U.S. international taxation: source rules, the origin basis, and the destination basis

My last post offered a rumination prompted by my teaching a class in Corporate and International Tax Policy. This time around, the sand in the oyster (as I’d like to think) comes from teaching Survey of U.S. International Taxation.

Yesterday in this class, I was slogging through the main source rules in U.S. international tax law. These are the rules that determine, for U.S. income tax purposes, whether a given taxpayer has U.S. source income or foreign source income (FSI).

Foreign taxpayers are potentially taxable in the U.S. only on what we classify as U.S. source income. As for U.S. taxpayers, be they individuals or resident corporations, source matters because they need sufficient FSI to claim all otherwise available foreign tax credit. So U.S. taxpayers may actually not care about source determinations, if they do not have to worry about running into the foreign tax credit limitation. (In some settings, however, the factors that underlie source determinations overlap with something that typically matters a lot more – whether a given increment of income is going to be treated as U.S. source income of a U.S. entity, or FSI of an affiliated foreign entity.)

Anyway, teaching the source rules can be deadly for all concerned because the rules are so tedious and empty. They follow the usual “cubbyhole” approach of tax law – you have a bunch of categories, so for each item on your tax return you decide where to shove it, and that determines how the source question will be handled.

For example, the source of dividend and interest income depends (in the general case) on the residence of the payor. Dividends and interest paid by, say, Apple or GE yield U.S. source income, whereas those paid by, say, Tim Hortons or Siemens yield FSI.

For personal services, source depends on where you render the services. But for rents and royalties, it depends on where the use occurs. For sales of personal property (other than business inventory), it generally depends on the residence of the seller. For example, if I sell a painting the gain is U.S. source, but, if Pablo Picasso sells it, it’s FSI.

Ho-hum. A natural question to ask about these rules is why they come out as they do, and what they are trying to implement or accomplish. No short answer, of course, and even the long answers aren’t very satisfying. One of their annoying features in practice is that, in many cases, the same thing economically can be structured to fit into one cubbyhole or another.

A classic example that we discussed in yesterday’s class was the Wodehouse case. Yes, that Wodehouse – Pelham Grenville, aka P.G., aka Plum. While Wodehouse was living in the French Riviera (and/or in German prison camp after France fell in 1940), he wrote Uncle Fred in the Springtime – one of my absolute favorites, a hilarious masterpiece – and also Money in the Bank, which is great fun although not quite as top-drawer for him – and sold their North American (i.e., mainly U.S.) rights for $40,000 each. So what was the source of his income?

Economically, this really was income received for personal services. He sat there in his little French cottage (or the less commodious German arrangements that followed for him until 1945) and beavered away, so to speak, ultimately to the great joy of his many readers. This would mean that he had FSI, not taxable by the U.S. But as a matter of legal form this characterization had no chance.

A second view held that he was getting royalties for the U.S. use of the intellectual property that he had created through his labors.  This was also true, unless we adopt View #3 below, and it would mean that he had U.S. source income.

A third view held that he had sold personal property, i.e., his U.S. rights. At the time, this would mean he’d have FSI, as a foreign national selling such property. (The law has no changed since then, so that he would lose under this view because he was selling a piece of the copyright, rather than other personal property.)  But under tax law at the time, he would win if one regarded all of the North American rights, but no other rights, as sufficiently an item of separate “property” to avoid its being treated as a mere advance sale of royalties. Obviously, whenever one sells property that will yield expected rents or royalties, one is in effect selling them in one lump, and yet in some cases this works as a matter of tax characterization – e.g., to create capital gains rather than ordinary income, where that is the issue presented. (The inevitably unsatisfying line-drawing cases here assess when capital gains "carve-outs" will work for tax purposes versus not working.)

What are all these source rules even about? The framework I came up with, for purposes of trying to make it more than just a list, involved the distinction between origin-based and destination-based rules for carving up the tax base when there are multi-jurisdictional transactions.

Suppose initially that you have just one jurisdiction and no cross-border trade. So everything produced there is also consumed there. Each item’s point of origin – where it was produced – is the same as its point of destination – where it was consumed. Leaving aside the intertemporal issues raised by the choice between income taxation and consumption taxation, it makes no difference whether one taxes everything on the origin basis or the destination basis. The source of each item is the same either way.

