SENATE SELECT COMMITTEE ON A NEW TAX SYSTEM

SUPPLEMENTARY REPORT

SENATOR ANDREW MURRAY

This first stage of the Senate's Inquiry has been an essential precursor to the more detailed examination of the components of the Tax Package by the Senate Committees. This is because it is necessary to first examine the credibility of the arguments presented by the Government in favour of tax reform generally, and their proposals in particular.

Public sector [1] and private sector [2] analysts appear to agree that the revenue forecasts underpinning the Government's package are likely to be met and are affordable. In assessing the package, the Committee has had to determine whether the Government's claims of substantial economic and equity benefits flowing from the package are justified.

As an opening comment, we must express disappointment with the paucity of quantitative analysis presented to the Committee by the Treasury. It has emerged in the course of the Inquiry that the Government has not modelled the impact of the package, has made assertions about the consequences of further changes (such as exempting food) which are simply not backed up by the evidence, and has quite inappropriately underestimated the impact of tax reform on low income earners.

These are important findings. The Government is insisting that the Senate pass the tax package intact, yet has not come clean with the Australian people about the true effect of the tax reform package.

To correct the failure to present information by the Government and Treasury, the Senate Committee has itself engaged modellers to look at the impact of the package. The macroeconomic modelling report, by internationally renowned economist Professor Peter Dixon of Monash University's Centre of Policy Studies looked at the impact of tax reform under a series of different assumptions. His report was refereed by prominent private sector modeller Chris Murphy of Econtech.

The distributional effect of the tax reform package and the price effect on different households is to be modelled by Australia's leading social modellers, Professor Ann Harding of the University of Canberra's National Centre for Social and Economic Modelling, and Professor Neil Warren of the University of New South Wales Australian Taxation Studies Program. That report is due in about six weeks.

These modelling reports are vitally important because they provide the information which the Government has failed to provide. The Democrats wish to commend the Senate President, Senator Reid, for agreeing to fund these important studies. The public needs to know what will be the real effect of the Tax Reform package. They need to know what the effect will be of changing that package.

The public needs to be better informed than the Government to date has been prepared to make them.

1. THE CASE FOR TAX REFORM

The Government has argued that the current tax system is failing Australia, that:

• The wholesale sales tax system is no longer appropriate in an economy dominated by services rather than goods;
• The direct tax system combines high marginal rates at low thresholds, and high effective marginal tax rates when social security payments are taken into account;
• The current tax system places Australian exporters and investors at a disadvantage in a globalised market place;
• The system is ineffective because of a declining tax base, raising questions about whether further social expenditure can be funded;
• The tax system has become too complex, opening up opportunities for avoidance and minimisation. [3]

Several witnesses to the Committee recognised the need for tax reform for various reasons. The Business Coalition for Tax Reform said:

“The current tax system detracts from our International competitiveness… and is also unfair…The structural weaknesses of our tax system, with its selective high rates, act as an enticement to avoidance and evasion, which is unfair and which ultimately undermines confidence in and respect for, our tax system.” [4]

The President of the Australian Council of Trade Unions said:

“An income tax cut for our people is warranted, particularly to compensate for bracket creep.” [5]

The President of the Australian Council of Social Services said:

“We believe that there is a need for tax reform. We believe that the revenue is at risk-in fact, revenue has declined. We believe that tax reform has to be comprehensive on the income tax side, on the consumption tax side, on the business tax side, and in federal-state financial relations. We believe strongly that, in all of that, we also need to address the situation of families and that the poverty traps have been addressed. The package that the government has put forward addresses a number of those things, but there are substantial problems with the package and the substantial modifications that need to be made to this package in order to get a balance between equity and efficiency.” [6]

The President of the National Farmers Federation said:

“The mess of our tax system has done more to hamper farmers than help them, largely because it takes their inputs long before they see a dollar of profit. A new tax system which will support them is crucial to improving the future viability of agriculture in Australia.” [7]

Tax expert Professor Neil Warren said:

“Those macro benefits (of tax reform) are marginal….But I re-stress that this is about integrity in tax systems, robust indirect tax systems and a concern that state tax proliferation needs to be addressed… That to me is the issue and I think it is unreasonable to say that you are going to create hundreds of thousands of jobs and that export services are going to boom: it is about housekeeping again”. [8]

Macro-economist Professor John Quiggin said:

“There is no general reason why we should think it desirable to tax goods and not tax services. We just get an excessively narrow tax base and that narrowing will continue as services take up a larger and larger share of economic activity, which will inevitably happen…..In the current circumstances, we are discouraging the consumption of goods and subsidising the consumption of services. That leads to a somewhat less efficient use of our resources – in the technical sense of the term – than would otherwise be the case.” [9]

The modellers commissioned by the Committee, Murphy and Dixon, agreed that the macro-economic benefits of the package were likely to be small, with the change in consumer welfare varying between –0.01% and +0.2%. These issues will be dealt with in more detail later. But, even if the macroeconomic effects are modest, the real need for tax reform stems from the need to repair the revenue base. This need applies to both direct and indirect tax.

2. INDIRECT TAX BASE DECLINE:

By any analysis, the indirect revenue base is in decline.

COMMONWEALTH TAXES SHARES % OF GDP

(* Ignoring increases in 1993-4 Budget in WST and Petrol excises)

Source 1998-99 Budget Paper No. 1)

This table shows that the indirect taxes collection has fallen from 7.2% of GDP to 5.8% of GDP over the last decade, and will fall to around 5.5% of GDP within four years. This fall would have been even more dramatic but for the $3 billion of sales tax and excise rises in the 1993 Budget. If these rises had not occurred, indirect tax revenue would be just 5.2% of GDP last year.

To restore indirect tax revenue to its 1995-6 level would require indirect tax increases by 2001-02 worth $2.8 billion, i.e. a near re-run of the 1993 Budget.

The Commonwealth has three main sources of indirect taxes – wholesale sales tax, customs and excises. The following table highlights how these taxes have performed over the last decade.

INDIRECT TAXES % OF GDP

(Source 1998-99 Budget Paper No. 1)

The largest fall has been in the collection of excises, due to unilateral tariff reductions. These have cost the Budget around $3 billion, despite surging import sales. Excises have also fallen sharply, by the equivalent of $5 billion, driven by changing consumption patterns of tobacco, alcohol and petrol. Indeed excises would have fallen further but for the 15% increases in the 1993 Budget, now adding around $1.6 billion a year to revenue.

