Professor Calvin Johnson’s argument that the wealth tax proposed by Senator Elizabeth Warren would clearly be constitutional, made in the Summer 2019 ABA Tax Times, is misleading in several respects. (Senator Bernie Sanders has now also proposed a wealth tax, and I assume Professor Johnson would see no constitutional problems with that proposal as well. Indeed, Sanders cites the Johnson essay.) The misleading aspects run from the title to how Johnson interprets the direct-tax clauses in the Constitution to a bewildering failure to discuss a Supreme Court decision from 2012, National Federation of Independent Business v. Sebelius (NFIB), in which Chief Justice John Roberts advanced propositions that contradict Johnson’s argument about the meaning of direct taxation.
Misleading proposition number one is the title, “A Wealth Tax Is Constitutional.” (I suppose the title might not have been Johnson’s choice.) Of course a wealth tax can be constitutional; that’s not news. The congressional taxing power set out in Article I, section 8, is quite broad, permitting Congress to tax almost anything. The question isn’t whether Congress can impose a tax on wealth; the question is whether such a tax would be a direct tax that, under two constitutional provisions, would have to be apportioned among the states on the basis of population. For example, the residents of a state with one-twentieth of the national population would have to pay, in the aggregate, one-twentieth of any direct tax. (To make Johnson’s intended point, the title should have been “An Unapportioned Wealth Tax Is Constitutional.”) If a wealth tax would have to be apportioned, it couldn’t work as Senators Warren and Sanders want it to. To make the numbers come out right, it would have to have, as Johnson notes, “substantially higher tax rates in poorer states” than in richer ones. As Johnson indicated, that would be absurd.
Misleading proposition number two is that the potentially absurd result—“higher tax rates in poorer states”—is an “injustice.” The point of apportionment was to make direct taxation difficult, to limit the national taxing power. If apportionment would make a proposed direct tax unjust, Congress generally shouldn’t enact it. No direct tax, no injustice. And direct taxation wasn’t supposed to be the primary revenue source for the United States. It was to be used only in emergencies like wartime, when revenue needs trump arguably unjust effects.
From the beginning, the Supreme Court assumed that a tax on real property is a direct tax—as distinguished from indirect taxes, generally taxes on articles of consumption—as had Alexander Hamilton writing as Publius in The Federalist. In Hylton v. United States, decided in 1796, the Court concluded that a national tax on carriages wasn’t direct and therefore didn’t have to be apportioned. But the three justices who wrote opinions agreed that a tax on land—the most significant form of wealth at the time—was direct. For example, Justice Samuel Chase said that the direct taxes “contemplated by the Constitution, are only two, to wit, a capitation, or poll tax, simply, without regard to property, profession, or any other circumstance; and a tax on LAND.”
Professor Johnson once treated Hylton as gospel because he liked the way the justices defined “direct tax” narrowly. The justices, founders all, were, Johnson wrote, “giants [who] walked upon the earth,” and what they said in Hylton was unquestionably correct: they “knew the Constitution far better than we do.” Apparently their stature has recently shrunk, however; now we must ignore what they wrote in Hylton about land taxes’ being direct. In contrast to Johnson, I believe the Hylton Court’s understanding of “direct tax” was too narrow, but even that narrow conception encompassed taxes on real property.
And it wasn’t just Publius and Supreme Court justices who understood that a tax on real property was direct. Between 1798 and 1861 Congress enacted several national real-estate taxes, all of which were apportioned. Congress took it for granted that apportionment was required.
Professor Johnson has a response to what I’ve just described, of course, which brings us to misleading proposition number three. He argues that apportionment of land taxes was permissible in the early years of the republic because “[i]t was presumed ... that real estate value was equal per capita across the states, and that a real estate tax would qualify as a direct tax.” In that case, apportionment was acceptable because it didn’t lead to injustice. But, he says, “the consequences of apportionment by population when the per capita tax base is uneven rebuts the use of apportionment.” He adds:
Real estate and wealth taxes ceased to be direct taxes because per capita wealth or land value so varied among the states that apportionment by population would require drastically higher tax rates in poorer states. Drastically higher tax rates required by apportionment by population entails that apportionment is not required because the tax is not direct.
So what was considered a direct tax in the late eighteenth century isn’t necessarily one today: “Over time taxes that could not reasonably be apportioned among the states were expelled from being considered . . . direct tax[es], by ordinary language usage or Supreme Court decision.”
