Chapter 3. The case for and against individual net wealth taxes

This chapter reviews the arguments for and against individual net wealth taxes. These arguments relate to the efficiency, equity and administrative implications of net wealth taxes. The effects of capital income taxes and taxes on wealth transfers are also discussed to examine how these taxes interact with net wealth taxes and whether they can be complements or substitutes to taxes on net wealth.

  

This chapter is based on the tax rules that were in place as of 1 September 2017. Since then, France has replaced its net wealth tax (“impôt de solidarité sur la fortune”) with a new real estate wealth tax (“impôt sur la fortune immobilière”), with effect from 1 January 2018.

This chapter is a review of the arguments both for and against individual net wealth taxes. The chapter discusses these arguments and assesses their validity. These arguments relate to the efficiency, equity and administrative implications of net wealth taxes. The effects of capital income taxes and taxes on wealth transfers will also be discussed to examine how these taxes interact with net wealth taxes and whether they can be complements or substitutes to taxes on net wealth. Overall, this chapter concludes that, both from an efficiency and equity perspective, there are limited arguments for having a net wealth tax on top of well-designed capital income taxes – including taxes on capital gains – and inheritance taxes, but that there are stronger arguments for having a net wealth tax as an (imperfect) substitute for these taxes.

Characteristics of net wealth taxes

Capital is typically taxed through both income and property taxes (Table 3.1). Capital income taxes are levied on the flow of income from assets. Property taxes, on the other hand, are levied on assets. Property taxes can be subdivided into two major categories of taxes – taxes on the transfers of property and taxes on the use and ownership of taxes. In this section, net wealth taxes, which are defined as recurrent taxes on individuals’ net wealth stocks, will be contrasted with both taxes on personal capital income and other types of property taxes. It should be mentioned, however, that capital income can also be taxed through social security contributions, as has been the case in France with the “contribution sociale généralisée” and that the boundary between capital income taxes and property taxes may sometimes be blurry as, for instance, taxes on immovable property levied on the basis of a presumptive net income and which take into account taxpayers’ personal circumstances, are classified as income taxes under the OECD’s Revenue Statistics classification (OECD, 2016).

Table 3.1. A typology of taxes on capital

Type of tax

Examples

Taxes on capital income

Corporate income taxes

Personal capital income taxes (on interest, dividends, rents, capital gains)

Property taxes

Taxes on property transfers

Inheritance/estate and gift taxes

Taxes on financial and capital transactions

Taxes on the use and ownership of property

Recurrent taxes on immovable property

Recurrent taxes on individual net wealth

Recurrent taxes on businesses’ net assets

Non-recurrent taxes on property (e.g. sporadic capital levies)

Source: Authors.

Comparing net wealth tax with personal capital income taxes

There are similarities between the taxation of net wealth and the taxation of capital income. For instance, if an individual taxpayer has a total net wealth of EUR 10 million that earns a rate of return of 4% (or a return of EUR 400 000), the tax liability will be the same whether the government levies a tax of 30% on the capital income of EUR 400 000 (EUR 120 000) or a wealth tax of 1.2% on the capital stock of EUR 10 million. This means that a capital income tax of 30% would equate to a wealth tax of 1.2% where the rate of return is 4%.

A key difference, however, is that a net wealth tax is imposed irrespective of actual returns. Net wealth taxes do not tax the actual return earned on assets but are equivalent to the taxation of a presumptive (i.e. fixed) return. This implies that, as opposed to a capital income tax, a net wealth tax implicitly imposes a lower effective tax on the return of high-yield assets compared to low-yield assets. Using the example above, if the return increases to 5%, the capital income tax liability will increase to EUR 150 000, while the wealth tax liability will remain the same, implying a drop in the effective tax on the return. This feature of wealth taxation is central to any discussion concerning its equity and efficiency effects.

A net wealth tax can be more comprehensive than a capital income tax. As opposed to capital income taxes, under a net wealth tax, even the assets that do not generate monetary returns are generally taxed. For instance, artworks that increase their owner’s wellbeing but do not generate any monetary returns until they are sold are often (at least partly) included in the tax base. Assets that generate returns that are not readily observable (e.g. owner-occupied housing generating an imputed return) are also taxed.

A third key difference is that a wealth tax is in theory levied on an accrual basis. Indeed, under the assumption that the wealth tax base is kept up to date through regular asset valuations, the appreciation in asset values is taxed every year under a wealth tax. On the other hand, under income taxation, taxes are levied on a realisation basis, meaning that taxes are levied when assets are sold on the increase in value between the time they were purchased and sold. In theory, it would be possible to tax capital gains on an accrual-equivalent basis but this has very rarely been implemented in practice. Accrual-based taxation has a number of advantages: it does not create lock-in effects and the resulting inefficiencies in capital allocation and it enhances fairness as appreciations in asset values are a better reflection of taxpayers’ current wealth. However, taxation on an accrual basis involves numerous practical difficulties, and if asset values are not regularly updated, the wealth tax becomes more comparable to a tax on a realisation basis.

Another major difference between the taxation of capital income and the taxation of wealth is related to their capacity to raise revenues in a volatile economic environment. If capital income is equal to zero or negative, the tax liability will also be zero under a capital income tax, while it will still be positive under a wealth tax if the capital value of the assets remains positive. As a result, net wealth taxes are a more stable tax revenue source than capital income taxes but they differ in their automatic stabilisation properties (Keen, 2014), which may have consequences on entrepreneurship and risk-taking as discussed below.

Interaction between capital income tax rates and the net wealth tax base

Taxes on capital income reduce the net expected return on the existing capital stock and therefore reduce the value of the assets in which the tax rate has been capitalised. For instance, the value of corporate shares reflects the net present value of the after-tax dividends to which the share owners are entitled. An increase in the taxes on dividend income will therefore lower the value of those shares, at least in the short run when firms cannot change their capital stock and investment behaviour in response to the tax increase (see also the discussion on the different views of dividend taxation in OECD, 2007). The more inelastic the supply of capital, the larger the extent to which taxes will be capitalised into asset prices. As immovable property is typically inelastic in supply, changes in the tax rates levied on housing will generally have a strong impact on house prices.

The capitalisation of taxes creates some interdependence between the taxes on capital income and the net wealth tax. An increase in taxes on personal capital income will reduce the value of income-generating assets. An increase in capital income taxes will therefore result in a smaller net wealth tax base, and potentially in a reduction in wealth tax revenues.

Similarly, a reduction in capital income taxes can be expected to increase asset values, generating windfall gains for existing asset owners, on the one hand, and increasing the net wealth tax base, on the other hand. This applies to mortgage interest relief as well. Mortgage interest relief lowers the income tax liability but may also increase the value of immovable property which, if housing is included in the tax base, will broaden the net wealth tax base. Depending on the design of the tax relief and the income and wealth taxes, the income tax advantage that high-wealth taxpayers obtain might be neutralised, to some extent, by the increase in wealth taxes as a result of the increase in housing prices. Whether the negative impact on income tax revenues of a capital income tax reduction will outweigh the positive impact on net wealth tax revenues when asset prices increase will depend on the design of income and wealth taxes in general, and on the degree of tax capitalisation and the level of the wealth tax rate in particular.

These observations not only apply to income taxes but also apply to the impact of the wealth tax itself. An increase in the wealth tax rate, for instance, will decrease the value of the assets whose price is predominantly set by individuals who are subject to the net wealth tax in that jurisdiction, thereby narrowing the net wealth tax base. The wealth tax is likely to affect its own tax base through tax capitalisation.

Comparing net wealth taxes with other taxes on personal property

Property taxes cover a wide variety of taxes levied on the ownership, transfer or use of property. According to the OECD Revenue Statistics classification (OECD, 2017), they include recurrent taxes on immovable property, which can be levied on property owners, tenants or both; recurrent taxes on net wealth, which include both individual and corporate taxes; estate, inheritance and gift taxes; taxes on financial and capital transactions; and non-recurrent taxes on property. These taxes have different goals as well as different effects in terms of revenue, efficiency, equity and tax administration. In practice, the mix of property tax instruments used by countries varies across the OECD (see Chapter 1).

As opposed to other taxes on wealth holdings, in particular recurrent taxes on immovable property, net wealth taxes are levied on a broad range of capital stock or property. Net wealth taxes are levied on immovable property, movable assets and financial investments. The main rationale is that total wealth stocks are a better reflection of taxpayers’ ability to pay. In addition, as financial assets make up a large share of wealth at the top of the wealth distribution, taxing total wealth is more progressive than taxing exclusively immovable property. However, a net wealth tax involves more practical challenges, largely because of the mobility of financial assets and the difficulty associated with valuing some categories of infrequently traded assets.

In contrast with sporadic capital levies or taxes on the transfer of capital, net wealth taxes are levied on a regular basis (usually annually). A sporadic capital levy has attractive features as a revenue raising instrument because it is a lump sum tax. This means that it taxes past wealth already accumulated, and therefore should not cause any distortions. While there are good arguments for such a tax, there have been in practice few successful examples of such taxes (Keen, 2014). Inheritance or estate taxes are also only levied once either on the deceased donor or on the recipient(s).

