Skip to main content Skip to navigation
Australian Government - GST Distribution Review

Chapter 6 - Greater stability of GST shares and methodology improvements

Along with the need to improve the frameworks for governance, communication and transparency of process covered in Chapters 4 and 5, a number of other issues have been put to the Panel, or have arisen indirectly out of our inquiries. Apart from mining (which is dealt with in Chapter 7), the most significant of these relate to concerns about the instability of GST shares as a side effect of decisions of the Commonwealth to fund large infrastructure projects in various States, and methodology changes and data revisions by the CGC.

Although methodology issues are rightly left to the wisdom of the CGC, the Panel has considered that two issues raised with us are worthy of being highlighted, for the CGC’s future consideration.

6.1 The treatment of Commonwealth capital payments

This section discusses a rationale and approach for the future HFE treatment of capital payments from the Commonwealth, and makes recommendations relating to capital payments for road and rail based transport infrastructure in particular.

How does the CGC treat capital payments?

As explained in the Panel’s first interim report1, Commonwealth National Partnership Payments (NPPs) for capital purposes are generally included in the HFE system at present, and therefore affect relativities. Capital payments are treated in the same way as NPPs for recurrent purposes, fully included in the year of receipt (when that year becomes an assessment year).

Commonwealth payments included in the HFE system are assessed for revenue needs and the corresponding expense needs are also assessed. Capital payments and capital expenditure needs are treated in a consistent fashion, both being assessed ‘upfront’.

The revenue side effect on the GST revenue distribution for included capital payments is easy to quantify, being the difference between an EPC distribution and the actual distribution of the payments (ignoring averaging and lag effects). On the other hand, primarily due to the inter-related nature of the capital assessments, the expenditure side effects can be problematic to quantify. Notwithstanding this difficulty, the main influence on capital needs is relative population growth.2

Broadly speaking therefore, when any State receives a capital payment from the Commonwealth, the amount it receives above its population share is ‘equalised away’ over time. While this might sound drastic at first, in context it has tended not to matter too much in the past — as prior to 2010 the equalisation period extended over 5 years and, over the very long term, all States get close to their population shares of Commonwealth capital payments.

State & other views from submissions on the interim reports

The Panel’s first interim report said that:

The Panel sees merit in:

- equalising all capital payments … over a longer time frame to recognise the lasting nature of the asset being funded and reduce the impact of the payments on GST shares in any one year, or

- equalising most capital payments, but excluding capital payments for nationally significant projects.

If [these] capital payments … are to be excluded, a process for identifying … the projects will [be needed].3

Summary of State views

All States addressed these issues in their final submissions. New South Wales considered them to be largely methodological, saying that any change to the treatment of Commonwealth payments would be a ‘piecemeal solution’ to a broader problem. Other State views on the issues raised by the Panel are summarised below.

Equalising capital payments over a longer period of time

Queensland and South Australia do not support equalising over a longer period because to do so would be inconsistent with the upfront approach to capital. The Australian Capital Territory supports equalising capital payments over time, but said this would be more consistent with an approach that recognises the use of infrastructure over its life. Tasmania and the Northern Territory do not rule out equalising over a longer period, although the Northern Territory does not support equalising over more than 10 years.

Excluding Commonwealth capital payments for nationally significant projects

All States support the general principle of including Commonwealth payments for capital purposes in the equalisation process. Victoria, Western Australia and the Australian Capital Territory did not support the introduction of a national significance principle. Victoria and the Australian Capital Territory said there should be consistent treatment of road and rail payments and they would prefer including 100 per cent of both. While supporting the general principle of inclusion, Queensland, South Australia and the Northern Territory see merit in a national significance principle for identifying Commonwealth capital payments to be excluded. South Australia notes the 50 per cent exclusion precedent for national network roads with a role in interstate freight.

On the question of how nationally significant projects should be identified, South Australia and the Northern Territory say that the CGC should be responsible for identifying eligible projects. Queensland said that the CGC was not the appropriate body and that identifying nationally significant projects should be a role for the Commonwealth Treasurer. Queensland and the Northern Territory say that the Infrastructure Australia priority list is not suitable for identifying relevant projects.

