, cl. 1, and from “the foundation of the Nation sharp differences of opinion have persisted as to the true interpretation of the phrase” “the general welfare.” Butler, 297 U. S., at 65. Madison, it has been said, thought that the phrase “amounted to no more than a reference to the other powers enumerated in the subsequent clauses of the same section,” while Hamilton “maintained the clause confers a power separate and distinct from those later enumerated [and] is not restricted in meaning by the grant of them.” Ibid.
The Court resolved this dispute in Butler. Writing for the Court, Justice Roberts opined that the Madisonian view would make Article I’s grant of the spending power a “mere tautology.” Ibid. To avoid that, he adopted Hamilton’s approach and found that “the power of Congress to authorize expenditure of public moneys for public purposes is not limited by the direct grants of legislative power found in the Constitution.” Id., at 66. Instead, he wrote, the spending power’s “confines are set in the clause which confers it, and not in those of section 8 which bestow and define the legislative powers of the Congress.” Ibid.; see also Steward Machine Co. v. Davis, 301 U. S. 548 –587 (1937); Helvering v. Davis, 301 U. S. 619, 640 (1937) .
The power to make any expenditure that furthers “the general welfare” is obviously very broad, and shortly after Butler was decided the Court gave Congress wide leeway to decide whether an expenditure qualifies. See Helvering, 301 U. S., at 640–641. “The discretion belongs to Congress,” the Court wrote, “unless the choice is clearly wrong, a display of arbitrary power, not an exercise of judgment.” Id., at 640. Since that time, the Court has never held that a federal expenditure was not for “the general welfare.”
One way in which Congress may spend to promote the general welfare is by making grants to the States. Monetary grants, so-called grants-in-aid, became more frequent during the 1930’s, G. Stephens & N. Wikstrom, Ameri- can Intergovernmental Relations—A Fragmented Federal Polity 83 (2007), and by 1950 they had reached $20 billion [ 8 ] or 11.6% of state and local government expenditures from their own sources. [ 9 ] By 1970 this number had grown to $123.7 billion [ 10 ] or 29.1% of state and local government expenditures from their own sources. [ 11 ] As of 2010, fed- eral outlays to state and local governments came to over $608 billion or 37.5% of state and local government expenditures. [ 12 ]
When Congress makes grants to the States, it customarily attaches conditions, and this Court has long held that the Constitution generally permits Congress to do this. See Pennhurst State School and Hospital v. Halderman, 451 U. S. 1, 17 (1981) ; South Dakota v. Dole, 483 U. S. 203, 206 (1987) ; Fullilove v. Klutznick, 448 U. S. 448, 474 (1980) (opinion of Burger, C. J.); Steward Machine, supra, at 593.
This practice of attaching conditions to federal funds greatly increases federal power. “[O]bjectives not thought to be within Article I’s enumerated legislative fields, may nevertheless be attained through the use of the spending power and the conditional grant of federal funds.” Dole, supra, at 207 (internal quotation marks and citation omitted); see also College Savings Bank v. Florida Prepaid Postsecondary Ed. Expense Bd., 527 U. S. 666, 686 (1999) (by attaching conditions to federal funds, Congress may induce the States to “tak[e] certain actions that Congress could not require them to take”).
This formidable power, if not checked in any way, would present a grave threat to the system of federalism created by our Constitution. If Congress’ “Spending Clause power to pursue objectives outside of Article I’s enumerated legislative fields,” Davis v. Monroe County Bd. of Ed., 526 U. S. 629, 654 (1999) (Kennedy, J., dissenting) (internal quotation marks omitted), is “limited only by Congress’ notion of the general welfare, the reality, given the vast financial resources of the Federal Government, is that the Spending Clause gives ‘power to the Congress to tear down the barriers, to invade the states’ jurisdiction, and to become a parliament of the whole people, subject to no restrictions save such as are self-imposed,’ ” Dole, supra, at 217 (O’Connor, J., dissenting) (quoting Butler, 297 U. S., at 78). “[T]he Spending Clause power, if wielded without concern for the federal balance, has the potential to obliterate distinctions between national and local spheres of interest and power by permitting the Federal Government to set policy in the most sensitive areas of traditional state concern, areas which otherwise would lie outside its reach.” Davis, supra, at 654–655 (Kennedy, J., dissenting).
Recognizing this potential for abuse, our cases have long held that the power to attach conditions to grants to the States has limits. See, e.g., Dole, supra, at 207–208; id., at 207 (spending power is “subject to several general re- strictions articulated in our cases”). For one thing, any such conditions must be unambiguous so that a State at least knows what it is getting into. See Pennhurst, supra, at 17. Conditions must also be related “to the federal interest in particular national projects or programs,” Massachusetts v. United States, 435 U. S. 444, 461 (1978) , and the conditional grant of federal funds may not “induce the States to engage in activities that would themselves be unconstitutional,” Dole, supra, at 210; see Lawrence County v. Lead-Deadwood School Dist. No. 40–1, 469 U. S. 256 –270 (1985). Finally, while Congress may seek to induce States to accept conditional grants, Congress may not cross the “point at which pressure turns into compulsion, and ceases to be inducement.” Steward Machine, 301 U. S., at 590. Accord, College Savings Bank, supra, at 687; Metropolitan Washington Airports Authority v. Citizens for Abatement of Aircraft Noise, Inc., 501 U. S. 252, 285 (1991) (White, J., dissenting); Dole, supra, at 211.
When federal legislation gives the States a real choice whether to accept or decline a federal aid package, the federal-state relationship is in the nature of a contractual relationship. See Barnes v. Gorman, 536 U. S. 181, 186 (2002) ; Pennhurst, 451 U. S., at 17. And just as a contract is voidable if coerced, “[t]he legitimacy of Congress’ power to legislate under the spending power . . . rests on whether the State voluntarily and knowingly accepts the terms of the ‘contract.’ ” Ibid. (emphasis added). If a federal spending program coerces participation the States have not “exercise[d] their choice”—let alone made an “informed choice.” Id., at 17, 25.
Coercing States to accept conditions risks the destruction of the “unique role of the States in our system.” Davis, supra, at 685 (Kennedy, J., dissenting). “[T]he Constitution has never been understood to confer upon Congress the ability to require the States to govern according to Congress’ instructions.” New York, 505 U. S., at 162. Congress may not “simply commandeer the legislative processes of the States by directly compelling them to enact and enforce a federal regulatory program.” Id., at 161 (internal quotation marks and brackets omitted). Congress effectively engages in this impermissible compulsion when state participation in a federal spending program is coerced, so that the States’ choice whether to enact or administer a federal regulatory program is rendered illusory.
Where all Congress has done is to “encourag[e] state regulation rather than compe[l] it, state governments remain responsive to the local electorate’s preferences; state officials remain accountable to the people. [But] where the Federal Government compels States to regulate, the accountability of both state and federal officials is diminished.” New York, supra, at 168.
