University of New South Wales Law Journal

SURI RATNAPALA[*]

[1] The idea of sovereignty persists in political discourse against all odds. It is difficult to define, and when defined, is hard to find in real life. The term is employed alternatively to signify the position of legal independence of states at international law (national sovereignty), the political supremacy of the people (popular sovereignty) and the attribute of a limitlessly empowered supreme authority within a national legal system (constitutional sovereignty). Sovereignty is an ideal and like all ideals is extraordinarily hard to achieve. National sovereignty is limited by overarching international law. Popular sovereignty, even when identified with majority will, is routinely subverted by elected and unelected governments. Constitutional sovereignty, as Professor Hart argued,[1] does not exist even in the United Kingdom whose Parliament was once thought to possess absolute power that included the power to pass Bills of Attainder and of Pains and Penalties. This paper focuses on sovereignty in the third or constitutional sense.

[2] Though sovereignty in the constitutional sense remains an ideal notion, it is one that perennially attracts the attention of governments, judges and significant sections of the public; particularly in times of social or economic upheaval, but also in times of peace and prosperity. The driving sentiment behind sovereignty in the constitutional sense is the belief that governments, particularly those responsible to the electorate, must not be restrained in the pursuit of the public good. It is the conviction that John Stuart Mill viewed with apprehension in his work On Liberty, that ‘the nation need not be protected against its own will’.[2] It is an idea that dominated the middle span of the 20th century. The ideal of sovereignty stands in contrast to the classical republican mistrust of absolute power even in the hands of popular assemblies. Republicans believe that even democracy must be tempered by constitutional restraints if the public good is to be advanced and private vice suppressed. Like sovereignty, the republican notion of limited and divided power is a compromised ideal. The desire for public goods such as the provision of personal security, national defence, essential infrastructure and social security generate demands for greater governmental powers. Thus, the paradigms of sovereignty and of republican limited government are in constant tension in a game of constitutional ‘tug-of-war’.

[3] Australia is a federation of States that formerly were colonies within the British Empire. The six States were endowed with very similar constitutions by the British Parliament. State Parliaments were granted plenary powers limited only by territorial boundaries and British laws that extended expressly or impliedly to the States by paramount force or which overrode State laws on the basis of repugnancy to fundamental rules of common law. The Colonial Laws Validity Act 1865 (Imp) conferred on the colonial legislatures the power to impose upon themselves ‘manner and form’ limitations with respect to the alteration of laws concerning the constitution, powers and procedure of such legislatures. On Federation, the Australian Constitution (‘Constitution’) imposed further limits on the legislative powers of the States. State constitutions continued subject to the Constitution (s 107) and State laws yielded to contrary Commonwealth laws (s 109).

[4] The Commonwealth arms of government are themselves limited in their powers. The Constitution imposes an incomplete but substantial division of powers among the judicial, executive and legislative branches. It disperses power territorially by the preservation of the States and the specification of the Commonwealth’s legislative powers. It imposes explicit limits on legislative power with respect to property takings, jury trials, inter-State trade, discrimination among States and state sponsorship of religion. Recently, the High Court of Australia (‘High Court’) has added a range of implied limitations on Commonwealth power derived from the liberal or republican orientation of the Constitution.

[5] It may seem, therefore, that the Commonwealth and State Constitutions contain in-built breaks against gravitation to sovereignty. However, constitutions in the living sense are not determined solely by texts. In the absence of transcendentally valid meanings, texts are but constructions of human minds and the barriers to sovereignty that the constitutional texts erect are only as strong as judges construe them to be. These constructions, in turn, are influenced by the judges’ own experiences, intellectual predilections and the ideological currents that run through political discourse. These comments are not meant as a criticism of judges, but as a reflection on the nature of their role, and an attempt to place in social context the constitutional fluctuations for which they bear formal responsibility.

