As noted yesterday, the US does not tax foreign income of foreign corporations, even when the foreign corporation is a wholly-owned subsidiary of a US corporation. Under this regime, a US parent corporation with a subsidiary in a low-tax (tax haven) jurisdiction has an incentive to park income in the tax haven by (i) paying the subsidiary too much for goods and services and (ii) charging the subsidiary too little for goods and services. The US polices this with two mechanisms: First, some very mobile income, like royalties, of a controlled foreign corporation is taxed to US shareholders as earned by the foreign corporation. Second, there are rules that attempt to insure that parent/subsidiary transactions are priced at “arm's length.” The arm's-length rules don't work, which calls the entire regime into question.
The arm's-length standard fails both in theory and in practice. The theoretical problem probably doesn't seem all that compelling (except to a hand full of theorists), but it helps understand the practical problems. Here is the idea: Income is a measure of individual wellbeing. It is not earned in any place or by a legal entity. This was illustrated in the Sue hypothetical. She performed her surgeries offshore. But, in the abstract, the earning process began the day that she was born. To try to figure out what she earned offshore –- to try to break a seamless earning process into artificial pieces –- seems doomed to failure.
And fail it does. To figure out an arm's-length price, one needs a comparable transaction between unrelated parties. In every interesting case, there is no useful comparable. Economists are beginning to figure out why. Multinational corporations are more complicated than initially thought. They are more than the sum of their parts. (Like The Beatles, I like to say.) This violates traditional economic theory, but the empirical work bears out that there is extra value in multinationals. Thus, comparables may not exist, as multinational groups can do things that a collection of independent local companies cannot. Also, a multinational has extra profit from being a multinational, which is not attributable to any local piece, and which is missed in traditional arm's-length pricing. (But yes, tax jocks, is dealt with toward the end of the 1994 regs somewhat.)
Tomorrow: Formulary apportionment
ah, now you are getting to the real problem areas, i.e. how to deal with offshore tax evasion schemes.
to me, the question is not so much whether such wholly owned foreign subsidiaries should be subject to US taxation, but how one discriminates between legitimate subsidiares that are a necessary part of doing business in many foreign nations, and those subsidiaries whose existence is predicated entirely upon avoiding taxation.
But again, to me the bottom line is that the US (and every nation) should regard foreign taxation as a business expense, no different from rent, or utility bills.
The whole “we can’t tax businesses or individuals twice” argument seems to be based on the idea that international trade is an absolute good that must be encouraged by any means possible. I don’t buy that—nations are far better off the more self-sufficient they are, and international trade leads to a lack of self-sufficiency that when times get bad exacerbates economic (and the resultant social) problems.
Paul:
Even a diehard free trader such as myself is not convinced by the double taxation argument in this context for a simple reason. Multinationals need not face double taxation on cross-border transactions precisely because of tax treaties and the OECD insistence that the arm’s length standard be the guiding rule for all national tax authorities – whether the IRS, Revenue Canada, Mexico’s SAT, UK’s Inland Revenue, or even Japan’s NTA. As one might have been able to tell, I have a little more faith in the arm’s length standard than George does. But I hope to post over at Angrybear something on transfer pricing, cost sharing, and employee stock options that suggests that it is the poor enforcement of section 1.482 that is the failure – not this notion that the benefits of integration (which George notes) creates a fundamental fatal flaw. Give me a couple of hours and check the Angrybear for another twist on this.
—–
Pingback: Hobson's Choice
Pingback: Hobson's Choice
Pingback: Hobson's Choice
Pingback: Hobson's Choice
Pingback: Hobson's Choice
Pingback: Hobson's Choice