Australia's Commonwealth government and states share power over key state-based services - notably human capital (such as health and education) and infrastructure. These services are increasingly seen as crucial national policy areas. And demand for them is growing faster than GDP.

The incoming Rudd government has made federal-state relations an early and prominent element of its agenda. The government has promised "a new model of cooperation" with the states, to put an end to "the blame game and buck passing". The Council of Australian Governments, COAG, has been given a central role in the process.

The result is increased attention on how the Commonwealth government and state governments jointly manage their relationship. In particular, many observers currently argue that:

• We have had or will soon have too much federal government involvement in traditional state fields of activity
• The states should implement uniform policies
• The states should minimise duplication and overlap with federal government programs
• The states should prevent further growth in their outlays
• Richer states should cease to subsidise poorer ones - "horizontal fiscal equalisation" is outdated
• Australia should seek to reduce the gap between state spending and the revenue that states raise themselves - "vertical fiscal imbalance".

Each of these points is either wrong or overstated in the current debate. In particular:

• The Commonwealth has an important role to play in:
• • coordinating policies
  • fostering competition among states, and between states and the Commonwealth to find the best policies, even where this results in overlap and duplication
  • centralising where appropriate
• State budgets will come under growing stress over coming years, even if states merely aim to maintain current service levels.
• Some horizontal fiscal equalisation is justified, though the present system delivers too much equalisation in too complex a fashion.
• The traditional argument against vertical fiscal imbalance - that the states will spend more wisely the revenue which they raise themselves - has a very weak basis in logic or data. The current arrangements, with the Commonwealth raising a large portion of the states' funds, brings substantial advantages.

The Australian federal system must be judged as one of the most successful in modern history. This paper argues that a key to its success is the evolution of Australian federal arrangements in response to changing circumstances. Federalism can continue to assist Australians in achieving their aspirations in the twenty-first century, by encouraging the choice of good public policies, and facilitating their efficient implementation.

Sources of budget stress

In 2005 the Productivity Commission estimated the effects of an ageing Australian population on the budgets of Australian governments. Assuming that tax collections will not rise as a proportion of GDP, the Commission found that, without policy changes by governments, the gap between government revenue and government spending will reach around 6.4 percentage points of GDP by 2044-45. It expected much of this would be borne by the Commonwealth, but with significant potential burdens facing state and territory governments.[1]

Changes in the age structure of the population represent only one source of budgetary stress, particularly for state governments, in the years ahead. By far the largest likely source of increase is the very process of economic growth itself. As citizen-voters grow richer, they tend to demand from government more services and better quality services. This includes more and better roads, public transport, public parks, public schools, public hospitals, police, courts, jails and so on.[2] Currently, many of these are provided by state (or state and local) government, frequently with Commonwealth money supplementing the states' own revenue sources. The Commonwealth also faces its own expenditure pressures, in part due to social security spending.[3]

A third factor can also increase the unit cost of provision of many state services, whatever the source of their funding. States provide services in health, education, justice and the like. These are intensive users of labour - often highly trained labour. In many of these services, it is hard to raise productivity simply by replacing people with non-human inputs such as information technology. Even with their workers efficiently deployed (not always the case), these services can be productivity laggards over time. The wage rates of state employees, however, tend to rise with the general increase in the economy's productivity, causing a kind of cost squeeze.[4] Similar comments apply to Commonwealth expenditures, and especially to human services (but not to cash transfers).

Additional state or Commonwealth expenditure does not have to be financed exclusively from taxes and charges; that is, appropriate accounting conventions and strong financial discipline should not prevent governments from borrowing for investment purposes. However, the recurrent spending will be largely tax funded, and that implies there will be pressure for taxes to rise as a share of GDP.

Though this section has focused on budgetary stress, much more is at stake. Through regulation, taxing and spending, governments help to create the social and economic environment or habitat within which Australians can best pursue their individual and collective goals. Although productivity isn't everything, it can facilitate the creation of that environment. Therefore, this paper gives some attention to schemes of microeconomic reform designed to improve national productivity.

Policy responses in a federal system

Our constitution has, over time, evolved to deliver the Commonwealth greater constitutional and fiscal power than the states. This gives the Commonwealth the scope to choose between:

• excluding the states from policy determination and implementation[5]
• co-opting the states, as agents of the Commonwealth
• cooperating with the states, as partners
• leaving matters entirely to the States.

