This article appears in the following Oxford Review of Economic Policy issue: BUSINESS TAXATION IN A GLOBALIZED WORLD [View the issue table of contents]
What problems and opportunities are created by tax havens?
* Department of Economics, University of Connecticut, e-mail: dhammika.dharmapala{at}uconn.edu
 |
Abstract |
|---|
Tax havens have attracted increasing attention from policy-makers in recent years. This paper provides an overview of a growing body of research that analyses the consequences and determinants of the existence of tax-haven countries. For instance, recent evidence suggests that tax havens tend to have stronger governance institutions than comparable non-haven countries. Most importantly, tax havens provide opportunities for tax planning by multinational corporations. It is often argued that tax havens erode the tax base of high-tax countries by attracting such corporate activity. However, while tax havens host a disproportionate fraction of the world's foreign direct investment (FDI), their existence need not make high-tax countries worse off. It is possible that, under certain conditions, the existence of tax havens can enhance efficiency and even mitigate tax competition. Indeed, corporate tax revenues in major capital-exporting countries have exhibited robust growth, despite substantial FDI flows to tax havens.
Key Words: tax havens • governance • tax avoidance • tax competition • FDI
I would like to thank the editor Mike Devereux, my discussant Ken Mayhew, an anonymous referee, and conference participants at the Oxford University Centre for Business Taxation Summer Symposium, June 2008, for helpful comments. Any remaining errors are, of course, my own.
1 DH (2006) begin with the list of jurisdictions in Hines and Rice (1994, Appendix 2, p. 178), all of which also appear in Diamond and Diamond (2002) and various other sources, and match these with countries and territories for which current data on economic and other characteristics are available. This set of countries is listed in Table 1 (under the heading: ‘Tax havens (DH)’).
2 Proximity to major capital exporters is measured by Gallup el al. (1999) as the distance by air from the closest of New York, Tokyo, or Rotterdam. The coastal population measure is also constructed by Gallup et al. (1999).
3 See DH (2006). An exception is UN membership, which is slightly higher for small havens than for small non-havens.
4 Some limited but intriguing supporting evidence for this view is that inbound foreign direct investment (FDI) appears to be much more sensitive to corporate tax rates in countries with stronger governance institutions (DH, 2006).
5 Slemrod (2008) views tax havens within a broader context of practices that involve the ‘commercialization’ of state sovereignty, a concept that also encompasses involvement in facilitating money-laundering activities, and the issuance of stamps designed to appeal to foreign collectors. Slemrod (2008) finds that better governance is related not only to being a tax haven, but also to ‘pandering’ stamp issuance (though not to involvement in facilitating money laundering). He suggests a more general interpretation, arguing that better governance leads not only to greater policy credibility but also to a greater capacity to undertake welfare-enhancing government activities.
6 The scandal also spread to the UK—see ‘UK in Liechtenstein Tax Data Deal’, BBC Online (24 February 2008) at http://news.bbc.co.uk/2/hi/business/7261830.stm
7 The tendency for a disproportionate share of foreign portfolio investment (FPI, i.e. cross-border investment by individuals, or by institutions such as pension funds on behalf of individuals) to be located in tax havens seems to support this view. Data from the US Treasury International Capital (TIC) reporting system (available at www.treas.gov/tic/ and described in more detail in Desai and Dharmapala (2007)) shows that 21 per cent of US FPI is located in tax havens (as defined in DH (2006)). However, as the TIC data is based on survey responses by US financial institutions, it is unlikely that investors would evade taxes on these investments. More generally, evaluating individuals’ use of havens is complicated by the wider context of understanding FPI. Economists have long argued that there are substantial gains available to investors from international portfolio diversification (e.g. French and Poterba, 1991). Historically, investors have failed to achieve these gains, although this ‘home bias’ phenomenon appears to be eroding over time (Dharmapala (2008) calculates that in 2004 US investors held only 12 per cent of their equity portfolios in foreign stocks; however, from 2004 to 2005, increases in holdings of foreign stocks represented 43 per cent of the total increase in holdings of equity). If the home bias is caused by an irrational aversion to foreign assets, then there may be a case for subsidizing FPI; lax enforcement of home-country taxes on foreign-source income may be one way to implement such a subsidy.
8 The data on US FDI are from the US Bureau of Economic Analysis, and are available at http://www.bea.gov. The data on UK FDI are from the UK Office for National Statistics, available at http://www.statistics.gov.uk/. Tax havens are defined as in DH (2006). Note that FDI, which is carried out by multinational firms, typically involves holding a controlling stake in a foreign affiliate (in contrast to FPI, which is carried out by individuals or by institutions such as pension funds on behalf of individuals).
9 Note, however, that the FTC is limited to the home-country tax liability on the foreign income.
10 The exemption applies to ‘active’ income generated by the firm's normal business operations. Generally, ‘passive’ income from firms’ cash holdings or portfolio investments is taxed by the home country.
11 Deferral also applies only to ‘active’ income generated by the foreign affiliate's business operations. Passive income generated by the foreign affiliate's cash holdings or portfolio investments is taxed immediately by the US under the Subpart F rules.
12 The deferral advantage applies only to the active income generated by the foreign affiliate; the interest earned on the passive assets held by the foreign affiliate is subject to immediate US taxation under the Subpart F rules.
13 Thus, the primary difficulty with the arm's-length standard is usually thought to be its administrative feasibility. However, Devereux and Keuschnigg (2008) argue that it may be inefficient, even if perfectly implemented. In their model, heterogeneous firms endogenously choose whether to engage in arm's-length trade with a foreign entity or to acquire it, based on factors such as agency costs and financing constraints; imposing arm's-length pricing on those firms that choose to acquire the foreign entity can potentially distort their choices.
