The Yale Law Journal


Conditional Taxation and the Constitutionality of Health Care Reform

25 May 2010

The recent enactment of major health care reform legislation has brought with it a welter of constitutional challenges to the legislation and its key provisions. Attorneys General in more than a dozen states have already filed suits seeking to enjoin the operation of the statute, arguing that its requirement that most individuals either purchase health insurance or pay a penalty tax exceeds Congress’s enumerated powers. And several prominent scholars have argued similarly that this “individual responsibility requirement” (IRR) ought to be unconstitutional, even if current case law does not clearly require that outcome.

These claims are mistaken because, among other reasons, they take too narrow a view of Congress’s constitutional power to “lay and collect . . . Taxes . . . to pay the Debts and provide for the common Defen[s]e and general Welfare of the United States.” Even if the IRR is not authorized by the Commerce Clause, or by some combination of the commerce power and the Necessary and Proper Clause, it would still fall squarely within Congress’s authority to set the terms on which it will collect revenues. To be clear, I believe that the commerce power permits Congress to require the purchase of insurance, and that, if not, the Necessary and Proper Clause would permit it instead. But for those who disagree, such as Professor Randy Barnett, the taxing power remains as an alternative source of congressional authority.

In brief, my claim is that “conditional” taxes—taxes used to achieve some regulatory end—are not limited only to those purposes covered by Congress’s other enumerated powers. Instead, Congress may condition exemptions from a tax on any criteria it chooses—other than those expressly prohibited by the Constitution, such as restrictions on free speech—so long as it is willing to pay the political price for carving out that exception. This structure exactly mirrors the well-established law of conditional spending.

Part I of this Essay explains why any court challenge to the use of the taxing power would fail. Part II explains why it should.

I. The Court and the Taxing Power

The existing doctrine on the scope of Congress’s taxing power is at first glance muddy, but the best reading is that courts will not impose any substantive limits on the uses to which Congress may put its taxing authority. Any confusion results from the Court’s failure to formally overrule outdated precedents that once suggested otherwise.

A. Article I, Section 8

Although the Supreme Court has never formally repudiated an early decision that Congress’s power to tax extends no further than the Commerce Clause, that holding is effectively dead letter. For example, in the most recent decision directly on point, United States v. Kahriger, the Court rejected a challenge to a federal tax on “the business of accepting wagers.” The challengers claimed that the tax was invalid on the basis that it was, in effect, a regulation exceeding the bounds of authority constitutionally granted to Congress. But the Court found that the tax was valid notwithstanding both its obvious purpose to restrict gambling and the uncertainty as to whether the commerce power authorized Congress to enact such a restriction. The Court explained: “Unless there are [penalty] provisions extraneous to any tax need, courts are without authority to limit the exercise of the taxing power.” That is, if a provision that is labeled as a tax raises any revenue, it is within the taxing power. For example, the Court pointed approvingly to a tax on adulterated butter that had raised $3,501 nationwide.

Similarly, modern cases interpreting the congressional power to spend funds acknowledge that any judicial limits on Congress’s power to tax and spend are largely hortatory. Again, although Lochner-era cases such as United States v. Butler suggested otherwise, Congress may now use its spending power to pursue ends other than those set out in its other enumerated powers. In the key case, South Dakota v. Dole, the Court listed some loose limits on what courts would demand from Congress’s spending decisions but admitted they could not be meaningfully enforced.

That the Court has found no justiciable limits on the spending power necessarily implies a similar lack of limits on the taxing power. For one thing, the absence of limits on the spending power deeply undercuts the logic of the Court’s earlier, pre-Kahriger ruling constraining the taxing power. In that decision, The Child Labor Tax Case,the Court reasoned that some bounds on the taxing power were necessary to prevent the loss of “all constitutional limitation” on Congress. The Butler Court later applied the same logic to the Spending Clause, but that reasoning was then rejected by Dole. Even if the rationale of Dole could not be extended to conditional taxes, The Child Labor Tax Case rule now would be senseless: there is little need to preserve strict categorical limits if Dole already has abandoned them.

