Changing Housing Taxation Composition: A Review of Policy in the Australian Capital Territory

ABSTRACT

A tax reform introduced by the Australian Capital Territory (ACT) Government in 2012 aimed to ease the barrier of owning a home by replacing stamp duty with a broad-based general rates and land tax. This article assesses on the impact of this reform on the ability of low-income families to buy a house using a microsimulation model. The results show that tax reform has increased property turnover and reduced the amounts paid for stamp duty and rates for most groups of vulnerable families in the ACT. However, extreme increases in house prices may offset this gain for vulnerable families.

摘要

澳大利亚首都地区（ACT）政府于2012年推出了一项税收改革，旨在通过用有着广泛基础的普通税率和土地税取代印花税，来缓解买房压力。本文运用微观模拟模型评估了这项改革对低收入家庭购房能力的影响。结果表明，税收改革增加了房地产成交量，降低了澳大利亚首都地区大多数弱势家庭群体为印花税和房价支付的金额。然而，对弱势家庭来说，房价的极端上涨可能会抵消这一收益。

KEYWORDS:

• Tax reform
• housing
• home ownership
• stamp duty
• general rates
• land tax

1. Introduction

The Australian Capital Territory (ACT) is a symbolic centre of the Australian Commonwealth government. After formally being established in 1913, its development was relatively slow with only 40,000 people in 1957, but it then boomed fivefold in 20 years after that. There were more than 400,000 people living in the ACT in the beginning of the 2020s. Housing affordability starting becoming an issue as the ACT has become the second most expensive place in Australia to buy a house after Sydney (Strahorn 2022). The need to have large savings to buy a high-priced house is not only due to the deposit required, but also the tax associated with the purchase. The ACT Government’s 2012 tax reform proposed the phased elimination of conveyance duty that has to be paid at the purchase of a house, with an increase in general rates charged for public services provision to offset the reduction in conveyance duty. This reform has meant that the ACT is now among the jurisdictions in Australia with the lowest stamp duty rates. New South Wales, the most populated state in Australia, plans to also phase out stamp duty (Morris 2021).

This article aims to assess the impact of this ACT tax reform. In light of the issues identified by Pawson et al. (2020), the article addresses the research question related to the impact of tax reform on the ability to purchase a house, especially for disadvantaged and vulnerable households. In considering this, all the available concessions and deferral programs are considered.

Shelter is one of the basic human needs. Beer et al. (2011) argue that the existence of stable shelter, especially owning it, gives people a “wellbeing dividend”, such as improving social relationships as well as enabling employment and education. In addition, owning a house creates emotional security and a sense of belonging (Colic-Peisker et al. 2015, Pawson et al. 2020). Purchasing a home also provides a means of wealth creation. This offers financial security, especially for older people as the age pension system does not provide enough support for private renting. Ownership of a house also provides self-esteem and independence (Colic-Peisker et al. 2015).

The importance of homeownership in old age cannot be underestimated. Older people that rely on the age pension (or pensioners) are more disadvantaged due to the low rate of pension compared to a full-time wage, but home ownership can help ensure adequate living standards for older people (Doling and Elsinga 2013, Colic-Peisker et al. 2015, Wood et al. 2020). Yates (2015) and Wood et al. (2016) argue that homeownership has been a pillar supporting retirement income policy. However, this may no longer be the case for the next generation (Colic-Peisker et al. 2015).

This deterioration in access to the housing market has affected the affordability of all tenures. Private renting has become a long-term form of tenure for many, and especially those with a low income, due to the rapid increase in house prices (Beer et al. 2011, Sissons and Houston 2019). Although the low-interest rates at the end of the 2010s resulted in relatively lower costs of housing finance, the potential buyer still faces a large deposit to pay (Pawson et al. 2020). As the homeownership rate has fallen among the young adult population, there has been an increase in the demand for private rental housing among this cohort. This, in turn, has led to increasing rent levels, resulting in unaffordability not only in homeownership but also in private renting.

Another group that are considered disadvantaged in term of housing is female-headed households. This group is considered disadvantaged and vulnerable because women still lag behind men in terms of homeownership rates, and are identified as having ownership challenges not only in Australia but also in the UK and the USA (Kupke 2014). Gandelman (2009) shows that after taking many factors into consideration, the female-headed household has a lower proportion of homeownership in the majority of Latin American countries.

Growing concerns about housing in Australia have also entered the public policy debate (Beer et al. 2011). The Australian Federal Government’s direct housing assistance to low-income households can be classified into three forms: public rental housing, subsidies for private rental housing and assistance with entry into homeownership. First Home Owner Grant (FHOG) schemes and access support to private financing for households that may not otherwise be eligible for a housing loan are examples of the assistance provided by Government to enable home ownership (Armstrong et al. 2003, Wood 2006). Ironically, research has also revealed that government policy intervention has often become the source of house price inflation (Beer et al. 2011).

