National Federation of Independent Business v. Sebelius (redux)

[There is no single vote in this case, and any number you attach to it will misdescribe something. The line-up differs part by part, and here it is the first thing to learn. Per the syllabus at 567 U. S. 524:

Roberts, C. J., announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II, and III–C, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined; an opinion with respect to Part IV, in which Breyer and Kagan, JJ., joined; and an opinion with respect to Parts III–A, III–B, and III–D — which no other Justice joined. Ginsburg, J., filed an opinion concurring in part, concurring in the judgment in part, and dissenting in part, in which Sotomayor, J., joined, and in which Breyer and Kagan, JJ., joined as to Parts I, II, III, and IV. Scalia, Kennedy, Thomas, and Alito, JJ., filed a joint dissenting opinion. Thomas, J., filed a dissenting opinion.

So, for the three passages this reading turns on:

• Part III–B, the saving construction — the Chief Justice alone. No other Justice joined it. The reporter’s running head reads “Opinion of Roberts, C. J.” Nothing in it is a holding of the Court.
• Part III–C, the taxing-power holding — the opinion of the Court, five Justices. The running head reads “Opinion of the Court.” This is the holding that sustained the individual mandate.
• Part IV, the Medicaid holding — a three-Justice plurality, Roberts, C. J., with Breyer and Kagan, JJ., only. The running head reads “Opinion of Roberts, C. J.” The Medicaid result commanded seven votes, because the four joint dissenters separately found the expansion unconstitutional; the remedy in Part IV–B drew Ginsburg and Sotomayor, JJ., concurring in the judgment. Part IV itself is not an opinion of the Court.

Every passage reproduced below carries a marker naming which of these it comes from. A student who cannot tell III–B from III–C has learned the case backwards.

This is the second of two readings in this case. The Commerce Clause and Necessary and Proper Clause portions — Part III–A, and Justice Ginsburg’s answer to it — are a separate reading in Module 3, at /conlaw/cases/national-federation-of-independent-business-v-sebelius. They are not repeated here. This reading carries the taxing power and Medicaid only.

Also omitted: Part II (the Anti-Injunction Act holding); the joint dissent of Scalia, Kennedy, Thomas and Alito, JJ., except for one passage counting the Medicaid votes; and Thomas, J., dissenting. Footnotes are omitted except those set out and labeled. Acquisition gaps — reporter pages that did not come back from the source and are therefore unavailable to this edition — are marked separately, by page number, wherever they fall inside material used here. Bracketed italics are editorial; everything else is verbatim.]

Part III–B — the saving construction

[Chief Justice Roberts, writing alone. Joined by no other Justice. Not an opinion of the Court. Running head: “Opinion of Roberts, C. J.” Reporter pp. 561–563. The syllabus verb is telling: Chief Justice Roberts “concluded in Part III–B,” not “delivered the opinion of the Court.”]

B

That is not the end of the matter. Because the Commerce Clause does not support the individual mandate, it is necessary to turn to the Government’s second argument: that the mandate may be upheld as within Congress’s enumerated power to “lay and collect Taxes.” Art. I, § 8, cl. 1.

The Government’s tax power argument asks us to view the statute differently than we did in considering its commerce power theory. In making its Commerce Clause argument, the Government defended the mandate as a regulation requiring individuals to purchase health insurance. The Government does not claim that the taxing power allows Congress to issue such a command. Instead, it asks us to read the mandate not as ordering individuals to buy insurance, but rather as imposing a tax on those who do not buy that product.

The text of a statute can sometimes have more than one possible meaning. To take a familiar example, a law that reads “no vehicles in the park” might, or might not, ban bicycles in the park. And it is well established that if a statute has two possible meanings, one of which violates the Constitution, courts should adopt the meaning that does not do so. Justice Story said that 180 years ago: “No court ought, unless the terms of an act rendered it unavoidable, to give a construction to it which should involve a violation, however unintentional, of the constitution.” Parsons v. Bedford, 3 Pet. 433, 448–449 (1830).

The most straightforward reading of the mandate is that it commands individuals to purchase insurance. After all, it states that individuals “shall” maintain health insurance. Congress thought it could enact such a command under the Commerce Clause, and the Government primarily defended the law on that basis. But, for the reasons explained above, the Commerce Clause does not give Congress that power. Under our precedent, it is therefore necessary to ask whether the Government’s alternative reading of the statute—that it only imposes a tax on those without insurance—is a reasonable one.

Under the mandate, if an individual does not maintain health insurance, the only consequence is that he must make an additional payment to the IRS when he pays his taxes. That, according to the Government, means the mandate can be regarded as establishing a condition—not owning health insurance—that triggers a tax. Under that theory, the mandate is not a legal command to buy insurance. Rather, it makes going without insurance just another thing the Government taxes, like buying gasoline or earning income. And if the mandate is in effect just a tax hike on certain taxpayers who do not have health insurance, it may be within Congress’s constitutional power to tax.

The question is not whether that is the most natural interpretation of the mandate, but only whether it is a “fairly possible” one. Crowell v. Benson, 285 U. S. 22, 62 (1932). As we have explained, “every reasonable construction must be resorted to, in order to save a statute from unconstitutionality.” Hooper v. California, 155 U. S. 648, 657 (1895). The Government asks us to interpret the mandate as imposing a tax, if it would otherwise violate the Constitution. Granting the Act the full measure of deference owed to federal statutes, it can be so read, for the reasons set forth below.

