Fixing federal financial relations

Saul Eslake has brought necessary attention to what he labels ‘the worst public policy decision of the 21st century’ – 2019 changes to the way GST is shared between the States and Territories.

Saul’s writing1 has inspired me to write a series of articles on fixing federal financial relations drawing on my experience. To start, I wanted to add to Saul’s writing by looking at some of the data on GST sharing.

Over its first 20 years, the GST was distributed to ensure states and territories had the capacity to provide the same standard of services. The relativities used to determine each state’s share of GST over time are plotted below.

Following the 2019 changes, the CGC has also been required to assess the capacity of states relative to the fiscally stronger of New South Wales or Victoria and apply a floor of 0.7 when calculating relativities. The CGC provides a comprehensive overview of the steps involved for those interested.

The effect of the 2019 decision can be seen by comparing the solid lines - the relativities with no change - and the dashed lines - the relativities that have been used. The changes will result in Western Australia receiving around $5.5 billion more GST in 2023-24 than without.

This chart also shows that some change to GST sharing was likely required. If the previous approach had been maintained, then the system of GST sharing was at risk of failing. Before we get to that, I will provide some background on what these relativities mean.

Horizontal fiscal equalisation (HFE) is a long-standing feature of Australia’s federal system and a common feature of many federal systems. But what does it mean to equalize the fiscal capacity of Tasmania and Western Australia, say.

HFE is often, and incorrectly, framed as the GST paid in a jurisdiction they get to keep. Media about the 2019 changes often mentions Western Australia’s share falling to around 30 cents of every dollar of GST paid within the state. But GST distribution isn’t based on how much GST is paid within a State. There is a simple reason for this - we don’t know where GST comes from.

The ATO knows who pays them GST, but the address the ATO records for any payer is unlikely to reflect where the value-add that was taxed took place. Consider a supermarket chain, their address is likely a headquarters in Sydney or Melbourne. There is no way to unpack whether the GST paid by them related to sales in Launceston or Leeton or was reduced by input credits related to milk from Gippsland or pineapples from Gympie.2

Even if we did know where GST originates, a distribution based on each state’s contribution would be at odds with the purpose of HFE and other government taxing and spending decisions. The income tax I pay isn’t returned to me in the form of services of equal value. This reflects governments having distributional goals that influence its spending and taxation decision, including for the GST and HFE.

So, what does a relativity of 0.3 mean, if not 30 cents of every dollar of GST paid within a state going back to them. The relativities are weightings applied to each state’s population. The weighted populations calculated using these relativities are then used to determine the share of total GST each state will receive.

To illustrate how these weightings work, consider the case of all states having the same capacity to provide similar services. All relativities would be equal to 1 and each state would get its population share of GST. The results of this equal per capita distribution (using population and GST data from MYEFO) is shown below. The amounts are equivalent to each state receiving $3 149 per person. The table also shows how much each state is expected to receive based on the relativities calculated by the CGC and what would have arisen from the old methodology.

The GST relativities are calculated with reference to this amount $3 149 per person. The Commonwealth Grants Commission calculates an amount referred to as a ‘need’ that represents the amount above the equal per capita distribution a state requires to have the capacity to provide the same standard of services. Prior to the 2019 changes, the GST relativity for each state was calculated using the following formula:

If the capacity of a state is below the average, like Tasmania, its need will be positive. If it is above the average, like New South Wales, its need will be negative. These needs are calculated by the CGC based on each state’s ability to raise revenue and the cost to each of them to provide the same level of services. More revenue raising capacity, or lower costs of service provision, result in lower needs.

The CGC’s assessment is based on three years of data and averaged. Under the new methodology, the additional steps of assessing capacity relative to New South Wales or Victoria and applying a floor are also undertaken.

The table below shows the need per capita for 2023-24 if the methodology hadn’t change. If the risk to the system of GST sharing weren’t obvious in the chart above, then these numbers below may help.

Compare the absolute value of Western Australia’s need $2 833 with GST per capita,? $3 149. If the former was to exceed the latter, then HFE under this method would require Western Australia to contribute revenue to the GST pool that is to be shared, rather than receive a share. The system does not allow for this scenario.

The introduction of the GST and the Commonwealth’s commitment to pay all of the revenue to the states was intended to provide them with a stable and growing source of revenue. However, the growth has been less than expected. At the same time, service costs and the quantity of services provided have grown faster. And, at least in the case of Western Australia, state revenue raising capacity has grown faster still.

The risk that a state’s need per capita could exceed the amount of GST per capita has been apparent for some time. The 2019 changes may have been an attempt to address the risk of these trends breaking our system of HFE (as well as some political risks). But, as Saul has noted, it has come at a high cost to the Commonwealth and left Australia with a system of HFE that isn’t working as well as it should.

There were (or are) other solutions. Two of these stand out if you consider the equation above through the prism of tax reform that informed the introduction of the GST. To prevent any state’s relativity from being able to turn negative either GST per capita has to get bigger or Needs per capita needs to get smaller.

Increasing GST per capita, which would be consistent with the original objective of its introduction - to provide the states with a stable and growing source of revenue - requires changes to the rate and/or base of the GST. There are several good reasons for making such changes (other than fixing HFE). Conversely, there are several reasons both real and imagined for not making changes. I will come back and explore these in a later piece.

Reducing need per capita can be achieved by either changing the actual relative fiscal capacities of the states or more simply how we assess these capacities. I particularly favour changing how we assess state revenue raising capacity along the lines recommended by Peter Abelson and featured in his submission to the Productivity Commission’s inquiry on HFE. The current approach is overly complex, is unlikely to be truly reflective of fiscal capacity and likely to be an impediment to tax reform. An alternative assessment approach will be the subject of my next piece.

Footnotes

1. Being free of the restrictions of public service employment for the first time in two decades and the opportunity to work with Quarto for the first time also helped motivate my writing. ??
2. It is worth noting that GST is not paid on iron ore exports. So, arguments that Western Australia’s share of the economy might be indicative of its contribution to the Commonwealth’s GST collections would likely be incorrect. ??