Raising Australia's market share

A submission concerning the 2008/9 Federal Budget, by Gavin R. Putland. (P.S.: See also Maximalist ‘fiscal devaluations’ for Greece and Australia.)

Assisted by the unpopularity of the previous government's workplace policies, the leader of the Labor Party was elected Prime Minister of Australia. His name was James Henry Scullin. It was October 1929 and the Wall Street bubble was at its height. Twelve days later the bubble burst, triggering global depression, hence a collapse in demand for Australian exports. Scullin struggled on for two years before defections from his party forced an early election, in which the party was reduced to 14 seats out of 75 in the House of Representatives.

In December 1972, assisted by the unpopularity of the Vietnam War, Edward Gough Whitlam was elected Prime Minister at the height of a global real-estate bubble. The bursting of that bubble, closely followed by the First Oil Crisis, led to recession and stagflation. Sensing Whitlam's vulnerability, the conservative parties used their numbers in the Senate to force him to an election, which he lost in the House of Representatives by 91 seats to 36. The 55-seat margin is still a record.

In November 2007, assisted by the unpopularity of the previous government's workplace policies and the Iraq War, Kevin Michael Rudd was elected Prime Minister in economic circumstances similar to those of 1972, but further advanced towards disaster. The Fourth Oil Crisis was already upon us, while the global housing bubble — the biggest asset bubble in history — was already bursting in several key countries; in particular, the U.K. housing market had just slipped into reverse, while the U.S. housing market was in depression territory, threatening the stability of the global financial system through devaluation of collateral. The conservatives had already positioned themselves for a swift comeback by claiming that Rudd's workplace policies would cause the next recession.

Devalued collateral means less credit, hence less effective demand, hence (directly or indirectly) less global demand for the kinds of products that Australia exports. Hence, to avoid job losses in Australian export industries, Australia must capture a greater fraction of the smaller market.

Scullin's predecessor, Stanley Melbourne Bruce, was the first Australian Prime Minister to lose his own seat in the House. Rudd's predecessor was the second. Could Rudd himself be the third? In 1996, when Rudd first contested his present seat for the Labor Party, he lost.

If Rudd does not want to be remembered as another Scullin, another Whitlam, and the first Labor Prime Minister to lose his seat in the House, he must take urgent and drastic action to raise Australia's market share — preferably without breaking any election promises. This article considers how he might do it.

Contents

1.  Defining the problem

1.1  Unaffordable housing hurts efficiency
1.2  Payroll tax: Discriminating against local products
1.3  Superannuation: The payroll tax that dare not speak its name
1.4  Compliance costs: Taxation without revenue
1.5  Income tax: The biggest reverse-tariff of all
1.6  Incentive traps and poverty traps

2.  Components of a solution

2.1  A minimum deemed income from property
2.2  Reinventing the first home owners' grant
2.3  Owner-occupants vs. investors
2.4  A further note on negative gearing
2.5  Ending payroll tax and VFI
2.6  Keeping the GST rate below 10%
2.7  Cutting compliance costs: Make the GST a retail tax
2.8  Mythical advantages of input credits
2.9  Taking superannuation out of export prices
2.10  Inferior option: Dumping PAYE compliance costs on the banks
2.11  Superior option: Getting rid of income tax

3.  Constitutional threats to existing taxes

3.1  Are compliance costs unconstitutional?
3.2  Are stamp duties on new vehicles unconstitutional?
3.3  Is payroll tax an excise?

4.  Q&A

4.1  How would the minimum deemed income (MDI) affect rents?
4.2  Is the MDI a land tax?
4.3  Is this a “beggar my neighbour” strategy?

5.  Conclusion

1.  Defining the problem

1.1  Unaffordable housing hurts efficiency

Jobs cannot be created unless:

(a) the employer can pay the rent or mortgage on the business premises out of the proceeds of the business; and

(b) the workers can pay the rent or mortgage on housing within commuting distance of those jobs, out of wages that the employer can pay out of the proceeds of the business.

Unaffordable housing violates condition (b). In “bad” economic times, any such violation is a barrier to job creation. In “good” economic times, any such violation is a “capacity constraint” and is therefore inflationary. If one tries to offset the cost of housing by raising wages without a commensurate increase in productivity, inflationary pressures increase and Australian products become less competitive. In summary, there is never a wrong time to do something about the housing problem.

The present housing emergency is often discussed in terms of low rental vacancy rates (e.g. 1.2% for Melbourne in September 2007). But the official vacancy rates published by the various Real Estate Institutes ignore speculative vacancies — i.e. vacant lots, boarded-up buildings, and other unoccupied buildings for which the owners are not seeking tenants. The 2006 Census found that almost 10% of Australian private dwellings are unoccupied. This figure, although more realistic than REI vacancy rates, omits vacant lots and includes holiday homes whose locations are unsuitable for full-time occupation. These faults are not found in Earthsharing Australia's “I Want to Live Here” Report, which surveyed a sample area in inner Melbourne and found that the elimination of speculative vacancies would add a conservative 11.7% to the housing supply (in terms of population housed).

When Kevin Rudd, as PM-Elect, told his MPs to visit shelters for the homeless in their electorates, he was showing them the symptom. If he wanted to show them the cause, he should have told them to count the vacant lots and boarded-up dwellings. So the challenge for his government is to encourage owners to build on vacant lots and to sell or let unoccupied buildings.

Simply putting more money in the hands of renters and buyers will not have the necessary effect, but will bid up rents and prices. From the viewpoint of private entities, the overall supply of residential sites (locations, addresses) is fixed, as is the supply within acceptable distance of any particular services, infrastructure, or job opportunities. Yet access to suitably located sites is essential. Therefore values of housing sites are competed upward until they absorb the capacity to pay. If wages and salaries rise due to a “strong economy”, so does the cost of housing. If childcare or petrol or food gets cheaper, the cost of housing eats up the savings. What the rest of the economy giveth, the housing market taketh away. If this were not the case, economic growth alone would have solved the housing problem. It hasn't.

Therefore if a benefit for “working families” is not to be competed away in the housing market, it must be delivered through the housing market. The Deputy Prime Minister's “war on poverty” will be won or lost in the housing market. Victory requires policies that maximize the supply of housing, and hence the bargaining power of renters and buyers, by inducing property owners to convert every available site into an available home.

