Restoring The Balance: Tax Reform for the Australian Federation

PART I:  BACKGROUND

1.1 INTRODUCTION

We gratefully acknowledges the assistance and advice of Peter Durack, Q.C., and Perry Shapiro who bear, of course, no responsibility for the final product.  It continues and expands work published by W&A over the last few years on various aspects of federalism and taxation policy.

Its purpose is to canvass the core issues which arise out of the now widespread desire to reform not only the taxation systems of the Australian Commonwealth and States, but also the fiscal arrangements which prevail between the States and the Commonwealth.  The two issues -- taxation and federal fiscal affairs -- are, in our view, inseparable.

Reform of both tax and federalism are subjects which have been much studied in the past;  and although those past studies have a good deal to teach us, they have – regrettably -- largely been in vain.  Potentially viable plans for reform have not been lacking (and some of these are alluded to in what follows).  What has been lacking in the past is a political window of opportunity, combined with broad and strong advocacy.  In the judgement of those supporting the Project, that opportunity now exists.

There has, in the first place, been a dramatic change in the fortunes of the States.  Where once they were, to varying degrees, seen as fiscally irresponsible and only too happy (despite rhetoric to the contrary) to be tied to the expenditure strings of the Commonwealth, now they lead the way in fiscal responsibility and other reforms.  Where once they whined their way to each successive Premiers' Conference, and allowed the Commonwealth to divide and rule, fiscal responsibility has increased their moral standing, and they are capable of co-operating on key issues as never before, and with more power of suasion than ever before.  To some extent, that perception of fiscal irresponsibility has been an illusion, as the following chart shows.  What matters now is that fiscal responsibility is now matched by perceptions of responsibility.

Chart 1:  General Government Recurrent Budget Balances

Source:  V.W. FitzGerald, "Short-Termism in Saving, Real Investment and Capital Flows", An Address to CEDA Forum, 11 December 1995, page 6, based on ABS Cat. No. 5501.0.

As well, circumstances at the Commonwealth level have changed.  The new government in Canberra, while it is not conspicuously federalist, is at least nominally in favour of reforms which will improve the working of the federation.  It is also obliged by its own unsustainable fiscal situation to contemplate the return to the States of some functions;  that in turn will necessitate some reconsideration of the grants process, at the very least.  It seems likely that it will discover early in its first term that the days of transferring fiscal pain to the States without political cost are over.

The Commonwealth is also committed to a widely consultative process in the area of Constitutional reform;  and while the immediate focus is the issue of "the republic", other issues may now be raised.  It is, further, committed to continuing microeconomic and competition policy reform.  The competitive context established by the Hilmer process is now clearly dominant.  And clearly -- although it is not usually perceived this way -- the reform of federal fiscal arrangements is now the last and most challenging frontier of microeconomic reform.

It may well be, then, that the next two or three years present us with the best opportunity since the 1930s for significant reform of these eminently reformable areas.  We hope that the present project will begin to provide the advocacy that the reform so badly needs.

The structure of what follows is relatively straightforward.  In Part I, by way of background, we briefly outline the relevant issues of taxation and federalism -- on the understanding that the economic issues cannot and do not exist in a political vacuum.  In Part II, we address the two major economic reform issues in the federal context:  the distribution of taxation and expenditure powers, and then equalisation.  It needs to be admitted that the issues are complex;  and while we have tried to make the discussion as straightforward as possible (by, for instance, simplifying references), some complexities remain.  Although the key areas for decision have been discussed with a relatively open mind, there are areas where our preferences emerge;  these are summarised in the brief Conclusion.

1.2 THE TAX BACKGROUND

The realisation that Australia's taxation system is deeply unsatisfactory is not new.  From the Asprey Report of 1975, through the Draft White Paper of 1985, up to the comprehensive programme of Fightback! in 1991, there have been a number of serious attempts at tax reform at the national level.  All of them have been unsuccessful, essentially for political reasons:  political timidity, political favouritism, or political hostility.  We have instead seen a drawn-out process of piecemeal legislative change, some of which has furthered the cause of genuine reform, some of which has set the cause back.  There has been a parallel, but less public and less committed, process of projected reform at the State level which has, if anything, been less successful.

It is easy to lay down the desirable features -- economic and political -- of a satisfactory tax system.  It should be:

• efficient,
• simple,
• equitable,
• transparent, and
• politically acceptable.

On all these counts, Australia's tax system does not do a very good job.

With a multiplicity of bases, hundreds of exemptions, often exacting and onerous reporting requirements, a complex legislative foundation, and sometimes high compliance costs, the criterion of simplicity is only poorly satisfied.  With a notionally progressive income tax, the system might be thought to perform better on the criterion of equity, but even here performance is very far from ideal:  there are serious problems of avoidance by the wealthy in respect of income tax, and the State taxes are conspicuously inequitable in their overall impact.  Again, transparency is very poorly served:  only the Commonwealth's income tax scores at all well under this criterion, while the wholesale sales tax, the tariff, and most of the State taxes perform poorly.  Worse, the distinction between Commonwealth and State taxes is unclear to taxpayers, making political accountability -- to which transparency is the necessary prerequisite -- difficult.  Again, the system increasingly fails by the standard of acceptability:  evasion and avoidance, particularly of the income tax, are a serious problem, creating an uneasy social distinction between PAYE and other taxpayers.

The real problems, however, emerge when we look at the Australia's tax arrangements in the light of efficiency:  the way in which tax impacts on economic activity.

The problems are manifold, and can only be briefly alluded to here. ⁽¹⁾  Most have in common the characteristic that they cause tax to influence economic choices.

Most notable here, perhaps, is the choice of working or not working.  The income tax, with high marginal rates cutting in at relatively low incomes, operates so as to distort the choice between work and leisure in favour of leisure.  Notable, too, are the effects on the choice between consuming and investing.  The operation of the same income tax discriminates strongly against saving in general, with some exceptions like superannuation and dividend income being favoured.  Some investments are particularly favoured, such as owner-occupied housing (and, to a lesser extent, negatively-geared housing).  The process of production is largely subject to a cascade of taxes on inputs (which effectively become a tax on exports), although again there are favoured exceptions, such as the diesel fuel rebate.  These are not mere defects measured against some notional theoretical ideal:  every one of these problems has its cost -- often very considerable.

(It is also of some importance to note that many tax measures, of course, exercise their influence over economic choices in conjunction with other non-tax factors, such as the welfare system, thus increasing many distortions.)

This brief summary of the shortcomings of Australia's tax system represents what is now the consensus view:  we rely too heavily on income tax, and we tax saving and productive activity too heavily for our own long-term good.  This view is, in its essentials, 20 years old, dating back in most respects to the Asprey Report.  To that view, we can add two very important factors not always taken into account in the reform efforts of the past.

The first is the role of the State taxation systems in the overall structure of Australian taxation.  The facts are simple:  State taxes now occupy about 30 per cent of all taxes, and about 12 per cent of GDP.  On the whole, State taxes have less satisfactory bases than the Commonwealth's taxes, as the accompanying table makes clear;  and even those which are theoretically reasonable (such as property taxes and payroll tax) are in practice characterised by numerous anomalies and exemptions.  No serious study aimed at reforming Australia's tax system can afford to omit State taxation.

Table 1:  Effective Incidence of State Taxes

Type of tax:	NSW (Collins)	Vic (Nieuwenhuysen)
Payroll tax	Proportional	Mildly Regressive
Franchise fees	Mildly Regressive	Regressive
Motor taxes	Regressive	Regressive
Contracts and conveyances	Mildly Regressive	Regressive
Other stamp duties	Regressive	Regressive
Gambling taxes	Uncertain	Progressive
Land tax	Mildly Regressive	Mildly Regressive
All taxes	Regressive	Regressive

Sources:  NSW Tax Task Force, 1988;  and Nieuwenhuysen, 1983.

The second factor is the absence in our political or constitutional system of any durable constraints on the size and manner of taxation.  The public sector now occupies about 40 per cent of our economy.  Given that it has hovered around this figure for some years now, it may be thought that some "natural" limit has been reached;  that, perhaps, government in Australia will not grow much more than that.  But as the political pain of the relatively modest reduction in expenditures which accompanied the 1996 Commonwealth Budget showed, however, we can make no such assumption.  Indeed, it seems unlikely that the combination of voter demand and electoral commitment which has produced the current situation will go away.  (This is not a comment on the composition of spending;  it is perfectly possible to have an equitably-organised welfare state which occupies much less than four-tenths of GDP.)

This consideration brings us to the vital nexus between federalism and tax reform.

Most debate among policy-makers or academics about tax reform effectively takes place in a policy vacuum.  Models for better tax design -- proposing simpler, more efficient, more equitable tax -- are put forward as if those who would act on the advice -- politicians and bureaucrats -- were perfectly benevolent and disinterested guardians of the public good.  To the political realist, this was always a curious assumption, although perhaps at some stage in human history there have been such creatures.

Natural scepticism is now reinforced by the advent of public choice theory, which (oversimplifying considerably) offers a more realistic assumption or model, in which politicians and bureaucrats are rational economic actors in the affairs of government in just the same way as private individuals are assumed to be.  Politicians and bureaucrats alike may often act in such a way as to advance their own interest, quite as much as the public interest.  The rewards may not be simple monetary rewards;  they may take the form of increasing power and influence, ensuring tenure or re-election, and so on.  It is a more useful and valid view of the way the world works than the assumption of benevolence.

The discarding of the "benevolent ruler" assumption is important for all economic advice.  It is acutely important in thinking about taxation.  In a completely centralised regime, for instance, one which had no real limits on the public sector's share of the economy, the conscientious policy adviser might well think twice about suggesting any improvement to the tax regime which would improve the system to the point of actually making tax easier to raise.  At the very least, in a relatively centralised regime like Australia's, proposals for tax reform should at the very least acknowledge the need for tax restraint.

In the absence of any other explicit constitutional arrangements, a strongly federalised tax system -- with tax competition at its heart -- is probably the best way of achieving such restraint.  Australia already has a federal form;  we need only to give it substance to achieve the desired end.

1.3 FEDERALISM MATTERS

Australia is, and was founded as, a federation.

In most political debate about changes to our federal arrangements, arguments about change and reform have tended to be based on that historical fact.  This has, fundamentally, been the "States' Rights" position:  the States were endowed by the original constitutional settlement with certain rights and sovereign powers, which cannot and should not be changed.

In this view, history is all.  There is still some force in the argument -- in so far as arbitrary and continuous changes to constitutional rules do not make for good government over the long term.  That aside, it is in fact fairly weak and far from self-sufficient.  For many years it has, in any case, been discredited by the unwillingness of its proponents (not least some past State Premiers) to accept the fiscal discipline which should have been its natural corollary.

But the main problem with the States' Rights argument is that it fails as a historical explanation.

The extensive changes which have been made over nearly a century to Australia's federal structure had many causes.  Some of those causes -- such as the desire for a fully-fledged welfare state, or the apparent necessity for centralised macroeconomic management -- are not now as convincing or durable as they once might have seemed.

But however questionable the means, the end was on the whole a good one:  out of six relatively fragile colonies, we did build a nation.  It may well be that we could have gone about our nation-building in other, more federalist, ways;  on the other hand, it may well be that, for Australia and Australia's peculiar circumstances (isolation, small population, geographical dispersion), a strong dose of centralism was precisely the glue we needed to bind the federation together.

Those are questions for historians;  their relevance here is that, having acknowledged the importance of nation-building, we can be more confident now in assessing the merits of both centralism and federalism, and in choosing what we wish to retain from both.  If we now choose to rediscover more of the potential of federalism, we can be confident that our nationhood will not be compromised.

To repeat the opening statement of this section, Australia is, and was founded as, a federation.  That is an irreversible political fact.  The States will not go away;  abolishing them is not an option.  We have, however, reached the point where the workings of centralism and federalism together are producing unacceptable difficulties in the execution of public policy.  That leaves us little choice but to improve the workings of Australian federalism.

What are the difficulties?

The most obvious one is well known:  the old problem of overlap and duplication, ⁽²⁾ the waste of resources which occurs when two levels of government (sometimes three) simultaneously occupy more-or-less identical policy areas.  This characteristically takes the form of copious bureaucratic intervention at the Commonwealth level into a significant policy area still nominally under State administration -- health, education and training are the most conspicuous examples.  In smaller areas -- culture, trade, sport, for example -- there may be parallel administration.

This represents a significant waste of resources, although quantification is not easy.

Further, it leads us into considering more subtle difficulties.

Accountability is an obvious one:  who do voters hold responsible for any given policy?  How do voters express their policy choices in an area such as health, for instance?

Responsiveness is another.  Policies decided on by a large electorate will clearly have less chance of being expressive of voter preferences than those in smaller electorates.  Uniformity of policy will inevitably mean that legislation is ill-suited to local differences.

Responsibility is yet another:  if the State governments are spending moneys not raised by them, the temptation to act less than responsibly will be ever-present.

If we were to start again from scratch, bearing these sorts of factors in mind, we might well try to redesign our political system in a somewhat better fashion than the present one.  Design factors might include such elements as better accountability, responsiveness and responsibility;  and they might take into account some current intellectual political fashions, such as "subsidiarity" and "community".  All these are indeed the kinds of consideration which exercise the minds of Western Europeans as they try to construct a new Europe.  We, fortunately, are spared that sort of radical and laborious political overhaul, since much of what we might want is already available to us.  All we have to do is take advantage of what our system already offers us.

All we have to do, in fact, is rediscover the virtues of federalism.

Federalism represents the final development of the Western liberal tradition of government. ⁽³⁾  The elements of that tradition are well-known:  representative democracy, sovereignty resting with the people, the rule of law, equality before the law, and the recognition of and respect for a number of basic rights and freedoms.

The circumstances of the American Revolution -- and to a certain extent the model of Switzerland -- added one more layer to that tradition:  the notion which we in Australia understand as "checks and balances".  Deeply imbued with a sense of republican history and theory, and profoundly concerned to construct a republic which would survive the vicissitudes which had caused other republics in history to decline into chaos or tyranny, the American founders added institutional structures which would guarantee the division of powers.  Their reading of history -- a pessimistic but accurate one -- was that political man tended toward tyranny.  Personal tyranny, of course, the tyranny which bred emperors and dictators;  but equally the tyranny of the majority, of ideas, of religion, of self-interest.

The converse of that fear of tyranny was a concern for freedom and for individual freedoms.  Not least of these is economic freedom.  In the words of one recent study of economic reform, "Economic freedom is a good in its own right.  Most people value the freedom to choose among rival suppliers and to accumulate, inherit and dispose of private property in the light of their own judgements, values and interests, within a framework of morality and law.  In this sense, any increase in economic freedom is intrinsically valuable". ⁽⁴⁾

The solution devised by the American founders was to balance the parts of the republic against each other, carefully separating legislature, executive and judiciary all from each other.

That was not necessarily a new idea;  what was new was the division of the powers allocated between the levels of government.  The federal government was left with those matters which most deeply concerned the national common interest, while virtually everything else rested with the states.  Power -- that dangerous commodity -- was as widely and evenly dispersed as was possible in the real world.

That is the essence of federalism.  But it only prescribes the forms (however important), and says little about the nature of federal governments.  It turns out that federalism has a number of important incidental benefits, some of which have already been foreshadowed:  responsibility, responsiveness and accountability.

In Australia's present circumstances, other benefits are more than relevant.  Federalism, for instance, offers a diversity of policy which not only better represents voter preferences, but also improves the ability to experiment without major risk.  In the context of ongoing microeconomic reform, this is very useful.  One State may decide to reform its delivery of public health services in one way, another State in another way.  One may decide that the privatisation of utilities is the best policy, another may decide on public ownership with accompanying disciplines.  Smaller-scale experiments in policy are more likely to yield better and more appropriate solutions, and at far smaller risk than full-scale national experiments.

Another benefit often overlooked is the way in which federalism can work to diminish the power of special interests.  Put briefly, the costs of policies designed to favour special interests are shared over all voters, while the gains are delivered to relatively few.  In a large polity, the costs to each voter are relatively slight and relatively invisible, while the gains may be considerable.  Decentralising policy makes the costs relatively greater and more visible.  This is why Canberra has been the preferred arena for so many interest groups -- environmental, welfare, arts, and so on -- over the last 15 years or so.

This catalogue of virtues could be extended considerably.  But what is most important in all this is that, properly constituted, federalism offers the voter and the citizen choice.  Citizens and voters within one federal nation have a choice of jurisdictions in which to live and work.  Either by formal voting, or by voting with their feet, citizens within federations can, over time, create different policy jurisdictions;  jurisdictions which reflect not only the given characteristics of natural and human resources, but also different tax and spending regimes, welfare regimes, or environmental regimes, or any of a host of different policy preferences.

