CESifo Economic Studies Advance Access originally published online on June 4, 2007
CESifo Economic Studies 2007 53(2):172-228; doi:10.1093/cesifo/ifm010
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Can Capital Income Taxes Survive? And Should They?
* Department of Economics, University of Copenhagen, Studiestraede 6, 1455 Copenhagen K, Denmark,e-mail: peter.birch.sorensen{at}econ.ku.dk
 |
Abstract |
|---|
 Top Abstract 1 Introduction2 Capital income taxation...3 Capital income tax...4 Summing up and...AcknowledgementsReferences |
|---|
The article surveys some main results in the theory of capital income taxation in the open economy; reviews recent trends in international taxation and discusses alternative blueprints for fundamental capital income tax reform from the perspective of an open economy faced with growing mobility of capital income tax bases. (JEL code: H21, H25)
Key Words: Capital income taxation • international taxation • tax reform • corporate income tax
|
1 Introduction |
|---|
|
TopAbstract1 Introduction2 Capital income taxation...3 Capital income tax...4 Summing up and...AcknowledgementsReferences |
|---|
Can capital income taxes survive? And should they? In recent decades these questions have been debated with increasing intensity among tax economists and policy makers, as growing capital mobility and the ensuing tax competition has driven down statutory tax rates on capital income throughout the world. Those who believe that returns to savings should not be taxed have typically welcomed capital tax competition, hoping that it will push governments towards greater reliance on consumption-based taxation. However, many other observers fear that an erosion of capital income taxation will undermine the integrity and political legitimacy of the tax system and lead to greater inequality and to an underprovision of public goods—the well-known spectacle of a "race to the bottom".
In this article I discuss whether capital income taxes are actually vanishing, and whether and how we should attempt to tax capital income in a globalizing economy. These are big issues, so my discussion will have to be selective. Throughout the article I will focus on tax policy in the open economy, but I will not address the problem of international tax coordination. Instead, I will ask how the national government of an open economy can and should design its system of capital income taxation, taking the policies of other countries as given, and assuming that extensive multilateral agreements on tax coordination are not politically feasible at the present stage of international political integration.2
The rest of the article falls in three parts. Part 2 presents some theory and evidence on capital income taxation in the open economy. Part 3 discusses alternative blueprints for capital income tax reform in a globalizing economy, and Part 4 summarizes and compares the main reform options.
|
2 Capital income taxation in the open economy: theory and evidence |
|---|
|
TopAbstract1 Introduction2 Capital income taxation...3 Capital income tax...4 Summing up and...AcknowledgementsReferences |
|---|
2.1 Taxing capital income—some basic distinctions
Taxes on saving versus taxes on investment
Capital income takes many forms such as interest, dividends, capital gains, business profits, the value of the housing services enjoyed by owner-occupiers, etc. Some types of capital income are harder to tax than others, technically and politically, so in practice capital income tax systems tend to be highly complex, incoherent and discriminatory. I shall return to this problem later on. For the moment, I will focus on some basic theoretical distinctions which are fundamental for understanding the effects of capital income taxes.
The first distinction is that between source-based and residence-based taxes on capital. Under the source principle (the return to) capital is taxed only in the country where it is invested. Source-based taxes may therefore be termed taxes on investment. Under the residence principle the tax is levied only on (the return to) the wealth owned by domestic residents, regardless of whether the wealth is invested at home or abroad. Since wealth is accumulated saving, residence-based taxes may also be termed taxes on saving.
The most important example of a source-based capital tax is the corporate income tax, since most countries only tax corporate income generated within their borders.3 In contrast, the personal income tax as well as the personal wealth tax are based on the residence principle, since domestic residents are liable to tax on their worldwide capital income and on wealth invested abroad as well as at home. As a rough approximation, we may therefore say that the corporation tax is a tax on investment, whereas the personal taxes on capital income and wealth are taxes on saving.
In an open economy with free international mobility of capital, the two types of taxes have very different effects on the domestic economy and on international capital flows.4 This is illustrated in Figure 1 where the horizontal axis measures the volumes of domestic saving and investment, while the vertical axis measures the real rates of return on saving and investment. The downward-sloping curve labelled ‘I’ indicates how the level of domestic investment varies with the required pre-tax rate of return. The lower is the required return, the greater is the volume of investment which will be deemed profitable. The upward-sloping curve denoted by ‘S’ shows how the level of domestic saving varies with the after-tax rate of return. The positive slope of this curve reflects the common assumption that a higher after-tax return will induce a higher volume of saving.5
|
The horizontal curve labelled ‘r’ measures the cost of finance to domestic corporations. This is the rate of return that the investors providing the finance require after payment of corporation tax but before payment of any personal tax. If we consider a corporate investment financed by equity, then r is the required return on shares before payment of personal taxes on wealth and on dividends and capital gains on shares. If the investment is instead financed by debt, r is the rate of interest before personal taxes on interest and wealth. In Figure 1 the corporation tax is denoted by tI, so a corporate investment has to earn a minimum return rg = r + tI to cover the corporation tax bill as well as the cost of finance. Hence the level of domestic corporate investment is given by Ie. Furthermore, when domestic personal taxes (denoted by tS) are deducted from the return r offered in the markets for shares and interest-bearing assets, domestic savers end up with an after-tax return equal to rn, so the level of domestic saving becomes Se. The excess of domestic investment over domestic saving has to be financed by capital imports from abroad, so in the example given in Figure 1 the volume of capital imports is given by the distance Ie – Se.
In an open economy where investors can invest in foreign as well as domestic securities, the cost of finance r is given by the rate of return required in the international capital market. When the domestic economy is small, it does not have any noticeable impact on the international interest rate or the rate of return on shares required by international investors. Hence the level of r may be taken as given from the viewpoint of the small open economy. This means that personal taxes on saving do not have any impact on domestic investment. In Figure 1 a rise in the savings tax tS implies a movement down along the S-curve, leading to a lower volume of domestic saving and a higher level of capital imports without affecting domestic investment. By contrast, a rise in the investment tax tI reduces the level of domestic investment and leads to lower capital imports but does not affect domestic saving.
This analysis has important implications for tax policy. In particular, it shows that if the government of a small open economy wishes to stimulate domestic real investment through lower taxes on capital, it should concentrate on lowering source-based taxes on investment such as the corporation tax. According to Figure 1, a lowering of savings taxes such as the personal taxes on dividends and capital gains on shares will not stimulate domestic investment. Rather, it will stimulate domestic saving and will most likely imply that some shares in domestic companies that were previously owned by foreign investors will be taken over by domestic investors, thus increasing the share of the domestic business sector controlled by domestic owners.6
Taxes on normal returns versus taxes on rents
The second fundamental distinction in the theory of capital income taxation is the one between taxes on the normal return to capital and taxes on rents. By definition, rents are "pure profits" in excess of the going market rate of return on capital. For debt capital, the normal return is the market rate of interest on debt in the relevant risk class, and for equity it is the required market rate of return on stocks with the relevant risk characteristics. If markets for risk pooling are underdeveloped, the required risk premia will tend to be higher, and so will the normal return.