1.2  Payroll tax: Discriminating against local products

To the extent that payroll tax is passed on in prices, it causes inflationary pressure, which is especially objectionable in “good” economic times; in this sense payroll tax is as bad as a consumption tax. To the extent that payroll tax is borne by employers, it deters employment, which is especially objectionable in “bad” economic times; in this sense payroll tax, which exclusively targets employment, is worse than a consumption tax. Payroll taxes imposed by the Australian States apply only to Australian labour, and consequently inflate prices of Australian products more than prices of foreign products at the same stage of refinement. In particular, the effect of Australian payroll taxes on Australian exports is similar to that of foreign tariffs!

Consumption taxes do not discriminate in this way, because they make no distinction as to the origin of the products “consumed”, and because exports, which by definition are not “consumed” within the taxing jurisdiction, are zero-rated (that is, excluded from the tax base). Zero-rating of exports avoids double-taxation of products that will be taxed in the countries of destination. Accordingly, indirect taxes that zero-rate exports are said to be border-adjusted. The Australian GST, the European VAT and the North American retail sales taxes are border-adjusted. Payroll tax isn't.

1.3  Superannuation: The payroll tax that dare not speak its name

The 9% Federally-mandated, employer-funded superannuation contribution is equivalent to a 9% Federally-funded contribution paid for by a 9% Federal payroll tax. I note in passing that a Federally-funded superannuation contribution would never have been supported by itself, because it defeats the ostensible purpose of universal superannuation, namely to reduce public expenditure on social security. But more importantly, a Federal payroll tax would never have been thought a sensible way to pay for anything, because of its obvious effects on employment and international competitiveness; a consumption tax, for all its faults, would have been preferable on both counts.

The compulsory superannuation contribution, which began under the Keating government and ramped up during the early years of the Howard era, is undoubtedly part of the reason why Australia's trade performance was poorer during the Howard era than during the Hawke-Keating era. If the Rudd government is to reverse this decline, it must break the nexus between superannuation and payrolls.

1.4  Compliance costs: Taxation without revenue

The compliance costs associated with taxes deter the establishment of businesses, the hiring of labour, and the expansion of businesses beyond thresholds that incur additional taxes or reporting requirements. All these effects amount to capacity constraints. Compliance costs also feed into prices, including prices of exports and of local products that compete with imports. The jobs “created” for purposes of compliance are not a net benefit or even a compensating benefit, because if the economy can afford to employ a certain number of people solely for compliance, it could afford to employ a larger number for productive purposes. In short, compliance costs are an unmitigated evil.

1.5  Income tax: The biggest reverse-tariff of all

Australia's 30% corporate income tax applies to export income. In terms of international competitiveness, this is equivalent to a 30% import tariff in every country of destination of Australian exports.

Exempting export income from income tax is prohibited by GATT/WTO rules; this is logical because export income is still income. But exempting exports from consumption taxes is permitted; this is also logical because exports are not consumed within the taxing jurisdiction. Hence there is a competitive advantage to be gained by replacing income tax with a consumption tax.

Replacing income tax with a consumption tax while preserving gross (pre-tax) wages and salaries would obviously be regressive. But replacing income tax with a consumption tax while preserving net (after-tax) wages and salaries would not have any redistributive effect and, as we shall see, would not raise prices.

We will never “sustain ourselves beyond the resources boom,” as the Trade Minister puts it, unless we correct the anti-export bias of the tax system. Payroll tax is part of the bias. Superannuation is a bigger part. Income tax is the biggest.

1.6  Incentive traps and poverty traps

There are still some working families who, due to the combination of income tax and income-tested welfare payments, face effective marginal tax rates of 100% or more, and who therefore cannot increase their disposable income no matter how hard they work. This is not learned helplessness; it is enforced helplessness. It is also a capacity constraint and is therefore inflationary. As long as this situation persists, the “war on poverty” will be better described as a war on effort.

It is quite possible for a welfare system, by itself, to produce combinations of income tests that result in total withdrawal rates above 100 cents in the dollar. The elimination of income tax therefore does not guarantee the elimination of poverty traps and incentive traps. But it certainly helps.

2.  Components of a solution

2.1  A minimum deemed income from property

For the purposes of the income test on the age pension, the financial assets of pensioners are currently deemed to be earning 3.5% per annum (or 5.5% above certain thresholds), regardless of what they are really earning. This policy is defended as giving pensioners an incentive to shop around for higher returns.

I therefore submit that for tax purposes, in order to give property investors an incentive to seek rental income, every investment property should be deemed to be earning a minimum of 3.5% of the site value — where the site value is the value of the associated ground and/or airspace, including any attached building rights, but excluding actual buildings so as not to discourage construction. The exclusion of buildings makes the deeming rate generous by comparison with that applied to age pensioners. That is why the deeming rate determines only the minimum taxable income; for a property earning more than the minimum, the deeming rate would be of no effect.

In practice, any property that is fully developed (as permitted by the land zoning) and fully occupied by tenants will be earning more than the minimum deemed income (MDI) and will therefore not be affected by deeming. But a property that is underdeveloped or under-occupied may earn less than the MDI, and a vacant lot or unoccupied dwelling will certainly earn less. Hence the owner of an unoccupied dwelling would be motivated to let it to a tenant (or sell it), while the owner of a vacant lot would be encouraged to build on it and seek tenants (or sell it to someone who will).

Under present policies, a speculator can buy a vacant property, leave it vacant, sell it, pay tax on only half the capital gain (if the property is held for more than one year), and deduct holding costs (including interest) against future capital gains. This practice not only fails to supply accommodation, but probably also frustrates someone else's plans to supply accommodation. The deeming rate, by making this practice uneconomic, would improve the supply of housing.

Notice that I have described the deeming as “for tax purposes” without specifying the tax. For residential sites, the only applicable tax under the present system would be income tax. But for non-residential sites, the deeming rate could also be used for GST. This would strengthen the incentive for owners of non-residential sites to build accommodation and seek tenants, and would consequently improve the availability and affordability of business accommodation.

The introduction of the MDI would be revenue-positive, not only because of the tax on the deemed income, but also because sites would be literally taxed into use, causing additional economic activity, which in turn would yield tax revenue (the “growth dividend”).