In any event, Dole’s rationale does apply squarely to conditional taxes, as the Court has recognized that conditional tax exemptions are the functional equivalent of direct spending. In most cases it is trivial to construct fiscally identical spending and taxing regimes. For instance, suppose we want to offer a $100 incentive to encourage production of widgets to a citizenry of two, A and B. We can offer a $100 grant to widget makers, funded by a $100 tax on all. Or we can impose a $100 tax on all, with an exemption for widget makers. Or we can impose a $100 tax on the failure to produce widgets. If A makes widgets and B refuses, all three leave A with a net balance of $0, B with a loss of $100, and $100 in revenue for the government. If the tax succeeds in changing B’s behavior, then all three players again end up the same, with a net of $0 for each. As such, it would make no sense to treat conditional taxing and spending differently, because they are economically identical. Moreover, because the two derive textually from the same words, it would be odd for those words to be limited in one setting and not in another.

B. The Apportionment Clause

In addition to arguments about the limits of Article I, Section 8, Professor Barnett and others claim that the IRR is a “direct” tax and thus subject to a separate constitutional requirement that the burden of “direct” taxes must be “apportioned,” or divided among the states in proportion to each state’s population. An income tax, however, need not be apportioned. The IRR is an income tax because whether a family pays $695 or some other amount depends on the household’s income. In addition, the obligation to pay the minimum $695 tax is subject to exemptions for personal hardship, which are also determined with reference to income.

Even if the IRR were not an income tax, it still would not need to be apportioned because it is not a “direct” tax. Though there is no agreement on the meaning of a “direct” tax, in all likelihood it includes only “capitation” taxes and taxes on the mere ownership (rather than use) of real or personal property. Taxes on a “particular use . . . of property,” or on “particular business transactions, vocations, occupations and the like,” are indirect and need not be apportioned. For instance, a tax on the privilege of doing business in corporate form is indirect.

Under the framework of these cases, the IRR is an indirect tax. It is imposed on a particular use of wealth: the use of personal wealth for purposes other than the purchase of health insurance. Alternatively, it could be seen as a tax on a particular form of insurance: the choice to shift the risk of future medical needs from oneself to the social safety net—in effect, a tax on the use of the existing systems of free care, Medicaid, debtor protection, and bankruptcy law. Even an advocate for an expansive view of “direct” taxation, Erik Jensen, concedes (in light of evidence from the Founding era) that “an indirect tax is one which the ultimate consumer can generally decide whether to pay.” Taxpayers can avoid the IRR simply by purchasing a qualifying insurance plan, and those who cannot afford to do so will not pay the IRR.

Finally, the Supreme Court has held that taxes (or, at least, taxes besides those on real estate) for which the rule of apportionment cannot “reasonably apply” are by definition not direct taxes. The fact that households’ tax would be dependent on their income means that the IRR could not be apportioned without gross national inequalities, which implies it would not be a “direct” tax.

Thus, it is clear that the IRR faces no serious court challenge. Whether or not the IRR is permissible under the Commerce Clause, there is little question that it would easily clear the low modern bar for exercises of the taxing power.

II. Textual and Structural Arguments

Although the doctrinal view of the IRR is likely well settled, it is fair to ask whether that outcome is a persuasive reading of the constitutional text. A number of commentators have attacked the wide latitude that the Court gives to the spending power, and at least some of them feel similarly about the power to tax. Most of these arguments are based on inferences from the text of the Constitution. The first is that reading the phrase “for the general welfare” to encompass broad regulatory aims makes the Taxing Clause’s grants of authority to provide for the common defense and to pay the government’s debts redundant. The second is that a broad reading of those powers, even if consistent with the text’s plain language, is absurd because it would make the specific enumeration of a limited set of other congressional powers meaningless. Both of these attacks are misguided.

A. Redundancy

The redundancy argument fails because it assumes that there could be no reason for the Constitution to emphasize the national power to pay debts and defend the country. But the Constitution is not just a set of instructions for text-parsing technocrats; it is also a political document. Its authors had tremendously important reasons for hammering home to readers the new nation’s status as a reliable borrower and dangerous foe. These were obviously vital messages both to the great powers of Europe and to possibly recalcitrant domestic constituencies. Additionally, they served as a sales pitch to colonial citizens whose votes were needed to ratify the document and who had witnessed the struggles of the Articles of Confederation on both these fronts.