Other views

The CGC notes that there could be ‘double-dipping’ where a State receives a Commonwealth capital payment that is excluded from the equalisation process, but gets GST revenue to meet the related expenditure needs.

The Business Council of Australia puts the emphasis on payments being made on a consistent and transparent basis that is neutral in its treatment of different modes — for example, neutral between road and rail. Other commentators4 say that Commonwealth payments for nationally significant projects should be excluded from the equalisation process, but provide no guidance on how to identify such projects.

As this important matter is not a simple choice between clear alternatives, the Panel has felt it necessary to work through a number of issues before coming to its position. These are set out below.

How are Commonwealth transport infrastructure payments allocated?

Commonwealth capital funding for infrastructure projects is made through project National Partnership Payments (NPPs). In the Commonwealth Government Budget Paper 3, payments are classified to sectors, such as Health, Education, Infrastructure and so on. Economic infrastructure payments (including for roads and rail) tend to be classified to Infrastructure payments, while other social infrastructure payments (such as for hospitals and schools) are generally classified to their relevant sector.

Table 6.1 shows the proportion of total transport infrastructure payments for each State currently identified by the Commonwealth. For comparison, the table also includes each State’s share of total population as at 2012-13.

Table 6.1: Transport Infrastructure payments, 2008-09 to 2015-16

  % % % % % % % % $m
2008-09 36 13 34 8 6 2 0 2 5,119
2009-10 27 13 37 7 10 3 0 3 4,961
2010-11 40 19 18 8 9 4 1 2 3,050
2011-12 35 19 27 7 10 1 1 2 7,006
2012-13 29 23 22 17 4 2 0 2 3,284
2013-14 31 30 20 13 3 2 1 2 4,441
2014-15 27 41 17 10 1 0 4 1 1,927
2015-16 15 39 17 10 17 0 2 1 1,455
Total 32 21 26 9 7 2 1 2 31,241
Pop (12-13) 32 25 20 10 7 2 2 1  

Note: Additional funding to New South Wales for the Pacific Highway not included.
Source: 2008-09 to 2010-11, Commonwealth Government Final Budget Outcome; 2011-12 to 2014-15, Commonwealth Government 2012-13 Budget Paper 3; CGC, 2012 Update Report.

Table 6.1 shows that, in any given year, States’ shares of Commonwealth transport infrastructure payments vary from their population shares, sometimes substantially. This difference leads to GST redistributions. Ignoring expenditure effects, if these payments were excluded from the equalisation process, many States would be largely unaffected, but States receiving less than their population share, for example Victoria, would receive less GST revenue, and States receiving more than their population share, for example Queensland, would receive more GST revenue.

Identifying nationally significant projects

The Panel’s first interim report noted the difficulties in identifying payments for nationally significant projects and recognising all relevant factors that influence State expenditure on them.

    … [A] satisfactory process to identify eligible [nationally significant] projects does not currently exist.

    … the current assessments do not recognise all the complex and multi-faceted factors that influence State infrastructure spending.5

These difficulties arise partly due to programs with similar aims being controlled or administered by different bodies. For example, the Nation Building Program (used to fund national and inter-regional land transport corridors of critical importance to national and regional growth) is administered by the Department of Infrastructure and Transport, while the allocation of the Building Australia Fund (similarly used to fund critical infrastructure in the transport sector, along with the communications, water and energy sectors) is primarily based upon the Infrastructure Australia priority list.

Infrastructure Australia

Infrastructure Australia’s priority list identifies projects that would help solve a nationally significant infrastructure problem and deliver economic benefits. In its 2012 report, Infrastructure Australia’s priority list included $76.5 billion of projects that it says will make a valuable contribution to addressing nationally significant issues. However, these projects cover a range of completeness and readiness to proceed. Projects progress through four stages — Early Stage ($48.1 billion), Real Potential ($10.1 billion), Threshold ($6.6 billion) and Ready to Proceed ($11.8 billion).

Early Stage projects may fail to proceed or change substantially by the time they progress through to the Ready to Proceed stage, so that not all projects on the list could be considered to have definite national benefits. However, if Infrastructure Australia’s priority list were in the future to become the primary determinant for the Commonwealth’s funding of infrastructure, it could possibly be used to identify nationally significant projects.