In a closed economy a tax on the normal return to capital will tend to reduce the volume of saving and investment (if the elasticity of saving with respect to the net return is positive), whereas a tax on pure rents will be non-distortionary. However, in the open economy a source-based tax on rents will reduce domestic investment if the business activity generating the rent is internationally mobile, that is, if the firm is able to earn a similar excess return on investment in other countries. In that case we may speak of firm-specific or mobile rents. In other words, when the economy is open, a source-based tax will be non-distortionary only if it falls on location-specific (that is, immobile) rents. Location-specific rents may be generated by the exploitation of natural resources, by the presence of an attractive infrastructure, or by agglomeration forces, whereas firm-specific rents may arise from the possession of a specific technology, product brand or management know-how.
A fundamental proposition on capital income taxation in the open economy
A fundamental theorem—originally derived by Gordon (1986) and forcefully restated by Razin and Sadka (1991)—states that in the absence of location-specific rents, a small open economy should not levy any source-based taxes on capital.7 Under perfect capital mobility a small open economy faces a perfectly elastic supply of capital from abroad, so the burden of a source-based capital tax will be fully shifted onto workers and other immobile domestic factors via an outflow of capital which drives up the pre-tax return. In this process, the productivity of the domestic immobile factors will fall due to a lower capital intensity of production. To avoid this drop in productivity, it is more efficient to tax the immobile factors directly rather than indirectly via the capital tax.
If governments pursue optimal tax policies, we should therefore expect to observe a gradual erosion of source-based capital income taxes in the recent decades marked by growing capital mobility. Is this actually happening?
2.2 Trends in corporate income taxation: a puzzle
To answer this question, let us look at the most important source-based capital income tax, i.e. the corporation tax. Here we are faced with a puzzle. On the one hand, the statutory corporate income tax rate has fallen significantly in almost all OECD countries since the early 1980s, as illustrated in Figure 2. On the other hand, corporate tax revenues relative to GDP have actually increased in most countries, the main exceptions being Germany, Japan and the UK, as shown in Figure 3. Hence the corporate income tax base must have grown sufficiently to (over)compensate for the drop in the statutory tax rate.
|
|
To explore these trends in greater detail, it is useful to decompose the ratio of corporate tax revenue to GDP in the following manner:8
|
| (1) |
In Figure 4 I have undertaken such a decomposition of the ratio of corporate tax revenue to GDP for a number of OECD countries for which the necessary national income statistics were available back to the early 1980s.10 Note that the concept of "profit" adopted here is the OECD measure of "operating surplus", defined as profit gross of interest and depreciation, i.e. a measure that is broader than the actual business income tax base. This explains why the estimated effective tax rates (R/C) are relatively low compared to the statutory tax rates in Figure 2. To eliminate some of the cyclical fluctuations in the data, Figure 4 shows 3-year moving averages centred on the year indicated on the horizontal axis.
|
We see that movements in the profit share of GDP have generally been very limited and hence cannot explain the tendency for the corporate tax revenue share to increase. However, Figure 4 also reveals that in most of the countries the ratio of corporate profits to total profits has tended to increase, reflecting the growing importance of the corporate organizational form. Thus part of the reason why governments have managed to stabilize or even increase the ratio of corporate tax revenues to GDP is that the corporate sector has expanded at the expense of non-corporate firms. In part this may reflect the secular decline of certain sectors such as agriculture where the non-corporate organizational form has tended to dominate, but it may also reflect income-shifting into the corporate sector induced by the large drops in statutory corporate tax rates.11 In both cases the gain in corporate tax revenue will tend to be offset by a loss of revenue from the personal income tax. However, none of the countries shown in Figure 4 have experienced a significant drop in their aggregate effective corporate tax rate (R/C), and in several countries such as Australia, Belgium, Denmark, Finland and Italy, the effective corporate tax rate even seems to have increased. In other words, it appears that governments have made up for the drop in statutory tax rates by broadening the corporate tax base, e.g. by eliminating special deductions and moving towards less generous rules for asset depreciation, etc. in line with the tax-cut-cum-base-broadening philosophy that became popular in the 1980s and early 1990s. In section 3.4 I will discuss some reasons why governments have moved in this direction.
2.3 Why do source-based capital income taxes survive?
Reasons for the viability of source taxation
Whereas the basic theory of optimal taxation suggests that globalization will gradually wipe out source-based capital income taxation, we have seen that corporation tax revenues have actually been very robust in recent decades. This section briefly discusses various reasons why source-based capital income taxes may be able to survive despite growing capital mobility.12
Location-specific rents. If firms can earn above-normal returns by investing in a particular location, the government of that jurisdiction may impose some amount of source tax without scaring investors away. Moreover, when location-specific rents co-exist with foreign ownership of (part of) the domestic capital stock, the incentive for national governments to levy source-based capital taxes is strengthened, since they can thereby export part of the domestic tax burden to foreigners whose votes do not count in the domestic political process [see Huizinga and Nielsen (1997)]. Since globalization implies increased international cross-hauling of investment and a resulting increase in foreign ownership shares, this may be an important reason why governments choose to maintain source-based capital income taxes, as suggested by Mintz (1994). Indeed, using European data, Huizinga and Nicodème (2003) found evidence of a significant positive relationship between average effective corporate tax rates and the share of domestic companies owned by foreigners. Nevertheless, since a source tax on the normal return can be fully shifted under perfect capital mobility, the factors mentioned earlier do not explain why governments choose to levy a distortionary source tax on the full return to capital rather than taxing only rents, e.g. by allowing a deduction for the normal return to capital. I return to this issue at the end of this subsection.
