Yesterday at the colloquium, Ruth Mason discussed her paper, Citizenship Taxation. I enjoyed reading it enough to conclude that I may want to write about this topic as well. (I generally agree with the paper, but it’s a rich topic and perhaps a fruitful area for me to deploy some of my interests and approaches.)
Here are some of the points from my notes that I thought worth discussing.
(1) The U.S. as outlier (again!).
Here we go again. To paraphrase Ray Davies in a 1960s Kinks song, “we’re not like anybody else.” The U.S. is the only country to impose (at least in theory) worldwide taxation on all non-resident citizens. Other countries may tax some non-residents on a worldwide basis, via the use of standards other than just current year physical presence to determine who is really still a member of the domestic community, but no one else does it flat out based on citizenship.
Other examples: we don’t have a VAT, our statutory rate for corporations is unusually high, and in the international realm we employ deferral / foreign tax credits far more extensively than anyone else.
Now, everybody else could be wrong and we could be right. Or our circumstances could be distinctive. But it is natural to wonder, given, for example, the “wisdom of crowds” line of argument.
(2) Benefit tax rationales.
The literature discussing citizenship-based taxation often employs a benefit tax rationale. On the one hand, U.S. citizens living abroad aren’t using the roads, etc. They also generally can’t get U.S. public assistance, or use healthcare subsidies that apply within the U.S., etc. On the other hand, in theory the U.S. Army is protecting them, plus they have the valuable option to return, which lots of other people might pay good money for if it were on sale.
All this might call perhaps for an intermediate U.S. tax burden on U.S. citizens living abroad - which we actually have, given section 911(b), which permits excluding about $100,000 of earned income, and the allowance of foreign tax credits. But benefit tax rationales are not widely accepted these days, as compared to ability to pay, which of course raises the core question here: Whose ability to pay?
Suppose that, by reason of my living abroad, the U.S. government saved a marginal $10,000 that it would otherwise have spent giving me services. Even without any normative attachment to benefit taxation, it might make sense to reduce the U.S. tax bill I would otherwise have incurred by $10,000, as this properly aligns my incentive regarding where to live, from the fiscal standpoint and assuming that there are no other relevant considerations. But not much of what the U.S. government does on my behalf invites this sort of marginal cost analysis.
(3) Defining “us” versus “them.”
Here is the nub of the issue. From a utilitarian standpoint, everyone in the world matters equally. Numerous other philosophical approaches agree that everyone counts the same in the relevant sense, although they have other ways of implementing the idea that everyone matters equally.
But just as individuals, rather than being perfect altruists, generally act for their own benefit and that of loved ones, so it is generally accepted that countries can care primarily just about “us” (the members of the national community), as opposed to “them” (everyone else). Indeed, just like a buyer and seller in a market negotiation, it is widely considered fine if everyone would like to extract as much money from each other as possible, albeit subject to rules of honesty and fair dealing, not doing bad things, and acting altruistically in extreme cases (e.g., rescue when you see someone drowning, seeking to prevent genocide abroad).
Let’s just take it as given that this personal, family, or community-based selfishness can be justified philosophically. Many regard it as a prudentially required exception to universal morality (we don’t have to be saints, especially when everyone else isn’t being a saint). I gather that Ronald Dworkin tried to fit it into his philosophical system more holistically, but I don’t generally find myself in agreement with his full approach.
But even taking all this as given, it is difficult to find firm normative footing for the determination of who fits into our community. Personally, I care about the people I care about (and for close family members or other associates, there’d be an argument that I ought to care even if I didn’t). But, in the impersonal setting of a national mass community, it is hard to establish definite guideposts re. what we are trying to do.
If I am a citizen who goes abroad for a month, I’m surely a full U.S. person for the year (not just 11/12). Arguably the same, whether or not we choose to impose full citizenship taxation, if I go abroad for three years but always plan to return.
If I am here illegally, I definitely should count normatively, at least to a degree. (Not to dive into the murky, for me, waters of immigration law and policy.) This is why most of us would insist on allowing illegals to get treated in the ER when they face dire medical problems. And we shouldn’t be happy about having a two-class society here, the Americans here and the others, whether the latter are lawful guest workers or illegals. But I digress. The point here is simply that residence inevitably counts – we care more about people who are right in our faces than about others – but it is not all that counts given that I can be away from what continues to be my community.