2.2  Reinventing the first home owners' grant

The first home owners' grant (FHOG) is available regardless of whether the recipients build new homes (thus adding to supply and improving affordability) or buy existing homes (thus merely adding to demand). In contrast, the temporary supplement to the FHOG (March 2001 to June 2002) known as the Commonwealth Additional Grant (CAG) was available only for buyers of previously unoccupied homes (including buyers who dealt directly with builders). Thus the CAG stimulated construction by causing first-time buyers to prefer new homes. The FHOG should have worked that way from the beginning. As it did not, the Howard government, instead of phasing out the CAG, should have kept the CAG and phased out the FHOG.

If the grant were also made available to investors in new homes — in other words, if it were turned into a New Home Builders' Grant (NHBG) — it would improve supply not only for first-time buyers but also for renters. Moreover it would stimulate supply at the affordable end of the market, because an investor could get more grants by building a large number of cheap homes than a small number of expensive ones.

Because first-time buyers and new construction traditionally account for similar fractions of turnover in the housing market, turning the FHOG into an NHBG would be approximately budget-neutral, and could be made exactly so by suitably adjusting the size of the grant.

2.3  Owner-occupants vs. investors

Ordinary home buyers (owner-occupants) cannot claim their interest payments against imputed rent, let alone other income, whereas investors can claim interest and other expenses against rent and other income. Consequently, prospective owner-occupants cannot borrow as much as prospective investors. Previous owner-occupants usually have a compensating advantage in the form of tax-free capital gains; but first-time owner-occupants do not.

If first-time owner-occupants were given the option of being taxed as investors — claiming deductions for interest and other expenses, and being taxed on their imputed rent and half their capital gains — they could overcome their competitive disadvantage. Because this arrangement would be optional, it would not break any explicit or implicit promise; one can break a promise by taking away options or changing options, but not by giving additional options. To ensure that this “additional option” stimulates construction of housing and does not simply drive up prices, the option should be available only to purchasers of previously unoccupied homes (like the CAG).

One might call this proposal “HomeChoices” or “TaxChoices”. But, as the word “choices” has been profaned by past abuses, I shall call it the interest deduction option instead.

Because owner-occupants are not speculators, the 12-month qualifying period for discounting of capital gains should be waived for owner-occupants who take up the interest deduction option.

Australia's property investors now claim more in deductions than they declare in rental income. In other words, the average Australian property investor is negatively geared, paying negative tax on a recurrent basis (that is, excluding capital gains tax). This average is across all investors, including those who have partly or fully paid off their loans. The average new investor is negatively geared by a greater margin. The clear implication is that the average first home buyer, mortgaged to the hilt, would be better off on a recurrent basis if taxed as an investor. In view of the discounting of capital gains, it is by no means certain that the inclusion of capital gains tax would tip the balance. The truism that the existing tax system favours owner-occupants is at best an exaggeration, and a worst a myth.

But if income tax were abolished, as is my preferred option, this entire subsection would become a dead letter.

2.4  A further note on negative gearing

Negative gearing, like the FHOG, could be modified so as to encourage new construction — e.g. by allowing full deductibility for investors in new homes, but not for future investors in previously occupied homes. However, because the present Treasurer promised before the election that “We won't be touching negative gearing,” I shall not pursue the matter any further. The minimum deemed income, the FHOG reforms, and the interest deduction option for owner-occupants of new homes will have to suffice.

2.5  Ending payroll tax and VFI

VFI (vertical fiscal imbalance) is the mismatch between the vertical distribution of taxing powers and the vertical distribution of spending responsibilities: the Commonwealth presently collects about 82% of all public revenue collected in Australia, but is directly responsible for only about 54% of public expenditure, while the States together collect about 15% of public revenue but are directly responsible for about 40% of public expenditure.

(Note: In this article, “State” usually means “State or Territory”.)

VFI is fundamentally incompatible with accountability. When a State spends revenue raised by the Commonwealth, each party will blame the other for any shortfall in delivery of the associated services: the State will say it is due to lack of funding, while the Commonwealth will say it is due to inefficiency or inappropriate use of funds.

The GST did not solve the problem, but made it worse because the revenue-sharing formula required the States to abolish some of their own taxes, leaving them with less control over their total revenue, hence more at the mercy of the Federally-imposed grants system, hence better able to blame the Commonwealth for inadequate services. Furthermore, the revenue-sharing formula itself became a scapegoat as various States claimed that they were subsidizing other States or not being subsidized enough by other States. Meanwhile the Commonwealth, by pointing to increased GST revenue and ignoring demand for State-delivered services, was better able to accuse the States of inefficiency or waste.

For similar reasons, giving the States a guaranteed share of income tax or of some other Federal tax would further exacerbate the problem.

Replacing tied grants with untied grants, or otherwise giving the States more discretion in their spending of Federal grants, would not solve the problem, but would merely make the claims and counter-claims more general. The States, instead of alleging insufficient funding for several specific purposes, would allege insufficient funding overall; and the Commonwealth, by giving the States more choice, would gain more scope for accusing the States of making bad decisions.

That said, the GST has one major advantage: because the GST base is relatively broad, a small and politically tolerable increase in the GST rate would give a large increase in revenue, hence a large and politically advantageous improvement in service delivery. Hence, if each State (or Territory) were free to set its own GST rate, the Commonwealth could easily wash its hands of the matter, saying that it is incumbent on each State to collect enough GST for its needs and to accept the political costs and benefits as they come.

(Of course, in view of Rudd's emphatic promise that the GST rate would not rise above 10%, the rate must first be reduced below 10%, and any subsequent “small” increase must not lift the rate above 10%. But this, as we shall see, can be arranged.)

Allowing each State to set its own GST rate raises a constitutional problem, namely that the GST, in so far as it applies to goods, is a duty of excise. Under s.90 of the Constitution, only the Federal Parliament may impose duties of excise.

The obvious solution is to let the Commonwealth impose and collect the GST, and refund it to the State in which it is collected.

Under s.51(ii) and s.99 of the Constitution, the Commonwealth cannot discriminate between the States in matters of taxation or revenue. This would rule out a Federally-imposed tax with different rates in different States if the rates were unilaterally determined by the Commonwealth, but not if the rates were requested by the respective States. Let the Commonwealth say to the Parliament of each State: “Pass a Request and Consent Act specifying the desired GST rate in your State, and we'll collect GST at that rate in your State and refund it to the consolidated revenue fund of your State, subject to the same conditions as in all other States.” As long as the Commonwealth makes the same offer to each State and keeps its side of the bargain, there is no discrimination.

