HIGH COURT OF AUSTRALIA
FRENCH CJ,
HAYNE, CRENNAN, KIEFEL, BELL AND KEANE JJFORTESCUE METALS GROUP LIMITED & ORS PLAINTIFFS
AND
THE COMMONWEALTH OF AUSTRALIA DEFENDANT
Fortescue Metals Group Limited v The Commonwealth
[2013] HCA 34
7 August 2013
S163/2012
ORDER
The questions reserved for the consideration of the Full Court on 5 November 2012 be answered as follows:
Question 1
Are any or all of s 3 of the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), s 3 of the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth) and s 3 of the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) invalid in their application to the plaintiffs on one or more of the following grounds:
A.they discriminate between the States of the Commonwealth of Australia contrary to s 51(ii) of the Constitution;
B.they give preference to one State of the Commonwealth of Australia over another State contrary to s 99 of the Constitution;
C.they so discriminate against the States of the Commonwealth or so place a particular disability or burden upon the operations or activities of the States, as to be beyond the legislative power of the Commonwealth?
Answer
No.
Question 2
Are any or all of the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) and the Minerals Resource Rent Tax Act 2012 (Cth) invalid in their application to the plaintiffs on the ground that they are contrary to s 91 of the Constitution?
Answer
No.
Question 3
Who should pay the costs of the reserved questions?
Answer
The plaintiffs.
Representation
D F Jackson QC with B Dharmananda SC and W A D Edwards for the plaintiffs (instructed by Corrs Chambers Westgarth Lawyers)
J T Gleeson SC, Solicitor-General of the Commonwealth and N J Williams SC with G J D del Villar and D F C Thomas for the defendant (instructed by Australian Government Solicitor)
Interveners
W Sofronoff QC, Solicitor-General of the State of Queensland with A D Scott for the Attorney-General of the State of Queensland, intervening (instructed by Crown Law (Qld))
G R Donaldson SC, Solicitor-General for the State of Western Australia with A J Sefton and J D Berson for the Attorney-General for the State of Western Australia, intervening (instructed by State Solicitor (WA))
Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.
CATCHWORDS
Fortescue Metals Group Limited v The Commonwealth
Constitutional law – Powers of Commonwealth Parliament – Constitution, s 51(ii) – "[T]axation; but so as not to discriminate between States or parts of States" – Minerals Resource Rent Tax Act 2012 (Cth), Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth), Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) ("Acts") established and imposed minerals resource rent tax ("MRRT") – Amounts paid as State royalties allowed under Acts as royalty credits – Available royalty credits which do not exceed mining profit deductible from MRRT as royalty allowance – Effect of Acts alleged to be that liability to pay MRRT varies between States and that reduction in State royalty increases liability to pay MRRT by the amount of the reduction – Whether Acts discriminate between States contrary to s 51(ii) of Constitution.
Constitutional law – Constitution, s 99 – Prohibition on Commonwealth, by any law of revenue, giving preference to one State over another – Whether Acts give preference to one State over another.
Constitutional law – Melbourne Corporation doctrine – Whether Acts discriminate against or place particular burden upon operations or activities of States, beyond legislative power of Commonwealth Parliament.
Constitutional law – Constitution, s 91 – "Nothing in this Constitution prohibits a State from granting any aid to or bounty on mining for gold, silver, or other metals" – Whether Acts contravene s 91.
Words and phrases – "discrimination", "Melbourne Corporation doctrine", "minerals resource rent tax", "preference in trade, commerce or revenue", "State royalties", "States or parts of States".
Constitution, ss 51(ii), 91, 99.
FRENCH CJ.
Introduction
Fortescue Metals Group Ltd and four subsidiaries of that company commenced proceedings against the Commonwealth by way of writ issued out of this Court on 22 June 2012. They assert that provisions of the Minerals Resource Rent Tax Act 2012 (Cth) ("the MRRT Act") and three related Acts imposing Minerals Resource Rent Tax ("MRRT") in relation to iron ore are not valid laws of the Commonwealth. The three related Acts are the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth) and the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth) (together referred to as "the Imposition Acts").
The stated object of the MRRT Act is to ensure that the Australian community receives an adequate return for its "taxable resources" having regard to their inherent value, their non-renewable nature and the extent to which they are subject to Commonwealth, State and Territory royalties[1]. The Act makes allowance, in fixing the MRRT liability of a miner, for mining royalties payable under State laws. Because the MRRT Act makes those allowances, the liabilities it imposes can vary according to State mineral royalty regimes. That potential for a differential operation from State to State underpins the plaintiffs' argument that the Act discriminates between States contrary to s 51(ii) of the Constitution and gives preference to one State over another contrary to s 99 of the Constitution. The plaintiffs also assert that, contrary to s 91 of the Constitution, the MRRT Act detracts from, impairs or curtails the grant by States of aid to mining for iron ore by the reduction of royalty rates applicable to the mining of iron ore. The Act is also said to detract from, impair or curtail the capacity of the States to function as governments contrary to the principles enunciated in Melbourne Corporation v The Commonwealth[2]. The Imposition Acts are challenged, along with the MRRT Act, because s 3 of each of them imposes the MRRT. A reference to the MRRT Act in these reasons is a reference to that Act read with the Imposition Acts.
[1]MRRT Act, s 1–10.
[2](1947) 74 CLR 31; [1947] HCA 26.
The limitations on Commonwealth legislative power imposed by ss 51(ii) and 99 of the Constitution protect the formal equality in the Federation of the States inter se and their people, and the economic union which came into existence upon the creation of the Commonwealth[3]. They must be read in their context with s 51(iii), which limits legislative power with respect to bounties on the production or export of goods by requiring that they shall be "uniform throughout the Commonwealth", and s 88, which requires that customs duties be "uniform". The scheme of economic unity which they support is reinforced by s 102, which empowers the Parliament, by laws with respect to trade or commerce, to forbid as to railways any preference or discrimination by any State. The relationship between those provisions, the exclusivity provided by s 90 for Commonwealth legislative power with respect to customs, excise and bounties, and the guarantee of freedom of trade, commerce and intercourse among the States made by s 92, was encapsulated in the joint majority judgment in Capital Duplicators Pty Ltd v Australian Capital Territory [No 2][4]:
"ss 90 and 92, taken together with the safeguards against Commonwealth discrimination in s 51(ii) and (iii) and s 88, created a Commonwealth economic union, not an association of States each with its own separate economy." (footnote omitted)
Importantly, the proscription of differential taxes avoided distortion of "local markets within the Commonwealth."[5]
[3]A formal equality which belied persistent economic and geographical inequalities: Anderson, "The States and Relations with the Commonwealth", in Else-Mitchell (ed), Essays on the Australian Constitution, (1961) 93 at 108.
[4](1993) 178 CLR 561 at 585; [1993] HCA 67. See also Clark King & Co Pty Ltd v Australian Wheat Board (1978) 140 CLR 120 at 153 per Barwick CJ; [1978] HCA 34; Hematite Petroleum Pty Ltd v Victoria (1983) 151 CLR 599 at 660 per Deane J; [1983] HCA 23; Australian Coarse Grains Pool Pty Ltd v Barley Marketing Board (1985) 157 CLR 605 at 647–648 per Brennan J; [1985] HCA 38; Philip Morris Ltd v Commissioner of Business Franchises (Vict) (1989) 167 CLR 399 at 426 per Mason CJ and Deane J; [1989] HCA 38.
[5](1993) 178 CLR 561 at 585.
At a more detailed level, the interactions between ss 51(ii), 51(iii), 88 and 99 are as summarised by Latham CJ in Elliott v The Commonwealth[6]:
"The sections mentioned operate independently, but they overlap to some extent. Laws of taxation, including laws with respect to customs duties, fall under sec 51(ii) and as laws of revenue they fall under sec 99. Laws with respect to bounties on the export of goods fall under sec 51(iii) and also, as laws of trade or commerce, under sec 99. A preference in relation to any of these subjects which infringed sec 99 would also be a prohibited discrimination or a prohibited lack of uniformity under one of the other sections. Preference necessarily involves discrimination or lack of uniformity, but discrimination or lack of uniformity does not necessarily involve preference."
[6](1936) 54 CLR 657 at 668; [1936] HCA 7.
The limitations imposed by ss 51(ii) and 99, which are in issue in this case, operate at a level of generality appropriate to their federal purposes. They do not prevent the Parliament of the Commonwealth from enacting uniform laws which have different effects in different States because of differences in the circumstances to which they apply, including different State legislative regimes. Nor do they apply to a law with respect to taxation merely because it provides for adjustments to the liabilities it imposes according to liabilities which might from time to time be imposed by differing State laws. The generality of the non‑discrimination and no-preference limitations permits differences between States in the application of the law, for which the law makes provision, if such provision is based upon a distinction which is appropriate and adapted to the attainment of a proper objective[7]. Such a provision neither discriminates nor gives a preference within the meaning of those terms in ss 51(ii) and 99.
[7]Austin v The Commonwealth (2003) 215 CLR 185 at 247 [118]; [2003] HCA 3; Permanent Trustee Australia Ltd v Commissioner of State Revenue (Vict) (2004) 220 CLR 388 at 424 [89]; [2004] HCA 53.
For the reasons that follow, the MRRT Act neither discriminates between States or parts of States nor gives preference to one State over another. For the reasons given in the joint judgment of Hayne, Bell and Keane JJ, s 91 of the Constitution has no effect on the validity of the Act. Nor, for the reasons given by their Honours, does the Act impair the capacity of the States to function as governments contrary to the principles explained in Melbourne Corporation v The Commonwealth[8] and more recently in Austin v The Commonwealth[9] and Clarke v Federal Commissioner of Taxation[10]. The plaintiffs' challenges to the MRRT Act fail.
[8](1947) 74 CLR 31.
[9](2003) 215 CLR 185.
[10](2009) 240 CLR 272; [2009] HCA 33.
The questions reserved
On 5 November 2012, the Court ordered that the following questions be reserved for determination by the Full Court (on the basis of the pleadings and documents referred to in the pleadings):
"(i)Are any or all of s 3 of the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), s 3 of the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth) and s 3 of the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) invalid in their application to the plaintiffs on one or more of the following grounds:
A.they discriminate between the States of the Commonwealth of Australia contrary to s 51(ii) of the Constitution;
B.they give preference to one State of the Commonwealth of Australia over another State contrary to s 99 of the Constitution;
C.they so discriminate against the States of the Commonwealth or so place a particular disability or burden upon the operations or activities of the States, as to be beyond the legislative power of the Commonwealth?
(ii)Are any or all of the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) and the Minerals Resource Rent Tax Act 2012 (Cth) invalid in their application to the plaintiffs on the ground that they are contrary to s 91 of the Constitution?
(iii)Who should pay the cost of the reserved questions?"
An outline of the scheme of the legislation follows.
The structure of the tax
A miner is liable to pay MRRT for an MRRT year equal to the sum of its MRRT liabilities for each of its mining project interests for that year[11]. MRRT liability for a mining project interest for an MRRT year is calculated by the following formula[12]:
"MRRT liability = MRRT rate x (Mining profit – MRRT allowances)".
The MRRT rate is 22.5%[13]. MRRT allowances are listed in Ch 3 of the MRRT Act. The "mining profit" for a mining project interest is the difference between "mining revenue" and "mining expenditure"[14].
[11]MRRT Act, s 10–1.
[12]MRRT Act, s 10–5.
[13]The rate is specified in s 4 of each of the Imposition Acts as "30% x (1 – Extraction factor)" where the extraction factor is 25%.