Seen in this light, federalism is a kind of competition policy for government.  Government by its very nature displays all the characteristics of monopoly behaviour, with the added twist that it is sanctioned by the rule of law.  But once choice is introduced, competition follows.

Given the steady and apparently irreversible growth of government in Australia, this may be of some importance to us:  it may be that federalism, through competition, offers us the only realistic option for stopping the unstoppable growth of government -- or, at the least, for making government efficient.  It is also the best available way of developing and guaranteeing our necessary economic freedoms.

1.4 PROBLEMS OF AUSTRALIAN FEDERALISM

Although the Australian Founders were well-acquainted with the workings of federalism in the United States, Canada and Switzerland, they chose a weaker form of federalism than their models offered.  They chose, in particular, a less than complete separation of powers, with the executive and the legislature more or less united.  They specified the powers of the Federal government, without specifying those of the States.  They specified that in a conflict of laws, the Federal law would prevail.  They allocated considerable tax powers to the Federal government.  And, although it could not reasonably have been foreseen, the judicial creativity of the High Court enabled the checks and balances to be very seriously eroded.

Some of the effects of this have already been noted;  more are examined in detail below.  To state the matter clearly, however, we can say this:  that, by damaging the vitality of Australia's federal arrangements, we have simultaneously damaged the quality of government, and denied ourselves the benefits of competition.  Even under the relatively weak form of federal competition which prevails in Australia, there is competition between the States, competition, in particular, over various aspects of the tax base, or over fairly blatant subsidies to selected investment.  This is not usually very productive;  at best, a pale shadow of the competition that a well-constituted federation can offer.

1.5 GOVERNING BY THE RULES

While the immediate purpose of this paper is to examine ways of restoring some authentically federal proportion to our overall scheme of taxation and expenditure, that is only a part -- although a very important one -- of the wider task of recovering the best that federalism has to offer.

A part of that wider task is the restoration of the importance of constitutionalism in Australian government.

One of the deeper flaws of Australian federalism is that it generates uncertainty.  The most obvious example of this is the annual spectacle provided by the negotiations surrounding the annual allocation of Commonwealth revenues to the States.  (This year's was particularly notable for the fact that what was basically at issue was not the level of the Commonwealth's generosity, but the level to which the States were willing to assist in the Commonwealth's need to reduce its own deficit.)  This uncertainty -- over negotiations involving about one-third of the States' revenues -- is completely counterproductive.  It makes it very difficult for the States to plan even one year ahead;  and, in turn, generates uncertainty at the State level about changes in taxes and expenditures, adding to the instability of the State tax base.

That is not all.  Because the Commonwealth also occupies a large number of policy areas, and tends constantly to expand its policy reach, there is a wider uncertainty generated by its freedom to expand.  It is very difficult for the States to maintain a consistent policy direction over a wide number of policy areas -- housing, health, education, industrial relations, training, the environment, sport and the arts, Aboriginal welfare, transport -- without having to allow for arbitrary, and often conflicting, Commonwealth intervention.  This amounts to a kind of sovereign policy risk -- and the cost of the risk is ultimately, of course, borne by State taxpayers and voters and consumers of services.

For these and other reasons -- not least the need to secure the conditions of authentic competitive federalism -- we need to establish (or re-establish) a framework for Australian federalism which is bound by relatively simple, and relatively few, rules.  Such rules can be formally enshrined in the Constitution, or can be the less formal result of co-operation or mutual agreement.  Model rules have already been suggested which would fit the Australian context. ⁽⁵⁾  This is not the place to canvass or develop these models.  It can be said, however, that reform which will lead to a closer and better re-alignment of taxation and expenditure will be a very good start to this important long-term goal.

---

PART II:  THE ECONOMICS OF FEDERALISM

SECTION 1:  THE DISTRIBUTION OF TAX POWERS

1.1. INTRODUCTION

To the economist, federal systems can be thought of as a compact between autonomous political units coming together to create a central government, to which selected powers are ceded so as to secure the benefits from co-operation on matters of common interest.

Three types of common interest have been identified.  First, they can derive directly from the provision of public goods by governments.  Such public goods may include defence and foreign relations.  Indeed, common interests over defence were important in the US federation process.  Second, there can be trade-related common interests, where the benefits of co-ordination are obtained through a customs union (federation) between states.  This is thought to be one of the main motives for Australian federation.  Third, common interests may emerge when factors of production (such as labour and capital) migrate across political boundaries.  These forms of interdependence between states provide rationales for federation.  The implication is that sub-national governments within the federal union are left to make independent, diverse and competitive decisions over taxation and expenditure policy in all areas where there are no shared interests.

These arguments centre around the concept of externalities:  that is, the creation of benefits, positive or negative, which spill over from one jurisdiction to another.

The fiscal federalism literature also uses the externality argument to find economic principles for the allocation of spending and taxation powers in existing federations. ⁽⁶⁾  Again, the general idea is that if states undertake policies for which there are externalities, their policy decisions may be inefficient.  This is taken to justify a role for the central government;  either in direct provision of public goods or the control of tax instruments, or through other means such as matching grants or inter-state equalising transfers.

Applied to the assignment of taxation powers, these arguments imply that tax bases which are prone to significant spillovers or externalities (usually due to the mobility of capital or labour), or which affect inter-state or overseas trade, ought to be assigned to the centre.  This leaves states with all the remaining taxes.

In Australia, the centralisation of taxation powers has gone beyond what is suggested by these assignment principles. ⁽⁷⁾  One index of this is shown in Chart 2, another in Table 2.  This is mainly because the States are excluded from the income tax base, while the High Court's interpretation of Section 90 of the Constitution has excluded them from much of the consumption tax base.  This contrasts strongly with the US, where the States are barred from imposing taxes which might impede free inter-state trade or external trade policy, leaving them free to impose most consumption, personal income and corporate taxes.

Chart 2:  Index of Sub-National Autonomy

Source:  A. Shah, The reform of Inter-governmental Fiscal Relations in Developing and Emerging Countries, Policy and Research Series No. 23, World Bank, Washington, 1994.

Table 2:  Australian Federation -- Trends in Revenue Raising and Spending Power

	1901-02		1993-94
	C/wealth	States	C/wealth	States
Share of Revenue	58%	42%	75%	25%
Share of Expenditure	13%	87%	54%	46%

Source:  Rebuilding the Nation, Richard Court, MLA, Premier of Western Australia,
February 1994, page 8.

Drawing on the principles for tax assignment provided by the fiscal federalism literature, we discuss the disadvantages (Section 1.2) and advantages (Section 1.3) of Australia's centralised tax powers.  This is supplemented by a case study of the distribution of tax powers in the US (Appendix 2), which provides a valuable comparison with Australia.  Finally, we examine the options for reform of Australia's distribution of tax powers and the issues raised by these options (Section 1.4) and draw conclusions in Section 1.5.

1.2. THE DISADVANTAGES OF CENTRALISED TAX POWERS

There is increasing concern about the adverse consequences of highly centralised taxation powers in Australia.  We now provide a brief survey of these concerns. ⁽⁸⁾

Less tax competition

Arguably the biggest economic cost of centralised taxes is that, by reducing tax competition between states, such a regime works like a cartel in private markets, where producers collude to raise price and restrict supply against the interest of consumers.  The OPEC oil cartel worked much like this in the 1970s.

In the context of federalism, it has been argued that, since Federation, the Commonwealth and the States have, effectively, colluded to centralise tax powers, eliminate tax competition, and raise taxes above what they would be in a more decentralised and competitive federation.  In other words, by eliminating competition, the tax cartel allows the public sector to pluck more feathers from the goose, making the public sector bigger than it otherwise would be.

A cornerstone of this view is the idea that governments are not beneficent, and in fact pursue their own self-interests, such as maximising the size of bureaucracies, the budgets under the control of ministers, their power and prestige, and the demands of interest groups.  In other words, governments are not benign and beneficent custodians of the public interest and will, once in power, pursue their own self-interests.  As noted in the introduction, this is our (realistic) view of how governments behave.

It has been vigorously argued ⁽⁹⁾ that a strong federation with decentralised tax and policy-making powers is the most effective constraint on self-interested governments.  When powers are centralised this constraint on government is weakened, making citizens or voters worse-off.  The conclusion here is that one might expect to see a lower tax and spending regime, with correspondingly smaller government, in strong federations with decentralised tax powers than in unitary countries or weak federations.

The States have been major beneficiaries of the tax cartel, receiving higher revenues (in the form of grants from the Commonwealth) than they would otherwise receive if they had to impose the taxes themselves in a competitive environment.  The nature of the tax cartel and the States' co-operation in the arrangement for most of the post-War period is captured in the "State tax officials' poem", supposed to have been written following a conference on the allocation of tax powers (and preserved for posterity by Bert Kelly):

We thank you for the offer of the cow,
But we can't milk so we answer now
We answer with a loud emphatic chorus,
You keep the cow and do the milking for us. ⁽¹⁰⁾

Of course, as we would expect, and can show from the theory of collusion, such arrangements are only stable if the surplus revenue from co-operation is fairly distributed among the participants;  in this case, in the form of Commonwealth-State transfers.  Without mutually-agreed distribution of the spoils, the arrangements can be inherently unstable.  Indeed, one can argue that the Commonwealth's action during the 1980s and 1990s of reducing grants to the States in real terms, while continuing to expand its own public sector, violated the agreement implicit in the tax cartel -- that is, the Commonwealth started taking too much of the surplus arising from collusion -- and thus precipitated the recent State demands for greater tax freedom and a decentralisation of tax powers.  Given this treatment, it is doubtful that the States would echo the sentiments of the State tax officials' poem today.

Vertical fiscal imbalance

While tax powers are highly centralised, policy-making powers remain relatively decentralised.  The States retain responsibility in major areas of public policy such as health, education and law and order.  Therefore the Commonwealth collects correspondingly more than it needs for its own expenditures and the surplus is returned to the States as grants, to enable them to meet their spending commitments.  This phenomenon -- a gap between expenditure and revenues at the State and Federal levels -- is known as a fiscal gap (or vertical fiscal imbalance -- VFI).  Although it can be argued that some fiscal gap is desirable, and although it is a feature of most federations, the fiscal gap is much bigger in Australia because our tax powers are more centralised. ⁽¹¹⁾

The large fiscal gap in Australia means that the Commonwealth makes extensive tax reimbursement grants to the States.  This divorces spending decisions from taxing decisions, and strongly suggests the argument that voters have difficulty in knowing which level of government is responsible for the provision of services.  This allows State governments to "pass the buck" to the Commonwealth for policy failures and accountability.  The Commonwealth, too, has honed its skills in attributing the blame for failure of its policies to the States.

The fiscal gap creates "fiscal illusions" in the perceptions of voters, bureaucrats and politicians regarding the true costs of spending and policy decisions.  Walsh illustrates the point by proposing that the gap creates the impression in voters' minds (the fiscal illusion) that the States spend "60 cent dollars", implying that voter demand for State spending programmes is higher than optimal. ⁽¹²⁾

Criteria for evaluating tax systems

As already noted, the economic criteria for evaluating the desirability and viability of a tax system include efficiency, equity, compliance and administrative costs, transparency and acceptability.

The Commonwealth's dominance of the two major broadly-based taxes means that the States are forced to rely heavily on a relatively narrow tax base, which includes payroll tax, property taxes, financial taxes, selected excises and levies, gambling taxes, insurance and motor vehicle charges and franchise fees.  Since the 1980s, this has become worse because the Commonwealth has cut back on grants in real terms, thus pushing the States to rely more heavily on this unsatisfactory base.  This move was motivated by the Commonwealth's desire to "solve" its own budgetary problems, and has been controversial.

The narrow State tax base is for the most part highly distortionary, increasing the total efficiency cost of the tax system, and is also more distortionary than the Commonwealth tax base.  Access Economics have promoted this view, with their estimates of the excess burdens of State versus Federal taxes (using their AE-CGE model).  Their conclusion is that

.... in accord with the generally accepted view, ... Federal taxes are less distortionary than State taxes. ⁽¹³⁾

The most distortionary State taxes are business franchise fees, motor vehicle registration charges, land taxes, and all other stamp duties -- which include taxes on the banking and investment industries, residential construction and the wholesale and retail trade. ⁽¹⁴⁾

The narrowness of the States' tax base has forced the States into taxing highly mobile tax bases (stamp duties on financial transactions, for instance).  A considerable proportion of State revenue is also potentially open to legal (Constitutional) challenge (the fuel, tobacco and alcohol franchise fees, for instance). ⁽¹⁵⁾  This increases the uncertainty associated with State revenues.

The evidence shows that many State and Federal taxes are highly distortionary (the Wholesale Sales Tax at the Federal level, for instance, and stamp duties on highly mobile bases at the State level), and that through judicious changes to the tax mix, giving the States access to more efficient taxes, the total efficiency cost of the tax system could be reduced.

When assessed against the other criteria for evaluating a tax system, many State taxes perform equally poorly.  For example, most are regressive and do not meet the goal of equity. ⁽¹⁶⁾  On the other hand, the administrative costs of State taxes compare reasonably favourably with the Commonwealth's personal income tax, with the exception of land taxes and stamp duties, which have relatively high collection costs per dollar of revenue raised.  Compliance costs are also high for some State taxes.  It is claimed, for example, that the costs of collecting stamp duty on cheques in Western Australia can exceed the revenue raised. ⁽¹⁷⁾  But for other State taxes, in particular the payroll tax, compliance costs are relatively low, even compared with the federal personal income tax. ⁽¹⁸⁾  Transparency is also a major problem for many State taxes.

Two State taxes which are often seen as relatively efficient, and reasonable when assessed against the more general tax criteria, also suffer difficulties.

One, the payroll tax, has the problem that the Commonwealth's Superannuation Guarantee Levy (the SGL) essentially taxes the same base.  Given that the SGL is already set to rise to 12 per cent, and probably to higher levels in future years, this raises questions about vertical tax competition, and whether there is any real scope for the States to make further use of this particular base.

The only other half-decent tax available to the States is the land tax;  but it is not generally realised that it is in fact more than fully exploited already.  State and local governments impose a multiplicity of taxes on land values -- apart from the land tax itself -- including metropolitan improvement taxes, fire brigade levies, stamp duties, water and sewerage charges, and local government rates.  In this broader sense, the total land tax take in 1994-95 (excluding water and sewerage charges) was about $10 billion, and equivalent to the revenue collected by the Commonwealth wholesale sales tax.  Land tax is by no means a perfect tax, even at rates lower than those levied in Australia.  It is very costly to administer -- the most costly of any tax -- and it creates severe distortions for asset-rich but cash-poor people.

Overall, then, State taxes are a hotchpotch of relatively inefficient and regressive taxes with high compliance costs -- with payroll tax being a possible exception.  Many of the State taxes are also highly unpopular for these same reasons, and hence fail on the notion that a desirable tax system should have wide acceptance in order to avoid the development of inefficient evasion/avoidance schemes.  Of course the reason that State taxes are in such a mess is that the more economically desirable taxes have been taken by the Commonwealth -- the income tax and most indirect taxes -- forcing the States to raise revenue from what is left.

State budget uncertainty

The process by which grants are determined each year creates a high degree of uncertainty, which restricts the ability of the States to prepare medium-term budget strategies.  The Federal government unilaterally decides upon the total pool to be distributed, and ignores or adjusts at will the relativities announced by the Commonwealth Grants Commission (CGC).  Further, in recent years, the Commonwealth has looked on the grants component of its own budget as a variable which may be conveniently adjusted to meet its own budgetary objectives without regard for the States.  Moreover, the States' limited access to taxation revenues has often forced them to adopt creative accounting techniques, such as the off-budget financing of capital projects.

Overlap and duplication

Finally, the financial powers of the Commonwealth have helped it encroach on the policy-making powers of the States, powers which were left to them (implicitly) by the Constitution.  This has created further economic costs (discussed below), ⁽¹⁹⁾ and has led to unnecessary duplication and overlap of functions. ⁽²⁰⁾

1.3. THE ADVANTAGES OF CENTRALISED TAX REGIMES

While there are undoubted costs associated with centralised tax powers, there are also well-known arguments about the efficiency and equity costs of a decentralised tax regime in a federal economy.