Wholesale Sales Tax peaked at 2.8% of GDP in 1988-89, declining to 2.4% in 1992-3, then, boosted by $1.2 billion of rates rises, rising to 2.6% of GDP in 1995-6 before commencing a further slow decline. If the rate rises had not occurred, WST would have continued to fall to 2.2% of GDP, representing a revenue loss of $2.3 billion over nine years. As Geoff Carmody of Access Economics put it:

“It should not be difficult to understand that, if you have a tax base that applies to a very limited part of private final consumption expenditure and that part of private final consumption expenditure is declining relative to the total, unless you continually raise the tax rates applying to that narrow base, you will lose revenue.” [10]

Consumption of goods subject to Federal sales taxes and excises is in long-term relative decline. WST relies heavily on a narrow band of goods – with motor vehicles purchases and parts, alcohol, furniture, drapes, floor coverings, household appliances, printing and stationery, toys, sporting and recreational equipment comprising 67 per cent of WST collections. Petrol, tobacco and alcohol comprise 100% of excises. As a percentage of Private Consumption data, spending on these goods has fallen from 25.4% in 1985-6 to 22.2% in 1996-7, and continues to fall. The non-taxed sector (i.e. services, clothing, food and hospitality sectors) have grown at almost twice the rate of the taxed sector of consumption over the past decade.

PRIVATE FINAL CONSUMPTION EXPENDITURE

TAXABLE COMPONENT

(Average 1989-90 prices, “Taxable” is categories Cigarettes and tobacco, Alcoholic drinks, Household appliances, Furniture and floor coverings, Purchase of motor vehicles, Operation of motor vehicles, Books & papers, and Toys & sporting goods, which account for 67% of WST and 100% of excises)

Source: ABS 5204.0.40.002)

It should be noted that State taxes have grown over the last eight years, from 6% to over 7% of GDP. This has been driven by a modest rise in payroll tax, a modest rise in financial taxes (including stamp duties) but mostly a very large rise in gambling taxes. State taxes are mildly regressive, and any increase in the state taxes will have negative effects for distributional equity [11] and carry high deadweight losses. Reliance on gambling taxes is hardly socially responsible.

MAJOR STATE TAXES % OF GDP

(Source: ABS 5506.0)

Our indirect tax system is also regressive. Federal and State indirect taxes constitute 27.4 per cent of the income of the bottom 20% of households, but just 16.1 per cent of income of the top 20% of households. [12] Nearly half of WST falls on business inputs [13], adding about 1.2 per cent to the final cost of exports. [14] While the decline in Federal indirect taxes has been offset by a rise in State indirect taxes, this has resulted in even greater reliance on regressive taxes collected from an even narrower range of goods and services. The real advantage of indirect tax reform comes from the fact that GST revenues are projected to grow over time at a faster rate than existing indirect taxes, with revenue flowing to the States projected to grow by $1.25 billion by 2004-5. [15]

3. DIRECT TAX BASE:

Little evidence was presented to the Committee on direct tax reform. ACOSS tabled a report highlighting the problems with growing tax avoidance and minimisation, while the ACTU strongly argued that ordinary workers needed relief from bracket creep. The Government in ANTS argues that there is a need to reduce the growing burden on income tax collections due to the reduction in indirect tax collections.

The second stage of the Inquiry is tasked with looking at the income tax system in more detail. Suffice for our purposes to note two points in this report.

First, there is strong evidence that the income tax base is being eroded due to tax loopholes, avoidance and minimisation.

Second, there is a need to reduce the tax burden on low and middle income earners due to bracket creep.

The following table highlights the exponential growth in some of the more obvious tax loopholes, concessions and minimisation schemes:

GROWTH IN TAX SHELTERED INCOME

# this tax break rose to $740m in 1996-7

*growth to 1996-7

(Source: ACOSS “But it is broke” Attachment to Analysis of Labor' Tax Package September 3 1998, Taxation Statistics, Tax Expenditures)

Closing loopholes, reducing concessions, and cracking down on avoidance and minimisation would free up large amounts of revenue to provide much needed tax relief for middle and low income earners. Having said that, the Democrats would not support using revenue from indirect tax increases to pay for income tax cuts, as occurred in the 1993 Budget and as proposed in the ANTS package. A tax mix change of this sort only adds to regressivity, as income taxes are much more progressive than indirect taxes. Further, such a switch may have a negative economic effect, as Peter Dixon pointed out:

“If you make a balanced increase in consumption taxes by increasing consumption taxes and reducing income taxes – so that is the move you make in a balanced way – that is slightly job destroying.” [16]

Part of the compensation for bracket creep could be paid from the surplus. But this should only occur if social expenditure commitments are fully met. The damage done to health, education and community services by the cuts in the 1996 Budget should be addressed by funding from the surplus in equal priority to funding of tax cuts. The spending cuts in the 1996 Budget hurt low income households hardest, costing low income households $13 a week loss of disposable income, four times the percentage loss of high income households. [17]

The proposed income tax scales will reduce the progressiveness of the tax system, with the largest percentage increase reserved for taxpayers on $75,000 a year or more. This highlights the unfairness of the proposed tax scales, which deliver more than half of the benefit of $13 billion of tax cuts to the top 20 per cent of taxpayers, and an increase in disposable income three times that proposed for low income earners in percentage terms, and seven times in actual dollar terms.

INCREASE IN DISPOSABLE INCOME DUE TO TAX CUTS

(IGNORING GST EFFECT) [18]

Income p.a. Tax Cut per week % Change

$15,000 $6.80 2.7%

$25,000 $12.31 3.2%

$35,000 $19.98 3.9%

$45,000 $39.74 6.4%

$55,000 $58.92 8.2%

$65,000 $72.34 8.8%

$75,000 $85.77 9.3%

4. MACROECONOMIC EFFECTS:

The ANTS document makes the following claims:

“The tax reform package will deliver substantial long-term improvements in the operation of the economy, to the benefit of all Australians.

