Corporate tax 2.0?

Sorry, I thought you meant... [Photo by Gillicious, CC, Flickr]

I promised I wouldn’t write about tax, either on this blog, or while I was away on leave; and am therefore failing on both counts. I’ve been tempted into this transgression by the surprisingly radical report released over the summer by the Economic Affairs Committee of the UK House of Lords, the latest in the deluge of UK parliamentary inquiries into tax avoidance and the integrity of the corporate profit tax. As you’d expect from a committee that includes both Nigel Lawson and Robert Skidelsky, its report is politically provocative (what report on tax can afford not to be these days?) But it’s also in parts a more theoretical and birds-eye view than the reports of its Commons sister, the Public Accounts Committee, can afford to be.

Unsurprisingly, the brief media coverage focused on the politically provocative bits: proposals on naming/shaming, HMRC ‘deals’, barring tax-scheme advisers and publishing large corporates’ tax returns. Far less commented upon, though equally radical, has been the Committee’s blue-sky consideration of alternative corporate tax systems – and particularly a recommendation for the Treasury to examine the possible advantages of taxing corporate sales instead of corporate profits - or, more properly, to replace the corporate profit tax with a ‘destination-based cash-flow tax’, or DBCT. (Michael Devereux of the Oxford Centre for Business Taxation, who was appointed the Committee’s specialist adviser for this enquiry, is the UK’s leading proponent of the DBCT, and has probably done more technical groundwork on it than anyone else).

That there’s political space in the parliamentary system for this kind of big-picture thinking, rather than simply considering particular tax controversies, is an unambiguously good thing. (Though the Committee’s grand vision isn’t limitless: other ‘big-picture’ solutions, particularly proposals for the formulary apportionment of the whole profit base of a multinational group, enjoy easily as large a fan-base as the Destination-Based Corporate Tax, albeit amongst a politically rather different crowd; yet the Committee’s report rejected such ‘unitary taxation’ proposals without much detailed consideration).

But what of the DBCT? (And we really need a better acronym…) A DBCT would have two features fundamentally different to the current corporate income tax. First, it would be ‘destination-based’: rather than being levied on a company’s income in that company’s country of residence, or by the ‘source’ countries where the value underlying a piece of income is deemed to be created (generally where it produces goods or services), it would be levied instead on income from the sales of goods and services in the country where that entity sells its goods or services. Close to a classic VAT, it would essentially apportion the tax base of a multinational according to where that multinational makes its sales.

Second, it would be levied only on ‘cash flows on real activities’: in other words, as Professor Devereux envisages it, the tax would only be levied on income from the ‘real’ sale of a good or service minus related expenditure, including the immediate deduction of spending on capital investment – in the case of many multinationals, a smaller tax base than the current corporate profit tax, although also ignoring things like deductions for interest payments on inter-company loans.

On first consideration, as redesigns go this has:
  • a sort of intuitive fairness in the world of Amazon et al (if we’re buying Amazon’s books in the UK, shouldn’t the income generated by those transactions be taxable in the UK?)
  • a sort of intuitive practicality as a means of taxing global corporations (profits and establishments may be easily shift-able to Switzerland or Luxembourg; customers are far less mobile).

In many cases I suspect a DBCT would reduce the overall tax burden of a multinational (including by potentially shrinking the tax rate on income from new corporate investment to zero), although its proponents argue that fiscs could compensate for resulting revenue loss by increasing overall corporate tax rates without fear of losing out to tax-competition-driven profit-shifting. Conversely, the DBCT’s proponents argue that it offers some obvious advantages in protecting against tax avoidance through intra-group transactions.

This may well be true for intra-group financing, which simply ceases to be deductible from the tax base; although since tax is still going to be levied on the basis of individual corporate entities rather than whole groups, I’m less clear about why a multinational couldn’t still avoid tax by using intra-group trading to suppress operating profits in its sales entities. For instance, let’s say that under a DBCT Amazon’s (low-tax, high-profit) Luxembourg sales entity is going to be taxed on its UK sales in (higher-tax) UK itself. Couldn’t Amazon just turn the Luxembourg sales entity into a low-profit distributor, with only a sliver of profit to be taxed in the UK; and have this Luxembourg entity in turn buy the books for sale in the UK from another (high-profit) subsidiary which, selling to its sister company in Luxembourg, will still have its hefty profits taxed in low-tax Luxembourg? (Thanks to Professor Sol Picciotto for pointing this out).

