Imaginechina via AP Images
Altera won unanimously in the U.S. Tax Court, then lost three times in the Ninth Circuit Court of Appeals, and now finally in the Supreme Court.
Last week, the Supreme Court declined to hear a major tax case involving a dispute between the IRS and Altera Corporation, a chip maker that is now owned by Intel. Usually, such denials are routine, since the Court hears very few tax cases. But this case is important, because it strikes at the heart of a technique that U.S. multinationals have used to shift trillions of dollars from the U.S. to low-tax countries. It also represents the first decisive victory for the IRS in litigation against the multinationals in a generation.
The technique at issue is called “transfer pricing,” and it involves setting the prices for goods or services transferred between a U.S. parent corporation and its foreign related entities so as to shift most of the profits offshore. In Altera’s case, the shifting was to the Cayman Islands, a notorious tax haven in the Caribbean.
Transfer pricing is a major problem for the IRS because it is so easy. Suppose a multinational manufactures a gadget in the U.S. at a cost of $20 and then sells it to a controlled subsidiary in a tax haven that incurs another $20 of marketing costs to sell the gadget to customers for $100. The net profit of the multinational is $100 - $20 - $20 = $60, but it is the price at which the gadget is sold from the parent to the subsidiary that determines where that profit is taxed: If the price is $80, the entire profit is taxable income of the U.S. manufacturer. But if the price is $20, the entire profit escapes U.S. tax because the U.S. manufacturer will show a net profit of $20 - $20 = $0, while the offshore marketer in the tax haven will report the entire profit ($100 - $20 - $20 = $60).
This simple technique is what enabled Amazon, Apple, Facebook, Google, Microsoft, and other U.S. multinationals to shift $3 trillion of income that economically was generated by intangible property created in the U.S. (and protected by U.S. patents and copyrights) to low-tax jurisdictions like Bermuda between 2005 and 2017. While this game is a bit less lucrative since the U.S. corporate tax rate was cut from 35 percent to 21 percent in 2017, 21 percent is still much more than Bermuda’s corporate tax rate of zero, and the U.S. corporate tax rate will hopefully be raised should the Democrats win the White House and the Senate in November.
Since the 1960s, the IRS has tried to combat such profit-shifting by arguing that the prices charged between the related parties are different than what would have been charged had they operated “at arm’s length.” But this argument usually fails, because in fact the related parties are not operating at arm’s length, and it is very hard for the IRS to find comparable transactions at arm’s length. As a result, the IRS lost a series of transfer-pricing cases in the 1970s and 1980s against large American multinationals like U.S. Steel, Bausch & Lomb, Merck, and Eli Lilly.
Most of these cases involved transfers of patents to low-tax jurisdictions, which almost never happen at arm’s length. In 1986, Congress revised the underlying provision of the tax law to state that in any such transfer, the foreign related party must pay the U.S. party a royalty that is “commensurate with the income” from the transferred patent. This was intended to strengthen the hand of the IRS in cases where comparable transactions could not be found.
The multinationals, however, found a loophole. They relied on an old regulation stating that if they shared the costs of developing a patent with their offshore affiliate, they can share the profits in the same proportion, even if all of the actual research that produces the patent is done in the U.S. Relying on such “cost-sharing arrangements” is what enabled a new generation of multinationals like Amazon, Apple, Facebook, and Google to transfer hundreds of billions of dollars offshore before 2017. The IRS recently lost a case involving such a transfer against Amazon and is currently in litigation against Facebook.
Altera involved a more specific issue: whether the costs that are included in a cost-sharing arrangement include the cost of employee stock options, which are a major form of compensation among tech companies. Altera argued that it should not include the cost of these options because unrelated parties would not have shared them—which is true, since if the parties are unrelated, the foreign party has no control over the performance of the stock of the U.S. party. If the parties are related, however, the same stock options are issued for both parent and subsidiary employees, since the performance of the stock reflects the profits of both parties.
The tax revenue at stake is at least $2 billion and could be much more.
Altera won unanimously in the U.S. Tax Court, then lost three times in the Ninth Circuit Court of Appeals, and now finally in the Supreme Court. This decision is important for several reasons.
First, the sheer sums involved. Affected companies include Facebook, Google, and Twitter. The tax revenue at stake is at least $2 billion and could be much more, because many companies did not record the effect on their books until the court case was resolved. Facebook said it paid $1.6 billion in November 2019 related to the appeals court ruling. Alphabet, the parent of Google, said the ruling cost it $418 million. Twitter reported an $80 million impact. Many other tech companies are affected as well. At a time when the government desperately needs revenue but lacks the political will to raise taxes on anyone, this is welcome news.
Second, the victory must be encouraging to the IRS, which has consistently lost transfer-pricing cases against major multinationals since 1979. Perhaps the IRS will finally win some cases against, for example, Facebook and Coca-Cola, which are currently in litigation on this issue. This, in turn, could induce other multinationals to settle rather than litigate, saving the government immense resources.
Third, perhaps the IRS would be willing to concede that in enacting the cost-sharing regulations, it made a crucial mistake, and would either repeal them or at least subject them to the “commensurate income” rule that Congress enacted in 1986 to fix this problem.
Finally, the ruling means that the courts are willing to look beyond arm’s length “comparables” in cases where these cannot be found. That is an encouraging sign that maybe the whole transfer-pricing tax avoidance game could be coming to an end, despite the immense resources invested in it by the multinationals and the Big Four accounting firms. If comparables do not exist, then the IRS can proceed with an analysis of the functions actually performed by the related parties, and allocate zero profits to offshore parties that do not perform any functions. Should the courts follow Altera in respecting this analysis, transfer pricing may finally become obsolete.