These improvements will be reflected in higher economic growth as a result of stronger, more productive, investment which, together with lowering industry costs, will yield better export outcomes. Such changes will be crucial in relaxing the balance of payments constraints that have for so long held back Australia's growth performance. The combination of higher growth and improved work incentives will deliver more jobs and lower unemployment.” [19]

In evidence, it became extremely clear that Treasury had not performed any modelling to back up these statements. In its written response to Senator Murray, Treasury stated:

“These are qualitative statements regarding the benefits of the Government's tax package.” [20]

In evidence, Mr Smith of Treasury conceded:

“We have not modelled specific point estimate impacts for employment or any of the other macro-economic variables, apart from the ones you know like the CPI.” [21]

1. The Modelling results:
2. The Committee has heard evidence from four prominent modellers on macro-economic variables – Peter Dixon from Monash University, Chris Murphy from Econtech, David Johnson from the Melbourne Institute and Geoff Carmody from Access Economics. Their basic findings are summarised as follows:
3. LONG TERM MACRO-ECONOMIC FINDINGS (%):
4. Variable

1. Dixon has also conducted short-term modelling, which finds that the fiscal stimulus will generate around 30,000 jobs in the short-run, an impact eventually bartered away as real wage increases. Murphy largely agreed with this analysis. Access economics has forecast that the fiscal stimulus will increase the inflation effect from the package to more than 3 per cent, but expects it also to lead to more employment and GDP growth in the short-term.
2. All of these models have their own assumptions and limitations. The Committee in its evidence focused on the Dixon and Murphy modelling, as two well respected and used long-term modellers. The modellers, within their assumptions and errors of margins, produced a similar result – that the macro-economic effect of the package is likely to be small but positive.
3. In this analysis, it is proposed to deal with only two of the macro-economic measures – economic welfare and employment, as these are the key indicators of whether Australians will be better off as a result of the tax. The Democrats note that the Government report contains further detail on other macro-economic variables, and do not propose to repeat the information provided in that report.
4. This analysis also looks at four downside risks to the tax package – wages (under employment), the pass through of indirect tax cuts, and the impacts on tourism and housing.
5. Welfare effects:
6. These differences were discussed in evidence at the hearing of February 15. The key differences between the two modellers focussed on the different calculation of the welfare gain/loss. Both modellers agreed that, in terms of deciding whether tax reform was in the national interest, the effect on consumer welfare (defined as the increase in spending power) was the key determinant. The discussion between the modellers centred on a small range – 0.01% to 0.2%. This was affected by:
7. i) tourism elasticity. Dixon's high elasticity was rejected by Murphy, Carmody and Treasury. Using a lower elasticity, Dixon revised his welfare estimate up to 0.22%.
8. ii) consumption substitution: Murphy's model picks up welfare gains from substitution between 313 consumption variables, Dixon's only uses 107 commodities, a point acknowledged by Dixon.
9. iii) capital investment choices: In Murphy's report, 10% of the welfare effect came from businesses using different investments (e.g. substituting equipment purchases for cars and vice versa). Dixon raised doubts about the significance of this substitution.
10. iv) export price effects: In Dixon's report, the export price effects of the Monash model had been changed from those that had applied over 15 years. This had the effect of reducing welfare by around $680m in Murphy's model, and likely to be a similar amount in Dixon's model. Murphy strongly disputed Dixon's figures, questioning the extent to which world export prices are affected by modest increases in Australian output.
11. v) definition of welfare: Murphy highlights that the formula used to determine welfare gains in the Dixon report is an approximation which always underestimates the result. There is a related over-estimate which should also have been reported to signify the range. In the Murphy model, this range is around plus or minus $500 million on the exact welfare figure cited by Murphy. Dixon acknowledged the error, but said the range estimates in his model would have been only $200 million (i.e. from a welfare gain of -$30m to plus $200m).
12. Employment Effects:
13. The Committee has not received much evidence on long-term employment effects of the package, as the three main private sector models, and Treasury in its PRISMOD modelling, all assume zero employment growth in the longer term. Dixon argued that a GST was “slightly job destroying” and that the short-term jobs gain came from the fiscal stimulus. Murphy argued that the jobs gains from the package were likely to flow predominantly from the elimination of poverty traps. Carmody suggested that further jobs, albeit a small effect, might flow from the efficiency effects. Dixon also made the point that employment would be reduced if workers failed to accept the tax cuts as adequate compensation for the imposition of the GST, with a jobs loss of up to 100,000 workers in the short term:
14.  
15. “If the tax package is to be implemented smoothly, it is vital that Australian workers allow their before-tax wages to decline relative to the CPI.” [26]
16. Other commentators were not as worried about the wages outlook. Murphy described the wage adjustment assumption as “implausible”, particularly in a more deregulated labour market. He concluded that the chances of ANTS producing a wages blow-out was remote, looking at international experience and the size of the tax cuts [27].
17. However, the ACTU had a clear message for the Committee that the compensation offered was inadequate and that the inflation effect would need to be factored into wage demands:
18. “Our primary responsibility is to protect real wage levels and living standards of working people. To the extent that I have already outlined that this (compensation) is inadequate, we would seek to address that inadequacy either through arbitral proceedings for those who have no capacity to bargain or though the bargaining mechanism.” [28]
19. However, the extent to which the unions, representing just 28 per cent of workers, can successfully push such an agenda is limited. Further, the Australian Industrial Relations Commission, in terms of arbitrating National Wages Cases or award rates, is required under section 88B(2)(b) to have regard for:
20. “economic factors, including levels of productivity and inflation, and the desirability of attaining a high level of employment.”
21. Indeed, in recent years, with the growth of enterprise bargaining, wage claims have tended not to follow consumer price index movements, but rather to be more closely related to productivity:
22. WAGE AND INFLATION MOVEMENTS
23. Year (to Sep)