Amici who support the Government argue that forcing state employees to implement a federal program is more respectful of federalism than using federal workers to implement that program. See, e.g., Brief for Service Employees International Union et al. as Amici Curiae in No. 11–398, pp. 25–26. They note that Congress, instead of expanding Medicaid, could have established an entirely federal program to provide coverage for the same group of people. By choosing to structure Medicaid as a cooperative federal-state program, they contend, Congress allows for more state control. Ibid.
This argument reflects a view of federalism that our cases have rejected—and with good reason. When Congress compels the States to do its bidding, it blurs the lines of political accountability. If the Federal Government makes a controversial decision while acting on its own, “it is the Federal Government that makes the decision in full view of the public, and it will be federal officials that suffer the consequences if the decision turns out to be detrimental or unpopular.” New York, 505 U. S., at 168. But when the Federal Government compels the States to take unpopular actions, “it may be state officials who will bear the brunt of public disapproval, while the federal officials who devised the regulatory program may remain insulated from the electoral ramifications of their decision.” Id., at 169; see Printz, supra, at 930. For this reason, federal officeholders may view this “departur[e] from the federal structure to be in their personal interests . . . as a means of shifting responsibility for the eventual decision.” New York, 505 U. S., at 182–183. And even state officials may favor such a “departure from the constitutional plan,” since uncertainty concerning responsibility may also permit them to escape accountability. Id., at 182. If a program is popular, state officials may claim credit; if it is unpopular, they may protest that they were merely responding to a federal directive.
Once it is recognized that spending-power legislation cannot coerce state participation, two questions remain: (1) What is the meaning of coercion in this context? (2) Is the ACA’s expanded Medicaid coverage coercive? We now turn to those questions.
The answer to the first of these questions—the meaning of coercion in the present context—is straightforward. As we have explained, the legitimacy of attaching conditions to federal grants to the States depends on the voluntariness of the States’ choice to accept or decline the offered package. Therefore, if States really have no choice other than to accept the package, the offer is coercive, and the conditions cannot be sustained under the spending power. And as our decision in South Dakota v. Dole makes clear, theoretical voluntariness is not enough.
In South Dakota v. Dole, we considered whether the spending power permitted Congress to condition 5% of the State’s federal highway funds on the State’s adoption of a minimum drinking age of 21 years. South Dakota argued that the program was impermissibly coercive, but we disagreed, reasoning that “Congress ha[d] directed only that a State desiring to establish a minimum drinking age lower than 21 lose a relatively small percentage of certain federal highway funds.” 483 U. S., at 211. Because “all South Dakota would lose if she adhere[d] to her chosen course as to a suitable minimum drinking age [was] 5% of the funds otherwise obtainable under specified high- way grant programs,” we found that “Congress ha[d] of- fered relatively mild encouragement to the States to enact higher minimum drinking ages than they would otherwise choose.” Ibid. Thus, the decision whether to comply with the federal condition “remain[ed] the prerogative of the States not merely in theory but in fact,” and so the program at issue did not exceed Congress’ power. Id., at 211–212 (emphasis added).
The question whether a law enacted under the spending power is coercive in fact will sometimes be difficult, but where Congress has plainly “crossed the line distinguishing encouragement from coercion,” New York, supra, at 175, a federal program that coopts the States’ political processes must be declared unconstitutional. “[T]he federal balance is too essential a part of our constitutional structure and plays too vital a role in securing freedom for us to admit inability to intervene.” Lopez, 514 U. S., at 578 (Kennedy, J., concurring).
The Federal Government’s argument in this case at best pays lip service to the anticoercion principle. The Federal Government suggests that it is sufficient if States are “free, as a matter of law, to turn down” federal funds. Brief for Respondents in No. 11–400, p. 17 (emphasis added); see also id., at 25. According to the Federal Government, neither the amount of the offered federal funds nor the amount of the federal taxes extracted from the taxpayers of a State to pay for the program in question is relevant in determining whether there is impermissible coercion. Id., at 41–46.
This argument ignores reality. When a heavy federal tax is levied to support a federal program that offers large grants to the States, States may, as a practical matter, be unable to refuse to participate in the federal program and to substitute a state alternative. Even if a State believes that the federal program is ineffective and inefficient, withdrawal would likely force the State to impose a huge tax increase on its residents, and this new state tax would come on top of the federal taxes already paid by residents to support subsidies to participating States. [ 13 ]
Acceptance of the Federal Government’s interpreta- tion of the anticoercion rule would permit Congress to dic- tate policy in areas traditionally governed primarily at the state or local level. Suppose, for example, that Congress enacted legislation offering each State a grant equal to the State’s entire annual expenditures for primary and secondary education. Suppose also that this funding came with conditions governing such things as school curriculum, the hiring and tenure of teachers, the drawing of school districts, the length and hours of the school day, the school calendar, a dress code for students, and rules for student discipline. As a matter of law, a State could turn down that offer, but if it did so, its residents would not only be required to pay the federal taxes needed to support this expensive new program, but they would also be forced to pay an equivalent amount in state taxes. And if the State gave in to the federal law, the State and its subdivisions would surrender their traditional authority in the field of education. Asked at oral argument whether such a law would be allowed under the spending power, the Solicitor General responded that it would. Tr. of Oral Arg. 44–45 (Mar. 28, 2012).
Whether federal spending legislation crosses the line from enticement to coercion is often difficult to determine, and courts should not conclude that legislation is unconstitutional on this ground unless the coercive nature of an offer is unmistakably clear. In this case, however, there can be no doubt. In structuring the ACA, Congress unambiguously signaled its belief that every State would have no real choice but to go along with the Medicaid Expansion. If the anticoercion rule does not apply in this case, then there is no such rule.
The dimensions of the Medicaid program lend strong support to the petitioner States’ argument that refusing to accede to the conditions set out in the ACA is not a realistic option. Before the ACA’s enactment, Medicaid funded medical care for pregnant women, families with dependents, children, the blind, the elderly, and the disabled. See 42 U. S. C. §1396a(a)(10) (2006 ed., Supp. IV). The ACA greatly expands the program’s reach, making new funds available to States that agree to extend coverage to all individuals who are under age 65 and have incomes below 133% of the federal poverty line. See §1396a(a) (10)(A)(i)(VIII). Any State that refuses to expand its Medicaid programs in this way is threatened with a severe sanction: the loss of all its federal Medicaid funds. See §1396c (2006 ed.).
Medicaid has long been the largest federal program of grants to the States. See Brief for Respondents in No. 11–400, at 37. In 2010, the Federal Government directed more than $552 billion in federal funds to the States. See Nat. Assn. of State Budget Officers, 2010 State Expenditure Report: Examining Fiscal 2009–2011 State Spending, p. 7 (2011) (NASBO Report). Of this, more than $233 billion went to pre-expansion Medicaid. See id., at 47. [ 14 ] This amount equals nearly 22% of all state expenditures combined. See id., at 7.