Fiscal stabilisation policies

The key argument for strong central tax powers is that they allow the central government to carry out effective fiscal stabilisation policies, and, further, that if states have sufficient fiscal autonomy to implement their own stabilisation policies, the policy of one state will create spillover effects, or externalities, for other states, with associated efficiency costs.  (The alleged effects become clear when we think of real-world examples:  a strongly expansionary or contractionary fiscal policy run by New South Wales, for instance, might leak over into other States, and cause their economies unwanted problems, or perhaps undeserved free benefits.)  This has been taken to justify the central government having control over the major tax instruments important in stabilisation policy. ⁽²¹⁾

But there are compelling counter-arguments. ⁽²²⁾  First, it is no longer clear that pursuing counter-cyclical fiscal targets is desirable, especially in a world of floating exchange rates.  Second, it has recently been argued that the need for stabilisation has no implications for the assignment of taxation and spending responsibilities. ⁽²³⁾  This argument is based on the following:

• Evidence from research on the European Union (EU) suggests that effective stabilisation can be accomplished by a central government with very limited tax powers -- certainly much less than the Commonwealth has.
• On the basis of empirical analysis of Australia and Canada, the most effective stabilisation takes place through automatic stabilisers.  The discretionary component is pro-cyclical.  This means that regionally-sensitive stabilisation occurs regardless of which level of government has the tax or spending powers, and that assignment has no impact on the effectiveness of stabilisation policy.
• Evidence from other federations, again principally the US and Canada, suggests that decentralised taxation powers provide no barrier to the successful conduct of national fiscal policy. ⁽²⁴⁾

Alternatives to fully-centralised stabilisation

But even if we assume that discretionary stabilisation is desirable, and take the original arguments in the federalism literature about fiscal spillovers as given, this may still only imply a need for fiscal policy co-ordination, not central control.  The basis for this co-ordination is already present in the form of the National Fiscal Outlook and the latest Australian Loan Council Arrangements.  If tax powers were returned to the States, it might be necessary to expand this co-ordination to allow the Commonwealth, the States and the Territories to promote short- and medium-term fiscal objectives, and to decide on how to co-ordinate their fiscal policies.  Walsh has suggested, as an alternative, that the Australian Loan Council be renamed and expanded into a National Fiscal Co-ordination Council to fulfil this role under any plan to reallocate tax powers and co-ordinate fiscal policy. ⁽²⁵⁾

Co-operative fiscal policy also has a distinct advantage over centralised policy:  it would allow much greater flexibility to cater for State diversity in macroeconomic fluctuations.  While national policy could be set to meet nation-wide objectives, State fiscal policy could be adjusted to meet local conditions.  This is a flexibility not catered for by the current, fully-centralised stabilisation policy.  We conclude that the possible need for discretionary fiscal policy does not necessarily justify centralised tax powers, most of all because it takes inadequate account of the potential for co-ordinated fiscal policy. ⁽²⁶⁾

Income redistribution

It is usually said that the central government may need strong tax powers to implement income redistribution policies.  The standard argument is that state income redistribution policies are self-defeating.  States which undertake greater redistribution than others attract low-income migrants and repel richer residents, thus undermining the intended effects of the policy.  As with stabilisation, state income redistribution generates spillovers which create inefficient outcomes.  Some take the view that income redistribution undertaken by states will be less than what is socially desirable.

On one view, this justifies centralised income redistribution and centralisation of the associated tax instruments. ⁽²⁷⁾  But, as with fiscal policy, these arguments do not necessarily imply that -- in the case of Australia -- the Commonwealth needs to collect 75 per cent of all public sector revenue:  it is likely that income distribution goals can be achieved by the centre with a much smaller fiscal presence than at present.  And to the extent that spillovers are created by sub-national redistribution, inter-governmental co-ordination again has the potential to achieve efficient outcomes, and may be an alternative to central control.  But there is still likely to be some efficiency and equity role for the Federal government if the States co-ordinate their income redistribution policies. ⁽²⁸⁾

Thus, the need for income redistribution does not necessarily justify centralised tax powers;  once again because there is potential for co-ordinated outcomes between States, monitored and facilitated by the centre.

Constitutional constraints on State tax powers

The High Court of Australia has used a set of complex arguments to justify its decisions in respect of Section 90 of the Constitution, and the consequent exclusion of the States from the consumption tax base.  We consider these arguments here, since they represent a defence of the current centralised regime and highlight some additional possible costs of decentralised tax powers.  In this regard, Section 90 reads:

On the imposition of uniform duties of customs the power of the Parliament to impose duties of customs and of excise, and to grant bounties on the production or export of goods, shall become exclusive.

The motives of the Founders and the interpretation of the meaning of "excise" were not specified in the Constitution, and have been difficult to glean from the Constitutional debates.  This omission is possibly one of the greatest gaps in the Constitution.  It has meant that the High Court has been left with the role of interpreting what is to be included under the term "excise", and hence, effectively determining the distribution of tax powers.  This has conferred enormous influence on the Court.  It has helped determine the strength of the States vis-à-vis the Commonwealth, and, as we shall see below, in a way which has diminished the States' role.  Indeed, the role of the Court has been highly controversial, and the issue of excise hotly debated.

Much of the debate has been over the Founders' motives and the meaning of excise.  It turns out that the various views on the meaning of excise depend on one's view of the motives behind Section 90, so that some, for instance, have argued that it was indeed intended to give the Commonwealth a broad general power over consumption taxes for more-or-less macroeconomic reasons.

Our view, a view probably held by most economists, is that the federal tariff, inter-state trade and federal unity motives were behind including the phrase "and of excise" in Section 90, and that excise should, therefore, be interpreted narrowly, as a discriminatory tax on local production, and not given the broad meaning ascribed to it by the High Court.  This interpretation would leave the States free to impose a general consumption tax (a retail sales tax, for instance), and makes the current broad interpretation of excise by the Court difficult to sustain on economic grounds.  (This is, however, a complex matter;  it is discussed at greater length in Appendix 1 to this paper.)

We do not believe, then, that the High Court's arguments, which lock the States out of the consumption tax base, are a valid defence of centralised consumption tax powers.  As we argue below, there is a case for the Court to review its decisions on Section 90.

Adequacy of public spending and taxation

We have argued that the main benefit of a decentralised tax regime is that it allows beneficial tax competition to constrain governmental powers by breaking up tax powers among competing governments.

Some hold the view that if the States were given greater tax powers, then such competition would constrain the ability of State governments to provide "adequate" or optimal levels of expenditure on public services and infrastructure.  This view has been expressed by the Commonwealth in the past, as shown by the following comment from Senator Bob McMullan in 1992:

One source of concern is that tax competition amongst different States may result in competitive bidding-down of tax rates while the same forces of inter-state competition may prevent a commensurate reduction in the level of services provided.  The inevitable consequence of such an outcome would be an increased call on the community's savings by the public sector or a need for higher levels of Commonwealth financial assistance to the States. ⁽²⁹⁾

There is a literature which shows that this argument may be valid if one adopts the view of government as a beneficent and benign entity implementing policies in the collective public interest.  In such a world, competition between governments over taxes in a decentralised federation may result in "under-provision" of public services relative to a theoretical optimum. ⁽³⁰⁾

But benevolence and governments acting in the public interest are exacting requirements not often met in practice.  As we have argued, in our view governments can also be self-seeking, and pursue their own interests, which bear little necessary relation to voters' best interests or the theoretical ideal of an optimum.  In such a world, taxes will be too high in relation to any optimum;  and constraining tax powers through a decentralised tax regime with competition between the states will bring public policies closer to the optimum -- not further away from it.  A decentralised tax regime acts as a brake on "over-spending" by government.

Even if one believes the beneficent view of government behaviour, it has been shown that inefficiencies created by competition between States would be minimal if States were given access to a broadly-based retail sales tax or income tax -- as is proposed below. ⁽³¹⁾  This is also borne out by evidence from the US and Canada.  Further, access to such broadly-based taxes would alleviate pressure on the States to attempt to tax highly mobile tax bases -- something which sub-national governments in federations ought not do on efficiency grounds.  This would also give the States a means of competition, better than handing out locational subsidies to industries.  Inefficiencies from competition could also be minimised by intergovernmental coordination over the relevant bases.

(Dis)economies of scale

Finally, there may be significant diseconomies of scale from decentralised tax collection, leading to higher per unit administration costs.  This is often cited as a benefit of centralised tax collection.  This is something we may expect on conceptual grounds, and has been confirmed in some empirical studies.  It is often why we find higher-level governments in federations collecting tax revenues on behalf of lower-level governments.

But as the discussion above on administrative costs shows, there is no evidence that the costs of collecting existing State taxes are any higher than those of federal taxes.  This casts some doubt on the importance of scale economies.

Summing up ...

In conclusion, decentralised tax regimes in federal economies can produce economic inefficiencies and adverse equity effects due to various externalities.  For these reasons, some centralisation of tax powers will always be desirable in a federation.  But the presence of these distortions should not be taken to imply that centralisation is the only way of correcting for the inefficiencies and inequities.  To do so is to overlook the potential for policy co-ordination, perhaps facilitated and monitored by the central government if self-enforcing agreements between states are not feasible.

In addition, once the costs of central tax powers are considered, especially the potential for the abuse of monopoly taxing powers by the centre, it may often be better to bear these costs rather than minimise them through central intervention.

As an interesting point of comparison, a closer look at the tax powers of the American states is included in Appendix 2 to this paper.

1.4. REFORM OPTIONS

The adverse consequences of Australia's tax assignment arrangements have led to claims that the federal system needs to be reformed, with a reallocation of tax powers back to the States.  These claims are based on an assessment that the costs of such a highly-centralised tax assignment outweigh the benefits.  There are four feasible policy options, most of which have been canvassed in the past.  They are:

• to reduce the spending powers of the States further, or to get rid of them completely;
• to introduce tax revenue-sharing arrangements on the income or consumption tax bases;
• to return income or consumption tax powers to the States;  and
• to expand existing relatively efficient State taxes.

The discussion of tax assignment in Sections 1.2 and 1.3 suggests that these policy changes should be assessed against the following criteria:

1. ensuring that administrative and compliance costs of the taxation system are no higher than at present (and, preferably, lower);
2. reducing the total excess burden of the tax system by changing the tax mix in favour of more efficient taxes (assuming that there is scope for this);
3. allowing stabilisation and income redistribution policy to be at least as effective as at present (and preferably more so);
4. increasing competition in the federal system;
5. improving political accountability and responsibility, and reducing voter fiscal illusion over tax prices (if this exists);
6. allowing States to implement medium-term budget strategies by introducing more stability into Commonwealth transfers;
7. reducing undesirable overlap and duplication;
8. being at least as equitable as the current tax system, and not imposing the costs or benefits of adjustment on any particular group;  and
9. being feasible, politically and constitutionally.

We also believe that in the current macroeconomic environment in Australia, reform to the distribution of tax powers should also:

• be revenue-neutral (that is, they should not raise tax revenues in total);  and
• result in no increase in State or Federal budget deficits.

We now assess each of our four reform options, using these criteria (1) to (9).

To make the assessment clearer, and to assist the reader in following the discussion of the criteria, we present the assessment in abbreviated form below.

	Unitary System	Tax sharing	Return Income Tax	Return Consumption Tax	Expand Existing State taxes
1. No increase in compliance or admin. costs	***	***	***	*	*
2. Greater economic efficiency	*	**	**	***	***
3. No macro policy effects	***	***	***	***	***
4. Increase beneficial tax competition	*	*	***	***	**
5. More accountability	***	**	***	***	***
6. Better State budget strategy	n a	**	***	***	*
7. Less overlap/duplication	***	**	**	**	**
8. No neg. equity effects	***	***	***	**	*
9. Feasibility	*	***	***	*	***

Note:
*** denotes high achievement of the specified goal;
** denotes fair achievement;  and
* denotes limited or zero achievement.

1.4.1 UNITARY SYSTEM

Further centralisation (or, in the extreme, complete abolition of the States) meets criteria 1, 3, 5 and 7.  But it does this by getting rid of the federal system and the ability to cater for diversity and to control the abuse of power by government.  Since we believe these to be significant benefits, and since it totally fails on the criterion of political feasibility, this "throwing the baby out with the bath water" option is not considered any further.

1.4.2 TAX REVENUE-SHARING

A tax revenue-sharing arrangement between the States and the Commonwealth could be introduced on the consumption or income tax bases.

There have been proposals along such lines going back to the 1970s.  They have been discussed and analysed elsewhere. ⁽³²⁾  Under tax-sharing, States would continue to receive grants from the Commonwealth, but the size of the grants would be determined as a share of the revenue from a nominated Commonwealth tax base.  Some have even suggested that the share going to the States be stated on taxpayers' assessments.

How does tax revenue-sharing rate against our criteria?  Clearly, it meets 1 and also 3, since the Commonwealth still controls the nominated tax base.  It may go some way towards meeting 5, though only in a limited sense.  Finally, tax-sharing would meet 8 and 9, and possibly 2, depending on how it was designed.

But tax-sharing fails to increase beneficial tax competition -- it retains the tax cartel, but in a slightly different guise -- or to reduce overlap and duplication.  While it offers some benefits over existing arrangements, the welfare gains are modest compared with the other reform options, to which we now turn.

1.4.3 REALLOCATING TAX POWERS

There are two possibilities here.  First, the States could be given a new broad-based consumption tax (a retail sales tax, for instance).  Second, the States could be given income tax powers.  Proposals for both have been put forward in Australia. ⁽³³⁾  It must be noted here that the possibilities are not mutually exclusive.  Indeed, any plan to devolve tax powers to the States may well involve some combination of expanded income and consumption tax powers at the State level.  In addition, such a plan might also include the expansion of the more efficient existing State taxes.  We now assess these two possibilities against our criteria.

Option A:  Income Taxes

In recent years a number of similar proposals have been put forward for returning income tax powers to the States.  In general, returning income tax powers has been a more popular option than allowing the States into the consumption tax base, principally because of the difficulties arising from the Constitutional prohibition.

It needs to be noted here -- to avoid any possible confusion -- that "income tax" refers only to the personal income tax;  the assignment of corporate taxation is not at issue.

The States' Tax Plan

During 1990-91, the States and the Commonwealth put forward a plan to hand back income tax powers to the States.  Although the first three years of the so-called States' Tax Plan looked more like tax revenue-sharing than a devolution of taxing powers, the arrangements proposed for the period beyond the initial three years had the hallmarks of a genuine reallocation of tax powers.  Key elements were that:

• the States would get an agreed share of personal income tax revenue to be collected by the Australian Tax Office in the first three years.
• the State tax rate for the first three years was set at 6 cents in the dollar, uniform across States.  After three years, the States would be allowed to vary their rate, with or without consultation with other States, but there was to be consultation with the Commonwealth when changes in tax rates, thresholds or the tax base were being contemplated.
• revenue neutrality, budget neutrality for the States and the Commonwealth, and tax liability neutrality for individual tax payers were guaranteed, at least initially, by the Commonwealth making a proportional cut in its own income tax rates (thus effectively making room for the States), and also cutting its general revenue grants (that is, for every dollar of State income tax revenue, there would be a cut in Commonwealth income tax of one dollar and a reduction in grants of one dollar).

The States' plan failed to get off the ground for political reasons and has been shelved;  though interest in the basic ideas remains, particularly among the States.

Walsh ⁽³⁴⁾ proposes a somewhat similar scheme, although his plan has no three-year period of tax sharing in which inter-state tax competition is suspended and budget and taxpayer neutrality are assured.  Elements of his proposal include:

• co-operation between the States to adopt common Income Tax Assessment Acts, so that there would be a common tax base, and common assessment of allowable costs of earning income, exemptions and rebates, so that there would be no competition over the tax base.
• a cut by the Commonwealth in its personal income tax rates across the board, with corresponding cuts in State grants -- a one dollar cut in Commonwealth tax revenue combined with a one dollar cut in grants to the States -- which would make room for the States in the income tax base.
• freedom on the part of the States, within agreed limits, to determine a flat rate of tax to apply to the incomes of their own residents above an income threshold jointly agreed with the Commonwealth.
• retention of centralised collection of revenue through the ATO.

Access Economics ⁽³⁵⁾ propose a return of income tax powers along these lines as well.  Their proposal includes:

• a reduction in Commonwealth income tax rates by a constant marginal rate above the tax-free threshold of fifteen percentage points.
• introduction by the States of their own income tax at a constant marginal rate above the Commonwealth tax-free threshold.
• the State rate to be higher or lower than the reduction made by the Commonwealth (so no need for individual tax payer neutrality)

The Business Council of Australia (BCA) has also suggested that a State income tax be introduced, with the Commonwealth making room on the base for the States. ⁽³⁶⁾  They suggest that the additional revenue be used by the States to reduce some of their more distortionary taxes, such as stamp duties.

Rating the proposals

Clearly, there is much similarity between all these proposals.  How do they rate against our criteria?  As Walsh notes, his scheme is closely related to the system in Canada whereby the Provinces "piggy-back" on the central government income tax system.  It is designed specifically to allow tax competition between the States over the rates they set, but not over the tax base.

The proposals therefore potentially score highly on 4, because they allow beneficial tax competition.  By directly reducing the fiscal gap, the proposals also increase accountability and responsibility, and have the potential to reduce overlap and duplication.  Compliance costs are unlikely to change (assuming that there is no change in the tax mix, as discussed below);  and although administrative costs of the ATO may rise, this could be offset by a fall in costs associated with the grant process.  Such schemes would also be revenue- and deficit-neutral.