1. (@ Department of Employment, Workplace Relations and Small Business “Wage Trends in Enterprise Bargaining” September 1998;
2. *Growth to August quarter in Full Time Average Weekly Ordinary Time Earnings, ABS 6302.0)
3. In short, in recent years, there has been relatively little relationship between a low CPI and annual wage movements. As Ms George herself conceded:
4. “From the point of view of our constituency, the protection of living standards has been largely fought out and bargained over in the marketplace in terms of the productivity growth that has been sustained or by the kind of argument we have put forward at the level of the commission for decent living standards.” [29]
5. It is noted that both Dixon and Murphy agree that the tax package will result in an increase in productivity, measured in part by a rise in real wages in both of their models.
6. Nevertheless, if there develops a strong belief among workers that the package does in fact undercompensate for them and that inflation will rise, this may affect wage expectations. The need to reduce short-term wage inflationary expectations must be a paramount consideration in framing the tax framework. In this respect, the Treasury finding that exempting food should reduce the second year CPI effect of the tax package from 1.9 per cent to 0.5 per cent is very significant. Using a narrower definition of food, Dixon found that exempting food would reduce by a third the employment losses from before tax bargaining.
7. The Democrats conclude that the benefits of the package could be lost if wages move with the CPI movement. While wages are moving more with productivity rather than CPI, workers' perceptions about the compensation package could inflate wage demands. To that extent, the substantial reduction in the CPI if food is excluded would help reduce potential wage pressure.
   1. Pass Through Issue:
   2. The Treasury price effects assume that all benefits of indirect tax cuts are passed through to consumers from day one. This has the effect of reducing the inflation effect from the package. Colin Hargreaves, in his evidence, argued that this assumption was unlikely to be met. Treasury responded by pointing to research in Canada and New Zealand showing that the vast bulk of tax cuts were passed through in the first two quarters. [30] Professor Warren cited a study by Bruce Freeland [31], by the OECD and the Brookings Institute to conclude that:
8. “As a rule, the Treasury is not too wrong in saying that, within 12 months, most of those effects have passed through. That is the general sentiment, but again it is a function of the environment from which it comes.” [32]
9. Professor Dixon argued that the effect would be negative, but temporary, in the short-term:
10. “Business could not withhold the tax cuts indefinitely. Competitive forces would eventually bring the tax cuts to bear on prices. So, in the long run, you would eventually wind up roughly where you would have been even if the tax cuts had been passed though immediately. There is no real long-run problem; it is a short-run problem if the tax cuts are not passed through.” [33]
11. Geoff Carmody from Access Economics pointed out that the expected CPI increase from this package of 1.9-2.5 per cent was a lot lower than that for Fightback (4.4%) or the 1985 RST (6.5%) because:
12. “It is…an environment which is much less conducive to not passing price reductions on.” [34]
13. Johnson also suggested that the pass-though of price effects into investment outcomes could take up to 8-10 years. [35]
14. The Committee also heard evidence from the ACCC on the procedures that they are putting in place to police pricing practices. It is difficult to see how the ACCC can hope to police 1.5 million GST tax collectors with 40 additional inspectors. However, education should encourage deterrence, and, combined with competitive forces in the market place, tax cuts should eventually be passed through.
15. The Democrats conclude that indirect tax cuts will eventually be passed through to consumers, in most cases within the first year, with major taxes like Wholesale Sales Tax. However the impact of other taxes may take longer to flow through, particularly for investment goods. This suggests that the short-term inflation outlook may be worse than the Government has indicated and highlights the need for compensation to be over-generous. The efforts of the ACCC to police prices and to educate retailers may need to be enhanced.
    1. Tourism:

Tourism is without question the most adversely affected export industry by the tax reform package. This is particularly concerning given that tourism is the most labour intensive of export industries. Dixon says that tourism exports would fall by 6.6 – 9% depending on the elasticity of demand. Murphy, using the lower elasticity figure, predicts a 6.9% fall in tourism exports. Modelling for the Tourism Taskforce described even Dixon's outer forecast as “conservative”, and predicted a fall of 11.75% in inbound tourism numbers, which they forecast would result in the loss of 24,500 jobs. [36] Domestic tourism was forecast to fall about 4.4%, contributing to a further 23,000 jobs losses. However, Murphy rejected these job loss forecasts as “highly exaggerated” instead predicting an employment loss of around 7,000 jobs in Accommodation, Cafes and Restaurants. [37] The Tourism Council, while arguing that the GST would increase the price of inbound tourism packages by 3.8-5.5% and domestic tourism by 5%, was reluctant to agree to the high price elasticities driving the Tourism Taskforce and Dixon modelling outcomes on overall effects. [38]

The Tourism Council, the Tourism Taskforce and the Inbound Tourism Association all called for inbound tourism packages to be GST-free, arguing that tourism exports should be treated as other exports and zero-rated. The Committee requested that this issue be modelled by Dixon and Murphy. Their findings are summarised:

ZERO-RATING INBOUND TOURISM EFFECTS

These tables assume a high elasticity in Dixon (-3) and a lower one in Murphy (-2). Geoff Carmody of Access Economics, who appeared with the Tourism Council, had grave doubts about both of these elasticities. He argued that while high elasticity figures were apparent for the inbound tourism packages, other parts of inbound tourism (e.g. business travel and personal travel) were not as price sensitive. [39]

In summary, the evidence to the Committee has been a little ambiguous on the response of the economy to the zero-rating of tourism. What is not disputed however, is that inbound tourism packages are highly price sensitive, the GST will significantly add to prices and that tourism exports, particularly of packages, must fall to some degree if they are taxed.

g. Housing:

The residential housing sector, which employs around 300,000 people, will also be adversely affected by a GST. Murphy has forecast a 2.5% decline in housing investment, close to the Monash estimate of 2.5%. Murphy has suggested that the price of house construction will rise around 6%, the price of a house and land package by about 5%, and the price of residential rents by 3-4%. Employment in housing is forecast to fall by 8,000 jobs, partly offset by a 7,000 job gain in commercial construction. The 6% rise in house prices compares with the 4.7% increase forecast by Treasury. The Housing Industry Association predicted that housing prices (including land) would rise by 8 per cent in the short-term [40]

It is inevitable that there will be an adverse effect on housing, which will be greatly accentuated by a likely boom to get in before the GST followed by a bust, according to the Master Builders Association submission. They recommended a three year phase in of the GST, a proposal which would be very expensive to revenue and difficult to police.

The Housing Industry Association instead recommended a re-arranging of the First Home Buyers Scheme into a rebate scheme. This is because, they argued, around $600m of the $830m earmarked for the current scheme would be used to support the purchase of GST-free houses. [41] By converting the scheme into a partial GST/stamp duty rebate scheme, at a cost of an additional $90 million, much of the negative effect on housing is reversed. Modelling of the plan by Econtech found:

HIA COMPENSATION PROPOSAL EFFECT

While the details of the compensation package proposed by the HIA raise some difficulties, the Econtech modelling suggests that restructuring the First Home Buyers Scheme could, with little extra expenditure, deliver significant benefits. The Democrats believe that this deserves further detailed consideration by the Community Affairs Committee in its second stage report, and draws attention to the HIA and MBA submissions.

5. DISTRIBUTIONAL ANALYSIS:

A key issue of dispute before the committee has been the adequacy of the Government's compensation package. The Government has rested its compensation arguments on the “cameos” published in ANTS. These cameos assume that household spending equals disposable income, and that all households have the same expenditure pattern. The GST impact is then calculated by multiplying the CPI effect for year 2 by disposable income.

By contrast, the Treasury in 1985 for the Draft White Paper and its analysis of Fightback in 1993 calculated price effects based on differing expenditure patterns for households derived from the Household Expenditure Survey. However, in ANTS, the Treasury has rejected this approach, arguing that the HES contains known biases in reporting of tobacco, alcohol and gambling, is affected by “lumpy” large household expenditures and that there are problems with HES savings data. [42] These arguments were expanded in detail in a paper by Carnahan [43] tabled by the Treasury and by Treasury officers in Committee.

However, Treasury's approach was rejected by six university professors in evidence to the Committee and by the Australian Bureau of Statistics.