[Still Roberts, C. J., alone — Part III–D, at pp. 574–575, restates the move and is worth reading with III–B:]

Justice Ginsburg questions the necessity of rejecting the Government’s commerce power argument, given that § 5000A can be upheld under the taxing power. But the statute reads more naturally as a command to buy insurance than as a tax, and I would uphold it as a command if the Constitution allowed it. It is only because the Commerce Clause does not authorize such a command that it is necessary to reach the taxing power question. And it is only because we have a duty to construe a statute to save it, if fairly possible, that § 5000A can be interpreted as a tax.

The Federal Government does not have the power to order people to buy health insurance. Section 5000A would therefore be unconstitutional if read as a command. The Federal Government does have the power to impose a tax on those without health insurance. Section 5000A is therefore constitutional, because it can reasonably be read as a tax.

Part III–C — the taxing power: the opinion of the Court

[Opinion of the Court. Roberts, C. J., joined by Ginsburg, Breyer, Sotomayor and Kagan, JJ. Running head: “Opinion of the Court.” Reporter pp. 563–574. This, and not Part III–A or Part III–B, is the holding that sustained the individual mandate.]

C

The exaction the Affordable Care Act imposes on those without health insurance looks like a tax in many respects. The “[s]hared responsibility payment,” as the statute entitles it, is paid into the Treasury by “taxpayer[s]” when they file their tax returns. It does not apply to individuals who do not pay federal income taxes because their household income is less than the filing threshold in the Internal Revenue Code. For taxpayers who do owe the payment, its amount is determined by such familiar factors as taxable income, number of dependents, and joint filing status. The requirement to pay is found in the Internal Revenue Code and enforced by the IRS, which must assess and collect it “in the same manner as taxes.” This process yields the essential feature of any tax: It produces at least some revenue for the Government — about $4 billion per year by 2017.

It is of course true that the Act describes the payment as a “penalty,” not a “tax.” But while that label is fatal to the application of the Anti-Injunction Act, it does not determine whether the payment may be viewed as an exercise of Congress’s taxing power. It is up to Congress whether to apply the Anti-Injunction Act to any particular statute, so it makes sense to be guided by Congress’s choice of label on that question. That choice does not, however, control whether an exaction is within Congress’s constitutional power to tax.

Our precedent reflects this: In 1922, we decided two challenges to the “Child Labor Tax” on the same day. In the first, we held that a suit to enjoin collection of the so-called tax was barred by the Anti-Injunction Act. In the second case, however, we held that the same exaction, although labeled a tax, was not in fact authorized by Congress’s taxing power. Drexel Furniture, 259 U. S., at 38. That constitutional question was not controlled by Congress’s choice of label. We have similarly held that exactions not labeled taxes nonetheless were authorized by Congress’s power to tax. [The Court cites the License Tax Cases and the nuclear-waste surcharge upheld in New York v. United States.]* We thus ask whether the shared responsibility payment falls within Congress’s taxing power, “[d]isregarding the designation of the exaction, and viewing its substance and application.” United States v. Constantine, 296 U. S. 287, 294 (1935); cf. Quill Corp. v. North Dakota, 504 U. S. 298, 310 (1992) (“[M]agic words or labels” should not “disable an otherwise constitutional levy”); United States v. Sotelo, 436 U. S. 268, 275 (1978) (“That the funds due are referred to as a ‘penalty’ … does not alter their essential character as taxes”).

Our cases confirm this functional approach. For example, in Drexel Furniture, we focused on three practical characteristics of the so-called tax on employing child laborers that convinced us the “tax” was actually a penalty. First, the tax imposed an exceedingly heavy burden—10 percent of a company’s net income—on those who employed children, no matter how small their infraction. Second, it imposed that exaction only on those who knowingly employed underage laborers; such scienter requirements are typical of punitive statutes. Third, this “tax” was enforced in part by the Department of Labor, an agency responsible for punishing violations of labor laws, not collecting revenue. 259 U. S., at 36–37.

The same analysis here suggests that the shared responsibility payment may for constitutional purposes be considered a tax, not a penalty: First, for most Americans the amount due will be far less than the price of insurance, and, by statute, it can never be more. It may often be a reasonable financial decision to make the payment rather than purchase insurance, unlike the “prohibitory” financial punishment in Drexel Furniture. Second, the individual mandate contains no scienter requirement. Third, the payment is collected solely by the IRS through the normal means of taxation—except that the Service is not allowed to use those means most suggestive of a punitive sanction, such as criminal prosecution. See § 5000A(g)(2). The reasons the Court in Drexel Furniture held that what was called a “tax” there was a penalty support the conclusion that what is called a “penalty” here may be viewed as a tax.

[Footnote 8, at p. 566, supplies the figures behind the first factor: in 2016 individuals making $35,000 a year are expected to owe the IRS about $60 for any month without insurance, and someone earning $100,000 about $200, against a projected qualifying policy price of around $400 per month.]

[Footnote 9, at pp. 566–567, marks the limit of the reasoning:]

We do not suggest that any exaction lacking a scienter requirement and enforced by the IRS is within the taxing power. Congress could not, for example, expand its authority to impose criminal fines by creating strict liability offenses enforced by the IRS rather than the FBI. But the fact the exaction here is paid like a tax, to the agency that collects taxes, suggests that this exaction may be viewed as a tax.

None of this is to say that the payment is not intended to affect individual conduct. Although the payment will raise considerable revenue, it is plainly designed to expand health insurance coverage. But taxes that seek to influence conduct are nothing new. Today, federal and state taxes can compose more than half the retail price of cigarettes, not just to raise more money, but to encourage people to quit smoking. And we have upheld such obviously regulatory measures as taxes on selling marijuana and sawed-off shotguns. Indeed, “[e]very tax is in some measure regulatory.” That § 5000A seeks to shape decisions about whether to buy health insurance does not mean that it cannot be a valid exercise of the taxing power.