The States devote a larger percentage of their budgets to Medicaid than to any other item. Id., at 5. Federal funds account for anywhere from 50% to 83% of each State’s total Medicaid expenditures, see §1396d(b) (2006 ed., Supp. IV); most States receive more than $1 billion in federal Medicaid funding; and a quarter receive more than $5 billion, NASBO Report 47. These federal dollars total nearly two thirds—64.6%—of all Medicaid expenditures nationwide. [ 15 ] Id., at 46.
The Court of Appeals concluded that the States failed to establish coercion in this case in part because the “states have the power to tax and raise revenue, and therefore can create and fund programs of their own if they do not like Congress’s terms.” 648 F. 3d 1235, 1268 (CA11 2011); see Brief for Sen. Harry Reid et al. as Amici Curiae in No. 11–400, p. 21 (“States may always choose to decrease expenditures on other programs or to raise revenues”). But the sheer size of this federal spending program in relation to state expenditures means that a State would be very hard pressed to compensate for the loss of federal funds by cutting other spending or raising additional revenue. Arizona, for example, commits 12% of its state expenditures to Medicaid, and relies on the Federal Government to provide the rest: $5.6 billion, equaling roughly one-third of Arizona’s annual state expenditures of $17 billion. See NASBO Report 7, 47. Therefore, if Arizona lost federal Medicaid funding, the State would have to commit an additional 33% of all its state expenditures to fund an equivalent state program along the lines of pre-expansion Medicaid. This means that the State would have to allocate 45% of its annual expenditures for that one purpose. See ibid.
The States are far less reliant on federal funding for any other program. After Medicaid, the next biggest federal funding item is aid to support elementary and secondary education, which amounts to 12.8% of total federal outlays to the States, see id., at 7, 16, and equals only 6.6% of all state expenditures combined. See ibid. In Arizona, for example, although federal Medicaid expenditures are equal to 33% of all state expenditures, federal education funds amount to only 9.8% of all state expenditures. See ibid. And even in States with less than average federal Medicaid funding, that funding is at least twice the size of federal education funding as a percentage of state expenditures. Id., at 7, 16, 47.
A State forced out of the Medicaid program would face burdens in addition to the loss of federal Medicaid funding. For example, a nonparticipating State might be found to be ineligible for other major federal funding sources, such as Temporary Assistance for Needy Families (TANF), which is premised on the expectation that States will participate in Medicaid. See 42 U. S. C. §602(a)(3) (2006 ed.) (requiring that certain beneficiaries of TANF funds be “eligible for medical assistance under the State[’s Medicaid] plan”). And withdrawal or expulsion from the Medicaid program would not relieve a State’s hospitals of their obligation under federal law to provide care for patients who are unable to pay for medical services. The Emer- gency Medical Treatment and Active Labor Act, §1395dd, requires hospitals that receive any federal funding to provide stabilization care for indigent patients but does not offer federal funding to assist facilities in carrying out its mandate. Many of these patients are now covered by Medicaid. If providers could not look to the Medicaid program to pay for this care, they would find it exceed- ingly difficult to comply with federal law unless they were given substantial state support. See, e.g., Brief for Economists as Amici Curiae in No 11–400, p. 11.
For these reasons, the offer that the ACA makes to the States—go along with a dramatic expansion of Medicaid or potentially lose all federal Medicaid funding—is quite unlike anything that we have seen in a prior spending-power case. In South Dakota v. Dole, the total amount that the States would have lost if every single State had refused to comply with the 21-year-old drinking age was approximately $614.7 million—or about 0.19% of all state expenditures combined. See Nat. Assn. of State Budget Officers, 1989 (Fiscal Years 1987– 1989 Data) State Expenditure Report 10, 84 (1989), http://www.nasbo.org/publications-data/state-expenditure-report/archives. South Dakota stood to lose, at most, funding that amounted to less than 1% of its annual state expenditures. See ibid. Under the ACA, by contrast, the Federal Government has threatened to withhold 42.3% of all federal outlays to the states, or approximately $233 billion. See NASBO Report 7, 10, 47. South Dakota stands to lose federal funding equaling 28.9% of its annual state expenditures. See id., at 7, 47. Withholding $614.7 million, equaling only 0.19% of all state expenditures combined, is aptly characterized as “relatively mild encouragement,” but threatening to withhold $233 billion, equaling 21.86% of all state expenditures combined, is a different matter.
What the statistics suggest is confirmed by the goal and structure of the ACA. In crafting the ACA, Congress clearly expressed its informed view that no State could possibly refuse the offer that the ACA extends.
The stated goal of the ACA is near-universal health care coverage. To achieve this goal, the ACA mandates that every person obtain a minimum level of coverage. It attempts to reach this goal in several different ways. The guaranteed issue and community-rating provisions are designed to make qualifying insurance available and affordable for persons with medical conditions that may require expensive care. Other ACA provisions seek to make such policies more affordable for people of modest means. Finally, for low-income individuals who are simply not able to obtain insurance, Congress expanded Medicaid, transforming it from a program covering only members of a limited list of vulnerable groups into a program that provides at least the requisite minimum level of coverage for the poor. See 42 U. S. C. §§1396a(a) (10)(A)(i)(VIII) (2006 ed., Supp. IV), 1396u–7(a), (b)(5), 18022(a). This design was intended to provide at least a specified minimum level of coverage for all Americans, but the achievement of that goal obviously depends on participation by every single State. If any State—not to mention all of the 26 States that brought this suit—chose to decline the federal offer, there would be a gaping hole in the ACA’s coverage.
It is true that some persons who are eligible for Medicaid coverage under the ACA may be able to secure private insurance, either through their employers or by obtain- ing subsidized insurance through an exchange. See 26 U. S. C. §36B(a) (2006 ed., Supp. IV); Brief for Respondents in No. 11–400, at 12. But the new federal subsidies are not available to those whose income is below the federal poverty level, and the ACA provides no means, other than Medicaid, for these individuals to obtain coverage and comply with the Mandate. The Government counters that these people will not have to pay the penalty, see, e.g., Tr. of Oral Arg. 68 (Mar. 28, 2012); Brief for Respondents in No. 11–400, at 49–50, but that argument misses the point: Without Medicaid, these individuals will not have coverage and the ACA’s goal of near-universal coverage will be severely frustrated.
If Congress had thought that States might actually refuse to go along with the expansion of Medicaid, Congress would surely have devised a backup scheme so that the most vulnerable groups in our society, those previously eligible for Medicaid, would not be left out in the cold. But nowhere in the over 900-page Act is such a scheme to be found. By contrast, because Congress thought that some States might decline federal funding for the operation of a “health benefit exchange,” Congress provided a backup scheme; if a State declines to participate in the operation of an exchange, the Federal Government will step in and operate an exchange in that State. See 42 U. S. C. §18041(c)(1). Likewise, knowing that States would not necessarily provide affordable health insurance for aliens lawfully present in the United States—because Medicaid does not require States to provide such coverage—Con- gress extended the availability of the new federal insurance subsidies to all aliens. See 26 U. S. C. §36B(c) (1)(B)(ii) (excepting from the income limit individuals who are “not eligible for the medicaid program . . . by reason of [their] alien status”). Congress did not make these subsidies available for citizens with incomes below the poverty level because Congress obviously assumed that they would be covered by Medicaid. If Congress had contemplated that some of these citizens would be left without Medicaid coverage as a result of a State’s withdrawal or expulsion from the program, Congress surely would have made them eligible for the tax subsidies provided for low-income aliens.