In its term so far, and notably at the 20 December 2007 meeting of COAG, the Rudd government has opted for cooperation.[6] This section discusses the grounds for choosing between these options, but gives most attention to the grounds for cooperation.

Centralisation and decentralisation

A functioning Australian federalism requires a balance between centralisation and decentralisation. Ultimately, the balance is decided by democratic politics - reflected in the unilateral decisions of the Commonwealth, or in inter-governmental agreements - and by the High Court. However, political philosophy and economic analysis can and should give some guidance.

The decentralisation-centralisation debate is also a debate about competition between governments. Centralisation is anti-competitive: the full centralisation of a government function brings to an end any competition between domestic governments (though not an end to competition between governments in a globalised world). Decentralisation, in contrast, is competitive. Agreements between governments for coordinated action are in the middle - they tend to lesson competition, but can leave room for it.

Centralisation of government decisions or administration can achieve citizens' collective and individual goals better than would decentralisation of government in many areas of governmental action. The right course depends on specific circumstances.

The Constitution offers the Commonwealth the power to centralise - that is, to exercise exclusive powers over - not only defence, external affairs and trade, immigration and other matters not directly relevant to this paper, but also the laws and regulations needed to create national markets; for example, telecommunications, weights and measures, currency and coinage, bankruptcy, copyright, insurance, and banking. As the Australian society and economy has grown and changed, strong cases have arisen for the extension of central power over management of the national economy and the financial security of its citizens.

Decentralisation works to the best advantage to Australians when the framework has been correctly set for productive competition between governments. As section on competitive federalism will suggest, the continued health of the federation requires the decentralisation of fiscal capacity, and of policy frameworks, policies and their implementation, unless there are strong reasons to centralise them.

Synergies and spillovers

This section now turns to the principles behind the adoption of the intermediate arrangement: inter-governmental agreements to cooperate and coordinate. When is there a case for Commonwealth grants to smooth the process of cooperation?

The work of national reform will involve coordinated and cooperative efforts of governments. Coordination of reform is advantageous when there are synergies in joint action, and when there are spillovers.

The 1995 National Competition Policy provides an example of synergies. One state reforming just its local electricity industry could not reap the full benefits of that reform. But increased interconnection of electricity grids in south-eastern Australia delivered substantial benefits to every state. Coordinated reform produced what no state could immediately gain acting alone.

Spillovers present a slightly different case. Here a reform by one state acting alone can produce benefits not only for residents of the reforming state, but also for residents in other states. When there are significant and uncompensated beneficial spillovers across state boundaries, no one state has sufficient incentive to put in the full reform effort that would be warranted from a national point of view.

A state will tend to capture less of the total benefits of its own efforts when reforms have long maturation periods, as from early childhood education, included in the 2006 National Reform Agenda. Because Australians are very mobile, and because of the importance of income tax in the revenue system, the benefits of reforms made by one state or territory government will not be confined within its own boundaries. More generally, the effects of enhanced competitiveness increased labour productivity and improved labour-force participation - the 2006 National Reform Agenda - will be widely felt throughout the economy and society.

Of the new elements in the December 2007 COAG reform agenda - climate change and water, infrastructure, housing, and funding for treating Indigenous alcohol and substance abuse - probably only the first two also satisfy the criteria of synergy and spillovers.

This is not to say that the policy decisions on housing and Indigenous funding are consequently bad and that the other two are good. Rather, the point is that these may be two instances in which national goals have been set by the central government unilaterally, and that the Commonwealth will be relying on the states merely for assistance in their implementation.

The case for incentive payments

The case was sketched above for compensatory payments to be made when a state's own reform effort produces significant beneficial spillovers to other states and to the Commonwealth. Such "spillover" payments act as an incentive to contribute appropriately to the jointly beneficial result. For practical reasons, it is sensible for these payments to be arranged as Commonwealth grants.[7]

There is a fiscal side to the spillover case for Commonwealth-funded reform-based payments to the states.[8] The fact that the Commonwealth stood to gain fiscally whenever a state privatised a publicly owned utility was an important argument made for the Commonwealth to share the fiscal gains of the 1995 National Competition Policy. And the recent quantitative estimates made by the Productivity Commission indicate that the fiscal benefits of implementing the 2006 National Reform Agenda will be shared between the Commonwealth (mostly through rises in income tax receipts) and the states and territories.

And there is also the matter of everyday state politics. Despite the case for coordinated reform, political and economic circumstances may tempt a state government to renege on previous agreements, or backslide. Incentive payments can be wielded as a stick, as well as offered as a carrot.