[14]MRRT Act, s 25–5.
The "mining revenue" for a mining project interest for an MRRT year is "the sum of all the amounts that, under this Act, are included in the miner's mining revenue for that interest for that year."[15] It includes revenue from taxable resources extracted from the project area for the mining project interest, to the extent that the revenue is reasonably attributable to the taxable resources in the form and place they were in when they were at their valuation point[16]. It is not necessary for present purposes to explore the full complexity of the definition of mining revenue.
[15]MRRT Act, s 30–5.
[16]MRRT Act, s 30–1(a). Section 40–5(1) provides that the valuation point for a "taxable resource" is the point just before the resource is removed from the run-of-mine stock pile on which it is stored.
The "mining expenditure" for a mining project interest for an MRRT year is "the sum of all the amounts that, under this Act, are included in the miner's mining expenditure for that interest for that year."[17] It does not include amounts designated as "excluded expenditure"[18]. Payment of a "mining royalty" is "excluded expenditure"[19]. The term "mining royalty" is defined and, relevantly for present purposes, is an expenditure which[20]:
"(a)is made in relation to a taxable resource extracted under authority of a production right; and
(b)is made under a Commonwealth law, a State law or a Territory law; and
(c)either:
(i)is a royalty; or
(ii)would be a royalty, if the taxable resource were owned by the Commonwealth, State or Territory (as the case requires) just before the recovery of the resource."[21]
Although mining royalties payable to a State are excluded expenditure, they are to be deducted from mining profit in calculating MRRT liability. That is because they fall into the category of "MRRT allowances"[22]. That category is dealt with in Ch 3 of the MRRT Act. It consists of a number of classes of allowances which are defined by the Act[23]. The class immediately relevant to these proceedings is the "royalty allowance"[24].
[17]MRRT Act, s 35–5(1).
[18]MRRT Act, s 35–5(2).
[19]MRRT Act, s 35–40.
[20]MRRT Act, s 35–45(1).
[21]Section 35–45(1)(c)(ii) covers the case where an amount is payable under an Australian law in relation to minerals owned by private landowners.
[22]MRRT Act, s 10–10, item 1 and see Pt 3–1.
[23]The MRRT allowances are listed in s 10–10 with cross‑references to the numbered Parts of Ch 3 which apply to them. They are: royalty allowance (Pt 3–1), transferred royalty allowance (Pt 3–2), pre-mining loss allowance (Pt 3–3), mining loss allowance (Pt 3–4), starting base allowance (Pt 3–5), transferred pre-mining loss allowance (Pt 3–6), transferred mining loss allowance (Pt 3–7).
[24]MRRT Act, Pt 3–1.
Royalty allowances are dealt with in Pt 3–1 of Ch 3. That Part consists of Div 60, also entitled "Royalty allowances". The overview of the Division states[25]:
"Mining royalties paid to the Commonwealth, States and Territories reduce a miner's MRRT liabilities for a mining project interest.
To work out the royalty allowance, the amount of the royalty is grossed‑up using the MRRT rate, in effect reducing the MRRT liability by the amount of the royalty."
The mechanism that is adopted for bringing royalties into account is that of "royalty credits"[26]. Royalty credits not applied in one MRRT year can be applied in later years[27]. They are reduced if a miner recoups an amount giving rise to a royalty credit[28].
[25]MRRT Act, s 60–1.
[26]MRRT Act, s 60–10.
[27]MRRT Act, s 60–25(2).
[28]MRRT Act, s 60–30.
A royalty credit includes a liability to pay a mining royalty in relation to a taxable resource extracted under the authority of the production right to which the relevant mining project interest relates[29]. The royalty credit arises at the time the miner incurs the liability and relates to the MRRT year in which it arises[30]. The amount of the royalty credit in the MRRT year in which the royalty credit arises in relation to a liability of a miner is calculated by determining how much of the liability gives rise to a royalty credit and dividing the result by the MRRT rate[31]. A "royalty allowance" is so much of the "royalty credits" as do not exceed the mining profit[32]. The mining project interests to which the MRRT Act applies are interests in relation to iron ore and coal and some related substances. They are called "taxable resources"[33]. If royalty credits in one year are not needed to offset the mining profit in that year, they can be carried over for use in subsequent years[34]. In that event, the amount of the royalty credits is uplifted to take account of the time value of money[35].
[29]MRRT Act, s 60–20(1)(a). There is an extended aspect of the definition which is not material for present purposes.
[30]MRRT Act, s 60–20(2).
[31]MRRT Act, s 60–25(1).
[32]MRRT Act, s 60–15(1).
[33]MRRT Act, ss 15–5(4) and 20–5.
[34]MRRT Act, s 60–25(2).
[35]MRRT Act, s 60–25(2).
The MRRT does not become payable until the miner's group mining profit for an MRRT year exceeds $75 million[36]. The full amount of MRRT does not become payable until the group mining profit reaches $125 million[37].
[36]MRRT Act, ss 10–15 and 45–5.
[37]MRRT Act, ss 10–15 and 45–10.
The plaintiffs submitted that the effect of the MRRT Act is that a miner's MRRT liability, when payable, is either inversely proportional to the miner's liability for State mining royalties or is directly related to the extent of the miner's liability for such royalties. That is to say, the MRRT Act is expressly designed so that if more State royalties are payable, less MRRT is payable, and vice versa. The plaintiffs submitted that, in the result, where MRRT is payable, a miner's liability will vary from State to State, depending upon the royalty rate applicable in that State. The Commonwealth took issue with the plaintiffs about the relationship between MRRT liabilities and State royalties. It did so by reference to the different times at which, and conditions under which, MRRT liabilities and State royalties could become payable.
There are undoubtedly a number of variables which can affect the liability of a miner for MRRT in a given year or over a number of years. One of those variables is the royalty payable from time to time under State law. It is not necessary for present purposes to explore hypothetical cases that might arise and differences in the liabilities which might attach or be attributed to mining projects in one State or another. The issues raised in the questions reserved can be decided on the basis that, all other things being equal, the MRRT Act can have the effect that a miner's liability for MRRT is greater in a State with a lower applicable royalty than in a State with a higher applicable royalty. It can therefore also have the effect that when a State reduces the applicable royalty, a miner's liability for MRRT, all other things being equal, will increase. The arithmetical gymnastics that, according to the plaintiffs, would enable the outcomes to be characterised as the application of different "effective" MRRT rates between States can be disregarded.
In the forefront of consideration in this case is the interpretation and application of ss 51(ii) and 99 of the Constitution. Their interpretation depends upon their text. It is informed by their drafting history and the decisions of this Court interpreting and applying them. Those decisions do not yield single, simply expressed and exhaustive explanations and definitions of the limitations on legislative power imposed by those provisions. The Court responds to the cases it is called upon, by the accidents of history, to decide. Judicial interpretation in particular cases must be seen in the context of the Court's function. As Windeyer J said in the Payroll Tax Case[38]:
"Exegesis must not be substituted for the text."
That observation should be read in light of his Honour's approach to constitutional interpretation in the Australian context[39]:
"In any country where the spirit of the common law holds sway the enunciation by courts of constitutional principles based on the interpretation of a written constitution may vary and develop in response to changing circumstances. This does not mean that courts have transgressed lawful boundaries: or that they may do so."
What Windeyer J said echoed the remarks of Alfred Deakin, first Attorney‑General of the Commonwealth, in his Second Reading Speech for the Judiciary Bill 1902 (Cth) in March 1902[40]:
"It is as one of the organs of Government which enables the Constitution to grow and to be adapted to the changeful necessities and circumstances of generation after generation that the High Court operates."
It is in that spirit that ss 51(ii) and 99 in their application to this case should be interpreted. That is a conservative spirit which nevertheless recognises that a written constitution should be able, consistently with textual limitations, to accommodate changing circumstances. That approach, in this case, requires consideration of the text and the drafting histories of ss 51(ii) and 99, their judicial exegesis and the particular questions to be decided about their application.
[38]Victoria v The Commonwealth (1971) 122 CLR 353 at 403; [1971] HCA 16.
[39](1971) 122 CLR 353 at 396–397.
[40]Australia, House of Representatives, Parliamentary Debates (Hansard), 18 March 1902 at 10967, cited in New South Wales v The Commonwealth (Work Choices Case) (2006) 229 CLR 1 at 73–74 [54] per Gleeson CJ, Gummow, Hayne, Heydon and Crennan JJ; [2006] HCA 52.
Constitution, s 51(ii) — drafting history
Section 51(ii) confers power on the Parliament of the Commonwealth, subject to the Constitution, to make laws for the peace, order and good government of the Commonwealth with respect to:
"taxation; but so as not to discriminate between States or parts of States".
The ambit of the power is expressly confined by the "positive prohibition or restriction" against discrimination between States or parts of States[41]. It stands adjacent to s 51(iii), which authorises the Parliament to make laws with respect to "bounties on the production or export of goods, but so that such bounties shall be uniform throughout the Commonwealth". Their drafting histories are closely connected. In the drafts proposed at the National Australasian Convention in Sydney in 1891[42] and the drafts reviewed at the Adelaide[43] and Sydney[44] sessions of the Convention in 1897, the powers in relation to both taxation and bounties were subject to a uniformity requirement. The provisions relating to customs and excise and bounties were separated into two clauses in 1898 and the former was overtaken by the general taxation power[45], a separation maintained in the final draft adopted by the Convention[46].
[41]Work Choices Case (2006) 229 CLR 1 at 127 [219]–[221].
[42]Inglis Clark Draft, 1891, cll 45(I), 55, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 85, 87; Kingston Draft, 1891, Pt XII, cl IV, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 129; First Official Draft, Sydney, 1891, cl 30(2), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 143; Final Draft, Sydney, 1891, cl 52(2), (3), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 446.
[43]Adelaide Draft, 1897, cl 50(II), (III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 509.
[44]Sydney Draft, 1897, cl 52(II), (III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 777.
[45]Melbourne Draft, 1898, cl 52(II), (III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 873.
[46]Final Draft, 1898, cl 51(II), (III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 1127.
From the first drafts of the Constitution considered at the National Australasian Convention in Sydney in 1891[47] up to those considered at the 1898 Convention session held in Melbourne[48], it was proposed that the Commonwealth Parliament's power to impose taxation should be "uniform" throughout the Commonwealth. That constraint appeared in cl 55 of Inglis Clark's draft[49], which informed much of the draft adopted by the 1891 Convention. It also appeared in Charles Kingston's draft[50]. It was inspired by Art I, s 8(1) of the United States Constitution, which required that "all Duties, Imposts and Excises shall be uniform throughout the United States".
[47]Inglis Clark Draft, 1891, cll 45(I), 55, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 85, 87; Kingston Draft, 1891, Pt XII, cl IV, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 129; First Official Draft, Sydney, 1891, cl 30(2), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 143; Final Draft, Sydney, 1891, cl 52(3), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 446.
[48]Adelaide Draft, 1897, cl 50(III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 509; Sydney Draft, 1897, cl 52(III), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 777; Melbourne Draft, 1898, cl 52(II), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 922; cf Final Draft, 1898, cl 51(II), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 1127.
[49]Inglis Clark Draft, 1891, cl 55, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 87.
[50]Kingston Draft, 1891, Pt XII, cl IV, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 129.
At the time of the 1891 Convention, the uniformity requirement in Art I, s 8(1), which applied only to indirect taxes, had been considered by the Supreme Court of the United States in the Head Money Cases[51]. Miller J, delivering the opinion of the Court, said[52]:
"The tax is uniform when it operates with the same force and effect in every place where the subject of it is found."