Another issue that has been identified in regard to government policy is that the housing tax system can be perceived as regressive. Stamp duty, which is a charge on the transfer of property, is one of the housing taxes that is often cited as being regressive, since its burden falls mostly on those who have not yet purchased a home (Carrington et al. 2019). Henry (2010), in their tax review, also highlights the inequitable nature of stamp duty since those who need to move more because of work requirements or certain life events also need to pay more tax. Therefore, there are growing calls to replace stamp duty with the introduction of some form of annual land tax on all residential property (see Carrington et al. 2019). This is not an easy task as the electorally dominant current homeowners may try to avoid having additional land tax (Morris 2021). In addition, there is still a lack of analysis on the impacts of government housing policy regarding the move from stamp duty to land taxes or rates.

In Australia, stamp duty can be charged by State and Territory governments on conveyances, mortgages, and mortgage insurance contracts (Wood 2006). For example, stamp duties on conveyancing are levied on the purchase price of the property, and therefore they increase the amount of the purchase cost and, hence, the necessary amount of the initial payment (deposit + stamp duty) to purchase a house. Therefore, stamp duty is seen to be a barrier to home purchase. Replacing stamp duty with an annual after-purchase payment, such as land tax, would increase the number of people able to purchase a home although there is always the question of whether those people can afford the stamp duty replacement of higher regular land taxes (Anacker 2019). In addition, replacing stamp duty with land tax may also help current homeowners to downsize their property (Colic-Peisker et al. 2015). Henry (2010) sees stamp duty as a barrier to downsizing as it increases the cost of downsizing. This means it is inefficient since it deters the property being allocated to the most efficient usage and prevents people moving to a more suitable property.

2. Data and Methodology

2.1. Data

The unit of analysis in assessing the impact of the tax reform is the income unit, which is also the tax unit in Australia. An income unit may consist of a single person or a couple, with or without any dependent children. As income units often overlap with families in Australia, we also use the term “family” to refer to an income unit in this paper. It is assumed that the individuals within an income unit share income and resources. This unit of analysis is used to separate homeowners who own at least one residential property and renters who do not own property but rent or share a dwelling with homeowners or other renters. In this case, a non-dependent person (aged 15 or above and not a dependent student) who is identified as a non-homeowner and who lives in their parents’ home is considered a renter, even though they may not pay any rent. However, they could become potential home buyers.

Given that one focus of this article is to assess the impact of the tax reform on disadvantaged and vulnerable families, some specific groups are analysed. As stated in the introduction, pensioners and female-headed households are disadvantaged in terms of housing. In this study, pensioners are defined as income units that have at least one person aged 65 years or above or who are receiving the age pension while female-headed income units are defined as those where the reference person (or the oldest person in the income unit) is female. These income units are classified into quintiles of low-, middle- and high-income groups based on equivalised disposable income per capita. The equivalising factor is based on the OECD-modified scale which assigns a value of 1 to the first adult in the income unit, 0.5 to the second and each subsequent person aged 15 and over, and 0.3 to each child aged 14 or below (Buhmann et al. 1988). Using this income variable, low-income homeowners and low-income renters are defined as those income units with incomes in the lowest 40% of all income units in the ACT. Middle-income homeowners and middle-income renters are defined as income units with incomes in the middle 41–60% of all income units in the ACT.

We also classify income units by net wealth level based on the Australian Bureau of Statistics (ABS) definition. Net wealth in our estimation includes financial assets (savings, offset and superannuation accounts, value of shares and trusts, business value, loans to other people); non-financial assets (property values, value of vehicles); net of liabilities (mortgages, personal and study loans, loans from other people, credit card debt). Similar to income, we classify families into low (bottom 40%), middle (middle 20%) and high (top 40%) wealth groups based on equivalised net wealth per capita using the OECD-modified scale.

Potential home buyers include renters and homeowners who intend to buy properties for investment or to live in. Home buyers will be a sub-group of these people, who can afford to buy properties without needing a loan or who meet the borrowing constraints. Following the findings of Wood (2006), it is assumed that renters would prefer to buy an appropriate dwelling instead of renting provided all borrowing constraints are met. Although it is also assumed that all renters purchasing are first-home buyers, they cannot all buy newly constructed homes and hence, not all are eligible to receive the First Home Owner Grant (FHOG). This is reflected in our model by limiting the number of potential buyers with the FHOG. Homeowners who may decide to change their housing (e.g. smaller homes for older people or bigger houses for couples with children) or to invest (e.g. rental properties for middle- and high-income homeowners) can also become home buyers in our model if they can afford to buy a property without a loan or if they meet the borrowing requirements.