Now suppose we allow for cross-border trade. Individuals who live in the jurisdiction now can swap some of their production for others’ production. Given trade’s reciprocity, the value of what they produce still equals (in market terms) the value of what they get to consume.  But tax bases defined, in source terms, using the origin basis and the destination basis tax bases no longer include exactly the same items. Exports but not imports are treated as domestic source via the origin basis, while imports not exports are treated as domestic source via the destination basis.

The equivalence underlies standard thinking about international trade. For example, export subsidies are pretty much the same as import tariffs. And this often has tax policy implications. For example, a destination-basis VAT is not distorting trade by reason of its exempting exports and taxing imports. By contrast, an origin-basis income tax that departs from its standard approach by including targeted export subsidies is getting just what it deserves when the World Trade Organization strikes down the subsidies.

How do the source rules relate to this? To some extent, they can be divided into those that are (at least kind of) origin basis, and those that are destination basis.

The rule that the source of dividend and interest income depends on the residence of the issuer makes these what I would call fake origin-basis rules. They’re origin basis in the sense that where the money came from – the residence of the counterparty that paid it to you, i.e., where it “originated” or the use of the underlying funds occurred – determines the source. What makes these rules only fake origin-basis is that there are need not actually be any significant connection connection between the payor’s formal residence (e.g., as a legal entity) and any actual set of facts about where the associated use of the underlying funds occurred.

The rule for personal services is clearly an origin-based rule. Wodehouse wrote his comic masterpieces in England and then France (before moving ultimately to Long Island), so that’s where the production occurred, and then his work was exported to the U.S. among other markets.

The rule for rent and royalties is, by contrast, a destination-based rule. Revenues from consumer use under the U.S. copyright to Uncle Fred in the Springtime would face a well-designed destination-basis U.S. VAT, but would not face income taxation here if we relied on where the production activity occurred.

Finally, the rule for sales of personal property is probably best-viewed as origin-based, insofar as it’s actually one or the other. It looks at the person who sold the property, and who thus perhaps “produced” the gain from sale (even if only in the sense of picking something that would appreciate in value). In a Wodehouse-type case, of course, this is especially clear, as he actually created the property that he is selling through his personal efforts.

It’s a truism that, for reasons I’ve discussed elsewhere, an income tax pretty much has to use the origin basis as its main method, whereas a consumption tax can use either the origin-basis or the destination-basis.  But nonetheless real world income taxes often use destination-basis rules, such as when determining the source of income from cross-border transactions. A good example, apart from certain of the source rules that I’ve discussed above, is the use of sales factors in formulary apportionment. These cause a business that is active in multiple jurisdictions to be taxed, in a given jurisdiction, based at least partly on its sales to consumers and others in that jurisdiction.

Why does the U.S., along with other countries in their source rules, build as much destination basis as it does into its source rules? Well, suppose initially we were thinking of this in a standard international trade context. Here it’s a bit like having an import tariff, on top of taxing domestic production even when exported. Import tariffs can be domestically popular, as a political matter, even when they’re good policy. But they can actually be good policy, from the standpoint of residents’ economic welfare, where the jurisdiction has market power, e.g., because importers would be enjoying rents (in the economic sense, as distinct from that of “rents and royalties” under the source rules). So one isn’t surprised to see, say, the U.S. adopting rules that might permit it to tax P.G. Wodehouse on his work in that French Riviera cottage, given that it led to the situation where U.S. consumers would pay for the reader’s privilege (at zero extra marginal production cost to him).

Does our mix between origin-basis and destination-basis source rules mean that we are effectively imposing tariffs on certain imports? Perhaps in some cases, but my sense of the rules’ overall tenor is somewhat different, for three main reasons.

First, destination-basis rules tend to apply to outbound as well as inbound transactions. Computer engineers in California who design IP to generate rents and royalties abroad would therefore be generating FSI even without access to the full panoply of tax planning tricks that have flourished in the last couple of decades.

Second, to the extent that different countries measure source consistently, the importer’s domestic source income under a destination-basis rule will be FSI in the exporting country. In such a case, the latter country may offer exemption or foreign tax credits that eliminates “double taxation” (or, more meaningfully, combined relative over-taxation).

Third, by structuring carefully, taxpayers have considerable ability to decide which rule will apply to their business income. Add in all the other tax planning opportunities that they have, and “stateless income” may loom considerably larger as an issue than tariffs. Indeed, even just the ability to choose between origin-based and destination-based rules may significantly move the effective overall regime in that direction.