Ordinary language usage? Come on. Think about what Professor Johnson is claiming: the apportionment requirement is to be imposed only when it doesn’t matter—only, that is, when the tax base is spread across the country in a way that is proportionate to population. What kind of rule is that? Why are direct taxes even mentioned in the Constitution, if they would, by definition, meet the apportionment requirement anyway and, for that matter, satisfy the uniformity rule that applies to indirect taxes—“duties, imposts, and excises”? (To be sure, capitation taxes, specifically denominated as direct taxes in the Constitution—and confirmed by Hylton—are close to being automatically apportioned, if it’s understood that a capitation is a lump-sum head tax. Even if that understanding is right, however, something I question, apportionment wouldn’t have been meaningless for a capitation when the Constitution was ratified.) In form—in ordinary language usage—the apportionment rule is a limitation on congressional power. It’s because the results of apportionment might be absurd that Congress generally shouldn’t enact direct taxes, particularly those with sectional effects. (Direct taxes aimed at property value concentrated in one part of the country are supposed to be disfavored.)
A special rule limiting Congress’s power was necessary because direct taxation was thought to be much more dangerous than indirect taxes, as Hamilton argued in Federalist 21.
It is a signal advantage of taxes on articles of consumption [i.e., indirect taxes] that they contain in their own nature a security against excess. They prescribe their own limit, which cannot be exceeded without defeating the end proposed—that is, an extension of the revenue. ... If duties are too high, they lessen the consumption; the collection is eluded; and the product to the treasury is not so great as when they are confined within proper and moderate bounds. This forms a complete barrier against any material oppression of the citizens by taxes of this class, and is itself a natural limitation of the power of imposing them.
In contrast, direct taxes contain no “natural limitation” on their use, and an explicit limitation on Congress’s authority to enact them was therefore thought to be desirable.
In a branch of taxation where no limits to the discretion of the government are to be found in the nature of the thing, the establishment of a fixed rule, not incompatible with the end, may be attended with fewer inconveniences than to leave that discretion altogether at large.
And, as Hamilton wrote in Federalist 36,
An actual census or enumeration of the people must furnish the rule, a circumstance which effectually shuts the door to partiality or oppression. The abuse of this power of [real-estate] taxation seems to have been provided against with guarded circumspection.
The Johnson understanding turns the apportionment rule on its head. His position also implies that the founders, those giants, were engaged in subterfuge: drafting and ratifying a document that in form seemed to constrain the direct-taxing power, but using words not intended to have meaningful effect. Does anyone other than Calvin Johnson think the Constitution would have been ratified if it had been understood that the taxing power was nearly unlimited?
Professor Johnson quotes and discusses the direct-tax clause in Article I, section 9, but there’s a second direct-tax clause in the Constitution. (Some commentators and Supreme Court justices seem to forget this.) In Article I, section 2, apportionment of direct taxes is tied to apportionment of representatives in the House: “Representatives and direct Taxes shall be apportioned among the several States which may be included within the Union, according to their respective Numbers.” That language makes it hard to conclude that the apportionment rule was intended to be meaningless. A serious compromise was involved: states entitled to more representation in the House because of their large populations would also bear a larger share of any direct-tax liability.
Madison made this point in Federalist 54. After noting that population would be used to govern both representation and direct taxation, he stressed that the rules are “by no means founded on the same principle.” For representation, the use of population protects personal rights; for taxation, it serves as a measure, however imperfect, of wealth and contributions. The tension between the two principles, despite the “common measure,” is a good thing.
As the accuracy of the census to be obtained by the Congress will necessarily depend … on the disposition, if not on the co-operation of the States, it is of great importance that the States should feel as little bias as possible to swell or to reduce the amount of their numbers. Were their share of representation alone to be governed by this rule, they would have an interest in exaggerating their inhabitants. Were the rule to decide their share of taxation alone, a contrary temptation would prevail. By extending the rule to both objects, the States will have opposite interests which will control and balance each other and produce the requisite impartiality.
It’s only if the apportionment rule was thought to have teeth that there was a real compromise.
The founders nevertheless understood that, even with no natural limitations to prevent congressional overreaching, direct taxation would sometimes be necessary. For example, at the Virginia ratifying convention, James Madison emphasized that national defense requires extraordinary taxing powers in times of emergency.
When ... direct taxes are not necessary, they will not be recurred to. It can be of little advantage to those in power to raise money in a manner oppressive to the people. … Direct taxes will only be recurred to for great purposes. … [I]t is necessary to establish funds for extraordinary exigencies, and to give this power to the general government; for the utter inutility of previous requisitions on the states is too well known.
Madison also noted that, if war came, when revenue needs would dramatically increase, imports—and therefore imposts—were almost certain to fall. Additional sources of revenue would be essential, so direct taxation had to be a possibility, even with arguably “unjust” consequences in the short run.
Of course, it has turned out that apportionment has been uncommon. No apportioned tax has been enacted since 1861, and the 1913 ratification of the Sixteenth Amendment made an unapportioned income tax clearly constitutional—regardless of source of income—lessening the need for consideration of other direct taxes when emergencies arise.