As opposed to inheritance taxes, wealth taxes are levied on both inherited and self-made wealth. Under a net wealth tax, there is no distinction between wealth resulting from personal effort and lifetime savings, inherited wealth, increases in asset values or luck (e.g. lottery). As discussed below, there may be both equity and efficiency justifications for taxing different sources of wealth differently.

Similar to inheritance or estate taxes, however, wealth taxes are imposed on net assets, meaning that debts are deductible. Net assets are a closer reflection of taxpayers’ ability to pay than gross wealth. They differ in this regard from recurrent taxes on immovable property, which are measured gross of debt. Not allowing for debt deductibility under recurrent taxes on immovable property may limit highly leveraged housing investment and the accumulation of high household debts but may raise distributional concerns: two taxpayers who own property of comparable value but with different levels of debt will pay the same amount of property tax. What countries do instead is to provide mortgage interest relief from PIT.

The case for net wealth taxes

This section reviews the arguments in favour of net wealth taxes. Arguments in favour of net wealth taxes broadly fall into two categories: (1) arguments for having a wealth tax on top of capital income taxes and other taxes on property and (2) arguments for having a wealth tax as a substitute for taxes on capital or capital income. This section reviews both sets of arguments.

Reducing wealth inequality and promoting equality of opportunity

As discussed in Chapter 2, private wealth is much more unequally distributed than income and there is some evidence suggesting that wealth inequality has increased in recent decades. While it is very difficult to assess wealth distribution trends over time, some studies point to increasing wealth inequality in recent decades. For instance, Piketty (2014) compiled data from eight OECD countries since the 1970s and concluded that, like income, private wealth has tended to become more unequally distributed in recent decades, reversing a long-term decline throughout much of the 20th century. Evidence of greater wealth concentration is particularly strong in the United States where Saez and Zucman (2016) find for instance that the share of total household wealth owned by the top 0.1% increased from 7% in the late 1970s to 22% in 2012. Bricker et al., estimated that the increase was from 11% to 15% (see Chapter 2).

These recent wealth distribution trends have strengthened the distributional case for taxing net wealth. Indeed, because wealth is highly concentrated at the top of the wealth distribution, even a low proportional tax on wealth holdings can increase progressivity. In a few OECD countries – in particular Nordic countries – which tax personal capital income at a flat rate, wealth taxes have been justified as a way of adding progressivity to the taxation of capital (Silfverberg, 2002). Data for Norway shows that the net wealth tax makes the overall tax system progressive at the top of the income distribution (Norwegian Ministry of Finance, 2017). Also as a consequence of the high concentration of wealth at the top and the large amount of private wealth, in theory even a low tax with a high exemption threshold excluding the lifecycle savings of most taxpayers could still raise a sizeable amount of revenue, although this is not what happens in practice (see Chapter 1).

A key aspect of wealth accumulation is that it operates in a self-reinforcing way; wealth begets wealth. High earners are able to save more (i.e. the marginal propensity to save increases with income), meaning that they are able to invest more and thereby ultimately accumulate more wealth. Moreover, returns tend to increase with wealth. Wealthy taxpayers, who tend to have more diversified asset holdings, are in a better position to invest in riskier assets which will tend to generate higher returns. The ability of the wealthiest taxpayers to generate higher average returns may also come from their access to wealth management and tax planning services as well as different investment opportunities (e.g. mutual funds that have entry requirements) (Fagereng et al., 2016). Rich taxpayers are also more likely to obtain loans, which will in turn allow them to invest more and accumulate more wealth. Finally, it may be argued (see below) that wealth begets more power, which may ultimately beget more wealth. Overall, this means that, in the absence of taxation, wealth inequality will tend to increase.

Capital income taxes alone will most likely not be enough to address wealth inequality. If a tax is only levied on the return to investment, the post-tax return will largely – at least for the wealthiest taxpayers – not be consumed, but be added to the principal and re-invested, thereby generating further (and likely higher) returns and allowing wealth to continue accumulating. For individuals at the top of the wealth distribution, even in the cases where a large portion of post-tax returns to investment are consumed, it will most likely be in the form of luxury purchases such as high-value immovable and movable property, which will also end up increasing their capital stocks. This suggests the need to complement capital income taxes with a form of wealth taxation in order to address wealth inequality and this report argues that there is a strong case for an accompanying inheritance tax (see below). Capital income taxes could also potentially be designed in ways that enhance their effectiveness in addressing wealth inequality but this would need to be explored in future work.

There is a clear case on distributional grounds for taxing wealth transfers at death. Although there is limited evidence on the relative importance of inherited wealth in total wealth and in the persistence of wealth inequality, there is a strong case for taxing wealth transfers to reduce intergenerational inequality and increase equality of opportunity by reducing and dispersing wealth holdings at death. A recent study in the United Kingdom (Hood and Joyce, 2017) shows that inheritances increase with income: lifetime inheritances are 4.4% of net lifetime income for the top quintile and 3.6% for the bottom quintile, compared with around 2% for the second and third lifetime income quintiles. In addition, they highlight that today’s elderly have more wealth to bequeath than their predecessors, largely because of higher homeownership rates and rising house prices, which means that the wealth of younger generations is more likely to depend on who their parents are than in the past. On the other hand, Elinder et al. (2015) find that inheritances in Sweden reduce inequality. Wealthier heirs inherit larger amounts, but less affluent heirs receive substantially larger inheritances relative to their pre-inheritance wealth. However, they find that if the revenues raised from inheritance taxes are redistributed to the less wealthy, then the total effect of inheritance taxation makes the wealth distribution more equal. Even though it shows that inheritances reduce inequality, the latter study can be viewed as supporting the case for progressive inheritance taxes, which would involve taxing large inheritances but not taxing or taxing at very low rates small inheritances received by poor taxpayers, to ensure that the equalising effect of inheritances identified in the study is not offset by inheritance taxes.

Additionally, there are meritocratic arguments for taxing inherited wealth more than self-made wealth (Piketty et al., 2013). Inheritances constitute an unearned advantage for recipients (Iara, 2015). From an equal opportunity perspective, wealth transfers can be viewed as a source of additional opportunity that is not linked to the recipient’s effort and that should therefore be taxed, regardless of whether the donor has already paid income tax or capital gains tax on the assets (Boadway et al., 2010). As with wealth taxes, it makes sense to have an exemption level that avoids taxing the majority of people who leave small inheritances. This reduces the number of people subject to tax without losing much of the potential revenue.

In practice, however, and as discussed in more detail later, there are many factors that reduce the positive effects of net wealth taxes on equity. The narrowness of the tax base limits the actual progressivity of net wealth taxes. Some of the assets which are widely held by the wealthiest are often exempt or preferentially taxed. Part of the limited effect of the wealth tax on redistribution also comes from tax avoidance and evasion opportunities, which allow the wealthiest taxpayers to minimise their wealth tax burden. Other design features, in particular tax caps (see Chapter 4), limit the progressive effect of wealth taxes. Finally, limited effects on redistribution come from the limited revenues raised through wealth taxes.

Wealth provides benefits above and beyond income

According to this argument, taxpayers with high wealth have greater resources to draw from and should be taxed at a higher rate than taxpayers with fewer assets even if they earn the same level of income. Simply put, there is a difference in ability to pay between a taxpayer who earns an annual income of EUR 20 000 from a EUR 200 000 investment, and a taxpayer who earns a salary of EUR 20 000 a year (Rudnick and Gordon, 1996).

Indeed, wealth confers advantages over and above the income derived from wealth. In addition to the income it generates, wealth may bestow social status, power, greater opportunities, satisfaction, or provide an insurance value against unexpected future needs, and it has been argued that such benefits should be taxed (Meade, 1978). Besides, wealth can provide income without having to sacrifice leisure (McDonnell, 2013). In some cases, assets do not generate income but still provide the benefits mentioned above. In that sense, a wealth tax can be seen as a complement to income tax, reflecting the additional advantages and capacity provided by wealth. Others have argued, however, that because the benefits of holding wealth are not measurable, it would be difficult for the tax system to take them into account, and that those types of benefits typically accrue to the very wealthy, so that a separate wealth tax would only be justified, if at all, on taxpayers at the very top of the wealth distribution (Boadway et al., 2010).

An alternative way of taking into account the benefits that wealth confers would be through the use of wealth-testing for broader tax and benefit purposes. Tax systems in OECD countries typically do not use information on household net wealth to determine taxpayers’ income tax liability. Information about taxpayers’ wealth could be used to strengthen the fairness of the income tax and the benefit system. Tax privileges for private pension savings, for instance, could be made dependent on the level of household wealth. In theory, capital income taxes could also be designed to increase with both income and wealth. These issues could be examined in future work.

Net wealth taxes could be an efficient substitute for capital income taxes by encouraging a more productive use of assets

In addition to arguments for having a wealth tax on top of capital income taxes, there is an argument for having a wealth tax as a substitute for capital income taxes, which could encourage taxpayers to use assets more productively. Given that a wealth tax is imposed on accumulated assets irrespective of the income they generate (see above), it may encourage taxpayers to use assets more productively. For instance, if a household owns land which is not being used and therefore does not generate income, no income tax will be payable on it. However, if a wealth tax is levied, the household will have an incentive to make a more productive use of their land or to sell it to someone who will (McDonnell, 2013). It has been argued by some that imposing a presumptive tax on capital income can be viewed as a tax on “potential income” (see Box 3.2). Guvenen et al. (2017) developed a theoretical model which suggests that replacing capital income taxes with a wealth tax shifts the tax burden onto unproductive entrepreneurs and that this reallocation increases aggregate productivity and output. Indeed, efficiency gains can occur because capital is reallocated to high-return individuals, and because the higher return of high-return individuals can motivate the accumulation of greater saving (Fagereng et al., 2016). The argument here is that wealth taxes do not discourage investment per se but discourage investments in low-yielding assets and reinforce the incentives to invest in higher-yielding assets because there is an additional cost to holding assets, which is not linked to the return they generate.

