Disney’s Transfer Pricing and the Destination Based Cash Flow Tax

Neil Irwin adequately addressed the trade policy debate over the Destination Based Cash Flow Tax (DBCFT) but his discussion of the transfer pricing angle leaves me cold:

Two prime examples are transferring intellectual property to overseas holding companies and engaging in corporate inversions that move a company’s legal headquarters to a country with lower taxes.

Corporation inversions do not impact transfer pricing as all they do is to allow a multinational to have a territorial system which many already effectively do. But I need to update my discussion to focus on intellectual property (IP) as this goes to the heart of my beef with what Auerbach has proposed. In my Trump Toaster Ovens example (Canadian production with U.S. distribution), I noted:

Total profits are $25 per oven with 80% going to the Canadian affiliate if the intercompany price is $100. US tax rates are now 35% and Canadian tax rates are close to 25% so with no repatriation tax involved (Canada is a territorial system), the effective tax rate is 27%. While currently Tiffany might want to raise the intercompany price – she knows the IRS could object. Of course Auerbach’s DBCFT would change her incentives as she might want to lower this price to only $80 to eliminate the Canadian income tax – assuming the Canadian Revenue Agency does not object. What’s going on here?

What’s going on per multinationals is that the U.S. becomes the tax haven per income taxation but also imposes sales taxes on imported goods. I know many U.S. centric transfer pricing types think all IP is created here but companies like the Japanese automobile manufacturers, Novartis, Adidas, Jimmy Choo, and Zara created IP abroad. DBCFT would be bad news for them as their IP faces sales taxes here as we consume their products but also face income taxes abroad. Let’s toss in Ikea even if Tim Worstall struggles to understand what an arm’s length royalty rate is. Of course we generate more IP income that most nations and DBCFT makes any IP income involved when foreigners consume our products tax free. Ricardo Hausmann and Federico Sturzenegger illustrated their “Dark Matter” idea with this:

Imagine the construction of EuroDisney at the cost of 100 million (the numbers are imaginary). Imagine also, for the sake of the argument that these resources were borrowed abroad at, say, a 5% rate of return. Once EuroDisney is in operation it yields 20 cents on the dollar. The investment generates a net income flow of 15 cents on the dollar but the BEA would say that the net foreign assets position would be equal to zero. We would say that EuroDisney in reality is not worth 100 million (what BEA would value it) but four times that (the capitalized value at our 5% rate of the 20 million per year that it earns). BEA is missing this and therefore grossly understates net assets. Why can EuroDisney earn such a return? Because the investment comes with a substantial amount of know-how, brand recognition, expertise, research and development and also with our good friends Mickey and Donald. This know-how is a source of dark matter.

In its latest fiscal year, Disney sourced over 94% of its worldwide income in the U.S. as its foreign affiliate pay the U.S. parent for the value of Mickey and Donald. While Starbucks has received some weird attention with respect to its transfer pricing, the 6% royalty rates paid by its foreign affiliate to the U.S. parent are consistent with what third parties pay and hence are arm’s length. In its latest fiscal year, over 84% of its worldwide income was sourced to the U.S. as a result. Both of these U.S. based multinationals are currently sourcing their IP income in the U.S. A switch to DBCFT would change that significantly reducing U.S. tax collections. Now you might ask what about those accused of Base Erosion and Profit Shifting such as Caterpillar? In her defense of their transfer pricing, Julie Lagacy noted that their Swiss affiliate paid the U.S. parent a 6% royalty rate for the use of Caterpillar’s technology but then other witnesses noted:

“I think the I.R.S. should have attacked this transaction on economic substance grounds,” said the other tax professor, Reuven S. Avi-Yonah, of the University of Michigan. To make the transaction bona fide, he said, the I.R.S. should have applied a 1986 law requiring the subsidiary to pay a “super-royalty” to its parent. That would have brought the profits back to the United States where Caterpillar’s higher rate, around 29 percent, would apply. Its negotiated rate in Switzerland was less than 6 percent. “There are all these opportunities that the I.R.S. had to go after this transaction, and unfortunately it didn’t,” he said.

The 29% percent effective tax rate was due to the fact that around 45% of Caterpillar’s income was U.S. sourced whereas only 30% of its sales were in the U.S. Again DBCFT would have meant even less U.S. taxes given the fact that a lot of the foreign generated IP income already was coming back to the U.S. This “super-royalty” would have the royalty rate raised from 6% to 9% as if the U.S. owned all worldwide intangibles, which strikes me as a very aggressive IRS view on this issue. In other words, one can make the case that the 6% royalty rate was arm’s length. If the IRS wants to disagree and pursue a transfer pricing challenge, then let it. But if we passed the DBCFT, then we would simply give up on taxing U.S. generated IP income when it is consumed abroad. This strikes me a very bad retreat from trying to enforce the transfer pricing rules.