It's all happening in the field of transfer pricing
Now we've got an equally forceful, and rather blunter, email from James Henry, a member of the board of TJN-USA who was also at the meeting.
Henry has quite some experience in business and in this particular field: among many other things, he has been Chief Economist, McKinsey & Co.; VP Strategy, IBM/Lotus; Business Development Manager, Chairman's Office (Jack Welch), General Electric - and he has written and spoken widely on key issues in the arenas of development finance and international tax. He is author of the acclaimed book Blood Bankers.
Below is the text of an email Henry circulated recently, in response to Spencer's outspoken presentation, and published by TJN with his permission. Anyone with an interest in this general area ought to read this, and we will only highlight a couple of snippets from the main email, plus supplementary ones from him:
"There is almost no empirical research on comparative transfer pricing systems in practice and workflow: "staffing levels on audit teams, case loads, throughput and back, life cycles and end states of typical cases, system costs per case, expected revenue per audit, etc....The overall TP policy brief is so "lawyer-driven" that it is almost impossible to tell a given country what the payoff would be from adopting system X Y or Z. And the OECD has insisted on the "one size fits all" version of ALS, with all the attendant costly processes. So, not surprisngly, most developing countries either have no TP laws at all, or don't bother to really enforce them.Or try this for size:
The OECD is hardly a role model for dispute resolution when it comes to transfer pricing: the US IRS now faces a backlog of some 800 transfer pricing cases involving cross border disputes, with most cases taking 3-5 years to resolve. [Data provided by tax expert David Rosenbloom.]Or this:
Far from providing invaluable expertise, many of these private sector spokespeople -- especially those from mining companies, big law firms, big accounting firms, and data companies like Bureau van Dijk (a big Dutch company that has a near-monopoly on the expensive "comparables pricing" data bases required to implement ALS -- are mainly involved to push their own jaundiced private interests. The complete absence of comparability data bases for many developing countries was mentioned repeatedly as a practical obstacle to implementing orthodox OECD ALS regimes.And this.
some huge MNCs may exist only because of their ability to arbitrage taxes and influence across borders. From this angle, the growth of MNCs may actually be a RESULT from transfer pricing and uneven taxation, not the other way around.That's enough snippets. The main text of Henry's mail is below. Although long and at times quite wonkish, focusing on the OECD's so-called Arm's Length Standard (ALS) method for addressing transfer pricing - it's fiery and hard-hitting.
Not for the faint-hearted.
David,TJN: And that's by no means all. More next week. Comments, from the OECD or anyone else, welcome.
Thanks for the final draft of your remarks. I apologize in advance for this long response, but you got me thinking.
While our frank presentations did produce a little consternation among the OECD drones, all NGO presentations, including yours, were very well received, especially by the developing countries.
We provided a brief dose of reality, by emphasizing actual real-world transfer pricing abuses, the overwhelming trend (outside the OECD) away from arms-length pricing orthodoxy, the value of crack investigators, and the longer term potential of Country by country reporting and formulary apportionment.
Overall, I was glad to have devoted the two days to participate, if only because of contacts that we made in Brazil, South Africa, India, Guatemala, and Ecuador -- and the huge amount we learned from them, as discussed below.
We should be grateful to UN staffers for standing up for the principle of hearing alternative points of view, at least for seven minutes per presentation.
1. Whether or not we will have much impact on the final outcome of the manual draft is not clear. As you have argued, a case can be made that the UN Subcommittee is more firmly than ever under the control of the OECD [see Spencer's speech for more on this].
Not only does the Subcommittee include a majority from OECD countries that already by definition support standards like ALS, but at least half of the attendees were orthodox: the OECD's technical representative, who had presided over Mexico's implementation of ALS in the 1990s under Angel Gurria (including the negotiation of over 1000 APAs with Multintional Corporations (MNCs) in Mexico before he did his first audit!!!); law firms like Baker-McKenzie and Caplan Drysdale that have huge for-profit ALS and APA practices;major accounting firms with similar stakes; and a variety of MNC business groups (Rio Tinto, Swissholdings, Maersk,the Intl Chamber of Commerce) who claim -- quite falsely in my view -- to speak for the diverse global MNC community at large.
Consistent with this "OECD dominance" view, there were several testy interventions to steer us back to the appointed task (in his view) of drafting a TP handbook that would basically just help developing countries GET ON WITH IMPLEMENTING the OECD's Holy Grail, the so-called arms-length standard (ALS) for transfer pricing.
2. Time and again we also heard from the OECD gallery that transfer pricing is a rather technical matter, not really about "development" or even "tax policy" per se, so much as about simply guaranteeing MNCs (OECD's true clients) "certainty" and "freedom from double taxation" when they venture into those nasty,corrupt, oh so poorly governed Third World environments.
