Australia's Future Tax System

Final Report: Detailed Analysis

Chapter C: Land and resources taxes

C2. Land tax and conveyance stamp duty

C2–3 Current taxes on land

There are currently three taxes on land in Australia. The first is property conveyance duties (stamp duties) levied on the transfer of land and buildings. In 2007–08 they raised $14.4 billion for State governments. A significant proportion of this revenue is raised on the transfer of building values, rather than of land. The second is local government rates levied on land (and also on building values by some councils). They raised $10.2 billion in 2007–08. Finally, State government land tax (mostly levied on unimproved land values) raised around $4.3 billion in 2007–08.

Stamp duty

Each of the States levies stamp duty on conveyances (the transfer of property), both residential and commercial. The duty is usually remitted to the State Revenue Office by the purchaser of the property, based on its reported sale price (or the market price, if that is deemed to be a fairer representation of the value). The value of the property includes the value of land and buildings.

The average rate of stamp duty across States has risen from 2.45 per cent in 1993 to 3.25 per cent in 2005, largely due to the non-indexation of the scales in the face of property value appreciation. However, rates are variable across States and different types of property. The highest rate of stamp duty is 7 per cent for residential properties valued above $3 million in New South Wales.

Each State sets different rates and thresholds for stamp duty on conveyances, and within the one State, different rates and thresholds can apply to the purchase of different types of property. For example, the purchase of non-residential property may be treated differently to the purchase of residential property. The States adopt a progressive rate scale for stamp duty; as the value of the property increases, so does the average rate of stamp duty. Chart C2–2 shows for each State the stamp duty payable on residential properties valued at $300,000, $500,000 and $1,000,000.

Chart C2–2: Stamp duty payable by State and property value

Chart C2–2: Stamp duty payable by State and property value

Note: Assumes residential property and that the purchaser is not eligible for a concessional rate of stamp duty.

There are programs in each of the States that provide discounted rates for first home buyers, often limited to less expensive homes. In NSW, for example, the First Home Plus Scheme provides eligible purchasers with an exemption from stamp duty on homes valued up to $500,000 and concessions on duty for homes valued between $500,000 and $600,000. There are other programs that provide concessions and exemptions for particular groups, such as pension card holders.

Stamp duty is a simple tax

Stamp duty is a relatively simple tax to collect, since it is levied on the sale price, which is easily observable. Administrative simplicity was one of the main reasons why stamp duties were first introduced. The maintenance of a property right system by governments — for example, the maintenance of title deed offices — made it administratively simple to levy a tax on transactions, particularly since land values needed to be reported (see Box C2–2). However, now that broad-based taxes on income and consumption are available, the relative simplicity of stamp duty is not a strong justification for retaining the tax.

Box C2–2: A brief history of stamp duty

A stamp duty is any tax levied on a legal document, like a contract for sale of a business or land. In Australia, stamp duties were first levied by the colonial governments before federation. A physical stamp had to be attached to or impressed upon the document to denote that stamp duty had been paid before the document became legally effective. This included documents relating to many items including wages, unemployment insurance, beer, cheques, cattle and pigs. Most of these duties have now been removed. Those on the transfer of a business or real property, the registration of a motor vehicle and insurance contracts are the only significant duties remaining. While the use of adhesive stamps on documents has now been abolished, the related tax obligation has not.

Stamp duty stamp used in Queensland in the 1950s

Stamp duty stamp used in Queensland in the 1950s

Source (image): Dave Elsmore, ozrevenues.com.

A large but volatile revenue source

In recent years, stamp duties on conveyances have been a significant source of revenue for the States. Indeed, in some States, stamp duty revenues have sometimes been the main source of revenue. As a proportion of gross domestic product, taxes on financial and capital transactions in Australia, which mainly comprise stamp duties, are twice the average of OECD countries.

Revenue from stamp duty is volatile. This is because the tax base is determined by two variables that can be subject to significant swings in short periods of time: the value of properties being transferred and the number of properties being transferred. For instance, around 52,000 established properties were turned over in Sydney in 2007, but only 42,000 in 2008, a fall of 19 per cent. The progressive nature of conveyance duty rates can add to this volatility. Chart C2–3 illustrates stamp duty volatility, both in terms of actual revenue and as a proportion of total State tax revenue, over the past decade.