In distinguishing penalties from taxes, this Court has explained that “if the concept of penalty means anything, it means punishment for an unlawful act or omission.” United States v. Reorganized CF&I Fabricators of Utah, Inc., 518 U. S. 213, 224 (1996). While the individual mandate clearly aims to induce the purchase of health insurance, it need not be read to declare that failing to do so is unlawful. Neither the Act nor any other law attaches negative legal consequences to not buying health insurance, beyond requiring a payment to the IRS. The Government agrees with that reading, confirming that if someone chooses to pay rather than obtain health insurance, they have fully complied with the law.

Indeed, it is estimated that four million people each year will choose to pay the IRS rather than buy insurance. That Congress apparently regards such extensive failure to comply with the mandate as tolerable suggests that Congress did not think it was creating four million outlaws. It suggests instead that the shared responsibility payment merely imposes a tax citizens may lawfully choose to pay in lieu of buying health insurance.

[The Court answers the argument from the statute’s wording by analogy to New York v. United States, where imperative language was read as “a series of incentives” and the charge sustained as a tax, and then offers a hypothetical:]

Suppose Congress enacted a statute providing that every taxpayer who owns a house without energy efficient windows must pay $50 to the IRS. The amount due is adjusted based on factors such as taxable income and joint filing status, and is paid along with the taxpayer’s income tax return. Those whose income is below the filing threshold need not pay. The required payment is not called a “tax,” a “penalty,” or anything else. No one would doubt that this law imposed a tax, and was within Congress’s power to tax. That conclusion should not change simply because Congress used the word “penalty” to describe the payment.

Our precedent demonstrates that Congress had the power to impose the exaction in § 5000A under the taxing power, and that § 5000A need not be read to do more than impose a tax. That is sufficient to sustain it.

[Still Part III–C — the Direct Tax Clause, at pp. 570–571.]

Even if the taxing power enables Congress to impose a tax on not obtaining health insurance, any tax must still comply with other requirements in the Constitution. Plaintiffs argue that the shared responsibility payment does not do so, citing Article I, § 9, clause 4. That clause provides: “No Capitation, or other direct, Tax shall be laid, unless in Proportion to the Census or Enumeration herein before directed to be taken.” This requirement means that any “direct Tax” must be apportioned so that each State pays in proportion to its population. According to the plaintiffs, if the individual mandate imposes a tax, it is a direct tax, and it is unconstitutional because Congress made no effort to apportion it among the States.

Even when the Direct Tax Clause was written it was unclear what else, other than a capitation (also known as a “head tax” or a “poll tax”), might be a direct tax. Soon after the framing, Congress passed a tax on ownership of carriages, over James Madison’s objection that it was an unapportioned direct tax. This Court upheld the tax. See Hylton v. United States, 3 Dall. 171, 174 (1796) (opinion of Chase, J.). The Court was unanimous, and those Justices who wrote opinions either directly asserted or strongly suggested that only two forms of taxation were direct: capitations and land taxes. That narrow view persisted for a century. In 1895, we expanded our interpretation to include taxes on personal property and income from personal property, in the course of striking down aspects of the federal income tax. Pollock v. Farmers’ Loan & Trust Co., 158 U. S. 601, 618 (1895). That result was overturned by the Sixteenth Amendment.

A tax on going without health insurance does not fall within any recognized category of direct tax. It is not a capitation. Capitations are taxes paid by every person, “without regard to property, profession, or any other circumstance.” The whole point of the shared responsibility payment is that it is triggered by specific circumstances—earning a certain amount of income but not obtaining health insurance. The payment is also plainly not a tax on the ownership of land or personal property. The shared responsibility payment is thus not a direct tax that must be apportioned among the several States.

[Still Part III–C — pp. 571–574, on taxing an omission.]

There may, however, be a more fundamental objection to a tax on those who lack health insurance. Even if only a tax, the payment under § 5000A(b) remains a burden that the Federal Government imposes for an omission, not an act. If it is troubling to interpret the Commerce Clause as authorizing Congress to regulate those who abstain from commerce, perhaps it should be similarly troubling to permit Congress to impose a tax for not doing something.

Three considerations allay this concern. First, and most importantly, it is abundantly clear the Constitution does not guarantee that individuals may avoid taxation through inactivity. A capitation, after all, is a tax that everyone must pay simply for existing, and capitations are expressly contemplated by the Constitution. The Court today holds that our Constitution protects us from federal regulation under the Commerce Clause so long as we abstain from the regulated activity. But from its creation, the Constitution has made no such promise with respect to taxes. Sustaining the mandate as a tax depends only on whether Congress has properly exercised its taxing power to encourage purchasing health insurance, not whether it can. Upholding the individual mandate under the Taxing Clause thus does not recognize any new federal power. It determines that Congress has used an existing one.

Second, Congress’s ability to use its taxing power to influence conduct is not without limits. A few of our cases policed these limits aggressively, invalidating punitive exactions obviously designed to regulate behavior otherwise regarded at the time as beyond federal authority. See, e. g., United States v. Butler, 297 U. S. 1 (1936); Drexel Furniture, 259 U. S. 20. More often and more recently we have declined to closely examine the regulatory motive or effect of revenue-raising measures. We have nonetheless maintained that “‘there comes a time in the extension of the penalizing features of the so-called tax when it loses its character as such and becomes a mere penalty with the characteristics of regulation and punishment.’” Because the tax at hand is within even our narrowest interpretations of the taxing power, we need not here decide the precise point at which an exaction becomes so punitive that the taxing power does not authorize it.