In effect, we were being asked to take it on faith that ALS would necessarily serve the ultimate goal of development by encouraging foreign direct investment.
Well... Now in fact, when I put on my economist's jump suit, i search the economics literature in vain for any empirical evidence WHATSOEVER that transfer pricing regimes or double tax treaties (DTTs) actually have any measurable impacts on direct investment one way or the other, relative to all the other crucial variables. Or, indeed, for any evidence that FDI necessarily even consistently benefits economic development! As usual, "it depends...." Of course the BRICS have done quite well with very different TP regimes in the last decade -- but that may have just been due to their size and strength.
Furthermore, developing countries clearly have no monopoly on corruption and home-team judges and lawyers. Indeed, one clear obstacle to TP legislation has been their lack of trust in the First World experts, arbitrators, data base, and rules -- a lack of trust that the OECD's outsized role in the UN Subcommittee has hardly helped to reduce.
Mentioning such uncomfortable real world facts at this unworldly draft-oriented confab, dominated by lawyers, bean counters, and lobbyists, would have been unfair. But these basic facts do underscore the unreality of the exercise. It also underscores the impression we have here of a crumbling orthodoxy that is in deep crisis, losing control over many of the most interesting developing countries on the planet.
3. Between the OECD's furious defence of the ALS, and a position of outright rejection of it, there be other approaches. This alternative would involve the creative expropriation (without compensation) and subtle redefinition of the OECD's simplistic formula.
This approach would recognize just how elastic "arms length" really always is, in practice. It would be the equivalent of declaring "OK, we are all "Keynesians now" -- as in " Of course we all support "arms length pricing" in theory. Now let's get on with the real issue: what (the hell) does that mean in practice?"
4. For example, IN PRACTICE -- except in perfectly-competitive markets at all dates and states, with full information and constant returns to scale and scope -- the fact is that any actual application of ALS always yields a WIDE RANGE of estimated "comparable" prices.
In thirty years as a professional economist, studying markets for everything from cars, banking, and electric power to software, tomatoes, wireless, and diamonds, i have yet to find a perfect market. Even for a supposedly pure commodity like steel, there really is no such thing as "a market price." There are many different prices that vary by location, grade, degree of finishing, market conditions, credit terms, and quantity discounts, not to mention competitive strategies and applicable taxes.
Indeed, the very existence of huge multi-market MNCs is prima facie evidence that the perfectly competitive markets required for the optimality and tax neutrality of the OECD version of ALS do not exist.
Such firms are able to succeed even though they abolish their internal markets either because of economies of scale and scope, and associated market power that is completely inconsistent with ALS.
Or, in the alternative, some huge MNCs may exist only because of their ability to arbitrage taxes and influence across borders. From this angle, the growth of MNCs may actually be a RESULT from transfer pricing and uneven taxation, not the other way around.
5. These complexities are especially prevalent in the highly imperfect global market for intellectual property. I was astonished to hear an OECD representative claim that intellectual property TP is (a) not an appropriate concern for developing countries, and (b) is the subject to a new OECD project that is well under way, to be released "real soon now."
Well.... I doubt that China --now the world' second largest source of new patents -- or India would agree. And as for the relevancd of IP transfers to developing countries, just cf the SABMiller case. On this issue in particular, the absence of developing country leadership may clearly be biasing the UN's approach.
6. Contrary to the OECD, such comparable price measurement problems are also endemic in global markets that are subject to increasing returns to scale. There, the use of prices based on supposed arms length comparables systematically overstates the value of services/goods purchased from haven-based subsidiaries, and systematically understates the value of inputs sold to such entities. As a result, haven entity profit is inflated, and tax revenue is lost.
7.We also heard from the conference attendees that estimating comparable prices is a very costly process, even in OECD countries where private market data bases and specialized experts are available at a price.
The complete absence of comparability data bases for many developing countries was mentioned repeatedly as a practical obstacle to implementing orthodox OECD ALS regimes. A handful of Dutch and other private European data base companies dominate this lucrative market. If the UN really wants to help developing countries with ALS, it should figure out how to provide cheaper data. Other many tax authorities will simply find it impossible to implement an orthodox ALS regime. One approach might be to require MNCs to pay the full cost of such comparability analysis. Another would be to follow Brazil and use simple rules of thumb for imperfect markets where no comparable data exists.
8. We also learned that even where such estimates can be made, under the OECD's preferred approach to "dispute resolution," they could be subject to prolonged, costly litigation and arbitration. Especially in a lower-tier developing country context, arbitration, if mandated, may be so expensive that only MNCs can afford it. So the practical impact of requiring it (as part of a DTT's "MAP" process, for example) would be to hand over case after case to MNC tax violators.
