The Constitution was intended to give the national government greater power to raise revenue—the Articles of Confederation had been a fiscal disaster—but many Framers remained fearful of taxation. Indirect taxes (generally understood as falling on articles of consumption) did not lend themselves to congressional abuse (for reasons that will be described presently), but the Framers believed that “direct taxes” needed to be cabined. The cumbersome apportionment rule, requiring that a direct tax be apportioned among the states on the basis of population (so that, for example, a state with twice the population of another state would have to pay twice the tax, even if the more populous state’s share of the national tax base were smaller), made the more dangerous taxes politically difficult for Congress to impose.

The effectiveness of apportionment as a limitation on congressional power obviously depends on the levies to which it applies, and students of the Founding disagree on this point. At one extreme, some scholars, citing Rufus King’s unanswered question at the Constitutional Convention (“Mr King asked what was the precise meaning of direct taxation? No one answd.”), have argued that “direct taxes” had no agreed upon meaning, or, much the same thing, that the Framers did not think through what they were doing. They created an apportionment scheme so unworkable that a cramped definition of “direct taxes” became necessary to prevent the collapse of the system.

Those views overstate the extent of the confusion in 1787. No interpretation of “direct taxes” can be consistent with all statements made at the time, but the Founding debates are full of references to two forms of taxation for which apportionment was clearly intended: capitation taxes (specifically denominated as direct in the Constitution) and taxes on land (generally including slaves as well). Although intended to be difficult, apportionment was not impossible. Between 1798 and 1861, Congress enacted several real estate taxes, all with complex schemes for apportionment.

Does “direct taxes” include anything beyond capitation and taxes on property? The conventional wisdom is that it does not, based on dicta in Hylton v. United States (1796), which held that a tax on carriages was an excise rather than a direct tax. Justice William Paterson, for example, thought the provision was designed to allay Southern fears of a federal tax on their lands and slaves—nothing more. Because the Federalist justices were themselves among the Framers, these dicta are often accepted as evidence of original understanding. Not all significant Framers thought the concept of “direct taxes” was so limited—James Madison voted against the carriage tax in Congress because he thought it needed to be apportioned—but, based on the Hylton dicta, the Supreme Court in the nineteenth century upheld unapportioned federal taxes on insurance company receipts, Pacific Insurance Co. v. Soule (1869); on notes of state-chartered banks, Veazie Bank v. Fenno (1869); on inheritances of real estate, Scholey v. Rew (1875); and on the Civil War income tax, Springer v. United States (1881).

What looked to be a revolutionary change occurred in 1895, when, in Pollock v. Farmers’ Loan & Trust Co. (1895), a divided Court accepted a broader conception of “direct taxes” and concluded that an unapportioned 1894 income tax—which largely reached income from property—was invalid. In rejecting the notion that nothing but a capitation or land tax could be direct, the Court stressed that a limitation on congressional power ought not to be interpreted in a way that destroys the limitation. Although the Court after Pollock continued to approve a large number of unapportioned federal taxes, calling them “excises,” Pollock unquestionably hampered the government’s ability to raise revenue. In 1913, the Sixteenth Amendment was ratified, exempting “taxes on incomes” from apportionment.

Over the years, until 2012 (see below), Supreme Court authority had provided little further guidance on the meaning of “direct taxes.” That was understandable because the Sixteenth Amendment made worrying about new sources of revenue less pressing for the federal government. Nonetheless, it seems that any direct tax other than an income tax should still be subject to the apportionment rule today. At a minimum, that would include capitation and land taxes. (In 1934, the Supreme Court emphasized, in dictum, that a tax on the value of real estate would be direct. Helvering v. Independent Life Insurance Co. (1934).)

Furthermore, after Pollock, a broader interpretation of “direct taxes” was certainly possible. In fact, a broad interpretation of “direct taxes” can reconcile the clear original understanding that capitation and land taxes were direct with the equally clear intention that apportionment should have bite. The Constitution divided governmental levies into two mutually exclusive categories: indirect taxes subject to the uniformity requirement, and direct taxes subject to apportionment. Indirect taxes, which the Framers assumed would fund the national government in ordinary circumstances, were “Duties, Imposts, and Excises”—generally taxes on articles of consumption. These taxes were considered safe because, regardless of who collected them, the burden was thought to be shifted to consumers. If Congress became greedy and raised rates too high, fewer taxed goods would be purchased and revenue would decrease. It was because taxes on articles of consumption “contain in their own nature a security against excess,” wrote Alexander Hamilton in The Federalist No. 21, that further constitutional protection against congressional overreaching was unnecessary.

Direct taxes, which were expected to be used only in emergencies, did not have the built-in protections characteristic of indirect taxes. Direct taxes were imposed directly on individuals, who, it was assumed, could not shift the tax burden away from themselves. If a tax was not indirect, the Framers thought it should be apportioned. Capitation and land taxes (the eighteenth century form of a wealth tax) were direct under this understanding, but so might other taxes be, whether known in 1787 or not.

Recent scholarship has provided another reason to think that the phrase “direct taxes” has a much broader meaning than is generally understood today. James Campbell has demonstrated that the term “capitation” was understood in the late eighteenth century to encompass much more than a lump-sum head tax; it would, for example, have included an income tax.

With the exception of Pollock, however, the Supreme Court has not blessed an expansive conception of what constitutes a direct tax. For one brief moment in 2006, a panel of the U.S. Court of Appeals for the District of Columbia Circuit did just that. In Murphy v. Internal Revenue Service, the panel initially held that an unapportioned tax on a whistleblower’s recovery, received because she was wrongly discharged from her governmental position and suffered emotional distress, was direct and was not exempted from apportionment by the Sixteenth Amendment. Facing criticism, however, the panel vacated that decision, reheard the case, and in 2007 made a 180-degree turn. As a matter of first principle, the panel seemed sympathetic to the notion of direct taxes outlined above, but it could find no support for that understanding in Supreme Court cases.

There continues to be no Supreme Court support for a broader conception of direct taxes. In 2012, in ruling on the constitutionality of the individual mandate in the Affordable Care Act (the requirement that individuals must purchase health insurance or pay a “penalty”), the Court held that the penalty provision was actually a tax. The opinion of Chief Justice John Roberts went further. It effectively held that the only “recognized” categories of direct taxes are capitations and taxes on the ownership of property, and that, as a result, the penalty for failure to acquire suitable health insurance, scheduled to come into effect in 2014, will not be a direct tax. National Federation of Independent Business v. Sebelius (2012). (To no avail, four dissenters complained that NFIB presented a case of first impression about the scope of direct taxes, and that that issue had been inadequately briefed and argued.) Moreover, writing for the majority, Chief Justice Roberts quoted Justice Samuel Chase’s opinion in Hylton that “[c]apitations are taxes paid by every person ‘without regard to property, profession, or any other circumstance’”—in effect, limiting capitations to lump-sum head taxes. With the Supreme Court’s decision in NFIB, we are left with a narrow understanding of direct taxes, encompassing at most capitation taxes (itself narrowly construed) and taxes on property, even though the original understanding of direct taxes may have been far broader.