Chart C2–3: Revenue from conveyance duty

Chart C2–3: Revenue from conveyance duty

While Chart C2–3 shows the aggregate for all States, recent experience suggests that the housing cycle can move at different times in different States, so the chart masks the volatility in revenue that can arise for individual States. For example, revenue from conveyance duty in Western Australia is expected to decrease from $2.3 billion in 2007–08 to $1.1 billion in 2008–09, a fall of over 50 per cent.

A tax on transferring property, not on land

Section C2–1 outlined how a tax on land value can be efficient because it is levied on an immobile base and is difficult to avoid. Stamp duties do not have these properties.

Stamp duty is triggered by the sale of a property. This creates the possibility for people to avoid stamp duties by choosing not to buy or sell property, which can result in people not living in the house they really want to live in or staying too long in a house that could be better used by somebody else. This probably results in a poor allocation of the housing stock. Though the efficiency impact of transactions taxes are difficult to estimate, one study suggests that stamp duties have efficiency costs more than ten times as great as those of a recurrent tax on the market value of houses (O'Sullivan et al. 1995).

Since stamp duty applies to the whole property value, to some extent it taxes the capital used to improve land. While land is immobile, the capital used to improve it is not. Discouraging capital owners from investing in property improvements — particularly improving and selling property — is particularly inefficient.

The most obvious way stamp duty biases decisions is that it discourages people from moving. The effect of stamp duty on the decision to move is determined by the size of the tax in comparison to the non-tax costs of moving, such as real estate agent fees, removal costs and search costs. Stamp duty can double these costs. Indicative estimates of the effective tax on the decision to sell one median sized home and buy another are depicted in Table C2–1 (based on a similar table in Hird 2007). Because stamp duty rates are progressive, the effective rate of tax is generally higher in cities with higher house prices.

Table C2–1: Stamp duty expressed as a tax on moving in capital cities

  Value of median home,
June 2009
($)
Stamp duty payable
($)
Other moving costs
($)
Total costs of moving
($)
Effective tax rate on moving
(%)
Sydney 544,000 19,970 21,320 41,290 94
Melbourne 441,900 18,484 18,257 36,741 101
Brisbane 419,000 5,915 17,570 23,485 34
Perth 450,000 15,390 18,500 33,890 83
Adelaide 359,000 14,280 15,770 30,050 91
Hobart 336,000 10,990 15,080 26,070 73
Canberra 458,000 18,240 18,740 36,980 97
Darwin 537,100 26,586 21,113 47,699 126

Note: Other moving costs assume real estate agent fees of 3 per cent on the value of the home as well as a flat $5,000 cost in all States. Stamp duty payable assumes that the buyer is not entitled to concessions such as first home buyer assistance. These estimates overstate the monetary non-tax costs of moving for those vendors who choose not to engage a selling agent or professional removalists.

Leigh (2009) finds that a 10 per cent increase in the level of stamp duty reduces the numbers of properties exchanged by 4–5 per cent if the increase is sustained over a three year period. This suggests that current rates of stamp duty prevent a substantial number of mutually beneficial housing exchanges. Reduced turnover of housing can have a significant impact on people's lives (see Box C2–3).

Box C2–3: The real-world effects of stamp duty

Making housing transactions more expensive means that people tend to move less (Van Ommeren & Van Leuvensteijn 2005; Van Ommeren 2008). This can have a range of efficiency and equity effects, including:

  • People may commute more, creating greater road congestion (Larsen et al. 2008).
  • People who want larger houses may choose to renovate, rather than move; or they may buy a larger house than they need in anticipation of eventually needing the space. This could lead to a housing stock that is larger than necessary, which may have environmental consequences.
  • Making housing transactions more expensive may lead to higher unemployment, as people are less likely to move to get a job, and to lower productivity, as there is greater impediment to shifting to a better-paying job (Van Ommeren 2008).
  • Some groups may have less access to the housing market since they need to save to pay the stamp duty.
  • Stamp duties may discourage older Australians from moving to a smaller home and reduce the amount of equity withdrawn from a home if they do downsize (Wood et al. forthcoming).

Stamp duties are a particularly bad tax on business

Stamp duties tax transactions in property, but also the value-add from capital investment. Stamp duties are a particularly inefficient tax when levied on business. This is because businesses face incentives to minimise their transactions and investment in property. For example, a business has incentive to use existing buildings rather than moving to a lower cost region and buying a new property. As businesses are more likely to be mobile than consumers, stamp duties are likely to be particularly inefficient. Consumers are worse off in two ways — goods and services are provided using less efficient processes, and higher tax rates apply to those goods and services that disproportionately depend on property for their production.