Imperfect capital mobility. The prediction that source taxes on capital will vanish assumes that capital is perfectly mobile. In practice, there are costs of adjusting stocks of physical capital so such capital cannot move instantaneously and costlessly across borders. Since adjustment costs tend to rise more than proportionally with the magnitude of the capital stock adjustment, the domestic capital stock will only fall gradually over time in response to the imposition of a source-based capital income tax [see Wildasin (2000)]. In present value terms, the burden of the tax therefore cannot be fully shifted onto domestic immobile factors, and hence a government concerned about equity may want to impose a source-based capital tax, particularly if it has a short horizon. Another reason for imperfect capital mobility may be that financial instruments (in particular shares) issued in different countries are imperfect substitutes because they have different risk characteristics, as will be the case when returns in different national stock markets are not perfectly correlated. Investors will then want to hold securities issued in different countries to diversify their portfolios, and individual countries will have some market power over the price of equity in domestic firms. Governments may exploit this market power by imposing withholding taxes at source on returns on domestic equity paid out to foreign investors, just as a country with international market power may want to impose an optimal tariff [see Gordon and Varian (1989)]. Gordon (2000, p. 31) points out that a roughly similar result may be achieved by imposing a source-based corporation tax and granting a personal tax credit for the corporation tax only to domestic shareholders, as many governments have actually done. Empirically, the imperfect mobility of capital is reflected in the well-documented "home bias" in investor portfolios, i.e. the fact that financial investors (especially small ones) tend to invest the bulk of their wealth at home rather than abroad [see, e.g. French and Poterba (1991)]. In part this home bias may reflect that (small) investors are less well informed about foreign than about domestic investment opportunities [Gordon and Bovenberg (1996); Westerhout (2002)].
Foreign tax credits. Some important capital exporting countries like the US, UK and Japan try to tax the worldwide profits of their multinational corporations while allowing a foreign tax credit for taxes paid abroad, up to a limit given by the amount of domestic tax on the foreign source income. In these cases the foreign host (source) countries have an obvious incentive to levy a source-based corporation tax, since they can do so without reducing the incentive of foreign multinationals to invest in the domestic economy, as long as the domestic corporation tax does not exceed the corporation tax levied by the home (residence) country of the multinational company. It is less clear why the home country government would want to offer a foreign tax credit which effectively amounts to a give-away of revenue to foreign host governments. Gordon (1992) argues that foreign tax credits are offered exactly because they give an incentive for foreign host countries to levy source-based taxes up to the limit given by the domestic tax rate. In this way the foreign tax credit enables the home country to tax domestic investment without inducing a capital flight from the domestic economy. However, as recognized by Gordon himself (Gordon 2000, pp. 27–28), this argument rests inter alia on the counterfactual assumption that home countries tax foreign source income upon accrual. In practice, home country tax is typically levied only when income is repatriated from abroad. As long as profits are retained and reinvested abroad, only the host country tax rate matters for the cost of capital, as Hartman (1985) showed, and taxation at the margin is then effectively source-based. Moreover, a large number of countries explicitly adhere to the source principle by exempting foreign source corporate income from domestic corporation tax. Overall, Gordon and Hines (2002) find it difficult to argue that crediting arrangements have a significant impact on host country corporate tax rates.
The corporation tax as a backstop to the personal income tax. One important function of the corporate income tax is to serve as a "backstop" for the personal income tax. The corporation tax falls not only on returns to (equity) capital but also on the labour income generated by entrepreneurs working in their own company. In the absence of a corporation tax, taxpayers could shift labour income and capital income into the corporate sector and accumulate it free of tax while financing consumption by loans from their companies. Still, while it is easy to see why protection of the domestic personal tax base may require a corporation tax on companies owned by domestic residents, it is not obvious why it requires a source-based corporation tax on foreign-owned companies whose shareholders are not liable to domestic personal tax. However, as pointed out by Zodrow (2006, p. 272), if foreign-owned companies were exempt from domestic corporate income tax, it might be relatively easy to establish corporations that are nominally foreign-owned but are really controlled by domestic taxpayers, say, via a foreign tax haven. Hence the backstop function of the corporation tax may be eroded if it is not levied on foreign-owned companies.
Political constraints. Finally, even though it may be inefficient to tax capital income at source, the voting public may not realize that such a tax tends to be shifted to the immobile factors, so levying a source-based corporation tax may be a political necessity, since abolition of such a tax would be seen as a give-away to the rich, including rich foreign investors. More generally, if there are political limits to the amount of (explicit) taxes that can be levied on other bases, it may be necessary for a government with a high revenue requirement to raise some amount of revenue via a source-based capital income tax, even if such a tax is highly distortionary.
In summary, there are a number of reasons why source-based capital income taxes have been able to survive the growing international mobility of capital, one of them being the existence of location-specific rents. The next subsection discusses whether such rents are likely to persist in the face of further international economic integration.
Agglomeration, trade costs and the future of capital income taxes
The modern theory of economic geography [e.g. Baldwin et al. (2003), Baldwin and Krugman (2002) and Kind, Knarvik and Schjelderep (2000)] highlights how location-specific rents are created and destroyed. It also suggests that the link between economic integration and taxation is more complex than previously thought. This theory seeks to explain the geographical pattern of economic activity by accounting for transportation costs as well as other barriers to international trade. The theory identifies factors which tend to disperse economic activities across the geographical space as well as factors working in favour of a geographical concentration (agglomeration) of activities. When firms deliver inputs to each other and there are costs of long-distance trade in inputs, they can save on transportation costs by locating near each other. Thus forward-linkages (input deliveries) among firms create an incentive for agglomeration. This incentive increases if agglomeration also stimulates technological spillovers among firms, for example, by making it easier for them to observe and learn from each others’ production processes.
On the other hand, firms also compete with each other in final product markets, so if product markets are to some extent geographically segmented due to transport costs, increased concentration of activity in one location will tend to intensify competition and drive down profit margins in that local market. This crowding effect creates an incentive for firms to operate in different locations to reduce the intensity of competition. However, if the crowding effect is not too strong, firms will tend to agglomerate (concentrate) in certain geographical areas. As a result of the cost savings arising from agglomeration, firms will earn location-specific rents in these areas, and these rents can be taxed away via the corporation tax without inducing firms to relocate to other jurisdictions.
Economic integration is driven by lower trade costs which reduces the individual firm's cost advantage of being close to the suppliers of its intermediate inputs. This tends to reduce the location-specific rents from agglomeration, thereby reducing the scope for source-based taxation. At the same time lower trade costs increases the scope for price competition via long-distance trade in final products. This will reduce the crowding effect from agglomeration, since the degree of competition becomes less dependent on the geographical closeness of the individual firm's competitors. A smaller crowding effect tends to increase the location-specific rent from agglomeration. It is quite possible that this latter effect of lower trade costs may be stronger than the former one, in which case lower trade costs will induce governments in countries with many agglomerations to increase their tax rates.