Even if taking the Constitution as intended solely as a set of instructions for later interpreters, there are legitimate reasons to state the borrowing and defense powers separately. Although the inference that every word in a legal text should be given independent meaning is a common interpretive technique, like all such techniques it is hardly universal. The Tax Code, for example, defines “gross income” as “all income from whatever source derived” and then goes on to enumerate some fifteen specific sources. This approach helps to provide notice to parties who might be uncertain about the meaning of a broad term, such as the Tax Code’s “income” or the Constitution’s “general welfare.” It also guarantees a minimum content for those broad terms.

Furthermore, in many instances courts will demand especially clear language before they recognize the grant of a particular power. Seemingly redundant terms may simply be a way of ensuring that the text is construed correctly even in the face of such demands by later interpreters. Considering that a national standing army and the status of sovereign debt were particularly contentious issues at the time of the framing, it is plausible that the text’s authors would have worried that, without a clear statement, the taxing and spending powers might not be found to authorize those tasks.

B. Absurdity

Thus, the text of the Taxing Clause by its terms is fairly unlimited: it authorizes any taxing and spending “for the general welfare” and without further limitation. But can that really be what the language means? Critics echo the Child Labor Tax Case, arguing there would have been no point to limiting Congress to some enumerated powers if Congress can use its powers to tax and spend for any purpose at all. So, they say, the Constitution cannot possibly be read that way.

As I have argued in the context of the Spending Clause, this structural inference about the meaning of the Constitution is misguided because it fails to recognize that the taxing and spending powers are not unlimited. They are simply limited in a different way than Congress’s other enumerated powers. The spending power has an inherent political constraint: Congress can only do what it can afford to pay for, and to pay for things Congress ultimately must impose taxes. Regulations need not be paid for (although someone must enforce them) and may be rather more hidden from the public’s view at the time of enactment than a tax increase. Thus, the commerce power is limited in scope but easy to invoke; the spending power is unlimited in scope but difficult to invoke. Whether or not this is the optimal way to design a government, it certainly is a plausible one. Thus, critics of the spending power cannot claim that an interpretation of the Spending Clause that leaves it without substantive limits is absurd or impossible. We are left with the plain text, which by its terms allows any tax and expenditure “for the general welfare.” This same logic applies to conditional taxes because, again, conditional tax incentives are indistinguishable from conditional grants.

Another way in which the taxing and spending powers may be limited is through judicial interpretation of statutes that rely on them. To the extent that conditional spending looks too easy for Congress, there are tools more nuanced than outright invalidation that courts can and do use to strike a better balance. Courts can demand that Congress speak with especially clear and particularized language or that Congress make particularized findings. And courts can limit Congress’s ability to delegate expansive lawmaking by limiting agency power to approach constitutional limits without clear authorization. These tools have almost entirely replaced invalidation as the means of limiting Congress’s enumerated powers, and rightly so: they strike a better balance between the limits of the courts’ competence to interpret terms such as “general welfare” and the strictures of the Constitution. Indeed, although he does not explicitly consider this framework at all, a form of the clear statement rule is Jensen’s ultimate answer to the question of whether a tax must be apportioned: courts could simply ask whether Congress expressly labeled it as an “income” tax.


Under present law, the IRR provisions of the new health care legislation are obviously constitutional. The leading critiques of present law are unpersuasive, and would apply equally to the spending power, where they have gained no traction. Challenges to the IRR may be good politics, but they are not warranted by existing law.

Brian Galle is a Visiting Associate Professor, George Washington University Law School; Assistant Professor, Florida State University College of Law. The author would like to extend thanks to Jack Balkin, Joseph Dodge, Rick Hills, and Ilya Somin for helpful comments and suggestions as well as to Sophia Brill and the staff at TheYale Law Journal Online for swift and insightful editorial assistance.

Preferred citation: Brian Galle, Conditional Taxation and the Constitutionality of Health Care Reform, 120 Yale L.J. Online 27 (2010),