Box 3.1. A presumptive tax on capital income or a net wealth tax as a tax on ‘potential income’?

According to Faulk, Martinez-Vazquez and Wallace (2006), presumptive income taxes do not necessarily have to be seen as an approximation of the tax liability under normal tax accounting rules but they could also be interpreted as a measure of the tax burden on “potential” income. The authors focus on the taxation of the return on human capital. Depending on the individual’s work effort, potential income may be more or less than earned income. A tax on potential income rewards those individuals that work harder and earn more than their potential income and penalises those individuals that earn less than their potential income. However, this clearly does not imply that a potential tax would not create any economic distortion. Faulk, Martinez-Vazquez and Wallace (2006) explain that this type of tax may have an impact on human capital accumulation, for instance.

A certain element of “potential” taxation on capital income at the business level could be found in France (the ‘forfait’ system) and Israel (the ‘tachisv’ system). In France, for instance, unincorporated businesses could agree with the tax administration to be taxed for a number of years based on estimated income, following sophisticated and detailed administrative procedures, instead of on actual income (Wallace, 2002).

The Dutch presumptive capital income tax has sometimes been referred to as a tax on “potential” income (see for instance Stevens et al., 2006). It could be argued that the Dutch government holds the presumption that, over a longer time horizon, the potential return that could be earned if savings would be invested with due diligence is 4%. Individuals who own less wealth (not taking the owner-occupied house into account) may want to take less risk and will therefore earn a lower return on their savings. This implies that government should impute a “potential” return which increases with wealth. In 2017, this element was introduced (see Box 4.1).

However, there are limitations to this argument. There may be cases where asset returns do not reflect higher productivity and where recurrent net wealth taxes may therefore not support an efficient allocation of resources. Above-market returns may for instance be the result of luck or privileged market access (Kopczuk and Shrager, 2014). Favouring high returns may also discourage potentially highly profitable investments, such as investments in start-ups which are likely to generate low returns in their first few years of operation. In addition, even if there might be efficiency gains from replacing capital income taxes with wealth taxes, in practice, wealth taxes generally come on top, rather than as a replacement of, capital income taxes. Finally, from an equity point of view, as discussed earlier, favouring the holders of high-return assets implies imposing lower effective tax burdens on the wealthiest households.

Net wealth taxes could be a substitute for taxes on capital gains

A net wealth tax could also be a substitute for capital gains taxes. If capital gains are lightly taxed or not taxed at all, and if it is easy for taxpayers to avoid capital gains taxes by converting taxable personal capital income into untaxed or lightly taxed capital gains, then a net wealth tax can be seen as a substitute for capital gains taxation on the appreciation of assets.

As mentioned already, a key difference between a capital gains tax and a net wealth tax is that a wealth tax is theoretically levied on an accrual basis. Under the assumption that the wealth tax base is kept up to date through regular asset valuations, the appreciation in asset values is taxed every year under a wealth tax. In contrast, under income taxation, taxes are levied on a realisation basis, meaning that taxes are levied when assets are sold on the increase in value between the time they were purchased and sold. Accrual-based taxation has a number of advantages. First, it prevents the deferral of realised capital gains and lock-in effects, which ultimately result in an inefficient allocation of capital. Second, accrual-based taxation enhances fairness as appreciations in asset values are a better reflection of taxpayers’ current wealth. However, taxation on an accrual basis involves numerous practical difficulties, and if asset values are not regularly updated, the wealth tax becomes more comparable to a tax on a realisation basis.

An alternative could be a wealth accretion tax. Instead of taxing total net wealth, governments could consider taxing only the changes in household wealth under a capital or wealth accretion tax on an accrual basis, or mark-to-market tax as the tax is also called. According to the Schanz-Haig-Simons income concept, the annual accretion of wealth, measured in real terms, is the most ideal income tax base (OECD, 2006; Cnossen and Bovenberg (2001)). Under mark-to-market accounting, all assets would be valued at their fair market value at the end of each fiscal year and the taxpayer would be taxed on this wealth increase upon accrual. Mark-to-market accounting is a significant departure from generally applied realisation accounting and would prevent timing distortions in that taxpayers would no longer have a tax-induced incentive to realise (tax-deductible) losses and to defer the realisation of taxable capital gains.

A mark-to-market tax would allow full loss offset in the fiscal year when the loss is incurred or would allow losses to be carried forward (Toder and Viard, 2016) and to be offset against future mark-to-market tax liability. Miller (2005) has suggested introducing a mark-to-market tax only for very high-income and high-wealth individuals, trusts and companies, where the threshold would be set such that only the very wealthy and highest income-earning individuals would be affected. Cnossen and Bovenberg (2001) have suggested introducing a mark-to-market tax to tax the returns on financial products but to tax the returns on real estate under a realisation-based capital gains tax, with interest on the deferred tax to reduce lock-in effects.

Another option would be to design the net wealth tax as a minimum tax which is creditable against any current or future capital income and/ or gains tax liability. As such, the wealth tax could be used to reduce the lock-in effects generated by a realisation-based capital gains tax. Alternatively, the wealth tax which would tax the normal return to investment could be combined with a capital gains tax levied only on the infra-marginal return on savings in order to prevent double taxation of the normal return on savings.

Net wealth taxes could be used as a substitute for inheritance taxes

By reducing the amount of net wealth that can be passed on to future generations, a net wealth tax has features that are common to inheritance taxes. Both taxes are levied on net wealth, but one is levied on a taxpayer’s total net assets on a recurrent basis, while the other is only levied on wealth transfers at death. They nevertheless both reduce the amount of wealth that can be transferred by donors to recipients. Of course, to be equivalent to an inheritance tax, a recurrent net wealth tax would have to be levied at a very low rate. Indeed, even a low recurrent wealth tax liability results in a high effective tax rate when the total amount of net wealth taxes paid is expressed as a share of taxpayers’ estates. If a household owns EUR 100 000 in net wealth over time, which is taxed at 0.5% under a recurrent net wealth tax, it will have to pay EUR 500 in wealth tax every year. If the same amount of net wealth is held for 40 years, the total amount of wealth taxes that will have been paid will amount to EUR 20 000 (for simplicity, the time value of money is ignored), which is equivalent to an inheritance tax of 20% on the EUR 100 000 transferred as a bequest (EUR 20 000/EUR 100 000 or 40 x 0.5%).

Nevertheless, a net wealth tax remains an imperfect substitute for inheritance taxes. A net wealth tax is levied on all accumulated wealth including the wealth that households use to finance consumption at a later stage in their life, in particular when they retire. The effective inheritance tax burden imposed by a yearly net wealth tax increases rapidly if part of the household’s wealth is consumed during retirement. This is less true, however, if a net wealth tax is levied only on very wealthy taxpayers, whose wealth will largely be transferred to recipients as opposed to consumed during the donors’ lifetime. It should be mentioned as well that if a country has a very high level of wealth inequality, reducing wealth gaps through income and wealth transfer taxes may take time and that, under such circumstances, there may be a role for a net wealth tax to address wealth inequality.

Promoting human capital investment

Human capital is always exempt under net wealth taxes. This results from a number of considerations, including the fact that human capital is very difficult to value, that it is not directly transferrable or convertible into cash, and that there is uncertainty about the durability of its value (McDonnell, 2013). Therefore, a wealth tax lowers the net return on real and financial assets relative to the returns on investments in human capital.

Thus, wealth taxes encourage investment in human capital, which may in turn have positive effects on growth. Human capital is a critical driver of long-run economic growth. This implies that a wealth tax may be less harmful to economic growth than commonly believed as it can encourage a substitution from physical to human capital formation (Heckman, 1976 in Hansson, 2002).

The case against net wealth taxes

This section assesses the arguments that have been made against net wealth taxes. This section starts by looking at the arguments against wealth taxes on efficiency grounds; it then looks at the equity-based arguments against net wealth taxes; and it concludes with the practical limitations of net wealth taxes. The practical limitations of net wealth taxes should be distinguished from the theoretical arguments against net wealth taxes, however, given that they may, at least to some extent, be addressed through good tax design and administrative improvements.

Double taxation

Double taxation is a popular objection to net wealth taxes, but it is far from unique to wealth taxes. One of the most common objections to individual net wealth taxes is that they are unfair because they generate double (or even triple) taxation. If wealth is accumulated from wage earnings, savings or personal business income, then these flows will have in many cases already been taxed. If households accumulate wealth in order to smooth consumption over their lifetime, their wealth will be taxed again when it is used for consumption. However, multiple levels of taxation are far from unique to wealth taxes. Consumption taxes, for instance, are paid out of post-tax income.