Third, although the breadth of Congress’s power to tax is greater than its power to regulate commerce, the taxing power does not give Congress the same degree of control over individual behavior. Once we recognize that Congress may regulate a particular decision under the Commerce Clause, the Federal Government can bring its full weight to bear: Congress may simply command individuals to do as it directs, and an individual who disobeys may be subjected to criminal sanctions. By contrast, Congress’s authority under the taxing power is limited to requiring an individual to pay money into the Federal Treasury, no more. If a tax is properly paid, the Government has no power to compel or punish individuals subject to it. But imposition of a tax nonetheless leaves an individual with a lawful choice to do or not do a certain act, so long as he is willing to pay a tax levied on that choice.

The Affordable Care Act’s requirement that certain individuals pay a financial penalty for not obtaining health insurance may reasonably be characterized as a tax. Because the Constitution permits such a tax, it is not our role to forbid it, or to pass upon its wisdom or fairness.

Part IV — the Medicaid expansion: a three-Justice plurality

[Chief Justice Roberts, joined only by Breyer and Kagan, JJ. — three Justices. Not an opinion of the Court. Running head: “Opinion of Roberts, C. J.” Reporter pp. 575–588.]

IV — A

The States also contend that the Medicaid expansion exceeds Congress’s authority under the Spending Clause. They claim that Congress is coercing the States to adopt the changes it wants by threatening to withhold all of a State’s Medicaid grants unless the State accepts the new expanded funding and complies with the conditions that come with it.

There is no doubt that the Act dramatically increases state obligations under Medicaid. The current Medicaid program requires States to cover only certain discrete categories of needy individuals—pregnant women, children, needy families, the blind, the elderly, and the disabled. The Medicaid provisions of the Affordable Care Act, in contrast, require States to expand their Medicaid programs by 2014 to cover all individuals under the age of 65 with incomes below 133 percent of the federal poverty line. The Federal Government will pay 100 percent of the costs of covering these newly eligible individuals through 2016, decreasing thereafter to a minimum of 90 percent.

[The plurality then sets out the Spending Clause framework — that such legislation is “much in the nature of a contract,” so that its legitimacy “rests on whether the State voluntarily and knowingly accepts the terms,” Pennhurst; that the Constitution “simply does not give Congress the authority to require the States to regulate,” New York; and that Congress may create incentives, but when “pressure turns into compulsion,” Steward Machine, the legislation runs contrary to our system of federalism. Reporter page 579 and the top of page 580 did not return in acquisition — that is the bridge between the Steward Machine and Dole discussions, cited later as “ante, at 579–580.” The text resumes at p. 580.]

Given the nature of the threat and the programs at issue here, we must agree. We have upheld Congress’s authority to condition the receipt of funds on the States’ complying with restrictions on the use of those funds, because that is the means by which Congress ensures that the funds are spent according to its view of the “general Welfare.” Conditions that do not govern the use of the funds cannot be justified on that basis. When such conditions take the form of threats to terminate other significant independent grants, they are properly viewed as a means of pressuring the States to accept policy changes.

In South Dakota v. Dole, we considered a challenge to a federal law that threatened to withhold five percent of a State’s federal highway funds if the State did not raise its drinking age to 21. We asked whether “the financial inducement offered by Congress” was “so coercive as to pass the point at which ‘pressure turns into compulsion.’” By “financial inducement” the Court meant the threat of losing five percent of highway funds; no new money was offered to the States to raise their drinking ages. We found that the inducement was not impermissibly coercive, because Congress was offering only “relatively mild encouragement to the States.” We observed that “all South Dakota would lose if she adheres to her chosen course as to a suitable minimum drinking age is 5%” of her highway funds. In fact, the federal funds at stake constituted less than half of one percent of South Dakota’s budget at the time. Whether to accept the drinking age change “remain[ed] the prerogative of the States not merely in theory but in fact.”

In this case, the financial “inducement” Congress has chosen is much more than “relatively mild encouragement”—it is a gun to the head. Section 1396c of the Medicaid Act provides that if a State’s Medicaid plan does not comply with the Act’s requirements, the Secretary of Health and Human Services may declare that “further payments will not be made to the State.” A State that opts out of the Affordable Care Act’s expansion in health care coverage thus stands to lose not merely “a relatively small percentage” of its existing Medicaid funding, but all of it. Medicaid spending accounts for over 20 percent of the average State’s total budget, with federal funds covering 50 to 83 percent of those costs. The Federal Government estimates that it will pay out approximately $3.3 trillion between 2010 and 2019 in order to cover the costs of pre-expansion Medicaid. In addition, the States have developed intricate statutory and administrative regimes over the course of many decades to implement their objectives under existing Medicaid. It is easy to see how the Dole Court could conclude that the threatened loss of less than half of one percent of South Dakota’s budget left that State with a “prerogative” to reject Congress’s desired policy, “not merely in theory but in fact.”

The threatened loss of over 10 percent of a State’s overall budget, in contrast, is economic dragooning that leaves the States with no real option but to acquiesce in the Medicaid expansion.