Arbitrage opportunities aside, one critical aspect of the DBCT that is less often considered (I don’t know of any published, systematic modeling) is its impact on the international distribution of the corporate tax base. Full disclosure: in my employer’s own evidence [pdf] to the Committee we raised the concern that a DBCT, adopted internationally, could undercut the tax base of countries – often poorer countries – whose economies are dominated by export-oriented primary production, like minerals or agricultural products, which are generally sold to parties in other countries rather than locally sold and consumed. (The Committee described our concern in their report more expansively, as a more general statement that it would disadvantage all developing countries).

And the Committee unsurprisingly disagrees with this latter formulation. They concede that a DBCT may broadly disadvantage countries with a trade surplus (i.e. strong exporters and/or those with weak consumer bases), but argue that many developing countries run a trade deficit, and that a DBCT will broadly advantage such countries. (A recent presentation [pdf] by Michael Devereux and Rita de la Feria puts it somewhat less bullishly, arguing that a DBCT “will not necessarily be unfavourable to developing countries: developed countries will most likely benefit; but so will big developing countries”. There’s something of a silence here about ‘smaller’ developing countries).

Our objection was about a particular category of countries that would, I think, self-evidently lose out: small developing economies broadly reliant on raw materials exports. Like Zambia, whose economy is dominated by copper mining, running a smallish trade surplus with over half its exports being sold to Swiss metals traders or affiliates of multinational mining companies. Zambia is unlikely to be thrilled by a major chunk of its corporation tax base being shifted from Zambia to Switzerland under a DBCT.

But I’m curious about the larger, less self-evident proposition: that the tax take of Majority World countries in general wouldn’t be disadvantaged by a DBCT.

Let’s accept, for a moment, the assumption that trade-surplus countries are generally losers under a DBCT, and trade-deficit countries are generally winners. It’s certainly true that poorer countries have – with some big exceptions – been more likely to run trade deficits over the last decade than richer ones:
Average annual GDP/capita (current US$) against average annual external trade balance on goods and services, 2002-12. Source: My calculations from World Bank World Development Indicators

The picture looks a little less clear when we consider whether those countries currently struggling to raise tax revenues (i.e. with low tax/GDP ratios) are more likely to run trade deficits or trade surpluses:

Average tax take/GDP (percent) against average annual external trade balance on goods and services, 2002-12. Source: My calculations from World Bank World Development Indicators
Nonetheless it’s clear that poorer countries aren’t necessarily a priori losers by tying their corporate tax bases to their external balance of trade.

The problem, though, is that a DCBT wouldn’t be charged on the value of imports and domestic sales of goods and services, but on the value of those sales minus costs (broadly speaking, the operating profits derived from those sales). So if you’re linking corporate taxation to the geographical location of sales on this basis, the two relevant questions for any given economy seem to be:
  1. Which corporate taxpayers are likely to be making larger operating profits on a given volume of sales – exporters or importers?

  2. Which corporate taxpayers are more likely to be subject to tax on those sales to begin with – exporters or importers?

The reality of most countries’ corporate tax bases is that taxable profits are highly concentrated, with the lion’s share of the corporate tax base generally contributed by the corporate giants in which capital and profits are concentrated. In the UK, 1% of companies pay 81% of all corporation tax. 86% of all UK corporation tax liabilities are found in companies that are part of a multinational group. (Larger companies pay a slightly higher rate of UK corporation tax, of course, but nowhere near enough to account for this disparity, and in any case large companies tend to pay out a smaller share of their trading profits in UK tax than smaller firms).

In developing countries, the picture is probably equally concentrated, if not more so. Figures specific to corporation tax are not widely available; but large taxpayers across sub-Saharan African countries, for example, typically constitute less than 5% of taxpayers and typically contribute three-quarters of all tax revenues; in Kenya, just 0.01% of all taxpayers pay 50% of all VAT. In addition, a large proportion of small businesses in many developing countries fall below the threshold for corporate profit tax entirely.