Professor Ann Harding said modelling using adjusted HES data would take into account different spending patterns of households, and provide important information in making decisions about the distributional effect of the package. She said that the two critical assumptions by Treasury about spending and saving patterns “…give you the most favourable outcomes for the distributional impact of the package.” [44]:

Professor John Quiggin also agreed that the Treasury approach provided the best possible scenario on compensation, and that rejection of the CPI-based model of distributional analysis was

“….unanimously held within the economic profession. No-one outside the Treasury has believed that the CPI calculations produced in the package were defensible.” [45]

Associate Professor Neil Warren said that the CPI relied on a narrower band of households – wage earners in capital cities – than did the HES. This population sample, about 30 per cent in the 12th series CPI used in the PRISMOD estimates, excludes from its analysis pensioners and social security beneficiaries, the self-employed and non-metropolitan Australia. While acknowledging the problems with the HES, he said that there was no better alternative to working with the data on distributional analysis [46].

Professor Peter Saunders, of the Social Policy Research Centre at the University of New South Wales, while acknowledging the limitations of the HES also rejected the Treasury assumption of common expenditure patterns. His analysis, using a Budget Standards approach of likely expenditures by households showed that price changes by the GST and accompanying tax changes have different effects according to the composition of the household at a given standard of living. [47]

Professor Peter Macdonald, Head of the Australian University Research School of Social Sciences also criticised use of the CPI for distributional analysis, which he said was particularly unfair on families with children, as PRISMOD does not take into account household composition and the different spending patterns associated with it. The ABS HES data, he argues, is a better source. [48]

Associate Professor David Johnson, of the Melbourne Institute, said he believed that the Treasury costing underestimated the effects of price changes on households because the overall price effect was under-estimated, that it assumed the same spending patterns for all households and that there would be no saving or dissaving. He concluded that the effect was to considerably underestimate the cost effect on low income households and over-estimate the effect on high income households. [49] :

His modelling, using three different assumption sets, shows the extent to which the Treasury assumptions under-state the effect on low income earners and over-state that on high income earners. In the attached table, Assumption A is the ANTS assumption of a standard CPI effect of 1.9% for each quintile of income, Assumption B, is a CPI effect of 2.44% (taking into account tobacco and housing) and expenditure patterns, and Assumption C is the same as B but with savings and dissaving appropriate to the household type:

NET CHANGE IN WEEKLY PURCHASING POWER FROM GST

ALL HOUSEHOLDS AVERAGE [50]

What the Melbourne Institute model shows is that adjusting for the full CPI effect, and expenditure and savings patterns, the impact of a GST on low income earners will be more than twice that claimed by the Government for low income earners, but 22 per cent less for high income earners. Geoff Carmody also supported the use of the HES for distributional analysis. [51]

In short, Treasury received no support whatsoever for its decision to use the CPI rather than the HES for its distributional analysis. Even within the official family, there was dissension. The Australian Bureau of Statistics expressed concern that Treasury's criticisms of the HES were overstated, with the Government Statistician, Mr W McLennan, advising the Committee that:

“The ABS consider statistical data from the Households Expenditure Survey can be used, at a suitably disaggregated level, after appropriate adjustments are made to reported expenditure information, to describe the expenditure patterns of population sub-groups. ABS itself uses the information in this way in its own `fiscal incidence studies;, in the calculation of the weights in the CPI, and in the compilation of the National Accounts. Also, HES data has been used over the years by many analysts for distributional analysis by population subgroups.” [52]

The ABS submission argued that provided the sample contained a sufficient number of records for the population groups to minimise sampling errors, the HES data on household income and expenditure was suitable for distributional analysis by household type, income decile and specific geographic regions, with appropriate adjustments. The ABS noted that the HES itself forms part of the basis for calculation of the CPI, a point acknowledged by Treasury.

Treasury has also drawn the Committee's attention to two other papers by Treasury officials on the use of HES. One, cited in the Carnahan paper, is the 1992 Wright and Dolan paper. However, this paper, while rejecting the use of the raw data from HES, concludes that the HES can be used in a modified form to correct for known errors [53]. All modellers agreed with this approach. The second paper is from Phil Gallagher from the Retirement Income Modelling Taskforce within Treasury. Mr Gallagher, too, defended the use of HES in distributional modelling [54]. In short, even within Treasury, the CPI approach has not unanimous support.

ACOSS called on the Senate Committee to conduct its own modelling using HES data. Mr Raper argued that the modelling needed to pass two tests: first, it must over-estimate the impact to ensure that Government assurances can be trusted. Second, it must take account of different households in terms of composition, income, expenditure and savings. [55]

In summary, the evidence to the Committee from all experts outside of Treasury was that the use of CPI for distributional analysis rather than the HES was not appropriate. Indeed, the use of the CPI by the Treasury picks up the most favourable assumptions for distributional analysis and underestimates the impact of the tax plan on low income earners. It is also likely to underestimate the impact on specific household types private sector renters, on the disabled, on pensioners, on families and on people in rural and remote localities.

The Senate Committee, in responding to these concerns, has commissioned Professor Harding and Professor Warren , using the STINMOD/STATAX model, to re-model the impact of the package on all cameos, using the first year full CPI effect of the package of 3.1 per cent, and taking into account the HES-identified differing expenditure patterns. When that report is received, on April 1, the Committee will be in a better position to determine whether indeed, any groups are left worse off by the package.

6. FOOD:

Exemption of food was a topic of considerable evidence before the Committee. The Treasury advised the Committee that an exemption for food (defined to exclude restaurants and takeaway food) would cost:

GST-FREE FOOD – REVENUE IMPLICATIONS:

2000-01 2001-02 2002-03 2003-04

-$4465m -$5230m -$5520m -$5820m

In 2000-01, the zero-rating of food would reduce GST revenue by about 17 per cent. This would need to be made up by reducing the $13 billion of income tax cuts, or closing tax loopholes or by raising the GST rate. The advantage of reducing the income tax cuts (52 per cent of which will flow to the top 20% of taxpayers) is that it has the added advantage of reducing the tax-mix change. Under ANTS, the GST raises $8 billion in 2000-01 (falling to $6.1 billion in 2002-3) more in indirect taxes than the indirect taxes it replaces. This is then used in part to fund income tax cuts.

This element of the package was criticised by a number of commentators. Geoff Carmody of Access Economics described this tax mix change as “an unfortunate element” of the package, and called for it to be eliminated either by abolishing more indirect taxes or by reducing the rate. The more the tax mix was reduced, the lower the net price effect will be. [56] Dixon also criticised the tax mix change, arguing that it will reduce employment. Reducing consumption taxes and increasing income taxes would result in a short-run increase in employment. [57] The Business Coalition for Tax Reform, the Australian Food and Grocery Council and ACOSS all said that an optimal tax reform would not have a tax mix change in it. ACOSS has argued that this should be achieved by exempting food.