In fact models of economic geography typically imply a non-monotonic relationship between trade costs and the rents from agglomeration: starting from relatively high trade costs, a fall in these costs will at first tend to increase the location-specific rents from agglomeration, but when trade costs fall below a certain level, a further fall will reduce the agglomeration rents, as illustrated in Figure 5. The intuition is the following: with very high trade costs, there is no scope for trade in intermediate and final goods, so local markets can only be served by local firms, leaving no room for a geographical concentration of production and the associated agglomeration rents. At the other end of the spectrum, if trade costs are zero the price of inputs and the degree of competition in product markets becomes quite independent of the geographical location of production, so again there are no agglomeration forces. It follows that if there is any potential for agglomeration rents at all, these rents must initially rise from zero to some positive level as trade costs start to come down from a very high level, but at some point if trade costs continue to fall, agglomeration rents must start to decline again.
|
This suggests that, in the early stages of economic integration, a move towards more integration may well increase the scope for taxation of mobile factors for a while. But as integration deepens and trade costs lose importance, the location-specific rents that enable governments to levy source taxes without deterring inward investment tend to get wiped out. Thus, if the recent wave of globalization has so far involved a leftward move from a starting point to the right of the maximum of the curve in Figure 5, rising agglomeration rents may help to explain the observed robustness of corporate tax revenues, but according to the theory of economic geography this is no guarantee that source-based capital taxes can be maintained in the future if trade costs continue to fall. Although those who fear a race to the bottom in capital income taxes may seem to have cried "wolf" too many times, the wolf may yet show up some day!
2.4 Should the normal return to capital be taxed?
From a theoretical viewpoint, there is an obvious case for a tax on rents, since such a tax is in principle non-distortionary, at least if the rents are location-specific. A long-standing debate in the theory of taxation is whether governments should also tax the normal return to capital. This section reviews the main arguments in this perennial debate.13
As we have seen, a source-based tax on the normal return tends to be shifted onto the immobile domestic factors, so in the open economy a genuine tax on the normal return to saving must be residence-based, falling on income from foreign as well as domestic sources. The next section discusses whether the residence principle can actually be implemented, but in order to distinguish normative theoretical arguments from administrative concerns, the present section assumes that a residence-based capital income tax can indeed be enforced.
A tax on the normal return to saving increases the relative price of future consumption by raising the amount of present consumption that has to be given up to allow a unit increase in future consumption. Hence a capital income tax (on the normal return) may be seen as a selective commodity tax on future consumption. As argued by Chamley (1986) and Judd (1985), in the long run such a tax seems very hard to defend in an economy where consumers behave as if they have an infinite time horizon, say, because generations are linked through altruistic bequests in the manner described by Barro (1974). To see why, note that in period zero the relative price of consumption in some future period T (the amount by which consumption in period 0 must be cut in order to increase consumption in period T by one unit) is
|
| (2) |
t is the capital income tax rate for that period. The last bracket in this expression measures the tax-induced distortion to the relative price of future consumption. This distortion to PT will be larger the longer the consumer's time horizon T, and as the horizon tends to infinity, the tax distortion becomes infinitely large. Clearly it is hard to think of circumstances in which it would be optimal to impose an infinitely large tax distortion, and Chamley (1986) and Judd (1985) do indeed find that in the long run the capital income tax rate should be lowered to zero even if the alternative public revenue sources are also distortionary, and even if the government is concerned about equity as well as efficiency.14 To be sure, in the short and medium run where the relative price distortion caused by capital income taxation is finite, it is typically optimal to levy a positive capital income tax, but in a standard Ramsey type infinite-horizon model the tax rate should gradually be lowered as the size of the distortion increases. This policy conclusion assumes that capital markets are perfect so that consumers may freely reallocate their consumption over time through borrowing and lending. If intertemporal substitution in consumption is restricted by borrowing constraints, consumers can respond only to a limited extent to the relative price distortions caused by capital income taxation. When labour income is uncertain and insurance markets are incomplete, it may then be optimal to levy a capital income tax even if consumers have infinite horizons [Aiyagari (1993) and Chamley (2001)]. With constraints on borrowing, consumers will engage in precautionary saving to hedge against future rainy days, and consumption at any given date will tend to be positively correlated with accumulated precautionary saving. A tax on (income from) capital that finances a lump-sum transfer will then help to redistribute income from high consumption states to low consumption states, thereby performing an insurance function that the market cannot provide.15
A limitation of infinite-horizon models is that many consumers may not behave as if they live forever. For example, they may not have any heirs or may not care sufficiently about them to leave them any positive bequests, or their bequest motive may be selfish rather than altruistic. Many scholars have therefore studied optimal capital income taxation in models with finite consumer horizons such as the Diamond life cycle overlapping-generations model. In such models several rationales for a positive capital income tax rate have been identified, including the following:
Complementarity between leisure and future consumption. As pointed out long ago by Corlett and Hague (1953), a commodity tax system which minimizes deadweight loss will impose relatively high tax rates on goods that are more complementary to (less substitutable for) leisure, since this will minimize tax-induced substitution towards leisure. Because a capital income tax is a tax on future consumption, the Corlett–Hague rule implies that we should tax (subsidize) capital income if future consumption is more (less) complementary to leisure than present consumption. In the benchmark case where preferences are separable in consumption and leisure so that current and future consumption are equally substitutable for leisure, we have the powerful theorem of Atkinson and Stiglitz (1976) that a government concerned about equity as well as efficiency will not want to differentiate taxes across commodities (and hence should not impose a capital income tax), provided it can use a progressive non-linear labour income tax to achieve its distributional goals. The intuition for the Atkinson–Stiglitz result is clear: there is no second-best efficiency case for distorting the choice between present and future consumption when these goods are equally substitutable for leisure; nor is there any equity case for imposing a tax on future consumption, since a wage income tax is a better targeted instrument for redistributing income in a world where innate differences in labour productivity is (assumed to be) the only source of inequality. A widespread view among public finance economists is that we do not know whether future consumption is more or less substitutable for leisure than present consumption, so based on the principle of insufficient reason, we should assume equal degrees of substitutability. The Atkinson–Stiglitz theorem then immediately leads to the prescription of a zero tax rate on the normal return to capital. However, this view has recently been challenged by Erosa and Gervais (2002). They show that for any reasonable parameterization, a standard life cycle overlapping-generations model implies that consumption and leisure generally move together over time at the level of the individual consumer. In other words, it is reasonable to assume that future consumption is more complementary to leisure than present consumption, and hence it is efficient to tax future consumption via a capital income tax. The simulations undertaken by Erosa and Gervais (op.cit.) indicate that the optimal capital income tax rate is indeed significantly above zero for any plausible parameterization of the life cycle OLG model.
Mimicking age-dependent labour income taxation. Erosa and Gervais (2002) also point out that the elasticity of labour supply with respect to the relative price of leisure tends to increase with age in a realistic life cycle model. Ideally, the government would therefore want to impose an age-dependent labour income tax, with a marginal tax rate that declines with age. If the labour income tax cannot be conditioned on age, the government can (imperfectly) imitate an age-dependent labour tax by levying a positive capital income tax. To see this, note that if the marginal after-tax wage rate in some future period T is WT, the relative price of leisure in that period (as seen from the present when the consumer plans the time profile of his labour supply) is PT WT, where PT is given by equation (2) above. By imposing a capital income tax t the government can thus raise the relative price of leisure in later stages of the consumer's life cycle where labour supply tends to be more elastic.