Stabilisation and income distribution goals can still be met, because the Commonwealth still has a substantial presence in the tax base and still determines the degree of progressivity in the rate structure.  As noted above, there is the potential for inter-governmental co-operation on fiscal policy to sort out these co-ordination issues.

Thus, such proposals are likely to yield significant federalism benefits because they reduce the fiscal gap, increase beneficial tax competition (while minimising harmful tax competition over the bases), and cut overlap and duplication.

What are the issues?

What issues are raised by reallocating income tax powers?

Improving the benefits:  changing the mix

There are doubts over whether these proposals offer any potential for tax-mix benefits.  This can be seen from considering a simple example.

Suppose the Walsh plan is implemented, and that for every dollar of income tax revenue raised by the States, the Commonwealth reduces its tax take by one dollar to retain individual tax payer neutrality.  We take this need for neutrality as an assumption in what follows, so that a dollar of income tax revenue previously raised by the Commonwealth is now raised by the States.

Assuming that federal spending is fixed, the Commonwealth has two options.  First, it can reduce grants to the States by one dollar.  This keeps the Federal deficit fixed, because the dollar of tax revenue forgone is regained as a dollar less in grants to the States (the policy is then "federal deficit neutral").  For the States, a dollar in grants forgone is gained as a dollar of State income tax revenue.  For given State spending, State deficits are unchanged also.

But this offers no prospect for the States to reduce any of their existing inefficient taxes.  There is simply a one-for-one substitution between grant dollars and income tax revenue, with complete deficit- and revenue-neutrality.  Hence, there is no potential for efficiency-enhancing changes in the State tax mix, and criterion 2 is not met.

The binding constraint here is the requirement for revenue neutrality with regard to individual taxpayers.  Of course, if we relax this assumption, then the story changes.  The Commonwealth could reduce its own income tax revenue by less than one dollar, say, by 80 cents, while the States raise an additional dollar.  This raises both the net income tax burden (by 20 cents) and the national excess burden associated with the income tax.  Assuming Federal budget neutrality, the Commonwealth could reduce grants to the States by 80 cents, leaving the States free to reduce tax revenue from other sources by 20 cents.  On the assumption that the excess burden of income taxes is less than the excess burden of the State taxes which fall, the total excess burden of taxation will be lower.  This results in an efficiency gain for the economy as a whole.

This option is, however, difficult, particularly in the first few years of any return of tax powers, because it would require a net increase in income tax to fund the fall in revenue from the inefficient State taxes.  As was found with the States' Tax Plan, this is probably not politically feasible (so criterion 9 is violated) simply because it is not revenue-neutral.  Indeed, that is the reason why the first three years of that Plan involved revenue and budget neutrality, with the political pain of having any net increase in income tax left for future governments to decide.  Hence, with a reallocation of income tax powers there seems to be a serious political constraint to achieving the tax-mix benefits of reform.

Alternatively, we can retain individual taxpayer neutrality, but allow non-neutrality in the Federal budget, to get a change in the tax mix.

For example, the Commonwealth could reduce its income tax burden by one dollar (while the States increase their take by one dollar), but reduce grants by only 80 cents.  For given Federal spending and other taxes, this increases the federal deficit by 20 cents.  The States are better off by 20 cents, which results in more State spending and lower taxes on existing State tax bases (again, we assume that the cuts fall on those State taxes with the high excess burdens).  There is a change in the national tax mix, but at the cost of a larger Federal deficit.  This option, too, is ruled out by our criterion of deficit neutrality.  It is inadvisable on economic grounds in any case, since it is effectively the same as borrowing at the Federal level to redress the fiscal gap and change the tax mix.

Finally, the only other way to get a change in the tax mix is if we maintain individual taxpayer and Federal budget neutrality, but reduce Federal spending.

For example, if the Commonwealth reduces its income tax revenue by one dollar and Commonwealth spending falls by 20 cents, it can reduce grants by 80 cents and keep the Federal deficit fixed.  Again, the States have an extra 20 cents, which results in a fall in State tax revenue and an increase in State spending, according to preferences between public and private spending.  But this option does not seem all that feasible either, particularly in the present macroeconomic context where the Commonwealth already faces a commitment to cut spending and the deficit over the next two years.  To add a further burden to this plan, in order to redress the fiscal gap and secure a change in the tax mix at the State level, seems heroic.

Thus, with a return of income tax powers, under conditions of revenue and deficit neutrality, there may be barriers to securing tax mix benefits, although it is clear that the federalism benefits are certainly achievable.  For this reason, such a proposal may not meet criterion 2 -- a possibly serious blow against the idea, because changing the national tax mix has, arguably, the potential to achieve additional welfare gains.

On the other hand, if handing back tax powers is seen strictly as a federalist initiative, and changing the tax mix is taken to be a separate policy problem, then this is not a concern, and returning income tax powers to the States according to any of the above formulas would rank highly in welfare terms.

Improving the benefits:  reducing VFI

A second issue to be resolved with a return of income tax powers is:  what degree of fiscal gap is to be eliminated?  The deeper the cuts to the fiscal gap the greater will be the federalism benefits.

Federal grants to the States have two components.  The first is general revenue assistance, which contains an unconditional "tax reimbursement" component (revenue collected from the citizens of a given State and returned to that State by the Commonwealth);  and a "fiscal equalisation" component (untied revenue which is redistributive across States).  The second type of grants are specific purpose payments or tied grants.

A minimum withdrawal of the Commonwealth from grants would see the complete elimination of tax reimbursement grants, so that the only general revenue assistance remaining would be the fiscal equalisation component.  The revenue lost would be replaced by a State income tax, with rates set at a level sufficient to replace the revenue lost.  Thus, the redistributive part of the grants system would remain, as would tied grants.  It is only the grant component arising out of the centralisation of tax powers which is eliminated under this minimal scenario.

A more adventurous strategy would go one step further, and see tied grants to the States eliminated as well.

Under this strategy, tax reimbursement and tied grants would be eliminated, leaving only the equalisation component of transfers.  Again, State income tax rates would be set to replace the lost revenue for each State.  Finally, what Walsh calls the "radical" option would have all grants, including the equalisation component eliminated, with State income tax rates being set accordingly.

Our preferred option would be to get rid of both tax reimbursement and tied grants, but to leave the fiscal equalisation transfers intact.  As we show in Section 3, we believe that equalisation has significant benefits, and should be retained.

Improving the benefits:  compensating the weaker States, and the role of equalisation

A third issue, one which is directly linked to the question above, is whether the income tax rates that States set competitively under a devolution of income tax powers would result in revenues at the State level which match what they get now from the fiscal cartel.  This question has been examined in the literature. ⁽³⁷⁾  Results suggest that it is likely that some States will be made worse-off in any reallocation of tax powers (as the tax cartel is broken up), even though such a policy change may be of benefit overall.  But the losing States have an incentive to resist the change.  For the policy to be feasible, there may need to be transfers between States following the devolution of tax powers.

To illustrate the potential problem, suppose that all tied and tax reimbursement grants are eliminated by the Commonwealth, and that it withdraws proportionately and uniformly (as it is constrained to do by the Constitution) from the income tax base, so that the policy is neutral with regard to the Federal budget.  This allows the States to enter the income tax base up to the extent of the Federal withdrawal (assuming that individual taxpayer neutrality is to be maintained).  Let us suppose further that the Walsh plan is implemented.

As noted, States will wish to impose different individual tax rates, according to local conditions and preferences.  Each State, too, will have a different taxable capacity -- some States are richer than others.  As a result, the rates and revenues in some States may exceed what is lost in grants, while in other States the reverse may be true.  For example, a poorer State might have trouble setting a flat income tax rate which is sufficient to replace grant revenue lost, without causing a sizeable portion of its tax base (especially the rich) to migrate interstate.  The data presented in Table 3 on page 52 indicate which States may be the losers in any devolution of tax powers.

Table 3:  Impact of Fiscal Equalisation on the Distribution of
the Pool of Funds Available to the States, 1995-96, $m

	Distribution with Equalisation (1)	Equal per capita Distribution (2)	Difference (1)-(2) (3)	Distribution on the basis of income tax paid (4)	Difference (1)-(4) (5)
NSW	5522	6322	-800	6736	-1214
VIC	3947	4645	-698	4799	-851
QLD	3545	3401	144	2890	655
WA	1886	1794	92	1777	109
SA	1830	1520	309	1386	444
Tas	751	487	264	426	325
NT	902	179	723	168	734
ACT	281	315	-34	483	-202
Total	18664	18664	-	18644	-

Source:  Commonwealth Financial Relations With Other Levels of Government,
1995-96, Budget Paper No. 3.  {Table reproduced from Petchey 1995.)

If the policy is to be neutral in its impact on State budgets, the Commonwealth will have to compensate the weaker States, presumably from the pool of grant dollars saved.  Without compensation, some States (probably the lower-income States) may not agree to a return of tax powers. ⁽³⁸⁾  This could pose a significant hurdle to reform.

Of course, under any return of the income tax power, the income tax bases of the States would also enter the equalisation process;  hence some compensation would be automatic, though with some time-lag.  States with comparatively weak income tax bases should benefit from the application of the equalisation formula, while those with strong income tax bases would contribute to the weaker States.  But whether this equalisation for differences in the strength of the income tax base across States is a sufficient compensating mechanism requires further research.  As we show in Section 3, this is one of the reasons why we recommend the retention of the equalisation process under any tax powers reallocation plan:  quite simply, it would help greatly in sorting out the compensation issues in a rational fashion, at least partly divorced from politics.

Further matters for decision

Other issues to be resolved include:

• Should there be a period during which there is no tax competition, with Federal and State budget neutrality as well as individual taxpayer neutrality (as in the States' Tax Plan)?
• Above what threshold would the State rate apply (with the Commonwealth rates being reduced accordingly to make room)?
• Should the States be allowed to impose a flat or progressive rate above the agreed threshold (and if the latter, then is the States' role in income redistribution to be greatly expanded)?
• How should any fiscal policy co-ordination issues related to income redistribution and stabilisation policies be handled?
• If tax-mix benefits are also to be realised, how is this to be achieved -- through violating either taxpayer neutrality or federal budget neutrality?
• If changes to the tax mix are possible, what taxes should the States withdraw from?
• How would the States cope with the fluctuations in revenue inherent in the income tax base?

We now turn to a discussion of the other option for a return of tax powers to the States:  State consumption taxes.

Option B:  State Consumption Taxes

The types of consumption tax are:

• general wholesale or retail sales taxes;
• selective sales tax, alternatively known as an excise or differential commodity tax;
• value-added taxes (VAT);  and
• personal consumption (or "direct expenditure") tax.

Each type of consumption tax has its own efficiency costs (excess burdens), equity implications, and compliance and administrative costs.  It should be noted that there is currently some interest in the expenditure tax, particularly in the US, at the present;  Wood & Associates will be releasing a paper on this issue in the future.  (It may be relevant to note here that although the expenditure tax is accompanied by some difficulties of implementation, it would appear that Constitutional constraints on the States are not a difficulty.)

Examining the costs and benefits of giving the States access to each tax would be a major exercise, taking us too far afield for present purposes.  Nevertheless, excluding the expenditure tax option for the time being, there are two broad options.

Consumption tax sharing

First, the States could be given access to one of the Commonwealth's existing consumption tax bases on a tax-sharing basis, along similar lines to the income tax plans discussed above.  But such a proposal has the weaknesses of the tax-sharing plans already discussed.  It has the added weakness of what tax base to choose.  A candidate is the Wholesale Sales Tax (WST).  But this tax base has multiple rates applied to a narrow base with many exemptions, leading to high efficiency and administrative costs.  The rate differentiation is arbitrary, and certainly not based on optimal tax considerations.  To introduce tax-sharing on this base would not be advisable from a tax-mix efficiency standpoint.  For these reasons, we do not consider tax-sharing on the Federal government's existing consumption tax base to be a feasible proposition.

States' own consumption tax

The second option is for the States to be allowed a new consumption tax, with competitive setting of rates across an agreed tax base.  As noted above, the possibilities would be a general or discriminatory retail or wholesale sales tax, a VAT or a personal consumption tax. ⁽³⁹⁾  We cannot examine in detail all these possibilities, so we proceed to analyse the issues that would be confronted with the introduction of a general State retail sales tax.  Of course, many of the issues we discuss would also apply to other types of consumption tax.

Suppose -- putting aside the Constitutional restriction for now -- that the States have a new retail sales tax.  In Appendix 3, we consider four scenarios which show how the new tax might be used to achieve different combinations of tax-mix efficiency gains and federalism benefits, while at the same time satisfying the criteria set out earlier -- particularly the need for revenue and deficit neutrality.  The option we settle on as deserving of further attention is Scenario 4, which we believe offers maximum federalism and tax-mix gains.

(Briefly, under this Scenario, for each dollar of State sales tax revenue, the Commonwealth, while preserving the neutrality requirements, would reduce its grants to the States by 80 cents, the difference of 20 cents being used by the States to reduce revenue from their most inefficient taxes.)

WHAT ARE THE ISSUES?

The Constitutional constraint

The main difficulty with consumption taxes is the standard one:  that they are generally unconstitutional, for reasons already outlined.  But in the last major High Court decision on Section 90, the Capital Duplicators case, the decision was 4-to-3 in favour of the broad definition of excise -- hardly a resounding majority.  This indicates that there is potential for a change in the Court's interpretation of excise, especially since two of the four judges in favour of this interpretation have since resigned and been replaced.

The Court may be reluctant to reopen any debate on its broad interpretation of excise, particularly to admit any rigorous economic debate, since this might raise inconsistencies in the broad interpretation of excise and present difficulties for the Court.  Second, the States may be reluctant to raise the issue, because of fears about being seen to be pushing a new State tax and the political costs this may involve.  Thus, there may be a reluctance on the part of both the Court and the States to consider the issue, even though a reinterpretation of excise has the potential to yield welfare benefits.

Nevertheless, there is a stream of Section 90 cases going before the Court, and the States regularly make submissions to hearings.  The States could co-ordinate their efforts on one of these cases, and present a joint submission which attempted to open up the Court's interpretation of excise to debate.  With co-ordinated action, no single State would be criticised for seeking a "new State tax", and the States could develop a strategy for justifying their joint action.  It may be up to one of the politically strong States -- at present Western Australia, South Australia or Victoria -- to lead the way.

Ways around the constraint?

If, despite concerted argument, the Court will not reverse its interpretation of excise, there may be another workable option for a State consumption tax:  that is, for the Commonwealth to collect the tax.  For this to be anything other than a tax-sharing arrangement, States would need a say in setting rates independently, implying that different rates would need to be applied in different States in response to State wishes.  This may, however, run into Constitutional problems if the Commonwealth applies the tax on the States' behalf, because Section 51 (ii) prohibits the Commonwealth from treating the States differentially with regard to taxation.  Also, a "bookkeeping" system -- similar to the so-called "Braddon Clause", implemented in the 10 years following Federation -- would be required, so that the revenue collected in a State could be returned to that State.  This may increase administrative costs.

The States could, of course, be admitted into the consumption tax base if the phrase "and of excise" were deleted from the Constitution -- as recommended by the 1988 Constitutional Commission.  This would require Commonwealth agreement, and on both sides of politics.  History suggests that referendums to change the Constitution are only successful if they have the broad support of both parties at the Commonwealth and State levels.  Such support would probably face insurmountable political hurdles, and it seems unlikely to us to have a great chance of success.

Barring unforeseen Constitutional initiatives, the main hope for independent State access to the consumption tax base thus seems to lie in a change of the High Court's interpretation of excise, perhaps encouraged by a joint State-Commonwealth effort to have the Court's interpretation of excise re-examined, not least in the light of economic rationales.  Commonwealth implementation of a sales tax regime on behalf of the States may also be a possibility, if some way can be found for genuine tax competition between the States to be preserved while getting around any Constitutional restrictions on the Commonwealth's treating the States differently.

Efficiency gains from the tax mix

A second issue is, what Federal and State taxes would be replaced by the general State sales tax in order to get the efficiency gains which can flow from a change in the tax mix?  Ideally, one would wish to reduce reliance on those taxes which have higher excess burdens, compliance and administrative costs, and inequities than would arise from the State sales tax.  It is unclear which taxes these might be, and more work is needed here.  But a candidate at the federal level would be the WST which, as noted above, is generally seen as highly distortionary;  while candidates at the State level would be stamp duties, the various financial institutions duties, and franchise fees.