The criterion of uniformity under Art I was understood to be geographical. However, that understanding was called into question shortly before the 1897 and 1898 Convention sessions by a separate concurring opinion of Field J in Pollock v Farmers' Loan and Trust Co[53]. Field J construed the uniformity requirement as forbidding a tax‑free threshold and the imposition of different rates of the same tax on property income according to whether the income was derived by natural persons or various classes of corporation[54]. That opinion raised a concern at the National Australasian Convention which led to a change in the text of what became s 51(ii).
[51]Edye v Robertson 112 US 580 (1884).
[52]112 US 580 at 594 (1884).
[53]157 US 429 (1895).
[54]157 US 429 at 595 (1895). The observations were not part of the ratio of the Court as the challenged law was held by the majority to be invalid on the basis that the tax was a direct tax, which did not attract the uniformity requirement: 157 US 429 at 583 (1895); see also at 607 per Field J.
The change from a requirement of uniformity to a prohibition against discrimination appeared in the draft Constitution produced at the Melbourne session of the Convention on 12 March 1898[55]. It was explained by Edmund Barton as a cautious response to the "expressions" in the opinion of Field J in Pollock. In moving his amendment, Barton said[56]:
"I think that although the word 'uniform' has the meaning it was intended to have—'one in form' throughout the Commonwealth—still there might be a difficulty, and litigation might arise about it, and prolonged trouble might be occasioned with regard to the provision in case, for instance, an income tax or a land tax was imposed. What is really wanted is to prevent a discrimination between citizens of the Commonwealth in the same circumstances."
He described the amendment as preventing discrimination "or any form of tax which would make a difference between the citizen of one state and the citizen of another state, and to prevent anything which would place a tax upon a person going from one state to another."[57] Professor Harrison Moore, writing in 1910, summed up the concerns enlivened by the opinion of Field J. The uniformity requirement, he said, "was more than the federal spirit required; it prevented not merely discrimination among the States, but discrimination in the case of individuals", so the Convention "adopted terms of geographical limitation."[58]
[55]Clause 52(II), reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 954 read with the statements at 802 regarding Document 31.7B.
[56]Official Record of the Debates of the Australasian Federal Convention, (Melbourne), 11 March 1898 at 2397.
[57]Official Record of the Debates of the Australasian Federal Convention, (Melbourne), 11 March 1898 at 2397.
[58]Harrison Moore, The Constitution of the Commonwealth of Australia, 2nd ed (1910) at 516.
Quick and Garran characterised the constraint in s 51(ii) as a "limitation ... provided for federal reasons"[59] being directed against "a system of taxation designed to press more heavily on people or property in some States than on people or property in other States."[60] Thus[61]:
"to impose a high tax on commodities or persons in one State and a low tax on the same class of commodities or persons in another State, would be to discriminate. Such discriminations are forbidden, and uniformity of taxation throughout the Commonwealth is an essential condition of the validity of every taxing scheme."
Quick and Garran characterised the constraint in s 51(ii) as "practically the same in substance as the requirement of Art 1, s 8, sub-s 1, of the United States Constitution"[62]. That conclusion rested upon the unstated but correct assumption that what Field J had said did not state the law in the United States before or after Pollock.
[59]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 550.
[60]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 550.
[61]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 550.
[62]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 550.
The connection between Art I, s 8(1) and s 51(ii), reflected in the drafting history of s 51(ii), led to submissions in this case about decisions of the Supreme Court of the United States on the uniformity requirement. Some of them should be mentioned. The Head Money Cases and Pollock have been referred to. The geographical character of the uniformity requirement, rejected by Field J in Pollock, was reaffirmed in Knowlton v Moore[63]. White J, delivering the opinion of the Court, quoted one of the delegates to the Constitutional Convention of 1787, Luther Martin, who observed that some duties might be laid on articles little used in some States and much used in others[64]:
"in which case, the first would pay little or no part of the revenue arising therefrom, while the whole or nearly the whole of it would be paid by the last, to wit, the States which use and consume the articles on which imposts and excises are laid."
Much, of course, depends upon the level of generality of the requirement for uniformity or non-discrimination. The requirement for geographical uniformity in the United States was pitched by the decisions of the Supreme Court at a level of generality permitting differences across State boundaries in specific applications of the law. At a level of generality appropriate to its federal purpose, the non-discrimination requirement in s 51(ii) excludes, from legislative power with respect to taxation, laws which make distinctions between States or parts of States which are inconsistent with the economic unity of the Commonwealth and the status of the States and their people as equals inter se in the Federation. That level of generality does not require the exclusion from the scope of the taxation power of a uniform rule incorporating adjustments of liabilities that take account of liabilities imposed by State laws.
[63]178 US 41 (1900): a case concerning death duties.
[64]178 US 41 at 106 (1900).
Reflecting that concept of uniformity informed by federal considerations, the Supreme Court of the United States in Florida v Mellon[65] rejected as "without merit" a contention that a federal inheritance tax was not uniform because it allowed for deductions of State inheritance taxes when not all States imposed such taxes. All that the Constitution required was that[66]:
"the law shall be uniform in the sense that by its provisions the rule of liability shall be the same in all parts of the United States."
An analogous issue arose in Phillips v Commissioner of Internal Revenue[67], which concerned a federal law for recovery of corporate taxes from stockholders who had received the assets of a dissolved corporation. The Court did not accept an argument that the law offended Art I, s 8(1) on the basis that the liabilities might differ from State to State because of differences in State laws. Brandeis J, delivering the opinion of the Court, said that[68]:
"The extent and incidence of federal taxes not infrequently are affected by differences in state laws; but such variations do not infringe the constitutional prohibitions against delegation of the taxing power or the requirement of geographical uniformity."
The "settled doctrine" of the Court that "the uniformity exacted is geographical, not intrinsic" was reaffirmed in Steward Machine Co v Davis[69].
[65]273 US 12 (1927).
[66]273 US 12 at 17 (1927).
[67]283 US 589 (1931).
[68]283 US 589 at 602 (1931).
[69]301 US 548 at 583 (1937) per Cardozo J, delivering the opinion of the Court.
What might be thought to be a limiting decision was reached in 1983 in United States v Ptasynski[70]. The Supreme Court held that the Crude Oil Windfall Profit Tax Act of 1980, which exempted from the tax which it imposed domestic crude oil produced from wells within a defined geographical area in Alaska, was valid. The exemption was found not to have been drawn on State political lines, but to reflect a legislative judgment that unique climatic and geographic conditions required that oil produced from the exempt area be treated as a separate class of oil[71]. The general principle was that[72]:
"The Uniformity Clause gives Congress wide latitude in deciding what to tax and does not prohibit it from considering geographically isolated problems."
[70]462 US 74 (1983).
[71]462 US 74 at 78 (1983) per Powell J, delivering the opinion of the Court.
[72]462 US 74 at 84 (1983).
The plaintiffs submitted that the United States decisions were "not on point". They quoted an observation of Dixon CJ in Deputy Federal Commissioner of Taxation v Brown[73] that s 51(ii) "may not be the same as art 1, s 8 of the Constitution of the United States"[74]. In its context, which concerned the application of s 79 of the Judiciary Act 1903 (Cth) in taxation recovery proceedings, the observation was not apposite to the plaintiffs' proposition. The plaintiffs went further and characterised the Supreme Court's decisions on Art I, s 8(1) as "illogical" and "appear[ing] to neuter the requirement for uniformity."[75] The plaintiffs' submissions on the utility of the decisions of the Supreme Court on Art I, s 8(1) should not be accepted. They reduce to a complaint about the level of generality at which the uniformity requirement in Art I, s 8(1) has been interpreted.
[73](1958) 100 CLR 32; [1958] HCA 2.
[74](1958) 100 CLR 32 at 39; the balance of the sentence being "but what the Supreme Court has said about State law in the collection of federal taxes seems to me to be true of our system."
[75]The plaintiffs cited, in support of their criticism of the United States decisions, a trenchant academic article: Claus, "'Uniform Throughout the United States': Limits on Taxing as Limits on Spending", (2001) 18 Constitutional Commentary 517 at 522‑529. Not surprisingly, a variety of academic perspectives have been expressed in relation to decisions of the United States Supreme Court on Art I, s 8(1): eg Lund, "The Uniformity Clause", (1984) 51 University of Chicago Law Review 1193 especially at 1200; Norton, "The Limitless Federal Taxing Power", (1985) 8 Harvard Journal of Law and Public Policy 591 at 604–605; Eggleston, "United States v Ptasynski: A Windfall for Congress", (1984) 61 Denver Law Journal 395 at 402.
The drafting history of s 51(ii) does not support an argument that the non-discrimination limitation differs fundamentally from the uniformity requirement in Art I, s 8(1) as it was understood before and after the "expressions" of Field J in Pollock. Quick and Garran's treatment of the two provisions as equivalent is supportive of that proposition, as are the observations of Harrison Moore. While decisions of the Supreme Court of the United States on uniformity cannot automatically be treated as applicable to the non‑discrimination constraint in s 51(ii), they are appropriate sources of comparative constitutional law in its construction. In each case the principle underlying the limitation is a federal principle. In this country it allows the Commonwealth Parliament to make laws with respect to taxation which, by reason of differing circumstances, including State legal regimes, may have different effects in different States. As appears below, the principle does not preclude the Commonwealth Parliament from incorporating in its taxation laws uniform provisions of general application providing adjustments to the liabilities which they impose by reference to liabilities imposed under State law. It has done so for very many years.
Drafting history — s 99
Section 99 of the Constitution provides:
"The Commonwealth shall not, by any law or regulation of trade, commerce, or revenue, give preference to one State or any part thereof over another State or any part thereof."
Section 99 was inspired by Art I, s 9(6) of the United States Constitution, which provides that:
"No Preference shall be given by any Regulation of Commerce or Revenue to the Ports of one State over those of another".
Clause 54 of Inglis Clark's draft in 1891, like the United States provision, prohibited preference to the "ports of one Province over those of another" and added "nor shall vessels bound to or from one Province be obliged to enter or clear or pay duties in another."[76] Kingston's draft contained a similar provision[77]. In the final draft, which emerged from the 1891 Sydney session of the Convention, the equivalent provision, under the heading "Equality of Trade", was cl 11 of Ch IV, entitled "Finance and Trade" and provided[78]:
"Preference shall not be given by any law or regulation of commerce or revenue to the ports of one part of the Commonwealth over those of another part of the Commonwealth."
[76]Inglis Clark Draft, 1891, cl 54, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 87.
[77]Kingston Draft, 1891, Pt XII, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 130.
[78]Final Draft, Sydney, 1891, Ch IV, cl 11, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 454.
The no-preference provision, which emerged from the 1897 sessions of the Convention as cl 95, prohibited preference to the ports of one State over the ports of another with the addition that any law or regulation derogating from freedom of trade and commerce between different parts of the Commonwealth should be null and void[79]. A number of amendments were debated at the Melbourne session of the Convention in 1898. There was substantive discussion at that session which, among other things, canvassed the necessity for the provision and whether it should apply to State laws. The final text of s 99 reflected in substance wording proposed by Edmund Barton[80].
[79]Final Draft, Adelaide, 1897, cl 95, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 606.
[80]Official Record of the Debates of the Australasian Federal Convention, (Melbourne), 22 February 1898 at 1329.