These features of the ACA convey an unmistakable message: Congress never dreamed that any State would refuse to go along with the expansion of Medicaid. Congress well understood that refusal was not a practical option.
The Federal Government does not dispute the inference that Congress anticipated 100% state participation, but it argues that this assumption was based on the fact that ACA’s offer was an “exceedingly generous” gift. Brief for Respondents in No. 11–400, at 50. As the Federal Government sees things, Congress is like the generous benefactor who offers $1 million with few strings attached to 50 randomly selected individuals. Just as this benefactor might assume that all of these 50 individuals would snap up his offer, so Congress assumed that every State would gratefully accept the federal funds (and conditions) to go with the expansion of Medicaid.
This characterization of the ACA’s offer raises obvious questions. If that offer is “exceedingly generous,” as the Federal Government maintains, why have more than half the States brought this lawsuit, contending that the offer is coercive? And why did Congress find it necessary to threaten that any State refusing to accept this “exceed- ingly generous” gift would risk losing all Medicaid funds? Congress could have made just the new funding provided under the ACA contingent on acceptance of the terms of the Medicaid Expansion. Congress took such an approach in some earlier amendments to Medicaid, separating new coverage requirements and funding from the rest of the program so that only new funding was conditioned on new eligibility extensions. See, e.g., Social Security Amendments of 1972, 86Stat. 1465.
Congress’ decision to do otherwise here reflects its understanding that the ACA offer is not an “exceedingly generous” gift that no State in its right mind would decline. Instead, acceptance of the offer will impose very substantial costs on participating States. It is true that the Federal Government will bear most of the initial costs associated with the Medicaid Expansion, first paying 100% of the costs of covering newly eligible individuals between 2014 and 2016. 42 U. S. C. §1396d(y). But that is just part of the picture. Participating States will be forced to shoulder substantial costs as well, because after 2019 the Federal Government will cover only 90% of the costs associated with the Expansion, see ibid., with state spending projected to increase by at least $20 billion by 2020 as a consequence. Statement of Douglas W. Elmendorf, CBO’s Analysis of the Major Health Care Legislation Enacted in March 2010, p. 24 (Mar. 30, 2011); see also R. Bovbjerg, B. Ormond, & V. Chen, Kaiser Commission on Medicaid and the Uninsured, State Budgets under Federal Health Reform: The Extent and Causes of Variations in Estimated Impacts 4, n. 27 (Feb. 2011) (estimating new state spending at $43.2 billion through 2019). After 2019, state spending is expected to increase at a faster rate; the CBO estimates new state spending at $60 billion through 2021. Statement of Douglas W. Elmendorf, supra, at 24. And these costs may increase in the future because of the very real possibility that the Federal Government will change funding terms and reduce the percentage of funds it will cover. This would leave the States to bear an increasingly large percentage of the bill. See Tr. of Oral Arg. 74–76 (Mar. 28, 2012). Finally, after 2015, the States will have to pick up the tab for 50% of all administrative costs associated with implementing the new program, see §§1396b(a)(2)–(5), (7) (2006 ed., Supp. IV), costs that could approach $12 billion between fiscal years 2014 and 2020, see Dept. of Health and Human Services, Center for Medicaid and Medicare Services, 2010 Actuarial Report on the Financial Outlook for Medicaid 30.
In sum, it is perfectly clear from the goal and structure of the ACA that the offer of the Medicaid Expansion was one that Congress understood no State could refuse. The Medicaid Expansion therefore exceeds Congress’ spending power and cannot be implemented.
Seven Members of the Court agree that the Medicaid Expansion, as enacted by Congress, is unconstitutional. See Part IV–A to IV–E, supra; Part IV–A, ante, at 45–55 (opinion of Roberts, C. J., joined by Breyer and Kagan, JJ.). Because the Medicaid Expansion is unconstitutional, the question of remedy arises. The most natural remedy would be to invalidate the Medicaid Expansion. However, the Government proposes—in two cursory sentences at the very end of its brief—preserving the Expansion. Under its proposal, States would receive the additional Medi- caid funds if they expand eligibility, but States would keep their pre-existing Medicaid funds if they do not expand eligibility. We cannot accept the Government’s suggestion.
The reality that States were given no real choice but to expand Medicaid was not an accident. Congress assumed States would have no choice, and the ACA depends on States’ having no choice, because its Mandate requires low-income individuals to obtain insurance many of them can afford only through the Medicaid Expansion. Furthermore, a State’s withdrawal might subject everyone in the State to much higher insurance premiums. That is because the Medicaid Expansion will no longer offset the cost to the insurance industry imposed by the ACA’s insurance regulations and taxes, a point that is explained in more detail in the severability section below. To make the Medicaid Expansion optional despite the ACA’s structure and design “ ‘would be to make a new law, not to enforce an old one. This is no part of our duty.’ ” Trade-Mark Cases, 100 U. S. 82, 99 (1879) .
Worse, the Government’s proposed remedy introduces a new dynamic: States must choose between expanding Medicaid or paying huge tax sums to the federal fisc for the sole benefit of expanding Medicaid in other States. If this divisive dynamic between and among States can be introduced at all, it should be by conscious congressional choice, not by Court-invented interpretation. We do not doubt that States are capable of making decisions when put in a tight spot. We do doubt the authority of this Court to put them there.
The Government cites a severability clause codified with Medicaid in Chapter 7 of the United States Code stating that if “any provision of this chapter, or the application thereof to any person or circumstance, is held invalid, the remainder of the chapter, and the application of such provision to other persons or circumstances shall not be affected thereby.” 42 U. S. C. §1303 (2006 ed.). But that clause tells us only that other provisions in Chapter 7 should not be invalidated if §1396c, the authorization for the cut-off of all Medicaid funds, is unconstitutional. It does not tell us that §1396c can be judicially revised, to say what it does not say. Such a judicial power would not be called the doctrine of severability but perhaps the doctrine of amendatory invalidation—similar to the amendatory veto that permits the Governors of some States to reduce the amounts appropriated in legislation. The proof that such a power does not exist is the fact that it would not preserve other congressional dispositions, but would leave it up to the Court what the “validated” legislation will contain. The Court today opts for permitting the cut-off of only incremental Medicaid funding, but it might just as well have permitted, say, the cut-off of funds that represent no more than x percent of the State’s bud- get. The Court severs nothing, but simply revises §1396c to read as the Court would desire.
