Final Report: Detailed Analysis
E4. Housing affordability
Infrastructure charges (sometimes called 'developer charges' or 'developer contributions') are fees levied on developers to compensate governments for providing facilities necessary for land development. The charges are often associated with basic infrastructure (such as local roads and water mains), but more recently this has sometimes been extended to include major headworks (arterial roads and pumping stations) and social infrastructure (parks and libraries).
Infrastructure charges are widely used by local government as well as some State governments, and are increasingly prevalent in other developed countries. There is limited information and few aggregate statistics relating to infrastructure charges in Australia. Chan et al. (2009) reported that in 2005–06, New South Wales councils collected $232 million and Victorian councils collected $454 million in charges.
In Australia, the practice of governments charging for infrastructure has been becoming more prevalent since the 1980s. This reflects increasing demand for infrastructure and fiscal constraints on local governments, but also a policy shift towards using economic instruments to allocate infrastructure and influence development decisions (Chan et al. 2009).
In principle, efficient provision of infrastructure would be encouraged where its users pay for the construction of infrastructure that would be avoidable (that is, not needed) if the development did not proceed. By levying infrastructure charges that reflect these costs, State and local governments provide signals to develop housing in ways and places of greatest value. The cost of infrastructure increases directly with distance from essential headworks and inversely with the density of development (Slack 2002). To the extent that a developer can respond to these costs, for example, by choosing to build closer to an existing development or by increasing the density of housing, charging the developer can improve housing supply.
Indeed, in the absence of pricing, developers build without regard to such costs, and governments are more likely to rely on other policy instruments, such as planning regulations, to limit the budget costs of infrastructure associated with housing developments. The absence of effective infrastructure pricing increases the need for development regulations.
In practice, infrastructure charges have a number of problems.
First, infrastructure charges can sometimes be used to raise tax revenue, rather than focusing on providing efficient user charging. Where the charge exceeds the cost of providing infrastructure, it acts like a tax and can discourage development. This is more likely to occur where the size of the charge is not set relative to the cost of infrastructure but the developer's capacity to pay. In these cases, the charges may attempt to capture part of the increase in value resulting from the provision of infrastructure or from changes in zoning, that is, to impose a betterment tax (see Box E4–2). However, the benefit to the developer is difficult to determine, and attempting to set charges on this basis can lead to negotiations that are protracted and non-transparent. This can slow down development processes and result in payments that are not effective as prices for infrastructure. In general, infrastructure charges will operate more effectively if they are set to reflect the cost of infrastructure, not to tax the profit of development.
A particular form of tax used when land is re-zoned for alternative use is a 'betterment tax' which attempts to capture some of the increase in land value. Betterment taxes are not infrastructure charges since the objective is to tax economic rent, although sometimes the revenues are hypothecated (that is, earmarked) to infrastructure provision.
In concept, betterment taxes are attractive since they aim to tax the economic rent from land rezoning that would otherwise accrue to the landowner. However, in practice, betterment taxes can increase the uncertainty associated with land development. To operate effectively, betterment taxes need to isolate the increase in value attributable to the zoning decision or the building of infrastructure from general land price increases at the local level. This is often difficult since the value of land will move in anticipation of a change in re-zoning. Sometimes this can occur many years before the re-zoning. Betterment taxes may be applied on an ad hoc basis and the rate of the betterment tax is sometimes left to discussions between developers and government as part of the planning approval processes, rather than being set in a transparent manner. Betterment taxation can involve lengthy disputes as, by setting the tax conditions, the dispute is really about how to share the economic rent.
Additionally, having a betterment tax in place may encourage governments to create economic rent through additional zoning restrictions or delays in land release, in order to raise more revenue. Where zoning is used in such a manner, it is likely to stop land being devoted to its most productive use — at least in the short run. A land tax applied to all types of land (see Section C2 Land tax and conveyance stamp duty), is likely to encourage governments to allow land to be used for its most productive use as this will increase the value of the land (and hence increase the revenue raised from land tax).
Second, infrastructure charges can be complex and costly to levy. Ideally, each individual development would be assessed for the avoidable costs of infrastructure, which can be different for a similar item in different developments (McNeill & Dollery 2003). There is therefore a trade-off between the accuracy of user charging and its administrative feasibility.
Some infrastructure items nearly always have benefits that are limited to those residing in the development, including local roads, drainage, stormwater, utilities provision, and land for local open space. Charges should generally be imposed for this type of infrastructure since it directly connects new properties to wider networks and would serve little purpose in the absence of the development.