I emphatically reject any suggestion that the “Request and Consent” procedure circumvents the intent of the Constitution. If the intent of s.90 is to give the Commonwealth complete power over the taxation of goods prior to consumption, as the majority of the High Court has held since the Parton case (1949), then that power includes the power to decide that goods may be taxed in each State according to the revenue requirements of that State. If, to the contrary, the intent of s.90 were only to prevent the States from frustrating Federal tariff policy, as was unanimously affirmed by the High Court in Peterswald (1904), then the States would be free to impose their own broad-based consumption taxes, because such taxes do not discriminate between local and imported goods. The issue of discrimination between the States would then not arise. Either way, collecting the GST in each State at a rate requested by the Parliament of that State is consistent with both the letter and the spirit of the Constitution.

Under the present system, GST revenue is shared under a needs-based horizontal fiscal equalization (HFE) formula, whereby the revenue returned to each State may be more or less than what is collected in that State. If, as proposed here, the revenue returned to each State were precisely equal to the revenue collected in that State, the HFE element in the GST distribution would be eliminated. This by itself would create winners and losers, hence political difficulties for the Commonwealth. However, the HFE adjustments presently incorporated into GST-sharing grants can be shifted into other grants; there is no need for winners and losers in the transition.

To minimize compliance costs, the modified GST should replace not only the old GST, but also other State taxes that feed into prices of goods and services; these include stamp duties on vehicles and insurance policies, and of course payroll tax. The elimination of payroll tax would be especially beneficial in terms of international competitiveness.

2.6  Keeping the GST rate below 10%

In the absence of any reduction in the States' spending responsibilities, the elimination of indirect State taxes would require an increase in the GST rate, contrary to Rudd's election promise. Indeed, allowing the States to set their own GST rates and eliminating HFE from the GST distribution would tend to push the GST rate above 10% in some States, even if indirect State taxes remained in place; eliminating the latter would ensure that the GST rose in all States.

To prevent this unacceptable outcome, some spending responsibilities must move from the States to the Commonwealth. Two possibilities readily come to mind. First, Rudd could carry out his threatened Commonwealth takeover of public hospitals. Second, there could be a Commonwealth takeover of all aspects of family law, including child protection.

A Commonwealth takeover of public hospitals would clearly improve accountability. As long as the States continue to pay for public hospitals while the Commonwealth pays for GP services through Medicare, there will always be opportunities for cost-shifting and blame-shifting. The demarcation between State-funded hospitals and Federally-funded aged-care facilities raises a similar set of problems. The only solution is to bring the entire system under one government. The Commonwealth can do this unilaterally; under s.51(xxiiiA) of the Constitution, the Federal Parliament has a plenary power to make laws with respect to the “provision of ... medical and dental services”.

For the sake of efficiency and transparency, the Commonwealth should also have full responsibility for family law. Section 51 of the Constitution, at paragraphs (xxi) & (xxii), empowers the Federal Parliament to make laws with respect to “marriage” and “divorce and matrimonial causes; and in relation thereto, parental rights, and the custody and guardianship of infants”. The restrictive wording leaves the States responsible for orphans, de-facto relationships and the children thereof, domestic violence, and child protection. These residual State powers become more pervasive as divorce becomes more frequent and as non-traditional relationships and reproductive methods become more prevalent. The resulting tangle of State and Federal laws adds to administrative and legal costs.

While the Constitution does not give the Federal Parliament a plenary power with respect to family law, s.51(xxxvii) allows individual State Parliaments to refer legislative powers to the Federal Parliament. Moreover, a substantial part of State budgets, including GST revenue, is dependent on s.96 of the Constitution, whereby the Federal Parliament can make conditional grants to the States — and specify the conditions down to the last detail. Hence the Commonwealth, using s.96, could cut off funding for those aspects of family law that are currently under the control of the States, leaving the States with no option but to refer the associated powers to the Commonwealth.

After the Commonwealth had taken over responsibility for health and family law, the States would have less need for GST revenue, so that the GST rate set by the Parliament of each State could be made subject to a Federally-imposed cap of 10%. The abolition of indirect State taxes would be the condition under which the GST collected in each State would be refunded to that State.

While the transition to this arrangement would involve a somewhat heavy-handed exercise of Federal power, the final arrangement would give the States far more flexibility in how much revenue they raise, and far more freedom in how they spend it, than they currently enjoy. Indeed it would be possible for the Commonwealth to impose far more conditions on the refunding of GST revenue without reducing the fiscal independence of the States. For example, in order to minimize compliance costs and deadweight costs, the Commonwealth could demand that the bases of State property taxes be harmonized in the manner suggested by Prosper Australia's submission to the NSW Review of State Taxation; but as long as the States remained free to set the rates and thresholds, their flexibility in how much revenue they raise (and how they spend it) would be undiminished.

Of course the Federal takeover of health and family law would require additional revenue or savings at the Federal level. Some of this would come from the DMI or the growth dividend thereof, while some would be a growth dividend from the reforms described below.

2.7  Cutting compliance costs: Make the GST a retail tax

If the GST remained a value-added tax (VAT), the emergence of different rates in different States would greatly increase compliance costs, mainly because an enterprise in one State would have to claim input credits on inputs from multiple States at multiple rates. The solution is to get rid of input credits by turning the GST into a single-stage retail sales tax, better known in the USA as a “General Sales Tax”.

Eliminating the need to record and claim input credits is not the only mechanism by which a retail tax would reduce compliance costs. Non-retail businesses would no longer pay or collect GST, and therefore would no longer need to be registered for GST purposes. The long-running legal uncertainty over the precise point at which exported goods become zero-rated (“GST-free”) would no longer matter, because none of the entities involved would be retailers. Entities with turnover below the registration threshold would no longer be forced to register simply because potential customers want to claim input credits. Non-registered (input-taxed) business customers would be treated the same as retail customers. But no tax would be collected from registered or exempt customers (where “exempt” now means “GST-free”, not “input-taxed”). If some enterprises, in order to be able to treat all customers alike, decided to deal exclusively with non-exempt retail or non-registered customers, or exclusively with exempt or registered customers, so be it; this would become the new boundary between “retail” and “wholesale” businesses. To avoid any new complications, the rules for determining which items are taxable and which are not, together with the averaging rules for small businesses with both taxable and non-taxable product lines, could all be left as they are.