The validity of the double taxation argument also depends on countries’ overall tax burden on capital and on the design of the wealth tax. For instance, if wealth primarily comes from asset revaluation, as Weale (2010) argues is the case in the United Kingdom, and if capital gains are not (or not adequately) taxed, then taxes on wealth do not constitute double taxation. The occurrence of double taxation also depends on how many taxpayers are subject to the net wealth tax. At lower wealth levels, as wealth is likely to be in large part accumulated for later consumption, wealth taxation indeed adds a third layer of taxation on a base that has been taxed as earned income and will be taxed as consumption. For very wealthy taxpayers, on the other hand, it is likely that part of their wealth comes from capital income, which is taxed at effective rates that vary widely and that can be low, and might not be used for later consumption, which means that double taxation is likely to be much more limited. Thus, a wealth tax that is only levied on the very wealthy might not generate much double taxation in practice. This underlines the importance of looking at wealth taxes as part of a broader tax system and of assessing how it interacts with other taxes (Brys et al., 2016).

In the case of inheritance taxes, it may be argued that the double taxation argument is weaker (Piketty et al., 2013). As with net wealth taxes, double taxation is a commonly stated objection to estate and inheritance taxes: people have already paid income tax or capital gains tax on their income before it was used to purchase assets which will be taxed again at death. However, in the case where the wealth transfer tax is levied on the recipient rather than on the donor (i.e. an inheritance tax rather than an estate tax), there is no double taxation of the donor himself and the inherited wealth is also only taxed once in the hands of the recipient. Moreover, as is the case with net wealth taxes, there might be instances where the inheritance tax will be the first time asset returns are taxed. For example, increases in the value of main residences are often exempt from capital gains tax. As a consequence, while the purchase price may well have been paid out of taxed earnings, any subsequent increases in value – which have been far greater than normal returns in recent years – will not have been subject to tax (Boadway et al., 2010).

Distortions to savings and investment

The main efficiency related argument against net wealth taxes is that – in a way that is comparable to capital income taxes – they distort saving behaviours. Standard economic models of optimal taxation assume that households save in order to consume tomorrow instead of today. Savings are therefore related to expenditure on future consumption. If the return on savings is taxed, the decision to postpone consumption and the intertemporal allocation of resources is distorted by the tax system, as the tax drives a wedge between the prices of consumption at different dates.

Two seminal models have concluded that the optimal capital income tax rate is zero. In their two-period model with one period of work, weak separability between consumption and leisure in each period, identical preferences across households and with the condition that non-linear taxes on labour earnings can be levied, Atkinson and Stiglitz (1976) show that there is no case for taxing future consumption and therefore the return to savings. The second major model concluding that capital should not be taxed was developed by Judd (1985) and Chamley (1986). They found that, in a dynamic Ramsey model, assuming infinitely-lived agents and no distortions in the economy, the long-run optimal tax on capital is zero.

However, these models rely on assumptions that are highly restrictive and that have to a large extent been empirically invalidated, with many new models concluding that positive capital taxes are optimal. Assumptions behind these models are highly stylised – including infinite time horizons, altruistic dynasties or the separability of preferences, for instance – and have often been questioned (e.g. Banks and Diamond, 2010). Many recent optimal tax theory models have refuted the optimality of zero capital taxation. For instance, Aiyagari (1995), by introducing non-trivial heterogeneity, assuming that markets are incomplete and allowing for uninsured idiosyncratic constraints, shows that there is a role for capital income taxation. Jacobs and Bovenberg (2010) show that it is optimal to tax capital income to reduce the distortions of the labour income tax on human capital investment. Straub and Werning (2014) also refute the optimality of a zero long-run tax on capital by revisiting and using the very logic of the models developed by Chamley and Judd.

As discussed above, wealth taxes do not operate exactly like capital income taxes and their effects on savings and investment will differ. While it may seem irrelevant for a taxpayer who has a total net wealth of EUR 10 million that earns a rate of return of 4% whether the government levies a tax of 30% on the capital income or a wealth tax of 1.2% on the capital stock – as both raise EUR 120 000 – there is an important distinction between the two. As discussed above, a tax on the stock of wealth is equivalent to taxing a presumptive return but exempting returns above that presumptive return. Where the presumptive return is set at the level of or at a level close to the normal - or risk-free – return to savings, a wealth tax is economically equivalent to a tax on the normal return to savings, which is considered to be inefficient. Indeed, the taxation of normal returns is likely to distort the timing of consumption and ultimately the decision to save, as the normal return is what compensates for delays in consumption (Mirrlees et al., 2011). As discussed below, it is also unfair that the wealth tax liability does not vary with returns, which implies that the effective wealth tax burden decreases when returns increase.

The potentially large distortive effect of wealth taxes on savings also comes from the fact that when wealth taxes are levied, they are often imposed on top of capital income taxes. If imposed on top of high income taxes, a net wealth tax can significantly increase marginal effective tax rates (METRs), in particular at higher inflation rates and lower real rates of return (see below), and in such circumstances not only discourage saving but potentially encourage dis-saving (i.e. consumption out of capital) (Messere et al., 2003). Evidence from a new OECD study (OECD, 2018) confirms that METRs can reach high levels in the presence of net wealth taxes (for details on the methodology, see Box 3.1). Figure 3.1 shows ETRs with and without net wealth taxes on different types of assets for taxpayers subject to the top PIT and net wealth tax rates in the countries that had net wealth taxes in 2017. The results show that net wealth taxes significantly raise the tax burden on capital income. In both France and Spain, METRs reached values above 100%, which means that the entire real return is taxed away and that by saving people actually reduce the real value of their wealth.

Box 3.2. Methodology for the calculation of METRs on savings

The marginal effective tax rates (METRs) presented in this report are extracted from a new OECD study (OECD, 2018) which estimates METRs across a range of savings vehicles for 38 OECD and key partner countries to assess the effect of tax systems on the incentives individuals face to save in different forms.

The METR methodology in this OECD study follows broadly the approach of the OECD’s 1994 Taxation and Household Savings study (OECD, 1994), which itself drew on the methods used by King and Fullerton (1984). As emphasised in the OECD (1994) study, the appropriate way to analyse the effect of taxation on savings decisions is to examine the incentives faced by the taxpayer at the margin. The analysis therefore focuses on a saver who is contemplating investing an additional currency unit in one of a range of potential savings vehicles. The investment is a marginal investment, both in terms of being an incremental purchase of the asset, and in terms of generating a net return just sufficient to make the investment worthwhile (as compared to the next best savings opportunity).

The approach assumes a fixed pre-tax real rate of return and calculates the minimum post-tax real rate of return that will for that asset, at the margin, make the investment worthwhile. The METR can then be calculated as the difference between the pre- and post-tax rates of return (the savings income tax wedge) divided by the pre-tax rate of return. The calculations take into account different assumptions for the real rate of return, the inflation rate and the expected holding periods.

Source: OECD (2018)

Figure 3.1. Marginal effective tax rates (METRs) with and without wealth taxes on different assets
picture

Note: The METR results are based on tax rules as of 1 July 2016.

Source: Data from OECD Taxation of Household Savings (OECD, 2018)

However, the cumulative nature of a net wealth tax and its potentially distortive effects depend on the rest of the tax system and interactions with other taxes on capital. Some of the countries that levy net wealth taxes do not impose taxes on the transfer of capital or on capital gains. For instance, in Switzerland, there is no capital gains tax; in Norway, the inheritance tax was repealed; and in the Dutch system, the tax on assumed income from savings and investments replaces the taxation of the actual income flows from these assets (Lawless and Lynch, 2016). In France1 and Spain, on the other hand, in addition to net wealth taxes, the government levies taxes on capital transfers and taxes on capital gains. Figure 3.1 shows that, in contrast to Norway and Switzerland, the combination of personal capital income taxes and net wealth taxes results in very high METRs in France and Spain.

Beyond effects on the overall level of savings, net wealth taxes are also likely to affect the composition of savings. Adverse effects on savings and investment may also come from the distortions in the choice between different types of savings vehicles that wealth taxes generate through exemptions and reliefs. It can be assumed that if the tax base is narrow, there will be strong effects on the composition of savings, while if the tax base is broad, the wealth tax might have stronger effects on the overall level of savings. In practice, many categories of assets are exempt under net wealth taxes or benefit from reliefs or preferential valuation, which provides incentives to alter portfolio allocation away from that which would be optimal in a no-tax world (e.g. investing in assets with the lowest tax liability or those where valuation is most difficult). In practice, METR results reveal significant variations across asset types. Some assets tend to be particularly tax-favoured compared to others. Figure 3.1 shows in particular that pension savings are not taxed under net wealth taxes in any of the countries that currently have net wealth taxes.

The provision of exemptions and reliefs will have particularly strong distortive effects if, as has often been the case, they tend to favour non-productive assets (e.g. housing) over more productive asset types (McDonnell, 2013). It may be argued that if tax-induced distortions favour more productive investments, a wealth tax could have positive growth effects. Such a tax would lower incentives to continue investing in unproductive assets and encourage a shift towards more productive investments. In fact, the idea of encouraging productive investment has sometimes been used to support a wealth tax that would be levied exclusively on high value immovable property as opposed to overall net assets. Nevertheless, drawing a clear distinction between productive and unproductive assets is challenging. Besides, in the case of a wealth tax levied exclusively on high-value immovable property, equity might be reduced as financial capital, which is mostly owned by the very rich, would be exempt.