Stamp duties are particularly complex for many businesses. Most residential properties involve one transaction within a single State jurisdiction. Business transactions in property (involving land and non-land assets) can involve changes in the ownership of people indirectly related to a transaction (such as unit dealings in a unit trust). Such complexities involve dealing in high-value commercial property transactions. For example, there are many differences between State corporate reconstruction legislation and how to treat unit trusts that are complex and influence business investment.

Stamp duty is inequitable

Given that higher valued properties are often purchased by people with higher incomes, it may appear equitable that the average rate of stamp duty increases as the value of the property increases. However, as property is just a part of a household's consumption and wealth, stamp duties are a poor mechanism for improving equity. The tax instead falls most heavily on people with a preference for housing consumption. For example, one person of considerable means might buy an expensive house and pay more tax, while another does not because they prefer an expensive motor vehicle.

A further equity dimension to stamp duty is that it will always result in a differential tax burden depending on people's desire to move. Chart C2–4 illustrates how the effective rate of stamp duty on housing differs according to the length of time spent in a property and the frequency of moving. The stylised chart shows the effective tax rate of stamp duty as a proportion of the cumulative imputed rent (the value from living in the home) over time. The effective rate of tax declines over time as the up-front cost of stamp duty is spread over more years of occupancy. If a household sells their house and purchases another, the cycle starts over again — meaning that people have an incentive to stay in the same house.

Chart C2–4: Stamp duty effective tax rate(a)

Panel A: Effective tax rate falls with occupancy duration

Chart C2–4: Stamp duty effective tax rate - Panel A: Effective tax rate falls with occupancy duration

Panel B: Increased tax rate on people who move

Chart C2–4: Stamp duty effective tax rate - Panel B: Increased tax rate on people who move

  1. The effective tax rates are calculated as the ratio of stamp duty (assumed to be $20,000) to the value of imputed rent over the period the property is owned (assumed to be $25,000 per annum). In Panel B, the 'flat rate' reflects a constant tax on imputed rent, with the rate equal to the effective rate faced by a person making two moves in 25 years (which is not average but intended to be indicative).

Source: Treasury estimates.

While around half of owner-occupiers have occupied their house for nine years or less, 18 per cent of owners have brought within three years and 26 per cent stay in their home for at least 20 years. People who have to move more frequently because of their work or large changes in their life (for example, birth of children, divorce, or a new partner) will face higher rates of tax, regardless of their means.

Stamp duties also make it more difficult for credit constrained potential home buyers to access the market. For example, Wood et al. (2006) found that stamp duty accounted for around 23 per cent of up-front cash costs of renters who may be potential home buyers. Though stamp duty is an unnecessary impediment, its removal would not be likely to lead to a large increase in access to owner-occupied housing for renters of limited means.

Finding

Existing State stamp duties on property conveyancing are highly inefficient, distorting both residential and business use of property.

Stamp duty encourages people to stay in houses when they would prefer to move, contributing to longer commuting times, larger average home sizes and lower labour mobility.

Stamp duty is also inequitable as people who move more regularly— such as those needing to change homes for work — pay more tax than those who do not. Stamp duties also directly reduce access to housing for people who are credit-constrained.

Council rates

Council rates are broad-based, low-rate taxes levied on the value of land. They raised $10.2 billion in 2007–08. Council rates are administered by local governments to fund certain services they provide, such as sanitation and planning administration (see Section G3 Local government for more details).

Land value is generally not directly observable from vacant land transactions. Valuation methodologies differ from council to council and can also differ from the method used to value land for State land tax (see Box C2–4). Some councils base the tax on the value of the land only, while others base the tax on total property value (land and buildings).

Box C2–4: Different approaches to levying ongoing land value taxes

Methods of valuing land for tax purposes vary from State to State. There are subtle differences in base definitions of value in each State, but the following broad categories are indicative.

Measures of the value of land itself

Unimproved value, unimproved capital value, land value and site value are currently the bases on which land-only taxes are determined. Each of these bases is the value of the land without 'improvements' (for example, buildings as well as, in some bases, draining, levelling or filling). Site and unimproved capital value are similar, as both include the value of merged improvements (such as draining) in their values, though do not include building values. All of these valuations are influenced by the effects of nearby infrastructure (such as access roads, schools and parks).