The overall reduction in compliance costs would lead to lower prices, hence reduced pressure on interest rates.

A retail business with shop fronts in multiple States would of course have to charge a different GST rate in each State; but that particular compliance cost would be minimal by comparison with the other costs of maintaining multiple shop fronts.

Should a mail-order retailer selling into multiple States from a single location be required to remit GST at the rate applicable in the buyer's State, or in the seller's State? Obviously the latter option would result in lower compliance costs. If this causes internet/mail-order retailers to congregate in the State with the lowest GST rate, so be it.

2.8  Mythical advantages of input credits

It is argued that the value-added type of GST encourages entities to register for GST in order to claim input credits. But by that logic, income tax deductions encourage entities to join the income tax system in order to claim the deductions, which are more comprehensive than input credits. And having declared sales for income tax purposes, one cannot hide them for GST purposes, whether the GST involves input credits or not.

It is argued that input credits encourage entities to declare all sales so that they can claim all associated input credits. But by that logic, income tax encourages entities to declare all sales so that they can claim all associated deductions. And again, having declared sales for income tax purposes, one cannot hide them for GST purposes, even if the GST is a retail tax.

It is argued that when a retailer evades a value-added GST, only the value added by the retailer escapes tax, whereas under a retail tax the entire value of the sales would escape tax. But if the paper trail from the wholesaler's income tax were followed, how could the retailer escape detection? Those who point to the paper trail created by tax invoices, which are used for claiming input credits, fail to explain what is wrong with the paper trail created by ordinary invoices used for income tax deductions. Indeed, for purchases under $75, any invoice that is good for income tax purposes is also good for GST purposes!

Furthermore, input credits create opportunities for two particularly nasty forms of fraud, namely (i) faking input credits and, if caught, accusing the alleged supplier of failing to declare sales, and (ii) failing to declare sales and, if caught, accusing the customer of faking input credits or deductions. The absence of input credits would eliminate problem (i).

The above observations may help to explain why the broad-based indirect tax proposed by Treasurer Howard in 1979 and 1981, like the one “preferred” by Prime Minister Hawke and Treasurer Keating in 1985, was a retail tax. Prime Minister Rudd and Treasurer Swan should note the precedents.

That said, I must concede one residual point in favour of a value-added tax: If, instead of a retail GST plus an income tax, we had a retail GST plus a VAT, then all of the aforesaid anti-evasion features of the income tax would be shared by the VAT!

2.9  Taking superannuation out of export prices

Recall that the compulsory employer-funded superannuation contribution is equivalent to a 9% Federal contribution paid for by a 9% Federal payroll tax. If that notional payroll tax were moved onto the same base as the GST and imposed at such a rate as to raise the same amount of revenue as the notional payroll tax, the shift would have no noticeable effect on the cost of living, but superannuation contributions would no longer feed into export prices or discriminate against local products relative to imports. The replacement for the notional payroll tax could be called the superannuation levy. It would be calculated on the GST base and remitted to the ATO, but not forwarded to the States.

Because the superannuation levy would be imposed pursuant to the Federal taxing power, it would be on a firmer constitutional footing than the present employer-funded contribution, which is not technically a tax.

Shifting the superannuation burden off payrolls and onto consumption would improve both equity and efficiency. Employers, simply by creating jobs, already contribute to the nation's capacity to save for retirement. To make employers pay for superannuation in proportion to their payrolls is not only to demand double duty, but also to deter employment, and thence to reduce the nation's capacity to save. Thus a superannuation contribution based on payrolls defeats its own purpose. A superannuation levy on consumption causes no such problems, because consumption is the opposite of saving; in so far as such a levy deters consumption, it promotes saving through other channels and thereby assists the purpose of the levy.

It might seem that the proposed shift, by taking part of the old superannuation burden off exports and imposing part of the new one on imports, would increase the cost of living within Australia. But to argue thus is to overlook two vital points. First, harmonizing the bases of the GST and the superannuation levy would reduce compliance costs; this saving, under the influence of competition, would be passed on in prices. Second, due to the floating exchange rate, the reduction in export prices would be partly — but only partly — offset by an upward movement of the currency, which in turn would tend to reduce the cost of living as expressed in the local currency.

One might object that the proposed superannuation levy would turn a private-sector transfer into a tax-and-spend transfer — in other words, that it would cause a one-off increase in revenue offset by a one-off increase in expenditure. But if the Howard government could take the GST off-budget, notwithstanding that the GST was imposed pursuant to the Federal taxing power and is unashamedly called a tax, then surely the Rudd government can keep the superannuation levy off-budget, notwithstanding that it too is technically a tax. The off-budget treatment would be more convincing if the Federal superannuation contribution were exactly funded by the levy; to facilitate this, the proportionality between the beneficiary's wages/salary and the Federal contribution could be allowed to differ slightly from the target value (currently 9%).

Detractors might allege that shifting the superannuation levy onto the GST base would amount to an “over my dead body” increase in the GST rate. But such an allegation would not be comparing like with like. The new GST, like the old, would be for the States, and the new rate would be no higher than the old. The new superannuation levy, like the old employer-funded contribution, would be for saving for retirement; and the new arrangement, in view of the effect on international competitiveness, would be less burdensome than the old. Alternatively, if the new superannuation levy is considered part of the new GST, then in fairness the old superannuation contribution must be considered part of the old GST.

Moreover, if the reform suggested in Subsection 2.11 (below) were adopted, the superannuation levy would not need to be on the same base as the GST (but would still be on a consumption base).

2.10  Inferior option: Dumping PAYE compliance costs on the banks

The administrative cost of deducting personal income tax from wages and salaries deters employment and feeds into prices, including prices of exports and of local products that compete with imports. The trend towards part-time and casual employment has compounded the problem by increasing the compliance cost per hour of employment.

Shifting this compliance burden off employers and onto financial institutions would not only remove the disincentive to hire employees, but would reduce the overall cost due to economies of scale and specialization. It would also make the banks earn their fees! Instead of nominating one employer (if you have more than one) from whom to claim your income tax threshold, you would nominate one deposit-taking account into which all your taxable income is paid. Each of your employers would pay your gross wages/salary into the nominated account, and your financial institution would cumulatively deduct tax from the cumulative taxable deposits. Any social security payments that are taxable would be treated the same way. At the end of each financial year, you would get one bank statement instead of one or more group certificates. The need for an end-of-year adjustment would be greatly reduced if the “cumulative” deduction of tax were calculated pro-rata since the beginning of the financial year; financial institutions could easily automate this process.

