Memorandum from Professor Michael Devereux,
Professor of Economics and Finance, University of Warwick
TAX HAVENS AND THEIR IMPACT ON THE UK
1. These notes consider a number of the economic
issues underlying recent initiatives by the OECD and the EU to
combat "harmful tax competition". The details of the
initiatives are not discussed. Instead the focus is on issues
such as: whether "harmful" forms of competition can
be differentiated from those which are not harmful; whether taxes
on capital have fallen and whether this combined with increasing
taxes on labour should be seen as adversely affecting unemployment;
and on the fundamental difficulties in the existing systems of
taxing international flows of income.
2. SPILLOVERS FROM
2.1 Countries are not independent of each other.
A change in the tax regime in one country is likely to affect
the welfare of other countries. There are at least two distinct
ways in which this might happen.
2.2 First, a tax cut in one country which increases
the probability of a multinational firm locating there necessarily
makes it less likely that the firm will choose some other location.
In practice, most leading industrial countries aim to encourage
inward real investment, which might have positive effects on domestic
welfare through raising total investment and employment, or by
increasing productivity, for example.
2.3 Second, introducing a new low tax regime
is likely to attract mobile taxable income (through a variety
of devices) from other countries. This may have little or no impact
on real economic activity, but serves to reduce the taxpayers'
overall tax liability, and possibly increase the tax revenue of
the host country, at the expense of higher-taxed countries.
2.4 These two have different effects on welfare:
the first concerns transfers of real economic activity, while
the second concerns transfers of tax revenue. In principle, "tax
havens" may belong to either or both groups. It is not clear
which of these transfers the OECD report "Curbing Harmful
Tax Competition" is concerned with. On the one hand, it explicitly
links tax havens with attracting "geographically mobile activities,
such as financial and other service activities". This appears
to tie tax havens into the first kind of transfer, but linked
to very mobile activity. On the other hand, the OECD defines a
tax haven in terms of a jurisdiction that "offers itself
. . . as a place to be used by non-residents to escape tax in
their country of residence". This ties it to the more popular
notion that tax havens engage in the second kind of transfer.
3. IS THERE
3.1 Measuring the extent of the first of these
spillovers is very difficult, although there is some evidence
to suggest that taxes do affect the location decisions of multinational
Some idea of the impact of the second of these spillovers can
be found by examining tax revenues. If the existence of tax havens
has become an increasing problem for the UK and other comparable
countries, it would be reasonable to expect this to be reflected
in, for example, the size of corporation tax revenues. Although
the European Union has frequently suggested that "taxes on
capital" have fallen while "taxes on labour" have
risen, the evidence actually suggests that corporation tax receipts
in the EU and in OECD countries have been roughly constant over
a long period.
This is not necessarily inconsistent with a growing threat from
tax havens, but it is at least suggestive that tax havens do not
pose a significant threat, at least to corporation tax revenues.
4. IS "TAX
4.1 In most economic markets, competition increases
efficiency and drives down prices to the consumer; this is why
we have the OFT and MMC. To be more precise, in perfect markets
prices are driven down to be equal to marginal costs; firms with
high marginal costs are driven out of business. Some commentators
have tried to apply these ideas in such a way as to suggest that
"tax competition" must also be beneficial. But can these
simple concepts be applied to "tax competition"? The
answer is no.
4.2 The "prices" involved in tax competition
are tax rates on mobile tax baseswhich can apply to the
location of investment, or the location of taxable income, given
some investment pattern. Reducing the "price" of locating
activity in one country makes that country more attractive relative
to other countries. But in the long run, other countries are likely
to respond by reducing their tax rates as well. Many economic
models suggest that the outcome of this process must be "prices"i.e.,
tax ratesof zero.
4.3 But there are costs of reducing tax rates:
the opportunity costs of raising the same tax revenue from other
sources, or going without the expenditure made possible by the
tax. These costs may vary from country to country. But high costs
do not necessarily imply "inefficiency"; they will reflect
the alternative means available to different governments of raising
the same revenue.
4.4 The fact that most OECD countries continue
to have high corporation tax ratesrelative to the theoretical
prediction of zerois consistent with them having relatively
high opportunity costs. Important in this calculation is the size
of the domestic tax base. Cutting the corporation tax rate in
the UK has a high cost in terms of the tax revenue generated from
domestic business. But there is a much lower cost in, say, Ireland,
and a lower cost still in, say, Jersey.
4.5 So from this discussion, there is no reason
to suppose that a process of tax competitionwhether competition
over the location of real economic activity or over taxable incomeis
beneficial. Attempts by the OECD and the EU to define "harmful"
tax competition appears to suggest by default at least that there
is such a thing as tax competition which is not harmfulor
even positively beneficial. But it is very hard to come up with
examples of such competition. I have some sympathy with the OECD
and the EU in attempting to address only some of the issues raised
by tax competition. However, I am not persuaded that the issues
which they address can really be characterised as those which
5. SHOULD WE
5.1 The EU has suggested that lowering tax rates
on capital income as a result of competition has led to a rise
in taxes on labour. It has linked this to the problem of high
unemployment in Europe. This argument is completely mistaken.
It is based on the fallacy that taxes on capital are effectively
borne by the owners of capital.
