Yesterday at the Tax Policy Colloquium, Rebecca Kysar of Fordham Law School presented Unravelingthe Tax Treaty. Here are some of my thoughts regarding this very interesting paper and the issues it raises.
Bilateral tax treaties are the subject of a flourishing sub-literature that has often, and perhaps increasingly, criticized their impact on developing countries in particular. The paper explores extending this critique to the United States, e.g., because, as a net capital importer these days, we may tend to lose current tax revenue from treaties’ standard practice of having the two treaty partners mutually surrender source taxing jurisdiction with respect to the other’s residents.
For purposes of thinking about the issues that the paper raises, I find the following outline of main treaty provisions helpful. Bilateral tax treaties frequently include the main six types of provisions, among others:
1) Residency/LOB rules – Treaty benefits are accorded to residents of the two treaty partners. However, reflecting corporate residence’s inherently limited meaningfulness as a concept, there may be “limitation on benefit” (LOB) rules that aim to defeat “treaty-shopping” by denying treaty benefits to, say, entities that are residents of one of the jurisdictions but that are functioning as mere conduits.
2) Permanent establishment (PE) rules – Treaty partners generally agree not to tax each others’ residents’ business profits on a source basis in the absence of a “permanent establishment” (PE) in the source jurisdiction – e.g., an office with dependent agents working there.
3) Withholding taxes – Countries frequently tax domestic source dividend, interest, and/or royalty income that is paid to nonresidents. Given the collection and enforcement problems that might otherwise arise, this is done via withholding taxes that have a fixed rate and are gross basis (i.e., no deductions are allowed). For example, the U.S. by statute has a 30% withholding tax. In treaties, however, countries may reciprocally agree to a charge a lower withholding tax rate (e.g., 15%, 5%, or even 0%) to each other’s residents.
4) Anti-“double taxation” rules – I’m putting “double taxation” in scare quotes here for a reason. As I discuss below, the concept at issue here is perhaps better described other than as one of double taxation. But there can be good reason for one not to like the result whereby cross-border investment is tax-discouraged, such as by reason of its being fully taxed in both the residence jurisdiction from whence it came, and the source jurisdiction to which it went. Treaties may address this problem by requiring that each treaty partner, with respect to its residents, either exempt foreign source income (FSI) earned in the other jurisdiction, or grant foreign tax credits with respect to the taxes imposed by the source country. The tax treaties also typically provide mechanisms for achieving consistent treatment if otherwise each country would treat the same income as earned within its own borders.
5) Nondiscrimination – Treaty partners may agree not to have tax rules discriminating against each others’ residents for tax purposes (although the provisions tend to be quite narrowly drawn).
6) Information exchange, etc. – Treaties also contain lots more types of provisions, which I will here cavalierly funnel into a catch-all, albeit noting in particular information exchange between the two sovereigns as it’s an important aspect that treaty critics often praise.
A number of further common treaty provisions might merit separate coverage due to their potential importance. For example, treaties may indicate that one should apply an arm’s length standard to transactions between commonly owned affiliates, supporting the use of transfer pricing. But these provisions are not as extensively discussed in the Kysar paper as those listed above.
The paper’s main arguments are to the effect that (a) 2 and 3 are bad, (b) 4 and 5 are poorly defined and/or not needed, and (6) can be done separately. For reasons of time and space, I’ll just discuss (a) here.
WHY CEDE SOURCE-BASED TAX JURISDICTION RE. NON-PE BUSINESS INCOME AND PASSIVE INCOME?
Features 2 and 3 above cede source-based tax jurisdiction with respect to non-PE business income and passive income. This leads to the question: Why would one ever reduce optionality / future flexibility by ceding something in advance? There are 3 main answers that one could offer in this context, pertaining in turn to pre-commitment, reciprocity, and coordination.
Pre-commitment – Retaining discretion for future policy choices is a bad thing if one is sufficiently likely to misuse it. (This of course is the “tie Odysseus to the mast” line of argument.) Governments may benefit, for example, from committing in advance against ex post expropriation of inbound investment. In the realm of free trade, if we posit that tariffs are generally a bad idea, but are too often produced by protectionist forces even when one doesn’t have a malignant clown at the helm, free trade treaties serve to pre-commit countries to do what they would benefit from doing anyway.
The best argument for applying that line of reasoning here is that small open economies may fail to benefit from taxing inbound capital. Absent market power, local rents, etc., the tax is likely to be borne by locals even if it is nominally paid by the foreigners, and may impose greater deadweight loss than the alternative tax instruments that might have been used instead. But the optics of literal tax payment by the outsiders may be unduly enticing.
This line of argument seems more likely to apply to withholding taxes on passive income, than to non-PE business income. For example, foreign multinationals with valuable IP or trade names may be in a position to earn local rents, so one might want to tax them without regard to whether they have been able to avoid crossing the PE threshold.
Reciprocity – A tax treaty’s two parties reciprocally recede. So one’s loss of source-based taxing jurisdiction is offset by the other side’s so receding with respect to one’s own residents.
In a pure symmetry case where investment flows, income earned, tax rates, etc., are completely the same, the revenues foregone equal those that the other side foregoes with respect to one’s own residents. The paper notes that, if net capital importers tend to lose in revenue terms (at least on a current basis) from the asymmetry between the value at stake on the two sides to the deal, that is an issue not just for developing countries, but also for developed net capital importers like the U.S. Agreed that this is relevant to the analysis.
Coordination – If countries attach positive value to coordinating their tax systems with each others’, and thereby avoiding peculiar interactions or combined effects, then tax treaties may offer a valuable coordination device. There has increasingly been concern that countries’ go-it-alone responses to profit-shifting concerns (e.g., digital services tax, UK diverted profits tax, the BEAT, etc.) may interact with each other in undesired ways. But admittedly existing tax treaties may not help all that much in coordinating responses.
How much difference does it make? – To the extent that the U.S. is losing tax base by reason of treaty concessions with respect to non-PE businesses and withholding taxation, it’s worth noting that the treaties’ marginal effect is seemingly reduced by relevant aspects of domestic U.S. tax law. For example, we don’t tax inbound business income on a source basis unless there is a “U.S. trade or business.” The legal concept here overlaps considerably with that of finding a “permanent: establishment.” Likewise, even if we presume that 30% is the truly “intended” withholding tax rate (rather than serving in part merely to give us something to negotiate away on behalf of U.S. taxpayers), the extent to which it can be avoided by, say, using derivative financial instruments (such as notional principal contracts) instead of directly realizing income that is subject to withholding tax – and the extent to which (despite recent regulatory tightening) this might be intended and reasonably so given the “small open economy” issue – is worth keeping in mind.