While the Democrats acknowledge the need to raise an additional $2.8 billion by 2002 to restore indirect taxes to their 1995-6 level, the tax mix change in ANTS goes much further than that, with an increase in indirect tax collections of more than twice that amount to $6.5 billion.

The tax mix change is likely to increase the regressivity of Federal taxes, and is not justified on equity grounds.

1. Efficiency
2. In tax economics theory, the three key criteria for tax reform are listed as efficiency, simplicity and equity. In assessing the exemption of food from the GST, it is appropriate to consider the change under these taxation principles. According to Chris Murphy, the macroeconomic models (e.g. MM303 and Monash) can tell a great deal about efficiency effects, but very little about simplicity or equity effects. PRISMOD and STINMOD, however, can be used to determine equity effects. Simplicity effects are almost impossible to model, and should not rate as high as efficiency and equity in the prioritisation of competing principles in the making of public policy.
3. The Committee has commissioned modelling by Dixon on the economic effects (i.e. efficiency effects) of modelling food, which was also re-modelled by Murphy. Their findings are summarised below:
4. EFFICIENCY EFFECTS OF GST-FREE FOOD
   1. Peter Dixon Chris Murphy
   2. Monash University Econtech
   3. Variable

1. (** For his modelling, Peter Dixon used the Treasury generated CPI estimates. Treasury found that CPI would fall from 1.9% to 0.5% with food exempted)
2. Dixon's Report also found that exempting food has an even more positive short term (i.e. over the first five years) effect, with 8-12,000 extra jobs, and higher GDP growth, exports and economic welfare than if food was taxed. Murphy's report found that the Agriculture industry's long term growth rate would double if food was exempted, notwithstanding the opposition of the National Farmers Federation's national leadership to the food exemption.
3. Both modellers concluded that the exemption of food would have a negligible effect on economic outcomes, although both made the point that their models did not take into account compliance costs, which are dealt with under the simplicity principle below.
4. The Democrats conclude that on efficiency grounds, there is no case for taxing food, and that not taxing food will not affect the marginal positive economic benefit flowing from tax reform. Further, not taxing food holds out the prospect of short-term employment gains and of considerably dampening wage pressures, as well as doubling the long-term growth rate of the agriculture industry.
5. Simplicity
6. All business groups appearing before the Inquiry and most economists opposed the exemption of food because it would add to business compliance costs, although the Australian Food and Grocery Council qualified its opposition by stating that if the Senate decided to exempt food, then the definition should be as wide as possible. [58]
7. Economists appearing before the committee, including Chris Murphy, Neil Warren and Geoff Carmody also opposed the exempting of food, citing compliance costs as a key reason.
8. Professor Warren provided the Committee with a paper on food which he presented at an ATAX conference on December 18, 1998, which provided the largest amount of information on the compliance issues associated with exempting food. [59] The paper outlines many of the difficult definitional problems associated with defining food in Canada and the United Kingdom. He cited a report on compliance costs with the UK VAT finding that the total compliance costs for the UK VAT (which exempts food) is around 4.3% of revenue for businesses and 2.1% for taxation authorities. After assessing the various benefits that businesses receive from the VAT in terms of managerial information and cashflow effects, he concluded that:
9. “It would therefore appear from the United Kingdom evidence that the burden on compliance of a GST, in aggregate is not unduly serious”. [60]
10. The report also noted that there were low estimates for compliance costs from Canada (which exempts food), reflecting the evolution of that tax. Indeed, Warren concluded:
11. “Just what can be learnt from these international comparisons about the design of a VAT is not clear….However, what is probably reasonable to conclude is that making food GST-free can only worsen the compliance cost for businesses, especially small businesses.” [61]
12. The UK findings can be contrasted with Australian research showing that the taxpayer compliance cost of Wholesale Sales Tax was about 4.7% of revenue in 1994-5. [62] In short, the evidence on compliance costs has probably been overstated by business, and that compliance costs with a food-free GST are unlikely to differ markedly from those of the current tax system.
13. It should also be noted that of the 27 OECD countries with a VAT/GST, 23 of them either exempt food or tax it at concesional rates. If 23 out of 27 OECD countries, some 700 million people, can manage to develop a tax system that achieves fair taxation of food, why is Australia so markedly different?
14. A further issue mitigating against a huge compliance cost for a food exemption is the growing computerisation of business accounts. As Chris Murphy notes in his report [63]:
15. “For businesses with computerised accounts, the effects of compliance costs of the changes to indirect tax should be minimal. For other businesses, ANTS can be expected to speed the shift towards computerised accounts that is already occurring in the absence of ANTS.”
16. The same can be said for the non-taxation of food, although it highlights the need for a simple and useable definition of food. This is an issue on which the Committee will need to conduct further work in its next stage.
17. Equity
18. What has not been disputed in evidence to the Committee has been that low income earners will be hardest hit by a GST, and that there will be a need for compensation. Even those economists who opposed the exemption of food (e.g. Murphy, Warren and Carmody) argued for adequate compensation through the social security and taxation systems.
19. The reason why low income earners will be most adversely affected by taxes on food is demonstrated in their expenditure patterns, with low income households allocating twice as much of their spending to food than high income households:
20. SHARES OF EXPENDITURE BY QUINTILE
21. RANKED BY TOTAL EXPENDITURE (%) [64]
22. Item

1. The importance of food is amply demonstrated in this table. It should come as no surprise to discover that taxing food is regressive, and strongly so. If food is taxed, low income earners will pay twice as much GST as a percentage of expenditure on food than high income earners.
2. Some witnesses have argued that, while in percentage terms a food exemption would benefit low income earners, in actual dollar terms high income earners would gain the greater benefit because of their greater level of aggregate spending. [65] However, tax policy, in looking at equity effects, has always worked in percentage terms rather than nominal terms.
3. Further, if high income earners gain from a food exemption, then that reduces considerably the need to deliver tax cuts to that group. The HES-based analysis if GST effects by the Melbourne Institute shows that this group is already over-compensated for the GST effect. Even the ANTS cameos show that a single person on $75,000 a year is already receiving an increase in disposable income six times the benefit of a person on $25,000 in percentage terms, and more than 13 times the benefit in dollar terms. [66] Further, the argument that high income earners obtain the greatest benefit may be overstated. Quiggin in his evidence to the Committee pointed out that high income households have a larger number of people per household, and that adjusting for that shows that food expenditure is fairly constant on a per capita basis [67]:
4. HES HOUSEHOLD EXPENDITURE ON FOOD:
5. Item