Capital-skill complementarity and endogenous factor prices. The analysis of Atkinson and Stiglitz (1976) was static and did not allow for the fact that saving for future consumption induces capital accumulation which in turn affects pre-tax factor incomes. However, subsequent research has shown that, provided the government has other instruments (such as debt policy or public investment) to steer the capital stock towards its socially optimal level, a zero steady-state capital income tax rate remains optimal in explicitly dynamic OLG models with capital accumulation, as long as utility functions are weakly separable in consumption and leisure [Ordover and Phelps (1979)]. But the reason why dynamic models of capital accumulation confirm the optimality of a zero capital income tax is that capital accumulation does not affect the distribution of pre-tax wage rates in these models: a higher capital stock simply raises the level of real wages in the same proportion across all skill types. Yet there is strong empirical evidence that skilled labour is more complementary to capital than unskilled labour [see, e.g. Krusell et al. (2000) and Lindquist (2005)], implying that capital accumulation will tend to raise the relative wages of skilled workers.
As shown by Salanié (2003, p. 143), it then becomes optimal to impose a positive capital income tax since this will redistribute pre-tax income in favour of low-income earners by discouraging capital formation. Intuitively, the introduction of a small positive capital income tax will only cause a second-order efficiency loss, but at the same time it will cause a first-order social welfare gain through a more equitable income distribution.
Offsetting the labour tax distortion to human capital investment. Bovenberg and Jacobs (2005a) have pointed to another rationale for a positive capital income tax: when human capital investment is endogenous and its opportunity cost is not fully tax deductible (say, because part of the opportunity cost is a non-pecuniary cost of effort), a labour income tax will discourage human capital formation and induce consumers to substitute financial capital for human capital. If the government can fully observe all human capital investments and their related opportunity costs, education subsidies are the most well-targeted instrument to offset the labour tax distortion to human capital investment, as Bovenberg and Jacobs (2005b) have shown in a companion paper. However, realistically governments cannot verify all human capital investments and their associated costs. In that case it becomes second-best optimal to supplement education subsidies by a positive capital income tax to offset the tendency of the labour income tax to favour financial investment over human capital investment. The more the government wishes to redistribute income through a progressive labour income tax, the higher is the optimal capital income tax rate needed to offset distortions to human capital investment.
Inherited wealth and heterogeneous preferences. If governments cannot observe inherited individual wealth (say, because of bank secrecy rules) and hence cannot tax such wealth directly, it may be optimal to impose a linear capital income tax (e.g. an impersonal withholding tax) as an indirect way of taxing inherited wealth, as shown by Boadway, Marchand and Pestieau (2000) and Cremer, Pestieau and Rochet (2003). Saez (2002) also argues that a positive capital income tax is part of an optimal tax system in the empirically relevant case where high-productivity individuals have a higher propensity to save than low-productivity individuals.
In summary, recent research has shown that a case for a positive capital income tax rate can be made both on efficiency and equity grounds.
2.5 Can taxes on saving be enforced?
While there may be a normative case for a positive residence-based tax on the normal return to saving, tax economists tend to argue that the residence principle cannot be enforced in practice, since governments cannot effectively monitor foreign source income in the absence of systematic international exchange of information among tax authorities. To be sure, bilateral tax treaties do provide for exchange of information on specific taxpayers upon request from the residence country, and sometimes treaties also encourage "spontaneous" provision of specific information by the source country, but it is widely agreed that information exchange on such limited scale is insufficient to bring all foreign source capital income into the tax net of residence countries. According to the pessimistic but widespread view expressed by Tanzi and Zee (2001), the basic problem is that source countries seem to have no interest in providing information since they thereby make themselves less attractive as a haven for foreign investors.
However, some authors have recently tried to identify circumstances in which it would indeed be in the mutual interest of source and residence countries to engage in systematic international information exchange [see, e.g. Bacchetta and Espinosa (2000), Eggert and Kolmar (2002), Huizinga and Nielsen (2003), and Keen and Ligthart (2006)]. In particular, Keen and Ligthart show that if the residence country is willing to transfer an appropriate share of its revenue gain from information exchange to the source country, both countries will benefit from systematic information exchange.
Moreover, recent years have seen significant progress in international cooperation on information exchange. As part of the OECD initiative to counter "harmful tax practices", 30 out of 35 jurisdictions originally identified as international tax havens have now committed themselves to providing information to foreign tax authorities, and under the recent EU agreement on the so-called savings directive, 22 out of 25 EU member states have implemented automatic information exchange from July 2005. The remaining three member countries will adopt information exchange from 2011, relying on withholding taxes until then. To make the switch to information exchange more attractive to these member states, the EU savings directive contains an innovative clause requiring them to transfer 75 percent of the revenue from withholding taxes to the residence country.16
Furthermore, a large part of household saving is channelled through institutional investors such as mutual funds, pension funds and life insurance companies that may hold significant amounts of foreign assets. Because of their limited number and public reporting requirements, the foreign source income of such institutional investors is much easier to monitor than the foreign income of individual households. Given the strong revealed household preference for channelling saving through these financial intermediaries, it should be technically possible to enforce residence-based taxation of the return to most of private sector saving by collecting the tax at the level of the institutional investors.17 In fact, Denmark and Sweden do tax the return to retirement savings undertaken via banks, pension funds and life insurance companies.18 Still, the international norm is to exempt the return to retirement saving from tax, but this suggests that the failures of governments in implementing the residence principle stem to a large extent from a deliberate political choice to exempt a big chunk of capital income from tax.
This last observation raises the issues: how much revenue do we actually collect from capital income taxes, and how much of this revenue represents a tax on normal returns as opposed to a tax on rents?
