Final Report: Detailed Analysis
A3. Wealth transfer taxes
A central question for any tax system is whether, and how, to tax savings. The Review sees a role for the taxation of savings in Australia's future tax system but one that is more neutral between different forms of saving and that imposes smaller efficiency costs than the existing system (see Section A1–3). One element of the future taxation of savings could be a bequest tax; that is, a tax on the assets a person leaves behind at their death.
A bequest tax levied at a low flat rate, and designed to affect only large bequests, could be an efficient and equitable component of Australia's future tax system.14
Most taxes introduce biases into economic decisions and thus impose efficiency costs on society overall. But a tax on bequests would impose lower efficiency costs than many other means of taxing savings. The motives for leaving bequests have important impacts on the biases that the tax would bring to savings behaviour and, consequently, on the efficiency costs of the tax.
The tax system should aim to raise revenue with low efficiency costs — while also taking into account issues of equity, simplicity, sustainability and consistency with other policy objectives. Consideration should be given to any tax that would raise revenue with low efficiency costs and no large adverse effects on other tax policy considerations.
While altruism towards recipients, predominantly children, is an important motivation for leaving bequests, there are two other important scenarios.
First, many bequests are unplanned. The length of a person's life is uncertain and in many cases a donor saves not to leave a bequest but solely to ensure their own financial security. If they live for as long as they expect to, they consume their savings and pay tax as they do so. If they die early, however, a substantial amount of savings may remain. These savings are usually bequeathed to someone, but bequeathing the assets is not the purpose for which they were acquired. In this sense, the bequest is unplanned.
Second, some bequests may be a form of compensation for services provided by a recipient to the donor. For example, an elderly donor may leave a bequest to a neighbour who has devoted time and effort to caring for them.
How taxing bequests would affect donors
An important consideration in designing a tax on savings is to limit the extent to which the tax encourages people to consume now rather than save and have their savings taxed in future (see Section A1–3). Taxing savings at the time of death largely avoids such biases, as a person has at that point finished saving in order to consume later in life. Taxing bequests would, however, have some impact on donor saving decisions — and, therefore, some efficiency costs — depending on the donor's motive for accumulating wealth.
- Saving undertaken solely to ensure the donor's financial security would not be affected by a bequest tax because it is motivated by the donor's concern for their own wellbeing, not the desire to leave a bequest.
- Economic theory suggests that the impact of a bequest tax on saving for altruistic motives is uncertain. The tax means that the donor must forgo more consumption to deliver the same after-tax benefit to the recipient. This discourages saving. However, it also means that the range of consumption and bequest options open to the donor has been diminished. Whether this will induce the donor to save less or more will depend on their individual preferences. If the donor is determined to leave a post-tax bequest of a certain size, the tax may encourage them to save more. If they are more concerned about their current consumption, they may save less.
- For the same reasons, the net impact of a bequest tax on the saving of donors who are seeking to compensate recipients for services rendered is not clear in theory.
Many donors are likely to be moved by a combination of these motives, so that estimating the impact of a bequest tax is essentially an empirical matter. There is no doubt that the bequest motive has some impact on savings decisions. Life insurance choices and patterns of giving before death show that altruism has some effect on donor choices. On the other hand, there is a considerable body of evidence that precautionary savings, which result in unplanned bequests, account for a significant proportion of savings (Parker & Preston 2005). As purely precautionary savings would be unaffected by a bequest tax, it is unlikely that taxing bequests would greatly affect donor savings (Gale & Perozak 2001).
This suggests that a bequest tax would be more efficient than some other means of taxing savings — particularly taxation of the returns to saving as they accrue (see Section A1–3).
Recipients would work and save more
Because they do not receive, or do not expect, such a large post-tax inheritance, recipients and prospective recipients tend to work more and save more in the presence of a bequest tax. Evidence from the US suggests that receipt (or expected receipt) of an inheritance reduces labour supply, although the impact is not great (Holz-Eakin et al. 1993). Consequently, the improvement to labour supply that would arise from reducing post-tax bequests through a bequest tax, while positive, is likely to be small.
Double taxation is not good or bad in itself
Bequest taxes are sometimes opposed on the grounds that they involve double taxation: much of the income saved to build up the bequeathed assets has been taxed at the time it accrued to the donor. 'Double taxation' is not good or bad in itself. Any system that taxes economic flows at more than one point will involve an element of double taxation. For example, the current system taxes an individual's labour income as it accrues and taxes the part that is consumed a second time, through the GST. There is no reason to try to avoid such double taxation by adopting a system with a single tax. A system that raised all its revenue from a single household income tax would impose very large costs on taxpayers despite the absence of double taxation. The important thing is to design and implement a system that raises enough revenue while limiting the costs of doing so.
The efficiency of a bequest tax depends on the motivation of the donor and the benefits the donor and recipient receive. Taxing unplanned bequests is particularly efficient, as people would not change their behaviour to avoid the tax. Taxing planned bequests involves efficiency costs but these are relatively low.
A bequest tax could be seen as improving horizontal equity — that is, it could help ensure that people in the same economic situation pay the same amount of tax. If two people have the same economic resources and are similar in all relevant respects, apart from the fact that one earned a particular sum as an employee and the other inherited the same sum, it may be regarded as inequitable that tax falls only on the person who earned the sum as an employee. In general, a bequest tax taxes income from unrequited transfers rather than from work and saving.