We should not accept the Government’s invitation to attempt to solve a constitutional problem by rewriting the Medicaid Expansion so as to allow States that reject it to retain their pre-existing Medicaid funds. Worse, the Government’s remedy, now adopted by the Court, takes the ACA and this Nation in a new direction and charts a course for federalism that the Court, not the Congress, has chosen; but under the Constitution, that power and authority do not rest with this Court.
The Affordable Care Act seeks to achieve “near-universal” health insurance coverage. §18091(2)(D) (2006 ed., Supp. IV). The two pillars of the Act are the Individ- ual Mandate and the expansion of coverage under Medicaid. In our view, both these central provisions of the Act—the Individual Mandate and Medicaid Expansion—are invalid. It follows, as some of the parties urge, that all other provisions of the Act must fall as well. The following section explains the severability principles that require this conclusion. This analysis also shows how closely interrelated the Act is, and this is all the more reason why it is judicial usurpation to impose an entirely new mechanism for withdrawal of Medicaid funding, see Part IV–F, supra, which is one of many examples of how rewriting the Act alters its dynamics.
When an unconstitutional provision is but a part of a more comprehensive statute, the question arises as to the validity of the remaining provisions. The Court’s authority to declare a statute partially unconstitutional has been well established since Marbury v. Madison, 1 Cranch 137 (1803), when the Court severed an unconstitutional provision from the Judiciary Act of 1789. And while the Court has sometimes applied “at least a modest presumption in favor of . . . severability,” C. Nelson, Statutory Interpretation 144 (2010), it has not always done so, see, e.g., Minnesota v. Mille Lacs Band of Chippewa Indians, 526 U. S. 172 –195 (1999).
An automatic or too cursory severance of statutory provisions risks “rewrit[ing] a statute and giv[ing] it an effect altogether different from that sought by the measure viewed as a whole.” Railroad Retirement Bd. v. Alton R. Co., 295 U. S. 330, 362 (1935) . The Judiciary, if it orders uncritical severance, then assumes the legislative function; for it imposes on the Nation, by the Court’s decree, its own new statutory regime, consisting of policies, risks, and duties that Congress did not enact. That can be a more extreme exercise of the judicial power than striking the whole statute and allowing Congress to address the conditions that pertained when the statute was considered at the outset.
The Court has applied a two-part guide as the framework for severability analysis. The test has been deemed “well established.” Alaska Airlines, Inc. v. Brock, 480 U. S. 678, 684 (1987) . First, if the Court holds a statutory provision unconstitutional, it then determines whether the now truncated statute will operate in the manner Con- gress intended. If not, the remaining provisions must be invalidated. See id., at 685. In Alaska Airlines, the Court clarified that this first inquiry requires more than ask- ing whether “the balance of the legislation is incapable of functioning independently.” Id., at 684. Even if the remaining provisions will operate in some coherent way, that alone does not save the statute. The question is whether the provisions will work as Congress intended. The “relevant inquiry in evaluating severability is whether the statute will function in a manner consistent with the intent of Congress.” Id., at 685 (emphasis in original). See also Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. ___, ___ (2010) (slip op., at 28) (the Act “remains fully operative as a law with these tenure restrictions excised”) (internal quotation marks omitted); United States v. Booker, 543 U. S. 220, 227 (2005) (“[T]wo provisions . . . must be invalidated in order to allow the statute to operate in a manner consistent with congressional intent”); Mille Lacs, supra, at 194 (“[E]m- bodying as it did one coherent policy, [the entire order] is inseverable”).
Second, even if the remaining provisions can operate as Congress designed them to operate, the Court must determine if Congress would have enacted them standing alone and without the unconstitutional portion. If Congress would not, those provisions, too, must be invalidated. See Alaska Airlines, supra, at 685 (“[T]he unconstitu- tional provision must be severed unless the statute cre- ated in its absence is legislation that Congress would not have enacted”); see also Free Enterprise Fund, supra, at ___ (slip op., at 29) (“[N]othing in the statute’s text or historical context makes it ‘evident’ that Congress, faced with the limitations imposed by the Constitution, would have preferred no Board at all to a Board whose members are removable at will”); Ayotte v. Planned Parenthood of Northern New Eng., 546 U. S. 320, 330 (2006) (“Would the legislature have preferred what is left of its statute to no statute at all”); Denver Area Ed. Telecommunications Consortium, Inc. v. FCC, 518 U. S. 727, 767 (1996) (plurality opinion) (“Would Congress still have passed §10(a) had it known that the remaining provisions were invalid” (internal quotation marks and brackets omitted)).
The two inquiries—whether the remaining provisions will operate as Congress designed them, and whether Congress would have enacted the remaining provisions standing alone—often are interrelated. In the ordinary course, if the remaining provisions cannot operate according to the congressional design (the first inquiry), it almost necessarily follows that Congress would not have enacted them (the second inquiry). This close interaction may explain why the Court has not always been precise in distinguishing between the two. There are, however, occasions in which the severability standard’s first inquiry (statutory functionality) is not a proxy for the second inquiry (whether the Legislature intended the remaining provisions to stand alone).
The Act was passed to enable affordable, “near-universal” health insurance coverage. 42 U. S. C. §18091(2)(D). The resulting, complex statute consists of mandates and other requirements; comprehensive regulation and penalties; some undoubted taxes; and increases in some governmental expenditures, decreases in others. Under the severability test set out above, it must be determined if those provisions function in a coherent way and as Congress would have intended, even when the major provisions establishing the Individual Mandate and Medicaid Expansion are themselves invalid.
Congress did not intend to establish the goal of near-universal coverage without regard to fiscal consequences. See, e.g., ACA §1563, 124Stat. 270 (“[T]his Act will reduce the Federal deficit between 2010 and 2019”). And it did not intend to impose the inevitable costs on any one industry or group of individuals. The whole design of the Act is to balance the costs and benefits affecting each set of regulated parties. Thus, individuals are required to obtain health insurance. See 26 U. S. C. §5000A(a). Insur- ance companies are required to sell them insurance regardless of patients’ pre-existing conditions and to comply with a host of other regulations. And the companies must pay new taxes. See §4980I (high-cost insurance plans); 42 U. S. C. §§300gg(a)(1), 300gg–4(b) (community rating); §§300gg–1, 300gg–3, 300gg–4(a) (guaranteed issue); §300gg–11 (elimination of coverage limits); §300gg–14(a) (dependent children up to age 26); ACA §§9010, 10905, 124Stat. 865, 1017 (excise tax); Health Care and Education Reconciliation Act of 2010 (HCERA) §1401, 124Stat. 1059 (excise tax). States are expected to expand Medicaid eligibility and to create regulated marketplaces called ex- changes where individuals can purchase insurance. See 42 U. S. C. §§1396a(a)(10)(A)(i)(VIII) (2006 ed., Supp. IV) (Medicaid Expansion), 18031 (exchanges). Some persons who cannot afford insurance are provided it through the Medicaid Expansion, and others are aided in their purchase of insurance through federal subsidies available on health-insurance exchanges. See 26 U. S. C. §36B (2006 ed., Supp. IV), 42 U. S. C. §18071 (2006 ed., Supp. IV) (federal subsidies). The Federal Government’s increased spending is offset by new taxes and cuts in other federal expenditures, including reductions in Medicare and in federal payments to hospitals. See, e.g., §1395ww(r) (Medicare cuts); ACA Title IX, Subtitle A, 124Stat. 847 (“Rev- enue Offset Provisions”). Employers with at least 50 employees must either provide employees with adequate health benefits or pay a financial exaction if an employee who qualifies for federal subsidies purchases insurance through an exchange. See 26 U. S. C. §4980H (2006 ed., Supp. IV).