A few empirical studies have looked at whether the taxation of wealth actually deters savings, often pointing to the limited effects of net wealth taxes on real behaviour (e.g. wealth accumulation, labour supply) and to their stronger effect on wealth reporting. Zoutman (2015) estimates the elasticity of taxable savings using the 2001 Dutch capital income and wealth tax reform and using a difference-in-difference approach comparing households that are similar in terms of income and wealth but that were treated differently by the tax reform. He finds that an increase in the capital income and/or wealth tax leads to a relatively small loss in accumulated wealth: depending on the specification and the sample, a 1% increase in the current Dutch wealth tax of 1.2%, leads to a reduction in household savings between 0.10-0.17%. Using tax record data from Sweden, Seim (2017) estimates that net-of-tax-rate elasticities of taxable wealth were comprised between 0.09 and 0.27. His analysis also finds that these small but positive elasticities appear to reflect tax avoidance and evasion. Brülhart et al. (2017), on the other hand, find a stronger sensitivity of wealth holdings to wealth taxation in their analysis of wealth taxes in Switzerland. A 0.1 percentage-point increase in wealth taxes leads to 3.4% lower wealth holdings in their cross-canton data, substantially exceeding standard estimates of the elasticity of taxable income. However, their study also seems to suggest stronger effects on wealth reporting than on real behaviour with their results showing that taxpayers bunch below the tax threshold, that observed responses are driven by changes in wealth holdings rather than mobility, and that financial wealth is somehow more responsive than non-financial wealth.

Unfortunately, these studies cannot be used to draw conclusions on what the real effects of wealth taxes would be if tax avoidance and evasion opportunities were severely restricted. These studies generally suggest that real responses to wealth taxes are not significant in the presence of tax avoidance and evasion opportunities but the effects of wealth taxes on real behaviours would likely be much stronger if avoidance and evasion opportunities were severely restricted. These findings are consistent with the hierarchy of behavioural responses to taxation in Slemrod (1992), with decisions regarding the timing of transactions being the most responsive to tax changes, followed by avoidance, and real behaviours exhibiting the lowest degree of responsiveness. Taxpayers tend to respond in real terms (e.g. savings, labour) as a last resort, when avoidance and evasion opportunities are not available.

Regarding taxes on wealth transfers, their effects on savings will largely depend on bequest motives (e.g. accidental, strategic, altruistic or ‘joy of giving’). For instance, unintentional transfers (e.g. when assets are accumulated or held for the owner’s personal use for retirement or risk prevention) by definition have no impact on behaviour and can therefore be taxed more heavily without generating an efficiency loss. However, given that not all bequests are accidental and that they are often at least partly planned by donors, taxes on inherited wealth will always affect donors’ saving and consumption decisions to some extent (Mirrlees et al., 2011). That being said, determining the intent of a donor is extremely difficult and there is usually more than one motive (Kopczuk, 2012), so for administrative reasons countries do not distinguish between bequest motives, leading to all wealth transfers being taxed together. In addition, inheritance taxes are expected to have an impact on recipients’ saving and labour decisions: by reducing post-tax inheritances, they may give recipients stronger incentives to work and save themselves. Empirically, the evidence on the effects of wealth transfer taxes is mixed, but generally suggests a negative – but small – overall effect on donors’ savings (Mirrlees et al., 2011).

Negative effects on entrepreneurship and risk-taking

Another efficiency related argument is that a net wealth tax reduces the amount of capital available, which may in turn affect entrepreneurship and business creation as access to capital is an important determinant of an individual’s propensity to start a business (Hansson, 2010). Negative effects on entrepreneurship are even greater if business assets are (partly) taxed under the net wealth tax.

A key difference between income and wealth taxes is the treatment of losses, however, which has implications for risk-taking and entrepreneurship. Taxation is often believed to discourage risk-taking by capturing part of the return to risky investments. A competing view, however, is that, in cases of risk-averse investors and (perfect) loss offset, the taxation of income may in fact encourage risk-taking by absorbing a portion of the risk associated with risky investments (Domar and Musgrave, 1944). Under a net wealth tax, however, if income is zero or negative, the tax liability will still be positive if the capital value of the assets remains positive. In practice, new entrepreneurs which tend to generate low, or even negative, profits in their first few years of operation would still face a wealth tax liability. Thus, a heavy net wealth tax which is unlinked to income might discourage entrepreneurship relative to an income tax with (perfect) loss offset.

However, there are ways in which a wealth tax replacing an income tax may actually stimulate risk-taking and entrepreneurship. Taxpayers starting a business are likely to do so because in the longer run they expect to earn high returns compensating for the risks they have taken. As discussed, a tax on the stock of wealth is equivalent to taxing a presumptive return but exempting returns above that presumptive return, including returns to risk. In that sense, a wealth tax may stimulate entrepreneurship and risk-taking (in particular if the net wealth tax replaces taxes on capital income). Overall, it is therefore difficult to firmly argue that wealth taxes would have negative effects on entrepreneurship. The magnitude of the effects of wealth taxes on entrepreneurship is also unclear as business assets are often excluded from the wealth tax base in practice.

How assets are valued in the wealth tax base is also important when it comes to assessing how a wealth tax affects risk-taking. If assets are taxed on book values, the wealth tax liability will not vary with the business cycle and therefore the effective tax rate will be low in good times (i.e. low book/market value ratio) and high in bad times. On the contrary, if the wealth tax is based on market values and assuming that market values are regularly updated, then the effective tax rate will be constant over the business cycle. So a wealth tax based on market values (e.g. for listed firms) will be more neutral with respect to investors taking risks than a wealth tax based on book values (e.g. for non-listed firms) which increases the required return of risk-averse investors. On the other hand, book values may be significantly lower than market values (as is the case in Norway, for instance) and therefore a wealth tax may give an incentive to invest in non-listed firms relative to listed firms. This highlights the importance of setting the right tax base and the possible consequences of not doing so.

Inheritance taxes may also have a detrimental impact on entrepreneurship, particularly on family-owned businesses. Estate or inheritance taxes may discourage entrepreneurship by reducing the post-tax value of wealth transfers. Indeed, entrepreneurs might be less likely to start a business if they anticipate that a large portion of their business will be taxed when it is transferred to their heirs. High inheritance taxes may also make it difficult for family businesses to survive the death of their founders. To address these concerns, countries often provide inheritance tax relief for close-held/family businesses. However, the corollary of special tax treatments and reliefs is increased risks of tax planning and avoidance.

Liquidity concerns

Liquidity issues are a major equity concern regarding wealth taxes. The relationship between income and wealth is imperfect (see Chapter 2), which means that some households – the so-called “wealthy hand-to-mouth” (Kaplan et al., 2014) – have valuable assets which make them liable to the wealth tax but limited realised income with which to pay the tax. A substantial wealth tax bill combined with a low current income may result in assets needing to be sold in order to pay the tax, although the magnitude of the liquidity issue depends on how liquid assets are and on the level of the wealth tax. As will be discussed in greater detail in Chapter 4, liquidity issues can be mitigated through different measures, including by allowing tax liabilities to be spread over time or to be accumulated until assets are sold or inherited.

Liquidity issues also arise in the case of estate or inheritance taxes. A case where liquidity might be a real concern is for family-owned businesses where recipients are forced to sell parts or all of the business to pay the tax but, as mentioned above, many countries have special tax provisions lowering the tax liability in the case of family-business wealth transfers. However, liquidity constraints are less of a concern when inherited property has multiple recipients and has to be sold anyway to divide the value amongst the recipients (Boadway et al., 2010).

Penalisation of low-return and less diversified assets

From an equity perspective, a net wealth tax penalises the holders of low-return assets. As discussed already, because net wealth taxes do not tax the actual return earned on assets but are equivalent to the taxation of a presumptive return, the effective tax rate decreases when actual returns increase. Using the example of France in 2016, Figure 3.2 shows ETR results under different return scenarios and confirms that the wealth tax burden is heavier in a low-return setting. This may have negative equity effects. Indeed, there is evidence of heterogeneous returns that are positively correlated with wealth (Fagereng et al., 2016), which may be explained by the fact that wealthy investors tend to be less risk averse and allocate a much larger share of their financial portfolios to risky assets (Bach et al., 2015), have better access to financial expertise (economies of scale in wealth management) or might have better financial education and access to more lucrative investment opportunities. This means that if the wealth tax applies to (part of) the middle class, it might have regressive effects. For instance, taxpayers with a large portion of their assets in regular savings accounts, for which the rate of return is close to zero, are taxed for a return they generally did not realise, while wealthier taxpayers who have invested a lot of their savings in shares tend to accrue higher gains than they are taxed for. To avoid these negative equity effects, this would imply that a country that has a wealth tax should exempt some amount of bank deposits or have a high overall exemption threshold to ensure that only the wealthy are subject to the tax and/or levy the wealth tax at progressive rates.

Figure 3.2. METRs in France under different real return scenarios (2%, 3% and 4%) in 2016
METRs for taxpayers at 500% of the average wage and for top income and wealth taxpayers
picture

Note: The METR results are based on tax rules as of 1 July 2016.