What we mean by the “us” whom we agree to care about, once we have accepted an approach that mainly limits altruism to the members of one’s own community, is most likely going to be multi-dimensional, rather than turning on a single metric such as citizenship. Specifying a legal rule to define the members of our community (who may be subject to at least some elements of worldwide taxation, even when they are abroad) may involve the usual tradeoff between complexity and accuracy. But as I discuss below, there may also be efficiency benefits to a multi-factor approach.
(4) What about “them”?
For those in the “them” category, in principle we might want to revenue-maximize – that is, get as much money from them (in real economic incidence terms) as we can. This is not diabolical; it’s just how transactional counterparties commonly deal with each other.
This of course leads to the Monty Python Principle: “Tax all foreigners living abroad.” So why don’t we try to do that? Well, obviously, it would not be a great idea to try, at least without more of a hook. We don’t have jurisdiction, we don’t have information, it would violate comity and get them really angry, they would retaliate in various ways that we wouldn’t like, etcetera. And more generally, cooperation for mutual gain is always a good idea when it can be done. But, when we turn to the treatment of “us,” it’s worth keeping in mind that there is a hypothetical perspective from which we might like to tax all foreigners living abroad.
(5) The basic paradox: if you’re “us” and we care about you, you lose.
Once we accept that we can’t, and generally won’t even try, to “tax all foreigners living abroad,” something that is at least facially paradoxical arises when we consider taxing at least some nonresident citizens on their foreign source income. This tax burdens the individual who has to pay it. So we are effectively saying: “If we care about you, we’re going to burden you. If we don’t care about you, because we have decided not to classify you as a member of our community, then we’re not going to burden you.
Now, one thing we clearly do care about, when taxing members of our community whether they are currently residents or not, is deadweight loss that we impose on them through our rules. We don’t necessarily care directly about deadweight loss that is imposed on non-members, since we are at least mainly excluding all “thems” from our social welfare function, but for “us” it matters. And this is pretty much the proof that the current U.S. regime for taxing nonresident citizens is defective and needs to be revised. It appears to impose a lot of deadweight loss, from compliance burdens and the like, relative to the revenue that is actually raised. This results, for example, from tax filing requirements who owe little or no U.S. tax (e.g., due to the earned income exclusion plus foreign tax credits). Now, we may gain valuable information from the filing (or, rather, we would if people actually filed), but this is still a serious concern.
A question meriting further thought: Why don’t we extend more transfers or other social welfare, safety-net style benefits to non-resident citizens? Perhaps there are good reasons for not doing so, but at the least it requires further thought. Presumably if we had greater tax-transfer integration, such as via a Mirrleesean demogrant / income tax system, we would at least consider giving the demogrant to non –resident “us.” Or perhaps not, if there are particular incentive issues to keep in mind here. But worth some thought in any event.
Circling back to “us” versus “them”: Suppose we can tax U.S. citizens living abroad, but we classify these individuals as “them” rather than “us.” Only, suppose that, because they are arguably “us,” we can get away with it – other countries won’t start to scream, retaliate, etcetera. Then we would have a “them”-based reason for imposing the tax (and also for withholding welfare benefits and not caring so much about deadweight loss). In short, it could be the one case of “Tax all foreigners living abroad” that we could actually pull off. And this would imply not caring so much about deadweight loss that we impose on nonresident citizens. But it does strike me as a bit harsh.
(6) Two-stage analysis for immigrants: before & after they become “us”
Let’s leave aside all the issues about illegal residents, as that’s a very different topic and neither raised by Ruth’s paper nor an area of personal expertise on my part. Instead, let’s consider people who might like to immigrate to the U.S., at least conditionally on comparing it to other places where they might choose to live, and whom we are considering admitting to our community voluntarily. They are not here yet, and suppose they will only come here if we let them.
I bring up these people because Ruth’s paper rightly expands the scope of the inquiry to them, by noting that, when we are considering the taxation of nonresident U.S. citizens (as well as green cardholders) on their worldwide income, prospective immigrants are among those whom we should have in mind. Suppose, for example, that immigrants whom we would like to attract would take note of the U.S. worldwide tax system for citizens and green cardholders, when they are deciding where to go.