The 1995 National Competition Policy introduced the concept of incentive payments to the states and territories in return for implementing agreed reforms. The payments seem to have facilitated the reform process.

The same would probably have held true for the 2006 National Reform Agenda. However, the Howard government was reluctant to go down that route. In contrast, the Rudd government at the December 2007 COAG meeting proposed that incentive payments will be a pervasive feature in the pursuit of the new reform agenda.

A redistribution of the fiscal benefits of the 2006 National Reform Agenda and its 2007 COAG extension would add to the financial strength of the states - this will be examined later in this paper - unless the Commonwealth offsets them in some way.

Recent federal state co-operation

Over the past 13 years, Australian governments have gone through three rounds of cooperation on economic reform:

National Competition Policy, 1995

After a review led by Professor Fred Hilmer, the states and the Commonwealth agreed in April 1995 to a nationally coordinated program of microeconomic reform. Among other things this program extended competition laws to government businesses and the professions, reformed public monopolies, and introduced third-party access regimes for monopoly infrastructure providers. It made specific pro-competition and pro-market reforms to the electricity, gas, water and road transport industries.[9] The Commonwealth made payments to the states and territories on the achievement of agreed reform targets or milestones. According to Productivity Commission estimates, the changes in infrastructure industries associated with National Competition Policy added 2.5 per cent to Australian living standards.[10]

In its review of National Competition Policy and subsequently, the Productivity Commission outlined a series of further nationally coordinated reforms and recommended the use of incentive payments.[11] The Council of Australian Governments meeting of February 2006 agreed to a National Reform Agenda to further reform competition, regulation and "human capital" development in health, education and work incentives.[12] The Productivity Commission estimated that gains to GDP from these reforms could be substantially larger than those from the 1995 National Competition Policy. Potential increases in government revenues were to be shared between the Commonwealth and the states and territories, However, during 2006 and 2007 under the Coalition the Commonwealth did not agree to make incentive payments to the states and territories, of the kind that drove National Competition Policy.

The Council of Australian Governments meeting on 20 December 2007 agreed to push ahead with microeconomic reform, starting with areas previously agreed for further enhancement, but on which not much seemed to be happening. New areas have been added, some of them with a broader focus than microeconomics. COAG itself is to become a more important forum, with four COAG meetings to be held in 2008. The meeting established seven working parties to look at:

• health and ageing
• the productivity agenda - including education, skills, training and early childhood
• climate change and water
• infrastructure
• business regulation and competition
• housing
• Indigenous reform

Importantly, the Commonwealth will drive the process with incentive funding to the states and territories. In March 2008 COAG is also scheduled to consider reform of specific purpose payments - a large array of payments by the Commonwealth, contingent on the satisfaction of conditions imposed by the Commonwealth. These payments are primarily in health, education and transport, but also in many other areas. They amount to over $30 billion, or over 40 per cent of Commonwealth payments to the states and territories.[13]

Three election promises of the incoming federal government are to receive immediate attention: additional funding to cut waiting times for elective surgery, provision of funds for student computers, and a doubling of funding available for substance and alcohol rehabilitation and treatment services for Indigenous Australians.

Competitive federalism[14]

The previous section was about cooperation to set and achieve national goals. This section is about competition, and is motivated by the belief that monopolies, even in government, can be bad for citizens.

Just as competition among firms safeguards consumers against high prices and shoddy goods and services, so competition among governments can safeguard citizens against bad government and encourage good government. We are used to electoral competition: periodically, the governments in power in Canberra and in each state or territory "go to the people". Federalism reinforces the strength of electoral competition, by pitting the central government against the states and territories, individually or collectively. This is vertical inter-governmental competition, and it is unique to federal systems.

Discussions of federalism often disparage the existence of "overlap and duplication" between the Commonwealth and the states and territories. However, overlap can be a sign that vertical competition is operating.

Vertical governmental competition in a federation

It is difficult to argue that waiting times for elective surgery in, say, Queensland are matters of great importance to most residents of Western Australia who are themselves not waiting in line, or with relatives and friends in line. Waiting times do not seem to satisfy the criteria suggested in this paper for an inter-governmental or national reform effort - they offer little by way of synergies; and few spillovers across state boundaries. Rather, their inclusion in the 20 December 2007 COAG agreement delivers vertical competition within the federal system. The state public hospital systems have not delivered the kinds of waiting times that are acceptable to citizens, and so there is a political payoff for the Commonwealth to step in. The Rudd government is in competition with the states for political support and voter approval, via Commonwealth intervention in a field generally regarded as a state responsibility. Such competition is only available to citizens in a federal system.