The database for this analysis is set up by creating a synthetic population for the ACT using a spatial microsimulation method (Tanton 2011). The spatial microsimulation model uses a sample of all households living in the Australian Capital Territory (ACT) and Northern Territory (NT) from the 2009–2010 and 2011–2012 Surveys of Income and Housing (SIHs) conducted by the ABS to be reweighted according to the various characteristics of the ACT population based on the 2016 Census of population and housing. The use of ACT and NT unit record data is because the samples share the same geographical identification code on the survey, so they cannot be separated (Nepal et al. 2010). However, Nepal et al. (2010) show this does not cause a major problem since the reweighting technique used in the spatial microsimulation ensures that the sample is representative of the ACT population. The changes in population characteristics between 2011–2012 and 2018–2019 are derived from the age and sex-specific population projections from the ACT government.

2.2. Methodology

This study simulates the different income units produced by the spatial microsimulation model based on the policy scenario. The base condition is the current condition where the reform has happened while the counterfactual condition in our modelling is one in which the tax reform to phase out stamp duty did not occur. The main simulation in the model is the home purchase decision, or assignment, which identifies whether renters and/or homeowners can afford to purchase an appropriate property, and, if so, assigns them as home buyers if they meet the borrowing requirements.

The assignment of expected housing to different income units is important for understanding the possible costs that those income units are facing to secure a house. Therefore, the database of income units in the ACT needs to include the type and value of housing that would meet the needs of these income units. To do this, an ordinary least squares regression analysis of a sample of all homeowners was conducted. The natural logarithm of house value was used as a dependent variable, with the characteristics of the income unit reference person (age, gender, education, employment status) and the characteristics of the income unit (number of income earners, number of dependent children, gross income) used as the independent variables. Two dummy variables were also added as independent variables: homeownership and newly constructed property. The estimated coefficients were then utilised to predict the expected house values of all potential home buyers based on their characteristics, their income and their assignment to purchasing a new property.

To purchase property, there are two requirements that need to be fulfilled by income units – deposit and repayment requirements. For the deposit requirement, potential home buyers should have sufficient assets (e.g. cash in the bank, other equity in housing) to pay for a certain proportion of their expected house price (the deposit requirement and transaction costs) This amount includes any tax, concessions and /or grants. In Australia, potential home buyers can borrow up to 90% of their expected house prices, and banks will allow a 10% deposit with lender mortgage insurance. While owners can use the equity in their current house as a deposit, the SIHs 2011–2012 to 2017–2018 showed that more than 20% of homeowners in the ACT borrowed more than 80% of their house value. This number was much higher for recent homeowners (more than 40%) and recent first-home buyers (more than 50%) over the period. Therefore, to check whether potential home buyers meet the deposit requirement given their predicted house prices, the rule used is whether the deposit (or savings) and the total value of the homeowner’s current property are higher than the sum of stamp duty (with concession). The transaction cost also includes the lender mortgage insurance premium for the maximum loan-to-value ratio of 90%, and 10% of the total predicted value of the new house. If there are enough savings to pay a deposit of 20% of the house value, then this is applied to reduce the repayment requirement. For renters and those with rent-free housing (a very small proportion of the population), both deposit and repayment requirements are applied to model whether they become homeowners.

To calculate the repayment requirement, the simulation ensures that potential home buyers can afford to pay the minimum expenditure on including tax as well as loan interest, loan fees and principal repayments every year. Therefore, the maximum monthly repayment amount is estimated based on gross annual income and expected rental income (if any), with minimum annual expenditure on costs like food and clothing, annual income tax (net of offsets), annual Medicare levy, general annual rates (net of rebates) and annual loan fees subtracted from this amount. Renters will transition to homeownership and the homeowners will buy a property if both conditions– deposit and repayment requirements – are satisfied.

Another important note is that the model applies inflation factors to income, savings, land and house values to inflate the data to the financial year of 2018–2019. ACT Average Weekly Earnings is the inflator for income and savings while the ABS House Price Index (HPI) is the inflator for house and land value.

As stated in the introduction, the ACT Government’s reform is mainly targeted at conveyance duty, which is commonly known as stamp duty. This is a tax that people have to pay when they buy a property. Higher duty rates are applied to higher property values. The property value thresholds and duty rates for the ACT were unchanged between July 2002 to June 2012. From July 2012, these thresholds and rates changed and reduced over time.