[6] Constitutions tend to oscillate between the sovereignty model and the republican model of limited powers without ever reaching either end. The High Court’s interpretation of the Constitution in its first century of operation reflected this pattern. The jurisprudence of the early High Court with its strong federalist agenda of vertically separated powers was displaced, in 1920, by the doctrine formulated in Amalgamated Society of Engineers v Adelaide Steamship Co Ltd (‘Engineers’ Case’).[3] This doctrine favoured the literal and expansive reading of Commonwealth legislative powers subject only to express limitations found in the constitutional text. There was no place within this doctrinal model for implied limitations drawn from the structure or basic values of the Constitution. The Engineers’ Case dominated the High Court’s treatment of the Constitution for much of the century, before the doctrine began to unravel owing to its own contradictions. That unravelling resulted in the heightened commitment to republican constitutionalism of the Mason and Brennan Courts in the last decade of the century. In that period, the High Court time and again departed from the Engineers’ doctrine in order to defend the structure and basic values of the Constitution from intended and unintended consequences of executive and legislative action. These judicial efforts generated a constitutional model that recognises limits on Commonwealth legislative powers drawn from not only express provisions but also from implications of the Constitution.

[7] The Constitution began its existence in the care of a High Court that was composed of two of its principal drafters, Sir Samuel Griffith and Sir Edmund Barton, and another ‘founding father’, Justice O’Connor. The interpretation of the Constitution by the early High Court naturally reflected the model of the Constitution that these authors set out to create, namely one of federal balance in which the States and the Commonwealth had their respective spheres of autonomy secured by the Constitution with neither entity being superior to the other. The Justices did not seek to defeat unambiguous provisions of the Constitution in favour of an ideal balance, but rather sought, where possible, to interpret provisions in the light of a federal arrangement that they considered was clearly intended. The model generated two doctrines: that of implied governmental immunities and that of the reserved powers of the States.

[8] The doctrine of implied governmental immunities prohibited the Commonwealth and the States imposing upon each others’ agents and instrumentalities burdens that fetter, control or interfere with the free exercise of legislative or executive power unless expressly authorised by the Constitution. The kernel of the doctrine of reserved powers was the proposition that where legislative power was not clearly granted to the Commonwealth, it belonged by necessary implication to the States. The reasoning of the reserved powers doctrine is typical of constitutional interpretation engaged in by courts having judicial review power. The reasoning does not negate clear words that derogate from a desired model of the Constitution. Rather, it brings into play wider policy and theoretical considerations, where words permit a construction that avoids violence to some basic value or structural feature of the Constitution. This much is apparent from Chief Justice Griffith’s exposition of the doctrine in Peterswald v Bartley, where he stated that a construction that allows the Commonwealth Parliament to interfere in the internal affairs of a State ‘will not be accepted by this court unless the plain words of provisions compel us to do so’.[4]

[9] The twin federalist doctrines of the early Court combined with the implications of the tripartite division of judicial, executive and legislative powers and the express limitations of the Constitution yielded a model of limited powers approximating to the classical republican ideal. However, in the Engineers’ Case, the High Court rejected this approach in favour of a doctrine purportedly based in strict legalism.

[10] The Engineers’ Case considered whether the Commonwealth Parliament’s power under s 51(xxxv) of the Constitution, with respect to conciliation and arbitration of industrial disputes extending beyond the limits of any one State, authorised it to make provisions for the making of arbitral awards binding on the agencies of a State government. Western Australia argued that it did not. The Court, by a majority of five to one (Gavan Duffy J dissenting), not only overthrew the implied governmental immunities doctrine, but also the reserved powers doctrine that was not really in issue before the Court. More importantly, the judgment articulated a new doctrine concerning the interpretation of the Constitution that became a principal dogma of the Court. The main joint judgement of Knox CJ, Isaacs, Rich and Starke JJ, reputed to have been authored by Isaacs J, was also supported by a separate opinion of Higgins J. With respect to s 51, the Engineers’ Case generated the following principles:
(a) Each paragraph of s 51 must be given its widest literal meaning.
(b) Each paragraph is declared by s 51 to be ‘subject to this Constitution’. These words only refer to clear words of the Constitution that limit the power granted in the paragraphs of s 51. It is not permissible to read down a paragraph of s 51 on the grounds that some fundamental feature or basic value of the Constitution is subverted by giving the paragraph its full literal effect.
(c) Under this approach, the Court would not recognise any implied limitations on Commonwealth power derived from such features of the Constitution as the federal arrangement, the representative principle and the separation of powers. The philosophy of the Griffith Court regarded such implications as part of ‘this Constitution’ within the meaning of the opening words of s 51. Herein lies the key departure effected by the Engineers’ Case.