Since prospective immigrants are not “us” yet, we might want to maximize the sum of (1) the present value of the money we could get by admitting them, plus (2) the positive externalities that their presence would generate for us.
In terms of (1), you can be crude about it, if you like, and charge an upfront fee for a passport. Malta and Cyprus do this, and can charge extra-high fees because they can offer EU passports. Several Caribbean countries also sell passports, albeit for less as they can’t offer an EU passport. The closest the U.S. comes to charging an upfront fee is via our EB-5 immigrant investor program.
For the most part, however, we are a bit more decorous about exchanging money for admission to our community, in that we allow immigrants – once admitted – to pay under the installment plan, rather than up-front. That is, immigrants generally can be expected to generate positive tax revenues for the U.S. by reason of paying income and other taxes once here.
We probably should admit more of these people than we do, especially if positive externalities are significant too (as I suspect they are). But from the standpoint of efficient pricing, telling prospective green cardholders in particular that at some point they would be getting taxed on their worldwide income even if they weren’t living in the U.S. might not be an optimal way to structure the “fee.” We’re telling them, “It’s going to cost you even if and when you don’t find yourself living in the U.S., which might be the scenario in which they would value it the least. I have heard it said that well-heeled foreigners who are thinking of spending significant time in the U.S., and who have access to temporary admission under various of our categories, often are advised against seeking a green card.
(7) Relevance of multiple margins (citizenship & residence)
As David Gamage notes in a paper that he presented at our colloquium last year, it is often preferable to have many small distortions, rather than a few large ones. Citizenship-based taxation may induce people to renounce U.S. citizenship. Residence-based taxation can encourage them to make sure they are out of the country for at least the requisite time each year. A rule based on domicile may discourage having one here. And so forth. But a standard that relies on several different factors may end up, with proper design, inducing less tax-motivated distortion overall. Obviously, the fiend or goblin (to avoid saying “devil,” as that’s become such a cliché) is in the details.
(8) Foreign taxes
I’ve written extensively (such as here, here, and here) against the desirability of providing full and immediate foreign tax credits in the setting of business taxation. My main argument is that foreign taxes are a cost from the unilateral domestic standpoint, hence we should want domestic taxpayers to be foreign tax cost-conscious. Indeed, straightforward deductibility for foreign taxes (so they will be treated as equivalent to other outlays or forgone inflows) is the way to go, from a unilateral national welfare standpoint, unless there is more to think about here. As it happens, I agree that there actually is more to think about in the setting of corporate income taxation, e.g., because if profits show up in a tax haven that may be a “tag” for discerning profit-shifting away from home, whence my conclusion that, for U.S. companies, we might favor a better-than-deductibility, worse-than-creditability bottom line effect. Well-designed anti-tax haven rules can have this effect.
Exemption is an implicit deductibility system. Of course, if you limit exemption for the foreign source income of resident companies via anti-tax haven rules, you may get back into that intermediate range (or even beyond, if the rules are poorly designed)
How do these arguments apply to taxing nonresident individuals on their foreign source earned income? While the topic may require further thought, my initial thought is that a similar basic analysis does indeed apply. Suppose, for example, that an American is deciding whether to work abroad in lower-tax Singapore or the higher-tax UK. It may be desirable, from a U.S. standpoint, if he or she bases the analysis on after-foreign tax income, rather than before-foreign tax income. After all, from our standpoint those taxes are just a cost, as we don’t get the tax revenues.
While it might be the case that the use of tax havens at the expense of the domestic tax base is not as much a problem for individuals’ earned income as it is for companies’ reported profits, that would only push harder towards the deductibility, as opposed to creditability, side of the spectrum. And even if we treated foreign taxes on U.S. individuals’ foreign earned income as effectively deductible, rather than creditable, then (assuming a treaty-compliant design, which might not be impossible) we might decide to apply a significantly lower tax rate to foreign than domestic earned income, e.g., for the reasons I will discuss next.