[Footnote 12, at p. 582, answering Justice Ginsburg on the size of the new burden:]

Justice Ginsburg observes that state Medicaid spending will increase by only 0.8 percent after the expansion. That not only ignores increased state administrative expenses, but also assumes that the Federal Government will continue to fund the expansion at the current statutorily specified levels… . More importantly, the size of the new financial burden imposed on a State is irrelevant in analyzing whether the State has been coerced into accepting that burden. “Your money or your life” is a coercive proposition, whether you have a single dollar in your pocket or $500.

Justice Ginsburg claims that Dole is distinguishable because here “Congress has not threatened to withhold funds earmarked for any other program.” But that begs the question: The States contend that the expansion is in reality a new program and that Congress is forcing them to accept it by threatening the funds for the existing one. We cannot agree that existing Medicaid and the expansion are all one program simply because “Congress styled” them as such.

[Still Part IV–A, the plurality — the new-program reasoning, at pp. 582–585.]

Here, the Government claims that the Medicaid expansion is properly viewed merely as a modification of the existing program because the States agreed that Congress could change the terms of Medicaid when they signed on, the Social Security Act having reserved “[t]he right to alter, amend, or repeal any provision” of that statute. So it does. But “if Congress intends to impose a condition on the grant of federal moneys, it must do so unambiguously.” Pennhurst, 451 U. S., at 17. A State confronted with language reserving the right to “alter” or “amend” might reasonably assume that Congress was entitled to make adjustments to the Medicaid program as it developed.

The Medicaid expansion, however, accomplishes a shift in kind, not merely degree. The original program was designed to cover medical services for four particular categories of the needy: the disabled, the blind, the elderly, and needy families with dependent children. Previous amendments merely altered and expanded the boundaries of these categories. Under the Affordable Care Act, Medicaid is transformed into a program to meet the health care needs of the entire nonelderly population with income below 133 percent of the poverty level. It is no longer a program to care for the neediest among us, but rather an element of a comprehensive national plan to provide universal health insurance coverage.

[Footnote 14, at pp. 583–584 — and read the attribution carefully. This footnote is the Chief Justice’s answer to Justice Ginsburg’s “Medicaid II” argument. Her own statement of that argument, at pp. 636–637 of her opinion, did not return in acquisition and is not available to this edition. What survives here is the Chief Justice’s characterization and partial quotation of it. Anyone assigning the argument should quote it from her pages, not from this footnote, and no sentence of this footnote should be attributed to Justice Ginsburg as her own writing:]

Justice Ginsburg suggests that the States can have no objection to the Medicaid expansion, because “Congress could have repealed Medicaid [and,] [t]hereafter, … could have enacted Medicaid II, a new program combining the pre-2010 coverage with the expanded coverage required by the ACA.” Post, at 636–637. But it would certainly not be that easy. Practical constraints would plainly inhibit, if not preclude, the Federal Government from repealing the existing program and putting every feature of Medicaid on the table for political reconsideration. Such a massive undertaking would hardly be “ritualistic.” Ibid. The same is true of Justice Ginsburg’s suggestion that Congress could establish Medicaid as an exclusively federal program. Post, at 630.

Indeed, the manner in which the expansion is structured indicates that while Congress may have styled it a mere alteration of existing Medicaid, it recognized it was enlisting the States in a new health care program: Congress created a separate funding provision for newly eligible persons, at a 90 percent federal share rather than the existing 50 to 83 percent, and the conditions on use of the different funds are distinct.

As we have explained, “[t]hough Congress’ power to legislate under the spending power is broad, it does not include surprising participating States with postacceptance or ‘retroactive’ conditions.” Pennhurst, supra, at 25. A State could hardly anticipate that Congress’s reservation of the right to “alter” or “amend” the Medicaid program included the power to transform it so dramatically.

The Court in Steward Machine did not attempt to “fix the outermost line” where persuasion gives way to coercion. The Court found it “[e]nough for present purposes that wherever the line may be, this statute is within it.” We have no need to fix a line either. It is enough for today that wherever that line may be, this statute is surely beyond it. Congress may not simply “conscript state [agencies] into the national bureaucratic army,” FERC v. Mississippi, 456 U. S. 742, 775 (1982) (O’Connor, J., concurring in judgment in part and dissenting in part), and that is what it is attempting to do with the Medicaid expansion.

[Part IV–B — the remedy. Same three-Justice plurality, but Ginsburg and Sotomayor, JJ., concur in the judgment as to this Part, which gives the remedy five votes. Reporter pp. 585–588.]

B

Nothing in our opinion precludes Congress from offering funds under the Affordable Care Act to expand the availability of health care, and requiring that States accepting such funds comply with the conditions on their use. What Congress is not free to do is to penalize States that choose not to participate in that new program by taking away their existing Medicaid funding. Section 1396c gives the Secretary of Health and Human Services the authority to do just that. It allows her to withhold all “further [Medicaid] payments … to the State” if she determines that the State is out of compliance with any Medicaid requirement, including those contained in the expansion. In light of the Court’s holding, the Secretary cannot apply § 1396c to withdraw existing Medicaid funds for failure to comply with the requirements set out in the expansion.

That fully remedies the constitutional violation we have identified. The chapter of the United States Code that contains § 1396c includes a severability clause confirming that we need go no further. Today’s holding does not affect the continued application of § 1396c to the existing Medicaid program, nor the Secretary’s ability to withdraw funds provided under the Affordable Care Act if a participating State fails to comply with that Act.

This is not to say, as the joint dissent suggests, that we are “rewriting the Medicaid Expansion.” When we invalidate an application of a statute because that application is unconstitutional, we are not “rewriting” the statute; we are merely enforcing the Constitution.