Quick and Garran viewed s 99 in its application to taxation laws as adding little, if anything, to s 51(ii). They said[81]:
"This section, therefore, extends to all laws and regulations of trade, commerce, and revenue, the condition which is elsewhere imposed with regard to laws dealing with taxation—viz, that they shall not discriminate between States or parts of States."
Its object was "to prevent federal favoritism and partiality in commercial and other kindred regulations."[82] A similar view of the relationship between the uniformity and no-preference rules in Art I, s 8(1) and Art I, s 9(6) of the United States Constitution had been expressed in Knowlton v Moore[83], in which the Supreme Court held that, although couched in different language, they had "absolutely the same significance."[84]
[81]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 877.
[82]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 877.
[83]178 US 41 (1900).
[84]178 US 41 at 104 (1900).
This Court's treatment of the relationship between the non-discrimination and no-preference limitations recognises that "while preference necessarily involves discrimination or lack of uniformity, the latter does not necessarily involve the former."[85] In this case that has the consequence that if the MRRT Act cannot be said to discriminate within the meaning of s 51(ii) it cannot be said to give a preference within the meaning of s 99.
[85]Permanent Trustee Australia Ltd v Commissioner of State Revenue (Vict) (2004) 220 CLR 388 at 423 [88] per Gleeson CJ, Gummow, Hayne, Callinan and Heydon JJ, citing Elliott v The Commonwealth (1936) 54 CLR 657 at 668 per Latham CJ, see also at 683 per Dixon J.
The difficulty of identifying a prohibited preference given to one State over another where there were dissimilar circumstances was recognised by Quick and Garran. They foreshadowed the application of a criterion of reasonableness to the characterisation of preferences[86]:
"If a difference of treatment is arbitrary, or if its purpose is to advantage or prejudice a locality, it is undue and unreasonable, and is accordingly a preference. If on the other hand the difference of treatment is the reasonable result of the dissimilarity of circumstances—or if it is based on recognized and reasonable principles of administration—it is no preference."
That approach to characterisation was reflected in the general observation about the concept of discrimination made by Gaudron, Gummow and Hayne JJ in Austin v The Commonwealth[87], and quoted by the majority in Permanent Trustee Australia Ltd v Commissioner of State Revenue (Vict) in its discussion of the application of s 99[88]:
"The essence of the notion of discrimination is said to lie in the unequal treatment of equals or the equal treatment of those who are not equals, where the differential treatment and unequal outcome is not the product of a distinction which is appropriate and adapted to the attainment of a proper objective." (footnotes omitted)
Their Honours' observation did not amount to a qualification justifying a law which would otherwise exceed the constitutional limitations. It set out a criterion for characterisation of a law as discriminatory for the purposes of s 51(ii). It was invoked by the Commonwealth in its submissions. The plaintiffs submitted that the reasoning of the majority in Permanent Trustee in this respect should not be followed. That submission should not be accepted. Before considering it further, however, it is desirable to consider the concepts of discrimination and preference in ss 51(ii) and 99 as they have emerged from the decisions of this Court.
[86]Quick and Garran, The Annotated Constitution of the Australian Commonwealth, (1901) at 878.
[87](2003) 215 CLR 185 at 247 [118].
[88](2004) 220 CLR 388 at 424 [89].
Sections 51(ii) and 99 — discrimination, preference and differential operation
The uniformity requirement in the draft Constitution, as it stood after the Convention session held in Adelaide in 1897, attracted a "friendly suggestion"[89] from the Colonial Office in the form of a question: "does 'uniform' mean uniform in law, or uniform in effect?"[90] When Edmund Barton moved his amendment to replace uniformity with non-discrimination he made clear that the answer was "uniform in law". Laws with respect to taxation were to be "one in form" throughout the Commonwealth[91]. That approach was reflected in the first reported judicial consideration of s 51(ii), which was undertaken by the Full Court of the Supreme Court of Queensland, in The Colonial Sugar Refining Co Ltd v Irving[92]. Sir Samuel Griffith, then Chief Justice of Queensland and one month short of his appointment as the first Chief Justice of this Court, observed, consistently with the drafting history of s 51(ii), that[93]:
"the discrimination must depend upon the geographical position, and not upon the accident of whether things happen to be found in one State or in another."
The Full Court held that a Commonwealth law, imposing liability to excise duty on goods and providing an exemption for goods which had been subject to excise duties under State laws, did not discriminate within the meaning of s 51(ii). The Privy Council agreed[94]:
"The rule laid down by the Act is a general one, applicable to all the States alike, and the fact that it operates unequally in the several States arises not from anything done by the Parliament, but from the inequality of the duties imposed by the States themselves."
[89]So described by Colonial Secretary Joseph Chamberlain in a covering letter to George Reid in July 1897, reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 714.
[90]Memorandum C, "Australian Federal Constitution. Criticisms on the Bill", reproduced in Williams, The Australian Constitution: A Documentary History, (2005) at 728.
[91]Official Record of the Debates of the Australasian Federal Convention, (Melbourne), 11 March 1898 at 2397.
[92][1903] St R Qd 261.
[93][1903] St R Qd 261 at 276–277, Cooper J agreeing at 277, Real J agreeing at 281.
[94]Colonial Sugar Refining Co Ltd v Irving [1906] AC 360 at 367.
The plaintiffs submitted that a Commonwealth tax cannot impose different tax rates on different taxpayers in different States even when the result is that the total tax burden, both Commonwealth and State, upon all taxpayers is the same. They argued that CSR did not apply to such a case because in CSR the impugned duty was made payable on all sugar on which customs or excise duty had not been paid pursuant to State laws before 8 October 1901. That criterion of liability was said to identify a class of goods in respect of which excise duty was payable and to which it applied uniformly. That may be one way of characterising the tax in CSR. But, as appears from the judgments of the Full Court and the Privy Council, the basis upon which the tax was upheld was not so narrowly framed[95].
[95][1903] St R Qd 261 at 276–277; [1906] AC 360 at 367.
It is not controversial that a law which is uniform across the Commonwealth and does not in terms discriminate between States or parts of States can nevertheless have different effects between and within the States because of the circumstances upon which it operates, including the different State legal regimes with which it interacts. An example in the latter category from the United States is a law of the kind considered in Phillips v Commissioner of Internal Revenue[96], referred to earlier in these reasons.
[96]283 US 589 (1931).
It may be accepted that a Commonwealth law with respect to taxation which expressly provides, in a uniform rule, for the adjustment of the liabilities it imposes by reference to liabilities imposed by State laws is not logically completely congruent with a law which has differential effects across State boundaries or between parts of States because of its interaction with particular State laws. That does not mean, however, that such a law discriminates between States or parts of States. The term "discriminate" may vary in its precise meaning according to its context and can be difficult to define and apply. However that may be, as interpreted by the decisions of this Court on s 51(ii), it does not place the MRRT Act beyond power. As the plurality said of the concept of discrimination generally in Bayside City Council v Telstra Corporation Ltd[97]:
"It involves a comparison, and, where a certain kind of differential treatment is put forward as the basis of a claim of discrimination, it may require an examination of the relevance, appropriateness, or permissibility of some distinction by reference to which such treatment occurs, or by reference to which it is sought to be explained or justified." (footnote omitted)
Their Honours went on to emphasise that judgments about relevance, appropriateness or permissibility of a distinction may be influenced strongly by context[98].
[97](2004) 216 CLR 595 at 629–630 [40]; [2004] HCA 19.
[98](2004) 216 CLR 595 at 630 [40].
The Commonwealth submitted that a correct formulation of the concept of discrimination in s 51(ii) was to be found in the judgment of Isaacs J in his dissent in R v Barger[99], the first reported decision of this Court on s 51(ii). Isaacs J said[100]:
"Discrimination between localities in the widest sense means that, because one man or his property is in one locality, then, regardless of any other circumstance, he or it is to be treated differently from the man or similar property in another locality."
Higgins J reasoned along similar lines and observed that it would not be discrimination between States or parts of States if a graduated income tax were introduced when incomes were higher in one State than in another[101].
[99](1908) 6 CLR 41; [1908] HCA 43.
[100](1908) 6 CLR 41 at 110.
[101](1908) 6 CLR 41 at 133.
Controversy later attended another observation made by Isaacs J, in the same judgment, that discrimination under s 51(ii) was "preference of locality merely because it is locality, and because it is a particular part of a particular State."[102] Although quoted and approved by the Privy Council in W R Moran Pty Ltd v Deputy Federal Commissioner of Taxation (NSW)[103], it was a proposition which many years later in Commissioner of Taxation v Clyne[104] Dixon CJ had the "greatest difficulty in grasping"[105] and which Professor Geoffrey Sawer critically characterised as establishing a special criterion of "Stateishness"[106].
[102](1908) 6 CLR 41 at 108.
[103](1940) 63 CLR 338; [1940] AC 838.
[104](1958) 100 CLR 246; [1958] HCA 10.
[105](1958) 100 CLR 246 at 266.
[106]"Commonwealth Taxation Laws—Uniformity and Preference", (1958) 32 Australian Law Journal 132.
The majority in Barger[107] took a stronger view of the prohibition against discrimination in s 51(ii) in its application than did Isaacs and Higgins JJ but did not in terms disagree with the "widest sense" of discrimination formulated by Isaacs J. Moreover, the primary finding of the majority was that the impugned legislation, which provided for the exemption from excise of certain articles according to labour conditions in the area in which the articles were manufactured, was not a law with respect to taxation. Their secondary finding, that it discriminated between States or parts of States, was necessarily made on the hypothesis that the primary finding was wrong. Their conclusion as to discrimination was reached in the shadow of the reserved powers doctrine, which the majority described as a rule which was "different, but ... founded upon the same principles."[108] In that setting a strong view of the prohibition was not surprising. In a passage relied upon by the plaintiffs, the majority also distinguished CSR, observing that[109]:
"if the Excise duty had been made to vary in inverse proportion to the Customs duties in the several States so as to make the actual incidence of the burden practically equal, that would have been a violation of the rule of uniformity."
[107]Griffith CJ, Barton and O'Connor JJ.
[108](1908) 6 CLR 41 at 72.
[109](1908) 6 CLR 41 at 70–71.
Despite their dissent in Barger, the differential operation of laws permitted under the approach taken by Isaacs and Higgins JJ did not differ markedly from that permitted in later cases and in decisions on Art I, s 8(1) by the United States Supreme Court. The formulation by Isaacs J of discrimination in the "widest sense" was expressly adopted and applied in Cameron v Deputy Federal Commissioner of Taxation[110]. Starke J drew the necessary distinction between a discriminatory tax law and a non-discriminatory tax law which has a differential operation or effect[111]:
"A law with respect to taxation applicable to all States and parts of States alike does not infringe the Constitution merely because it operates unequally in the different States—not from anything done by the law‑making authority, but on account of the inequality of conditions obtaining in the respective States."
Knox CJ and Powers J, in James v The Commonwealth[112], also expressly adopted the formulation by Isaacs J in Barger and his equation of the non-discrimination limitation in s 51(ii) with the no‑preference rule in s 99[113]. Higgins J, "[a]fter twenty years", adhered to what he had said in Barger and asserted its relevance to s 99 "as one cannot conceive of any preference without discrimination"[114]. Starke J adhered to what he had said in Cameron[115]:
"if a law is not applicable to all States alike, then it operates unequally between the States, and discriminates as a law between them."