It can be demonstrated -- indeed, is now well known -- that by international standards Australia's taxation system is biased in favour of income taxes.  There is, further, a widespread consensus that there ought to be a shift from income taxes to indirect taxes in Australia;  this was the major thrust of the Fightback! plan of 1991.  This is a view that we find compelling;  and it seems that there is a strong case for using revenue from a consumption tax at the State level to reduce our reliance on the income tax.

Similarly, whether the States ought to use any additional sales tax revenue to reduce payroll tax is debatable, and there is no clear consensus on whether a sales tax has lower efficiency costs than payroll tax.  Indeed, there is a well-known thesis that, under certain circumstances, a payroll tax is equivalent to a consumption tax.  Again, however, the Fightback! proposal of 1991 took the view that the payroll tax system as it exists (with excessive exemptions) is highly inefficient, and contained a strategy for elimination of this tax and its replacement with revenue from the GST. ⁽⁴⁰⁾

Our view is that greater consideration needs to be given to all these issues, especially the question of whether the reliance on payroll taxes should be reduced.  There is already clear potential to achieve tax-mix benefits by abolition of some of the smaller but highly distortionary Federal taxes such as the Commonwealth's WST and State taxes such as stamp duties on financial transactions (where tax bases are highly mobile).

Uniform or differential tax rates?

A third issue is whether the State sales tax ought to be a differential tax or a uniform tax.  The theory of optimal taxation shows that if the goal is to collect a given amount of revenue as efficiently as possible, then tax rates should be inversely related to compensated price elasticities of demand (the inverse elasticity rule or Ramsey taxes).  This implies that to minimise the excess burden of the tax, rates should differ across commodities.  Efficiency does not require a general sales tax with the same rate for all goods.  Indeed, it requires that goods with highly inelastic demand ought to have a higher rate imposed on them.

But to the extent that government cares about equity as well, then optimal tax theory requires a departure from the inverse elasticity rule.  This is because many of the goods with inelastic demand (such as food, medical services and other necessities) make up a higher proportion of the budgets of the poor.  Governments with egalitarian goals may wish to tax such goods lightly or not at all (as in the US, where states exclude certain basics from the sales tax base).

Unfortunately we do not know with any certainty the price elasticities for all goods, and it is generally accepted that a uniform rate is not a bad approach.  If a sales tax is to be considered for the Australian States, then this is probably the best feasible option available.

Equity effects

The fourth issue we consider here is the equity effects of sales taxes for individuals.  A common view is that they are highly regressive, because the poor allocate a higher proportion of their income to consumption than do the rich (who save more of their income).  This has led to analysis of how the poor, and others who may bear a greater incidence of the tax, should be compensated upon the introduction of sales taxes.  The general view is that such compensation ought to be made through the personal income tax system, rather than by introducing exemptions on the sales tax base itself, since such exemptions raise the efficiency costs of the tax.  This was what the Fightback! proposal of 1991 was to do:  adjust the income tax scales and tax-free threshold for those on low incomes, to compensate for the effects of the GST on prices.

Thus, if a State sales tax were introduced, this whole question of equity would need to be addressed, as it needed to be confronted during the debate over Fightback! and the GST.  Herein lies a constraint to the introduction of a State sales tax -- the compensation issue is a significant political problem.

Recent evidence suggests that the conventional view about the regressiveness of a general sales tax may be misplaced.  This is so for three reasons.

First, the standard view looks at consumption as a proportion of annual income.  Lifetime income is more relevant, however, and there is quite convincing evidence that lifetime consumption is similar across all income groups.  This has led some ⁽⁴¹⁾ to argue that a general sales tax is actually progressive when measured with regard to lifetime income.  Second, the standard argument implicitly assumes that economic and legal incidence are the same, and ignores the whole theory of tax incidence.  But the sales tax may be shifted in a way which depends on supply and demand conditions in particular markets.  Finally, the rich find it relatively easy to avoid paying income tax, and invest vast sums in tax avoidance and tax minimisation.  These are opportunities less available to the poor.  Such avoidance mechanisms are not possible with a sales tax.  It is possible that a general sales tax could force the rich to pay more tax than under an income tax regime.

These complications suggest that the conventional story on sales taxes and regressiveness is too simplistic, and that, rather, the effect of a general sales tax on income distribution is still an unresolved question.  Nevertheless, any State sales tax regime would have to confront the issue of equity effects and compensation.  Both questions need further research.

Other issues

A fifth issue is whether there would be any ongoing inflation effects from the one-off price stimulus generated by the introduction of a State sales tax.  Again, this was a key concern with the Fightback! plan, and we only note the issue here as one that would require some resolution.

The final issue, one discussed above in relation to the return of income tax powers, is whether for each State the competitive sales tax rates and revenues would be greater or less than the grants lost under any plan to reduce the fiscal gap.  This raises again the issue of compensation to certain States (as opposed to compensation to individuals), and the role that equalisation would play in such compensation.  Since the conceptual issues are the same as for the income tax, there is no need to revisit them here.

CHOOSING BETWEEN THE OPTIONS: AN INCOME OR CONSUMPTION TAX FOR THE STATES?

One drawback of allowing the States into the income tax base is that to get the tax-mix benefits of reform, the criteria of deficit or revenue neutrality would need to be violated.  But in other respects, the State income tax is compelling, since it maximises the federalism benefits from reducing or eliminating the fiscal gap, increases competitive federalism with regard to taxation by allowing beneficial tax competition, and minimises harmful competition over the tax base by allowing appropriate co-operation.  Also, it does not face Constitutional restrictions to the same degree as a State consumption tax.

A new State consumption tax, on the other hand, achieves the federalism benefits and any possible tax-mix efficiencies, while maintaining total tax revenue and budget neutrality at the Commonwealth and State levels.  (See Scenario 4, in Appendix 3.)  This option also provides all the benefits of beneficial tax competition, while minimising the harmful effects of destructive competition over tax bases.  The State sales tax, however, faces a Constitutional hurdle.

Thus, each option offers considerable scope for economic efficiency gains, and it is our view that both should be given further consideration in any plan to reform and devolve tax powers.  Perhaps the eventual solution would involve a combination of income tax and indirect tax powers for the States.

1.4.4 EXPAND EXISTING STATE TAXES

As noted above, a few State taxes are reasonably efficient, and it has been argued ⁽⁴²⁾ that there is scope for the States to increase revenue from existing tax bases.

The most likely candidate for expansion would indeed be the payroll tax base, since, under certain conditions, the long-run effects of the consumption and payroll taxes are identical.  One might thus argue that the States ought to expand their use of the payroll tax base, since its efficiency effects are the same as those associated with a consumption tax.  Further, the payroll tax has relatively low compliance and administrative costs.  Why then, would we want a new consumption tax, with all the complications this would bring?  The fiscal gap could be reduced just as easily if the States reformed their payroll tax bases and increased revenue from this source, with a consequent reduction in grants.

If one were comparing an ideal broad-based consumption tax with an ideal broad-based labour tax, then it is true that there would be no efficiency gains in introducing a new State consumption tax.  But there are three problems with this argument.

First, the conditions required for equivalence between the payroll and consumption tax are stringent and extremely unlikely to be met in practice.  Given this, the payroll tax would perform less efficiently than the consumption tax.

Second, the payroll taxes imposed by the States are also far from the ideal:  they are not broadly-based, and have many exemptions and different thresholds.  This is probably partly due to destructive tax competition between the States over the base itself.  The tax has also been used by governments to hand out industry assistance, particularly to small business, often on an ad hoc basis.

Finally, as noted above, the SGL competes with the payroll tax, implying that there is a good case for returning the payroll tax to the Commonwealth and for using its revenue for superannuation and other similar "earmarked" purposes, as in other federal and unitary systems.  This raises further doubts about whether the States should become increasingly reliant on the payroll tax and, indeed, whether it is an appropriate State tax at all.

1.5. CONCLUSION

We have examined the costs and benefits of Australia's centralised taxation regime and noted that many people have called for a reallocation of tax powers to the States.  The reform options were examined and many issues were raised.  Perhaps the first is whether to reallocate income or consumption tax powers, or some combination of them.  In contrast to those who have gone before us over this ground, we have argued that both income and consumption tax powers for the States deserve more serious consideration.  We have highlighted the potential for consumption taxes on the grounds that they yield the same federalism benefits as a devolution of income tax powers, but also offer greater potential for efficiency-enhancing changes in the tax mix at the State and Federal levels.

The main reason that reformers have rejected this option in the past, and favoured returning income tax powers, is the Constitutional hurdle.  But this may not be such a difficulty as it has been in the past, given recent decisions of the Court and changes in its membership.  In addition, recent economic research on the excise tax issue has provided added weight to the economic arguments against the broad Constitutional interpretation of excise.

This suggests that we may be at a point in our federal history where we have the best chance of changing the Court's interpretation of excise;  but doing so would require concerted action by one State or a collection of States in challenging the Court to reopen its debates over the interpretation of excise.  Of all the issues discussed above with regard to a State consumption tax, including the equity and tax competition issues, this is probably the most important of all, since without change at this level the other issues remain of academic interest only.

Finally, our work here suggests that further research is needed on the income tax and consumption tax options.  In particular, estimates of rates and revenues under both scenarios need to be made, along with an analysis of all the issues which we have raised here.

---

SECTION 2:
THE DISTRIBUTION OF SPENDING POWERS:
ROLES AND RESPONSIBILITIES IN THE FEDERATION

2.1. INTRODUCTION

Centralisation of tax powers has also helped the Commonwealth to encroach on policy-making decisions over areas of responsibility implicitly left to the States at Federation, including the big policy and spending areas of health and education.  Thus, the economic effects of centralised tax powers reach far beyond the immediate consequences of the large fiscal gap, accountability, responsibility and limited beneficial tax competition.

Interestingly, this aspect of Australian federalism has received much less attention than the tax question.  This is in contrast to other federations, where the distribution of roles and responsibilities between central and sub-national governments is a major research and policy issue. ⁽⁴³⁾  As also noted in Section 1, in the US and Canada there are moves to devolve spending powers to the states and provinces, a trend which is not evident in Australia.

We now provide a brief summary of how roles and responsibilities -- which we refer to, from now on, as spending powers -- have become centralised, discuss the disadvantages of centralisation, and go on to examine the main argument used in defence of this trend -- the 'national interest' argument.  We show that it may have been greatly overstated as a justification for the degree of centralisation seen in Australia, and that policy coordination between States, facilitated and monitored by the Commonwealth, would be likely to achieve equal or superior outcomes when there is a genuine national interest at stake.  We conclude that there is a case for devolving policy-making powers, principally through a programme of reducing both tied grants and the legislative dominance of the centre.

2.2. HOW HAVE SPENDING POWERS BECOME CENTRALISED?

Most expenditure powers in Australia remain, in spirit at least, decentralised.  Section 51 of the Australian Constitution gives about forty powers to the Commonwealth.  All powers not listed in Section 51 are, by implication, left to the States as a residual, although they are not specified explicitly as in the Canadian Constitution.

But the Commonwealth has gained substantial influence over State policies in these areas because of the way it has been able to use conditional grants to control State behaviour, and because of its legislative dominance.

Tied grants

Conditional or tied grants have been used to change the distribution of spending powers as follows.  As discussed in Section 1, tax power centralisation means that the Commonwealth must make large tax reimbursement grants to the States.  In the years immediately following Federation, these grants were intended to give back revenue to the States which properly belonged to them but was collected by the Commonwealth because it had dominant tax powers.  If this was as far as it went, then centralisation of tax powers, per se, would not result in any centralisation of spending powers.  The Commonwealth would simply collect revenue on behalf of the States and hand it back to them as untied grants, as it did initially.

This, however, is not the way things developed.  Because of a key High Court decision, Commonwealth v. Victoria (1926), the Commonwealth was given the power to attach any conditions it liked to these grants, regardless of whether the conditions had anything to do with the Commonwealth's Constitutional powers.

As a result, it became possible for the Commonwealth increasingly to encroach on State policies by attaching more and more conditions to what were previously unconditional grants made necessary by centralised tax powers;  that is, tied grants enabled the Commonwealth to achieve back-door centralisation of spending powers.  The use of these grants by the Commonwealth picked up markedly from the early 1970s onward.  As a proportion of total State revenue, they increased from 26 per cent in 1972-73 to 53 per cent in 1993-94, with the biggest expansion coming during the Whitlam Government years.  There are now considerable Commonwealth and State bureaucracies devoted to conditional transfers, and their extensive coverage is a contentious issue in State-Federal relations.

Legislative dominance

Commonwealth legislative dominance has also contributed to this process.  Section 109 of the Constitution provides that where a State law is inconsistent with a valid Commonwealth law, the latter prevails, and the State law is invalid to the extent of the inconsistency.  Thus, the Commonwealth is able to overrule the States in any area where it enacts valid legislation, that is, legislation where it has a Constitutional responsibility.

One of the areas defined as a Commonwealth responsibility, in Section 51 (xxix) of the Constitution, is the external affairs power.  Supported by High Court decisions, the Commonwealth has used this power to ratify international treaties in areas such as human rights, labour relations and the environment.  The terms and conditions of these treaties are often determined by people in other countries, and the Commonwealth, with little or no consultation with the States.  By appealing to Section 109, the Commonwealth has then used these treaties as a basis for national legislation which overrules State laws.  This strategy has been used extensively by the Commonwealth to overrule State policies, especially in environmental matters and industrial relations. ⁽⁴⁴⁾  It is another source of friction between the States and the Commonwealth.

Many believe that the High Court has also had an important role in allowing the Commonwealth to gain greater influence over State policy-making.  Certainly, its decisions have generally supported the Commonwealth's expansion of influence into areas of State responsibility (for example, the decisions over tied grants).

2.3. HOW IS FEDERAL INTRUSION JUSTIFIED?

Nation-building and national interest

The major justification for central control is the 'national interest' argument.  The basic idea is that State decisions in areas such as environmental policy, education and health, have broader ramifications which go beyond State borders.  Thus, States can be said to have common or joint interests in particular policies because their interests are interdependent.  It is then argued that States ignore these common interests because they only care about citizens within their own borders, and that the national government, because it has the interests of all citizens of the federation at heart, will take account of common interests in its policy-making.  That is, only the federal government is supposed to act in the national interest, and to pursue nation-building goals.  In this way, the national vs. State interest story can be used to generate a case for federal intervention, with the centre taking over direct control of particular functions, or making matching (tied) grants to the States in the specific areas involving the common interests. ⁽⁴⁵⁾

Whether they realise it or not, this is what is meant by policy-makers and politicians when they talk about the need for a strong central government in the national interest.  The following comment from former minister Senator Bob McMullan makes the point nicely:

In my view, which I believe is also the Prime Minister's view, the most compelling argument for the Commonwealth maintaining its current revenue position vis-a-vis the States relates to the need for adequate representation of national interests across the full range of major economic and social interests facing Australia.  The Commonwealth's fiscal power enables it to pursue national priorities in such areas as education and health, to name just two. ⁽⁴⁶⁾ (italics ours)

2.4. TREATIES AS AN ALTERNATIVE TO FEDERAL CONFROL

But is central control the only way to make policies in the national interest and to build a unified nation?  What has been overlooked, but is now receiving attention, is that the assignment issue should really be thought of in three stages:  competitive policy-making by states;  competitive policy-making by the states with inter-state cooperation (that is, inter-state treaties) to take account of common interests;  and central control through either tied grants or direct intervention.  Traditionally, the literature has dealt with the first and third options as the only alternatives.

But ideas on cooperation as an alternative to central action are now being developed.  They are likely to change the way we think about the assignment issue by highlighting the role of inter-state cooperation as an alternative to centralisation, emphasising the cases where cooperation offers superior outcomes to centralisation, and cases where competitive outcomes are preferable.  Some work has already been done, ⁽⁴⁷⁾ and further research is progressing.

The potential for inter-state 'treaties' as a substitute for central control seems high in areas such as education, health, law and order, and environmental policy (to the extent that externalities are important in these areas).  Recent examples of successful inter-state treaties where the Commonwealth played a key coordinating role were the agreement on uniform gun laws, and much of the legislation and reform to emerge from the Council of Australian Governments (COAG) process.

Of course, there may be certain instances where self-enforcing treaties are not feasible.  For example, the costs of bargaining between States may be too high, or States may be unable to agree due to a lack of trust.  These problems are well known.  Even then, however, central control is not necessarily justified.  Rather, what may be called for is federal monitoring and facilitating of inter-state agreements with control remaining with the States.

Unfortunately, the Commonwealth has abrogated its key role as a neutral arbitrator and facilitator of inter-state agreement in the Australian federation, and opted instead for direct control through tied transfers and legislative dominance.

2.5. REFORM OPTIONS

All this suggests that there is a case for redistributing spending powers in Australia, and for increasing inter-governmental co-operation on matters of common interest.

