I'm back in the NYC area after 4 quick days at the annual summer conference of the Oxford University Center for Business Taxation.
The travel part had its annoying moments. The Dial 7 Carmel car service told me I needed to assume that the ride from Greenwich Village to Newark on the evening of July 4 would take 2 hours. I seriously doubted this, but figured they must know best. Actual trip time: 20 minutes. Very few people leave NYC through the Holland Tunnel or fly out of Newark on a holiday. So I got to Newark more than 3-1/2 hours before my scheduled overnight red-eye flight. Thanks a lot, Dial 7.
The way back was awful in the opposite direction. It took me nearly 4 hours to get home from Newark, counting from the time when my plane landed but was still on the runway. This was partly due to a huge passport & customs backlog from simultaneous flights (although the INS or Customs people did a good job considering the underlying problem). But this time, when I wanted a fast car trip, there were immense Holland & Lincoln Tunnel backups, this being early Friday evening. Suffice it to say that I got quite vexed, and all the more so because my 3 (?) year-old iPhone is showing extreme battery fatigue and thus I was unable to call home saying I was on my way.
It was nice to see lots of old friends at the conference itself. Attendees were a mixed group - mainly economists but also a number of lawyers and accountants. Mainly Europeans but also lots of North Americans, Australians, and several from Asia. Ed Kleinbard and I were the only lawyers to present papers (Stateless Income from him, Corporate Residence Electivity from me), and it was odd how we agree about a great deal yet offer significantly different reform ideas. He wants a worldwide tax without deferral but still retaining foreign tax credits; I'd go for exemption with a transition tax on U.S. companies' existing unrepatriated earnings plus significant improvement of the source rules, in particular to apply to worldwide groups, whether American-headed or not, based on the concept of a unitary business. A key difference in perspective is that he is more focused on data concerning existing U.S. companies, whereas I am more focused on thinking about incentives & opportunities that tax-concerned players may have in the future.
Dhammika Dharmapala presented his co-authored paper (with Mihir Desai) proposing that transfer pricing continue to be used, in lieu of switching to formulary apportionment, but with one big change: the company's internal transfer prices, ostensibly used for purposes such as executive compensation, would be determinative in lieu of looking at ostensibly comparable arm's length prices.
Nearly everyone in the room agreed that (a) the comparable price method is seriously flawed (actually, 2 people disputed this, but they were thinking more about the method in principle than in actual recent practice) and (b) if one is continuing to use a transfer price approach in lieu of switching to formulary apportionment (and in practice the two overlap), then the internal numbers that Desai and Dharmapala want to rely on should indeed be examined as relevant evidence.
As it happens, the discussion brought out the fact that these numbers aren't nearly as widely available as D & D posit. While it may be true, as they assert, that companies need to keep track of internal value (or a decent estimate thereof) for substantive investment and/or compensatory purposes, the dividing lines they use in allocating internal income often have nothing in common with the legal entity lines that the U.S. tax authorities would want to see for transfer pricing purposes. For example, they may break down their "true" internal transfer prices by line of business or by geographical region, rather than on a per-country separate-entity basis.
Still, the paper could have made a modest but unambiguously positive contribution by emphasizing that these numbers may sometimes be available and could well be illuminating. Instead, however, it swings for the fences and claims a lot more. This led at Oxford to a decidedly mixed reception. (The reception would have been worse, except there was a widespread view in the room that Dharmapala was not the party most responsible for the paper's aspects of reckless over-claiming.)
The paper purports to identify a brand-new inefficiency that policymakers need to take very seriously in thinking about transfer pricing. It is as follows: The company accurately transfer-prices for its own internal purposes because it MUST do so in order to make rational economic decisions. But then the darn U.S. government comes along and forces the company to use transfer prices for U.S. tax purposes that place too much income in the U.S. and less, say, in Ireland, Luxembourg, or Singapore. The result is that the U.S. share of taxes on the company's worldwide activities is too high. This in turn causes the poor duped company to think the American operations are less profitable than they really are, because it subtracts taxes ultimately levied that the U.S. affiliate should never have had to pay. The company therefore under-invests in the U.S. and under=pays U.S. executives, while over-investing in Ireland and over-paying its Irish executives, because the U.S. transfer pricing shift has duped it into misjudging the actual economics of its worldwide operations. (NOTE: A careful reading of the paper might suggest modifying this story a bit - e.g., is the point that the company isn't duped but responds rationally to the U.S. over-tax? I only had a chance to read it very quickly and then listen to the presentation.)
But I think I have it right. If so, this is not a real world empirical story that I for one find enormously compelling. I don't agree that companies are in the main constrained to transfer-price accurately (even given the admitted difficulty of getting away with two distinct sets of books). And even if we see a given number in their books, we can't really tell how they are actually using all available information to make their various decisions. What's more, even if it's boring conventional wisdom to say the transfer pricing problem goes the other way (with income being shifted from high-tax countries to tax havens), every now and then the CW is correct. Sometimes the dog actually did bite the man, not the other way around. I would also find it quite startling if the companies couldn't figure out what is going on economically even if the U.S. forces the adoption of transfer prices that diverge from what the company sincerely believes is correct. The new inefficiency that the paper detects is one that I don't expect to hear a lot more about. And more generally, identifying ostensibly brand-new inefficiencies and coming up with new magic bullet solutions isn't always all that it's cracked up to be.
One other presentation at the conference gave me considerable pleasure, albeit purely as an observer and consumer. Unfortunately, I need to be a bit circumspect in describing it. Suffice it to say that a senior and well-known (though not to me personally) empiricist who is clearly passionate about econometrics, and who presents as having very high self-regard in this area, made a seemingly dazzling presentation that the likes of me couldn't assess very well, purporting to find strong empirical results on a contested issue. The commentator who followed to offer a response was very junior, and in tone very modest and understated, along the lines of "I wonder what X would look like if you had tested for that," or "it would be good to have run a control concerning Y, and I wonder what the results would have been." But by the time the commentator was about 2 minutes in (out of 10 minutes total), I realized that something rather extraordinary was going on. The paper was having its throat cut, and huge hunks of flesh were being torn out of it while its blood gushed onto the pavement - all by reason of one gentle knife thrust after another. By the time the full 10 minutes had passed, I would not even say the bloody carcass was lying on the ground twitching - what was left of it came closer to being an inert corpse. The author whose paper had gotten this treatment was scowling and head-shaking, but it got no better when the question period began. In sum, it felt like seeing Ali knock out Liston the first time in 1964, only without the underdog's having any braggadocio whatsoever. From my seat in the peanut gallery, I thought it was great fun, and all the more so because (not previously knowing the commentator) I had so little seen it coming.