In short: in most countries the subsidiaries and affiliates of large multinationals constitute most of the current corporate tax base. And it seems reasonable to assume (though it would need testing empirically) that in smaller developing countries – those with limited domestic consumer markets - the majority of large multinationals operating there are export-oriented. In the absence of the multinational retail giants prevalent in developed and emerging economies, most of the large multinationals in smaller developing countries are those involved with mining; agribusiness; manufacturing for export, taking advantage of inexpensive labour; or service exports, like call-centres for foreign consumer markets. (There are a few obvious exceptions – particularly financial services, which is somewhat sui generis in tax terms anyway).

By contrast, the firms importing goods and services for domestic third-party sales in smaller developing countries are often smaller businesses, whose profit base furnishes just a sliver of the corporation tax take. I think it’s likely, therefore, that the taxable corporate profits in smaller developing countries derive more from export sales than from imports and domestic sales, regardless of whether the country as a whole is running a trade surplus or not. This is why I’m not as convinced as the Committee that the balance of trade is a good proxy for tax revenues from a DBCT. And why I suspect that a DBCT may risk shifting the tax base from poor to rich countries even for those developing economies that are not dominated by extractives exports.

(There are some proposed versions of the DBCT that tip up this calculus by proposing that countries cooperate to share DBCT revenues internationally – so that Country A, home to Company A exporting goods to Country B, sends a slice of the tax on profits from the export sales of Company A to Country B. This would presumably advantage trade-deficit countries: but outside EU VAT, there’s no mechanism I can think of where such cross-border revenue-sharing actually takes place, and it seems administratively and politically ambitious, to say the least…)

All that said, this argument is impressionistic. I haven’t been able to find good data about the firm-level distribution of taxable profits in developing economies. And above the ranks of the poorest economies, there will be obvious exceptions to the rule – particularly those emerging Asian economies where there is a growing domestic consumer market that foreign multinationals are increasingly trying to tap into. If anyone knows of good relevant data, or if cash-flow-tax theorists have done any modelling of the geographical distribution of the corporate tax base, I’d be grateful to hear about it?


  1. Interesting post. I think I agree with you on DBCT. It looks like it could well reduce tax for developing countries, and not sure it solves other problems.

    I am not quite sure the problem it is meant to solve.... if it is the 'Amazon problem' of companies with big revenues and small tax bills - that seems to be as much a problem of perception/misunderstanding as it is a real problem. Surely if taxes were levied on sales minus costs Amazon would still have a low tax bill as their business model is run on ultra low margins.

    If it is the Google problem how to tax virtual goods - I'm not sure it is possible to say where the location of a consumer of virtual products is - if a company in the UK buys ads from google which are served across multiple google sites to consumers in different countries, it is not obvious that the tax on the profits from those ads should go to the UK. And if the UK company is itself a multinational it could just relocate its ad-buying function to another country and deal directly with Google-Ireland if there was a discount involved. As more and more goods and services go virtual what's to stop consumers choosing to spend their money in different locations?

    Ultimately if the problem of 'multinationals not paying their fair share of taxes' (which is the problem as it is presented to the man/woman in the street) your statistics could solve that problem - If you asked people on the what % of corporation taxes are paid by MNCs I guess most would guess something lower than 86% (similarly to when people say that too much money is spent on international aid, based on a wildly exaggerated perception of how much of the government budget goes on aid).

  2. Thanks Maya. I agree that the DBCT is designed to do a number of different things, not always made very clear. In the current climate, it's being pitched as a solution to intra-group profit-shifting by reducing the deductibility of intra-group transactions (in this sense, in popular parlance it might be said that it's supposed to tackle the so-called 'Starbucks problem' rather than the 'Amazon problem' or the 'Google problem'?!) Whereas back in the Mirrlees Review, it was being pitched as an investment-friendly tax, intended only to tax economic rents and not the 'normal' return to investment.

    I like the 'Amazon problem' and 'Google problem'! But I think I would probably characterise them differently. I’m not sure the issue with Amazon is just about low margins leading to low tax bills (indeed historically, as retailers go, the Amazon group’s trading profit margins – around 3-4% in most years since 2006, until taking a more recent dive with expansion since 2011 – don’t seem that small to me? It’s no Apple, but it’s no slouch. In any case there are *some* profits, and with Amazon’s scale their quantum is considerable). The issue is surely with European profits being booked substantially elsewhere from where its customers, physical infrastructure and ‘fulfilment’ operations are actually located? And this seems only possible by having very limited PE definitions in your tax treaties to start with, and then going to their limit in practice.