In short, the Act attempts to achieve near-universal health insurance coverage by spreading its costs to individuals, insurers, governments, hospitals, and employers—while, at the same time, offsetting significant portions of those costs with new benefits to each group. For ex- ample, the Federal Government bears the burden of paying billions for the new entitlements mandated by the Medicaid Expansion and federal subsidies for insurance purchases on the exchanges; but it benefits from reductions in the reimbursements it pays to hospitals. Hospitals lose those reimbursements; but they benefit from the decrease in uncompensated care, for under the insurance regulations it is easier for individuals with pre-existing conditions to purchase coverage that increases payments to hospitals. Insurance companies bear new costs imposed by a collection of insurance regulations and taxes, including “guaranteed issue” and “community rating” requirements to give coverage regardless of the insured’s pre-existing conditions; but the insurers benefit from the new, healthy purchasers who are forced by the Individual Mandate to buy the insurers’ product and from the new low- income Medicaid recipients who will enroll in insurance companies’ Medicaid-funded managed care programs. In summary, the Individual Mandate and Medicaid Expansion offset insurance regulations and taxes, which offset reduced reimbursements to hospitals, which offset increases in federal spending. So, the Act’s major provisions are interdependent.
The Act then refers to these interdependencies as “shared responsibility.” See ACA Subtitle F, Title I, 124Stat. 242 (“Shared Responsibility”); ACA §1501, ibid. (same); ACA §1513, id., at 253 (same); ACA §4980H, ibid. (same). In at least six places, the Act describes the Individual Mandate as working “together with the other pro- visions of this Act.” 42 U. S. C. §18091(2)(C) (2006 ed., Supp. IV) (working “together” to “add millions of new consumers to the health insurance market”); §18091(2)(E) (working “together” to “significantly reduce” the economic cost of the poorer health and shorter lifespan of the uninsured); §18091(2)(F) (working “together” to “lower health insurance premiums”); §18091(2)(G) (working “together” to “improve financial security for families”); §18091(2)(I) (working “together” to minimize “adverse selection and broaden the health insurance risk pool to include healthy individuals”); §18091(2)(J) (working “together” to “signif- icantly reduce administrative costs and lower health insurance premiums”). The Act calls the Individual Mandate “an essential part” of federal regulation of health insurance and warns that “the absence of the requirement would undercut Federal regulation of the health insurance market.” §18091(2)(H).
One preliminary point should be noted before applying severability principles to the Act. To be sure, an argument can be made that those portions of the Act that none of the parties has standing to challenge cannot be held nonseverable. The response to this argument is that our cases do not support it. See, e.g., Williams v. Standard Oil Co. of La., 278 U. S. 235 –244 (1929) (holding nonseverable statutory provisions that did not burden the parties). It would be particularly destructive of sound government to apply such a rule with regard to a multifaceted piece of legislation like the ACA. It would take years, perhaps decades, for each of its provisions to be adjudicated separately—and for some of them (those simply expending federal funds) no one may have separate standing. The Federal Government, the States, and private parties ought to know at once whether the entire legislation fails.
The opinion now explains in Part V–C–1, infra, why the Act’s major provisions are not severable from the Mandate and Medicaid Expansion. It proceeds from the insurance regulations and taxes (C–1–a), to the reductions in reimbursements to hospitals and other Medicare reductions (C–1–b), the exchanges and their federal subsidies (C–1–c), and the employer responsibility assessment (C–1–d). Part V–C–2, infra, explains why the Act’s minor provisions also are not severable.
The Act’s Major Provisions
Major provisions of the Affordable Care Act—i.e., the insurance regulations and taxes, the reductions in federal reimbursements to hospitals and other Medicare spend- ing reductions, the exchanges and their federal subsidies, and the employer responsibility assessment—cannot remain once the Individual Mandate and Medicaid Expansion are invalid. That result follows from the undoubted inability of the other major provisions to operate as Congress intended without the Individual Mandate and Medicaid Expansion. Absent the invalid portions, the other major provisions could impose enormous risks of unexpected bur- dens on patients, the health-care community, and the federal budget. That consequence would be in absolute conflict with the ACA’s design of “shared responsibility,” and would pose a threat to the Nation that Congress did not intend.
Insurance Regulations and Taxes
Without the Individual Mandate and Medicaid Expansion, the Affordable Care Act’s insurance regulations and insurance taxes impose risks on insurance companies and their customers that this Court cannot measure. Those risks would undermine Congress’ scheme of “shared responsibility.” See 26 U. S. C. §4980I (2006 ed., Supp. IV) (high-cost insurance plans); 42 U. S. C. §§300gg(a)(1) (2006 ed., Supp. IV), 300gg–4(b) (community rating); §§300gg–1, 300gg–3, 300gg–4(a) (guaranteed issue); §300gg–11 (elimination of coverage limits); §300gg–14(a) (dependent children up to age 26); ACA §§9010, 10905, 124Stat. 865, 1017 (excise tax); HCERA §1401, 124Stat. 1059 (excise tax).
The Court has been informed by distinguished economists that the Act’s Individual Mandate and Medicaid Expansion would each increase revenues to the insurance industry by about $350 billion over 10 years; that this combined figure of $700 billion is necessary to offset the approximately $700 billion in new costs to the insurance industry imposed by the Act’s insurance regulations and taxes; and that the new $700-billion burden would otherwise dwarf the industry’s current profit margin. See Brief for Economists as Amici Curiae in No. 11–393 etc. (Severability), pp. 9–16, 10a.
If that analysis is correct, the regulations and taxes will mean higher costs for insurance companies. Higher costs may mean higher premiums for consumers, despite the Act’s goal of “lower[ing] health insurance premiums.” 42 U. S. C. §18091(2)(F) (2006 ed., Supp. IV). Higher costs also could threaten the survival of health-insurance companies, despite the Act’s goal of “effective health insurance markets.” §18091(2)(J).