Although vertical competition can result in one level of government vacating the contested field, more commonly the Commonwealth comes to share the field with the states, in areas formerly regarded as state responsibilities. This is happening now in the area of hospital waiting lists. (The states risk High Court action if they attempt to encroach too boldly on areas that the Commonwealth believes its own. But state competition with the Commonwealth in shared fields is common.)

Overlap and duplication

The simultaneous involvement of more than one government in a single area is often criticised as wasteful overlap and duplication. Because no single government is clearly responsible, vertical competition is also criticised as the source of blame shifting and a reduction in political accountability.

These are costs - although they may have been exaggerated.[15] But they need to be weighed against the benefits of vertical competition.

Competitive processes and a rationally planned monopoly produce different results. The overlap that arises from inter-governmental competition may be worth the cost if the inter-governmental competitive process sufficiently enhances governmental responsiveness to citizens' preferences; for example, by promoting improvements to service delivery.

The alternative, often advocated, is a clear and unassailable assignment of responsibilities within the federation. This would, almost by definition, completely eliminate overlap, duplication and blame shifting - but only by completely eliminating the possibility of vertical competition itself.

What specific federal division of governmental powers and responsibilities best suits Australia and Australians, today and in the coming decades, cannot be determined completely on general principles. Competitive political processes influence that division, and rightly so. However, competition yields better outcomes when the playing field is level, and currently the national government's political power is backed by vastly superior financial resources, enabling it to "go it alone" or to make grants to the states contingent upon the actions or performance of the states. The excessive financial dominance of the Commonwealth over them is detrimental to vertical competition - an issue for the section dealing with the fiscal capacity of the states below.

Horizontal competition

Horizontal competition encourages governments to get the "economic fundamentals" right in:

• laws and regulation
• infrastructure, human and non-human, constructed and environmental
• social and economic services to persons and businesses
• revenue raising - both the rates and the structures of taxes and charges.[16]

Migration is the enforcer of horizontal competition. When government decisions really matter to people and businesses, they can and do "vote with their feet", moving to a country or state offering a political, social and economic environment more to their liking.

Movement is easier within a federation than between countries, and probably better informed, because federalism makes it easier for citizens to compare the performance of different governments - a process the Productivity Commission calls "yardstick competition". Governments, of course, also study these comparisons.[17]

Governments compete by experimenting with novel policies, or with variations on what they or others have tried. Naturally enough, this has costs, including for firms facing different regulations in different jurisdictions. And, paradoxically, to the extent that horizontal competition works, differences between governments tend to vanish. But don't be fooled - without competition, there is a weaker guarantee that the common result will be as beneficial for people and businesses.

Tax competition

Inter-governmental competition to give people good laws and regulation, and good value for their taxes and charges, seems an unambiguous good. Evaluating competition over tax rates and tax bases, however, is more complicated.

As is often the case in federalism, there is a question of achieving the right balance, this time between the twin terrors of a tax "race to the bottom" and a fully coordinated fiscal cartel. The former would leave governments with revenues inadequate to satisfy reasonable expectations for public services; the latter would leave citizens too little cash for the reasonable satisfaction of their private wants.

The problem is well illustrated by the history of payroll tax. After land taxes, payroll tax is regarded by many economists as the second-best tax available to the states. It is a growth tax, in that payrolls tended to grow along with national income. If a payroll tax is levied on a broad base and at a uniform rate, then most economists, against popular and business opinion, will judge it to be an economically efficient tax.

An important difference between the GST and payroll tax is that the states and territories do not control the base and rate of the GST. The GST tax base and tax rate is set by the Commonwealth parliament. The states can remit their GST revenue in the form of tax relief - in fact, the opportunity to reduce "bad" state taxes was part of the motivation for the introduction of the GST. Also, the states can and probably have used GST revenue to further erode what most economists think of as good or efficient taxes, like the payroll tax and land taxes.