Besides tax, government also provides discounts that are available according to family characteristics. Some of these grants include the FHOG, the home buyer concession scheme (HBCS) and the pensioner duty concession scheme (PDCS). Eligible income units can also receive rebates on rates. Even before the reform, home buyers could pay lower stamp duty if they were eligible for the HBCS. The HBCS applies to purchasers of a new home which has not been previously occupied and which is sold as a place of residence. Before 2014–2015, this concession was also available for an established home. The eligibility criteria for the HBCS include the home value, total gross income of all buyers, the buyers not having owned any other properties in the two years leading up to the transaction date, and a requirement for buyers to live in the home continuously for at least one year from the settlement date. In 2011–2012, home buyers who were eligible for the HBCS only had to pay $20 in stamp duty if the property value was lower than the lowest eligibility threshold. If it was higher than the lowest threshold but still below the upper threshold, they had to pay stamp duty based on a concession duty rate. The PDCS provides discounts with similar rules for eligible pensioners who purchase a new or established home. These concessions are applied in both our baseline and counterfactual models.

One concession or discount that is not directly related to stamp duty is the FHOG, which provides financial assistance to eligible people who are buying their first new or substantially renovated home. To be eligible for the FHOG, the total value of the property (home plus land) must be $750,000 or less and the purchasers need to move into the home within 12 months of the settlement date and live in the home as a principal place of residence for a continuous period of at least one year.

While stamp duty is a cost that needs to be paid up front, there are also costs that need to be paid after purchasing a house. General rates are levied on property owners according to the ACT government’s Rates Act 2004. In 2011–2012, general rates for residential properties were a combination of (i) a fixed charge (FC) of $555; and (ii) a valuation-based charge (or variable charge) calculated based on the latest three-year average of unimproved value of the parcel for a house (unit) (AUV) calculated for 2009, 2010 and 2011, with a rate-free AUV threshold of $16,500 and a rate of 0.2727%. From 2012, general rates become more progressive with the introduction of a number of tax brackets and increasing marginal tax rates. In 2012–2013, the FC remained unchanged ($555) and the AUV threshold of $16,500 was abolished. Instead of the previous single bracket, four tax brackets, along with increasing rates, were implemented.

The owner of a parcel of land can apply for a rates rebate if they are receiving a Commonwealth Government pension with entitlement to a Pensioner Concession Card, a Department of Veterans’ Affairs or a War Veteran’s pension. All eligible persons who have their names on the property title will receive a 50% rebate of their rates payment up to the value of the rebate cap (see the Rates Act 2004, section 64). This means that if there are two property owners and they both receive an eligible pension, the family will receive the full rebate. If there is only one person eligible for the pension, 50% of the rebate will be applied. The rebate cap was $481 in 2011–2012 and increased to $565 in 2012–2013 to cushion the impact of reform. This cap increased to $700 from 2015–2016 onwards.

Another cost incurred after purchasing a house in the ACT is land tax, but this is only charged, for residential properties, where the property is not the principal place of residence. Like the rates rules, the amount of land tax is made up of two components: an FC and a valuation charge based on AUV and its corresponding rates. According to the ACT Government’s Taxation Administration (Land Tax) Determination, the FC was not applied until the financial year of 2014–2015. The land tax rules were kept unchanged for the period from 2005–2006 to 2011–2012.

Another important component of the reform is that it tried to maintain overall revenue. We know how the rates change after the reform but we don’t know what they would have been before the reform (the counter-factual). Therefore, the 2012–2013 counterfactual general rates revenue is calculated based on the revenue target for that year in the absence of tax reform. The residential rates revenue target is then used to calculate the fixed charge and rating factor that would have applied in 2012–2013 for residential properties in the absence of tax reform, assuming a 50/50 revenue split between the fixed charge and rating factor elements.

3. Results

Following Pawson et al. (2020), the impact of the tax reform is assessed by comparing the decisions after tax reform and those for its counterfactual. This is mainly conducted by analysing whether the reform enables people to buy a house. Further analysis is also needed to determine whether any increase in house purchasing is occurring among various disadvantaged groups. Table 1 shows the impact of the reform on the number of properties purchased. This impact is an additional 2263 properties purchased over the period 2012–2018.

Table 1. Impact of tax reform on the number of residential properties purchased, ACT, 2012–2018.

	2012	2013	2014	2015	2016	2017	2018	Total
New tax policy
Meet income and deposit constraints	11,494	12,569	13,317	14,509	15,822	12,423	9634	89,768
Old tax policy
Meet income and deposit constraints	11,163	12,413	13,162	14,220	15,315	11,853	9379	87,505
Difference (impact of new policy)	331	156	155	289	507	570	255	2263

Source: Author calculations from microsimulation model.

