What Kind of Constitutional Mess Might a Wealth Tax Create?

by Neil H. Buchanan

Republicans and some nominally non-Republican billionaires are becoming increasingly alarmed by the popularity of proposals to increase taxes on the rich.  More accurately, they are apoplectic because many Democrats have finally started to take seriously progressive tax proposals that have in fact been popular for a long time.

In my Verdict column today, "Can We Tax Wealth? Yes, and Even if Not, Still Yes," I gingerly wade into a constitutional claim that some conservatives have floated recently, which they believe will allow the Supreme Court's newly fortified conservative bloc to invalidate anything that a future -- perhaps very near-future -- Democratic president and Congress might pass.

I will get into as few details as possible here, allowing readers with further questions to link to today's column, but the short version is that the Constitution requires that direct taxes be apportioned whereas indirect taxes need not be apportioned (but must be uniform).  Do you suspect that the meanings of each of the italicized terms is highly contestable?  Ubetcha!

Again, my purpose here is not to dive back into the constitutional details that I cover in the column (although some of that is necessary) but rather to explore how a victory on what conservatives view as their devastating argument would actually lead to something unexpected and objectively worse than a standard wealth tax, though still better than the status quo.  The bottom line is that we definitely can have a wealth tax, but conservatives might force us to choose the least sensible version of one.

The somewhat odd phrasing of the title of my Verdict column attempts to capture my two-part response to a one-part argument from conservatives.  They claim that a wealth tax, such as Senator Elizabeth Warren's recently proposed Ultra-Millionaire Tax, would be a direct rather than an indirect tax (and not an income tax, which is covered by the Sixteenth Amendment)  Warren's critics seem to think that if their assertion is true, there can be no wealth taxes at all.

I point to persuasive arguments that wealth taxes are indirect taxes or are a type of income tax.  The estate tax, after all, has never been struck down (and has been in effect since 1916), and it is merely a non-annual version of Warren's plan.  But if my column has any added value, it is in taking the second step, noting that even if a wealth tax is a direct tax, it is still allowable because the Constitution explicitly contemplates direct taxes so long as they are apportioned.

If all (or even any) of these terms were self-defining, the debate would be an easy one; but because direct/indirect, apportionment, and uniformity are terms of art at best and incoherent at worst, readers can be forgiven for being somewhat lost at this point.  Luckily, my argument from this point forward relies only on the concept of apportionment.  That is, I take as a starting point that conservatives have won the first step, with their friends on the Supreme Court declaring that the wealth tax is a direct tax.

Presumably, the Court would then say that Warren's tax is not apportioned and thus is unconstitutional, but the latter part of my column today points out (relying on some interesting scholarship by Professor Dorf) that Warren could add a "fallback" version of a wealth tax to her bill.  That is, she could say: "We hereby enact x, but if the Supreme Court holds that x is a direct tax and must be apportioned, we instead enact y," where y is an apportioned tax.

Even if she did not do that, Congress could respond to a loss in the Supreme Court by passing a new, apportioned version of the wealth tax.  That involves a fairly large amount of waiting time (at least a year, and maybe more), however, so anticipating this would be valuable.

OK, but even though I have set aside the direct/indirect mess, that still leaves us with apportionment, which is also not an obvious concept and is not defined in the Constitution.  As I note in today's Verdict column, part of the problem is that the passage in 1913 of the Sixteenth Amendment has made it unnecessary for people to know what direct/indirect and apportionment mean.  No one seems to be able to offer even the beginnings of a coherent explanation that reconciles various existing types of taxes.

And I am genuinely not kidding about the confusion.  Although one must always concede that such statements must at least implicitly include the qualifier "as far as I can tell, and I might be wrong," this is an area of law that is easily as incoherent as any other legal question that I have come across.  Habeas?  Complicated decision trees, but ultimately logically coherent (although not as a matter of policy).  ERISA?  Again, complicated but not unknowable.  Maybe the Dormant Commerce Clause comes closest, but in any event, this is weird stuff about which plenty of smart people cannot provide clear answers (even setting aside ideology).

What, then, is apportionment, and how would it apply to a wealth tax?  The basic idea is that the Constitution requires direct taxes to be collected from the states in proportion to the population of the states.  (This happens indirectly as a technical matter, because the provision in question ties apportionment to representation in the House of Representatives.  But House representation is based on population, so QED.)  If California has 12 percent of the population, then an apportioned tax must collect 12 percent of its total revenues from California.