The states are widely regarded by economists as having "abused" payroll tax - so much so that some previous supporters of the tax have joined popular and business condemnation of it.[18] In 1971, the Commonwealth handed the payroll tax field to the states. Initially, the states all reacted by raising rates of payroll tax, broadly in unison, from 2.5 or 3.5 per cent to over 6 per cent. Since then all states have eroded the payroll tax base by increasing the tax-free thresholds faster than the rate of wage inflation. Since the early 1990s, various states have provided payroll tax exemptions or reductions as a general means to attract employers, especially small employers, or to attract specific firms or operations - for example, in the case of Queensland, the headquarters of Virgin Blue.

In an even more famous instance of Australian tax competition, the states - led by former Queensland Premier Bjelke-Petersen - competed away death duties, another tax regarded by many economists as relatively efficient. These inducements are examples of what the Productivity Commission calls "destructive" horizontal competition between the States.[19]

If tax competition is too destructive, then sensible federations can centralise selected taxes, as did Australia with the income tax and the GST. That does not imply they should eliminate all tax competition.

Fiscal capacity of the states

One prerequisite for a well-functioning federation is continuing strong democratic support for the states. In turn, this requires the states to maintain adequate levels of fiscal capability and fiscal independence; in a federal fiscal system that is comprehensible and transparent, so that citizens know who is funding additional state spending.

Two remarkable features of the current fiscal system seem inconsistent with these objectives:

• The degree of "vertical fiscal imbalance" - the Commonwealth dominates tax collections, but then hands a substantial portion of that tax revenue to the states.
• The extent of "horizontal fiscal equalisation" - poorer states get their tax revenues topped up.

The first of these, vertical fiscal imbalance, is a false bogeyman. The second, horizontal fiscal equalisation, has indeed damaged intergovernmental competition, but there are strongly mitigating circumstances. Once these two features of the system are acknowledged as having useful aspects, then the debate over the states' finances changes substantially.

Alternative federal futures

In one possible future for Australian fiscal federalism, the states would cede further functions to the central government. That would reduce pressures on state budgets and also decrease the extent of vertical fiscal imbalance, or moderate its increase. In another possible future, the central government moves further into sharing expenditure responsibilities with the states in various fields. Yet a third future blends these two possibilities, with the states ceding functions and the Commonwealth relying on the states to deliver new Commonwealth programs rather than centralising the provision of such services. Such an arrangement is in operation in a number of unitary countries; including France, where the regional Departments deliver many public services.

Full centralisation of power has happened less frequently in Australia than many people believe. Income support was centralised after the passage of the 1946 social services amendment to the Australian constitution, but no such centralisation has occurred since the passage 40 years ago of the 1967 amendment regarding laws with respect to people of the Aboriginal race, s. 51 xxvi. In recent decades, cooperation and competition have been more common than outright and complete centralisation (with Work Choices the notable exception).

In an environment dominated by cooperation and competition, there are three ways to increase the tax funds available for state spending:

• increase Commonwealth grants to the states, funded largely by the GST and the income tax
• raise the rates on the states' taxes or extend their tax bases, so as to increase state tax collections from the existing set of state taxes
• extend the range of state tax sources to include some taxes presently exclusive to the Commonwealth, such as income tax.

The last of these three appears very unlikely, and may be undesirable. This leaves us with two alternatives from which to choose. To evaluate them, we need to decide how important is the problem of vertical fiscal imbalance.

In defence of vertical fiscal imbalance

The Commonwealth collects much more revenue than it spends for its own purposes. The state, territory and local governments, in contrast, do more public spending that their own-source erevnues can cover. They depend on Commonwealth grants for the difference. This is vertical fiscal imbalance.

Some commentators believe that any vertical fiscal imbalance at all is a bad thing. Cliff Walsh, adapting the American Revolutionary cry, coined the slogan "No representation without taxation", to indicate that he thought that the States should spend only what they earn. Along similar lines, former Premier Bjelke-Petersen of Queensland is alleged to have said that the only good tax is a Commonwealth tax - the Commonwealth takes the odium and the premier takes the credit for the (presumable profligate) spending.

This paper argues the opposite: that some degree of vertical fiscal imbalance actually benefits Australian fiscal federalism, and that eliminating it would damage the federal system and harm the economy. The GST arrangements, sometimes claimed as a fix for vertical fiscal imbalance, are neither a remarkable new beginning for federalism, nor the harbinger of its demise, but rather a change in how the Commonwealth is funding its grants to the states.

Finding the revenue

It is clear that to retain relatively independent control of policy determination and to afford to deliver something like the existing range of state services in the future, the states and territories will need more revenue.

So, with a lump of state spending to fund from taxes, from where should the taxes come? Who should be levying them, and at what rates?