The question is whether this estimated increase in buyers includes those experiencing disadvantage. To address this, results for some sub-groups of the population are presented: first-home buyers, low-income families, middle-income families, pensioners, and families with a female head of household. These groups are not mutually exclusive; for example, low-income renters can also be pensioner renters. The results in Table 2 show that the number of residential properties purchased by first-home buyers increased by 5.4%, which was mainly due to the stamp duty changes. Reducing stamp duty has therefore worked as a policy to increase purchases among first-home buyers. An increase in house purchasing is also seen for renters and for income units with a female head (based on the definition used in the survey). These female-headed families are usually single-parent families, a group with lower incomes, likely to have benefited from the reduced stamp duty. This means that the reform has, to some extent, fulfilled the goal of helping disadvantaged families.

Table 2. Impact of tax reform on the number of residential properties purchased by sub-group, ACT, 2012–2018.

	New tax policy – Number of purchases	Old tax policy – Number of purchases	Difference	Difference %
First home buyers	34,601	32,843	1758	5.4
Low-income homeowner	9491	9423	68	0.7
Low-income renter	4629	4500	129	2.9
Middle-income homeowner	11,726	11,638	88	0.8
Middle-income renter	11,169	10,830	339	3.1
Pensioner homeowner	5952	5887	65	1.1
Pensioner renter	2397	2355	42	1.8
Female-headed homeowner	20,246	20,060	186	0.9
Female-headed renter	12,309	11,708	601	5.1

Source: Author calculations from microsimulation model.

The reform may also have helped those that had savings. This can be seen by the impact for different income quintiles. Table 3 shows that the number of purchases increased the most for middle- and high-income quintiles, with increases of 3.6% and 2.8%, respectively. Nevertheless, income units in the lowest income quintile also increased their purchases of housing by 1.5%.

Table 3. Impact of tax reform on the number of residential properties purchased by income and wealth quintile, ACT, 2012–2018.

	New tax policy – Number of purchases	Old tax policy – Number of purchases	Difference	Difference %
Q1 of income	4376	4312	64	1.5
Q2 of income	9744	9611	133	1.4
Q3 of income	22,895	22,468	427	1.9
Q4 of income	25,668	24,766	902	3.6
Q5 of income	27,085	26,348	737	2.8
Q1 of wealth	92	92	0	0.0
Q2 of wealth	20,849	19,486	1,363	7.0
Q3 of wealth	21,697	21,318	379	1.8
Q4 of wealth	22,840	22,450	390	1.7
Q5 of wealth	24,290	24,159	131	0.5

Source: Author calculations from microsimulation model.

Given that home ownership is related more to wealth accumulation than income (Beer et al. 2011, Pawson et al. 2020), it is necessary to also assess the impact of the reform by wealth quintile. Results for the number of properties purchased with and without the tax reform by wealth quintile are also shown in Table 3. It can be seen that low-wealth households make very few purchases in either scenario, possibly due to the wealth test applied in the model. Although most of the purchases occur in wealth quintile 5, the largest increase in purchases is in the second lowest quintile. This indicates how the reform has helped spread homeownership across wealth quintiles, but unfortunately still not to the lowest wealth quintile. This is possibly due to a purchaser still needing to own a certain level of savings to buy.

There is a question mark over whether this tax reform has made the housing tax system more equitable. In general, low-income families spend the least on stamp duty and rates while higher income families pay more which supports the theory that lower stamp duty and higher rates is a progressive policy. However, lower income families are also spending more on stamp duty and rates as a proportion of their income, and this proportion is higher under the new tax policy. Under the new tax policy, families in the lowest income quintile are spending 4.69% of their income on rates and stamp duty, whereas under the old system this was 4.08%. Those in income quintiles 3 and 4 benefit from the new tax system, paying 4.64% and 4.37% of their income, down from 4.90% and 4.55%, respectively, in rates and stamp duty under this system.

The results above indicate that the tax reform was able to increase the level of homeownership in the ACT. In particular, the findings suggest that the reform allows disadvantaged or vulnerable families to purchase homes.

4. The Possible Impact of the Price Increase

Government support to own a house can become the source of house price inflation (Beer et al. 2011). Therefore, it is necessary to assess the impact of the reform on house price increases. The assessment using data in this period has failed to produce any significant impact as shown by Breunig et al. (2020). Using difference-in-difference econometric analysis, the coefficient that represents the impact of the reform is negative without any control variable. Adding population and income as control variables in addition to capital cities as a control group increases the impact of the reform on housing, but it is still statistically insignificant. The possible reason is that in the period 2011–2018, house price inflation (as measured by the Housing Price Index) in the ACT was lower than for other major cities such as Sydney, Melbourne and Brisbane, and also lower than the average for capital cities in Australia. Furthermore, the increase in rates as the stamp duty was phased out may reduce the demand for land and therefore land values and lower house prices. Despite this, the price increase scenarios are still introduced to check the stability of the result and gain a better understanding of the possible consequences of this price increase.