Upon a moment's reflection, this is obviously a less than optimal way to set up a tax.  Californians own a much larger proportion of the nation's wealth than a mere twelve percent, and New York and Florida are similarly going to be in the group with more wealth than people (proportionally).  Similarly, even though states like Mississippi, Iowa, and West Virginia are lightly populated, their share of the nation's private wealth is almost certainly even smaller.

In order to satisfy the Constitution's apportionment requirement, then, a wealth tax would have to impose higher rates on the wealthiest people in poorer states and lower rates on the wealthiest people in richer states.  So long as the proportions end up aligned, the tax will pass constitutional muster.  And because the Constitution only requires uniform tax treatment across states for indirect taxes, that would not be a barrier in this hypothetical world where Warren has already lost the direct/indirect battle in the Supreme Court.

As I noted above, however, this is a less-than-ideal (one might even say silly) way to design a wealth tax -- and the incoherence would almost surely be as offensive to conservative ideologues as to anyone else.  Consider just three oddities:

(1) If the wealth tax is set at lower rates in some states than others (as it would have to be), that will create incentives for wealthy people to move (at least as a matter of form) to low-tax states.  That kind of concern, which conservatives often tout regarding California's high income taxes (usually contrasted with Florida's lack of an income tax), turns out not to be empirically large.  After all, California is not emptying out, and although Florida has a lot of billionaires, other low-tax states like Mississippi are not hot destinations.

Even so, it is notable that this system would effectively require cross-state differences in a federal tax.  At the very least, this would be a Full Employment Act for public finance economists, who would lustily investigate any possible effects of these differences.  Those empiricists could compete with constitutional tax lawyers for the title of "group most inadvertently benefited by the Supreme Court's ruling."

(2) Warren's plan imposes a two percent rate only on wealth above $50 million and an additional one percent surtax on wealth in excess of $1 billion.  I do not have state-level wealth distribution tables in front of me, but I suspect that there are many states where the numbers of people with wealth exceeding those thresholds are quite small.

Depending on those numbers, it might actually be possible that Warren could only set up an apportioned tax by lowering the thresholds, because only then could she find enough people in the poorer states from whom to collect sufficient revenues to meet the apportionment targets.

That would probably be just fine with Warren.  After all, she is not talking about putting a two percent tax on a person with a net worth of $50 million.  She would tax the excess above $50 million.  (Similarly, the billionaires' surtax is only on amounts above the first billion.)  I would imagine that she did so to be able to argue that she was only aiming at the 0.1 percent, which her press release emphasized.

If apportionment effectively requires lowering the thresholds, then, Warren could honestly say that she did not want to tax mere thirty-millionaires but was forced to do so by the apportionment rules.  Is this the big win that conservatives would be looking for?

(3) In part because of the mobility concerns in (1) above, and in part simply because wealth (like all other economic variables) is not perfectly predictable, there will be an ex post versus ex ante problem.  If, for example, the Treasury sets rates for the year based on predictions of who will have to pay and how much, we could find ourselves at the end of the year with higher or lower revenues from some states than anticipated.  Ex post, then, the tax will not be apportioned.  Is that acceptable, or does the government then have to refund some money to the taxpayers in the wealthier-than-expected states?

Texas Law's tax professor Calvin Johnson has suggested, based on an early Supreme Court case, that these types of concerns tautologically prove that most taxes cannot be direct taxes.  That is, he says that the constitutional requirement of apportionment necessarily limits what the Framers meant by the term direct tax: non-apportionable taxes simply cannot be direct taxes, because the Constitution would not have told us to do the impossible.

There is much to Johnson's argument, and the process of sorting out a wealth tax (whether Warren's or any other) might well lead courts to that conclusion.  After all, if Warren writes a fallback apportioned wealth tax and people get a chance to look at it, they might well conclude that there is no coherent way to set up the rules of such a tax.

And even if that does not convince certain conservative ideologues, it would be educational to the public to be able to say: "Here is a proposal for a wealth tax, which polls suggest that large numbers of you support.  And here is what we would have to do instead, if this lawsuit against us succeeds.  That would still be better than allowing inequality to continue to worsen, but can't we agree to do this in the most coherent way possible?"

Again, this analysis is an attempt to make sense of a miasma of unsettled legal principles.  Even more than usual, the caveat, "as far as I can tell, and I might be wrong," is apt.