The actual cost of the regulations and taxes may be more or less than predicted. What is known, however, is that severing other provisions from the Individual Mandate and Medicaid Expansion necessarily would impose significant risks and real uncertainties on insurance companies, their customers, all other major actors in the sys- tem, and the government treasury. And what also is known is this: Unnecessary risks and avoidable uncertainties are hostile to economic progress and fiscal stability and thus to the safety and welfare of the Nation and the Nation’s freedom. If those risks and uncertainties are to be imposed, it must not be by the Judiciary.
Reductions in Reimbursements to Hospitals and Other Reductions in Medicare Expenditures
The Affordable Care Act reduces payments by the Federal Government to hospitals by more than $200 billion over 10 years. See 42 U. S. C. §1395ww(b)(3)(B)(xi)–(xii) (2006 ed., Supp. IV); §1395ww(q); §1395ww(r); §1396r–4(f)(7).
The concept is straightforward: Near-universal coverage will reduce uncompensated care, which will increase hospitals’ revenues, which will offset the government’s re- ductions in Medicare and Medicaid reimbursements to hospitals. Responsibility will be shared, as burdens and benefits balance each other. This is typical of the whole dynamic of the Act.
Invalidating the key mechanisms for expanding insurance coverage, such as community rating and the Medicaid Expansion, without invalidating the reductions in Medicare and Medicaid, distorts the ACA’s design of “shared responsibility.” Some hospitals may be forced to raise the cost of care in order to offset the reductions in reimbursements, which could raise the cost of insurance premiums, in contravention of the Act’s goal of “lower[ing] health insurance premiums.” 42 U. S. C. §18091(2)(F) (2006 ed., Supp. IV). See also §18091(2)(I) (goal of “lower[ing] health insurance premiums”); §18091(2)(J) (same). Other hospitals, particularly safety-net hospitals that serve a large number of uninsured patients, may be forced to shut down. Cf. National Assn. of Public Hospitals, 2009 Annual Survey: Safety Net Hospitals and Health Systems Fulfill Mission in Uncertain Times 5–6 (Feb. 2011). Like the effect of preserving the insurance regulations and taxes, the precise degree of risk to hospitals is unknowable. It is not the proper role of the Court, by severing part of a statute and allowing the rest to stand, to impose unknowable risks that Congress could neither measure nor predict. And Congress could not have intended that result in any event.
There is a second, independent reason why the reductions in reimbursements to hospitals and the ACA’s other Medicare cuts must be invalidated. The ACA’s $455 billion in Medicare and Medicaid savings offset the $434-billion cost of the Medicaid Expansion. See CBO Estimate, Table 2 (Mar. 20, 2010). The reductions allowed Congress to find that the ACA “will reduce the Federal deficit between 2010 and 2019” and “will continue to reduce budget deficits after 2019.” ACA §§1563(a)(1), (2), 124Stat. 270.
That finding was critical to the ACA. The Act’s “shared responsibility” concept extends to the federal budget. Congress chose to offset new federal expenditures with budget cuts and tax increases. That is why the United States has explained in the course of this litigation that “[w]hen Congress passed the ACA, it was careful to ensure that any increased spending, including on Medicaid, was offset by other revenue-raising and cost-saving provisions.” Memorandum in Support of Government’s Motion for Summary Judgment in No. 3–10–cv–91, p. 41.
If the Medicare and Medicaid reductions would no longer be needed to offset the costs of the Medicaid Expansion, the reductions would no longer operate in the manner Congress intended. They would lose their justification and foundation. In addition, to preserve them would be “to eliminate a significant quid pro quo of the legislative com- promise” and create a statute Congress did not enact. Legal Services Corporation v. Velazquez, 531 U. S. 533, 561 (2001) (Scalia, J., dissenting). It is no secret that cutting Medicare is unpopular; and it is most improbable Congress would have done so without at least the assurance that it would render the ACA deficit-neutral. See ACA §§1563(a)(1), (2), 124Stat. 270.
Health Insurance Exchanges and Their Federal Subsidies
The ACA requires each State to establish a health-insurance “exchange.” Each exchange is a one-stop marketplace for individuals and small businesses to compare community-rated health insurance and purchase the policy of their choice. The exchanges cannot operate in the manner Congress intended if the Individual Mandate, Medicaid Expansion, and insurance regulations cannot remain in force.
The Act’s design is to allocate billions of federal dollars to subsidize individuals’ purchases on the exchanges. In- dividuals with incomes between 100 and 400 percent of the poverty level receive tax credits to offset the cost of insurance to the individual purchaser. 26 U. S. C. §36B (2006 ed., Supp. IV); 42 U. S. C. §18071 (2006 ed., Supp. IV). By 2019, 20 million of the 24 million people who will obtain insurance through an exchange are expected to receive an average federal subsidy of $6,460 per person. See CBO, Analysis of the Major Health Care Legislation Enacted in March 2010, pp. 18–19 (Mar. 30, 2011). Without the community-rating insurance regulation, however, the average federal subsidy could be much higher; for community rating greatly lowers the enormous premiums unhealthy individuals would otherwise pay. Federal subsidies would make up much of the difference.
The result would be an unintended boon to insurance companies, an unintended harm to the federal fisc, and a corresponding breakdown of the “shared responsibil- ity” between the industry and the federal budget that Congress intended. Thus, the federal subsidies must be invalidated.
In the absence of federal subsidies to purchasers, insurance companies will have little incentive to sell insurance on the exchanges. Under the ACA’s scheme, few, if any, individuals would want to buy individual insurance policies outside of an exchange, because federal subsidies would be unavailable outside of an exchange. Difficulty in attracting individuals outside of the exchange would in turn motivate insurers to enter exchanges, despite the exchanges’ onerous regulations. See 42 U. S. C. §18031. That system of incentives collapses if the federal subsidies are invalidated. Without the federal subsidies, individ- uals would lose the main incentive to purchase insurance inside the exchanges, and some insurers may be unwilling to offer insurance inside of exchanges. With fewer buyers and even fewer sellers, the exchanges would not operate as Congress intended and may not operate at all.
There is a second reason why, if community rating is invalidated by the Mandate and Medicaid Expansion’s invalidity, exchanges cannot be implemented in a manner consistent with the Act’s design. A key purpose of an exchange is to provide a marketplace of insurance options where prices are standardized regardless of the buy- er’s pre-existing conditions. See ibid. An individual who shops for insurance through an exchange will evaluate different insurance products. The products will offer different benefits and prices. Congress designed the exchanges so the shopper can compare benefits and prices. But the comparison cannot be made in the way Congress designed if the prices depend on the shopper’s pre-existing health conditions. The prices would vary from person to person. So without community rating—which prohibits insurers from basing the price of insurance on pre-existing conditions—the exchanges cannot operate in the manner Congress intended.
The employer responsibility assessment provides an incentive for employers with at least 50 employees to provide their employees with health insurance options that meet minimum criteria. See 26 U. S. C. §4980H (2006 ed., Supp. IV). Unlike the Individual Mandate, the employer-responsibility assessment does not require employers to provide an insurance option. Instead, it re- quires them to make a payment to the Federal Government if they do not offer insurance to employees and if insurance is bought on an exchange by an employee who qualifies for the exchange’s federal subsidies. See ibid.