[110](1923) 32 CLR 68 at 72 per Knox CJ, 76 per Isaacs J, 78–79 per Higgins J, 79 per Rich J; see also at 79 per Starke J; [1923] HCA 4 (in which the Court held invalid a regulation under the Income Tax Assessment Act 1915 (Cth) fixing the value of various classes of livestock by State for the purpose of calculating profits and assessable income).
[111](1923) 32 CLR 68 at 79.
[112](1928) 41 CLR 442; [1928] HCA 45.
[113](1928) 41 CLR 442 at 455–456.
[114](1928) 41 CLR 442 at 460.
[115](1928) 41 CLR 442 at 464.
As Dennis Rose wrote in 1977, what Isaacs J said in his often quoted definition of discrimination in the "widest sense" had nothing to do with the proposition in the same judgment that discrimination for the purposes of s 51(ii) is limited to discrimination between localities as States or as parts of States[116]. Rose correctly observed that much of the subsequent support for what Isaacs J had said referred to the "widest sense" formulation.
[116]Rose, "Discrimination, Uniformity and Preference—Some Aspects of the Express Constitutional Provisions", in Zines (ed), Commentaries on the Australian Constitution, (1977) 191 at 194.
In Elliott v The Commonwealth[117], which involved the trade and commerce aspect of s 99, the Court, by majority[118], held valid regulations for the licensing of seamen applicable only to ports specified by the Minister. Latham CJ followed the approach taken by Isaacs and Higgins JJ in Barger, by the majority in Cameron, and by Knox CJ and Powers J in James. The Commonwealth, he held, was empowered to adjust its legislation to the varying circumstances of particular ports[119]. Rich J reasoned similarly, but more briefly[120]. Starke J held that legislation which discriminated between localities and made special rules for various occupations was "often desirable, but ... by no means preferences prohibited by sec 99."[121] Dixon J, in dissent, also took the view that discrimination between States did not necessarily involve a preference of one over the other[122]. Evatt J acknowledged that Barger remained the leading authority on s 99, but preferred the view of the majority of the Court in that case, that s 99 "forbids all preferences which arise solely as a legal consequence of association with or reference to any locality in 'Australia,' ie, 'one or more of the States of Australia.'"[123]
[117](1936) 54 CLR 657.
[118]Latham CJ, Rich, Starke and McTiernan JJ, Dixon and Evatt JJ dissenting.
[119](1936) 54 CLR 657 at 676 — on the basis that discrimination forbidden by s 99 was "not merely locality as such, but localities which for the purpose of applying the discrimen are taken as States or parts of States": at 675.
[120](1936) 54 CLR 657 at 678.
[121](1936) 54 CLR 657 at 680.
[122](1936) 54 CLR 657 at 683.
[123](1936) 54 CLR 657 at 690.
In Deputy Federal Commissioner of Taxation (NSW) v W R Moran Pty Ltd[124], little was said by the High Court about discrimination. The legislative scheme in issue comprised a Commonwealth law which imposed a uniform tax coupled with a law appropriating money for grants to Tasmania. The grants were made to enable the State to pay rebates to Tasmanian flour and wheat producers on the tax which they had paid to the Commonwealth. Once it was accepted that the relevant taxation laws, applying as they did a common rule, did not discriminate, the conclusion that the scheme as a whole did not contravene s 51(ii) did not require exegesis of the concept of discrimination. To that extent, the approval by the Privy Council of what Isaacs J said in Barger was not directly apposite to its reasoning, which rejected the attack on the scheme as one "really based on the exercise by the Commonwealth Parliament of its powers under sec 96."[125]
[124](1939) 61 CLR 735; [1939] HCA 27.
[125](1940) 63 CLR 338 at 349; [1940] AC 838 at 857.
Commissioner of Taxation v Clyne[126] involved, inter alia, a question whether s 79A of the Income Tax and Social Services Contribution Assessment Act 1936 (Cth), which prescribed allowable deductions in different amounts for residents of different geographical zones, offended against the restrictions in ss 51(ii) and 99. That question was not answered for reasons to do with the way the issues fell out in the case. However, Dixon CJ, with whom McTiernan, Williams, Kitto and Taylor JJ agreed, rejected the proposition, derived from the judgment of Isaacs J in Barger, that taxing legislation would not discriminate unless in some way the parts of the State in respect of which it discriminates were selected by virtue of their character as parts of a State[127]. Dixon CJ said[128]:
"I find myself unable to appreciate the distinction between the selection by an enactment of an area in fact forming part of a State for the bestowal of a preference upon the area and the selection of the same area for the same purpose 'as part of the State'."
That observation did not involve any rejection of the formulation by Isaacs J of "discrimination" in its "widest sense" as used in s 51(ii).
[126](1958) 100 CLR 246.
[127](1958) 100 CLR 246 at 266.
[128](1958) 100 CLR 246 at 266.
Under the general principle that a non-discriminatory law may have different effects according to its interaction with different State laws, Taylor J in Conroy v Carter[129], with the concurrence of Kitto and Windeyer JJ, characterised as non-discriminatory the deductibility under Commonwealth income tax laws of sums paid by taxpayers for land tax imposed under any law of a State. He said[130]:
"This is a provision which operates generally throughout the Commonwealth and the fact that in some States there may be no legislation imposing land tax does not mean that it discriminates between the States."
The asserted discrimination in Conroy was related to liability for certain Commonwealth levies, which depended upon the existence or otherwise of arrangements between the Commonwealth and particular States. The Court divided evenly and, by a statutory majority, held the impugned provision invalid. However, nothing in the reasons of Menzies J, who wrote the principal judgment for that majority, conflicted with the observation of Taylor J concerning the deductibility of sums paid under State law from income assessable for the purposes of the Commonwealth law. The Commonwealth relied upon the statement by Menzies J, with which Barwick CJ and McTiernan J agreed[131]:
"in determining whether a law imposes such a discriminatory burden, it is to the law itself that attention must be paid, not to the laws of any State or States."
[129](1968) 118 CLR 90; [1968] HCA 39.
[130](1968) 118 CLR 90 at 101, Kitto J agreeing at 96, Windeyer J agreeing at 104.
[131](1968) 118 CLR 90 at 103.
The passage from the judgment of Taylor J, including what his Honour said about the deductibility of State land tax from assessable income, was footnoted by Gleeson CJ, Gummow and Hayne JJ in support of their Honours' observation in Austin v The Commonwealth[132]:
"A law with respect to taxation, in general, does not discriminate in the sense spoken of in s 51(ii) if its operation is general throughout the Commonwealth even though, by reason of circumstances existing in one or more of the States, it may not operate uniformly."
The inclusion, under the rubric of differential but non-discriminatory operation, of a taxation law providing for the deductibility of expenditures incurred under State laws may unite categories of differential operation which are not precisely logically congruent. It nevertheless reflects an interpretation of the non-discrimination constraint at a level of generality which is consistent with its federal purpose.
[132](2003) 215 CLR 185 at 247 [117]. Their Honours also cited Deputy Federal Commissioner of Taxation (NSW) v W R Moran Pty Ltd (1939) 61 CLR 735 at 764 and W R Moran Pty Ltd v Deputy Federal Commissioner of Taxation (NSW) (1940) 63 CLR 338 at 349; [1940] AC 838 at 857.
The Commonwealth invoked the longstanding deductibility, for income tax purposes, of State payroll tax, State land tax, State royalties and "indeed any State impost that is an expense or outgoing incurred by a taxpayer in the circumstances identified in s 8–1(a) or (b) of the Income Tax Assessment Act 1997 (Cth)" ("the ITAA 1997"). The plaintiffs argued that there is a critical difference between the way in which royalty credits affect the imposition of the MRRT and the way in which deductions for State imposts are permitted by the ITAA 1997. That distinction was, with respect, an irrelevant matter of form rather than of substance. It may be accepted that the longstanding provision in taxation laws for deductions for expenses which may include liabilities under State laws does not itself provide the determinative answer to the constitutional question in any given case: does a law of taxation which makes such allowances impermissibly discriminate between States? Nevertheless, the subsistence of such laws over a long period of time, reflecting a practical and legitimate interaction in Commonwealth and State financial relationships, may constitute "circumstances" of the kind to which Windeyer J referred in the Payroll Tax Case which in turn inform the contemporary interpretation and application of the Constitution. They may, on that basis, be relevant to the application of a criterion of the kind foreshadowed by Quick and Garran in determining whether an impugned law discriminates or gives a preference within the meaning of the limitations imposed by ss 51(ii) and 99. That question is considered in the next section of these reasons.
Reasonable differences
The Commonwealth submitted that even if the MRRT Act gave rise to differential treatment or unequal outcomes as between States, it did not follow that it was a law made "so as to discriminate between States or parts of States". Relying upon the passage from Austin quoted in Permanent Trustee and set out earlier in these reasons[133], the Commonwealth submitted that:
•the MRRT being a tax on profits, not on revenue, Parliament was entitled to conclude that profits could not accurately be identified without regard to costs and outgoings incurred in the course of deriving revenue — one such class of costs and outgoings being royalty payments made to the relevant State Government;
•the MRRT being a tax on above normal profits or economic rents, the Act proceeds on the basis that royalties may indirectly and at least in part constitute charges on the economic rents which the Act makes subject to taxation. To ignore State royalties in the calculation of the MRRT liability would be to risk imposing a tax on economic rents at a higher rate than intended or on profits that were merely necessary to preserve the economic viability of a mining project.
On that basis, the Commonwealth submitted that any differential treatment or unequal outcome under the MRRT Act was the product of a distinction which was appropriate and adapted to the attainment of the objectives identified, each of which was a proper objective of the Parliament. The plaintiffs submitted, in effect, that such reasoning had no place in the characterisation of the MRRT Act as discriminatory or otherwise. If the law were unequally imposed it was prohibited by s 51(ii) regardless of the objectives.
[133]See above at [31].
It should be noted that although the Commonwealth put its argument on the hypothesis, which it denied, that the MRRT Act had a differential treatment or unequal outcome as between States the constitutional question is one of discrimination or preference. What the Commonwealth seemed to argue as a matter of confession and avoidance was in truth an aspect of characterisation of the MRRT Act for the purposes of ss 51(ii) and 99.
As explained earlier in these reasons, the constraints imposed by ss 51(ii) and 99 of the Constitution serve a federal purpose — the economic unity of the Commonwealth and the formal equality in the Federation of the States inter se and their people. Those high purposes are not defeated by uniform Commonwealth laws with respect to taxation or laws of trade, commerce or revenue which have different effects between one State and another because of their application to different circumstances or their interactions with different State legal regimes. Nor are those purposes defeated merely because a Commonwealth law includes provisions of general application allowing for different outcomes according to the existence or operation of a particular class of State law. A criterion for determining whether that category of Commonwealth law discriminates or gives a preference in the sense used in ss 51(ii) and 99 is whether the distinctions it makes are appropriate and adapted to a proper objective.
The Commonwealth Places (Mirror Taxes) Act 1998 (Cth) ("the Mirror Taxes Act") fell into the category just described, applying as it did the different tax laws of each State to Commonwealth places within that State. As this Court held in Permanent Trustee, s 51(ii) did not apply at all to the Act because it was a law made under s 52(i)[134]. As a law of revenue, however, the Act did attract the no‑preference limitation in s 99. On reasoning applicable to s 51(ii), the Court held that the Mirror Taxes Act did not give a preference to one State or any part thereof over another State or any part thereof. The majority said[135]:
"The scheme of the Mirror Taxes Act may produce differences in revenue outcomes between States, but that mirrors the differences that exist between the different taxation regimes from State to State. The differential treatment and unequal outcome that is involved here is the product of distinctions that are appropriate and adapted to a proper objective."