14 Outside the context of intellectual property, Swenson (2001) and Clausing (2003) find evidence of tax-motivated transfer pricing using data on international trade prices.
15 Huizinga et al. (2008) argue that the MNC's aggregate capital structure should itself depend not just on the tax rate in its home country, but on a weighted average of tax rates in the countries in which it operates; they find evidence consistent with this view.
16 Desai and Dharmapala (2008) use the introduction of the CTB regulations as a source of exogenous variation in their study of the impact of corporate tax avoidance on firm value: they construct instruments for tax-avoidance activity by interacting a dummy for the post-CTB period with time-varying firm characteristics that affect the incentives for tax avoidance.
17 As noted above, about a quarter of US and UK FDI is located in tax havens. By way of comparison, tax havens are home to only 0.7 per cent of the world's population when havens are defined as in DH (2006), or just 0.2 per cent when using the OECD definition.
18 In addition to these 35 countries, another six countries (listed in Hishikawa, 2002, p. 397, fn. 72) that otherwise satisfied the OECD's tax-haven criteria were not included on the list because they provided ‘advance commitments’ to eliminate allegedly harmful tax practices. The complete set of 41 OECD-designated havens is listed in Table 1 (under the heading ‘Tax haven (OECD)’).
19 OECD (2004) lists five recalcitrant tax havens that had failed to make such commitments as of 2004—see Table 1. The preferential regimes identified by the OECD have also generally been abolished or modified to remove the features that the OECD found objectionable.
20 These data are available at http://laborsta.ilo.org/. Unlike the investment data used by Kudrle (2008), employment data are not subject to the problem that evasion may occur through asset holdings that are unreported not just to tax authorities but also to statistical agencies. Unfortunately, the coverage of tax havens in the ILO data is very limited, rendering a cross-country longitudinal study of the impact of the OECD initiative difficult.
21 The OECD initiative initially included provisions directed at corporate activities, but this element was soon dropped—see Kudrle (2008, p. 7).
22 Klautke and Weichenrieder (2008) find evidence that another recent initiative directed at tax havens—the EU's Savings Directive—has been ineffective. This measure also targets individual evasion rather than corporate avoidance.
23 This prediction is, of course, consistent with the evidence on population size shown in Figure 1(a).
24 In addition, their equilibrium entails that the (small and relatively powerless) havens impose significant welfare costs on the populations of larger and more powerful countries. For example, many prominent tax havens are British dependent territories, yet the ‘negative’ view of havens implies that they pursue policies that are harmful to the UK. There are possible explanations for this state of affairs—for example, the UK may prefer to subsidize its territories by allowing them to become havens, the costs of which fall partly on other capital exporters, rather than by direct subsidies. However, this apparent disjuncture between theory and reality is not directly addressed in the literature critical of tax havens.
25 Slemrod and Wilson (2006) also model small open economies. They derive positive optimal corporate tax rates through a different route. Specifically, taxes on wages can be evaded, at some resource cost. Thus, governments prefer (up to a point) to tax workers indirectly by imposing taxes on capital rather than to tax them directly and thus incur these costs of evasion.
26 Equivalently, a country's pay-off can be defined as the (pre-tax) profits generated within its borders (whether by domestic or foreign firms), plus the social benefits (0.2 per dollar) of the revenue transferred from firms to the government. The profits generated can be viewed as a proxy for the amount of capital employed within the domestic economy, and hence for the wages of domestic workers (which, given the complementarity of capital and labour, are obviously higher when more capital is employed domestically). Incidentally, the assumption that tax revenue is socially valuable highlights that this is not a ‘Leviathan’ framework, in which tax competition can be beneficial by constraining a government inclined towards excessive spending.
27 C is thus a ‘financial’ haven, rather than a ‘production’ haven. Admittedly, this limits the generality of the example, but it none the less captures an important element of the characterization of tax havens in this literature.
28 Note that there is now no incentive for the mobile firms to move their real operations between A and B, as this incurs a real resource cost, while using the tax haven for financial transactions is assumed to be costless.
29 Slemrod and Wilson (2006) allow for heterogeneity among firms in their cost of using tax havens. However, the tax rates set by the government are not contingent on this cost.
30 The data on FDI and on US corporate tax revenues are from the US Bureau of Economic Analysis, and are available at http://www.bea.gov. In Figure 5(a), tax havens are defined as in DH (2006), while Figure 5(b) uses the OECD definition (see Table 1).
31 See Auerbach (2007) for an analysis of why US corporate tax revenues have increased.
32 Indeed, since 2001, personal rates have fallen while the corporate rate has been unchanged.
33 It is, of course, possible to imagine a counterfactual world in which corporate tax revenues would have grown even faster in the absence of tax havens. This, however, appears unlikely. For instance, in the USA, the recent increases shown in Figure 5(a) represent the reversal of a longer-term decline going back to the early 1960s (Auerbach, 2007). A counterfactual involving significantly higher growth in corporate tax revenues than has actually occurred would represent a dramatic departure from past experience. Incidentally, the long-term decline from the early 1960s until the 1990s cannot easily be attributed to tax havens, as it predates the emergence of the degree of international economic integration that would presumably be required for havens to be a major factor.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. P. Devereux Business taxation in a globalized world Oxf. Rev. Econ. Policy, December 1, 2008; 24(4): 625 - 638. [Abstract] [Full Text] [PDF]
|
|