[11] The policy of literal and expansive interpretation proclaimed by the Engineers’ Case has been applied primarily in relation to the empowering clauses of the Constitution. The Court has forsaken literalism when construing provisions that limited legislative powers. This divergence is best exemplified by the Court’s interpretation of the words ‘trade, commerce and intercourse among the States ... shall be absolutely free’ in s 92 of the Constitution. These words when literally interpreted maintained the so-called ‘individual rights theory’ of the freedom that postulated an individual’s freedom to engage in inter-State trade free from government control. This approach was rejected in a series of cases commencing with R v Vizzard; Ex parte Hill[5] in favour of the so-called ‘free trade theory’ that sustained legislation limiting the freedom, provided it was not protectionist in character as between the States. Similarly, the majority in Attorney-General (Cth); Ex Rel McKinlay found the requirement in s 24 of the Constitution, that the members of the House of Representatives shall be ‘directly chosen by the people of the Commonwealth’, insufficient to prevent Parliament from grossly distorting vote value or from denying suffrage if it chose to do so.[6] Thus, the overall effect of the Engineers’ doctrine was to drastically reduce the restrictive potential of the Constitution, thereby shifting it unmistakeably towards the sovereignty model.

[12] Legislative powers conferred by a written constitution are limited in two ways. First, there are intrinsic limitations in the empowering language. For example, language that confers power to make laws with respect to inter-State trade does not confer power to make laws with respect to intra-State trade. If the latter power exists, its source is elsewhere in the Constitution. Second, there are extrinsic limitations on power. Even if the conventional meaning of language taken by itself confers power over a subject, that power may be curtailed by the force of other provisions of the Constitution. The Engineers’ doctrine weakened the extrinsic limitations by rejecting all but express limitations on empowering clauses. It also whittled away the intrinsic limitations.

[13] The intrinsic limitations on empowering clauses were loosened by the Engineers’ doctrine with respect to: (i) nexus, (ii) purpose and (iii) proportion. This policy, when strictly applied, produced major accretions of power to the Commonwealth. The Engineers’ doctrine, while eliminating implied limits on Commonwealth power, allowed the Commonwealth to extend itself to matters over which it has no express constitutional authority. The literalism that the Engineers’ doctrine demanded led the High Court to uphold Commonwealth laws that covered fields left to the States provided that such laws also dealt with matters within one or more heads of Commonwealth power. Laws often fall within more than one subject and sometimes they concern both subjects within and outside Commonwealth jurisdiction. The fact that the law was mainly concerned with a subject denied to the Commonwealth and that it had only a partial nexus to a Commonwealth power was held not to be a ground of invalidity. As Stephen J stated in Actors and Announcers Equity Association v Fontana Films Pty Ltd, the position was that

[14] According to his Honour, constitutional validity did not require that the law related to a subject within Commonwealth legislative power by virtue of its predominant character.[8]

[15] The Commonwealth’s legislative power under s 51(xxxv), with respect to the conciliation and settlement of industrial disputes, encompasses only those disputes that extend beyond the limits of any one State. However, in Huddart Parker Ltd v Commonwealth,[9] the Court used the Engineers’ doctrine to allow the Commonwealth to regulate waterside employment generally by resort to its powers under s 51(i), with respect to inter-State and overseas trade and commerce. From the literalist viewpoint, the Court had no difficulty in construing the labour law initiative as falling within the subject of inter-State and overseas trade as waterside work invariably impacted on trade. Similarly, the Court used the Engineers’ doctrine to allow the Commonwealth to promote conservation goals,[10] to control the investments of superannuation funds,[11] and to regulate intra-State trade.[12] In all these cases, the reach by the Commonwealth into what would otherwise be State territory was not incidental but manifestly purposeful. Yet, under the Engineers’ doctrine, the Court regarded purpose as irrelevant to the classification of the law. As Dixon J stated in Melbourne Corporation v Commonwealth (‘Melbourne Corporation Case’):

[16] Connected with the refusal to question purpose was the Court’s reluctance to question the proportionality of means to ends. Under the British sovereignty model, courts show great vigilance in controlling the abuse of powers by officials acting under the authority of Acts of Parliament, but they leave to the political process the task of checking abuses of legislative power by Parliament itself. The Justices in the Engineers’ Case followed the British tradition in rejecting, as a ground for limiting power, the possibility of its abuse. Under the influence of Engineers’ doctrine, the High Court repeatedly declined to examine the nexus between laws and subjects of power and thereby left to the Commonwealth a sizable discretion to determine the limits of its own legislative power. The Court’s position was articulated by Kitto J in Herald and Weekly Times Ltd v Commonwealth with the statement:

[17] The reluctance to question purpose and proportion grew exponentially when applied to s 51(xxxix), which allows Parliament to legislate with respect to ‘matters incidental’ to its other constitutional powers and functions and those of the executive and judicial branches of government.