The objective of the impugned provision in that case was non-discriminatory. So too are the objectives of the impugned provisions of the MRRT Act. In general terms, they are those set out in the stated objectives of the Act referred to at the commencement of these reasons. The differences in the operation of the MRRT Act which arise out of its interaction with different royalty regimes serve those objectives. They are proper objectives, to which the impugned provisions are appropriate and adapted. The text, history, purpose and judicial exegesis of s 51(ii) require that the question whether the MRRT Act discriminates impermissibly be answered in the negative. It follows for reasons given earlier that the question whether the MRRT Act gives a preference contrary to s 99 is also to be answered in the negative.
[134](2004) 220 CLR 388 at 421 [79], applying Allders International Pty Ltd v Commissioner of State Revenue (Vict) (1996) 186 CLR 630 at 662, 678–680; [1996] HCA 58.
[135](2004) 220 CLR 388 at 425 [91] per Gleeson CJ, Gummow, Hayne, Callinan and Heydon JJ.
Conclusion
The questions reserved should be answered:
(i) No.
(ii) No.
(iii) The plaintiffs.
HAYNE, BELL AND KEANE JJ. The minerals resource rent tax ("MRRT") is imposed by the Minerals Resource Rent Tax (Imposition—Customs) Act 2012 (Cth), the Minerals Resource Rent Tax (Imposition—Excise) Act 2012 (Cth) and the Minerals Resource Rent Tax (Imposition—General) Act 2012 (Cth) (together "the Imposition Acts"). The assessment of the MRRT is provided for by the Minerals Resource Rent Tax Act 2012 (Cth) ("the MRRT Act").
In conformity with the intention declared in s 1‑10 of the MRRT Act to tax "above normal profits" from certain mining operations, MRRT is not exigible until a miner's group mining profit exceeds a prescribed threshold. Under the MRRT Act, a liability to pay MRRT arises only when a miner derives an annual profit of a given amount after taking into account all deductions for expenditure (including of capital), all allowances (including those carried forward at uplifted rates) and any applicable tax offsets. Once MRRT is payable, however, the formula by which its amount is calculated operates so that a reduction in the mining royalty payable to a State government would, other things being equal, result in an equivalent increase in the amount of the MRRT liability, and an increase in the royalty would, other things being equal, result in an equivalent decrease in the miner's MRRT liability. As it happens, State mining royalties differ between the States within the federation.
The plaintiffs, who are members of a group of companies which mine iron ore in Western Australia, brought proceedings in the original jurisdiction of this Court challenging the validity of the MRRT Act and of those provisions of the Imposition Acts which impose the tax. Pursuant to s 18 of the JudiciaryAct 1903 (Cth), questions were reserved for determination by the Full Court on the basis of the parties' pleadings and documents referred to in the pleadings.
The issues
The plaintiffs founded their challenge to the validity of the MRRT Act and s 3 of each of the Imposition Acts (together "the MRRT Legislation") principally on the ground that s 51(ii) of the Constitution expressly precludes the imposition by the Commonwealth of a tax which would exact a greater amount of tax from a taxpayer whose mining operations are conducted in a State with a lower mining royalty rate than would be exacted from the same miner if the same mining operations were conducted by it in a State with a higher State royalty rate. The plaintiffs also contended for the same result by invoking the constitutional implication associated with this Court's decision in Melbourne Corporation v The Commonwealth[136] and by reference to s 99 of the Constitution and its prohibition against the Commonwealth, by any law or regulation of trade, commerce or revenue, giving "preference to one State or any part thereof over another State or any part thereof". Finally, the plaintiffs argued that the MRRT Legislation is invalid because it is inconsistent with s 91 of the Constitution. The Attorneys‑General for the States of Queensland and Western Australia intervened to support the plaintiffs' challenge.
[136](1947) 74 CLR 31; [1947] HCA 26.
These reasons will demonstrate that the plaintiffs' challenge fails and the questions reserved should be answered accordingly. The reasons will first provide a summary of the relevant legislative provisions and then deal, in turn, with s 51(ii), s 99, the Melbourne Corporation principle and s 91.
The MRRT Legislation
Each of the Imposition Acts provided in s 3(1) that MRRT payable under the MRRT Act "is imposed". Section 4 of each of the Imposition Acts provided for an "MRRT rate" of 22.5 per cent. The Imposition Acts operated in the alternative to each other: see s 3(2). It was not disputed that the Minerals Resource Rent Tax (Imposition—General) Act 2012 was the relevant Imposition Act for present purposes.
The Revised Explanatory Memorandum to the Bills for the MRRT Legislation explained[137] that the MRRT is a tax on "economic rents", which constitute "the return in excess of what is needed [by miners engaged in extracting iron ore, coal and some gases from the ground] to attract and retain factors of production in the production process". It went on to explain[138] that "[a]s the MRRT taxes profits from minerals that are commonly subject to State and Territory royalties, it provides a credit for royalties".
[137]Australia, Senate, Minerals Resource Rent Tax Bill 2011, Minerals Resource Rent Tax (Consequential Amendments and Transitional Provisions) Bill 2011, Minerals Resource Rent Tax (Imposition—Customs) Bill 2011, Minerals Resource Rent Tax (Imposition—Excise) Bill 2011, Minerals Resource Rent Tax (Imposition—General) Bill 2011, Revised Explanatory Memorandum at 3.
[138]Australia, Senate, Minerals Resource Rent Tax Bill 2011, Minerals Resource Rent Tax (Consequential Amendments and Transitional Provisions) Bill 2011, Minerals Resource Rent Tax (Imposition—Customs) Bill 2011, Minerals Resource Rent Tax (Imposition—Excise) Bill 2011, Minerals Resource Rent Tax (Imposition—General) Bill 2011, Revised Explanatory Memorandum at 8 [1.25].
The MRRT Legislation is complex; and it is unnecessary to grapple with all of its complexities. It is sufficient for the purposes of this case to refer only to the central provisions that bear upon the calculation of the MRRT.
Calculating MRRT
Section 1‑10 of the MRRT Act provides that:
"The object of this Act is to ensure that the Australian community receives an adequate return for its taxable resources, having regard to:
(a)the inherent value of the resources; and
(b)the non-renewable nature of the resources; and
(c)the extent to which the resources are subject to Commonwealth, State and Territory royalties.
This Act does this by taxing above normal profits made by miners (also known as economic rents) that are reasonably attributable to the resources in the form and place they were in when extracted."
MRRT is payable for an "MRRT year"[139] by a miner in an amount equal to the sum of its MRRT liabilities for each of its "mining project interests"[140] for that year. Section 10‑1 of the MRRT Act provides that:
"A miner is liable to pay MRRT, for an MRRT year, equal to the sum of its MRRT liabilities for each of its mining project interests for that year."
Mining project interests are associated with "production rights" and, for present purposes, it is enough to notice that "production rights" include[141] extraction rights conferred by a State government in respect of a particular geographical part of the State. The mining project interests to which the MRRT Act applies are interests in relation to iron ore and coal (and some related substances), located in areas covered by a production right, which together are called "taxable resources"[142].
[139]Each MRRT year is a financial year and commences on 1 July: s 10‑25.
[140]This and other terms used in the MRRT Act are defined in the Dictionary set out in s 300‑1. For the most part, it is sufficient to indicate, by the use of quotation marks, that a term is defined in s 300‑1 without setting out the content of its definition.
[141]See ss 15‑5(2) and 15‑15 together with the definition of "Australian law" in s 300‑1 of the MRRT Act, which refers to the definition of that term in s 995‑1(1) of the Income Tax Assessment Act 1997 (Cth).
[142]ss 15‑5(4) and 20‑5.
Section 10‑5 of the MRRT Act provides that a miner's MRRT liability for a mining project interest for an MRRT year is to be worked out as follows: "MRRT liability = MRRT rate x (Mining profit – MRRT allowances)". Thus the amount of the MRRT liability for each mining project interest is calculated by subtracting from the "mining profit" certain "MRRT allowances". The sum so arrived at is then multiplied by the MRRT rate to establish the MRRT liability for each mining project interest.
A miner's "mining profit" is calculated[143] by deducting the miner's "mining expenditure" from its "mining revenue". The "mining revenue" for each mining project interest is determined in accordance with the provisions of Div 30 of the MRRT Act. The "mining expenditure" for each mining project interest is determined in accordance with Div 35 of the MRRT Act. The amounts to be deducted from mining revenue as mining expenditure do not include[144] "excluded expenditure". Mining royalties payable to a State are one form of "excluded expenditure"[145].
[143]s 25‑5.
[144]s 35-5(2). See also subdiv 35‑B.
[145]s 35‑40(1)(a).
If a miner's "group mining profit" for an MRRT year is less than $125 million, the miner is entitled[146] to an "offset" for that year. If the group mining profit is less than or equal to $75 million, the amount of the offset is the sum of the miner's MRRT liabilities for each mining project interest, with the consequence that no MRRT is payable[147]. If a miner's group mining profit is greater than $75 million, but less than $125 million, the amount of the offset is to be calculated in accordance with s 45‑10 and the miner will be liable to pay less than the amount that would be payable if MRRT at the rate of 22.5 per cent were to be applied to the full amount of the profit.
[146]ss 10‑15 and 45‑10.
[147]ss 10‑15 and 45‑5.
Taking account of royalties
Mining royalties are then included amongst the MRRT allowances which are to be deducted from the figure for mining profit. Royalty allowances appear as item 1 in the order of the allowances which are to be applied in the calculation of MRRT liability[305]. Other allowances include pre-mining and mining loss allowances, a starting base allowance, and transferred pre-mining and mining loss allowances.
[305]Minerals Resource Rent Tax Act 2012, s 10-10.
Chapter 3 deals with MRRT allowances and Pt 3-1, Div 60 with royalty allowances. As a guide to Div 60, it is said that "[m]ining royalties paid to the Commonwealth, States and Territories reduce a miner's MRRT liabilities for a mining project interest"[306]. The stated objects of the Division, by s 60-5, include reducing "a miner's MRRT liability relating to profits relating to taxable resources, to the extent those taxable resources are subject to Commonwealth, State and Territory royalties". A "mining royalty" is defined[307] as expenditure made under a Commonwealth, State or Territory law in relation to a taxable resource extracted under authority of a production right.
[306]Minerals Resource Rent Tax Act 2012, s 60-1.
[307]Minerals Resource Rent Tax Act 2012, ss 35-45(1), 300-1.
A miner has a royalty allowance for a mining project interest if the miner has a mining profit for that interest for the year and one or more "royalty credits" relating to the interest[308]. A royalty credit arises when a miner incurs a liability, inter alia, by way of a mining royalty[309]. A royalty credit may be transferred to the miner's other mining project interests and may be used in subsequent years[310].
[308]Minerals Resource Rent Tax Act 2012, ss 60-10, 300-1.
[309]Minerals Resource Rent Tax Act 2012, s 60-20(1)(a).
[310]Minerals Resource Rent Tax Act 2012, s 60-25(2).
Section 60-25 explains how a royalty credit is calculated. The amount is determined first by reference to the liability incurred for mining royalties. That liability is then divided by the MRRT rate. In the example given in the section, where a miner pays to a State mining royalties of $22.5 million in an MRRT year, the royalty credit is:
$22.5 million
= $100 million.