The answer is that a sensible federation draws on both centralised and decentralised taxes to fund state spending.

Two principles of efficient taxation can help us decide the right tax mix:[20]

• If it stands, tax it.
• If it moves, catch it first, and then tax it (but not so heavily that it leaves).

From these two principles, we also get a recipe for setting the various tax rates:

Utilise taxes so that the cost to the economy of an extra dollar of tax revenue is (about) the same for every type of tax being employed.

The story of US funding of state government welfare exemplifies what happens when the second point is ignored in a federal system. Heavier state income taxes and state sales taxes, imposed to fund more welfare recipients and/or more generous welfare, led to US business and households not dependent on welfare to move to states where taxes were lower.

Australia's federal system has neatly avoided this problem: the fields of sales taxes (that is, the GST) and income taxes are both centralised.

It is worth noting that for the purpose of federal-state financial relations, the GST and Commonwealth income tax revenues look almost identical to a recipient state. They differ only in that the GST flows without negotiation with the Commonwealth.

Commonly in economics, what matters is not so much the average, but what happens on the margin. Commonwealth money, whether sourced from income tax or provided automatically from a GST, gives a base to state funding - about 40 per cent of state recurrent spending. To fund extra state spending, the states need to find additional revenue from their own sources.

The ideal tax types to assign to the states are taxes for which what you tax (in technical language, the "tax base") grows in response to state spending and regulation, and taxes that are not likely to trigger significant interstate migration to avoid them. These include taxes on land values, and payrolls (the latter not fully immobile, as the Virgin Blue example shows.) And these are, of course, among the taxes that the states have in Australia today.

In the light of this analysis, vertical fiscal imbalance appears a good deal less sinister, and much more an appropriate trade-off. In a federal system, state government spending should be funded from both centralised (that is, Commonwealth) sources, and decentralised (that is, state) sources in a way that maximises economic efficiency. Federal grants to the states are likewise for public spending, and should be calibrated to minimise the economic costs of tax revenues.[21]

And a simple, transparent system for their distribution is needed to ensure that electoral competition does its job. For this defence of vertical fiscal imbalance (VFI) is most robust when there is strong electoral competition.[22]

In defence of (some) fiscal equalisation

The second feature of Australian fiscal federalism is an almost obsessive attempt to equalise the fiscal capacities of the states, through the distribution of the GST revenues. The objective is to ensure that each state, if it chose, could provide state services equal to the average across the states, and finance those from the proceeds of state taxes and charges levied at the average rate across the states.

Some degree of horizontal fiscal equalisation is beneficial to the efficient operation of the fiscal system, and accords with Australian notions of "a fair go". But there is a strong case for reform to the purposes and calculation of fiscal equalisation grants.

As Ross Garnaut remarked,[23] the extent extreme of horizontal fiscal equalisation in Australia is surprising:

…when, of all of the federations amongst developed countries - Germany, Canada, the United States - Australia starts with the least unequal distribution of income across States. Yet we do far more, and far more elaborately, to equalise. …The only case that compared with it in ambition was the former Soviet Union, but in practice the former Soviet Union did not go as far as Australia. …But one general point I would make is that it is very hard for electorates, or for commentators, or for anyone, to assess outcomes of [inter-governmental] competition when right at the centre of federal-state relations you have these huge transfers on the basis of principles that probably only a dozen people in Australia outside government understand.

Australians are mobile, and it is best if they settle where they can be provided government and other goods and services at the least cost, and where, if employed, they are most productive. However, they can be induced by a poorly designed federal fiscal system to make inefficient settlement decisions. Surprisingly, fiscal equalisation grants can be designed to reduce inefficient mobility and settlement.

Consider the tax side of the equalisation problem. Prominent among the tax bases suitable for assignment to states are land and resource taxation, and the payroll tax. In states rich in taxable natural resources, wage earners would enjoy a lower rate of payroll tax, or more state spending. Therefore, a mobile worker could find some financial advantage by moving to a resource-rich state, even if his or her productivity and wage rate were lower in the destination state. In effect, through the state fiscal system, such a worker would become a part owner of mineral resources without buying into the stock market. Migration for such a purpose would be inefficient and would reduce national output.

Calibrating the GST grants so as to "compensate" the budgets of resource-poor states, as occurs in Australia, removes this incentive for inefficient migration.