    I agree that the ‘Google problem’ seems to be about the arbitrary location of intangibles. But the intangible that’s ultimately sucking profit out of Europe and the USA is not so much advertising revenues as royalties for Google’s search technology (In that respect it’s not really an ‘Ireland’ problem, it’s a ‘Bermuda’ problem, surely?) With respect to online advertising, I’m not sure there’s that much uncertainty about the location of the customer in the example you give? If a UK company is buying the advertising, then the ‘consumer’ of the advertising service is that company, in the UK, no? The question of where an advert’s viewers (‘consumers’) are located is surely immaterial, since they’re not paying for it. (There’s a more interesting question about whether people viewing and clicking on Google ads are in some respects the ‘labour force’ of Google, by doing the ‘work’ of refining Google’s algorithm; but I don’t think that’s what we’re discussing here). The ambiguity, therefore, is not about where online advertising can be said to take place, but about where online *sales* of advertising can be said to take place. And that’s an empirical question about whether it is Google staff in Ireland that are doing the selling and concluding of advertising contracts, or Google staff in the UK; or if it’s a Google bot doing the selling and contracting rather than a human being, where that bot could be said to be located.

    So in both cases I think you *can* fairly straightforwardly identify the physical location of the consumer of Amazon’s goods and Google’s advertising services. Purely in terms of administrability, that would argue in favour of sales-based or destination-based taxes, I think?

    Fair shares… while the 86% sounds impressive, it obviously tells us nothing about the comparative tax burden of MNCs and domestics i.e. whether MNCs or domestics are paying a greater share of their revenues or pre-tax profits in tax. It may simply tell us that business profits are fantastically concentrated in MNCs, I suppose...

  3. Hi Mike,

    I think it would be good if the Amazon/Google/Starbucks problems were more clearly articulated -In the public debate at least the 'Amazon problem' is expressed as big revenues and small tax bills, but you are right, is really something more similar to the Google problem about how to tax stuff in the age of intangibles and virtual sales. (although the point about revenues vs profits makes clear that there isn’t a huge pot of gold out there to be divied up, as the headlines tell us)

    I don't think it is at all obvious that the bulk of Amazon’s profits are made in the place where the warehouse is. If you think about the Amazon experience – the breadth of products, the cataloguing, reviews and lists, pricing, 3rd party Marketplace etc… none of this is embedded in the decisions made by managers of the UK fulfillment operation (although as you say, some of it is based on the work of consumers in the UK writing reviews etc…). Even the way the warehouse is organised is based on technology and ‘how to’ which was presumably developed in the US. So If the UK operation (and all the other European operations) were taxable as individual trading entities, all of these elements could be recognized as internal transfers, covered by transfer pricing rules? (and therefore as with Google the real issue may not be UK-Ireland but US-Bermuda-Luxembourg – or wherever they park their IP)

    It seems administratively simpler to have one transfer (UK company charges Amazon EU for fulfillment services) than hundreds going the other way (Amazon EU charges UK company for loads of bits of IP, inventory risk, centrally procured stock etc…). The fact that Amazon do it this way, rather than the other way does not seem to me to be Machiavellian tax planning so much as the logical way to structure a multi-country business based on the same technology. Either way the question is the same isn’t it – How much of the profit from Amazon sales arises from its warehousing and fulfillment operation? I don’t think the answer is 100% (as in the Guardian’s estimates), but its not 0% either – If the current answer is wrong could it not be challenged within existing rules (i.e. Amazon UK is not charging enough to Amazon EU) rather than needing to flip the whole taxation system around and examine hundreds of transfer prices?

    On the location of Google’s advertisers I don’t think this is straightforward – most of the big ones will be international companies – so if Google has to set up separate adwords billing centres in every country it sells to, what is to stop them buying their adwords in which ever country is cheapest? And again, each national adwords billing centre would then need some mechanism for paying for those ads back to Google.

    The Amazon/Google problems seem to be: As more and more things that used to be done by people are done by bots and algorithms which don’t need to be physically located either at the location of the consumer or the original producer (coder etc…) and which involve diffuse networks in place of individual knowledge (e.g. Amazon reviewers vs the knowledgeable bookshop manager) how should the tax system evolve? (and perhaps even more importantly what will happen to employment and what to do about it).

    Important questions, but public debate will need to go beyond the fair shares/avoidance rhetoric I think