Source: Data from OECD Taxation of Household Savings (OECD, 2018)

In addition, taxing a presumptive return below the actual return earned on savings provides greater tax savings to more wealthy households. Although wealth tax liabilities will increase with the amount of household wealth, the tax savings as a result of the taxation of a presumptive return below the actual return on savings will increase with the amount of wealth. One might therefore argue that, in this case, replacing personal capital income taxes with a net wealth tax will have regressive effects on the amount of tax savings as it provides higher tax savings to wealthier households. The opposite result holds if the presumptive return exceeds the actual return on savings.

Net wealth taxes may also be perceived as unfair given that the assets held predominantly by the wealthiest taxpayers often benefit from a more favourable tax treatment. Equity issues will arise with narrow tax bases, in particular if broadly-held assets are taxed but more mobile assets held primarily by the wealthiest taxpayers are exempt or taxed favourably. For instance, taxing favourably financial and business wealth – which is mostly held by households at the top of the income distribution (see discussion in Chapter 2) – while imposing a high tax burden on residential property may have detrimental equity effects as most of the middle class’s wealth is tied up in immovable property. This is another strong argument for having a high exemption threshold to avoid possible negative equity effects.

Horizontal inequity over the lifecycle

A wealth tax might generate horizontal inequities over the lifecycle. If the net tax base is broad and low levels of wealth are taxed (i.e. the tax exemption threshold is low), households who save a lot when they are young to consume when they are older will pay more tax than households who spread their income more equally over their lifecycle. Chapter 2 shows a dynamic pattern in savings and wealth accumulation over the lifecycle and that on average households accumulate wealth throughout their working years, with wealth peaking in the years just prior to retirement, before declining.

Capital flight and fiscal expatriation

Capital flight has been a key argument against wealth taxes. In theory, the capital flight argument only applies in the case of non-residents because they are taxed on the assets they own within the taxing jurisdiction (source-based taxation) which will affect the international allocation of capital, but it does not apply in the case of residents as they are taxed on their worldwide assets, which should not affect the international allocation of capital. However, the possibility of holding assets abroad and not declaring them as well as the difficulty of valuing offshore assets, in particular for non-listed shares and other non-frequently traded assets, means that capital flight is in practice a significant concern for residents as well.

In addition, because wealth taxes are residence-based for residents, there is a risk that wealthy individuals can relocate to avoid the tax (i.e. fiscal expatriation). Indeed, a high wealth tax burden may encourage taxpayers to change their tax residence to a lower tax jurisdiction to reduce their tax burden. Risks of fiscal expatriation are likely to be more prevalent in countries whose neighbouring jurisdictions offer more favourable tax conditions. Regarding the potential effects of fiscal expatriation, on top of the immediate revenue losses, it might lead to a reduction in investment. However, whether fiscal expatriation has significant economic consequences on taxpayers’ country of origin remains a question and will depend on whether fiscal expatriates maintain activities in their country of origin.

Empirical studies on the effects of wealth taxes on capital flight and fiscal expatriation are very limited. Pichet (2007) found evidence of significant capital flight out of France since the introduction of the net wealth tax. Zucman (2008), on the other hand, finds that wealth tax evasion in France is limited compared with the revenue generated by the tax. In the case of Switzerland, Brülhart et al. (2017) find evidence of low wealth tax-induced mobility. As mentioned above, tax-induced incentives for individuals to change their residency will depend on a variety of factors including the effective tax rate differentials with other countries or regions.

The evidence collected by governments on these issues tends to be anecdotal and difficult to interpret. In France, the Ministry of Finance has tracked the number of taxpayers subject to the wealth tax who leave and return to France. In 2014, 780 taxpayers subject to the net wealth tax left France, while 300 returns were registered. However, it is difficult to determine the extent to which decisions to move are motivated by tax factors or other personal or professional reasons. In the case of France, the significant increase in the number of net wealth taxpayers leaving France coincided with tax changes which generally lowered tax burdens on the very wealthy (decrease in top marginal PIT rates and introduction of the net wealth tax cap), which may suggest that taxpayers’ departures were not primarily driven by tax considerations (Conseil des prélevements obligatoires, 2011). These studies also fail to capture the taxpayers who move abroad in anticipation of future wealth tax burdens, before they become liable to the wealth tax. Finally, as mentioned already, the economic repercussions of such fiscal exile are uncertain as taxpayers changing their residency for tax reasons can still continue to invest in their home country.

Taxpayers’ locational decisions, in particular for the elderly, may also be affected by estate or inheritance taxes, although there is no clear empirical evidence of that. Bakija and Slemrod (2004) find that, in the United States, state estate taxes have a statistically significant negative effect on the number of federal estate tax returns filed in a state. This evidence seems consistent with the idea that some rich individuals flee states that tax them relatively heavily, although it may reflect other forms of tax avoidance as well. However, a number of other studies have found very limited effects of inheritance or estate taxes on migration patterns. For instance, Conway and Rork (2006), find no statistical evidence that bequest taxes affect inter-state migration patterns of elderly taxpayers in the United States. Looking at a much smaller country which is characterised by a greater degree of heterogeneity in sub-national bequest taxation, Brülhart et al., (2014) also find that cuts in bequest tax burdens across Swiss cantons have had little noticeable impact on the migration patterns of elderly taxpayers.

Tax avoidance and evasion

Increasing capital mobility has also enabled tax avoidance and evasion. The increasing mobility of financial assets as well as the use of tax havens, combined with the development of information and communication technology and the elimination of barriers to cross-border capital transfers (such as capital controls), have allowed taxpayers to move their capital offshore without declaring it and made the enforcement of capital income taxes and wealth taxes much more difficult (Krenek and Schratzenstaller, 2017). In fact, capital mobility has been a major factor behind the reduction of taxes on capital in the last few decades. The mobility of capital also has significant implications on the incidence of wealth taxes, as wealth taxes will likely end up bearing more heavily on less mobile forms of wealth, i.e. immovable property.

However, the recent progress made on international tax transparency and the exchange of information is increasing countries’ capabilities to tax capital effectively. Information exchange agreements as well as further international cooperation on the exchange of information on request (EOIR), the automatic exchange of information (AEOI) and areas like beneficial ownership will reduce opportunities for tax evasion and ultimately allow countries to tax both capital and capital income more effectively. However, such efforts need to take into account that high-wealth individuals can change their tax residency and even their citizenship in response to high taxes, and that by limiting opportunities for tax avoidance and evasion, the real effects of taxes on capital – in particular on savings and investment – may be stronger (see above).

Domestic tax avoidance and evasion is also possible under a net wealth tax as there are a number of asset classes that are highly susceptible to non-disclosure or underreporting. As discussed below, some forms of wealth are difficult to value or can easily be hidden from tax authorities and the capacity of tax authorities to check non-disclosure and underreporting is often limited. Typical examples include household goods, vehicles, jewellery, artwork, etc. Relying on self-reporting also makes non-disclosure or underreporting more likely. This differs from withholding at source and third party reporting which are well-developed for many forms of capital income taxation such as dividends and interest; although in theory the same tools could be put in place for the taxation of capital stocks (Keen, 2014).

In addition, avoidance strategies are encouraged by the many exemptions and reliefs that are provided under net wealth taxes. As discussed in Chapter 4, there are different justifications for keeping certain assets out of the tax base or for taxing certain assets preferentially but tax exemptions and reliefs, in addition to narrowing the tax base, open up tax planning and avoidance opportunities. An important question, for instance, is whether the assets that taxpayers accumulate in their corporations should be taxed. This would be highly distortive but, if such assets are not taxed, wealthy taxpayers may avoid taxes by setting up corporations to accumulate their wealth tax-free. As mentioned already, from an equity perspective, tax planning opportunities through tax reliefs or exemptions are predominantly available to the wealthiest taxpayers who have diversified asset holdings and possibilities to restructure their assets and income.

Debt deductibility under net wealth taxes provides incentives to borrow and can encourage tax avoidance. While from an equity perspective it makes sense to tax net wealth, as net wealth is a better reflection of taxpayers’ ability to pay, it implies that individuals will have an incentive to keep on borrowing funds for investment purposes as long as the return on the investment exceeds the interest that has to be paid. If the wealth tax base is narrow, taxpayers will have an incentive to avoid the tax by borrowing and investing in exempt assets or – if debt is only deductible when incurred to acquire taxable assets – taxpayers will have an incentive to invest part of their savings in tax-exempt assets and finance their savings in taxable assets through debt. In addition to opening up opportunities for tax avoidance, debt deductibility may raise financial stability concerns, by encouraging highly leveraged investments and the accumulation of high household debts, especially in economic downturns.

Empirical studies show clear evidence of wealth tax avoidance and evasion. As mentioned above, empirical studies on behavioural responses to wealth taxes tend to show that taxpayers respond more through tax avoidance and evasion than through changes in real behaviour. As mentioned already, in his analysis of the behavioural response to the wealth tax in Sweden, Seim (2017) estimates that the net-of-tax-rate elasticities of taxable wealth were between 0.09 and 0.27 and finds that these small but positive elasticities reflect tax evasion and avoidance rather than changes in savings. Using a panel of tax return micro-data from Catalan taxpayers for the 2011-14 period, Durán-Cabré et al. (2017) examine taxpayers’ responses to the re-introduction of the Spanish net wealth tax in 2011. Their results provide empirical evidence that taxpayers facing higher tax rates reorganise the composition of their wealth in order to benefit from exemptions, also suggesting tax avoidance rather than real behaviour. More generally, Zucman (2015) finds evidence of a considerable and increasing amount of private wealth being hidden and managed in tax havens. His estimation is that around 8% of the global financial wealth of households is held in tax havens, three-quarters of which go unrecorded (Zucman, 2015).