One way for this to happen is through a reduction in tied grants, less direct Commonwealth involvement in policy-making, and more independent policy-making by the States, either competitively or with cooperation when there are genuine national interests at stake.  The discussion suggests a case for a complete withdrawal of the Commonwealth from tied grants and the direct provision of policy and services, except where there is a genuine national interest which cannot be handled through inter-governmental coordination.  This would leave the Commonwealth with the major areas of foreign policy, defence, trade and commerce, and an expanded role in facilitating and monitoring inter-governmental cooperation in all other common interest matters.  The mechanisms for this cooperation are already in place, through COAG.  The conceptual arguments above suggest that the role of such bodies needs to be expanded as an alternative to continued centralisation.

The legislative dominance of the Commonwealth should also be constrained.  With regard to the foreign affairs power, this might be done by allowing the States to have a say in Australia's involvement in international treaties.  The new Commonwealth government has recently taken some encouraging small steps in this direction.

But ultimately the Commonwealth's ability to centralise spending powers depends on its financial powers.  If tax powers are not given back to the States, it is unlikely that there will be any reallocation of spending powers, especially through lower tied grants.  For this reason we propose that tax and expenditure powers be reduced simultaneously.

Of course, federal bureaucracies and politicians whose power bases depend on the dominance of Canberra will resist any devolution of spending powers (and tax powers).  No matter what the conceptual logic of devolving powers, this practical 'public choice' constraint to welfare-enhancing policy reform remains a formidable hurdle to overcome.

2.6. CONCLUSION

We conclude by noting that there needs to be further research on the conceptual principles for how spending powers ought to be assigned in a federation;  with particular emphasis on the circumstances under which policies should be implemented by states independently, when there should be policy co-ordination, and which functions should be centralised.  Though much has been done in this area (as discussed above), there are still many issues to be resolved.

In conjunction with this conceptual work, there needs to be a practical assessment of Australia's federal assignment of policy powers against the conceptual criteria.  This work could be done, for example, for each major area of responsibility, commencing with the big spending areas where there is large-scale federal intervention, such as health and education.

---

SECTION 3:  FISCAL EQUALISATION ⁽⁴⁸⁾

3.1. INTRODUCTION

In many federations, for example Australia, the UK, Denmark, Sweden, Canada and Germany, importance is attached to inter-state equalising transfers;  while other countries, the US included, have no equalisation schemes at all.  At a practical level, the equalisation schemes which determine these transfers allow states in federations to provide some 'standard' level of service while still imposing average tax burdens on their citizens.  Without such schemes, it is likely that levels of state service provision would vary across states, even if all states had the same preferences over policies, simply because some states are richer than others.

We now provide a brief discussion of the arguments against and for equalisation, followed by conclusions about reform to Australia's equalisation system.

3.2. ARGUMENTS AGAINST EQUALISATION ⁽⁴⁹⁾

Australia has a system which equalises for both expenditure and revenue differences across States.  Since the fiscal gap is so large in Australia, the pool of funds to which the equalisation methodology is applied is substantial:  $18,664 million in 1995-96.  Application of the equalisation process to the pool means that there are large transfers between the States.  This can be seen from Table 3 (on the following page).

Column 1 shows what each State will actually receive in 1995-96 from the total revenue pool under the equalisation methodology.  The remaining columns, 2 and 4, show what the States would receive under two alternative methods of distributing the pool:  an equal per capita distribution, and an allocation based on income tax paid.

It is clear from this comparison that New South Wales and Victoria are substantial losers from equalisation, while all the remaining States and Territories are net beneficiaries, with the Northern Territory, South Australia and Tasmania being the largest winners.

Equalisation has always been controversial.  It is subject to regular review by the Commonwealth Grants Commission (CGC), acting in consultation with the States.  The main bone of contention surrounds 'disabilities' -- the factors used to estimate the extent to which a particular State suffers a cost penalty relative to other States.  The States commit considerable resources to these reviews, and many critiques of the Commission's results are made.  Another of these reviews is currently under way.

But most of these criticisms are limited in scope, in the sense that they take the basic methodology as given, and argue over parameters to gain maximum advantage for a particular State.  Recently, researchers have raised more fundamental questions about the theoretical basis of equalisation.

The key issue has been the efficiency effects of equalisation.  Many studies have been conducted into the efficiency implications of these transfers. ⁽⁵⁰⁾  With one exception, ⁽⁵¹⁾ the conclusion to emerge from this work is that fiscal equalisation, and the inter-state transfers which result, have economic efficiency costs.  The arguments range from the inefficient interstate migration caused by the transfers, to their negative incentive effects.

So pronounced is this view that the Grants Commission apparently accepts it with little question. ⁽⁵²⁾  Moreover, two official reports examining equalisation and efficiency were released in 1993-94.  The first, by the Heads of Treasuries, noted the conflicting views about equalisation and efficiency, and concluded that negative incentive effects of the type highlighted in the literature were probably minor in practice.  The second, from the Industry Commission, reviewed equalisation, efficiency and equity, and concluded that the evidence on the efficiency effects of equalisation is inconclusive.

3.3. ARGUMENTS FOR EQUALISATION

Balanced against this are arguments in favour of equalisation in federations.  These arguments are technically complicated, numerous and varied, and we do not have time to go into them all here. ⁽⁵³⁾  Some of the stronger notions which emerge from recent literature are, however, relevant to the Australian situation, which has tended very much to be focused on the minutiae of the Grants Commission methodology, and which perhaps needs a broader theoretical justification.

We can mention here the idea that equalisation actually encourages stability and cohesion in federations, by reallocating the benefits of federalism between the inevitable winners and losers among the participating states.  Bearing in mind the concept of winning and losing states, we can also see that some degree of equalisation may well be necessary to preserve and encourage tax and policy competition between participating states.  Equalisation is also important in overcoming some of the problems of equity mentioned earlier.  Finally, there is also the argument that equalisation can enhance economic efficiency. ⁽⁵⁴⁾  These are useful concepts, but full discussion is not feasible here;  instead, we point the interested reader to the discussion in the paper by Petchey and Walsh (1993).

There is one further issue which needs brief discussion here.

It is often overlooked that a great deal of equalisation occurs 'implicitly' in federal and unitary systems.  This is because central governments use uniform national tax systems to fund not only uniform provision of national public goods and services, such as defence, foreign relations and welfare benefits, but also uniform access to services such as education, health and hospitals, law and order, that are the province of sub-national governments in federal systems.  This leads to significant implicit transfers between states which have different average per capita taxable capacities -- for instance, because average per capita incomes differ, or because of different costs of service provision.

One scholar ⁽⁵⁵⁾ has estimated that in Australia these implicit transfers exceed in magnitude the explicit formula-based equalisation transfers.  These results indicate that States such as Tasmania are large beneficiaries of these implicit transfers, while other States such as Western Australia are large net contributors.  In short, implicit transfers, which occur because of the other operations of the federal budget, penalise the richer States and benefit the poorer States.

But the Grants Commission formula for equalisation takes no adequate account of this 'implicit' redistribution which is already going on.  In addition, even if we assume that the explicit equalisation transfers are at the precise level required for efficiency (that is, they establish an optimal distribution of mobile factors across the States) they are unlikely to be so if implicit transfers are taken into account.  In other words, the aim of efficiency-based equalisation might be to stop inefficient migration out of, say, Tasmania;  but together with the implicit transfers already going to that State, the effect may be the opposite -- the system as a whole may act to keep too many people in that State.

Apart from this implicit transfer issue, there are also continuing questions over the effect of the formula on the incentives of States to act strategically, and inefficiently, to improve on the transfer that they receive.  These questions have been addressed by a number of scholars, ⁽⁵⁶⁾ but it seems to us that they are unresolved.

3.4. REFORMS AND CONCLUSIONS

Arguments have been advanced in favour of equalisation which generally point to a case for expenditure and revenue equalisation, though not necessarily along the exact lines of Australia's equalisation system.

But as the criticisms of equalisation on efficiency grounds -- and the brief discussion above of these issues -- highlight, these matters remain controversial.  As yet there is no consensus view.

Given this, we propose that further research on the economic rationale for equalisation and the efficiency consequences of Australia's system of equalisation is needed.  The issue of implicit transfers should also be investigated.  In the meantime, we believe that the system should be retained in its present form, at least in the medium term.  This does not preclude changes to 'fine tune' the existing methodology as part of the normal CGC review process.

This approach is consistent with moving ahead with the other reforms we have proposed in Sections 1 and 2.  Indeed, it is likely, as pointed out above and in Section 1, that equalisation will have a crucial role in facilitating the decentralisation of tax powers by being an integral part of the compensatory process which will be necessary if this reform is to be supported by all States, particularly the economically weaker States.

Of course, if tax powers are reallocated to the States, the pool of funds available for redistribution will be reduced, so that the absolute level of transfers will fall.  This, along with Western Australia's and Queensland's imminent change in status from claimant to contributing States, is likely to take the heat off equalisation as a political issue, though it will, no doubt, remain a topic of general research interest.  The mechanics of the formula will also continue to be important as the States compete with each other for advantage, as in the past.

---

SECTION 4:  SUMMARY OF CONCLUSIONS

It seems clear that there is now an opportunity to put forward serious proposals for the reform of Australia's State taxation systems.  State taxes, however, go hand in hand with State-Federal fiscal relations:  the two must be reformed together.  That in turn will strengthen Australia's federal institutions, and pave the way for authentic competitive federalism.

It is also clear that the centralisation of both taxing and spending functions has gone too far in Australia, further than is necessary in practice and further than theory would suggest is ideal.  The necessity for reform is clear.

The options for tax reform have been measured against a set of clear criteria.  Against those criteria, the devolution to the States of the powers to raise a consumption tax or an income tax stand out.  Other options fail to yield the same clear benefits.

The consumption tax option suffers from what we have called the 'Constitutional constraint';  but that is perhaps not as binding a constraint as it once might have appeared.

The alternatives are not, of course, entirely exclusive.  For instance:  while we reject the alternative of increasing the efficiency of State taxes as a stand-alone option, there is no reason why it should not, in part at least, be combined with a better option;  nor is there any compelling reason why there need be a choice between consumption and income taxation, rather than a mix of the two.

It follows that devolution of taxing powers must be accompanied by devolution of policy and spending roles.  It may be noted here that this is already, to a limited extent, taking place.

To some extent, these considerations will be determined by the tax mix which we would wish to see in place post-reform, itself clearly a subject for further research and discussion.  We have pointed to Australia's excessive dependence on the income tax;  and there would clearly be little point in devolving comprehensive tax powers to the States if the mix were to be left unchanged.  (Here again, of course, some elements of the mix would be decided in the longer term by competition.)

Three features of this paper are, to some extent, unusual.

First, as noted above, there is the clear preference for the consumption tax and income tax options.

Second, there is a strong defence of the importance of equalisation, on both political and economic grounds.  It is not just that the maintenance of equalisation in some form will be necessary to secure the adherence of all the States to a coherent programme of reform.  As well, we have shown that equalisation has an important role in securing the full benefits of competitive federalism.

Finally, in contrast to most of the literature on reforming Australian federalism, we have repeatedly pointed to the benefits of the States (sometimes with the Commonwealth) acting in a cooperative manner to achieve desired goals.  Clearly there will need to be comprehensive cooperative strategies adopted by the States to secure the achievement of reform:  cooperation is such areas as united negotiation with the Commonwealth, joint action as needed over the Constitutional problems, sensible agreement over key aspects of the chosen tax base, and over implementation.

More than that, the fullest possible resumption by the States of their appropriate policy and expenditure roles will frequently require the States to cooperate in novel ways, ways which will allow them to consider and express the national interest more thoroughly than can the Commonwealth.  There will frequently be a delicate balance between securing the benefits of cooperation and securing the benefits of competition.  Even in an imperfect world, we believe that the benefits secured under a system of competitive federalism -- combined with co-operation over matters of common concern -- will be superior to those we presently see in our lopsided centralist regime.

This paper has canvassed only the broadest aspects of the questions involved.  In the conclusions to each section, we have been careful to indicate those topics which require more study.  Some of these -- such as the question of which taxes most need reform -- are major;  some much less so.  Work has already been commissioned on a number of these, and will be published in due course.

---

APPENDIX 1:  STATE TAXES AND SECTION 90

According to Coper, one of the major motives of the Founding Fathers in making the excise tax exclusive to the Commonwealth had to do with Federal tariff policy.  He notes that:

The consensus of opinion amongst informed scholars and right-thinking judges is that, historically, the ban on State excise duties was intended, as an adjunct or corollary to the ban on State customs duties and State bounties on production or export, to protect Federal tariff policy from State interference. ⁽⁵⁷⁾

This argument, though it does not emerge strongly from the Federation Debates, has merit.  State power over excise duties would enable them to change the intended effect of federal tariff policy.  A State could, for example, by imposing an excise duty on its producers, reduce the effect of a federal tariff designed to lower the domestic price of the good relative to the price of imports.  In other words, State excises could offset the intended effect of a federal tariff policy.  A free-trading State could use power over excise to frustrate a protectionist Commonwealth.

What is more, this added level of protection (or reduced level, if the excise tax is positive) would differ across States according to how each State set its excise tax policy.  It would, therefore, have an impact on inter-state trade, and run counter to the intent of Section 92, which guarantees free trade between States.  Since one of the main motivations for Australian federation was to obtain the benefits of a customs union, with a common external tariff and free inter-state trade, it seems reasonable that the power to impose excises should also be made exclusive to the Commonwealth.

But if this was indeed the motive behind centralising the excise power, then the meaning of excise to be adopted in order to satisfy the intent of the Founders would be extremely narrow.  It would include production taxes on internationally-traded goods subject to federal tariffs.  Other indirect taxes might be available to the States.  For example, it might be possible for States to impose production taxes on goods and services for which there are no competing imports or for which there is no Commonwealth tariff (that is, goods that were internationally traded but imported tariff-free). ⁽⁵⁸⁾

It is likely, however, that under this interpretation States would be free to impose taxes that affected imports (from inter-state and overseas) and domestically-produced goods equally.  This would leave the States free to impose, for example, a general non-discriminatory sales tax, because such taxes would not interfere with federal tariff policy nor would they affect the freedom of inter-state trade (since relative prices between States would not change).

Another view is that the Founders intended, in giving the Commonwealth control over both customs and excise duties -- two of the most important taxes of the time -- to grant the Commonwealth a general control over taxes on goods for macroeconomic policy reasons.  Proponents of this view argue that this means that we must adopt a wide interpretation of the meaning of excise to include taxes on the production, distribution or sale of goods, because taxes on the latter have the same effect as excise taxes on production:  that is, they all affect the price of goods.  If the Founders intended, in Section 90, to give the Commonwealth general control over taxes in order to have control over fiscal policy, then it follows that taxes on sales and distribution would have to be centralised.

Further recent economic analysis of Section 90 has identified an additional motive behind the centralisation of excise duties:  a desire to ensure free internal trade and to ensure the unity of the Federation. ⁽⁵⁹⁾  In this view, it is possible to discern such motives from the Constitutional debates.  Since a discriminatory excise tax has the potential to distort inter-state trade patterns, and since Australian federation was primarily motivated by a desire to promote free internal trade, together with a common external tariff (that is, to achieve the benefits of a customs union), the Founders centralised excise to ensure that the free internal trade end of the federal compact was upheld.

Thus, centralising excise taxes went hand-in-hand with centralising the customs tariff:  one secured a common external tariff and the other helped secure free inter-state trade.  Without a centralised excise tax, there was always the possibility that States could use discriminatory excise taxes to interfere with trade across State boundaries.  It is reasonable, then, to argue that the concern of the Founders for free inter-state trade derived from their fears over federal unity and the need for the States to act cooperatively.  But proposing this motive for Section 90 still leads to a narrow view of an excise tax, as a discriminatory tax on production.  General non-discriminatory taxes would not distort inter-state trade, and would, therefore, be allowable.

Given the lack of clarity about what was to be included as an excise, it was not long after Federation that the High Court was asked to adjudicate and interpret the Constitution on this matter.  In the first case to be considered, Peterswald v Bartley (1904), the Court adopted the narrow view of excise as a tax on the quantity or value of goods produced in a State.  This left the States free to impose taxes on goods that did not discriminate between States;  that is, the narrow definition did not pose a threat to State tax powers.  But this was gradually expanded to a broader interpretation of the meaning of excise in a series of cases during the 1920s and 1930s;  in particular, John Fairfax and Sons Ltd v New South Wales (1927), A.G. (NSW) v Homebush Flour Mills Ltd (1937) and Mathews v Chicory Marketing Board (1938).  The accepted, modern 'broad' definition can be seen in Dixon J.'s judgment in Parton v Milk Board (Victoria) of 1949.  By 1963, the Court unanimously stated that 'It is now established that for Constitutional purposes duties of excise are taxes directly related to goods imposed "at some step" in their production or distribution before they reach the hands of the consumers'.  (Bolton v Madsen).

