BMW Transfer Pricing and Trump Trade Accounting

The Tax Justice Blog features a critique of the Destination Based Cash Flow Tax (DBCFT) by the Institute on Taxation and Economic Policy (ITEP):

the border adjustment likely would make the corporate income tax substantially more regressive ... One of the major arguments that proponents of the border adjustment tax make is that it would stop corporate tax avoidance. It is certainly true that the border adjustment would remove companies’ incentive to use certain existing loopholes in our current system, but it would create numerous new opportunities for tax avoidance through the shifting around of sales. For example, Microsoft could avoid the tax by selling its software to consumers in the United States directly from servers in Ireland or another tax haven. At this point there is no reason to believe that following a tumultuous transition to a border adjusted tax that our tax system would end up less prone to tax avoidance than our current system.

The list of people realizing that will create new opportunities to game transfer pricing is growing. Permit me to address a somewhat related transfer pricing issue with respect to the proposed Trump Trade Accounting idea:

Let’s take an example based on Ford selling its cars to Canadian customers and having them manufactured in Mexico. Suppose Ford Canada sold a car for $20,000 that cost Ford Mexico $16,000 to produce and cost Ford Canada $2000 to distribute. Ford worldwide made $2000 in profits off of that car. The balance of trade statistics currently would take Ford’s transfer pricing as given so let’s speculate on how this might work. Suppose Ford US paid Ford Mexico cost plus 5% or $16,800 but then charged Ford Canada $17,600 on the premise that the Canadian distributor deserved a 12% gross margin as its operating expenses were 10% of sales. Ford US would retain $800 per car in profits for effectively doing nothing. Of course Ford might argue that this represents the value of Ford’s intangibles. I can see the tax authorities of Canada and Mexico disagreeing on this allocation of income. My simple point, however, is that current balance of trade accounting relies on the intercompany pricing of multinationals which at times can be suspect.

Dean Baker prefers to talk about BMWs:

The classic example would be if we offloaded 100 BMWs on a ship in New York and then 20 were immediately sent up to Canada to be sold there. The way we currently count exports and imports, we would count the 20 BMWs as exports to Canada and also as imports from Germany. These re-exports have zero impact on our aggregate trade balance, but they do exaggerate out exports to Canada and our imports from Germany.

I’m not sure why this is the “classic” example but I suspect there is a very different supply chain envisioned in this BMW example versus my example. I will admit that automobile multinationals would prefer not to have taxable income in the U.S. but they are also aware that the IRS and other tax authorities in places like Canada, Germany, Japan, and Mexico have extensive knowledge of the transfer pricing aspects for their sector. While Ford might want Ford Canada to pay only $16,800 as income tax rates in Canada are lower than they are in the U.S. currently, Ford has to let this $800 remain in the U.S. if that represents the value of the Ford intangible assets owned in the U.S. So the Trump proposal would have something other than a “zero impact” in my example. So what is Dean’s example about? Let’s assume that these BMWs are sold in Canada and the U.S for $30,000 (all figures U.S. dollars), cost Germany $21,000 to design and manufacture, and incur $4500 in local selling and marketing costs. Consolidated profits are therefore $4500 per car. Through negotiations with the IRS, the Canadian tax authorities, and the German tax authorities, the U.S. and Canadian distribution affiliates will receive a 20 percent gross margin – that is Germany receives $24,000 per car which leaves them with $3000 in profits and $1500 in profits for the local distribution affiliate. Now if a ship landed in New York with 100 BMWs where 20 of them would be re-rerouted to Canada, then Dean is likely right – BMW Canada will pay $24,000 per car. But would this have to be first invoiced to BMW U.S. as he assumes? Not necessarily as doing so could cause all sorts of confusion for our customs agents. But let’s grant Dean this narrow example under the current tax law. But what would likely happen under DBCFT? Karl Keller, George Korenko, and Lori Hellkamp note:

rather than eliminating transfer pricing, the border adjustments will likely shift its focus, incentivizing multinationals to minimize the cost of imports by U.S. affiliates (because such costs would no longer be deductible expenses) and maximize U.S. affiliates’ revenue from exports (because such income would escape U.S. taxation and possibly even result in tax rebates)

They and ITEP talk about multinationals altering their supply sides. In our BMW example, imagine if BMW US took over the selling and marketing efforts in Canada eliminating BMW Canada as the U.S. has become an effective tax haven. In this case, they would import the 20 BMWs for $24,000 per car and export them for $30,000. Now you might say this is not the current U.S. tax system which is true. But consider nations like Hong Kong, Ireland, and Switzerland that are tax havens. This kind of transfer pricing activity is widespread and the implications for balance of trade accounting is considerable.