National network roads

For transport infrastructure (and economic infrastructure more broadly) — as compared with social infrastructure — there is greater uncertainty around both the balance of national/State benefits and how well the CGC can capture ‘needs’ relating to the infrastructure. While individual school buildings, for example, have a clear and direct benefit to the jurisdiction in which they are located and a more intangible benefit to national education, almost all major transport infrastructure facilitates national economic growth and productivity directly to some degree.

In its 2010 review report, in relation to Commonwealth payments for national network roads, the CGC said:

We believe that part of the Commonwealth support for these [national network] roads and the consequent investment is influenced by Commonwealth considerations which are not captured in our State based disability measures. These include the need to develop an efficient national transport network to facilitate national economic growth and productivity gains in the long-term.6

In the absence of a better basis for determining the proportions of national and State benefit, the CGC decided that 50 per cent of national network road payments (and related expenditure) should not affect relativities to recognise the clear dual purpose. After consideration, the CGC did not make any similar findings for other Commonwealth payments for transport infrastructure, in particular for rail based infrastructure.

Conclusions on the treatment of Commonwealth capital payments

In the Panel’s view, while there may be some merit in treating all nationally significant projects in a special, but consistent, fashion under the equalisation system, we do not believe there is currently a comprehensive way to identify these ‘nationally significant’ projects. In practice, as the examples of problems seem to relate to the treatment of transport infrastructure, this raises the question of the appropriateness of the differences in treatment of Commonwealth funding for road and rail based projects.

The Panel’s main concern in this regard is that the CGC’s current approach does not recognise the inter-related nature of national transport infrastructure and the long-term view required to ensure appropriate investment decisions. Different equalisation treatment of road and rail infrastructure has the potential to distort State decisions about what transport projects to undertake and when. Investing in rail infrastructure, especially in Australia’s major cities, would allow more efficient use of the existing national road network, facilitating the movement of road-based freight, including to major capital city ports.

By treating Commonwealth funding for national network roads and rail based infrastructure in the same way, the equalisation system would be less likely to influence State decisions about what transport projects will proceed and when. This reasoning would indicate consistent treatment, either both fully included, or both excluded to the same extent. On balance, the Panel has favoured excluding both to the same extent, as full inclusion would still leave an unnecessary degree of lumpiness after equalisation.

There is no clear way of determining precisely the appropriate proportion of any payment that relates to its national benefits (or non-State needs). For some payments it may be greater than 50 per cent, in some cases it may be less than 50 per cent. However, similarly to the position adopted by the CGC, the Panel’s view is that, recognising the clear dual purpose and having regard to simplicity, a uniform 50 per cent of identified payments not affecting relativities seems suitable.

The Panel notes that, if 50 per cent of national network road and rail based payments did not affect relativities, this would substantially reduce volatility in GST shares arising from the lumpy nature of these payments. This approach would therefore negate any further need for the HFE system to include these payments over longer time frames.

The Panel has also noted that, as significant rail payments to States commenced in 2009-10, to ensure States that received payments since this time are not disadvantaged, the change in treatment should commence from the CGC’s 2013 Update.

None of the above is intended to exclude the possibility that the CGC, through its usual processes, might decide that any Commonwealth payment for transport infrastructure other than for national network roads or rail might have national significance. For example, a road to facilitate access to a remotely located port might be considered to have national significance. In such a case the CGC may well decide to assess these payments similarly to national network road and rail payments.

Recommendation 6.1

In recognition of the inter-related nature of transport networks and the national benefits that accrue from increasing the efficiency of these integrated transport networks, the CGC should identify all Commonwealth payments relating to national network road infrastructure and rail based transport infrastructure.

All identified payments should affect the relativities on a 50 per cent basis, to recognise their dual national/State purpose. To ensure that States that have previously received rail based transport payments are not disadvantaged, this change in treatment should apply from the CGC’s 2013 Update.