5.2 Consider a company which pays corporation
tax. Who bears the effective burden (or effective incidence) of
the tax? The shareholders may see a reduction in dividends and
in the share price. But in a competitive market, it is likely
that firms will attempt to pass on the tax in terms of higher
prices to consumers, or lower wages paid to employees. The effective
incidence is uncertain.
5.3 Now consider a small economy with capital
inflows. Non-residents have the opportunity to invest anywhere
in the world. Suppose they could earn 10 per cent at home or on
world markets. And suppose that this country charges a 50 per
cent tax rate on capital income. The non-resident investors will
still demand a return of 10 per centbut this must be after
tax. This means that they require a pre-tax rate of return of
20 per cent in that country. This is most likely to be achieved
by reducing total inward investmentselecting only the better
5.4 So who has borne the cost of the tax? Not
non-residentswho still earn 10 per cent post-tax. It must
therefore be borne by residents: there is less investment in the
country, and this will be reflected either in higher unemployment
or lower wages, or both. A more efficient alternative would be
to tax residents directlya tax on labour. Such a tax would
not distort inward investment decisions. So inward investment
would be higher and overall welfare would be higher. Local residents
still bear the tax burden. This argument largely carries over
when considering larger countries; except that for a large country
there may be some impact on the "world" rate of return.
5.5 So the distinction between taxes on capital
and taxes on labour turns out to be far from clear cut. In fact
it has been argued that a process of tax competition which drives
taxes on capital income to zero may after all be goodbecause
it prevents distortions to investment patterns, while not affecting
the effective incidence of taxes. But this is going too far. Countries
which accept the analysis which suggests that zero capital income
tax rates are beneficial always have the option to reduce such
taxes; that they choose not to do so presumably reflects other
factors. Tax competition, however, may leave them no choice.
6. THE OECD APPROACH
6.1 Both the EU and the OECD focus on tax regimes
with a low or zero effective rate of tax. The EU's code of conduct
is aimed at persuading Member States not to engage in such tax
measures, especially in the context of essentially discriminatory
treatment, with benefits given only to non-residents, ring-fenced
from the domestic economy, or without there being any real economic
activity. The OECD supplements its definition with a consideration
of the exchange of information, and makes a number of recommendations
concerning combating such regimes.
6.2 Of course, both initiatives face very difficult
problems. For example, for the EU, even if such practices were
completely removed from the EU, there remain plenty of jurisdictions
outside the EU which offer scope for reducing the tax on international
investment flows. To be effective, these initiatives must be supplemented
with penalties applying to taxpayers who make use of identified
regimes. Even then, the problem of definition of the regime is
very difficult and likely to become increasingly more so.
7.1 Some international income flows are taxed
primarily in the country in which they originate (the source country);
this is true of profit earned by multinational corporationsalthough
such profit may also face further tax when repatriated. But some
international income flows are taxed primarily in the country
of residence of the recipient. This is true of most interest flows,
for example. Each of these approaches creates difficulties; further
difficulties are created since both are used simultaneously.
7.2 The problem of tax havens primarily results
from income being taxed on a source basis. Suppose instead that,
say, UK resident multinational firms were taxed purely on a residence
basisthey would account for their world-wide income to
the Inland Revenue, and the country in which the income was earned
would be irrelevant. If this system could be made to work, there
would be no incentive to locate in a tax haven, or any low tax
country. (However, note that on its own this would not remove
the possibility of tax competition, since countries may seek to
give their firms a competitive advantage by reducing their tax
7.3 Such a system makes sense in economic terms.
Suppose I have an idea in London, I develop that idea in a laboratory
in Paris, I produce the subsequent good in Milan, financing it
in Frankfurt, and I sell it all over Europe. Where did I make
the profit? There is no logical way to allocate the total profit
earned between different locations. Yet that is precisely what
the international tax system attempts to do.
7.4 Of course, the principal problem with a
residence based tax is evasion: domestic tax authorities find
it difficult to identify income earned abroad (and quite possibly
not repatriated). This problem is at the heart of the EU's proposal
to introduce a withholding tax on individual savings income.
7.5 Modern financial instruments combine elements
of both debt and equity finance. As such, they create problems
even in a domestic settingfor example, interest payment
are deductible for corporation tax; so corporation tax is levied
on the return to shareholders, but not to creditors. But these
problems are multiplied in an international setting where alternative
financial arrangements make it relatively easy for taxpayers to
choose a preferred location in which to declare income.
7.6 So problems of tax havens should be seen
in the light of the fundamental problems of the international
tax system. It is easy to have sympathy with both the OECD and
the EU: attempting to reform the basis on which international
income is taxed is a gargantuan task. But in that context, their
recent initiativeseven if reasonably successfulwill
leave these fundamental problems untouched.
29 April 1999
1 One estimate is that a reduction of one percentage
point in the effective average corporation tax rate in the UK
increases the probability of a US multinational locating in the
UK instead of elsewhere in Europe by 1 per cent (Michael P Devereux
and Rachel Griffith, "Taxes and the location of production:
evidence from a panel of US multinationals", Journal of
Public Economics (68)3, 1998, 335-367). Back
See Lucy Chennells and Rachel Griffith, Taxing Profits in a
Changing World, Institute for Fiscal Studies (1997). Back