1. Quiggin concludes that the tax on food “…is effectively a poll tax, which is the most regressive of all taxes”.
2. The key question then is whether compensation or tax exemption is the best means of addressing equity. Quiggin argues that the cost of adequate compensation would be greater than the revenue collected from the tax on a non-targeted basis, and would create distortions sufficient to offset the beneficial effect of reductions in other taxes financed by the net revenues from the food tax if it was targeted. [68]
3. But, the real issue about compensation is the lack of faith in compensation being sustained.. As Warren, a supporter of food taxes puts it:
4. “The calls for food to be GST free have everything to do with the politics of tax reform. In particular, concern about the maintenance of the compensation package. If there is any justification for such an action, then politicians only have themselves to blame. It is therefore incumbent on politicians to overcome this lack of confidence in the guarantees and to implement an approach which is capable of restoring public confidence in their assurances.” [69]
5. Mitchell Hooke of the Australian Food and Grocery Council expressed a similar sentiment:
6. “We see the proposal to exempt food as symptomatic of a lack of confidence in the adequacy of fiscal compensation and their political durability for lower income and socially disadvantaged people.” [70]
7. The Anglican Church, in a statement issued this week calling for food to be exempted from the GST also expressed the view that:
8. “…Compensation packages are always likely to be fragile and are usually inadequate. Moreover, they tend to be eroded over time and some needy groups would receive no compensation at all.” [71]
9. The Australian Council of Social Services and Quiggin [72] also point to the New Zealand experience, where the compensation for the 1987 GST were wiped out in the across the board welfare cuts in the 1991 Budget.
10. The Democrats share the lack of faith in the guarantees provided by Governments. What one Government promises, another, or even the same Government can take away. The experience of the first Howard Government confirms this trend. In 1995, as Liberal Leader, John Howard told the ACOSS National Conference that:
11. “I want to make it unambiguously clear today that we will not be eroding the safety net that underpins our social security system. I repeat, so that even my Labor opponents can understand: We will not be eroding the safety net that underpins our social security system”. [73]
12. The Coalition then followed that statement up with detailed commitments in its Social Security, Health, Housing, Employment and Training, Aged Persons, Aboriginal and Torres Strait Islander and Higher Education Policies dealing with essential income support and safety net programs. However, not withstanding the “iron clad guarantees” that Mr Howard gave at the ACOSS conference and the specific policy commitments, the 1996 and 1997 Budget were riddled with broken promises that seriously undermined the safety net:
13. CUTS TO THE SOCIAL SAFETY NET 1996 AND 1997 BUDGETS

1. (NOTE: These are not all the Budget cuts, , but only those that directly breach commitments given in the Coalition's 1996 Election Policies. Source: Budget Papers)
2. Indeed, the Budget cuts constituting breaches of the 1995 safety net commitment are worth more in the 1999-00 year ($1.82 billion) than the social security compensation offered in the tax package in 2000-01 ($1.79 billion) [74].
3. It should be pointed out that, according to research commissioned by ACOSS from the Melbourne Institute, the 1996 Budget cuts hit low income households hardest. The cuts cost the average household the equivalent of $7 a week in lost disposable income, rising to $13 a week for households with an income of less than $400 a week, and rising to $38 a week where the chief breadwinner was unemployed. [75]
4. In summary, Governments simply cannot be trusted to hold to their guarantees on compensation. The best form of compensation is to reduce the need for it. As Treasury concedes that excluding food reduces the price effect of a GST from 1.9 per cent to 0.5 per cent in 2001/02 (and Murphy says it will make the CPI effect negative in the long run), the necessity for compensation is greatly reduced if food is excluded. That dramatically reduces the risk of harm to low income earners that would flow from watering down of the safety net.
   1. Food – Conclusion:

Reviewing the three principles of tax reform in respect of food, the Democrats conclude that:

1. on efficiency grounds, the exemption of food has no significant economic effect, other than a short-term boost to the economy from reducing the tax mix change and probably dampening down wage demands;
2. on simplicity grounds, the exemption of food will increase compliance costs, although overseas experience suggests that this will not be substantial and may be reduced with computerisation of business accounts and by supporting a simple, wide definition of food;
3. on equity grounds, there is a compelling case that the taxation of food is regressive, and that compensation is likely to be understated and subject to withdrawal at a later time.

In balancing the three principles, the substantial equity concerns must outweigh the more modest simplicity concerns, and food should be exempted. But, to minimise simplicity costs, a simple and workable definition of food needs to be found.

On political grounds, neither the Australian people (judging from opinion polls) nor the Australian welfare lobby trust this or any Government not to take away compensation later. By contrast, the more marked aversion of Governments to raising taxes makes it likely that if food becomes tax free, it will stay that way.

7. OTHER NECESSITIES OF LIFE:

1. Clothing:
2. Treasury was not asked to provide a costing on exempting clothing. However, based on Murphy's report that clothing constitutes around 7 per cent of the GST tax base [76], the cost of exempting clothing is likely to be $1.9 billion in 2000/01, rising to $2.3 billion by 2002/3. [77] Zero rating clothing would actually be regressive, because it constitutes only 3.4% of the spending of the poorest quintile and 5.2% of the richest quintile [78].
3. The Democrats conclude that there are no compelling reasons for zero-rating clothing.
4. Housing:
5. Housing has been dealt with earlier in the report, and the Democrats have noted concerns already with the taxation of housing. But, as research by the Melbourne Institute shows, tax treatment of housing, to be progressive, needs to focus on rental costs. Rents make up 19.3 per cent of the spending of low income households, but only 8.7 per cent of the spending of low income households. By contrast, renovations and additions to housing constitute 0.5 per cent of the spending of low income households and 13.56 per cent of the spending by high income households. To maximise the progressiveness of the tax system, rentals should be concessionally taxed and renovations fully taxed. This is how the ANTS is currently broadly structured.
6. However, as noted earlier, Murphy has forecast a 3-4 per cent rise in private sector rents as a result of the tax reform package. This could be alleviated considerably through restructuring of the housing rebate scheme, and the Democrats will turn more attention to this in the second stage of the Inquiry.
7. Books:

The Treasury was asked to cost the zero-rating of books. The following costing was prepared:

GST-FREE BOOKS – REVENUE IMPLICATIONS

2000-01 2001-02 2002-03 2003-04

-$180m -$200m -$220m -$230m

No evidence was entered during this stage on books, and this matter will be dealt with in the next stage of the inquiry.