6.2 Data revisions

The core of the CGC’s assessment of States’ fiscal capacities relies on data. The CGC uses data from a range of sources, such as the Australian Bureau of Statistics (ABS), State provided data, and data from third party sources, such as the Australian Institute of Health and Welfare (AIHW).

The CGC seeks to use the data that is most recent and relevant to the assessment year in question, which sometimes requires revisions to data used in previous years. Previously used data may be subject to revision for a range of reasons, from the fact that more recent relevant data has become available, or to correct errors subsequently discovered, or following from changes to statistical collection methods.

The Panel’s main area of concern is with revisions that cause undue volatility in GST shares. Revisions that cause the greatest volatility generally occur where data are revised for more than one of the three assessment years. This tends to occur when using data sets that are not produced or released on an annual basis. A recent example of this was in the 2011 Update where new data from the ABS Survey of Education and Training (SET) led to changes in the assessment of interstate wages, resulting in much larger revision effects (changes to relativities) than is usual in an Update.7

How are data revisions dealt with?

Annual data

The CGC uses a range of data from various providers (predominantly the ABS) collected on an annual basis, that are subject to revision from time to time.

ABS revisions occur where there have been improvements in reporting of survey results, additional administrative data are provided to the ABS, through changes in statistical and/or survey methods, or international standards or classifications. In most cases the CGC will use any revised ABS data available as that will constitute the most up to date and relevant data for the CGC’s purposes. In determining whether to use revised data, the CGC investigates the reasons for the revisions and establishes that the data remain the most relevant to its task.

However, in many instances States are the only source of data for the CGC to make its assessments. Some State data are provided directly to the CGC, through annual data collection processes as part of each update of the relativities. These are reviewed by the CGC in an effort to ensure their fitness for purpose, that is, the data are comparable with that provided by other States, and are reliable (such as being consistent with known trends in the activity to which the data relate). The main driver of revisions to State data is the availability of later administrative data. For example, as revenue compliance tasks are completed more accurate data becomes available. On occasion, States may identify errors in the data previously supplied to the CGC, and revise the data to be incorporated by the CGC into its assessments.

In some cases data may be collected and published annually, albeit with a lag.8 For example, the AIHW morbidity data set is published annually, but with a two year lag. The CGC applies the most recently available data to more recent years, which are then revised once the data set relating to that year becomes available. The resulting revisions can lead to changes in State’s GST shares.

Non-Annual data

Some data used by the CGC are not produced annually, but less frequently. For example, the main sources of non-annual ABS data used by the CGC are Census data, collected every five years. While Census data themselves are not generally subject to revision, new Census data can lead to revisions to population estimates between Census years, thus affecting CGC assessments and possibly resulting in changed GST shares.

Similarly, for other non-annual data sets used by the CGC, when more recent data become available, the CGC will generally revise its estimates for years between the actual years to which the data relate. This was the issue in the 2011 Update, with the CGC revising its estimates for relative wage levels across States for the four years in between two SET collections, leading to larger than usual changes to GST shares.

Accuracy and volatility

The incorporation of more recent (and more accurate) data, leading to revisions to earlier CGC estimates and consequently changes to GST shares, increases the volatility of GST shares. Currently, in its assessments the CGC aims to reflect State capacities as accurately as possible in earlier assessment years, without regard to any consequent volatility of GST shares. The Panel has considered ways that might lead to a reduction in GST share volatility without unduly affecting the accuracy of the CGC’s assessments. The options considered by the Panel included:

  • freezing relativities
  • using the latest non-annual data for the year it applies to and for future years
  • using the latest non-annual data for the final assessment year only.

Freezing relativities would mean that no data would be revised, regardless of whether errors were identified and whether the data set was annual or non-annual. While this approach would produce the most stable GST distribution between methodology reviews, it has several drawbacks. It would least reflect State capacities and could lead to greater changes in GST shares at methodology reviews. It would risk encouraging States to be conservative in their own interest when supplying data from their own sources. The current approach more closely reflects State capacities, but leads to the greatest volatility in GST shares when errors are identified and corrected.