The exclusion under Code section 911(b) for about $100,000 of qualifying foreign earned income has the desirable effect of pushing away from effective creditability. You can’t claim foreign tax credits with respect to foreign earned income that we don’t tax. But that doesn’t tell us whether zero is the right rate (perhaps even without being so capped), or is too low a rate. That depends on other considerations.
(9) Foreign earned income
Even leaving aside the treatment of foreign taxes, should we tax U.S. individuals’ foreign earned income at the full domestic rate? An “ability to pay” perspective would suggest that the answer is yes. But there may be an efficiency argument (from a national welfare perspective) in favor of a lower rate. This argument is borrowed from Mihir Desai’s and Jim Hines’ analysis of “national ownership neutrality” (“NON”).
Suppose the following, as Desai and Hines do in the NON context: Each “outbound” dollar is replaced by an “inbound” dollar from a foreigner who earns income that is taxable in the U.S. by reason of the outflow. Translating it from their context to mine, let’s take a simple case, grossly overstated by me just so one can grasp the argument. Suppose that, if I leave NYU for a year to teach at a foreign university (which will pay me in lieu of NYU’s doing so), NYU will hire a foreigner to teach in the U.S. in my place, and pay U.S. tax on the substitute salary. At a rough approximation, even if the U.S. gave me an incentive to visit abroad by not taxing my foreign earnings (and again, the U.S. may have no direct reason to view payments to the foreign tax authority as equivalent to paying U.S. tax), it would not have lost any domestic tax base. U.S. salaries would be the same, and the IRS would be taxing that foreigner, who would otherwise have escaped its clutches.
Suppose we take this hypothetical to be appropriate – as Desai and Hines do in the setting of cross-border business investment. (They of course do not make any such claim, or at least they haven’t yet, with regard my hypothetical, which they might join other readers in finding fanciful.) They then argue that the optimal U.S. tax rate on the foreign source income is zero, so that there will be no distortion at the margin of U.S. companies deciding how much to invest abroad rather than at home. (Again, the claim is that the lost domestic investment is actually zero, net of the inbound flows it induces that would not otherwise have occurred.)
I think this argument is wrong, because we are trying to balance distortions at multiple margins, not reduce them to zero at some arbitrarily chosen margins while they remain high at other margins. But I will go so far as to agree with Desai and Hines that it may support a lower U.S. tax rate on resident companies’ foreign source income than that on all companies’ U.S. source income.
Does this argument apply to foreign earned income of U.S. individuals? I suspect that the extent to which it applies is greater than zero, albeit less than 100 percent. Americans who work abroad may indeed often leave slots that get filled by someone else, including a foreigner. (There may also be a positive spillover if the result is to increase other Americans’ domestic earnings by reason of the slot I left open for the year.) So there may be an argument for taxing foreign earned income of U.S. individuals at a lower rate than their U.S. earned income, even wholly leaving aside the issue of how to treat foreign taxes.
There may also be positive externalities when U.S. individuals work abroad. One hopes that they are effective ambassadors, spreading goodwill abroad towards those of our ilk, rather than being the stereotypical “ugly Americans.” Plus, if they learn about life abroad and then return here, the things they learn may enrich life for other Americans. One of the absolutely greatest things about our country is the extent to which we’ve been a global melting pot, not only because people come here from abroad and are accepted, but also because it makes us (at least on the coasts) more cosmopolitan and international, to our great hedonic benefit. Or at least that’s how I view it.
(10) Exit tax
Code section 877A generally requires tax expatriates to pay tax on a deemed sale of their assets for fair market value. The tax only applies to net gain in excess of $600,000. This exit tax makes up for the fact that, once you’re gone, the built-in gain from the period when you were still a U.S. citizen is unlikely ever to be taxed here. So one could view it as anti-windfall gain, rather than as aiming to be punitive.
That characterization of the provision would be more solid and beyond dispute if we taxed net asset appreciation at death, rather than allowing a tax-free step-up in basis at that time. But, for what it’s worth, the provision can indeed reduce lock-in on the part of individuals who are considering expatriating.
Here is a further issue that might be worth thinking about. Suppose someone who might owe estate tax liability at death expatriates before that point. Proponents of the estate tax might support treating expatriation as a triggering event for levying the tax. Not just because "You're dead to us now," but to avoid creating the incentive to expatriate for that reason.