Render unto Caesar

March 25, 2013

An obscure reference to reforming the taxation of multinationals in the UK budget speech might be more important for developing countries than the big increase in aid that was announced at the same time. Mandatory ‘combined reporting’ by multinationals could enable countries to tax multinationals properly.

For developing countries, the most important part of last week’s UK budget might not be the confirmation that Britain is at last reaching the 0.7% international aid target. In the long run, the more significant announcement might turn out to be this part of the Chancellor of the Exchequer’s statement:

"And this year, we are leading international action on tax avoidance, through our presidency of the G8, with the OECD and at the G20.

We want the global rules governing the taxation of multinational firms to be updated from the 1920s when they were first written, and made relevant to the global internet economy of the twenty first century.”

Academics and tax experts have long been aware that the international tax rules need to be updated but it marks a striking shift to hear such direct language from a senior policymaker from an OECD country – especially from the country currently setting the G-8 agenda.  A shift in international tax policy could yield substantial benefits for the great majority of countries, regardless of their level of per capita income.

[caption]The IF campaign is targeting George Osborne to deliver tax reform for developing countries through the G8. Photo: CAFOD CC BY-NC-ND [/caption]


Reform would not involve countries giving up their sovereignty over tax: on the contrary, it would enhance the ability of countries to set meaningful corporate tax rates, reflecting the preferences of their citizens; and at the same time reduce the economic distortions caused by the current system, including the substantial cost of lawyers and accountants paid to help companies minimise their tax within the current rules.  We don’t need a global agreement on tax rates and allowances: but we do need some common agreements and information sharing about the tax base on which those rates are applied.

The challenge is to find an appropriate way for national governments to tax global businesses. In the past tax authorities have treated multinational companies as if they were a collection of separate companies which happen to have a common owner. But this does not reflect the reality that profits are realised by the global business as a whole, and so cannot readily be attributed to the activities of any particular part of the company. 

Tax is not supposed to be voluntary

The current OECD rules attempt to separate the different parts of the business on the “arm’s length principle”. This requires that when there is a transaction between two companies which are part of the same multinational group, they should be accounted as if the transaction were occurring between unrelated parties (hence, at “arm’s length”). But in reality it is very hard to find or construct these arm’s length prices.  In practice companies working within the same group have broad discretion over the way internal transactions are recorded, enabling them to organise their affairs so that their profits are recorded in low tax jurisdictions, even if this does not reflect the underlying reality of their businesses.

Allowing this degree of discretion so that companies can record their profits where they choose is bad for both developed and developing countries. If companies can easily shift their profits to places where taxes are very low, even if that is not where the economic activity is taking place, then no country has real sovereignty over its tax. No less a radical firebrand than the Financial Times notes that corporate tax has become ‘a largely voluntary gesture’, which it thinks ‘scandalous’. The problem is particularly acute for developing countries. Because OECD countries have more political power and more technical capacity, they can exert more pressure on multinational companies. If rich countries are able to demand that multinational companies make at least some contribution in tax, global companies will tend to shift their reported profits to those countries, at the expense of countries with poor tax administration and weak bargaining power.

How big a problem is this really?

We do not know for certain how big a problem this is, and I’ll be participating in a detailed research project on this starting soon. In the meantime here are some indicative statistics. A couple of years ago (see the first two minutes here) I used some data on US multinationals’ profits to explore whether they might be under-declaring in developing countries.  

I compared the distribution of where US companies declare their overseas profits with the distribution of world GDP. Of course, there are lots of reasons why companies might be earning a bigger share of their profits in one country rather than another, but overall the results do suggest that many companies are declaring profits in places other than where those profits are being earned.* In “secrecy jurisdictions” such as Switzerland profits are about 6 times more than we might expect, given the distribution of world GDP, and in the small secrecy jurisdictions (basically, the small island financial havens) the profits are 14 times larger. The biggest losers are not G-7 countries, in which the share of profits is only a little lower than their share of world GDP. The problem is in the rest of the world, where companies are declaring less than half the profits we might expect.