However, it is much less clear whether charges should be imposed for other forms of infrastructure, including community facilities such as schools, libraries and child care, regional improvements such as parks, open space and capital repairs, public transport capital improvements, regional road improvements and conservation of natural resources. When used in such circumstances, the charges should reflect only the additional costs that the development imposes on society, not the total cost. In many cases, governments would need to expand infrastructure, such as libraries, whether or not it is located in a new development. In such cases there should be no charge to the developer. Similarly, some forms of infrastructure that have network effects, such as electricity generation, should be paid for by all users. The higher prices necessary to recover capital costs serve a useful function in rationing such services to the highest value users (whether in a new development or not).
The scope of infrastructure charges varies significantly between States and internationally (see Table E4–2).
Table E4–2: Public infrastructure eligible for mandatory contributions (excluding basic infrastructure)
|Child care centres||(e)||(f)|
- Relates to infrastructure eligible for local infrastructure contributions mandated under s.94 to s.94EH of the NSW Environmental Planning and Assessment Act 1979 and includes proposed changes announced by the NSW Premier on 12 October 2007. Under these reforms, mandatory contributions will be limited to the infrastructure and land directly required to support land developments.
- Contributions generated by the ACT change of use charge flow into consolidated revenue and can be used to finance any government objective.
- Dedication of land only.
- Within the sub-division.
- Restricted to infrastructure that services the development site or precinct.
- In Victoria, contributions for community infrastructure are capped at $900.
Source: Adapted from Chan et al. 2009.
There is some community concern about the impact that infrastructure charges may have on the affordability of housing. Several submissions to the Review have proposed using tax financing for new infrastructure to improve the affordability of housing. Deliberately charging below cost for new infrastructure would effectively involve ratepayers providing a subsidy for the provision of new housing.
Subsidised infrastructure is a high-cost way to lower house prices because it encourages the delivery of infrastructure to areas where it is of relatively low value. Further, developers have less incentive to build housing that uses infrastructure efficiently (for example, by building more high-density housing).
More importantly, in the face of restricted land supply, an infrastructure subsidy is unlikely to achieve the intended goal of lower house prices. Instead, the infrastructure subsidy is likely to be capitalised into higher land prices for sellers (see Chart E4–6 Panel B). Where new developments are restricted in supply, their prices will be set by the availability of nearby existing dwellings.17 Sellers are likely capture most of the subsidy as they face high demand for their land. Similarly, if the subsidy is removed, sellers would be unlikely to convince buyers to purchase fringe dwellings at prices exceeding those in neighbouring established areas. When land supply is already restricted, developer charges are borne by the original land holder by reducing the above-normal return (economic rent) they would otherwise receive when selling their land. This is currently more likely to be the case in Australia, suggesting that infrastructure charges are unlikely to affect housing affordability substantially (Productivity Commission 2004).
However, where infrastructure charges are implemented poorly or are designed to operate as taxes, they can discourage housing supply and contribute to higher house prices.
Difficulties in administering infrastructure charges can increase uncertainty, potentially deterring investment. For example, charges on proposed developments are sometimes used to offset local objections, such as community concerns about traffic congestion or overcrowding of public transport. In some cases, councils or State governments have responded to these concerns by imposing additional charges on proposed infill development to upgrade local infrastructure, such as railway stations and pedestrian bridges. When development approval is contingent on development charges of uncertain size, this can also add risk to projects and affect their viability.
Applying infrastructure charges through use of simple flat prices that do not well approximate actual avoidable costs can sometimes reduce housing supply. For example, where charges are levied at a flat rate per dwelling, high-density developments are likely to face higher prices for the infrastructure they require, compared to lower density developments.
Where developer charges are set in an ad hoc fashion or are subject to unexpected changes, they can create uncertainty around new developments. If infrastructure charges are increased after a developer has bought land from its original owner, they cannot be factored in to the price previously paid for the raw land. In this case, the charge would lower the expected return from the development. In addition, general uncertainty about charging is likely to discourage development activity, which could reduce the overall supply of housing and increase the price of housing.
Box E4–3: Who pays infrastructure charges?
Who bears the burden of developer charges depends on the relative elasticity of the demand and supply of land (Neutze 1997). Following Wood et al. (forthcoming), Chart E4-6 illustrates the impact of infrastructure charges on the price and allocation of land, depending on whether land for housing is freely available or restricted.