8. CONCLUSIONS:

The Democrats make the following conclusions about this stage of the tax Inquiry:

1. The modelling before the Committee concludes that the long term macro-economic effects of the tax package are likely to be positive, albeit small, with increases in capital investment and traditional exports, a small but positive rise in economic welfare, a negligible effect on employment, and negative effects on industries such as tourism, housing and personal services.
2. The short-term effects of the package will be a positive boost to jobs driven by the fiscal stimulus. A wages blow-out would have a short-term negative effect, although the risk of this in a deregulated labour market should not be overstated.
3. While the macro-economic effects are slight, the need for tax reform is accentuated by the decline in the Federal indirect tax base due to changing consumption patterns.
4. The direct tax base is also under some threat due to the growth in tax minimisation. There is a clear need also for tax relief, particularly for low and middle income earners, to correct the effects of bracket creep, although the Government's tax cuts are poorly targetted.
5. The Government's modelling of the impact of the GST on households clearly understates the impact on low income earners. Treasury's arguments for failing to use the Household Expenditure survey rather than the CPI was unanimously condemned by all expert witnesses, including the ABS itself. The Government's claim that no-one will be made worse off will be tested by the Committee's next stage of modelling.
6. Food should be exempted from the GST because the equity case is compelling, the economic efficiency case of taxing food is neutral and the simplicity argument for taxing food is somewhat overstated and should always be sub-ordinate to the equity and efficiency principles.
7. The exemption of food would increase the short-term boost to jobs, reduce the risk of a wages blow-out by reducing the inflation effect of the package and thereby reduce the need for compensation.
8. Zero-rating clothing would be regressive, providing the greatest benefit to high income households.
9. Taxation of housing will need to be carefully considered in the next stage to minimise employment effects and the effect on vulnerable private sector renters.
10. Zero-rating inbound tourism packages would significantly reduce the negative effect of the GST on the price-sensitive inbound tourism packages sector of the market.

Senator Andrew Murray

Australian Democrats

Footnotes

[1] Treasury “Mid-year Economic and Fiscal Outlook” December 1998

[2] Access Economics “Budget Monitor” No.39 December 1998

[3] Treasury “A New Tax System” (ANTS) September 1998 pp.5-8

[4] Business Coalition for Tax Reform submission (no.104A) pp. 6-7

[5] Jennie George, evidence 4/2/99 p. 671

[6] Michael Raper evidence 4/2/99 p.706

[7] Ian Donges evidence 28/1/99 p.230

[8] Neil warren evidence 17/12/98 p.117

[9] John Quiggin evidence 3/2/99 p.599

[10] Evidence 4/2/1999 p.655

[11] Johnson D et al.Tax Reform Report No 2 “Indirect Taxes” Evaluation of Options for Reform” Melbourne Institute March 1998 p. 22

[12] NATSEM “Income Distribution report No. 8” April 1998 p.2

[13] Warren N. “Tax facts and Tax Reform” CPA Discussion Paper, June 1998 p.22

[14] Anderson P “Options for relieving the impact of indirect taxes on industry costs” Australian Tax Forum 10 (2) 1993 p.177-199

[15] ANTS p.25

[16] Peter Dixon evidence 3/2/99 p.530

[17] Johnson D and Hellwig O “Evaluation of the distributional impact of the 1996-7 budget on Australian households” Melbourne Institute-NIEIR, Perth January 1997

[18] ANTS p.178

[19] ANTS p.155

[20] Treasury response to Senator Murray 28/1/99 p.13

[21] Greg Smith evidence 28/1/99

[22] Dixon report January 1999, the range represents the effect of reducing the elasticity of tourism

[23] Econtech Submission no.57

[24] Melbourne Institute “Quarterly Bulletin of Economic Trends” 3.98 p.86

[25] Access Economics “Budget Monitor” No.39 December 1998

[26] Dixon report 25/1/99 p.ii

[27] Murphy report p.21

[28] Jennie George evidence 4/2/99 p. 681

[29] Evidence p.684

[30] Treasury responses to the Committee 28/1/98 p. 9

[31] evidence 17/12/1998 p. 106

[32] evidence p.121

[33] evidence 3/2/1999 p.555

[34] evidence 4/2/1999 p.653

[35] David Johnson evidence 17/12/98

[36] Tourism Taskforce submission no. 472 p. 22

[37] Murphy Report p.28

[38] Tourism Council submission p.13-6

[39] Geoff Carmody evidence 5/2/99

[40] Housing Industry Association submission tabled 5/2/99 p. 7.

[41] HIA Submission p.3-4

[42] ANTS pp.162-3

[43] Carnahan, Michael “Does Demand create poor quality supply: A critique of alternative distributional analyses” ATO Tax Policy Group 1998

[44] Ann Harding evidence p.222

[45] John Quiggin evidence p. 593

[46] Neil Warren evidence pp.91-3

[47] Peter Saunders submission 614 p. 5

[48] Peter Macdonald evidence p.307-8

[49] David Johnson evidence pp.114-5

[50] Johnson D et al. “The Government' Tax Reform Package – Equity and Efficiency Effects of Indirect Tax Reform” Melbourne Institute conference paper 1/9/98 p.10

[51] Geoff Carmody evidence 4/2/99 p.662

[52] ABS Submission No. 731 p.1

[53] Wright J and Dolan A “The Use of HES data in Distributional Analysis” Conference of Economists 8-10 July 1992, tendered as an attachment to the ACOSS submission no. 68

[54] Gallagher, Phil “Comments on Carnahan's Paper” notes for a seminar on 22/7/98 tabled in the House of Representatives on 11/11/98

[55] Michael Raper evidence p.695

[56] Geoff Carmody evidence pp. 653, 664

[57] Dixon report p.20

[58] Mitchell Hooke evidence 4/2/99 p.652

[59] Neil Warren “Food: Staple of life or Staple of the GST” ATAX 16/12/98

[60] ibid p.16

[61] ibid p. 18

[62] ATAX research cited in ATO publication “A Report into Taxpayer Costs of Compliance” AGPS November 1997 p. 75

[63] Murphy report p.3

[64] Melbourne Institute “Tax Reform:Equity and Efficiency” Report No 2 March 1998 p.9

[65] e.g. Food & Grocery Council evidence 4/2/99 p.652, Warren ibid p.7

[66] ANTS p.178

[67] Professor John Quiggin submission No. 78 p.8

[68] ibid p.17

[69] Warren ibid p.19

[70] David Hooke evidecen 4/2/99 p.653

[71] Statement by Archbishop Peter Hollingworth 16/2/99

[72] Quiggin ibid p.10

[73] John Howard address to the ACOSS National Conference 13/10/95 p.6

[74] ANTS p.67

[75] Johnson D and Hellwig O “Evaluation of the distributional impact of the 1996-7 budget on Australian households” Melbourne Institute-NIEIR Perth January 1997

[76] Murphy report p.6-7

[77] ANTS p.100

[78] Melbourne Institute ibid