A parallel argument arises on the expenditure side. Tasmania, with the highest proportion of elderly residents, needs higher per capita tax revenue to fund the same level of state services to its elderly folk as New South Wales. If this means higher Tasmanian tax rates, then some taxpayers of Tasmania may be induced to leave. If Tasmania receives additional GST monies to compensate, this problem is avoided. (The Tasmania government is not required by the Commonwealth to actually spend the extra grants on elderly folks.)

Funding additional state grants

Accepting, then, that state spending should be funded in part from Commonwealth grants, how should these grants be financed? On the marginal principle set out earlier, the answer is from both income tax revenue and GST revenue.

The first main effect of the introduction of the GST was that it enabled the Commonwealth to reduce income tax collections.[24] For economic efficiency, this was an advantage. The cost to the economy of an extra dollar of income tax revenue was high, relative to the cost of an extra dollar of GST revenue, because the GST is a tax on spending, not earning and saving, and has a broad base and a uniform rate (especially when compared with the sales taxes it also replaced).

Since the implementation of the GST, overall Commonwealth tax receipts have powered ahead, growing faster than GST receipts. Some rebalancing, therefore, may be needed within the Commonwealth's tax sources.

There is a second consequence of the introduction of the GST, of some importance for my argument. Previously, from the same pool of general Commonwealth tax revenue, the Commonwealth funded two kinds of grants to the states - unconditional grants and conditional grants. Unconditional grants are for the states to spend as they wish. Conditional grants are for the states to spend as the Commonwealth wishes. The change is that the GST monies are hypothecated to the states and territories, and fund unconditional grants. Buoyant GST revenue gives some scope for state autonomy, but at the expense of reducing their incentives to raise more of their own tax revenues, through the redesign of payroll taxes and land taxes.

More specific purpose grants?

This paper has defended the idea of revenue grants from the Commonwealth to fund state spending at the states' discretion. Another possibility - more a probability - is that the states will be come to rely more heavily on specific purpose payments (SPPs). These are currently subject to review by treasurers, under COAG. To the extent that such payments accord with the principles of coordination and competition set out above, then they pose no great threat to the long-term vitality of the states. (The payments contemplated in the COAG statement of 20 December 2007 are of this type, offering payment on achievement of outcomes or outputs that the states would, in any case, generally want to achieve.)

But SPPs can be used to induce the states to spend on things they would not want to spend their own money on, and for which there is no valid coordination or competition purpose.[25] At that point, the expansion of SPPs represents centralisation by stealth - and can pose a long-term threat to the satisfactory operation of the Australian federal system.

Conclusion

Contrary to received wisdom, Australia has got most of the big things right in the financial relationship between its states and its central government. The federation has achieved a good - though not perfect - balance between centralisation, cooperation and competition between governments. It has managed to realise significant microeconomic reforms over recent years as a result of governmental cooperation and central coordination. The major sources of tax revenues have been centralised, avoiding excessive tax competition among the states. And fiscal equalisation occurs across the states, albeit too comprehensively and too opaquely.

Both the Commonwealth and the states will need additional tax revenue as the population ages and incomes rise. For the states, part of that extra revenue is likely to continue to come in the form of Commonwealth grants. Despite what many analysts claim, that is not a calamitous result.

[1] Productivity Commission 2005, "Economic Implications of an Ageing Australia, Research Report", Canberra (especially Appendix F). In its submission to the Productivity Commission, the Tasmanian Treasury argued that a severe or worse expenditure load will fall on that state - a negative primary budget balance for Tasmania of 3.8 per cent in 2042.

[2] The ABS publication, "Government Finance Statistics" (5512.0) shows a fall in state and local outlays, expressed as a ratio to GDP, from a peak of over 22 per cent in 1982-83, to 16.2 per cent in 2004-05. Much of this fall could be due to the privatisation of public enterprises in transport, energy, water and sewerage, for example; and a little to the shift of students and patients from state to private schools and hospitals. The bulk of the evidence suggests that the services mentioned are, in economic terms, superior goods.

[3] See the Treasury's second Intergenerational Report, http://www.treasury.gov.au/igr/at_a_glance/IGR_2007_at_a_glance.htm.

[4] This is known to economists as Baumol's cost disease, after William Baumol, who helped describe it.

[5] Crediting Justice Kirby with the phrase, Anne Twomey and Glenn Withers call some such moves as "opportunistic federalism": Federalist Paper 1. Australia's Federal Future, a report for the Council of the Australian Federation, April 2007.