Chart E4–6: Incidence of infrastructure and land supply
Panel A: Unrestricted supply of housing land
Panel B: Restricted supply of housing land
The amount of total land (OQ) is divided between demand for housing land (Dh) and agriculture, which is valued at a fixed price (OA). To make land useable for housing, infrastructure worth AI must be added to the land.
If there are no restrictions on the use of land for housing at the fringe of a city and infrastructure is provided free, Qs land is used for housing (Panel A). With infrastructure for new housing effectively subsidised, the price of land for housing is equal to the value of agriculture (OA). Where there is a charge imposed for the infrastructure, the amount of land used for housing will fall to Qi. The cost of land used for housing at the fringe increases to OI, so that it equals the cost of agriculture plus the value of infrastructure. In this example, subsidised infrastructure can reduce the cost of housing because additional land can be acquired and converted to residential use at a price OA. However, this is not a good outcome from the viewpoint of efficient resource allocation: if the full social cost of housing is reflected in appropriate infrastructure charges demand for housing is Qi.
In contrast, if the supply of land available at the fringe is limited to the amount Qr due to zoning or planning restrictions (as in Panel B), free infrastructure has no impact on housing supply. The restriction in land supply means that prices at the fringe (OP) are higher than the value a marginal agricultural user would be willing to pay. Landholders receive an 'economic rent' (AP) when developing their land for housing. The price is determined by the demand for land in relation to the constrained supply. An infrastructure charge (IA) levied on land holders who are developing their land for housing reduces the size of this rent (from AP to IP), without affecting the price of housing. Only if the infrastructure charge exceeded AP would it raise house prices.
This is a stylistic comparison of two extreme cases, where land for housing is either fully elastic or fully inelastic. The real situation is likely to fall somewhere between these two extremes. That said, the value of agricultural land at the fringe of cities generally exceeds its opportunity cost in agricultural production, reflecting restrictions on the supply of land for housing.
Infrastructure charges can be an effective way of encouraging the efficient provision of infrastructure to areas where it is of greatest value and of improving housing supply. Charging for infrastructure may be a more effective means of allocating resources than regulating land release.
Where land supply is constrained, well-designed infrastructure charges are more likely to be factored in to the price that developers pay for raw land, than to increase the price of housing in the development where the charge is levied. However, where infrastructure charges are poorly administered — particularly where they are complex, non-transparent or set too high — they can discourage investment in housing, which can lower the overall supply of housing and raise its price.
COAG should review infrastructure charges (sometimes called developer charges) to ensure they appropriately price infrastructure provided in housing developments. In particular, the review should establish practical means to ensure that these charges are set appropriately to reflect the avoidable costs of development, necessary steps to improve the transparency of charging and any consequential reductions in regulations.
In general, improving the use of infrastructure charges and relying on general land taxation (rather than betterment taxes) is likely to reduce the need for regulations and planning, which can restrict housing supply.
For infrastructure charges, the Review is not in a position to be definitive about which costs should be met by developers, and which costs should be met by governments. However, there is a trade-off between relying too much on a high-level principle and difficulties of applying it in practice. One way of resolving which charges should be borne by developers and which charges should be borne by governments would be to establish principles that set out the funding responsibility for each type of infrastructure. However, the Review has not recommended specific pricing guidelines for future infrastructure charges. This is an issue that requires extensive consultation.
Further, making the charges and the process for setting them public should help ensure they are set close to the right level.
Where charges are transparent they are likely to encourage more efficient provision of infrastructure. If developers understand how charges are determined, they will respond by minimising the costs they face, which is the desired outcome. Setting charges publicly and in advance enables developers to make more effective plans about where to develop and can provide greater certainty to the process. Non-transparent and inscrutable infrastructure charging processes increase risk to developers and reduce valuable infrastructure investment.
Transparency also serves an important accountability role. The Review has received several submissions from housing industry organisations concerned with the lack of transparency in price setting. Because only a small number of taxpayers pay such charges there may be an incentive for governments to rely on infrastructure charges to fund services beyond infrastructure where general taxation would be a more efficient means of raising revenue. Public disclosure may encourage more consistent principles and their application in practice and reduce the chances of under- or over cost-recovery.
17 In contrast, where land supply is deregulated, subsidised infrastructure is likely to reduce the price of housing in fringe areas, as the price of land can be reduced through competition among buyers (see Chart E4-6: Panel A in Box E4-3). However, this reduction in the price of housing occurs because infrastructure is being supplied in areas where it is of low value.
Next Page – E5: Alcohol taxation >>
<< Previous Page – E4–4: The effect of non-tax policies on housing supply