2.6 Do we actually collect any revenue from taxing normal returns?
Table 1 makes a first attempt to estimate the importance of capital income taxes in the tax structures of OECD countries. The difficulty is that official OECD revenue statistics do not decompose the total revenue from personal income taxes into tax falling on capital income and tax levied on labour income. The estimates in the first and fourth columns of Table 1 assume that the fraction of personal taxes imposed on capital income equals the capital income share of total household income. The capital income share was estimated by Carey and Rabesona (2004) using OECD national accounts and imputing a "normal" labour income to the self-employed. With this decomposition we see from Table 1 that the personal tax on capital income typically contributes between 5 and 10 percent of total tax revenue in OECD countries, but in several countries the contribution is significantly lower. In most countries the corporate income tax is a more significant revenue raiser than the personal capital income tax, while the importance of property taxes—a mixed bag including a number of taxes on the ownership and transfer of real and financial assets—varies quite a lot across countries.
|
The decomposition of personal tax revenues into capital income tax and labour income tax in Table 1 implicitly assumes that capital income is taxed at the same effective rate as labour income, as would be the case under a comprehensive income tax. But as already noted, capital income often benefits from favourable tax treatment. Table 2 seeks to adjust for the most important tax preferences for capital income for a group of OECD countries for which the necessary data were available. Based on the estimates in Carey and Rabesona (2004), Table 2 accounts for tax reductions due to tax preferences for retirement saving, reduced or zero tax on the imputed rent from owner-occupied housing, relief from double taxation of dividends and the existence of low flat tax rates on (certain types of) interest income. With these adjustments, which are necessarily rough, we see by comparing the first columns in Tables 1 and 2 that the revenue share collected from personal capital income taxes falls substantially in all the countries considered, being almost completely wiped out in countries like Germany, the UK and Spain. In all of the countries in Table 2, the personal tax on capital income contributes considerably less to total revenue than the corporation tax and taxes on property.19
|
Since the corporate income tax allows a deduction for interest on debt, it falls to a large extent only on pure rents. Moreover, the finding in Table 2 that the personal tax on capital income raises very little revenue suggests that taxes on normal returns do not matter much for the financing of the public sector. Against this background, it is not surprising that several studies have found that current tax systems collect practically no revenue from taxes on the normal return to capital. For example, Gordon and Slemrod (1988) estimated that in 1983 tax revenue in the United States would have slightly increased in present value terms if the US had eliminated all personal taxes on capital income and all deductions for interest expenses and had allowed expensing rather than depreciation of investment spending on physical capital (thus effectively switching to a cash flow business tax which is known to fall only on pure rents). For 1995, Gordon, Kalambokidis and Slemrod (2004) estimated a slightly positive net revenue from the corporation tax on the normal return, but the revenue only amounted to 4 percent of total corporate profits. Sørensen (1988) found that personal capital income taxes in Denmark generated significant net revenue losses in the mid-1980s (due to liberal provisions for interest deductibility), and Becker and Fuest (2005a) recently estimated that a switch to a consumption tax in Germany—which would amount to exemption for the normal return to saving—would at most cause a revenue loss of 1.6 percent of GDP and might even yield a net revenue gain.
The fact that so little revenue is raised from taxes on normal returns does not mean that existing capital income taxes are roughly non-distortionary. On the contrary, since the capital income tax base is not adjusted for inflation, the marginal effective tax rate on some forms of saving and investment may be quite high, whereas other forms of saving and investment may receive big tax subsidies. For example, if all of the nominal interest income is subject to full personal tax, the effective tax rate on the real interest income may be quite stiff even at modest rates of inflation. On the other hand, in countries that allow deduction for mortgage interest payments without taxing imputed rents, investment in owner-occupied housing will tend to be heavily subsidized by the tax system. Thus, even though on average the tax wedge between the marginal pre-tax return on investment and the marginal after-tax return to saving may be quite low, current systems of capital income taxation probably imply large distortions to the pattern of saving and investment. These observations explain why so many tax economists tend to favour some form of consumption-based taxation which would leave the normal return to capital free of tax. If so little revenue is raised at the cost of such big distortions, why don't we simply give up the idea of taxing the normal return?
In the next part of this article I will discuss how the tax system of an open economy could be redesigned if governments only wish to tax rents, but I will also discuss how policy makers could reform the system of capital income taxation if they want to make a serious attempt at taxing normal returns.
|
3 Capital income tax reform in the open economy |
|---|
|
TopAbstract1 Introduction2 Capital income taxation...3 Capital income tax...4 Summing up and...AcknowledgementsReferences |
|---|
3.1 A classification of alternative blueprints for capital income tax reform
In most OECD countries the corporate income tax is a cornerstone of the system of capital income taxation. As indicated in Table 2, the corporation tax typically raises a lot more revenue than any other form of capital income tax, and its importance is even bigger than suggested by the revenue statistics because it serves as a backstop to the personal income tax. Hence most of the discussion subsequently will focus on the corporate income tax, although section 3.3 will also consider the role of other capital income taxes.
Since interest payments are deductible from the corporate tax base and are often lightly taxed in the hands of the recipient, say, because they accrue to tax exempt institutional investors, the system of capital income taxation tends to favour debt finance over equity finance. Further, dividends are often taxed at a higher effective rate than the capital gains on shares arising from retained corporate profits, so the tax system also tends to favour equity finance through retained earnings rather than finance via new equity.20 Apart from distorting financing decisions, the corporate tax system may distort real investment decisions through depreciation schemes that allow depreciation for tax purposes to deviate from the true economic depreciation. Finally, the corporation tax may distort the choice between corporate and non-corporate forms of business organization.
Over the years, many capital income tax reforms in OECD countries have been motivated by a desire to reduce these well-known distortions. In recent years policy makers have also paid increasing attention to reform proposals intended to make the corporate tax system more robust in a world of growing mobility of capital and of taxable profits. Against this background, Table 3 offers a classification of alternative ways of taxing corporate source income in an open economy.21 The rows in the table categorize the different tax systems according to the location of the tax base. Specifically, the tax base may be the corporate income earned in the source country where production takes place; it may be the income earned in the residence country of the company's corporate or personal shareholders, or it may be the sales (net of costs) to the destination country where the final consumption of the company's products occurs. The choice between these alternative tax bases may have important implications for the allocation of capital and economic activity between the domestic and the foreign economy.
|
The columns in Table 3 classify the different capital income tax systems according to the type of income subject to tax. Traditionally, the corporation tax has been levied on the full return to corporate equity income which includes the normal return as well as pure rents. Alternatively, the corporation tax might be levied on the full return to all capital invested in the corporate sector, including debt capital. Finally, a number of reform proposals would impose corporation tax only on rents, leaving the normal return free of tax. Alternative policy choices in this dimension will affect the degree to which the corporation tax distorts the investment, financing and production activities of companies, and they may also affect international location decisions.
Most existing corporate income taxes belong to categories 1 or 2 in Table 3. Under these tax regimes the source country has the prime right to levy tax on corporate profits after deduction for interest payments. The residence country of a multinational parent company may then either exempt the income from foreign corporate affiliates from domestic tax (category 1, exemplified by countries like France, Germany, Italy, Spain, the Benelux countries and the Nordic countries), or it may subject foreign source profits to domestic corporate income tax but grant a credit for the source country tax against the domestic tax bill (category 2, which includes the Anglo-Saxon countries and Japan). Under the credit system the domestic tax on "active" business income is typically deferred until the time of repatriation of foreign source income. For this reason, and because the foreign tax credit typically cannot exceed the amount of domestic tax on the foreign income, the effects on investment incentives of such "worldwide" taxation are rather similar to the effects of the pure source-based system in category 1.22 Some countries integrate the corporate and the personal income tax by offering a credit for (part of) the corporation tax against the personal taxes levied on shareholders. However, these countries typically do not extend such tax credits to foreign owners of shares in domestic companies.
