Caterpillar v Comm'r and the Rule of Law in International Tax

James P. Fuller has an interesting summary of the recently-filed Caterpillar case in his latest U.S. Tax Review [gated], in which he laments the competent authority breakdown and argues that the case would have been better off going to treaty-based arbitration, rather than to domestic judicial decision-making channels. I disagree with this conclusion because, given the structure of tax treaty arbitration today, it would--at best--provide a remedy for only one taxpayer at great cost, while the judicial route potentially creates rule of law upon which all can rely. If competent authority arbitration were instead to create a publicly viewable resolution, I could agree with Fuller because what is needed is (1) a multilateral solution to a multilateral problem and (2) a solution with precedent-making force.

Per Fuller:

Caterpillar Inc. has petitioned the Tax Court for a redetermination of income tax deficiencies that resulted from the IRS's allocation of royalty income to it from its Belgium and French subsidiaries. 

While the case was only recently docketed and has not yet been decided, I thought it worth discussing the case as it results from a breakdown in the competent authority process. It is, of course, the very process that is designed to prevent the consequences faced by Caterpillar. There should be no need for the taxpayer to litigate in one country or the other when treaty relief is an available remedy and the countries involved can settle the issue between themselves. 

Following a 1990 reorganization that pushed management power, responsibility, and accountability to subsidiaries such as Caterpillar Belgium and Caterpillar France, the taxpayer entered into amended license agreements with those subsidiaries limiting the maximum royalty in any given year to each subsidiary's net income from the sale of Caterpillar products. Also, if the subsidiary suffered a post-effective-date net operating loss, it would pay no royalty for that year and could carry the loss forward as a negative adjustment to future years' income for purposes of the royalty calculation. 

... The IRS ... concluded that the relief-from-royalty provision in the Belgian and French license agreements did not comport with the arm's-length standard under section 482.

 Fuller points out that there are "several Tax Court cases that permitted related-party license agreement provisions" like Caterpillar's, and as to which the IRS subsequently acquiesced. He goes on to discuss Caterpillar's attempt to obtain competent authority relief, which failed because the French and Belgian tax authorities disagreed with the US position, and US Appeals simply went along with the IRS decision despite these other rulings that would suggest reconsidering the issue, for litigation hazard purposes at minimum:

The Belgian and French tax authorities, having a thorough knowledge of the local operations of the Caterpillar subsidiaries in their respective countries, looked at the same facts that were addressed by the IRS exam team and found that Caterpillar's royalty limitation provision required no adjustment. Caterpillar subsequently had the adjustment reviewed in an IRS Appeals office proceeding. The IRS Appeals officer simply accepted the IRS exam team's economist's report. 

...it appears from Caterpillar's Tax Court petition that the IRS exam team's economist simply used that agreement as grounds for asserting that such a relief-from-royalties' provision is inappropriate, and not at arm's length.

Fuller notes the dual-bureaucracy created by two simultaneous review procedures, i.e., competent authority and internal appeals, and concludes that binding arbitration of the competent authority procedure, rather than resort to the US judiciary, would have been the better approach:

If the two countries' competent authority negotiators cannot reach an agreement, then there should be some form of compelled arbitration to bring about an agreement. Otherwise, the taxpayer is stuck in the middle. A taxpayer should not have to litigate its case in one country or the other (or both) simply because the U.S. competent authority negotiators could not reach an agreement with the foreign country's competent authority negotiators. This is especially true in a situation such as that faced by Caterpillar: The Tax Court has already held that such provisions are appropriate in related-party license agreements.

I appreciate Fuller's argument about the taxpayer's bind, but as I have noted before, if the case went the way of binding arbitration, in the long run the issue could potentially never be settled, since the decision in arbitration would be completely confidential and therefore not accessible or applicable to other taxpayers. That makes international tax dispute resolution much more expensive than it has to be, all because the powers that be have prioritized absolute taxpayer confidentiality over the rule of law. I think that is a miserable trade-off as well as being an unnecessary one: as this case shows, the taxpayer is willing to sacrifice some measure of its own confidentiality in order to get resolution in domestic law, so it is not clear why international law should be so different. (The arguments for difference are weak--see my analysis in the link above.)

The competent authority route, which (by being duplicative as in this case) already increases costs for producing the rule of law, would only be exacerbated by arbitration, because Caterpillar's problem would be perfectly preserved for another day, another taxpayer, and another expensive litigation involving multiple parties and governments.

Therein lies the conundrum for international tax law in the current status quo: either we can get taxpayer-specific outcomes but no rule of law (arbitration) or we can get unilateral rule of law but no international resolution (domestic appeals). I am not sure which to prefer, since both are bad for international tax law.

Accordingly, I am glad to see the Caterpillar case go forward in a forum which is open to public view and that, if not settled in the interim, then becomes a part of the body of law, creating more certainty for taxpayers going forward. However, I am unhappy that the forum is unilateral and potentially preserves an unsolvable problem for the taxpayer if the Court agrees with the US position, since France and Belgium will not be consulted in the process and can be expected to continue to disagree with the IRS view of things.

As a result, I can only agree with Fuller that the better route would be bilateral/multilateral decision-making via arbitration if that decision-making is, like internal judicial decision-making, open and accessible to public view.