Measures of land and buildings

Capital value and capital improved value include the total market value of the land, including any buildings or other improvements.

Annual value, annual assessed value and gross rental value estimate the sum of all rental payments that are paid to the landlord in a year or would be if the property was rented. These measures give a similar tax result to capital improved value. However, they do not allow for the deduction of the costs a landlord would incur in maintaining the land.

Net annual value is also the rental value of the property but allows the deduction of landlord's costs, including land taxes and maintenance costs.

Table C2–2: Current valuation methodologies for council rates and land tax

  NSW VIC QLD WA SA TAS NT ACT
Council rates LV SV, NAV, CIV UV Rural: UV
Non‑rural: GRV
CV, SV, AV LV, CV, AAV UCV, AV, ICV UV
Land tax LV SV UV UV SV LV Not levied UV

Notes: AV = Annual value, AAV = Assessed Annual Value, LV = Land Value, CV = Capital value, CIV = Capital Improved Value, GRV = Gross Rental Value, NAV = Net Annual Value, SV = Site Value, UCV = Unimproved Capital Value, UV = Unimproved Value, ICV = Improved Capital Value.

Sources: Productivity Commission (2008); Mangioni (2006); NSW Treasury (2009).

Overall, council rates are relatively efficient, simple and fair taxes. This is consistent with the indicative modelling of efficiency costs of taxes calculated for the Review (see Part 1.7). Rates are generally applied to all land uses with limited exemptions and apply equally to all properties within the council area.

However, the efficiency of council rates is likely to be reduced in councils that use improved values to assess the tax, as this discourages capital improvements. Further, councils often levy rates based on the zoning of land, with higher rates for commercial, compared to residential and rural property (see Chart C2–5). While these differential rates may be used as a proxy for imposing higher rates on higher value land, a direct method could achieve the same result without the risk of influencing the zoning process. Further, a segmented approach to land value taxation is more likely to result in the tax burden being passed to users rather than being borne by landholders. This reduces the efficiency of the tax.

Chart C2–5: Effective property rates

Rates by decile of property value (2004–05)

Rates by decile of property value (2004–05)

Source: Productivity Commission 2008.

The variation in valuation bases and methods from State to State and from council to council may be a source of complexity for landholders across different jurisdictions. For most payers, however, rates involve minimal compliance effort. The State governments' Valuer-General typically generates the valuation, the State Revenue Office (SRO) generates the assessment and, as long as the taxpayer pays the assessment, there is no risk of penalty. The low rates, lack of thresholds and limited range of concessions provide limited tax planning opportunities.

Local government rates are also a stable revenue source. This is especially when a moving average of recent valuations is used to determine the tax base, which minimise short-run fluctuations in land values. They are also a sustainable base as land values tend to climb steadily over the long run.

Land value taxes are a good base for local governments as there is a direct connection between the level of services delivered and the residents who benefit (see Section G2 State tax reform).

When comparing average tax rates across a State, council rates can appear regressive, as higher rates are generally levied in councils with lower property values (see Chart C2–5). This is likely to reflect the fact that many of the services provided by local governments cost the same regardless of the means of its recipients. In addition, these costs can be higher in rural or remote communities, which often have lower land values. As local government services benefit residents of particular areas, it is appropriate that their residents pay for them through rates. However, the provision of Financial Assistance Grants to all councils — even those with significant local revenue-raising capacity — may reduce the average tax rate in councils with high land values. This issue of Financial Assistance Grants is considered in greater detail in Section G3 Local government.

Land tax

Land tax is a general revenue tax levied by all States except the Northern Territory. Depending on the State, it is calculated on the 'unimproved' or 'site' value of land. Although the details, thresholds and tax rates vary between States, it generally applies only to a limited range of commercial land and investor-owned residential land. A range of land uses are exempt, including primary production, owner-occupied residential, child care and aged care. Land tax raised $4.3 billion in 2007–08.

Land taxes are levied according to a progressive rate scale. In all States (other that the ACT), these rates are based on an entity's total land holdings. Many States also apply substantial minimum thresholds before any tax is levied. Chart C2–6 reflects the thresholds and average rates applied to land holdings in each State.

Chart C2–6: Thresholds and average rates of land tax

Chart C2–6: Thresholds and average rates of land tax

Note: Land tax in the ACT is determined on a value per property, not on aggregate holding.

Source: NSW Treasury (2009).

A narrow-based tax

The thresholds applied to land tax and the wide ranging exemptions reduce the efficiency and equity of the tax.