Anti-avoidance provisions would obviously include treating all electronic deposits from unrecognized sources as taxable. Anyone who tried to evade tax by using more than one account would be caught in the same way as anyone who now tries to claim the tax-free threshold from more than one employer — and would be caught more easily, because financial institutions are less numerous than employers.

That said, the whole PAYE system would be unnecessary if there were no income tax.

2.11  Superior option: Getting rid of income tax

Question: What's the difference, in terms of assessment, between an income tax and the broadest possible VAT?

Wrong answer: From the viewpoint of the enterprise, export income is assessable for income tax but not for VAT, while outgoing wages/salaries are deductible for income tax but not for VAT. From the viewpoint of the employee, wages/salaries are assessable for income tax but not for VAT. Any further differences are peculiar to the jurisdiction, not essential to the natures of the two taxes.

The answer is “wrong” because under the present PAYE system, the primary burden of assessing and remitting employees' personal income tax falls on the employer, i.e. the enterprise. The “viewpoint of the employee” is relevant only for the end-of-year income-tax adjustment, which is minor unless the employee also happens to be conducting an enterprise (e.g. property investment); for the employee per se, the net wage/salary is real, while the gross wage/salary and the income tax paid therefrom are merely notional. So the viewpoint of the enterprise is the one that matters. Moreover, even from the enterprise's viewpoint, the income tax remitted on behalf of each employee may be understood as a function of net pay, in which case it is equivalent to a payroll tax — even to the extent of being deductible against taxable income! So, taking the viewpoint of the enterprise and treating net wages/salaries as “given”, we restate our answer as follows:

Right answer: Under a VAT, export income and outgoing wages/salaries are ignored; but under an income tax, export income is assessable while outgoing net wages/salaries are effectively subject to a payroll tax and, together with the payroll tax, are deductible from assessable income. Any further differences are peculiar to the jurisdiction, not essential to the natures of the two taxes.

(The fact that income tax treats capital transactions in terms of interest, depreciation and capital gains, whereas a VAT treats them in terms of purchase prices and resale prices, is taken as a jurisdictional peculiarity in the above “answer”, but accepted in the remainder of this article.)

Notice that under the income tax, the deduction for wages/salaries tends to cancel the effect of the “payroll tax”. So, for given net wages/salaries, substituting a VAT for income tax would have little effect on the tax implications, hence the price implications, of hired labour. The main effect of the substitution would be that export prices would be taken out of the tax base while import prices would be brought into it; in other words, part of the old tax burden would be shifted from exports to imports. This by itself might be expected to increase the cost of living within Australia. But, as with the substitution of a consumption tax for payroll tax, the effect would be offset by a reduction in compliance costs and by an automatic appreciation of the currency. The old truism that substituting a VAT for income tax would be “regressive” is based on the assumption that gross wages/salaries would be held constant; if, instead, net wage/salaries were held constant, the cost of living would hardly change, so the purchasing power of every employee's net wages/salary would be maintained.

Let us enumerate some of the savings in compliance costs. With a VAT there is no accounting burden for “consumers”, including wage/salary earners and ordinary home owners. But residential landlords would be treated as “enterprises” because residential rents would be assessable under the “broadest possible” VAT. As residential landlords are already outside the PAYE system, the requirement to register for VAT would not increase their compliance costs. With a VAT there is no need for depreciation allowances, because business inputs are deducted up-front unless there are insufficient sales to deduct them against, in which case the excess deductions are carried forward; but, to ensure that the deemed minimum income (DMI) serves its purpose, the cost of an investment property must be deductible only against the resale price thereof or actual rental income therefrom — not against any unrealized portion of the DMI. Similarly there is no need for specific provisions concerning capital gains; if an asset is bought and later resold for a higher price, the tax on the sale will outweigh the input credit on the purchase. Under a VAT there is no case for special provisions concerning fringe benefits, inheritances, or gifts, because the recipients are taxed as they spend, not as they earn or receive. In particular, an heir who reinvests an inheritance or continues to manage an inherited business automatically escapes tax on the inheritance per se, whereas one who sells an inheritance and spends the proceeds will pay tax on the spending. The needs of superannuation are also automatically covered: superannuants are taxed only as they spend, and therefore automatically escape tax if they choose to “roll over” a payout; but there would be no “devaluation” of existing superannuation accounts because, as explained above, the transition to a VAT would be managed with no overall increase in prices.

A further reduction in compliance costs, together with a reduction in the GST rate, could be achieved by harmonizing the retail GST base with the broader VAT base. This could be done on a State-by-State basis: the Parliament of each State could be invited not only to specify the GST rate in that State, but also to request that the base be changed from the old GST base to the new VAT base. In Canada, a provincial retail tax on the same base as the Federal “GST” (VAT) is called a harmonized sales tax (HST).

The problem of preserving net wages/salaries in the transition to a VAT has been made much easier by the Howard government's Federal takeover of workplace relations. Because the takeover was not complete, there would be some cases in which the cooperation of the States would still be needed. But such cooperation would be easily obtained under “wall-to-wall” Labor governments, and in any case could be secured using s.96.

In the short term, the most obvious method of preserving net wages/salaries is a comprehensive no-disadvantage rule: no disadvantage for current workers who keep their present hours; no disadvantage for current workers who change their hours, compared with those who were already working the alternative hours; and no disadvantage for new workers compared with current workers on the same hours.

By itself, a no-disadvantage rule probably cannot be made rort-proof, because under the present system the “net” wage corresponding to given gross wage from a given employer depends on whether the worker is claiming the tax-free threshold from that employer. Presumably employers would preferentially hire workers who do not claim the threshold, and who therefore can be paid less. To limit such abuses, the no-disadvantage rule must be backed by a safety net. Furthermore, under the present income tax, a part-time worker keeps a greater fraction of his/her pay than a full-time worker employed under otherwise identical conditions, so that (for example) a half-time worker get more than half the net pay of a full-time worker. To preserve this feature, the safety net should take the form of a minimum hourly rate (or piece rate) plus a shift bonus — that is, a separate payment for a completed shift or day's work.