As a first approximation, it is therefore fair to say that existing corporate income taxes are source-based, or at least their incentive effects are very similar to those of a pure source-based tax system, due to the common practice of deferring residence-country tax until the time of repatriation of the foreign-source income [Hartman (1985)]. Given the high mobility of capital and the opportunities for international profit-shifting through transfer-pricing, royalty payments and thin capitalization, a source-based tax system is vulnerable to tax competition from foreign jurisdictions. This may be termed "the capital flight problem". All the reform proposals included in Table 3 can be seen as an attempt to reduce this problem in one way or another (as well as an attempt to create greater tax neutrality in the domestic sphere). Thus, a pure residence-based tax seeks to avoid the problem of capital flight by imposing the same effective tax rate on residents’ investment at home and abroad and by exempting inward foreign investment from domestic tax. A source-based corporation tax with an equity allowance and a source-based cash flow tax potentially reduce the capital flight problem by exempting the normal international return to capital from tax, recognizing that this tax is likely to be shifted onto domestic immobile factors anyway. Furthermore, a destination-based cash flow tax provides no incentive to locate investment abroad rather than at home, and in its VAT type form it also avoids the problem of international profit-shifting, as we shall see. Finally, the Comprehensive Business Income Tax and the Dual Income Tax seek to address the capital flight problem by broadening the tax base to keep the tax rate low.
The following sections briefly discuss the various options for capital income tax reform summarized in Table 3. I start in section 3.2 by introducing some notation and by describing a conventional source-based corporate income tax. Section 3.3 considers alternatives to source-based taxation, i.e. residence-based and destination-based taxes. The main rationale for these reform options is that they tend to minimize the distortionary impact of taxation on corporate location decisions. In section 3.4 I proceed to discuss alternative forms of source-based taxation which are arguably easier to implement than residence-based taxes. In both sections I consider taxes that fall on the full return to (equity) capital and taxes that fall only on rents. Throughout the discussion I am assuming that corporate tax reforms are unlikely to be undertaken in an internationally coordinated manner. Thus I consider reforms that are carried out by one country in isolation.23
3.2 A conventional source-based corporation tax
To clarify the distinction between the alternative tax bases discussed subsequently, I shall use the following notation:
R = net cash flow from "real" transactions, excluding net capital spending (sales of goods and services minus purchases of goods and services minus labour costs)
I = net capital spending (purchases of capital goods minus sales of capital goods)
K = stock of real assets invested
= rate of depreciation
B = net financial debt
i = interest rate on debt
= corporate income tax rate
T = corporate tax bill
In applying the above notation, I consider a domestic parent company earning income from domestic operations as well as from affiliates operating abroad. Variables relating to the domestic economy are indicated by a superscript d while variables relating to the foreign economy are denoted by the superscript f. For simplicity, I assume that the parent company and its affiliates do not earn financial income from non-debt instruments.
With these assumptions, the domestic corporate tax bill under a conventional source-based corporate income tax (category 1 in Table 3) may be written as
|
| (3) |
Thus a traditional source-based corporate income tax allows deduction for depreciation and interest and exempts foreign source income from the domestic tax base. Note that, as long as the foreign source income is not repatriated, or if the foreign tax rate exceeds the domestic tax rate, equation (3) also describes a conventional residence-based corporation tax with deferral and a limitation on the foreign tax credit (category 2 in Table 3).
As already mentioned, the tax system summarized in equation (3) favours debt finance over equity finance, distorts the international location of investment and is vulnerable to international profit-shifting. In the absence of domestic double tax relief, it also discriminates against investment in the corporate sector of the economy. The sections subsequently consider alternative options for corporate tax reform seeking to address some or all of these problems.
3.3 Alternatives to a source-based corporation tax
Taxing the full return to equity: a residence-based shareholder tax?
One radical option for reform, once advocated by McLure (1979) and Feldstein (1987), might be to abolish the source-based corporate income tax altogether and to impute all of the income of a company to its shareholders, subjecting that income to the residence-based personal income tax (category 3 in Table 3). Under such a tax regime, the domestic corporate tax bill would always be zero (Td = 0), and the tax bill of domestic personal shareholders (Tpd) would be
|
| (4) |
The main rationale for imputing all corporate income to the individual owners of the corporation is that only individuals can bear a tax burden and that the impersonal corporate income tax is ill suited to achieve distributional goals. A further argument is that such full imputation of corporate income to shareholders would ensure a fully identical tax treatment of corporate and non-corporate firms and a neutral tax treatment of the different sources of investment finance (debt, retained earnings and new equity). Third, since a residence-based tax can be avoided only by moving abroad, and since individuals are much less mobile across borders than capital, a switch to worldwide income taxation at the personal shareholder level would greatly reduce the capital flight problem, provided that tax on foreign source income can be enforced. Nevertheless, there are a number of arguments against such a radical reform. For example, subjecting retained corporate profits to personal tax at the shareholder level on an accruals basis would create a liquidity problem for shareholders who receive little or no dividend from the company. Moreover, it is hardly realistic to expect that foreign corporations or foreign tax authorities would be willing to provide the information necessary to impute the retained profits of foreign companies to domestic holders of foreign shares.24 In the absence of an accruals-based tax on capital gains on (foreign) shares, the domestic shareholder could then defer his tax on retained profits in foreign companies until the time of realization of the shares. This would favour the holding of foreign shares relative to investment in domestic shares. To eliminate tax discrimination between domestic and foreign shares, it would be necessary to give up imputing retained profits in domestic companies to their domestic shareholders and to rely instead on a personal tax on realized capital gains on shares. However, in the absence of tax at the corporate level, this would leave a loophole in the tax system, allowing deferral of personal tax through the shifting of income into the corporate sector. The deferral advantage would also violate domestic investment neutrality and cause a "lock-in" of capital in existing corporations by reducing the cost of capital for corporate investment financed by retained earnings. Because of these problems, abolishing the corporation tax and imputing corporate income to shareholders is hardly a realistic reform option.
Taxing the full return to equity: a residence-based corporation tax?