[The plurality then concludes that the holding does not bring down the rest of the Act: the question is “whether Congress would have wanted the rest of the Act to stand, had it known that States would have a genuine choice whether to participate,” and “we do not believe Congress would have wanted the whole Act to fall, simply because some may choose not to participate.”]

[The concluding statement of the Chief Justice’s opinion, at p. 588 — running head “Opinion of Roberts, C. J.”:]

As for the Medicaid expansion, that portion of the Affordable Care Act violates the Constitution by threatening existing Medicaid funding. Congress may offer the States grants and require them to comply with accompanying conditions, but the States must have a genuine choice whether to accept the offer. The States are given no such choice in this case: They must either accept a basic change in the nature of Medicaid, or risk losing all Medicaid funding. The remedy for that constitutional violation is to preclude the Federal Government from imposing such a sanction. That remedy does not require striking down other portions of the Affordable Care Act.

[One passage from the joint dissent of Scalia, Kennedy, Thomas and Alito, JJ., at p. 689, because it is the dissent’s own count of the Medicaid votes and explains why the result had seven:]

Seven Members of the Court agree that the Medicaid Expansion, as enacted by Congress, is unconstitutional. 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. We cannot accept the Government’s suggestion.

Justice Ginsburg on Medicaid

[Justice Ginsburg, with whom Justice Sotomayor joins, and with whom Justice Breyer and Justice Kagan join as to Parts I, II, III, and IV, concurring in part, concurring in the judgment in part, and dissenting in part. The Medicaid discussion sits in her Part V. Breyer and Kagan, JJ., did not join Part V — on Medicaid she writes for herself and Justice Sotomayor only. Her Commerce Clause Parts I–IV are the Module 3 reading.

This section of the acquisition is thin. Reporter pages 630–632, 635–642 (except 642) and 644 did not return, and are marked where they fall below. Her own text of the “Medicaid II” argument, at pp. 636–637, is among the missing pages; it appears in this edition only as quoted in the Chief Justice’s footnote 14, above, and is attributed there to the footnote.]

I agree with The Chief Justice that the Anti-Injunction Act does not bar the Court’s consideration of these cases, and that the minimum coverage provision is a proper exercise of Congress’ taxing power. I therefore join Parts I, II, and III–C of The Chief Justice’s opinion. Unlike The Chief Justice, however, I would hold, alternatively, that the Commerce Clause authorizes Congress to enact the minimum coverage provision. I would also hold that the Spending Clause permits the Medicaid expansion exactly as Congress enacted it.

[Her single-program argument, at pp. 625–626:]

Medicaid is a prototypical example of federal-state cooperation in serving the Nation’s general welfare. Rather than authorizing a federal agency to administer a uniform national health-care system for the poor, Congress offered States the opportunity to tailor Medicaid grants to their particular needs, so long as they remain within bounds set by federal law. In shaping Medicaid, Congress did not endeavor to fix permanently the terms participating States must meet; instead, it reserved the “right to alter, amend, or repeal” any provision of the Medicaid Act.

The Chief Justice acknowledges that Congress may “condition the receipt of [federal] funds on the States’ complying with restrictions on the use of those funds,” but nevertheless concludes that the 2010 expansion is unduly coercive. His conclusion rests on three premises, each essential to his theory. First, the expansion is, in his view, a new grant program, not an addition to the Medicaid program existing before the ACA. Second, the expansion was unforeseeable by the States when they first signed on. Third, the threatened loss of funding is so large that the States have no real choice but to participate. The Chief Justice therefore—for the first time ever—finds an exercise of Congress’ spending power unconstitutionally coercive.

Medicaid, as amended by the ACA, however, is not two spending programs; it is a single program with a constant aim—to enable poor persons to receive basic health care when they need it. Given past expansions, plus express statutory warning that Congress may change the requirements participating States must meet, there can be no tenable claim that the ACA fails for lack of notice. Moreover, States have no entitlement to receive any Medicaid funds; they enjoy only the opportunity to accept funds on Congress’ terms. The Federal Government, therefore, is not, as The Chief Justice charges, threatening States with the loss of “existing” funds from one spending program in order to induce them to opt into another. Congress is simply requiring States to do what they have long been required to do to receive Medicaid funding: comply with the conditions Congress prescribes for participation.

[Her notice and federal-share argument, at pp. 628–629:]

Compared to past alterations, the ACA is notable for the extent to which the Federal Government will pick up the tab. In 2014, federal funds will cover 100% of the costs for newly eligible beneficiaries; that rate will gradually decrease before settling at 90% in 2020. By comparison, federal contributions toward the care of beneficiaries eligible pre-ACA range from 50% to 83%.

Nor will the expansion exorbitantly increase state Medicaid spending. The Congressional Budget Office projects that States will spend 0.8% more than they would have, absent the ACA… . Whatever the increase in state obligations after the ACA, it will pale in comparison to the increase in federal funding.

Finally, any fair appraisal of Medicaid would require acknowledgment of the considerable autonomy States enjoy under the Act. Far from “conscript[ing] state agencies into the national bureaucratic army,” Medicaid “is designed to advance cooperative federalism.” The ACA does not jettison this approach. States, as first-line administrators, will continue to guide the distribution of substantial resources among their needy populations.

[Reporter pages 630–632 did not return in acquisition. They carry her argument that the Federal Government could have operated Medicaid as an exclusively federal program and that States have no entitlement to federal funds; it survives here only in the condensed form at p. 626 above and by her cross-reference at p. 633. The text resumes at p. 633.]