Regardless of the present relativities between full-time and part-time workers, a mandatory shift bonus is several thousand years overdue. A shift bonus compensates workers of shorter shifts for the greater fraction of their wages that is consumed by travel costs, and the greater fraction by which their paid time is augmented by unpaid time, such as travel time and preparation time. The need for a shift bonus was made more acute by the Howard government's workplace reforms, which allowed employers to demand arbitrarily short shifts. For these reasons, a mandatory shift bonus should be included not only in the safely net, but also in all enterprise-bargaining agreements negotiated thereafter.

That said, one should also remember that the package of reforms proposed herein would be highly conducive to job-creation, so that competition between employers would minimize the need for no-disadvantage tests and safety nets.

Concerning the rate of the VAT, I make two points:

• The rate should be such as to make the overall package revenue-neutral;

• Because income tax rates are quoted for a tax-inclusive base, the quoted VAT rate should also be tax-inclusive in order to avoid a misleading impression of an increase in taxation.

Given those specifications, Treasury should be able to crunch the numbers and come up with a rate.

If the proposed VAT were introduced in lieu of income tax, the superannuation levy could be imposed on the VAT base instead of the retail GST base, making it harder to claim that the levy amounted to an increase in the GST (and reducing the rate of the levy, thanks to the broader base and tax-inclusive rate). But it would still be possible to keep the superannuation levy off-budget.

3.  Constitutional threats to existing taxes

3.1  Are compliance costs unconstitutional?

Section 82 of the Constitution declares:

The costs, charges, and expenses incident to the collection, management, and receipt of the Consolidated Revenue Fund shall form the first charge thereon...

At face value, this would seem to imply that the compliance costs of Federal taxes are fully recoverable from the Commonwealth, and not merely deductible against taxable income or sales. The implication is especially clear in respect of any tax which one private entity is required to collect from another — such as the GST, which sellers collect from buyers, and personal income tax, which employers remit on behalf of employees.

Compliance costs of State taxes are similarly problematic in view of, e.g., s.40 of the NSW Constitution. Again this is especially the case when one entity collects tax from another, as with the stamp duties on insurance premiums and new vehicle registrations.

If a High Court judgment were to confirm my suspicions, governments would have a strong incentive to minimize compliance costs. Accordingly I regard such a judgement as highly desirable. In the mean time, the Commonwealth would be well advised to do everything in its power to minimize compliance costs, in order not only to minimize the risk of a constitutional challenge thereto, but also to minimize the fiscal impact if such a challenge were to succeed.

[P.S. (July 6, 2009): See “Making the tax system comply with s.82 of the Constitution”.]

3.2  Are stamp duties on new vehicles unconstitutional?

Under s.90 of the Constitution, only the Federal Parliament may impose duties of customs or excise.

The currently prevailing definition of duties of excise was given by a 4-3 decision of the High Court in Ha v. NSW (1997), hereinafter called Ha. The majority held that:

... duties of excise are taxes on the production, manufacture, sale or distribution of goods, whether of foreign or domestic origin. Duties of excise are inland taxes in contradistinction from duties of customs which are taxes on the importation of goods. Both are taxes on goods, that is to say, they are taxes on some step taken in dealing with goods.

The minority drew a different distinction between customs and excise duties, namely that customs duties discriminate against imported goods in favour of locally produced goods, while excise duties do the opposite:

A State tax which fell selectively upon imported goods would, of course, be a customs duty and be prohibited by s.90. A State tax which fell selectively upon goods manufactured or produced in that State would be an excise duty and be prohibited by s.90. A State tax which discriminated against interstate goods in a protectionist way would offend s.92 and be invalid.

In the majority view, the Constitution gives the Commonwealth exclusive control of the taxation of goods and therefore prevents the States from taxing goods. In the minority view, the Constitution gives the Commonwealth exclusive control of tariff policy (be it protection or free trade) and therefore only prevents the States from taxing goods in ways that discriminate between local and non-local goods.

The stamp duty on the sale or registration of a new vehicle, especially if collected and remitted by the dealer, would appear to be a “tax” on the “sale” of a “good” (the vehicle), or at least on “some step taken in dealing with” the good, in which case it is an excise according to the majority view. But because it does not discriminate between imported and locally-produced vehicles, it would not be a duty of customs or excise according to the minority view. In either case, it is sufficiently clear that taxes on second-hand goods, including stamp duties on second-hand vehicles, are not excises. [See P. Sampathy, “Section 90 of the Constitution and Victorian Stamp Duty on Dealings in Goods”, Journal of Australian Taxation, vol.4, no.1 (2001), pp.133–155.]

The majority and minority definitions in Ha essentially agreed with the majority and minority definitions, respectively, in the 3-2 decision of the High Court in Parton (1949), in which the majority ruled that a tax of 1/8 of a penny per gallon of milk sold or distributed in Melbourne was an excise.

Paradoxically, the minorities in Ha and Parton could claim support from the unanimous judgment in the first case in which the scope of “excises” came before the High Court, namely Peterswald (1904). In this case the “discriminatory” definition of excises was not decisive, because the disputed tax, as applied to the case in question, was a “fee” for a licence to brew beer, and as such discriminated against local brewers. The tax was held not to be an excise because it was only loosely connected with goods; in particular, it was not apportioned to the quantity or value of goods sold. That the definition adopted in Peterswald was not critical to the outcome may help to explain why the majority in Parton did not feel bound by it. That the Peterswald judgment was written by Sir Samuel Griffith — who, as the author of the first draft of the Constitution, may be presumed to have had some idea what the founding fathers meant by “excise” — may help to explain the lingering support for the definition contained therein. But, as that definition has not prevailed, the Commonwealth would be well advised to demand the abolition of vehicle stamp duties in any reform of Federal-State financial relations.

3.3  Is payroll tax an excise?

Payroll tax falls on labour, including labour engaged in the “production, manufacture, sale or distribution of goods”, and feeds into prices, including prices of goods. Furthermore, because it applies to labour engaged in the “production, manufacture, sale or distribution of goods” only within the State imposing the tax, it taxes local goods more heavily than imported goods at the same stage of refinement, and therefore has as much capacity to frustrate Federal tariff policy as any other tax on goods. Hence a constitutional challenge to payroll tax would attract sympathy from both the Peterswald and Parton factions of the High Court.

The Commonwealth would therefore be well advised, not only for economic reasons but also for legal reasons, to demand the abolition of payroll tax in any reform of Federal-State financial relations.