All subsequent Section 90 cases before the High Court have seen the adoption of this very wide definition of an excise as any tax on the production, distribution or sale of goods.  This has meant the end of State taxes on goods, with three notable exceptions:  the States are allowed to levy franchise fees on tobacco, alcohol and petroleum products, though their hold on these tax bases has been tenuous in recent years.

The wide view of excise has been highly contentious.  A detailed discussion of the Court's reasons for taking the wide view would take us too far afield for present purposes.  An excellent survey can be found in Coper. ⁽⁶⁰⁾  Briefly, the main reason for the reversal of the earlier interpretation of excise is that the Court changed its mind over the motives behind Section 90.  In the early years, judges accepted the tariff policy motive.  This, as discussed before, implies a narrow interpretation of the meaning of excise.  Eventually, however, the 'general control over taxation' motive gained prominence, and the wide interpretation of excise that this implies was adopted.  It remains this way today.

We already know that the Founders realised that Section 90 gave the Commonwealth considerable financial power.  If, however, they intended that 'and of excise' be included in Section 90 purely because of the tariff motive, or because they were concerned about inter-state free trade and federal unity, then they would be shocked if they could see how it has contributed to the financial dependence of the States.  On the other hand, if they really meant to give the Commonwealth a general power over taxation of commodities, then maybe what we have today is what was intended.

There are at least two points against this last interpretation.

First, why would the Founders go to all the trouble of specifying the division of powers in the Constitution (the Constitution of Australia was settled on after 30-odd years of tough inter-colonial negotiations from the 1870s onward), and then give the Commonwealth so much control over taxation that it is effectively able to destroy that division of power?  This seems to make a mockery of the Constitution.  Second, if the Founders did intend to give the Commonwealth a broad power over taxation of goods, why then did they not centralise all the other powers that States use to influence the price of goods?  State decisions still have plenty of scope for influencing prices without consumption taxes.

---

APPENDIX 2:  TAX POWERS IN THE UNITED STATES

Having surveyed the main ideas on the costs and benefits of decentralised tax regimes in federations, we here offer a discussion of the distribution of tax powers in the US.  This provides a useful comparison with Australia, and highlights some stark differences over how tax assignment issues are handled elsewhere.

The taxing authority of the states in the US is limited only by certain provisions of the US Constitution.  Constitutional authority over relationships with foreign countries, international and interstate trade are exclusively the domain of the Federal Government, specifically the US Congress.  Exclusive of taxes that would interfere with this authority, everything else can be part of the base that can be taxed by individual states.  There are very few limits on state taxing authority as in Australia.

These few constitutional restrictions mean that states cannot charge tariffs on imported foreign goods or on goods exported to foreign countries.  They cannot impose special charges on goods imported from or exported to other states, for that would involve state interference in interstate trade and would subvert Congressional authority.  These restrictions, however, exhaust the list.  In the US, individual states may and do impose taxes on income, whether that earned by state residents within the state or that earned by state residents out of state and in foreign countries.  They can tax business and corporate income as well as extracted natural resources within state borders.  Property values are part of the tax base, as well as inheritance.  Most states have an ad valorem retail sales tax and all have license taxes of various sorts.

Various taxes are used by the states, even though states use different sets of taxes and administer them differently.  An illustrative case is California.  It is particularly salient, as it is the largest state in the US, with an economy larger and more complex than most independent nations'.

There are three major revenue sources for the state of California:  the individual income tax, the retail sales tax and the corporate income tax.  Local government within the state also has access to a property value tax and various license and permit fees.

The individual income tax is assessed against all income earned in the state, whether by residents or non-residents.  Out-of-state income of California residents is also subject to the state income tax, although, through agreements between some states (not all), the income tax paid to one state can be claimed as a credit against the amount owed to California.  Income earned abroad by California residents is subject to California taxes, even though much of the income might be excluded from federal taxation.

In most states, California certainly among them, the retail sales tax is a major source of general fund revenue.  For California, it represents one-third of the total of such revenues. ⁽⁶¹⁾  The retail sales tax is an ad valorem tax on retail sales, with many exclusions from the base.  Most notably, food and prescription drugs are excluded as part of life's necessities.  Without such exclusions, the tax would be unacceptably regressive.

There is an interesting constitutional issue in the assessment of sales tax.  A robust catalogue business in many different products has sprung up in the US.  Many items, including clothing and computers as well as more specialised commodities, are sold via catalogue sales.  One of the major attractions of this form of purchase for consumers is that the sales escape state sales taxes.  For California residents, where sales are taxed at a rate of 7.5 per cent, the saving is substantial, particularly on items such as computers.  States are constitutionally prohibited from charging taxes on such sales, as they are considered part of interstate commerce.  Taxing them would be the equivalent of charging a tariff, and the exclusive right to do this constitutionally resides with the Congress.  Interestingly, in this regard, Congress could have voted to allow states to charge sales taxes on inter-state catalogue sales, and it has been petitioned by the states to do so;  but due to intense lobbying by the mail order industry, the prohibition on the states remains.

Mail order sales represent only one way that states can lose taxes on retail sales.  Another is the direct transfer of business between states as a result of sales tax differentials.  If the rate differential between neighbouring states is large, the residents from the high-tax states will manage their purchases so that the bulk of them can be made across the state border.  Obviously, this tax leakage is more intense with residents closest to the borders, but for some states it represents a significant loss in revenue.

States can also charge a corporate income (earnings) tax.  This tax is an important source of revenue for California, and the source of an interesting Constitutional problem.  The state, of course, can only charge tax on California earnings, and that is not a problem for corporations which reside exclusively in the state.  But it is common for corporations to have operations in many different states and, not uncommonly, in many different countries.

These present a difficult problem in assessing a state tax, in that the operations of the corporation are often interconnected in complicated ways.  One of the most vexing problems is that due to 'transfer pricing' -- the assigning of costs to various parts of the production process leading up to the final sale.  An obvious example is an auto manufacturer which produces an automobile in one state to be sold in another.  The opportunity of assigning cost geographically by corporations gives them a large incentive to assign the costs to the states with the highest tax rates (to get maximum deductions) and the sales to the lower tax rate states.  This practice affected California in the 1930s, when motion picture companies established distribution subsidiaries in the neighbouring state of Nevada, which had no corporate profits tax.  They then assigned all of their production costs to California and all of their profits to Nevada.

In order to combat this accounting abuse, California adopted the Uniform Assessment Method, in which part of the entire world-wide profits of a corporation doing business in California is assigned to the state on a formula basis which accounts for the fraction of sales and employment in California.  This method has raised some vexing constitutional questions when applied to corporations whose natural headquarters are in foreign countries.  The Japanese, with many firms doing business in California, have raised the point that such a method of assessment brings California into the area of foreign relations from which states are constitutionally prohibited.  Pressure has been put on the state to reassess these arrangements, but they still remain in force.

Most states also charge severance taxes for mineral extraction.  The rate of tax varies with different states and different minerals.  It is generally assessed on an ad valorem basis on the extracted values, rather than value in the ground.  From a particular state's point of view, the severance tax is an ideal one, in that a large part of it can be exported from the taxing state itself.  To the extent that the mineral is used elsewhere, the price it fetches on external markets will reflect the state taxes paid.  States with small populations and large mineral wealth, such as Alaska and Wyoming, can have a well-financed public sector supported by severance taxes and, thus, subsidised, to a large extent, by citizens of other states.

States are also able to impose excise taxes on selective commodities.  California, for instance, imposes a large per unit excise tax on both cigarettes and motor vehicle fuel.  Only a part of these taxes is part of the general revenue budget:  a large part of the fuel tax is earmarked for highway maintenance and construction, and the tobacco tax for health-related expenditures.

Finally, most states now have their own lottery.  Depending on the state, the money can be part of the general expenditure revenue, or (as in California) it can be earmarked (for primary and secondary publicly-financed education).

Thus, the US states have far greater taxation powers than states in Australia.  While it is true that such taxation powers raise issues related to tax competition, tax exporting and other problems, it is also the case that the beneficial effects of decentralised tax powers are more highly felt in the US.  Moreover, there seems to be greater weight put on the value of these benefits, in particular the ability of financially-strong states to constrain the ability of Washington to abuse its powers.  This has traditionally been a greater concern in the US than in Australia, mainly because the US has a long history, beginning with the Articles of Confederation and the Federalist Papers, of grappling with the central problem of how to constrain the power of the state.  By contrast, in Australia, where we have a mixture of a Westminster system and a federation, we appear to have been less concerned with controlling the power of government.

---

APPENDIX 3:  THE CONSUMPTION TAX SCENARIOS

SCENARIO 1:

Each dollar of revenue raised by a new State sales tax can be used to reduce, by one dollar, existing State taxes on the most inefficient State tax bases (say, stamp duties).  The Federal budget and grants do not change (for Federal budget neutrality).  There are, potentially, tax-mix efficiency gains, but no federalism benefits, since the fiscal gap is unchanged.  Total State tax revenue is unchanged.  But there are equity effects, since relatively more State tax revenue is raised from the general sales tax and less from the inefficient State taxes (each with a different economic incidence).  The policy is also 'State budget neutral', and the effects of the policy change are confined to the State level, since all variables which interact with the Commonwealth are kept fixed.  With this policy, the federal budget is quarantined from the changes at the State level.

The following options allow State-Federal fiscal variables to change (in particular, grants), so that some or all of the effects of the State sales tax are shifted to the Commonwealth.

SCENARIO 2:

For every dollar of general sales tax raised by the States, the Commonwealth could reduce its grants by one dollar, giving the States no opportunity to reduce their inefficient taxes -- hence State tax revenue neutrality is violated.  There is an increase in State taxation, and the Commonwealth confiscates the extra revenue by reducing its grants.  There are no efficiency gains from changes in the tax mix at the State level.  Of course, there are federalism benefits, since the fiscal gap is reduced.

Since the revenue from the new State sales tax has been transferred to the Commonwealth, any tax-mix efficiency gains will depend on what the central government does with the dollar of grants it saves.  It can use the dollar to reduce its taxes by one dollar, with an unchanged Federal deficit (for a given level of spending).  To the extent that this tax cut falls on taxes with a higher excess burden than the general State sales tax, then tax-mix benefits are realised at the national level -- that is, the Commonwealth takes a dollar of grant saved and reduces its own inefficient taxes.

This policy is equivalent to the Commonwealth directly introducing its own general sales taxes, and withdrawing from existing inefficient taxes (the highly distortionary and inequitable WST, for instance);  except that here the States collect the revenue and the fiscal gap narrows as that revenue accrues to the Commonwealth while it reduces grants to the States.

Alternatively, the Commonwealth can use the dollar to increase its own spending by one dollar (leaving the deficit unchanged) -- but this has no tax-mix benefits at the national level, because Commonwealth taxes do not fall.  It may not be desirable on macroeconomic grounds either, since all that has happened is that taxation and Federal spending have risen proportionately.

There could also be a combination of reduced taxes (to give national tax-mix benefits) and increased spending (with the effect on the deficit being ambiguous, and depending on how the dollar is divided between spending and lower Federal taxes) -- again probably not a desirable outcome.

Finally, the Commonwealth could keep its taxes and spending fixed, and use the grants saved to reduce its deficit.  Unless the Federal government passes all the grants saved into lower Federal taxes, this option results in higher total tax revenue.  There will also be equity effects, since those bearing the economic incidence of the State sales tax will be worse-off, while those who bear the incidence of the reduced Commonwealth taxes are better-off.  But as noted below, increasing taxes to lower the deficit is politically difficult, and on economic grounds it would be better for the Federal government to reduce the deficit through spending cuts at the Federal level, not by increasing State taxes.

SCENARIO 3:

If the Commonwealth cuts its grants by less than one dollar -- that is, if it does not appropriate all the additional revenue from the State sales tax, and leaves some for the States -- the opportunity exists for the States to reduce inefficient State taxes, and also perhaps to increase State spending (though this might not be a desirable result).  This would improve the tax mix, as inefficient State taxes are replaced by less inefficient general sales tax revenues.  The federalism benefits are also achieved, because the fiscal gap narrows, and the Commonwealth has the opportunity to reduce its own taxes and/or the federal deficit in some combination.  Thus, under this option, the extra revenue from the State sales tax is shared between the States and the Commonwealth and used to reduce taxes which are more inefficient than the retail sales tax at both the national and State level.

But of course, the total tax take increases, unless the Commonwealth cuts its taxes by the full amount of its reduced grants, and the States do not increase their spending -- both of which are unlikely scenarios.  There are also equity effects, even if the total tax take is unchanged, because the economic incidence of taxation changes.  As with the other options, there is no way of knowing, without detailed analysis, what these equity effects might be, and what redistribution policies would be needed to satisfy the equity criterion.  Thus, tax revenue neutrality is violated.

SCENARIO 4:

Another option lies within the third scenario and is as follows.

For each dollar of State sales tax revenue, the Commonwealth would cut grants by less than a dollar, say 80 cents.  The difference, 20 cents, would be used by the States to reduce revenue from their most inefficient taxes (by 20 cents), yielding tax-mix benefits at the State level.  The cut in grants to the States also reduces the fiscal gap, thus yielding federalism benefits.  At the Commonwealth level, the grant saving of 80 cents should be used entirely to reduce revenue from inefficient Commonwealth taxes (for a fixed level of federal spending), with the Federal deficit unchanged.

In other words, every dollar of State sales tax revenue raised is used to fund a one dollar decline in more inefficient taxes:  80 cents at the Federal level and 20 cents at the State level.  And in so doing, the fiscal gap is also reduced by 80 cents, yielding federalism benefits.

This scenario maximises the federalism and tax-mix benefits of tax reform.  Moreover, tax revenue neutrality is retained overall, while for given Federal spending, the Commonwealth deficit is unaffected, as are State deficits (again for given spending).  For every dollar of State sales tax revenue collected, other State taxes fall by 20 cents, and the most inefficient Federal taxes fall by 80 cents.

The temptation for the Commonwealth to use some of the grants saved to fund deficit reduction (explained within scenario 3 above) should also be resisted.  All of the reduction in grants should go toward Federal tax reduction.  If the Commonwealth used grants saved to fund deficit reduction, it would effectively be solving its deficit problems with higher taxes -- applied at the State level.  It would be far preferable for the Commonwealth to address its deficit problems separately from federalism and tax-mix reform, and to do so via Federal spending cuts.

We believe that Scenario 4 is superior to the Fightback! proposal -- to introduce a new broad-based consumption tax at the Federal level and to allow the States a share of this base.  It is true that this would allow tax-mix efficiency gains (as Fightback! and the GST were primarily designed to do).  But it would probably ⁽⁶²⁾ increase the fiscal gap and hence reduce beneficial tax competition, and make the federal system even less competitive than at present.  Scenario 4 allows the tax-mix gains of the Fightback! proposal, but by introducing the broad-based consumption tax at the State level, it also permits federalism reform with its associated welfare gains.  Arguably, the federalism benefits of our reform proposal might outweigh the benefits from tax-mix changes.  If this is so, then the Fightback! strategy actually missed the more important aspect of tax reform!

---

REFERENCES

Access Economics (1995), 'An Access Economics Study on the Distribution of Federal/State Financial Powers', September.

Albon, R.P. (1990), 'The Efficiency Implications of Locational-Based Factors in Commonwealth Grants Commission Determinations', Appendix to the Victorian Submission to the Inquiry into Issues of Methodology, Victoria's Response to the Discussion Papers, Inquiry into Grants Commission Methodology, 1990.

Albon, R.P. (1996), 'An Overview of Australia's Taxation Structure', Commissioned Paper for the Stocktake of Progress in Microeconomic Reform, Productivity Commission, Canberra.

Albon, R.P., and J.D. Petchey (1993), 'Federalism Aspects of Fightback!', in J.G. Head, (ed.), Fightback!:  An Economic Assessment, Australian Tax Research Foundation, Conference Series No. 12, Sydney, 551-565.

Boadway, R.W., and F. Flatters (1982), 'Efficiency and Equalisation Payments in a Federal System of Government:  A Synthesis and Extension of Recent Results', Canadian Journal of Economics, 15.

Boadway, R.W., and M. Keen (1994), 'Efficiency and the Fiscal Gap in Federal Systems', Paper Presented at the International Seminar on Public Economics (ISPE) Conference on 'Fiscal Policy in Emerging Federations', Nashville, Tennessee, USA, August 26-28.

Boothe, P., and J.D. Petchey (1996), 'Assigning Responsibility for Regional Stabilisation:  Evidence from Canada and Australia', Forthcoming in P. Boothe (ed.), Reforming Fiscal Federalism for Global Competition, University of Alberta Press, Edmonton, Canada.