[6] The full communiqué is available at http://www.coag.gov.au/meetings/201207/index.htm#infra. Local governments were present at the COAG meeting, but the agreement was between the senior governments.

[7] For the spillovers enjoyed by other states, some kind of purely inter-state agreement on payments is conceivable, but maybe impractical. The synergies of policy coordination make it difficult to apportion credit to each state for its contribution: see references to Productivity Commission research, notes 8, 9 and 11 below.

[8] For those few outside government who understand the operations of the Commonwealth Grants Commission (CGC), what matters is whether the CGC treats the payments by "inclusion", not "exclusion" - that is, pools the payments with the GST revenues, to be distributed as fiscal equalisation grants.

[9] Productivity Commission 2005, "Review of National Competition Policy Reforms", Report no. 33, Canberra, February 2005.

[10] Productivity Commission 2005, "Modelling Impacts of Infrastructure Industry Change over the 1990s", Supplement to Review of National Competition Policy, Productivity Commission Report No. 33, Canberra, February.

[11] These reviews and recommendations are discussed in the Productivity Commission chairman's 2004 address to CEDA on microeconomic reform: Gary Banks, "NCP and Beyond: Agenda for National Reform", address to CEDA, Sydney, 29 November 2004, and Melbourne, 6 December 2004.

[12] Productivity Commission 2006, "Potential Benefits of the National Reform Agenda", Report to the Council of Australian Governments, Canberra.

[13] See "A framework to Guide the Future Development of Specific Purpose Payments (SPPs)". A discussion paper by the ALP Advisory Group on Federal-State Reform, 31 July 2007, Sydney, http://www.alp.org.au/download/now/070731_alp_advisory_group_on_federal_state_reform.pdf

The final report of the Advisory Group had not been published, at the time of writing. See also Meredith Edwards, "Federalism Panel Presentation", IPAA National Conference, Perth, 18 September 2007, available at http://www.wa.ipaa.org.au/documents.php?sort=name& category=National+Conference+2007.

[14] This section draws heavily on Productivity Commission 2006, "Productive Reform in a Federal System", Roundtable Proceedings, Productivity Commission, Canberra. The interpretations offered here are strictly my own.

[15] See Anne Twomey and Glenn Withers 2007, Federalist Paper 1. Australia's Federal Future, a report for the Council of the Australian Federation, April, Appendix 1, for a critical analysis of estimates of the costs of overlap and duplication (made by Drummond, and by Access Economics for the Business Council).

[16] Productivity Commission 2005, Annual Report 2004-05, Annual Report Series, Productivity Commission, Canberra, p. 8.

[17] Steering Committee for the Review of Government Service Provision 2007, Report on Government Services 2007, Productivity Commission, Canberra (the "Blue Book") covers 60 per cent of recurrent public expenditure, about 10 per cent of GDP.

[18] O. Gabbitas and D. Eldridge 1998, Directions for State Tax Reform, Productivity Commission Staff Research Paper, AusInfo, Canberra, May.

[19] Gary Banks 2002, "Inter-state Bidding Wars: Calling a Truce", a speech to the Committee for Economic Development of Australia in Brisbane, 6 November. See http://www.pc.gov.au/speeches/cs20021106

[20] These are vulgate versions of the famous Ramsey tax principles of 1927.

[21] G. Brennan and J.J. Pincus 1996, "A minimalist model of federal grants and flypaper effects", Journal of Public Economics,61, pp. 229-46.

[22] In South Australia, an election was fought largely over how to pay for the State Bank debt - cut state spending, or increase state taxes and charges.

[23] Productivity Commission 2006, Productive Reform in a Federal System, Roundtable Proceedings, Productivity Commission, Canberra, pp. 86 and 88.

[24] David Hamill (2005) argued that the GST arrangements were a further step towards a "creeping centralism [that] is threatening the fundamental characteristics that distinguish Australia's system of government as a federation." See David Hamill 2005, "The Impact of the New Tax System on Australian Federalism" (June) at http://www.ipaa.org.au/01_cms/details.asp?ID=146.

[25] The Commonwealth Grants Commission (CGC) treats many specific purpose grants as though they contribute dollar-for-dollar to the capacity of the states to fund their own spending - as though they were the states' own revenue. The CGC calls this "inclusion" into the pool of grants to be equalised. It may be a practical way to deal with these grants, but it assumes that the Commonwealth goes to a great deal of effort to influence state spending, to no avail.