For two reasons, the employer-responsibility assessment must be invalidated. First, the ACA makes a direct link between the employer-responsibility assessment and the exchanges. The financial assessment against employers occurs only under certain conditions. One of them is the purchase of insurance by an employee on an exchange. With no exchanges, there are no purchases on the exchanges; and with no purchases on the exchanges, there is nothing to trigger the employer-responsibility assessment.
Second, after the invalidation of burdens on individuals (the Individual Mandate), insurers (the insurance regu- lations and taxes), States (the Medicaid Expansion), the Federal Government (the federal subsidies for exchanges and for the Medicaid Expansion), and hospitals (the reductions in reimbursements), the preservation of the employer-responsibility assessment would upset the ACA’s design of “shared responsibility.” It would leave employers as the only parties bearing any significant responsibility. That was not the congressional intent.
The Act’s Minor Provisions
The next question is whether the invalidation of the ACA’s major provisions requires the Court to invalidate the ACA’s other provisions. It does.
The ACA is over 900 pages long. Its regulations include requirements ranging from a break time and secluded place at work for nursing mothers, see 29 U. S. C. §207(r)(1) (2006 ed., Supp. IV), to displays of nutritional content at chain restaurants, see 21 U. S. C. §343(q)(5)(H). The Act raises billions of dollars in taxes and fees, including exactions imposed on high-income taxpayers, see ACA §§9015, 10906; HCERA §1402, medical devices, see 26 U. S. C. §4191 (2006 ed., Supp. IV), and tanning booths, see §5000B. It spends government money on, among other things, the study of how to spend less government money. 42 U. S. C. §1315a. And it includes a number of provisions that provide benefits to the State of a particular legislator. For example, §10323, 124Stat. 954, extends Medicare coverage to individuals exposed to asbestos from a mine in Libby, Montana. Another provision, §2006, id., at 284, increases Medicaid payments only in Louisiana.
Such provisions validate the Senate Majority Leader’s statement, “ ‘I don’t know if there is a senator that doesn’t have something in this bill that was important to them. . . . [And] if they don’t have something in it important to them, then it doesn’t speak well of them. That’s what this legislation is all about: It’s the art of compromise.’ ” Pear, In Health Bill for Everyone, Provisions for a Few, N. Y. Times, Jan. 4, 2010, p. A10 (quoting Sen. Reid). Often, a minor provision will be the price paid for support of a major provision. So, if the major provision were unconstitutional, Congress would not have passed the minor one.
Without the ACA’s major provisions, many of these minor provisions will not operate in the manner Congress intended. For example, the tax increases are “Revenue Offset Provisions” designed to help offset the cost to the Federal Government of programs like the Medicaid Expansion and the exchanges’ federal subsidies. See Title IX, Subtitle A—Revenue Offset Provisions, 124Stat. 847. With the Medicaid Expansion and the exchanges invalidated, the tax increases no longer operate to offset costs, and they no longer serve the purpose in the Act’s scheme of “shared responsibility” that Congress intended.
Some provisions, such as requiring chain restaurants to display nutritional content, appear likely to operate as Congress intended, but they fail the second test for severability. There is no reason to believe that Congress would have enacted them independently. The Court has not previously had occasion to consider severability in the con- text of an omnibus enactment like the ACA, which includes not only many provisions that are ancillary to its central provisions but also many that are entirely unre- lated—hitched on because it was a quick way to get them passed despite opposition, or because their proponents could exact their enactment as the quid pro quo for their needed support. When we are confronted with such a so-called “Christmas tree,” a law to which many nongermane ornaments have been attached, we think the proper rule must be that when the tree no longer exists the ornaments are superfluous. We have no reliable basis for knowing which pieces of the Act would have passed on their own. It is certain that many of them would not have, and it is not a proper function of this Court to guess which. To sever the statute in that manner “ ‘would be to make a new law, not to enforce an old one. This is not part of our duty.’ ” Trade-Mark Cases, 100 U. S., at 99.
This Court must not impose risks unintended by Congress or produce legislation Congress may have lacked the support to enact. For those reasons, the unconstitutionality of both the Individual Mandate and the Medicaid Expansion requires the invalidation of the Affordable Care Act’s other provisions.
* * *
The Court today decides to save a statute Congress did not write. It rules that what the statute declares to be a requirement with a penalty is instead an option subject to a tax. And it changes the intentionally coercive sanction of a total cut-off of Medicaid funds to a supposedly noncoercive cut-off of only the incremental funds that the Act makes available.
The Court regards its strained statutory interpretation as judicial modesty. It is not. It amounts instead to a vast judicial overreaching. It creates a debilitated, inoperable version of health-care regulation that Congress did not enact and the public does not expect. It makes enactment of sensible health-care regulation more difficult, since Congress cannot start afresh but must take as its point of departure a jumble of now senseless provisions, provisions that certain interests favored under the Court’s new design will struggle to retain. And it leaves the public and the States to expend vast sums of money on requirements that may or may not survive the necessary congressional revision.
The Court’s disposition, invented and atextual as it is, does not even have the merit of avoiding constitutional difficulties. It creates them. The holding that the Individual Mandate is a tax raises a difficult constitutional question (what is a direct tax?) that the Court resolves with inadequate deliberation. And the judgment on the Medicaid Expansion issue ushers in new federalism concerns and places an unaccustomed strain upon the Union. Those States that decline the Medicaid Expansion must subsidize, by the federal tax dollars taken from their citizens, vast grants to the States that accept the Medicaid Expansion. If that destabilizing political dynamic, so antagonistic to a harmonious Union, is to be introduced at all, it should be by Congress, not by the Judiciary.
The values that should have determined our course today are caution, minimalism, and the understanding that the Federal Government is one of limited powers. But the Court’s ruling undermines those values at every turn. In the name of restraint, it overreaches. In the name of constitutional avoidance, it creates new constitutional questions. In the name of cooperative federalism, it undermines state sovereignty.
The Constitution, though it dates from the founding of the Republic, has powerful meaning and vital relevance to our own times. The constitutional protections that this case involves are protections of structure. Structural protections—notably, the restraints imposed by federalism and separation of powers—are less romantic and have less obvious a connection to personal freedom than the provisions of the Bill of Rights or the Civil War Amendments. Hence they tend to be undervalued or even forgotten by our citizens. It should be the responsibility of the Court to teach otherwise, to remind our people that the Framers considered structural protections of freedom the most im- portant ones, for which reason they alone were embod- ied in the original Constitution and not left to later amendment. The fragmentation of power produced by the structure of our Government is central to liberty, and when we destroy it, we place liberty at peril. Today’s decision should have vindicated, should have taught, this truth; instead, our judgment today has disregarded it.
For the reasons here stated, we would find the Act invalid in its entirety. We respectfully dissent.