In our model, we applied two possible scenarios: an increase of 0.8% per year above the ABS House Price Index (HPI) and an increase of 2.1% per year above HPI, with HPI representing the amount of price increase under the old policy. To calculate this, for the new policy, the base house price increase was set at the HPI, and the adjustments were made by reducing the price increases for the old policy. The average reduction in stamp duty as a ratio of the average house price is 0.8%; this price change assumes that the reduction in stamp duty was absorbed in house prices. The 2.1% increase is based on the rate that is just below the annual HPI increase in Canberra between 2012 and 2018. Therefore, the 2.1% price increase scenario is an extreme or upper-bound case.

After taking these price increases into account, the income and deposit constraints mean that fewer houses will be purchased under the reform. This is shown in Table 4. It can be seen that because of the price increases incorporated in calculating the impact of the new policy, the number of properties purchased is slightly higher when a 0.8% price increase is assumed, although lower than when the price increase for the old policy was the same as for the new policy. The number of properties purchased was slightly lower with a 2.1% price increase. This means this level of price increase may more than offset any increase in purchases as a result of reduced stamp duty.

Table 4. Impact of tax reform on the number of residential properties purchased with price effect, ACT, 2012–2018.

	0.8% price effect
	2012	2013	2014	2015	2016	2017	2018	Total
New tax policy
Meet income and deposit constraints	11,494	12,569	13,317	14,509	15,822	12,423	9634	89,768
Old tax policy
Meet income and deposit constraints	11,364	12,507	13,290	14,322	15,588	12,180	9503	88,754
Difference (impact of new policy)	130	62	27	187	234	243	131	1014
	2.1% price effect
	2012	2013	2014	2015	2016	2017	2018	Total
New tax policy
Meet income and deposit constraints	11,494	12,569	13,317	14,509	15,822	12,423	9634	89,768
Old tax policy
Meet income and deposit constraints	11,505	12,803	13,621	14,817	15,840	12,244	9647	90,477
Difference (impact of new policy)	−11	−234	−304	−308	−18	179	−13	−709

Source: Author calculations from microsimulation model.

The important question is how this price increase affects vulnerable families. Table 5 shows that a price increase of 0.8% would still allow more first-home buyers to purchase homes although the increase in the number falls from 5.4% without the price increase to only 0.1%. The number of potential buyers among female-headed renter families would also still increase. There is also a very small reduction among existing homeowners. When a price increase of 2.1% is modelled, however, the capacity of most families to buy a house is reduced. In percentage terms, the impact of this price increase is strongest on those who already own a home; in absolute numbers, however, the estimates show that more middle- and low-income renters would be affected.

Table 5. Impact of tax reform on the number of residential properties purchased by sub-groups with price effect, ACT, 2012–2018.

	New tax policy – Number of purchases	Old tax policy – Number of purchases	Difference	Difference %
	0.8% price change
First home buyers	34,601	33,794	807	0.1
Low-income homeowner	9491	9498	−7	0.0
Low-income renter	4629	4554	75	0.0
Middle-income homeowner	11,726	11,687	39	0.2
Middle-income renter	11,169	11,066	103	0.1
Pensioner homeowner	5952	5958	−6	0.0
Pensioner renter	2397	2397	0	0.0
Female-headed homeowner	20,246	20,193	53	0.1
Female-headed renter	12,309	12,058	251	0.1
	2.1% price change
First home buyers	34,601	35,010	−409	−0.1
Low-income homeowner	9491	9658	−167	−0.4
Low-income renter	4629	4944	−315	−0.1
Middle-income homeowner	11,726	11,859	−133	−0.5
Middle-income renter	11,169	11,472	−303	−0.2
Pensioner homeowner	5952	6059	−107	−0.3
Pensioner renter	2397	2397	0	0.0
Female-headed homeowner	20,246	20,426	−180	−0.4
Female-headed renter	12,309	12,746	−437	−0.2

Source: Author calculations from microsimulation model.

In terms of income quintile, the price increases affect the low-income groups more, as shown in Table 6. House price increases mean that these groups are excluded from buying housing due to affordability issues, because they do not meet the deposit and repayment tests in the model. An increase of 0.8% means that the number of houses purchased by the low-income group decreases by 0.3%, and an increase of 2.1% means that the lowest three income quintiles reduce their purchases, with the reduction in purchases greatest in Q2, at 4.1%. High-income earners are not as affected, with a reduction in houses purchased of only 1.0% for Q4 and 0.2% for Q5.