Once again, the OECD is hardly a role model for dispute resolution when it comes to transfer pricing: the US IRS now faces a backlog of some 800 transfer pricing cases involving cross border disputes, with most cases taking 3-5 years to resolve.
This only helps to underscore the fundamental absurdity of the whole orthodox approach to transfer pricing -- it basically presumes that all 192 countries in the world are capable of setting up not only their own TP auditing departments, but their own tax courts -- or the equivalent expertise in their existing judicial systems. This reminds me of the "original sin" committed at Bretton Woods in 1944, and reiterated by the IMF in 1973, when it was presumed that every developing country was capable of operating its own central bank. That assumption proved to be very lucrative for First World banks and country elites, but it was disastrous for most developing country residents. And now a similar mistake is about to be made with global taxation.q
9. We also learned that many countries are struggling with the basic step of deciding how to select particular companies for audit. This is a highly political decision even in OECD countries. We made several suggestions, based on investigative experience.
First, even without Country by Country reporting, NGOs and outside investigators have made surprising progress just by mining public data sources, and by doing "gumshoe" investigations of particular industries. NGOs should continue to play this role, partly just because they can help to reduce the inevitable political pressures on internal auditors, once big fish targets get hooked.
Second, for all the UN's talk about "training" and "capability building," NGOs have learned that with respect to transfer pricing investigations, a surprising amount can be accomplished with tiny teams of dedicated people and a few whistleblowers --assuming they get "political air cover." (EG one investigator equivalent for 3 months, plus a body guard.) But they need to be prepared for the career aftershocks. Indeed, for local residents pursuing careers in tax enforcement, THAT is a real issue. And part of the "corruption" is subtle -- if you play ball, you can look forward to a lucrative career in an accounting firm AFTER the (low paid) IRS. But if you don't, perhaps not...
Third, joint cross border investigations, while complex, can take some of the pressure off individual countries. This requires information sharing agreements among countries and basic data security procedures. But once these are in place, there are positive externalities. (The US IRS is exploring such joint investigations in certain international arenas right now such as with Australia - as we have just blogged.)
Fourth, there are also situations where the focus should really be on whole industries, or on all of the companies "enabled" by a particular haven or by a particular professional service firm, not just on individual MNCs -- for the sake of experience, productivity, and maximum deterrence. Given inter firm mobility and competitive pressures, it is a rare company that is the only one engaged in shady practices in its industry. "This IS the oil industry we talkin' about here, right?....(Syriana)
10.Not surprisingly, given such realities,countries like Brazil that can afford to be more independent have opted for more streamlined, fast-paced rules of thumb in the majority of interesting cases where markets are not perfect. According to the Brazilian rep from the Ministry of Finance, most MNCs are happier with this simpler system than the litigation-intensive orthodox ALS alternative.
11. Of course most poorer countries don't even yet have any TP legislation or tax treaties at all. Guatemala is a good example. As you may have heard, an incredible TWELVE NGOs and FOREIGN TAX DEPARTMENTS are now reportedly advising this poor rich country on tax policies to help it solve such problems. But the real problems are not technical, but political: this is an election year, the budget deficit is already 3 % of GDP, and Guatemala's business community is opposed to any taxes on corporate income or dividend payments, domestic or foreign. It is seeking to abolish all income taxes, and substitute excise taxes. No one wants to toy with new transfer pricing legislation and the costs of setting up tougher enforcement in such an environment.
To get such countries to invest in MNC taxation, new Presidential-level leadership will be needed, and it will require a first-rate fact-based business case, based on actual experience with implementing transfer pricing in other countries -- not generic guide books. To date, the BUSINESS CASE for the superiority of the full-blown, lawyer+data+courts+experts OECD version of ALS for developing countries simply does not exist. Will UN provide one?
I think that we could find support for doing such fact-based analysis of how TP systems in developing countries actually work among more "enlightened" MNCs, which not only understand the value of more efficient states, but might well actually prefer a simpler rule-of-thumb based system. We might also encourage the IMF's new leadership to tackle this assignment, because that organization does have the economic perspective, the deep bench of talent, the tax knowhow, and the cross-border range that is required. However, the IMF suffers from almost as much perceived "First World Club" bias and distrust as the OECD. In principle, someone at the UN -- perhaps UNDP-- should take this on, but whether it is prepared to step up is unclear.
Meanwhile,these issues may also deserve further study by NGOS, given that no one else is doing this. We might begin with a closer look at how the existing TP systems in Brazil, Mexico and India actually work in practice.
Unfortunately, the budding UN manual skips over such vital empirical questions completely. It is being invited to dinner and served pictures of food. I will be really surprised if any developing country that does not already have a transfer pricing regime in place -- e.g. most of them -- will find it all that helpful.
If David is right, maybe that's by design.