MRRT rate [22.5 per cent]
In summary, for every $22.5 million paid by a miner by way of mining royalties, a credit of $100 million is given. The royalty credit will thus be 4.4 recurring times each dollar paid, or otherwise incurred, by way of State mining royalties. As the note to s 60‑25(1) says, the calculation "grosses-up the royalty payment to an amount that will reduce the ultimate MRRT liability by the amount of the royalty payment". The plaintiffs' argument does not depend upon the provisions which have the effect of grossing up mining royalties. Their argument is the same regardless of those provisions. It centres upon full credit being given for mining royalties actually paid or incurred.
A miner's royalty allowance is so much of the royalty credits as does not exceed the mining profit[311]. The royalty allowance, together with all other allowances, is deducted from the figure for mining profit in accordance with the formula in s 10-5, which is set out above[312]. A miner's liability for MRRT is then determined by multiplying that figure by the MRRT rate. By way of example, if the mining profit is $500 million and the MRRT allowances are $200 million, a miner's liability will be:
$300 million × 22.5 per cent = $67.5 million.
[311]Minerals Resource Rent Tax Act 2012, s 60-15(1).
[312]See [188] above.
Because the calculation of a royalty credit, and therefore the royalty allowance, gives full credit for mining royalties in fact incurred by a miner, it follows that where the rate of royalty charged by one State varies from other States there will be differences in miners' liability for MRRT. The plaintiffs provided a series of equations transforming the formula in s 10-10 that they say demonstrate the impact of royalty allowances on the ultimate liability for MRRT. Those equations, which refer to a so-called "effective rate" of MRRT, are not of particular assistance to the issues before the Court and may be put to one side.
There will be other differences in liability for MRRT, as between miners generally and as between miners in different States, resulting from the other allowances provided for in the MRRT Act. Further, MRRT liability will differ as between miners because their mining expenditure, which is deducted from revenue to ascertain mining profit, will be different. Such expenditure may include State taxes and levies other than mining royalties, such as payroll tax, workers' compensation premiums and the like. These taxes and levies may also differ as between States.
Consideration of the s 51(ii) issue
Discrimination
Discrimination is a concept that arises for consideration in a variety of constitutional contexts[313]. Section 51(ii) prohibits a Commonwealth taxation law discriminating "between States or parts of States". The discrimination of which it speaks is discrimination on account of locality. Section 51(ii) requires that the States be treated alike and that a Commonwealth law relating to taxation not differentiate in its effect between the States.
[313]Bayside City Council v Telstra Corporation Ltd (2004) 216 CLR 595 at 629 [40]; [2004] HCA 19.
In R v Barger[314], Isaacs J said that "[d]iscrimination between localities in the widest sense means that, because one man or his property is in one locality, then, regardless of any other circumstance, he or it is to be treated differently from the man or similar property in another locality"[315]. Although his Honour was in dissent in Barger, with Higgins J, this view of s 51(ii) was subsequently cited with approval in Cameron v Deputy Federal Commissioner of Taxation[316], a case where a different standard was to be applied to the value of livestock solely by reference to "their State situation"[317].
[314](1908) 6 CLR 41.
[315]R v Barger (1908) 6 CLR 41 at 110.
[316](1923) 32 CLR 68 at 72 per Knox CJ, 79 per Rich J; [1923] HCA 4.
[317]Cameron v Deputy Federal Commissioner of Taxation (1923) 32 CLR 68 at 76 per Isaacs J.
Another statement by Isaacs J in Barger as to s 51(ii) is worthy of mention. It was referred to with approval by Evatt J in Deputy Federal Commissioner of Taxation (NSW) v W R Moran Pty Ltd[318] and on appeal by the Privy Council in that case[319]. Isaacs J said[320] that the "pervading idea" of the discrimination to which s 51(ii) refers is "the preference of locality merely because it is locality … It does not include a differentiation based on other considerations, which are dependent on natural or business circumstances". Although his Honour was speaking of the reference in s 51(ii) to "parts of States", what he said applies generally to the notion of discrimination with which s 51(ii) is concerned.
[318](1939) 61 CLR 735 at 781; [1939] HCA 27.
[319]W R Moran Pty Ltd v Deputy Federal Commissioner of Taxation (NSW) (1940) 63 CLR 338 at 348; [1940] AC 838 at 856-857.
[320]R v Barger (1908) 6 CLR 41 at 108.
Although discrimination can be an abstract concept, working out whether the effect of legislation is discriminatory is largely a practical question involving the consideration of unequal treatment[321]. It involves a comparison[322]. If a Commonwealth taxation law provides that the same measure is to apply to all persons or things subject to the tax, it would not generally be regarded as likely to discriminate in fact. Where a difference results from the operation of a taxation law, the question arises whether that difference is accounted for by the geographical situation of the subject of the tax. Importantly, for there to be the discrimination of which s 51(ii) speaks, the difference must be produced by the Commonwealth law itself and by reference to that geographical situation. There may not be discrimination where the difference results from the provisions of a State law. Section 51(ii) does not prohibit a taxation law from operating differentially in all respects. It does not require that a taxation law control the effect of other, external, factors which may be productive of a difference.
[321]Street v Queensland Bar Association (1989) 168 CLR 461 at 510; [1989] HCA 53.
[322]Street v Queensland Bar Association (1989) 168 CLR 461 at 506.
A general deduction?
The reference in Barger to "business circumstances"[323] brings to mind the possibility that the MRRT Act, in giving full credit for State mining royalties, does little more than permit a miner something in the nature of a deduction of a business expense from mining profits before those profits are subjected to taxation. As the Commonwealth points out, miners are able to deduct State imposts when calculating their liability for income tax. It is not suggested that Commonwealth income tax legislation, in the provision for general deductions which it allows for business expenses from income, operates so as to discriminate in any relevant respect. The allowance for mining royalties in the MRRT Act does not appear to be so different from deductions of this kind. In common with them, the allowance for mining royalties operates generally and does not discriminate between miners in different States. The allowance is provided whenever mining royalties are incurred by a miner, regardless of the miner's locality.
[323]See [201] above.
As the plaintiffs concede, their argument could not succeed had mining royalties been treated as mining expenditure under the MRRTAct. But they point out that the MRRT Act treats mining royalties differently from other mining expenditure. It expressly excludes them as items of mining expenditure. It further differentiates mining royalties from expenditure by allowing for them in full.
The State of Queensland, intervening in support of the plaintiffs, submits that the royalty allowance is to be distinguished from other, more general deductions because what the MRRT Act deducts is not the royalty paid, but a product of the operation of "grossing up" upon an operand. What is extracted by the division by the MRRT rate, in the formula in s 60-25, is not the amount of royalties paid, but the product of the calculation. This would seem to suggest that the royalty payment has, in the process, lost its quality as an item of business expenditure.
It must be accepted that the MRRT Act treats the payment of mining royalties separately, even from other mining allowances, in order that they may be "grossed up" and allowed for. Nevertheless, it is the payment of mining royalties upon which that calculation is based. Putting aside the inflated figure for royalty allowance, upon which the plaintiffs' argument does not rely, it is the payment of mining royalties for which full credit is given. The fact that it is allowed for in full is a distinction without a point. The MRRTAct plainly acknowledges mining royalties as a sum which is likely to have been paid by miners to a State in the course of mining operations. The relevant effect of the royalty allowance is to give credit for what has been paid. It applies whenever such an expense is incurred and regardless of where it is incurred.
The credit of royalty payment as a standard
On its face, the calculation provided for by ss 10-5 and 25-5 would appear to operate uniformly. MRRT liability is determined by first identifying mining revenue, and then deducting certain mining expenditure and mining allowances. A uniform MRRT rate is applied to the figure arrived at. But the plaintiffs submit that it would be wrong to think that, because the Imposition Acts provide for a rate of 22.5 per cent, MRRT is levied uniformly.
The plaintiffs contend that the legislation in Cameron is analogous in effect to the MRRT Act. The Income Tax Regulations 1917 (Cth) there considered provided that, so far as concerned profits made on the sale of livestock, different values were to be placed on stock of the same kind in different States. For example, a horse in New South Wales was to be valued at £8, but a horse in Victoria at £15 and in Queensland £4.
The Deputy Federal Commissioner of Taxation proffered the explanation that the standard adopted was not arbitrary, but the actual average value of livestock in each State, which was merely recognised and enforced by the Regulations as a convenient and just method of valuing stock[324]. That explanation was rejected. The reasons of Isaacs J[325] disclose that the value attributed to stock according to the Regulations was not in fact a "fair average value". Previously, the value of stock was to be that as determined by the Commissioner, but an amendment to the Regulations specified set values in a schedule[326]. Had the Commissioner retained discretion to make the determination, the Commissioner would have had to take account of values on either side of the relevant State borders. What the Regulations produced was a value rigidly fixed for the State in which the stock was located. Thus a horse in Albury was worth £8, whereas the same horse, located across the river at Wodonga, was to be valued at £15.
[324]Cameron v Deputy Federal Commissioner of Taxation (1923) 32 CLR 68 at 73-74.
[325]Cameron v Deputy Federal Commissioner of Taxation (1923) 32 CLR 68 at 74-77.
[326]Statutory Rules 1918, No 315.
Different standards were applied to different States by the Regulations in Cameron[327]. This was the source of the discrimination. What was produced by the Regulations was a standard which was identified not by reference to value in fact, but by reference to locality. Unsurprisingly, the Regulations were held to offend s 51(ii). As was pointed out by Isaacs J[328], the only discrimen provided by the Regulations was "which State?" Cameron provides an unusually clear example of s 51(ii) discrimination.
[327](1923) 32 CLR 68 at 77.
[328]Cameron v Deputy Federal Commissioner of Taxation (1923) 32 CLR 68 at 73; see also at 72 per Knox CJ.
The royalty allowance provided for in the MRRT calculation does not operate in this way. The standard it applies is the actual amount of mining royalties which a miner has incurred. In contrast to the facts in Cameron, any variation in MRRT payable from miner to miner results from that fact and not from any State-based standard applied by the MRRT Act. The only causal connection between the royalty allowance and a State is that mining royalties are only incurred by a miner because of a State law. The MRRT Act says nothing about the quantum of mining royalties except that they are to be allowed in full when incurred. This tells against notions of discrimination.
The cause of the difference?
In Colonial Sugar Refining Co Ltd v Irving[329] ("Irving") and in Barger, it was acknowledged that duties and levies in different parts of the Commonwealth would produce a differential effect for a Commonwealth law. It had been accepted by the framers of the Constitution that taxation may produce an indirect effect which was not uniform[330].
[329][1906] AC 360.
[330]R v Barger (1908) 6 CLR 41 at 69-70.
The Excise Tariff 1902 (Cth), considered in Irving, exempted from the duties thereby imposed, goods upon which State customs duties had already been paid. However, the scale of duties differed as between the States so that the exemption operated unequally. An analogy with the provisions of the MRRT Act in this case is evident. Lord Davey, speaking for the Judicial Committee of the Privy Council, said[331]:
"The rule laid down … is a general one, applicable to all the States alike, and the fact that it operates unequally in the several States arises not from anything done by the Parliament, but from the inequality of the duties imposed by the States themselves."
[331]Colonial Sugar Refining Co Ltd v Irving [1906] AC 360 at 367, cited with approval in Permanent Trustee Australia Ltd v Commissioner of State Revenue (Vict) (2004) 220 CLR 388 at 434 [127]-[128]; [2004] HCA 53.
A different view was taken of a similar exemption by the majority in Barger and this drew strong dissents from Isaacs and Higgins JJ. Barger is also authority for a view concerning powers reserved to the States which has long since been discredited[332]. This aspect of the case may be put to one side.