Tax avoidance and evasion are also common with taxes on wealth transfers. As with net wealth taxes, tax avoidance is facilitated by the existence of estate or inheritance tax reliefs. Another tax avoidance strategy is the possibility to transfer wealth through lifetime gifts. That is why an inheritance tax needs to be complemented with a gift tax. A third major issue is related to trusts which, although they can be set up for perfectly legitimate reasons, can also potentially be used to avoid inheritance tax since they confer the benefits of wealth without transferring property ownership. Indeed, they are often used to separate the entitlement to the income that property generates from the entitlement to the property itself (Mirrlees et al., 2011). Thus, rules are also needed to prevent inheritance tax avoidance through the use of trusts. Addressing these issues is all the more critical as inheritance or estate taxes are levied at death and therefore leave significant time for tax planning.

Valuation and other administrative issues

In addition to the difficulties associated with tracing back wealth ownership, many forms of wealth are difficult to value. Valuation is difficult in the case of non- or infrequently traded assets (e.g. personal and household effects, pension rights, etc.). Partly as a consequence of valuation issues, many of these hard-to-value assets have been exempted from wealth taxes, eroding the tax base, distorting the choice of savings vehicles and creating opportunities for tax avoidance. Valuation issues are also significant in relation to non-listed firms and closely-held companies and are even greater for assets held overseas. Overall, it is much easier to determine the value of income flows than the value of capital stocks (Brown, 1991). As discussed in Chapter 4, however, there are some practical ways to address certain valuation issues – for instance, insured values can be used in the case of high-value jewellery or artwork, and exemptions can be granted for household effects under a certain threshold value (McDonnell, 2013).

Regularly updating asset values is an additional difficulty. Indeed, there is a trade-off between regularly updating asset values, which is costly both in terms of tax compliance and administration, and updating them less frequently, which may increase distortions and reduce fairness. There are, however, some ways to minimise the administrative and compliance burden associated with regularly updating asset values. For instance, asset valuations used for the residential property tax or the inheritance tax can be used for net wealth tax purposes as well. In addition, the value of taxpayers’ total net wealth – or alternatively the value of particular asset classes – can be treated as fixed for a few years before being re-assessed (McDonnell, 2013).

The date of valuation can also raise issues. If assets are valued on 1 January, then the net wealth tax is partly levied on wealth that will be consumed later in the year. This distorts the timing of consumption decisions as taxpayers will have an incentive to bring their consumption forward to the end of the previous year. This argument is less convincing, however, if the net wealth tax base is broad. A lot of consumption occurring at the top of the wealth distribution is likely to consist in buying assets that would be taxed under a broad-based net wealth tax (e.g. cars, jewellery, artwork). On the other hand, if assets are valued at the end of the year, taxpayers may be taxed on wealth that they have accumulated during the year (i.e. savings) which implies that savings would be taxed twice in the same year.

In comparison, the taxation of wealth transfers at death tends to be less administratively costly. As opposed to net wealth taxes which require regularly updating asset values, valuing assets under an estate or inheritance tax, which involves determining their market value (or their realistic selling price) only occurs once, at the time of the transfer of assets between donors and recipients. Nevertheless, there are still some important valuation and administrative issues involved, including for instance complexities in relation to jointly held assets or due to the presence of two parties with different jurisdictional affiliations (Iara, 2015).

Summary and policy implications

This chapter has reviewed the arguments that have been made both for and against net wealth taxes. Those arguments are related to efficiency, equity and tax administration considerations. Many of the efficiency and equity effects of net wealth taxes are linked to the fact that they are levied regardless of actual returns and function like a tax on a presumptive (i.e. fixed) return. This chapter also discussed the effects of capital income taxes and taxes on wealth transfers to examine how these taxes interact with net wealth taxes and whether they can be complements to or substitutes for taxes on net wealth.

Overall, the chapter suggests that broad-based capital income taxation – including the taxation of capital gains – combined with well-designed inheritance taxes may be a more efficient and less administratively costly way of addressing wealth inequality. Overall, this chapter finds that, from both an efficiency and an equity perspective, there are limited arguments for having a net wealth tax on top of well-designed capital income taxes – including taxes on capital gains – and inheritance taxes, but that there are arguments for having a net wealth tax as an (imperfect) substitute for these taxes.

While there are important similarities between personal capital income taxes and net wealth taxes, the report shows that net wealth taxes tend to be more distortive and less equitable. As discussed in the chapter, a tax on the stock of wealth is equivalent to taxing a presumptive return to assets but exempting returns above that presumptive return. Where the presumptive return is set at the level of or at a level close to the normal - or risk-free – return to savings, a wealth tax is economically equivalent to a tax on the normal return to savings, which is considered to be inefficient. Indeed, the taxation of normal returns is likely to distort the timing of consumption and ultimately the decision to save, as the normal return is what compensates for delays in consumption (Mirrlees et al., 2011). This equivalence with a tax on a presumptive return also raises equity concerns as a wealth tax will favour the holders of high-return assets which tend to be wealthier taxpayers. There are also a number of practical difficulties associated with net wealth taxes, including valuation and liquidity issues. Finally, wealth taxes are generally imposed on top of capital income taxes, which can result in very high METRs on capital income. One of the areas where a wealth tax has advantages over capital income taxation, however, is that a net wealth tax is in theory levied on an accrual basis, which avoids the lock-in effects of taxation on a realisation basis, although this issue could possibly be addressed by taxing capital gains on accrual.

To strengthen progressivity, the way countries tax personal capital income could be revisited. In particular, progressive tax rates could be applied to personal capital income. As argued in previous OECD work (Brys et al., 2016), countries could consider introducing “dual progressive income tax” systems which would tax capital income under a separate rate schedule at progressive rates. The rate schedule could exempt or tax at low rates total household capital income below a minimum threshold. This could also encourage taxpayers at the bottom of the income and wealth distribution to save more, which could ultimately contribute to reducing wealth inequality. Finally, as mentioned above, consideration could be given to taxing capital gains upon accrual, noting the practical difficulties of doing so.

Inheritance taxes are also central to addressing the persistence of wealth gaps and tend to be less distortive than net wealth taxes. The report argues that capital income taxes alone will most likely not be enough to address wealth inequality and suggests the need to complement capital income taxes with a form of wealth taxation. The report finds that there is a strong case for an accompanying inheritance tax. The double taxation argument is weaker in the case of inheritance taxes than for net wealth taxes, as there is no double taxation of the donor and the inherited wealth is also only taxed once in the hands of the recipient. Effects on savings are also likely to be smaller than in the case of recurrent taxes on personal net wealth, and have generally been found empirically to be negative but small. Inheritance taxes are also easier to administer and comply with as they are only levied once. Finally, and perhaps more importantly, there are convincing meritocratic arguments for taxing inherited wealth more than self-made wealth. However, further work is needed to determine how to design inheritance taxes in a way that makes them both more efficient and fairer.

Overall, the report suggests that the merits of a net wealth tax cannot be assessed in isolation but depend on countries’ overall tax systems and broader economic and social circumstances. Previous OECD work has already highlighted the need to look at tax systems as a whole and in the context of countries’ economic and social circumstances (Brys et al., 2016). For instance, a net wealth tax may have more limited distortive effects and be more justified as a way to enhance progressivity in countries where the taxation of personal capital income is comparatively low. In practice, this implies that in countries with dual income tax systems that tax capital income at low and flat rates or in countries where capital gains are not taxed (e.g. Switzerland), there is a stronger justification for levying a net wealth tax. A similar argument can be made for countries that do not levy taxes on inheritances (e.g. Norway). Beyond tax considerations, there might also be greater justification for a wealth tax in a country exhibiting high levels of wealth inequality as a way to narrow wealth gaps at a faster pace.

References

Alvaredo, F. and E. Saez (2009): “Income and Wealth Concentration in Spain from a Historical and Fiscal Perspective”, Journal of the European Economic Association, Vol. 7, No. 5, pp. 1140-67.

Aiyagari, S. R. (1995), “Optimal Capital Income Taxation with Incomplete Markets, Borrowing Constraints, and Constant Discounting”, Journal of Political Economy, Vol. 103, No. 6, pp. 1158–75.

Bach, L., Calvet L. E. and P. Sodini (2017), “Rich pickings? Risk, return, and skill in the portfolios of the wealthy”, CEPR Discussion Paper, No. 11734.

Bakija, J. and J. Slemrod (2004), “Do the rich flee from high state taxes? Evidence from Federal Estate Tax Returns”, NBER Working Paper, No. 10645.

Banks, J. and P. A. Diamond (2010), “The Base for Direct Taxation” in The Mirrlees Review. Dimensions of Tax Design, edited by James A. Mirrlees et al., Oxford University Press, 548–648.