As indicated earlier, an important function of the corporation tax is to ensure that at least some tax is levied on retained corporate profits. However, imposing tax on domestic-source corporate profits accruing to foreign investors is not necessary if the main function of the corporation tax is to serve as a backstop to the domestic personal income tax. Thus one could imagine that the corporation tax were based on a consistent residence principle where all corporations headquartered in the domestic country are taxed on their worldwide income without any deferral, but where no tax is levied on domestic-source corporate profits accruing to foreign owners of corporate capital operating in the domestic economy. This would be a "pure" version of option 2 in Table 3. In practice the elimination of source taxation could be implemented by granting a credit to foreign shareholders for (their proportionate share of) any corporation tax paid by foreign-owned companies operating in the domestic country. By abolishing source taxation in this way, the incentive for inward investment by foreign investors would greatly increase. Of course, this comes at the price of giving up the revenue from the taxation of inward foreign direct investment.
To avoid international double taxation, the residence-based corporation tax would be combined with a credit for taxes paid abroad. In the pure version of the system, the credit would be given without limitation, ensuring full capital export neutrality. The domestic corporate tax bill would then be
|
| (5) |
However, company headquarters may be quite mobile internationally, so a residence-based corporation tax may not enable a country to maintain an (average) effective tax rate significantly above that prevailing in other countries, since companies can escape a high tax rate by moving their residence to a low-tax country. Moreover, it would be very hard to implement a consistent residence-based corporation tax [Sørensen (1993)]. The main problem would be the auditing of and enforcement of domestic tax on profits of foreign affiliates that are retained abroad. Like a residence-based shareholder tax, a consistent residence-based corporation tax thus does not seem a viable reform option.
Taxing rents: a destination-based cash flow business tax?
An important argument for maintaining the corporation tax is that it serves partly as a non-distortionary tax on pure rents. If such rent capture is seen as the main function of the corporation tax, there is a case for designing it so as to leave the normal return to capital free of tax. This case is strengthened in a small open economy where capital mobility implies that a source tax on the normal return to capital tends to be shifted onto the less mobile factors of production, as I explained earlier.
A cash flow tax on the real and/or financial surplus of firms allows an immediate expensing of investment. Because the present value of the cash flows from a marginal investment is just equal to the initial investment outlay, the cash flow tax therefore leaves marginal investment projects free of tax, falling only on pure rents. In a closed economy this feature would ensure that a cash flow tax would be non-distortionary. However, in an open economy the rents earned by multinational companies often derive from firm-specific assets and may be generated in many alternative locations. When the fixed costs of doing business are so large that multinationals choose to serve several national markets from a single location rather than producing in all countries, a cash flow tax on internationally mobile rents will therefore affect the international location decisions of multinationals, even though it will not reduce the privately optimal scale of local investment once a company has decided to locate in a particular country.
A cash flow tax on the net total of the firm's real and financial flows is essentially a dividend tax with a deduction for new share issues. In principle this tax base is more narrow than the conventional corporate tax base. In recent years most discussions of the cash flow tax have assumed that the tax would be levied on the so-called R-base which includes only cash flows from the firm's "real" transactions, leaving out financial transactions, including interest payments. This is also the assumption made here (although it may be worth considering whether the tax base for firms in the financial service industry should include financial cash flows).
As indicated by options 7 and 8 in Table 3, there are alternative ways of designing a cash flow corporation tax in an open economy.25 One unconventional option is a VAT-type destination-based cash flow tax. The tax base under this scheme is sales to domestic customers minus purchases from domestic suppliers and minus labour costs. Thus, export sales are not taxable, whereas all imports are taxed. In formal terms, the total domestic tax collected is
|
| (6) |
is the total export sale of domestic firms, and 
is the total sale of foreign firms to domestic customers (i.e. total imports). The tax base is therefore equal to the current VAT base minus labour costs, assuming that the tax is levied on all firms, and not just on corporations. Because of the formal similarity with the VAT, it is possible that such a tax could get the status of an indirect tax consistent with current international tax law so that domestic tax could also be levied on the domestic sales of foreign-based firms. In that case the tax would not only fall on firms located in the domestic economy; it would also fall on firms servicing the domestic market from abroad. This is one of the attractions of the destination base: because the tax on sales to the domestic market cannot be avoided by moving production abroad, the system minimizes the incentive to relocate. Because it allows deduction for wage costs, the VAT-type cash flow tax is basically a tax on domestic consumption out of non-wage income. Pure rents are taxed only to the extent that they are consumed by residents in the domestic jurisdiction. Hence the VAT-type cash flow tax will not distort the investment and location decisions of firms, but at the same time it will not enable the domestic government to capture any of the rents accruing to foreigners.
A very attractive feature of the VAT-type destination-based cash flow tax is that it eliminates the transfer-pricing problem: since the proceeds of a sale to a foreign customer are not included in the tax base, the price that related companies within a multinational group use to account for an export transaction has no impact on the amount of tax paid. The same holds for an import from a related foreign party: the price set by the parties does not matter for tax purposes because there is no deduction.
On the other hand a destination-based cash flow tax raises a special transition problem: since exports are tax exempt whereas imports are taxed, the domestic-currency prices of exports have to fall (or the domestic price level has to increase) by the amount of the tax to restore equilibrium in domestic and foreign product markets. Because of short-run nominal wage and price rigidities, this adjustment process could cause considerable friction unless the exchange rate is flexible. But even if the adjustment were handled smoothly through an appreciation of the domestic currency, there would be windfall gains and losses as domestic residents with net claims on foreigners (denominated in foreign currency) would experience an erosion of the real value of these claims, whereas domestic residents with net liabilities vis á vis foreign countries would benefit from a fall in the real value of their debts. If investors dealing with foreigners anticipate the switch to the destination-based cash flow tax, they will try to rearrange their portfolios so as to reap a gain and/or to avoid a loss. The resulting capital flows could cause disruptions in capital markets and foreign exchange markets around the time of reform. Any future anticipated tax rate changes after the introduction of the tax would tend to cause similar disruptions.
Because of these problems with a destination-based tax, it seems worthwhile to consider the alternative of a source-based cash flow tax.
3.4 Alternative source-based business tax systems
Taxing rents: a source-based cash flow business tax?
As already mentioned, there is an efficiency case for taxing domestic location-specific rents accruing to foreigners. This may be achieved through a source-based cash flow tax where tax is levied on the net cash flow from domestic production, i.e. the cash flow from domestic and foreign sales minus the cash expenses on the purchase of domestic and imported inputs, including the purchase of capital goods, and minus labour costs (option 7 in Table 3). Formally, the domestic business income tax bill would be
|
| (7) |
Such a tax will cut into all pure rents earned from domestic production. As long as the tax does not induce companies earning mobile rents to relocate their production, the source-based cash flow tax is a non-distortionary means of shifting rents from foreigners to domestic residents (via the public budget). However, if the tax becomes too high, it will cause a shift of production out of the domestic e