Using the latest non-annual data for the year it applies to and for future years would potentially reduce volatility in GST shares compared with the current situation, but may still result in revisions leading to significant changes in GST shares where there is a lag between the release of the data and the year to which it relates. For example, if data relating to the 2010-11 year were not available until after February 2012, these data would not be applied to the 2010-11 year when that year first becomes an assessment year. The following year these data would be applied to two of the three assessment years, 2010-11 (as revisions) and to 2011-12 (as the most recently available data).

Although technically the Government Finance Statistics (GFS) data pertaining to the most recent assessment year are not available until after the CGC has completed its update processes, the work the CGC has undertaken with States to obtain their Uniform Presentation Framework (UPF) data is considered sufficiently robust that reliable estimates are available for the most recent assessment year. Any revisions to these data included in the actual ABS release reflect legitimate corrections of errors and misallocations and should be reflected in the assessments.

On balance, the Panel considers that revising earlier assessment year estimates using lagged annual data or revised inter-survey interpolation for non-annual data sets can produce undue volatility in GST shares, beyond that reflecting changes in States’ circumstances. A better balance between accuracy and volatility would be found by:

  • for annual data sets (including ABS GFS data) — allow revisions as per currently
  • for all non-annual data sets and annual data sets with a lagged release — allow new data to only inform changes in States’ circumstances (the most recent assessment year) and not be used to revise previous estimates of earlier inter-survey years.

Two examples show how new data would be incorporated into the assessments using this approach. The first is for a lagged annual data set, for example the AIHW morbidity data set. In this case the newly available data would be included for the most recent assessment year (2011-12 in the 2013 Update) only, and not be used to revise the preceding year (to which it relates). Over time, each annual data set will have its full effect on relativities, albeit lagged. The second example is for population estimates, based upon Census data. The new 2011 Census data estimates would only be applied to the 2011-12 year, while the previous estimates extrapolated from the 2006 Census estimates would not be revised by any new inter-census year interpolation.9

Recommendation 6.2

Where data are updated or released annually with a lag, or updated or released less frequently than annually, the CGC should allow the newly available data to only inform changes in States’ circumstances in the most recent assessment year and not be used to revise previous estimates of earlier inter-survey years.

6.3 A simplified and integrated assessment framework

The Panel understands that the introduction of the 2010 Review capital assessment, emphasising ‘up front’ needs of high population growth States, was a contentious matter in the 2010 Review. While some States regard the capital assessment as remaining among the more complex of assessments (with South Australia and the Australian Capital Territory in particular favouring further simplification) this view is not universal, with some States regarding the assessment as relatively straightforward.

This section outlines a proposal for a simplified and integrated assessment framework for equalisation. The Panel considers this high level methodology change could improve simplicity, transparency and stability while addressing concerns about the treatment of subsidised public trading enterprises (PTEs), for example, public transport and social housing PTEs, in the current framework.

While acknowledging that the CGC is the appropriate body to consider the feasibility of this approach, the Panel suggests that when conducting its next methodology review, the CGC include the concept of a simplified and integrated assessment framework as one of the issues to focus on.

Why consider an alternative framework?

The current framework

HFE outcomes are derived in the context of a representative State budget. The representative State budget is used to calculate per capita revenues and expenditures used in assessing State fiscal capacities.10 The scope of the representative, or adjusted, budget has changed over time reflecting, amongst other things, the move to accrual accounting and changes to the treatment of capital needs and PTEs.

In the 2010 Review the CGC moved from an operating result framework to a net lending statement framework. The main motivation for the change was an encouragement from Heads of Treasury (HoTs) to recognise State capital needs in a simpler and more direct way than via an assessment related to net interest expense.11 In the Review, the CGC also completed the transition to a subsidy based approach to recognising needs for PTEs. Under this approach, needs related to State budget support for PTEs are assessed through the grants and subsidies paid to PTEs.

Problems arising from the current framework

The Panel considers that the changes to the capital assessment in the 2010 Review — including the population growth needs assessment — were a positive step forward. In adopting this approach, the CGC said it preferred an approach to capital that recognises the financial consequences of population growth when needs arise. There are three aspects to the current capital assessment: depreciation, net investment (which is gross fixed capital expenditure less depreciation) and net borrowing/lending. While the 2010 changes were intended to simplify the assessment of State capital needs there are aspects of the net investment assessment that are still unnecessarily complex and volatile. In addition, the CGC’s decision to ‘equalise States’ net financial worth’ in the net lending assessment12 imposes a constraint on the recognition of capital needs for subsidised PTEs.

