OPINION: This week saw the publication of a discussion paper on the reform of the federation. This document and next month’s Council of Australian Governments leaders retreat are timely given most agree the commitment to tax and federalism reform seems to be flagging.

Not only is GST revenue growth slowing, but the prospect of the commonwealth freezing (in real terms) special purpose grants to the states after 2017 will result in commonwealth funding to the states dropping from 28 per cent of federal spending in 2001 to a projected 21 per cent by 2024 — cost shifting on an unprecedented scale.

Although NSW and Victoria are benefiting from a short-term fiscal reprieve courtesy of their spectacular and equally unsustainable housing booms, the longer-term outlook for all states remains bleak.

Clearly something will have to give.

Tuesday’s discussion paper raised the possibility of a ­commonwealth takeover of key state responsibilities such as funding public hospitals as an alternative. While this would restore balance to the federation, there is little enthusiasm for such courageous reform in Canberra.

As a result of the emergence of this “fend-for-yourself federalism” state governments will need to develop efficient and sustainable strategies to fund key public services into the future.

Increasing the rate or broadening the base of the GST have been the most widely discussed responses to this challenge. However, advocates of a revamped GST are increasingly thin on the ground. Not only does the ALP, most premiers and the community sector remain opposed to increasing its rate, but tax experts are increasingly of the view the GST will never be the growth tax many had hoped.

As the discussion paper points out, personal income tax is ­Australia’s largest and most ­robust tax with clear potential to fund state governments. Moreover, according to Treasury, it’s ­almost as efficient in economic terms as the GST. From a political perspective, it’s more progressive than a GST and should be more palatable to those concerned about equity considerations.

In practical terms there are numerous precedents for income tax sharing in federal systems. Such arrangements exist in Canada, the US and Germany.

The fact any new state income tax levies could be harmonised with the existing federal tax and administered by the Australian Taxation Office would minimise administrative and compliance costs.

Given these benefits it is unsurprising Tuesday’s discussion paper argues that allowing states and territories to access the personal income tax base could be a central element of broader reform package.

Proposals to share the income tax base across the federation have real merit; however, there are design issues that have to be analysed carefully.

The first question is whether the states and the commonwealth should enter into a simple revenue-sharing agreement whereby the states get a fixed percentage of personal income tax revenue. Although this would give the states access to a growth tax, it would do little to improve political accountability.

A more ambitious model would involve granting the states the power to set their own income tax levy or surcharge on the federal income tax base. Such a ­regime would improve accountability and allow state governments to set state income tax levies based on local preferences and demand for services.

Despite improved account­ability, allowing the states to set their own income tax rates has significant distributional implications that have hitherto been ignored in the debate.

Quite simply, a per capita revenue from any residency-based state income tax levy will vary considerably between states because income and employment levels vary across the states.

The bottom line is wealthier states would benefit more from such a regime compared with poorer jurisdictions.

Our analysis suggests a 3.2 per cent income tax levy on income above $37,000 could raise $10 billion for the states and territories.

However, the distribution of such a levy favours wealthier jurisdictions whereas poorer states would have to exert a greater tax effort to raise equivalent per capita revenue.

One way of representing the magnitude of these distributional consequences is by considering the income tax levy different jurisdictions would have to impose to raise the same per capita revenue.

For example, Western Australia would have to impose a levy of 2.1 per cent to raise its per capita share of a $10bn levy. In contrast, this figure for poorer states such as Tasmania and South Australia would be 5 per cent and 4.3 per cent respectively.

What is clear is the design details of income tax sharing models as outlined in this week’s federalism discussion paper need to be analysed and debated in the coming months. The future of our federation depends on it.

Richard Eccleston is director of the Institute for the Study at the University of Tasmania. Neil Warren is in the school of taxation and business law, UNSW Business School.

This opinion piece was first published in The Australian.

The full paper is available here