This litigation does not present the concerns that led the Court in Dole even to consider the prospect of coercion. In Dole, the condition—set 21 as the minimum drinking age—did not tell the States how to use funds Congress provided for highway construction, and in view of the Twenty-First Amendment it was an open question whether Congress could directly impose a national minimum drinking age.

The ACA, in contrast, relates solely to the federally funded Medicaid program; if States choose not to comply, Congress has not threatened to withhold funds earmarked for any other program. Nor does the ACA use Medicaid funding to induce States to take action Congress itself could not undertake. The Federal Government undoubtedly could operate its own health-care program for poor persons, just as it operates Medicare for seniors’ health care. That is what makes this such a simple case, and the Court’s decision so unsettling.

The starting premise on which The Chief Justice’s coercion analysis rests is that the ACA did not really “extend” Medicaid; instead, Congress created an entirely new program to coexist with the old. The Chief Justice calls the ACA new, but in truth, it simply reaches more of America’s poor than Congress originally covered.

Medicaid was created to enable States to provide medical assistance to “needy persons.” By bringing health care within the reach of a larger population of Americans unable to afford it, the Medicaid expansion is an extension of that basic aim. The Medicaid Act contains hundreds of provisions governing operation of the program; the expansion leaves unchanged the vast majority of them. It adds beneficiaries to the existing program and specifies the reimbursement rate for services provided to them. The ACA does not describe operational aspects of the program for these newly eligible persons; for that information, one must read the existing Medicaid Act.

[Reporter pages 635–642 did not return in acquisition, except p. 642. Those pages carry her “Medicaid II” passage (pp. 636–637 — see the note at footnote 14 above) and her Part V–C–2 and V–C–3 discussion of foreseeability and fair notice, including the sentence the Chief Justice quotes as “a State would be hard put to complain that it lacked fair notice,” post, at 641. The text resumes at p. 642.]

The Chief Justice ultimately asks whether “the financial inducement offered by Congress … pass[ed] the point at which pressure turns into compulsion.” The financial inducement Congress employed here, he concludes, crosses that threshold: The threatened withholding of “existing Medicaid funds” is “a gun to the head” that forces States to acquiesce. The Chief Justice sees no need to “fix the outermost line,” “where persuasion gives way to coercion.” Neither do the joint dissenters.

[Her unadministrability argument, at p. 643:]

When future Spending Clause challenges arrive, as they likely will in the wake of today’s decision, how will litigants and judges assess whether “a State has a legitimate choice whether to accept the federal conditions in exchange for federal funds”? Are courts to measure the number of dollars the Federal Government might withhold for noncompliance? The portion of the State’s budget at stake? And which State’s—or States’—budget is determinative: the lead plaintiff, all challenging States (26 in this litigation, many with quite different fiscal situations), or some national median?

[Reporter page 644 did not return in acquisition. It carries the continuation of the unadministrability argument between p. 643 and the start of her Part V–D at p. 645. The text resumes at p. 645.]

Congress has delegated to the Secretary of Health and Human Services the authority to withhold, in whole or in part, federal Medicaid funds from States that fail to comply with the Medicaid Act. The Chief Justice, however, holds that the Constitution precludes the Secretary from withholding “existing” Medicaid funds based on States’ refusal to comply with the expanded program. For the foregoing reasons, I disagree that any such withholding would violate the Spending Clause.

But in view of The Chief Justice’s disposition, I agree with him that the Medicaid Act’s severability clause determines the appropriate remedy. The Court does not strike down any provision of the ACA. It prohibits only the “application” of the Secretary’s authority to withhold Medicaid funds from States that decline to conform their plans to the ACA’s requirements. Thus the ACA’s authorization of funds to finance the expansion remains intact.

Even absent § 1303’s command, we would have no warrant to invalidate the Medicaid expansion, not to mention the entire ACA. For when a court confronts an unconstitutional statute, its endeavor must be to conserve, not destroy, the legislature’s dominant objective. In this instance, that objective was to increase access to health care for the poor by increasing the States’ access to federal funds. I therefore concur in the judgment with respect to Part IV–B of The Chief Justice’s opinion.

[Two further acquisition gaps should be recorded. Reporter pages 543 and 545, in Part II of the Chief Justice’s opinion (the Anti-Injunction Act), did not return. Page 543 states the Act’s terms; page 545 carries the “assessed and collected in the same manner as taxes” discussion that Part III–C cross-refers to as “supra, at 545–546.” Part II is omitted from this reading in any event, but that cross-reference points into a page this edition does not hold.]

Notes & Questions

1. Big picture — the first time in seventy-six years that a spending condition failed, and the reason it is here rather than in Module 3. You read this case once already, for the Commerce Clause, and what you learned there was that the Chief Justice’s Commerce Clause analysis commanded no other vote. Here the arithmetic runs the other way and produces two very different things. On the taxing power Roberts speaks for a majority: the shared-responsibility payment, whatever Congress called it, functions as a tax, and Congress may impose it. That is how the individual mandate survives — not on the power everyone argued about for two years, but on the alternative ground the Government pleaded second. On Medicaid, Roberts writes for himself, Breyer and Kagan, and holds that conditioning a State’s existing Medicaid funds on its acceptance of the expansion is unconstitutionally coercive. It is the first time since Butler — the case that opens this module — that the Court has invalidated an exercise of the conditional spending power, and it is the first time ever that Dole’s fourth restriction has done any work. So the module has a shape: Butler announces a coercion idea it cannot define, Steward Machine and Helvering bury the result while keeping the idea, Dole states it as a restriction that never bites, and here it bites. Read the Medicaid section with Dole open beside it. The Court is not applying a new rule; it is finally putting a number on an old one.