Larger secrecy jurisdictions accounted for 5% of world GDP, but 30% of US companies’ profits; and smaller secrecy jurisdictions accounted for less than 1% of world GDP but 14% of profits. It seems most unlikely that US companies are really making 14% of their profits in Liechtenstein or Bermuda, even if they are acting legally by organising their accounts in such a way that it looks as if they do.

It is possible that these difference partly reflect genuine differences in the distribution of where profits are earned around world; but it certainly suggests the possibility that companies are under-declaring their profits in developing countries, and shifting them to the secrecy jurisdictions.  

Two years after I presented this data, the Congressional Research Service uncovered an even more dramatic picture: showing that in 2008, US companies declared 43% of their profits in Bermuda, Ireland, Luxembourg, the Netherlands and Switzerland, on the basis of just 4% of their overseas workforce and 7% of their overseas investment.

Is there a better way?

There is broad and growing recognition that the ‘arms length principle’ is no longer fit for purpose as the basis for working out where multinational companies should pay tax.  The alternative is to tax multinationals as a single unit rather than continue to treat them as if they were a series of unconnected national businesses. This approach - which is known as ‘unitary taxation’ – requires some common approach to how much of the companies’ profits should be taxed in each jurisdiction so that it is fair to the countries in which the businesses operate and is also fair to the companies and their shareholders.

The OECD already offers a ‘profit-split’ approach, which allows allocation either (i) by a percentage based on an evaluation of the contributions made by each affiliate using suitable `allocation keys’, or (ii) by applying one of the other methods and then apportioning the `residual’.  The Chancellor of the Exchequer’s statement about the reform of international taxation appears to be taking is further in this direction.

That doesn’t mean an immediate switch to a globally agreed ‘formulary apportionment’, whereby all countries commit to using the same formula (perhaps a combination of sales, employment and assets) to allocate each multinational’s profits between them. But the OECD’s ‘Base Erosion and Profit Shifting’ initiative, set up in response to growing frustration among member states at cases like Starbucks, Amazon and Google, may set the course for a gradual move towards unitary taxation, without requiring formal coordination. This would reflect the reality that many countries are increasingly relying on profit-split methods, given the manifest problems with arm’s length pricing.  

If the world is moving in this direction, it is important that the OECD countries do not develop a new set of rules for international taxation that protect their own interests but neglect the need for developing countries to be able to defend their tax base, since it appears to be developing countries who are currently suffering most from the absence of an effective system of international tax. It’s no coincidence that the UN Economic Commission for Africa was writing about the potential benefits back in 2004.

What needs to happen next?

The most important next step is to require multinationals to provide a ‘combined report’ setting out global profits and the scale of their economic activity in each jurisdiction – hardly a wild step in transparency, though long resisted. Interestingly, this possibility was raised recently (£) by Professor Paul Collier, who is currently advising the UK government on their agenda for the G-8 meeting. Combined reporting has been required by US States such as California for several decades, but is voluntary elsewhere; it has never been required by national governments. Requiring combined reports would provide the basis to determine the tax base on which nation states could exercise their tax sovereignty.

British Prime Minister David Cameron says that the UK priorities for the G-8 meeting are ‘three Ts’: tax, trade and transparency. Requiring a combined report from multinationals is a basic step towards transparency that would start to tackle a major distortion in world trade and the allocation of the tax base, and to reverse a significant obstacle to development. What’s not to like?

* This comparison of profit shares with shares of global GDP is of course illustrative. We are implicitly assuming that any difference between the distribution of the share of US companies overseas profits from the distribution of world GDP reflects a choice about where to declare profit, rather than differences in where those profits actually arose.  This makes the heroic assumptions that US multinationals are sufficiently large in number and in investment volume, and sufficiently profit-driven that we could expect their total overseas investments to be proportional to the amount of economic activity in each non-home market; and that with efficient markets the (expected) profit rate on those investments would be the same in each location.  The revealed deviations are sufficiently large, and sufficiently closely related to financial secrecy, to suggest that the distribution of (declared) profit is determined at least in part by choices made by multinational companies.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.