Table 6. Impact of tax reform on the number of residential properties purchased by income quintile, ACT, 2012–2018.

	New tax policy – Number of purchases	Old tax policy – Number of purchases	Difference	Difference %
	0.8% price effect
Q1 of income	4376	4387	−11	−0.3
Q2 of income	9744	9665	79	0.8
Q3 of income	22,895	22,753	142	0.6
Q4 of income	25,668	25,268	400	1.6
Q5 of income	27,085	26,681	404	1.5
Q1 of wealth	92	92	0	0.0
Q2 of wealth	20,849	20,083	766	3.8
Q3 of wealth	21,697	21,488	209	1.0
Q4 of wealth	22,840	22,680	160	0.7
Q5 of wealth	24,290	24,411	−121	−0.5
	2.1% price effect
Q1 of income	4376	4442	−66	−1.5
Q2 of income	9744	10,160	−416	−4.1
Q3 of income	22,895	23,331	−436	−1.9
Q4 of income	25,668	25,410	258	1.0
Q5 of income	27,085	27,134	−49	−0.2
Q1 of wealth	92	185	−93	−50.3
Q2 of wealth	20,849	21,058	−209	−1.0
Q3 of wealth	21,697	21,650	47	0.2
Q4 of wealth	22,840	22,788	52	0.2
Q5 of wealth	24,290	24,796	−506	−2.0

Source: Author calculations from microsimulation model.

In terms of wealth quintile (shown in Table 6), an increase in house prices of 0.8% due to the new policy would still see an increase in purchasing for several quintiles, particularly for the second lowest quintile (Q2), while only the highest quintile experiences a reduction in purchases. This suggests that, for most groups, the benefit of the reduction in stamp duty in increasing purchases more than offsets the reduction in purchases as a result of the price increase. However, where the upper bound of the increase is 2.1%, purchasing capacity is severely limited for the lowest quintile. This underlines the importance of controlling the house price impact of such policy reform.

5. Implications

Assessing the new tax system in the ACT, our modelling shows that the policy can be expected to benefit potential homebuyers. This can be seen from the increasing number of purchases for most of the disadvantaged sub-groups modelled once stamp duty is reduced. This would mean a greater likelihood of people owning a home, creating greater financial security at old age and hence, higher wellbeing (Colic-Peisker et al. 2015, Pawson et al. 2020). Another implication is their financial position after buying a house. The results recognise the possibility of higher housing costs due to mortgage repayments and increasing rates. However, the impact is relatively small for ACT residents due to low-interest rates, relatively high rents and low rental vacancy rates, which are currently below 1% (ACT Government 2021). It is important to further investigate whether the stamp duty phase out reduces the rental vacancy rate as well as housing rent in the ACT.

Although there is no evidence that strongly supports the assumption that this tax reform increased house prices in Canberra, it is necessary to simulate the price increase to anticipate the estimated increasing demand that may not be matched by supply (Beer et al. 2011). When the price effect is modelled, a slightly different picture of the impact of the new tax policy emerges. If house prices increase by 0.8% annually due to the new policy, the number of total purchases halves, while under an annual 2.1% price increase the number of purchases falls rather than rises – so the reduced demand due to the increased price has more than offset any benefit of the stamp duty reduction. A price increase of 0.8% reduces the purchases of the lowest quintile by 0.3%, while the 2.1% price increase reduces the purchases of the second lowest income quintile the most, by 4.1%. Essentially a higher house price increase means that fewer low-income families can afford to buy under the new system, and the old system would have served this group better. First-home buyers and female-headed renter income units are the groups most affected if house prices increase, as their purchasing behaviours are mainly driven by the deposit requirements and income constraints.

Australia-wide, about 18% of the state and local government revenue comes from property stamp duty and this can be as high as 22% for Victoria in 2019–2020 (ABS 2021). Property stamp duties are often affected by market conditions, and are thus a volatile source of financial revenue. Therefore, many state authorities are considering broadening the tax base to ensure its stability and sustainability. Due to the COVID pandemic worldwide, dramatic market change has also created volatilities in transaction-based taxation both within and outside Australia (Hu et al. 2021, Liu and Su 2021, Qian et al. 2021).

Our analysis acknowledges the possibility of house prices increasing due to moving towards the land tax-based property tax system. While this might not have a direct negative consequence for government revenue, its distributional impact should not be ignored as the higher house prices will likely offset any benefit towards the low-income families that may benefit from the progressive stamp duty scheme at the moment. In addition, the government needs to be ready for opposition to the policy, especially from the homeowners that will bear the increase in rates (Morris 2021). This means that the reform might need to be managed by increasing the supply of new housing when necessary, including new land releases and building approvals as well as new amenities and infrastructure that shift the demand from inner cities area (Gurran and Phibbs 2013, Morris 2021).