2. Raw specific knowledge — learn the parts before you learn the holdings. This case cannot be stated correctly without stating who joined what, and the reading labels every passage for exactly that reason. Memorise the following. Part III–C is the opinion of the Court — Roberts with Ginsburg, Breyer, Sotomayor and Kagan — and it holds the payment may be sustained under the taxing power. Part III–B is Roberts alone, joined by no one; it contains the saving construction, the move that reads the statute as imposing a tax rather than commanding a purchase. Part IV is a three-Justice plurality — Roberts with Breyer and Kagan only. Then the substance: (a) the functional test for what counts as a tax, and the factors drawn from Drexel Furniture — the magnitude of the exaction relative to the cost of compliance, the absence of a scienter requirement, and collection by the IRS through the ordinary means; (b) the Court’s insistence that a thing may be a penalty for one purpose and a tax for another, and what that does to the Anti-Injunction Act question; (c) the Direct Tax Clause analysis and why the payment is not a direct tax requiring apportionment; (d) on Medicaid, the new-program / existing-program distinction and the remedy — the withholding provision is unenforceable, the expansion itself stands. And note the reporting convention: there is no vote line on this reading, because a single number would misdescribe a case whose line-up changes part by part.

3. Practical application — the two numbers, and the drafting problem they create. The coercion holding runs on a comparison you can state in one sentence. In Dole, what was at risk was less than half of one percent of South Dakota’s budget. Here, federal Medicaid funds are over 10 percent of a State’s overall budget. That contrast is the argument; everything else is characterisation. So: (a) Plot the two figures on a line and mark where you think the constitutional boundary falls. Then state what your boundary does to the Federal-Aid Highway Program, to Title I education funding, and to federal disaster relief. (b) You are drafting the Medicaid expansion in 2009 with this opinion in hand. Redraft § 2001 so that it achieves the same coverage and survives. Your options include making the expansion a separate program with separate funds, phasing the federal share, offering the expansion as an unconditional option, or conditioning only new money. Draft one and then apply all four Dole restrictions plus this case. (c) Now the harder question. The plurality says the vice is that Congress attached a condition to money the States had already been receiving for decades and planned around — a change in the terms of an existing bargain rather than an offer of a new one. But Congress expressly reserved the right to amend the Medicaid Act, and the States took the money knowing that. Does the reservation answer the argument? If a reserved power to amend can be exercised only in ways that are not “surprising,” what is left of it — and what should Congress write into the next grant statute?

4. Attack the reasoning — the saving construction, and a remedy nobody voted for as a whole. (a) III–B is Roberts alone, and it is the load-bearing wall. The reasoning is that the statute is not naturally read as a command to buy insurance, but that it can fairly be read as a tax on going without, and that every reasonable construction must be resorted to in order to save a statute. Two objections to weigh. First, the Chief Justice reaches the constitutional question under the Commerce Clause before adopting the saving construction — but a saving construction is supposed to avoid deciding constitutional questions, not to follow one. Is the Commerce Clause discussion therefore dictum in the strictest sense, and does it matter that the reading you did in Module 3 was of an opinion joined by no one? Second, “fairly possible” is doing enormous work on a statute that calls the exaction a penalty in its text and was defended in Congress as not a tax. State the limiting principle that keeps the saving canon from becoming a licence to rewrite. (b) The Medicaid vote count. Seven Justices concluded the withholding was unconstitutional — the three-Justice plurality plus the four joint dissenters — but the joint dissenters would have struck the entire Act, while Ginsburg and Sotomayor, who thought the expansion perfectly constitutional, agreed that if it were coercive the right remedy was to bar the withholding rather than to invalidate the expansion. So the holding on the violation and the holding on the remedy rest on different majorities, assembled from Justices who disagree about the premise. Work out both coalitions on paper. Then ask the jurisprudential question: is an outcome supported by two different five-or-more-vote coalitions, neither of which agrees with the other about why, a holding of the Court at all? (c) Justice Ginsburg’s answer on Medicaid is thin in this reading — several reporter pages did not return in acquisition and are marked as gaps on the page. Note in particular that her “Medicaid II” argument appears here only as quoted inside Roberts’s footnote 14, and must be attributed to the footnote. That is not a stylistic nicety: a party’s argument as characterised by the Justice rejecting it is not the argument. If you need it, go to the reporter.

5. Creative thinking — what makes an offer a command? You have now met this question four times: the cotton farmer in Butler, the unemployment-compensation credit in Steward Machine, the highway funds in Dole, and the Medicaid expansion here. Two exercises. (a) State the theory. The plurality’s image is a gun to the head. But a gun to the head is a threat to make you worse off than the status quo, and no State was threatened with anything it had a legal entitlement to keep. Reconstruct the argument without the metaphor. The strongest version probably runs through reliance rather than through size — that the States built institutions, budgets and constituencies around a program over forty-five years, and that Congress may not use the leverage those investments created to extract agreement to something new. Write that argument in one paragraph, and then say what other federal programs it destabilises. (b) Now break it. Construct the case that coercion is not a workable constitutional concept in this setting at all: that the size of an offer cannot convert it into a command, that “existing program” and “new program” is a distinction Congress controls by drafting, and that the real objection in this case was to the size of the policy change rather than to any feature of the spending power. If you find that persuasive, you owe an account of what Dole’s fourth restriction is for. If you do not, you owe a rule that a district judge could apply to the next grant statute without waiting to see how big the political reaction is. Whichever you choose, this is the question the module has been building toward since Butler, and it is still open.