Net investment assessment13

According to the CGC, the net investment assessment aims to provide States with capacity to fund the investment in ‘new’ infrastructure required to maintain an assessed level of infrastructure. It does this by recognising two effects. The first is that of differential population growth on the stock of physical assets and the second is differences between the States in the quantity of infrastructure required to provide services and its cost.14 There are issues with the net investment assessment:

  • The assessment of net investment and depreciation appears to involve a double count because ‘new’ investment in one year gives rise to depreciation expenses that are then assessed over the life of the assets.
  • While in concept the net investment assessment is simple, in practice, applying current period capital stock disabilities to the large stock of physical assets estimates gives rise to a volatile element. The CGC took steps in the 2010 Review to reduce the amount of volatility by adopting a three year moving average for the capital stock disability factor. Nevertheless, small changes in disabilities continue to have an effect on assessment outcomes that is difficult to predict and explain, and resulting in unnecessary volatility in GST shares.
Net lending assessment15

Through the net lending assessment, the CGC seeks to recognise the impact of differential population growth on a States’ net financial worth (NFW) and therefore their capacity to earn revenue in the form of interest and dividends. Since the CGC considered the equalisation of NFW central to the measurement of equalisation, any assessment of State capital subsidies to PTEs could not recognise differential needs other than the impact of total State population growth.16 Therefore, capital needs for subsidised PTEs may not be fully recognised. This is notwithstanding that the CGC treats most Commonwealth capital grants to PTEs (including transport PTEs) by inclusion.

Conclusions on a simplified and integrated assessment framework

In its next methodology review, the Panel encourages the CGC to consider a simplified and integrated assessment framework that:

  • returns to an operating statement framework while retaining the population growth needs assessment
  • includes the net operating deficits of subsidised PTEs including depreciation and before subsidies.

The effect of these changes would be to move from the net lending statement for the State general government sector to a modified net operating statement. The Panel considers this change could improve simplicity, transparency and stability without any significant impact on HFE outcomes. The benefits of a modified operating statement framework might include that:

  • an operating statement framework is more accessible and familiar than a net lending statement framework
  • transparency, predictability and stability are improved through the removal of a volatile element in the net investment assessment
  • capital needs for subsidised PTEs can be fully recognised
  • the approach is consistent with the upfront inclusion of Commonwealth capital payments
  • population growth needs, based on population growth dilution of net worth,17 are unchanged
  • outcomes are expected to be largely unchanged in the long term because the largest component of the current capital assessment is retained and a user financial cost of capital element ‘scales up’ the depreciation assessment.18

Under the modified operating statement framework, revenues would be as per the GFS operating statement, but expenses would be expanded to reflect the net operating deficits of the subsidised PTEs. Population growth disabilities would be applied in respect of the net worth of the State general government sector. Therefore, adopting this type of framework means that the assessment framework would not strictly align with the GFS operating statement for the State general government sector, although it would be possible to reconcile the two frameworks.

Recommendation 6.3

That the CGC examine the merits of adopting a simplified and integrated assessment framework in its next methodology review.

6.4 Cost equalisation


Differences in fiscal capacity on the expenditure side of State budgets can arise from demand (or use) factors or from cost factors.

The CGC currently equalises two types of location costs:

  • Intrastate costs, which reflect differences in the number of high cost locations within States
  • Interstate costs, which reflect the fact that some State governments face higher costs as a whole (predominantly because of higher wage costs).

The CGC also equalises for costs of administration scale. This relates to the fixed costs of State governments arising from the existence of States and Territories (which are unaffected by population settlement patterns.)

Even those who argue that HFE promotes efficient population settlement patterns generally concede that there is a question mark over equalising for location costs. The argument goes that if people faced the full cost of providing services in high cost areas (through higher taxes or lower services) they would be less likely to live there, which would create fiscal savings. Therefore, by equalising these costs HFE promotes more expensive migration/settlement patterns.19 One way to test these concerns is to consider what would happen if location costs were not equalised.