The major exemption from land tax is owner-occupied housing. This exemption removes around 60 per cent of land by value from the tax base. Another significant exemption is land used for primary production. Despite the significant amount of land that this exemption covers, it represents only around 10 per cent of the total land value (see Chart C2–7 Panel A). Significantly, these exemptions have excluded from the tax base the land with the fastest recent growth in value (see Chart C2–7 Panel B).

Also excluded from the base are leasehold land, State and Commonwealth-owned land and land owned and used by non-profit organisations and charitable institutions.

Chart C2–7: Land values and growth

Panel A: Land value by category (2007–08)

Chart C2–7: Land values and growth - Panel A: Land value by category (2007–08)

Panel B: Growth in aggregate land values

Chart C2–7: Land values and growth - Panel B: Growth in aggregate land values

Source: ABS cat. no. 5204.

Substantial exemptions harm the efficiency of any tax by encouraging economic activity to move to the untaxed sector. In this regard, the large thresholds applied in some States have the effect of exempting small landholders from the tax. For land tax, the efficiency cost is also likely to be compounded by the burden of the tax shifting from landholders to land users.

The exemption is likely to have particular influence on land for residential property. The exemption of owner-occupiers rules out around 75 per cent of residential land and, for the remainder, high thresholds in some States effectively exempt many small-scale investors. As land can shift in and out of the tax base depending on who owns it, it is unlikely that the tax will be fully reflected in lower land prices for residential property. The portion of tax that is not reflected in lower land prices is borne by investors through lower returns, or by their renters through higher rent. This means the tax, to some extent, has been passed forward to workers and the owners of capital. Further, it is likely that, in the long run, much of the burden of the tax is shifted to renters, as rents adjust to ensure that investors achieve an adequate return. This may be inequitable, as renters generally have low income and wealth.

Significant exemptions also make land tax more complex to comply with and to administer.

Higher tax on aggregate holdings discourages large-scale investment in land

When the Australian colonies introduced land taxes in the late nineteenth century, higher tax rates on aggregate holdings were introduced to encourage large rural landholders to subdivide their land and sell it to settlers (Smith 2004). As rural land is no longer in the base, this rationale for higher rates on larger aggregate holdings is no longer applicable.

Today these rules lead to higher taxes on larger landholders. The most significant consequence of this approach is a bias against large investments in residential property. The land tax scales tax more heavily any corporation or individual that seeks to make a large investment in land, such as for residential housing. For the States that levy land tax on an aggregate basis, Table C2–3 depicts the different rate of tax per dwelling for a small and large investor in each State. The much larger share of rent that land tax represents places large investors at a significant competitive disadvantage. This is likely contribute to the investment housing market being dominated by small investors.

Very few institutional investors invest in private rental housing. The aggregate holding approach deters these potential long-term investors from the market, as do a number of elements of the existing income tax system. Policies that discourage such investment are particularly perplexing given that such investors may be a better match for private tenants who desire long-term tenure.

Table C2–3: Effect of aggregation returns to rental property investment

  Sydney Melbourne Brisbane Perth Adelaide Hobart Average
Median home ($'000) 544,000 441,900 419,000 450,000 359,000 336,000 424,983
Rent at 5% ($'000) 27,200 22,095 20,950 22,500 17,950 16,800 21,294
Small investor              
Land Tax per property ($) 0 0 0 0 209 837 174
Proportion of rent 0% 0% 0% 0% 1% 5% 1%
Large investor              
Land Tax per property ($) 4,848 3,270 3,190 2,232 5,618 3,844 3,834
Proportion of rent

18%

15%

15%

10%

31%

23%

18%

Note: The small investor is assumed to hold one median priced dwelling, with 25 held by the large-scale investor. The land value is assumed to be half the value of the property. Median house price is for a 3 bedroom dwelling at June 2009. In Perth, the calculation of land tax includes the Metropolitan Regional Improvement Tax. In Brisbane, the investor is assumed to be a corporation.

Source: Real Estate Institute of Australia (2009); NSW Government (2009) and Treasury calculations.

Finding

Several features of current land taxes, in particular their narrow base, make them less efficient and fair than they could be.

By levying the tax at increasing rates on an entity's total holding, land tax discourages large-scale investment in land, particularly for rental housing.

Because owner-occupied housing is exempt, the burden of land tax on residential investment properties is probably borne by renters through higher rents.