4.  Q&A

4.1  How would the minimum deemed income (MDI) affect rents?

The minimum deemed income (MDI) from an investment property, in combination with income tax or an ultra-broad VAT (and sometimes GST), would impose a minimum holding cost on the property. For a fully developed and occupied property, the actual rental income would exceed the MDI, which therefore would have no effect on the tax payable, and hence no possible effect on the rent. For an unoccupied property, the holding cost would pressure the owner to seek tenants in order to cover the cost (or sell the property to an owner-occupant or to someone who will seek tenants). This pressure would increase the availability of rental accommodation and therefore tend to reduce rents across the board. For a partly occupied property, the actual rental income might fall short of the MDI, in which case the owner would be pressured to seek additional occupants — which might involve extending or rebuilding the accommodation so as to take maximum advantage of the zoning. This too would tend to increase the availability of accommodation and reduce rents.

4.2  Is the MDI a land tax?

The land taxes levied by the several States are calculated on site values. As the MDI would also be calculated on the site value, one might suggest that the MDI amounts to a Federal land tax. It does not, for the following reasons:

• A Federal land tax would be imposed in addition to the tax on any actual rent from the property, whereas the MDI merely places a floor under the tax liability; if the actual rental income exceeds the MDI, only the tax on the actual income is payable. Alternatively, we may think of the MDI as imposing an additional holding cost, albeit one which is phased out as rental income increases, and which falls to zero as the rental income rises to the MDI; but in the case of land tax, the additional holding cost is not phased out in this way.

• In combination with an income tax (but not an ultra-broad VAT) the MDI is less burdensome than a land tax for asset-rich, income-poor taxpayers, because the MDI is taxed at the marginal income-tax rate, whereas a land tax liability is independent of income; even under a so-called “progressive” land tax, the tax rate increases with the total value of the taxpayer's sites, not with income.

That said, the comparison with land tax yields further insights concerning the effect of the MDI on rents.

First note that a sales tax can be passed on in prices because a seller who delays a sale in search of a higher price also delays the realization of the tax liability, and therefore can afford to delay the sale. But a landlord who tries to withhold a property from the market in search of a higher rent does not thereby delay realization of land tax, because the tax is payable whether the property is let or not. So a land tax, far from being passed on in rents, tends to reduce rents by pressuring landlords to let their properties sooner, in order to obtain income to cover the tax. Even if the land tax, like the MDI, exempts owner-occupied residential properties, it is still fully payable in respect of unoccupied properties, so its efficacy in forcing properties onto the market is undiminished.

Now if the MDI is considered as an additional holding cost which is phased out as actual rental income rises to the deemed level, a landlord who obtains a tenant not only covers the additional holding cost, but avoids it. But in the case of land tax, the landlord who obtains a tenant merely covers the additional holding cost. Thus, for a given minimum holding cost, the MDI is more efficacious than land tax in encouraging landlords to seek tenants, and therefore has a greater tendency to reduce rents.

4.3  Is this a “beggar my neighbour” strategy?

One might allege that the policies proposed here, in so far as they would raise Australia's market share, would help Australia only at other countries' expense. That would be a gross distortion, for the following reasons:

• The benefit to Australia would not be unrequited; the increased flow of currency into Australia would be matched by an increased flow of goods and services in the opposite direction.

• The aforesaid exchange would be mutually beneficial in that Australians would want the currency more than the goods and services, while the foreign customers would want the goods and services more than the currency; if this were not the case, the exchange would simply not take place.

• Any policy making Australian products cheaper in other countries would tend to reduce inflationary pressures in those countries, allowing them to maintain more accommodating monetary policies, hence lower unemployment.

• There is nothing to stop other countries from following Australia's lead or otherwise removing taxes from export prices (see, e.g., the proposal by FairTax.org — which is not what I am proposing here, and which I discovered only two days before completing the first draft of this article). Tax competition of this kind is not harmful; it neither reduces funding for essential services nor shuts down international trade. And while it is not possible for all countries to increase their fractional shares of global markets, it is possible for all of them to increase their absolute shares if they all adopt policies that increase trade. Such an outcome is highly desirable, especially in a time of threatened or actual recession.

The real “beggar my neighbour” policies are not border-adjusted taxes such as VATs and retail taxes, which open the borders to trade, but “protective” tariffs which close the borders to trade and thereby inhibit the pursuit of economies of scale, economies of specialization, and comparative advantage.

Moreover, the policies proposed herein would make indirect taxes less damaging to employment, and would reduce compliance costs, hence inflationary pressures, hence interest rates and the “natural rate” of unemployment — that is, the minimum unemployment rate consistent with stable inflation. These benefits do not depend on increased market share.

5.  Conclusion

I submit that in order to to ease capacity constraints, remove barriers to employment, reduce compliance costs (hence inflationary tendencies and interest rates), and make Australian products more competitive in global and domestic markets, the Federal government should:

• Deem every investment property to be earning a minimum of 3.5% of the site value per annum, for tax purposes;

• Deem every non-residential investment property to be earning a minimum of 3.5% of the site value per annum, for GST purposes;

• Turn the first home owners' grant into a new home builders' grant, available to all purchasers of new (previously unoccupied) homes;

• Complete the Federal takeover of health services (including public hospitals) and family law (including child protection);

• Turn the GST into a retail tax;

• Set the GST rate in each State according to the request and consent of the Parliament of that State (with a 10% cap), and return the GST collected in each State to the Treasury of that State, subject to the abolition of indirect State taxes (including payroll tax, and stamp duties on goods and services);

• Separate HFE from the GST distribution;

• Replace income tax with a VAT on the broadest possible base (with the rate quoted for a tax-inclusive base) while preserving net wages/salaries by means of a no-disadvantage rule and a safety net, including a per-shift payment in addition to the minimum hourly rate or piece rate; give the Parliament of each State the option of harmonizing the GST base in that State with the VAT base; and retain the DMI for VAT purposes, with the proviso that the purchase price of an investment property is claimable only against the resale price thereof or actual rental income therefrom.

• If the preceding step is deemed to be too hard, shift the burden of withholding personal income tax off employers and onto financial institutions, and give first home buyers the option of being taxed as investors provided that they purchase new homes;

• Replace the 9% employer-funded superannuation contribution with a Federal contribution funded by a Federal “superannuation levy” on the VAT base (or, if there is no VAT, the GST base).

[Reposted with updated links, August 18, 2012.]