Scenario 1 — not equalising intrastate costs

Intrastate costs reflect differences in the number of high cost locations within States (primarily due to population dispersion). These costs (that is, cost differences within States) are recognised to allow high cost locations to be treated the same way regardless of what State they are in. This does not give everyone the same service standard — it just allows the similar high cost locations to have the same services in different States.

If intrastate costs were removed from the equalisation process the States that lose GST would have to raise taxes, cut services or improve efficiency, while other States could lower taxes or expand services (see Table 6.2 for indicative impacts).

Table 6.2: Indicative changes in GST from removing intrastate costs from the HFE (Scenario 1)

Change in GST shares ($m) 606 884 -411 -706 -175 86 211 -494

Source: CGC, Report on GST revenue sharing relativities — 2012 Update, Table 7.

The impact of these changes could fall on relatively higher cost (generally remote) areas (scenario 1a) or relatively lower cost (generally metropolitan) areas (scenario 1b).

Scenario 1a

Relatively higher cost remote areas may bear the brunt of lower services/higher taxes because States save from discouraging people from living there or because cuts to services in cities are more difficult since most of the population lives there. If this occurred then people would be less likely to live in regional areas of Queensland, Western Australia, South Australia and the Northern Territory. However, since the other States now have higher capacity (and can provide lower taxes or better services) the incentive could be to re-locate interstate not intrastate. This would create some fiscal savings due to people leaving high cost locations, but it would also create costs due to distorted interstate migration patterns. The net effect may be a small positive, but it is hard to say. In terms of equity, it could seem inequitable to treat high cost locations better in some States unless the number of high cost locations is policy driven.

Scenario 1b

Alternatively, relatively lower cost metropolitan areas may bear the brunt of lower services/higher taxes because services in remote areas need to be cut by too much to achieve the savings (given relatively few people live there) or because services in remote areas are already much lower, so the impact of further cuts is much higher. If this occurred then people would be less likely to live in low cost metropolitan areas of Queensland, Western Australia, South Australia and the Northern Territory and people in high cost areas may be even less likely to move (since services in nearby cities are now lower). Movement out of low cost areas would further reduce State fiscal capacity and create inefficiency due to fiscally motivated interstate migration (which is what HFE is supposed to prevent). Under this scenario, the efficiency effects of not equalising for intrastate costs are probably negative, although the effect is probably small. In terms of equity, it seems inappropriate to give unequal treatment to low cost areas just because a State has more high cost areas.

Scenario 2 — not equalising for interstate costs

Interstate costs are recognised to allow different States to deliver the same services regardless of State level cost differences. For example, this allows Western Australia to provide the same education service as Victoria, even though Western Australia has to pay its teachers more to compete with higher private sector wages.

If these costs were not equalised States that lose GST revenue would have to raise taxes, cut services or improve efficiency — other States could lower taxes or expand services (see Table 6.3).

Table 6.3: Indicative changes in GST from removing interstate costs from the HFE (Scenario 2)

Change in GST shares ($m) - 402 636 462 - 671 133 71 - 110 - 119

Source: Secretariat calculation.

In this case, the distribution of tax and service cuts within States is less important since the idea is that some States as a whole are more expensive. People would simply be less likely to live in high cost States, which would produce overall fiscal savings. In terms of equity, perhaps those living in high cost States should bear the additional costs of their location decisions. On the other hand, if high cost sub-State regions are partly subsidised by lower cost sub-state regions it may be unfair if high cost States do not effectively get the same treatment.


The Panel believes there are some valid questions raised about fully equalising for interstate cost differences without allowing for the possibility that expenditure on high cost inputs should be economised in a similar way as service standards are reduced for high cost intra state locations.

There seems to be an efficiency and equity rationale for discounting the interstate wage and non-wage costs assessment to bring that assessment into line with the CGC’s treatment of intrastate costs, and potentially promote more efficient settlement patterns. People living in high cost areas within a State face lower service standards, even though States spend more per person in those areas. On the face of it high cost States should be treated the same way — that is, high cost States should have needs recognised, but not so much as to provide the same services as low cost States without an additional local contribution.