Bastani, S. and D. Waldenström (forthcoming), “What drives preferences for wealth redistribution? Evidence from linked survey and register data”, preliminary version.

Boadway, R., Chamberlain E. and C. Emmerson (2010): “Taxation of Wealth and Wealth Transfers”, in J. A. Mirrlees, S. Adam, T. Besley, R. Blundell, S. Bond, R. Chote, M. Gammie, P. Johnson, G. Myles and J. Poterba, eds., Dimensions of Tax Design: the Mirrlees Review, 8:737-836, Oxford University Press.

Brown, R. D. (1991), “A Primer on the Implementation of Wealth Taxes”, Canadian Public Policy / Analyse de Politiques, Vol. 17, No. 3, pp. 335-350.

Brülhart, M., Gruber, J., Krapf, M. and K. Schmidheiny (2017), “Taxing Wealth: Evidence from Switzerland”, NBER Working Paper Series, No. 22376.

Brys, B., Perret S., Thomas, A. and P. O’Reilly (2016), “Tax Design for Inclusive Economic Growth”, OECD Taxation Working Papers, No. 26, OECD Publishing, Paris.

Chamley, C. (1986), “Optimal Taxation of Capital Income in General Equilibrium with Infinite Lives”, Econometrica, Vol. 54, No. 3, pp. 607–622.

Cnossen, S. and L. Bovenberg (2001), “Fundamental tax reform in the Netherlands”, International Tax and Public Finance, No.7, pp. 471-484.

Conseil des prélèvements obligatoires (2011), “Prélèvements obligatoires sur les ménages : progressivité et effects redistributifs”, http://www.ladocumentationfrancaise.fr/rapports-publics/114000255/index.shtml

Conway, K. S. and J. C. Rork (2006), “State “death” taxes and elderly migration: the chicken or the egg?”, National Tax Journal, Vol. 59, No. 1, pp. 97–128.

Domar, E.D., and R.A. Musgrave (1944), “Proportional income taxation and risk-taking”, Quarterly Journal of Economics, Vol. 58, pp. 388-422

Durán-Cabré, J.M., Esteller-Moré, A. and M. Mas-Montserrat (2017), “Behavioural Responses to the (Re)Introduction of Wealth Taxes: Evidence from Spain”, preliminary version.

Elinder, M., Erixson, O. and D. Waldenström (2015), “Inheritance and wealth inequality Evidence from population registers”, Uppsala Center for Fiscal Studies Department of Economics Working Paper No. 2015:3.

Fagereng, A., Guiso, L., Malacrino, D. and L. Pistaferri (2016), “Heterogeneity and Persistence in Returns to Wealth”, NBER Working Paper Series, No. 22822.

Faulk, D., Martinez-Vazquez, J. and S. Wallace (2006), “Taxing potential income: a second look at presumptive taxes”, Georgia State University, Andrew Young School of Policy Studies, paper presented at the Annual Conference on Public Finance Issues.

Guvenen F., Kambourov, G., Kuruscu, B., Ocampo-Diaz, S., and D., Chenk, “Use It or Lose It: Efficiency Gains from Wealth Taxation”, preliminary version.

Hood, A. and R. Joyce (2017), “Inheritances and Inequality across and within Generations”, IFS Briefing Note BN192, The Institute for Fiscal Studies, London.

Iara, A. (2015), “Wealth Distribution and Taxation in EU Members”, Taxation Papers, Working Paper No. 60, Brussels.

Jacobs, B. and L. Bovenberg (2010), “Human Capital and Optimal Positive Taxation of Capital Income”, International Tax and Public Finance Vol. 17, No. 5, pp. 451-478.

Judd, K. L. (1985), “Redistributive Taxation in a Simple Perfect Foresight Model”, Journal of Public Economics, Vol. 28, No. 1, pp. 59–83.

Hansson, A. (2010), “Is the Wealth Tax Harmful to Economic Growth?”, World Tax Journal, 2010, Vol.10, No.1.

IMF (2014), “Fiscal Policy and Income Inequality”, IMF Policy Paper, Washington, D.C.

Kaplan, G., Violante, G. L. and J. Weidner (2014), “The Wealthy Hand-to-Mouth”, Brookings Papers on Economic Activity, Vol. 2014, No. 1, pp. 77-138.

Keen, M. (2014), “Taxing Wealth: Policy Challenges and Recent Debates”, presentation for the ECFIN Taxation Workshop - Taxing Wealth: Past, Present, Future, 13 November 2014.

Kopczuk, W. (2012), “Taxation of Intergenerational Transfers and Wealth”, NBER Working Paper Series, No. 18584.

Kopczuk, W. and A. Schrager (2014), “The Inequality Illusion: Why a Wealth Tax Won't Work”, Snapshot, Foreign Affairs, https://www.foreignaffairs.com/articles/united-states/2014-05-15/inequality-illusion

Krenek, A. and M. Schratzenstaller (2017), “Sustainability-oriented Future EU Funding: A European Net Wealth Tax”, FairTax WP-Series No.10.

Lawless, M. and D. Lynch (2016), “Scenarios and Distributional Implications of a Household Wealth Tax in Ireland”, ESRI Working Paper, No.549.

McDonnell, T.A. (2013), “Wealth Tax: Options for its Implementation in the Republic of Ireland”, NERI Working Paper Series, WP 2013/6.

Messere, K., De Kam, F. and C. Heady (2003), Tax Policy: Theory and Practice in OECD Countries, Oxford University Press.

Miller, D. S. (2005), “A progressive system of mark-to-market taxation”, Tax Notes, November 21.Mirrlees, J., S. Adam, T. Besley, R. Blundell, S. Bond, R. Chote, M. Gammie, P. Johnson, G. Myles and J. Poterba (2011), Tax by Design: The Mirrlees Review, Oxford University Press for Institute for Fiscal Studies, Oxford.

Norwegian Ministry of Finance (2017), “The Norwegian Tax System - Main Features and Developments”, Chapter 2 of the Budget Proposal on Taxes 2017: https://www.statsbudsjettet.no/upload/Statsbudsjett_2017/dokumenter/pdf/main_features_tax_system.pdf

OECD (2017), Revenue Statistics: 1965-2016, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264283183-en

OECD (2018), Taxation of Household Savings, Tax Policy Studies, OECD Publishing, Paris.

OECD (2007), Fundamental Reform of Corporate Income Tax, OECD Tax Policy Studies, No. 16, Paris, OECD Publishing. http://dx.doi.org/10.1787/9789264038127-en

OECD (2006), Fundamental Reform of Personal Income Tax, OECD Tax Policy Studies, No. 13, OECD Publishing. http://dx.doi.org/10.1787/9789264025783-en

Pichet, E. (2007), “The Economic Consequences of the French Wealth Tax”, La Revue de Droit Fiscal, Vol. 14, pp. 1–25.

Piketty, T., Saez, E. and G. Zucman (2013), “Rethinking Capital and Wealth Taxation”, preliminary version.

Rudnick, R. S. and R. K. Gordon (1996), “Taxation of Wealth”, in V. Thuronyi (ed.), Tax Law Design and Drafting, International Monetary Fund.

Saez, E. and G. Zucman (2016), “Wealth Inequality in the United States since 1913: Evidence from Capitalized Income Tax Data”, Quarterly Journal of Economics, Vol. 131, No. 2.

Schnellenbach, J. (2012), “The Economics of Taxing Net Wealth: A Survey of the Issues”, Freiburger Diskussionspapiere zur Ordnungsokonomik, No. 12/5.

Seim, D. (2017), “Behavioral Responses to Wealth Taxes: Evidence from Sweden”, American Economic Journal (forthcoming).

Silfverberg, C. (2002), “The Swedish Net Wealth Tax: Main Features and Problems”, Stockholm Institute for Scandinavian Law.

Slemrod, J. (1992), “Do Taxes Matter? Lessons from the 1980s”, American Economic Review, Vol. 82, No. 2, pp. 250–256.

Stevens, L.G.M., Albregtse, D.A. , Kavelaars, P., Smeets M.H.M. and Y.M. Tigelaar-Klootwijk (2006), “Fiscaal Commentaar,” Inkomstenbelasting, 2006, Kluwer, Deventer.

Straub, L. and I. Werning (2014), “Positive Long Run Capital Taxation: Chamley-Judd Revisited”, NBER Working Paper Series, No. 20441.

Toder, E. and A. D. Viard (2016), “A proposal to reform the taxation of corporate income”, Tax Policy Center, Urban Institute & Brookings Institution, Washington D.C.

Weale, M. (2010), “Taxation of Wealth and Wealth Transfers: Commentary by Martin Weale” in Dimensions of Tax Design: The Mirrlees Review, Oxford University Press for the Institute for Fiscal Studies, Oxford University Press.

Zoutman, F. T. (2015), “The Elasticity of Taxable Savings”, CESifo, 9th Norwegian-German Seminar on Public Economics.

Zucman. G. (2015), The Hidden Wealth of Nations: The Scourge of Tax Havens, University of Chicago Press.

Note

← 1. As mentioned previously, this chapter is based on the tax rules that were in place as of 1 September 2017. Since then, France has replaced its net wealth tax with a new real estate wealth tax, with effect from 1 January 2018.