A bequest tax could potentially be a highly progressive element in the overall tax system. The richest 10 per cent of households headed by a person aged 65 or older hold 43 per cent of such households' total wealth, while the top 20 per cent hold 58 per cent (see Chart A3–1). As the distribution of wealth is very uneven, a bequest tax could apply only to the largest wealth holders in Australia and still raise a large proportion of the available revenue. Such a tax would make the wealth distribution a little more even but no feasible rate of tax would have a major impact in this regard.
Chart A3–1: Percentage of total household wealth by wealth decile, Australia, 2005–06
Note: 'Older than 65' denotes households headed by a person at least 65 years old. 'Older than 65' figures show the distribution of wealth held by older households.
Source: ABS 2007, unpublished data.
Another important equity issue is the parity of treatment between married and unmarried donors. The primary goal of a bequest tax is to raise revenue while moderating the passing of economic resources between generations. This suggests that concessional treatment should be extended to bequests received by spouses. These are not intergenerational transfers, and if bequests to spouses were fully subject to a bequest tax, the assets of a couple would usually be taxed twice before reaching the next generation, while the assets of a single person would not.
Overall, a well-designed bequest tax would improve equity as it would help to distribute opportunities more evenly across the community.
A core theme for the Review has been to reduce complexity in the tax and transfer system. The introduction of a new tax on savings would inevitably involve some new complexity, and present at least some tax-planning opportunities for taxpayers approaching the end of their lives. Tax planning need not be illegal nor immoral, and in many cases it constitutes rational behaviour on the part of the taxpayer. But it contributes nothing to the wellbeing of Australians overall. Rather than creating value, it is entirely concerned with how much value should be transferred from the taxpayer to the government.
Any estate, inheritance or accessions tax would need to be accompanied by a means of taxing gifts, as an anti-avoidance measure. Otherwise, many donors would be able to avoid the tax by transferring their assets to the intended recipients shortly before the time of death. Other anti-avoidance provisions would also be necessary, although these may not need to differ greatly from existing provisions in other areas of the tax law.
US experience with its estate tax suggests that the best available tax planning is able to remove around one-third of a medium-size estate from the estate tax net (Schmalbeck 2001). Many of the expedients that minimise tax are undesirable to donors on other grounds; for example, they may oblige the donor to relinquish control of their assets many years before their death. Leakage from the US estate tax base is significant but falls well short of wholesale avoidance.
It is important to remember that the Review is proposing consideration of a bequest tax that falls only on the largest estates or inheritances. The large majority of estates or inheritances would not be subject to tax. Nevertheless, it is clear that the introduction of a tax on bequests would involve provisions of significant complexity. The tax base would need to be defined, avoidance countered and rates specified. This complexity and its attendant costs must be weighed against the efficiency of the tax in raising revenue and the contribution it could make to the redistributive goals of the tax and transfer system, its sustainability and its policy coherence.
The introduction of a new tax on bequests would involve some new complexity, and present some tax planning opportunities for taxpayers approaching the end of their lives. This additional complexity must be weighed against the efficiency and equity arguments in favour of taxing bequests.
Ideally a tax base should be sustainable, in the sense that it should continue to yield a predictable revenue stream over time. A tax on bequests would fit well with Australia's demographic circumstances over the next 40 years. Between 2003 and 2030, the proportion of all household wealth held by older Australians is projected to increase from 22 per cent to 47 per cent (Kelly & Harding 2003). Mortality rates among early baby boomers will begin to increase significantly from around 2015. Large asset accumulations will be passed on to a relatively small number of recipients. As a result, the amount of bequests passed on in Australia is estimated to rise from $22 billion in 2010 to $85 billion in 2030, in real terms (see Chart A3–2). This amounts to a projected increase from around 2 per cent of GDP to around 4 per cent.
Chart A3–2: Projected bequests, Australia, 2000–2030
Source: Kelly and Harding (2003) updated by Kelly. HILDA aggregates compiled by Kelly.
It is also worth noting that with longer life spans, children are inheriting from their parents much later in life, when they are often already well-established financially, so that the bequest is of decreasing importance as an economic support for children in early adult life.15
A tax on bequests would fit well with Australia's demographic circumstances over the next 40 years. Between 2003 and 2030 the proportion of all household wealth held by older Australians will more than double. Large asset accumulations will be passed on to a relatively small number of recipients, many of them in middle age.
14 This report uses 'bequest tax' as a generic term applying to any tax levied on assets at death. An 'estate tax' is levied on the whole of the donor's estate. An 'inheritance tax' is levied on the inheritances received by individual recipients. An 'accessions tax' applies on a cumulative basis to the gifts or inheritances received by an individual throughout their life. 'Donor' is used for the person who leaves a bequest or makes a gift and 'recipient' for the person who receives a gift or bequest.
15 Based on current mortality rates, a boy born in 2007 can expect to live 79.0 years while a girl can expect to live 83.7 years (ABS 2009i). The figures for children born in 1965–1967 were 67.6 and 74.2 respectively (ABS 2008).
Next Page – A3–2: The current system has some impact on bequests >>
<< Previous Page – A3: Wealth transfer taxes