[332]See Permanent Trustee Australia Ltd v Commissioner of State Revenue (Vict) (2004) 220 CLR 388 at 434 [129].
Barger was concerned with the Excise Tariff 1906 (Cth), which imposed duties of excise upon specified goods at specified rates. However, the tariff did not apply to goods manufactured by any person in any part of the Commonwealth under conditions of remuneration of labour which satisfied any one of four prescribed matters. The majority held there to be discrimination in the taxing scheme by reference to the criterion of locality because of the possibility that goods of the same class would be excisable in some parts of the Commonwealth, but not others[333]. Isaacs J saw the exemption as a general rule operating unequally only because of the inequality of industrial circumstances[334]. There may be something to be said for the view[335] that the approach of Isaacs J in Barger is more consonant with the decisions in Cameron and in Irving. It is not necessary to resolve the differences of opinion in Barger for the purposes of this matter.
[333]R v Barger (1908) 6 CLR 41 at 80.
[334]R v Barger (1908) 6 CLR 41 at 110-111.
[335]Deputy Federal Commissioner of Taxation (NSW) v W R MoranPty Ltd (1939) 61 CLR 735 at 780-781 per Evatt J (in dissent).
In Conroy v Carter[336], Taylor J, with whom Barwick CJ, McTiernan, Kitto, Menzies and Windeyer JJ relevantly agreed, after referring to decisions concerning the United States Constitution and its command of geographical uniformity, said that it was not necessary, for a tax to be lawful by reference to s 51(ii), that it select objects which exist uniformly in all States. A law cannot, in general, be said to discriminate if its operation is general throughout the Commonwealth even if, by reason of circumstances existing in one or more States, it may not operate uniformly. This reflects the view expressed in Irving. Taylor J gave the example of a State land tax, which Commonwealth income tax legislation allows as a deduction from income in the relevant year. That provision for deduction, his Honour observed, operates generally throughout the Commonwealth. The fact that in some States there may be no legislation imposing land tax does not mean that the Commonwealth income tax legislation discriminates between the States.
[336](1968) 118 CLR 90 at 100-101; [1968] HCA 39.
In the decision of the Full Court of the Supreme Court of Queensland in The Colonial Sugar Refining Co Ltd v Irving[337], Griffith CJ concluded that "[i]f the imposition of these duties leads to an inequality, it is not a defect in the Federal law; it arises from the fact that the laws of the States were different, which is quite another thing". His Honour also observed that, were inequality to be viewed by reference to the operation of State law, the power of the federal Parliament would be limited by the laws of the States and by the mode in which the States had exercised their legislative powers. These observations point up the difficulties inherent in the plaintiffs' argument, which identifies differences in State laws as relevant to Commonwealth laws, particularly given the supremacy of the latter by reason of s 109 of the Constitution.
[337][1903] St R Qd 261 at 277.
The plaintiffs argue that the MRRT Act cannot be viewed in the same way as the legislation in the abovementioned cases because the MRRT Act itself provides for differential rates. It makes State mining royalties incurred an essential integer in the calculation of a miner's ultimate liability for MRRT. The calculation of royalty allowance is critical to that liability. The importance of this aspect of the plaintiffs' argument may be seen from their statements that: the discrimination resides in the calculation; the MRRT Act is discriminatory because the royalty allowance is the basis of the Act's structure; and "the tax is one calculated directly by reference to the amount of the royalty".
In substance, payment of mining royalties to a State is treated no differently from any other allowance or deduction by the MRRT Act, save in the respects previously mentioned. Those differences do not detract from the fact that mining royalties incurred are an amount for which credit is given and which reduces the amount of mining profits to be subjected to the MRRT rate. Mining royalties are essential to the calculation of a miner's ultimate MRRT liability in the same way as are other items of mining expenditure and mining allowances.
An equalised tax burden?
The State of Queensland contends that the MRRT Act seeks to bring about equality between miners in different States. It was the object of the MRRT Act to equalise a miner's overall tax burden. This can be seen by it operating so as to increase the MRRT liability in States with lower royalty rates, and vice versa.
In support of that argument, reliance is placed upon a statement made by Griffith CJ for the majority in Barger. After referring to what had been said in Irving, that it was the effect of the State, not the Commonwealth, laws that created the unequal burden, his Honour said[338]:
"E converso, if the Excise duty had been made to vary in inverse proportion to the Customs duties in the several States so as to make the actual incidence of the burden practically equal, that would have been a violation of the rule of uniformity."
[338]R v Barger (1908) 6 CLR 41 at 70-71.
The MRRT Act does not operate as does the hypothetical law referred to in Barger. The law to which Griffith CJ referred is a law which itself adjusts according to the amount of State duties paid, so that the overall amount of Commonwealth and State taxes is equalised. By way of example, if mining royalties of $2 were paid by a miner in one State and $4 by a miner in another, a Commonwealth law would operate in the way contemplated by Griffith CJ if it provided that the firstmentioned miner pay Commonwealth tax of $4 and the second $2. The MRRTAct does not operate in this way. It is not structured to ameliorate the effect of the State mining royalties for miners, but rather makes provision for miners' business circumstances, which may be affected by various State laws. It does not breach the constitutional prohibition in s 51(ii).
Conclusion on the MRRT Act and s 51(ii)
In Conroy v Carter, Menzies J[339], with whom Barwick CJ and McTiernan J agreed, spoke of the discrimination to which s 51(ii) refers in the context of a taxation law which imposes a taxation burden. It may be accepted that s 51(ii) also prohibits a benefit which applies differentially as between the States. Expressing what his Honour said more generally, s 51(ii) forbids a taxation law which operates to benefit or burden a person because of some connection with a State, but which would not be granted to or imposed on other persons not having that connection. In determining whether a law operates in that way, it is to the Commonwealth law itself that attention is directed.
[339]Conroy v Carter (1968) 118 CLR 90 at 103.
The MRRT Act provides generally for a royalty allowance, the calculation of which includes a credit for the whole amount incurred by a miner by way of mining royalties paid to a State. There is no standard of locality, of connection to a State, in the allowance made and in the deduction for which it provides. The standard is the fact and amount of payment. Any difference in the amount of the deduction for mining royalties results not from the MRRT Act but from the State legislation.
Those who drafted the MRRT Act may be taken to have been aware that rates of mining royalties differ as between the States. The point, however, is not that there is some underlying assumption of difference on which the MRRT Act operates, as the plaintiffs and the State of Queensland suggest, but rather that the MRRT Act allows for whatever mining royalties are required to be paid under State legislation. It is not the Commonwealth Act that creates any inequality or difference, but State legislation. The Commonwealth is entitled to do what the States do and base its taxation measures on considerations of fairness, so long as it adheres to the constitutional injunction not to prefer States[340].
[340]R v Barger (1908) 6 CLR 41 at 108; Deputy Federal Commissioner of Taxation (NSW) v W R Moran Pty Ltd (1939) 61 CLR 735 at 781-782; W R Moran Pty Ltd v Deputy Federal Commissioner of Taxation (NSW) (1940) 63 CLR 338 at 348; [1940] AC 838 at 856-857.
A State royalty is treated by the MRRT legislation as an amount which is likely to have been incurred by a miner in connection with its mining activities. A miner's MRRT liability will be affected by the expenses which it incurs, the other allowances for which the MRRT Act provides and whether, in a given MRRT year, a royalty credit has been transferred or carried over. This brings to mind what was said by Griffith CJ in the Supreme Court of Queensland in The Colonial Sugar Refining Co Ltd v Irving[341], that the difference effected by the Excise Tariff 1902 was not discrimination created by Commonwealth law, but "a difference in the individual incidence of taxation".
[341][1903] St R Qd 261 at 276.
The MRRT legislation does not discriminate between States and does not create a preference for one over another.
The Melbourne Corporation doctrine
The plaintiffs' claim that the MRRT legislation affects a State's capacity to control its sovereign territory and to deal with its natural resources, and Western Australia's submissions to similar effect, appear to reflect arguments which were put by Western Australia, and which were rejected, in Western Australia v The Commonwealth (Native Title Act Case)[342], as Hayne, Bell and Keane JJ observe in their reasons.
[342](1995) 183 CLR 373 at 478-479, 481; [1995] HCA 47.
The MRRT legislation is not directed to the States and does not affect the government of a State. It does not deny the ability of a State to fix a rate of mining royalty. Any effect upon a State's ability to offer incentives, by reducing that rate or providing an exemption, is not a burden or limit respecting a State's constitutional functions. I agree with the reasons of Hayne, Bell and Keane JJ on this issue.
Section 91
Section 91[343], the plaintiffs submit, takes effect as a prohibition directed to any law made under a head of power in the Constitution which may hinder a State from providing the abovementioned incentives to a miner, as an encouragement to mining activity. In this regard, the plaintiffs rely upon the express provision, made in s 91, that "[n]othing in this Constitution prohibits a State from granting any aid".
[343]See [185], fn 296 above.
Section 91 must be read with s 90[344]. In Seamen's Union of Australia v Utah Development Co[345], the only decision of this Court which has been concerned with s 91, this Court discussed the relationship between ss 90 and 91. Mason J, with whom Jacobs and Aickin JJ agreed, observed[346] that s 90 contains a prohibition which arises from the exclusive conferral on the Commonwealth Parliament of a power to impose duties of customs and excise and to grant bounties on the production and export of goods. The function of s 91 was said to relax that prohibition. The words "[n]othing in this Constitution" were held to refer back primarily, if not exclusively, to s 90 because there was no other provision in the Constitution which contained a relevant prohibition.
[344]Section 90 of the Constitution provides:
"On the imposition of uniform duties of customs the power of the Parliament to impose duties of customs and of excise, and to grant bounties on the production or export of goods, shall become exclusive.
On the imposition of uniform duties of customs all laws of the several States imposing duties of customs or of excise, or offering bounties on the production or export of goods, shall cease to have effect, but any grant of or agreement for any such bounty lawfully made by or under the authority of the Government of any State shall be taken to be good if made before the thirtieth day of June, one thousand eight hundred and ninety‑eight, and not otherwise."
[345](1978) 144 CLR 120; [1978] HCA 46.
[346]Seamen's Union of Australia v Utah Development Co (1978) 144 CLR 120 at 147.
What was said in Seamen's Union provides the answer to the plaintiffs' argument. Contrary to the plaintiffs' contention, that case did decide that the purpose of s 91 is to qualify the prohibition in s 90. Section 91 does not itself operate as a prohibition on Commonwealth laws. Section 91 confirms that a State may grant aid, which is to say a State Parliament may authorise expenditure by this means; it does not speak of how Commonwealth laws might interact with that grant.
It is therefore not strictly necessary to point out that the plaintiffs do not refer to "aid" in the nature of a parliamentary grant of money, which is the sense in which it appears to have been understood in Seamen's Union[347]. Stephen J, in particular, appears to have taken up the argument put by the State of Queensland there, that aid refers to a pecuniary payment authorised by Parliament. His Honour said that history supports a narrower view of aid than as general assistance[348]. On that view, the incentives to which the plaintiffs refer would not amount to aid within the meaning of s 91.
[347]See (1978) 144 CLR 120 at 126 per Barwick CJ, 135 per Gibbs J, 140 per Stephen J, 148 per Mason J.
[348]Seamen's Union of Australia v Utah Development Co (1978) 144 CLR 120 at 140.
Conclusion and orders
I agree with the orders proposed by Hayne, Bell and Keane JJ.