Brennan, G., and J.M. Buchanan (1980), The Power to Tax:  Analytic Foundations of a Fiscal Constitution, Cambridge University Press, Cambridge.

Brennan, G. (1990), 'Commentary', in C. Walsh (ed.), Issues in State Taxation, Centre for Research on Federal Financial Relations, The Australian National University, Canberra, 93-100.

Breton, A. (1985), 'Supplementary Statement', in Report of The Royal Commission on the Economic Union and Development Prospects for Canada, Ministry of Supply and Services, Ottawa.  (Reprinted, 1987, as 'Towards a Theory of Competitive Federalism', European Journal of Political Economy, Vol. 3:1 and 2.)

Breton, A. (1991), 'The Existence and Stability of Interjurisdictional Competition', in D. Kenyon and J. Kincaid (eds), Competition among State and Local Governments, Urban Institute Press, Washington.

Brosio, G. (1992), 'The Balance Sheet of the Australian Federation:  Some Tentative Estimates', Working Paper, Federalism Research Centre, Research School of Social Sciences, Australian National University, Canberra.

Buchanan, J.M. (1990), 'Europe's Constitutional Opportunity', in Institute of Economic Affairs, Europe's Constitutional Future, London.

Buchanan, J.M., and C.J. Goetz (1972), 'Efficiency Limits and Fiscal Mobility:  An Assessment of the Tiebout Model', Journal of Public Economics, 1.

Buchanan, J.M., and D.R. Lee (1994), 'On a Fiscal Constitution for the European Union', Journal des Economistes et des Etudes Humaines, Volume 5, No. 2/3, June/ September, 219-232.

Burbidge, J.B., and G. Myers (1994a), 'Federation as Coalition Formation', Paper Presented at the International Seminar on Public Economics (ISPE) Conference on 'Fiscal Policy in Emerging Federations', Nashville, Tennessee, USA, August 26-28.

Burbidge, J.B., and G. Myers (1994b), 'Redistribution within and across the regions of a federation', Canadian Journal of Economics, 27(3), August, 620-36.

Cassing, J.M., and A.L. Hillman (1982), 'State-Federal Resource Rent Tax Rivalry:  The Queensland Railways and the Federal Export Tax', The Economic Record, 58 (162).

Chessell, D. (1993), 'Fightback!:  A Public Finance Banquet', in J.G. Head (ed.), Fightback!:  An Economic Assessment, Australian Tax Research Foundation, Conference Series No. 12, Sydney, 1-20.

Coase, R.H. (1960), 'The Problem of Social Costs', Journal of Law and Economics, 3, 1-44.

Commonwealth Grants Commission (1990), Report on Issues in Fiscal Equalisation, Vols I and II, Australian Government Publishing Service, Canberra.

Commonwealth Grants Commission (1993), Report on General Revenue Grant Relativities, Volumes I, II and III, AGPS, Canberra.

'Commonwealth Financial Relations With Other Levels of Government', 1995-96 Budget Paper No. 3, AGPS, Canberra.

Constitutional Commission (1988), 'Final Report', AGPS, Canberra.

Coper, M. (1987), Encounters With the Australian Constitution, CCH Limited.

Courchene, T. (1971), 'Interprovincial Migration and Economic Adjustment', Canadian Journal of Economics, 3.

Courchene, T. (1984), Equalisation Payments;  Past, Present and Future, Ontario Economic Council, Toronto.

Deakin, A. (1901), 'The Chariot Wheels of the Central Government', in W. Prest and R. Mathews (eds), Development of Australian Fiscal Federalism (1980), Australian National University Press, Canberra, 13-18.

Dixon, P.B., J.R. Madden and M.W. Peter (1993), 'The Effects of Reallocating General Revenue Assistance Among the Australian States', The Economic Record, Vol. 69, No. 207, December, 367-81.

Fisher, R.C. (1993), 'Macroeconomic Implications of Sub-national Fiscal Policy:  The Overseas Experience', in D.J. Collins (ed.), Vertical Fiscal Imbalance, Conference Series No. 13, Australian Tax Research Foundation, Sydney, 125-152.

Freebairn, J. (1993), 'The GST and Payroll Taxation', in J.G. Head (ed.), Fightback!:  An Economic Assessment, Australian Tax Research Foundation, Conference Series No. 12, Sydney, 97-110.

Galligan, B. (1995), A Federal Republic:  Australia's Constitutional System of Government, Cambridge University Press, Melbourne.

Gordon, R.H. (1983), 'An Optimal Taxation Approach to Fiscal Federalism', Quarterly Journal of Economics, November, 567-586.

Gramlich, E. (1987), 'Subnational Fiscal Policy', Perspectives on Local Public Finance and Public Policy, 3, 3-27.

Head, J.G. (1985), 'Interventions' at the National Taxation Summit, Record of Proceedings, AGPS, Canberra.

Heads of Treasuries (1994), 'Report on Fiscal Equalisation', March.

Industry Commission (1993), Report on Impediments to Regional Industry Adjustment, December.

Inman, R.P., and D.L. Rubinfeld (1994), 'Designing Tax Policy in Federalist Economies:  An Overview', Burch Working Paper No. B95-2, University of California, Berkeley.

Institute of Public Affairs (1995a), 'Overlap and Duplication Among Australian Governments:  Do Specific Purpose Grants Achieve Their Purpose?', Paper prepared for the Department of Premier and Cabinet, Victoria, October.

Institute of Public Affairs (1995b), 'Roles and Responsibilities in the Australian Federation', Paper prepared for the Department of Premier and Cabinet, Victoria, October.

Kelly, C.R. (Bert) (1978), One More Nail, Brolga Books, Adelaide.

Kerr, I.A. (1993), 'Taxation Measures to Reduce Vertical Fiscal Imbalance in Australia:  A Review of the October 1991 Report of the Working Party on Tax Powers', in D.J. Collins (ed.), Vertical Fiscal Imbalance, Australian Tax Research Foundation, Conference Series No. 13.

Mansoorian, A., and G.M. Myers (1993), 'Attachment to home and efficient purchases of population in a fiscal externality economy', Journal of Public Economics, 52, 117-132, North Holland.

McMullan, B. (1993), 'Vertical Fiscal Imbalance:  A Federal Government Perspective', in D.J. Collins (ed.), Vertical Fiscal Imbalance, Australian Tax Research Foundation, Conference Series No. 13, 11-17.

Maxwell, J. and C. Pestieau (1980), Economic Realities of Contemporary Confederation, C.D. Howe Research Institute, Montreal, Canada.

May, R.J. (1971), Financing the Small States in Australian Federalism, Oxford University Press, Melbourne.

Metcalf, G.E. (1993), 'The Lifetime Incidence of State and Local Taxes:  Measuring Changes during the 1980s', Working Paper 4252, Cambridge, MA:  National Bureau of Economic Research, January.

Moy, P. (1993), 'Vertical Fiscal Imbalance:  The States' View', in D.J. Collins (ed.), Vertical Fiscal Imbalance, Conference Series No. 13, Australian Tax Research Foundation, Sydney, 301-314.

Myers, G. (1990), 'Optimality, Free Mobility and Regional Authority in a Federation', Journal of Public Economics, 43.

New South Wales Tax Task Force (1988), 'Tax Reform and NSW Economic Development:  Review of the State Tax System', NSW Government Printer, Sydney.

Nieuwenhuysen (1983) = Committee of Inquiry into State Government Revenue Raising (1983), 'Report of the Committee into Revenue Raising in Victoria', Victorian Government Printer, Melbourne.

Norman, N.R. (1995), 'Refocussing Fiscal Reform', Paper Prepared for the Business Council of Australia, Melbourne.

Norfie, K.H. and M.B. Percy (1984), 'Province-building and Industrial Structure in a Small, Open Economy', in D.D. Purvis (ed.), Economic Adjustment and Public Policy in Canada, Kingston:  John Deutsch Memorial for the Study of Economic Policy.

Oates, W.E. (1972), Fiscal Federalism, Harcourt, Brace and Jovanovich, New York.

Oates, W.E., and R.M. Schwab (1988), 'Economic Competition Among Jurisdictions:  Efficiency Enhancing or Distortion Inducing?', Journal of Public Economics, 35.

Petchey, J.D. and C. Walsh (1993), 'Conceptual Foundations in Fiscal Equalisation', in Horizontal Fiscal Equalisation:  Efficiency and Equity Issues in Historical, Political and Economic Perspectives, Submission to the South Australian Treasury by the South Australian Centre for Economic Studies.

Petchey, J.D. (1993), 'Fiscal Equalisation in a Federal Economy with Free Inter-State Migration and Different Factor Endowments', Australian Economic Papers, Volume 32, December, 336-52.

Petchey, J.D. (1995), 'Inter-State Cost Differences, Resource Endowments and Equalising Transfers', The Economic Record, Vol. 71, No. 215, December, 343-353.

Petchey, J.D. and P. Shapiro (1995a), 'Chariot Wheels and Section 90', Policy, The Journal of the Centre for Independent Studies, Sydney.

Petchey, J.D. and P. Shapiro (1995b), 'Inter-State Cooperation and the Role of the Central Government:  An Economic Analysis of Spending Assignment in Federations', Paper Presented at a Conference on Regional Economic Issues Within Australasia, The University of New South Wales, Sydney, December 7-8.

Petchey, J.D., P. Shapiro and C. Walsh (1996), 'Transfers in Federal Systems:  A Critical Survey', forthcoming, in R. Fisher (ed.), Intergovernmental Fiscal Relations:  Perspectives and Prospects, Kluwer.

Petchey, J.D. and P. Shapiro (1996), 'One People One Destiny:  Centralisation and Conflicts of Interest in Australian Federalism', forthcoming, in D.E. Wildasin (ed.), Fiscal Policy in Emerging Federations, Cambridge University Press, New York.

Pope, J. (1992), 'The Compliance Costs of Taxation in Australia:  An Economic Policy Perspective', Curtin University, School of Economics and Finance, Working Paper No. 92.07.

Rosen, H.S. (1995), Public Finance, Fourth Edition, Irwin.

Ryan, M. (1995), 'What future for payroll taxes in Australia?', Treasury Research Paper No. 10, Treasury Research Paper Series, September.

Shapiro, P. and J.D. Petchey (1994), 'Shall Become Exclusive:  An Economic Analysis of Section 90', The Economic Record, Vol. 70, No. 209, 171-182.

Shapiro, P., J.D. Petchey and B.T. Coram (1995), 'Federal Stability, Secession and the Core', Federalism Research Centre, Australian National University, Canberra.

Shapiro, P., J.D. Petchey, and R.C. Cornes (1996) 'Federation and Confederation:  The Assignment of Authority With Inter-State Population Mobility', Mimeograph.

Stigler, G. (1957), 'The Tenable Range of Functions of Local Government', in USA Joint Economic Committee, Federal Expenditure Policy for Economic Growth and Stability, US Government Printing Office, Washington DC.

Swan, P. and G. Garvey (1993), 'The Equity and Efficiency Implications of Fiscal Equalisation', An Independent Study Undertaken on Behalf of the New South Wales and Victorian Treasuries, Swan Consultants Pty Ltd, Sydney.

Tiebout, C.M. (1956), 'A Pure Theory of Local Expenditures', Journal of Political Economy, 64.

Usher, D. (1995), The Uneasy Case for Equalization Payments, The Fraser Institute, Vancouver.

Walsh, C. (1990), 'State Taxation and Vertical Fiscal Imbalance:  The Radical Reform Options', in C. Walsh (ed.), Issues in State Taxation, Centre for Research on Federal Financial Relations, The Australian National University, Canberra, 53-92.

Walsh, C. (1993), 'Vertical Fiscal Imbalance:  The Issues', in Vertical Fiscal Imbalance, D.J. Collins (ed.), Australian Tax Research Foundation, Conference Series No. 13, 31-53.

Walsh, C. (1996), 'Refocussing Commonwealth-State Financial Relations:  Tax Powers, Microeconomic Reform and Intergovernmental Relations', forthcoming, in P. Boothe (ed.), Reforming Fiscal Federalism for Global Competition, University of Alberta Press, Edmonton, Canada.

Wildasin, D.E. (1988), 'Nash Equilibria in Models of Fiscal Competition', Journal of Public Economics, 35, 229-240.

Wildasin, D.E. (1991), 'Income Redistribution in a Common Labour Market', American Economic Review, 81, 757-74.

Wood, R.J. (1996), 'Land of the Long White Benchmark:  The New Zealand Reforms', Backgrounder, Vol. 8, No. 5, July 1996, Perth.

Wood, R.J. (1995), Competitive Federalism:  Promoting Freedom and Prosperity, Federalism Project Issues Paper No. 3, Perth.

---

ENDNOTES

1.  For a recent comprehensive summary of the problems, see Albon 1996.

2.  See Institute of Public Affairs 1995a.

3.  There are endless quantities of very good writings on federalism.  For a longer version of the arguments put here, see Institute of Public Affairs 1995b.  Also of interest is Galligan 1995.

4.  See Wood 1996.

5.  See, for instance, Wood 1995.

6.  See Oates 1972 and Gordon 1983.

7.  Although in the US the Republican majority in the Congress has focused recently on devolving authority from the federal government to the states, reversing the historical trend towards the centralisation of power in the US.  Similar trends are apparent in Canada.

8.  For a more detailed and extensive discussion, see Walsh 1990, 1993, 1996;  and Petchey and Shapiro 1996.

9.  By Brennan and Buchanan 1980, Buchanan and Lee 1994.

10.  Kelly 1978.

11.  Walsh 1993.

12.  Walsh 1996.

13.  Access Economics 1995, page 15.

14.  The States and the Commonwealth also currently tax the same tax bases in certain areas.  As shown by Cassing and Hillman 1982, such competition over the same base raises the potential for further inefficiencies, for example, between the federal Fringe Benefits Tax (FBT) and payroll tax (see Walsh 1996).

15.  See Moy 1993 for further discussion.

16.  See Kerr 1993, New South Wales Tax Review 1988, and Nieuwenhuysen 1983.

17.  Kerr 1993.

18.  Pope 1992.

19.  Petchey and Shapiro 1996.

20.  On which see IPA 1995a.

21.  See Oates 1972.

22.  See Fisher 1993 for a more detailed discussion.

23.  Boothe and Petchey 1996.

24.  Fisher 1993 and Gramlich 1987.

25.  This implies a loss of power by the Canberra economic bureaucracy and may meet resistance from this group even if such a change is in the national interest.  See Walsh 1996.

26.  Of course, co-operation may not be feasible for various reasons to be examined in more detail in section 2.

27.  Wildasin 1991.

28.  Shapiro, Petchey and Cornes 1996.

29.  McMullan 1993.

30.  Wildasin 1988.

31.  See Petchey and Shapiro 1995a, and Shapiro and Petchey 1994.

32.  Walsh 1990 and Moy 1993.

33.  Walsh 1990.

34.  Walsh 1996.

35.  Access Economics 1995.

36.  See Norman 1995.

37.  By Bruce Coram, in Shapiro, Petchey and Coram 1995, and Shapiro and Petchey 1994.

38.  In other words, some States may face a participation constraint,

39.  See Rosen 1995 for discussion of each.

40.  See Freebairn 1993.

41.  For example, Metcalf 1993.

42.  See Brennan 1990 and Ryan 1995.

43.  See Oates 1972.

44.  See Coper 1987.

45.  For example, see Oates and Schwab 1988.

46.  McMullan 1992.

47.  See Shapiro, Petchey and Cornes 1996.

48.  Parts of the discussion in this Section draw on Petchey, Shapiro and Walsh 1996.

49.  The discussion here draws on Petchey 1995.

50.  See Albon, 1990;  Commonwealth Grants Commission, 1990;  Swan and Garvey, 1993;  Dixon, Madden and Peter, 1993;  and Petchey, 1993, 1995.

51.  Petchey 1993.

52.  Commonwealth Grants Commission, 1993, vol. 1, pages 70-72.

53.  See Petchey, Shapiro and Walsh 1996.

54.  Petchey 1993, 1995.

55.  Brosio 1992.

56.  See, among others, Swan and Garvey 1993, and Usher 1995.

57.  Coper 1987, page 226.

58.  The only difficulty with such taxes is that they would still interfere with interstate trade by altering relative prices between states.  They could be struck down for contravening Section 92 guaranteeing free inter-state trade.

59.  Shapiro and Petchey 1994.

60.  Coper 1987.

61.  General Fund revenues are distinguished from special purpose revenues in that they are unrestricted funds that can be used for any public purpose.  Special funds revenues -- as for instance those collected from the tax on motor fuel, which must be used for highways, and lottery revenues, which must be used for public education -- are restricted in their use.  These taxes are discussed later.

62.  As argued by Albon and Petchey 1993.