Murphy (2012) reports that there should be full equalisation for costs, including for diseconomies of small scale and remoteness (on a stratified spend gradient basis), but only partially for interstate wage levels. This is consistent with Pincus’ (2011) observation that little or no allowance should be made for interstate differences in the unit costs of public provision of goods and services.

Identifying a discount or elasticity factor is difficult as it would be hard to measure how much services would appropriately decline with cost, and it would depend on the service being considered. The Panel has not been able to do this, or provide an indication in the time available.

Recommendation 6.4

That the CGC investigate whether it is appropriate and feasible to equalise interstate costs on a ‘spend gradient’ basis. This investigation should occur in the context of the assessment of other cost disability factors including costs of remote locations, and administrative scale.

1 GST Distribution Review, Interim Report, March 2012, pages 83-85.

2 For some areas of State expenditure, such as transport services and housing, the effects of population growth (known as population dilution needs) are the only capital needs assessed.

3 GST Distribution Review, Interim Report, March 2012, page 94.

4 Such as Mr John Hill and Mr Peter Emery. See John Hill, Final submission to the GST Distribution Review, August 2012. Also, Queensland Treasury Corporation in its submission to the Review.

5 GST Distribution Review, Interim Report, March 2012, page 93.

6 CGC, Report on GST revenue sharing relativities2010 Review, volume 2, chapter 21, page 444.

7 The ABS has since advised the CGC that the SET is unlikely to be conducted in 2013 and the CGC is considering which other data sets may be suitable.

8 Lagged annual data sets in the context of this Chapter are those that at the time of a CGC Update do not relate to the most recent assessment year, but to an earlier year.

9 The approach may not work smoothly in all aspects. For example, the population dilution disability is dependent upon population growth across years. Any such implementation issues should be approached by the CGC with the view to minimising volatility in GST shares.

10 The main source of data for the representative State budget has also changed over time. Currently, ABS GFS is the starting point. These data are supplemented by data from the States and other sources to improve comparability between States and to ensure the representative State budget is suitable for the CGC’s purposes.

11 The treatment of capital was a major issue in the 2010 methodology review. The treatment of capital is discussed in the CGC Report on GST revenue sharing relativities2010 Review, volume 1 — Main Report, 2010, pages 55 to 61.

12 See CGC Report on GST revenue sharing relativities — 2010 Review, volume 1 — Main Report, 2010, page 7.

13 For a description of the net investment and depreciation assessments, see CGC Report on GST revenue sharing relativities — 2010 Review, volume 2 — Assessment of State Fiscal Capacities, 2010, Chapter 21.

14 The second effect is proxied by using recurrent service use and cost disabilities.

15 For a description of the net lending assessment, see CGC Report on GST revenue sharing relativities — 2010 Review, volume 2 — Assessment of State fiscal capacities, 2010, Chapter 22.

16 A further complication arising from moving to a net lending assessment framework and the equalisation of NFW was the need to ensure consistent treatment of equity injections and capital subsidies to PTEs. The CGC said different treatments of equity injections and capital subsidies could influence how States make these contributions to PTEs because it could affect their assessed differences and thus their relativity. See Report on GST revenue sharing relativities — 2010 Review, volume 1 — Main Report, 2010, page 62.

17 Net worth for the State general government sector comprises physical assets and net financial worth and is also equal to net worth of the total State public sector.

18 It is likely that similar outcomes would be achieved if a financial cost of capital or lease equivalent payment is included in the modified net operating statement. Including an amount would more accurately reflect the full user cost of capital which is not captured in the depreciation estimate alone, and effectively replaces those elements of the net investment assessment other than pure population growth. Ideally, the depreciation estimates used to calculate the net deficit of subsidies to PTEs should also be ‘scaled up’.

19 This is not an issue for non-location related cost equalisation – high cost people (e.g. the elderly) cost more regardless of where they live. It is also a conceptual concern about whether location costs should be equalised. It does not deal with methodological concerns about whether they are accurately and appropriately measured in the current CGC assessment.