One of the aspects of the ACT Government’s reform was that there was no choice. All owners paid higher rates, whether they had paid stamp duty or not. So those who had already paid stamp duty when they purchased their property payed higher rates, but did not benefit from the reduction in stamp duty until they sold and purchased another house. This affects low-income families who own their homes but are not in the market for a replacement. Many of these are low-income retirees.

The approach taken by the NSW Government has been to suggest a gradual implementation of a policy to reduce the use of stamp duty, giving buyers an option to pay stamp duty, or move onto a property tax system. Those who have already paid stamp duty when they purchased their property do not pay property tax. Further, when buying a property, the buyer can decide whether to pay stamp duty or go onto a property tax system. Once a decision to move to a property tax is made for a house, it stays in this system. This means the number of houses on a stamp duty system will slowly decrease over time.

This does create complexity over the long term, but also introduces choice so a buyer with some wealth in their own home but low income (e.g. retirees) can decide to pay the stamp duty to downsize and not be subject to a regular property tax. This impact of stamp duty on mobility especially among older people has been recognised in other countries such as in the UK and China (Wang 2012, Burgess and Quinio 2020) and has been considered as a financial incentive to purchase an environmentally friendly house (Goodchild and Walshaw 2011). This shows that stamp duty is not only a barrier to buying a house but it also prevents people from making an optimal choice, and is therefore inefficient for the economy. Nevertheless, its existence can also be the tool to incentivise people to make certain choices in their housing options.

6. Conclusion

This article aimed to review the impact of a tax reform of stamp duty and rates introduced by one jurisdiction in Australia (the ACT), using a microsimulation approach. The main measures used to assess the impact are the ability to purchase a house, especially for vulnerable and disadvantaged families.

After taking into account all available concessions and deferral programs, and ignoring the reform’s effect on house prices, the modelling demonstrates that the tax reform increases property turnover especially for most groups of vulnerable families in the ACT. The new tax system results in greater access to housing and increased purchases for low- to middle-income and mid-wealth families, although it has no impact on the ability of low-wealth families to purchase housing. If the reform is accompanied by a certain level of price increase, however, this could result in inequitable access to housing for first-home buyers, middle-income earners (Q2–Q3) and mid-wealth (Q2–Q3) families. Nevertheless, in one of the price increase scenarios modelled, first-home buyers and renters could still be expected to benefit from the new tax system.

One of the biggest concerns is whether any price increase has a greater impact on vulnerable families. During the analysis period of 2012–2018, the HPI was only increased by around 2% annually in the ACT, lower than around 3% in Australian capital cities. However, the HPI has increased by almost 4% annually in the ACT during 2019–2022, higher than around 3% annually in Australian capital cities. Furthermore, the price of new dwellings purchased increased by more than 5% annually in 2019–2022 after only increasing by around 1.5% annually during 2012–2018. The simulation shows that this high price increase will affect female-headed and low-income families more than other type of families. Furthermore, the impact will be greatest on those that have the lowest wealth accumulation.

Another dimension is the changing deposit requirement. Although not related to the analysis period, the Australian government introduced the First Home Guarantee initiative in 2022 to enable first-home buyers on low and middle incomes to purchase a home with lower deposit requirements, with the government covering the lender mortgage insurance and lowering the deposit requirements. This could improve the accessibility of the housing market for lower-income households. However, as the initiative has limited availability nationwide, accounting for less than 6%¹ of the total sales, it is likely to have a very limited impact in the ACT.

Whilst our analyses focus on a particular territory within Australia, the findings may have implications more broadly as they are closely associated with the underlying market responses that are likely to be similar across the country. Other states and regions that endeavoured such reform may find a similar distributional impact given the similar tax systems and the incentive to reduce reliance on the transaction-based taxes for properties. Internationally, there have been various modifications of the stamp duty and similar taxes. The reduction in stamp duty has increased housing options for people while other policies may be used to affect choice, for example, more environmentally friendly housing.

While the results outlined in this paper provide a useful analysis of the complex impact of housing tax reform, it is important to note that the model used here rests on the assumption that everyone would prefer to buy rather than rent when they can afford to do so. Although the literature suggests that this is a reasonable assumption, it is not necessarily always the case. However, one of the great advantages of microsimulation modelling is its ability to assess what-if scenarios: for example, being able to identify what the impact on low-income groups would if we changed rates in some particular way in the future, or added a certain concession.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 The Australian Government made 35,000 First Home Guarantee places available in 2022 financial year. The annual house and unit sales across Australia were around 598,000 in 2021 according to Corelogic estimates.