But that doesn’t mean it was assumed at the founding that apportioned direct taxes would never be enacted. (Remember that Congress did enact several apportioned direct taxes in wartime or in anticipation of war.) And it doesn’t mean that the apportionment rule has had no effect. One reason that we haven’t had a national property tax since 1861 is that apportionment has been understood to be required. That should give Professor Johnson (and Senators Warren and Sanders) pause.
Here’s where the Johnson article is at its misleading best. Although he claims that “Supreme Court decision[s]” support his position, he makes no reference to Chief Justice John Roberts’s controlling opinion in the 2012 Obamacare case, National Federation of Independent Business v. Sebelius. That’s a shocking omission, and this dereliction of duty is evident as well in the two letters from prominent academics (including Professor Johnson) that Senator Warren made public with her proposal—letters supporting the proposition that an unapportioned wealth tax would pass constitutional muster. Those letters also ignored NFIB, as did the press release announcing the Sanders proposal.
In NFIB, the Court concluded that the individual mandate penalty as originally enacted in the Obamacare legislation—a penalty for failure to acquire suitable health insurance—was authorized as a tax under the Taxing Clause, and it wasn’t a direct tax that would have to be apportioned. (If apportioned, the penalty couldn’t have worked as intended.)
The Chief Justice cited, with approval, Hylton’s narrow definition of direct taxes—land taxes and capitations—and noted that the Court in the 1895 income tax cases, Pollock v. Farmers’ Loan & Trust Co., had extended the Hylton understanding to include a tax on any property, not just real estate, in the category of direct taxes. The Chief described the history:
That narrow view of what a direct tax might be [i.e., capitations and land taxes] persisted for a century. In 1880 [in Springer], for example, we explained that “direct taxes, within the meaning of the Constitution, are only capitation taxes, as expressed in that instrument, and taxes on real estate.” . . . In 1895 [in Pollock], 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. That result was overturned by the Sixteenth Amendment, although we continued [in Macomber, in 1920] to consider taxes on personal property to be direct taxes.
Taxes on property, both real and personal, thus remain direct, and this passage from NFIB wasn’t dictum. The Chief was writing for a Court majority, and construing the meaning of “direct tax” was necessary to the result. The penalty may have been a valid exercise of the taxing power, but if it were a direct tax, it would have been struck down because it hadn’t been apportioned.
How can commentators on the constitutionality of a wealth tax, a tax on property, ignore this critical passage from the Roberts opinion? Yes, the direct-tax issue in NFIB wasn’t fully briefed and argued. The Chief may have been winging it, with little help from his clerks; the apportionment rule isn’t a staple of law school curricula. And yes, the four justices who joined this part of the Chief’s opinion (Breyer, Ginsburg, Kagan, and Sotomayor) weren’t enthusiastic about the taxation analysis. They thought the individual mandate penalty was authorized by the Commerce and the Necessary and Proper Clauses, and it should have been unnecessary to decide whether the penalty was an exercise of the taxing power. They went along with the Chief because that was the only way to keep the individual mandate penalty from being struck down.
But the lack of enthusiasm for Chief Justice Roberts’s position doesn’t make the holdings in NFIB go away. You can argue about the NFIB analysis and explain why you think the Chief got it wrong. You can argue that Pollock was wrongly decided and should be given no weight today. You can argue until you’re blue in the face that the apportionment rule was a mistake to begin with. (It was a clunky way to limit the taxing power.) But, as a good lawyer and academic, you can’t ignore contrary authority from the Supreme Court, in a case decided less than a decade ago, about the meaning of “direct tax.” Ignoring NFIB borders on the disingenuous.
Enacting an unapportioned wealth tax without consideration of the relevance of NFIB would be foolhardy. Doing so might be seen as a direct attack on the Supreme Court, and that wouldn’t be helpful if the wealth tax is challenged in court (as it would be).
One final point is relevant here. The Sixteenth Amendment, ratified in 1913, did exempt “taxes on incomes” from apportionment, making the modern, unapportioned income tax possible. Even if an income tax is a direct tax, as the Supreme Court had concluded in 1895, it therefore needn’t be apportioned. But a wealth tax is unaffected by the Amendment; it’s a tax on property, not income. The income tax was directed at the wealthy, but late nineteenth-century and early twentieth-century debates regularly distinguished taxes on income from taxes on wealth. And Senator Norris Brown of Nebraska, who in 1909 introduced the resolution that ultimately became the Amendment, refused to extend the amendment’s scope beyond taxes on incomes. Many members of Congress wanted to do away with apportionment altogether—to make the meaning of “direct tax” irrelevant—but Brown said no, and he prevailed. As a result, a direct tax that is not a tax on incomes remains subject to apportionment.