[18] It became clear that the Engineers’ doctrine, when applied strictly, allows the Commonwealth to reach almost any subject. Chief Justice Latham perceived this danger when he observed in Bank of New South Wales v Commonwealth (‘Bank Nationalisation Case’):

[19] If taxation was one method by which the Commonwealth could extend its powers, licensing was another. The Commonwealth could prohibit inter-State and international trading except under licence conditioned to the observance by the licencee of rules prescribed with respect to matters that lay outside the Commonwealth’s legislative power. In Herald Weekly Times v Commonwealth, Kitto J declared that:

[20] Perhaps the greatest expansion of Commonwealth power through the application of the Engineers’ doctrine occurred as a consequence of the High Court’s literal reading of the two words constituting s 51(xxix), namely, ‘external affairs’. The High Court’s expansive construction has handed the Commonwealth executive an extraordinary means of conferring upon the Commonwealth Parliament legislative power over virtually any subject, by assuming treaty obligations under international law with respect to such subject. In a series of cases beginning with R v Burgess; Ex parte Henry[17] and continuing through Koowarta v Bjelke-Peterson[18] and Commonwealth v Tasmania (‘Tasmanian Dams Case’),[19] the High Court established the power of the Commonwealth Parliament to make law for the implementation of treaties even when the subject matter of the treaty was one in respect of which the Constitution did not otherwise confer power. A majority of the judges in the Tasmanian Dams Case (Murphy, Mason, Brennan and Deane JJ) took the view that the power arose in relation to any treaty obligation assumed bona fide under international law, rejecting the minority position that the power existed only in relation to treaties dealing with matters of international concern. The majority acknowledged that the Commonwealth cannot use a treaty as a device to create legislative power,[20] but nevertheless refused to set a substantive limit on subject matter.

[21] The Engineers’ doctrine continued to be influential sporadically, even in the later part of the 20th century, though it was already unravelling as a result of its internal contradictions. The 1975 case of Attorney-General (Cth); Ex rel McKinlay[21] provides a good example. The High Court was invited to consider whether Parliament’s capacity to take away adult suffrage or to depart from the ‘one vote one value’ principle was limited by the requirements in ss 7 and 24 that the Senate and the House of Representatives be ‘directly chosen by the people’. Only the equal vote value question was directly before the Court, but all the judges found the question concerning franchise sufficiently relevant to be explicitly addressed. The majority, comprising Barwick CJ and Gibbs, Stephen and Mason JJ, found no such limitation in the words of ss 7 and 24 or in the democratic structure of the Constitution. A major theme in the majority’s reasoning was that in the absence of clear limiting words, these were matters left to Parliament’s discretion as in the case of Britain’s sovereign Parliament. It was even suggested by two of the Justices that Parliament could deny suffrage on the grounds of gender, race or lack of property.[22]

[22] It was clear from the outset that the Engineers’ doctrine was unsustainable as a consistent principle of constitutional interpretation. It is at odds with the nature of a written and entrenched constitution. While there was certainly a case for tightening the doctrines of implied immunities and of reserved powers, there was no real prospect of the Court disregarding, in the longer term, the fundamental features of the Constitution. To do so is to expose the Constitution to legislative dismantling and ultimate destruction. With respect to the doctrine’s effect on the federal structure, Professor Zines states that it is

[23] Even in its efforts to enunciate the doctrine, the joint judgment was compelled to resort to two non-explicit features of the Constitution; namely, the institution of responsible government and the indivisibility of the Crown. These two features were used by the Justices to repudiate American authorities supporting the doctrines of implied immunities and reserved powers.[24]

[24] The Engineers’ doctrine subjects the Commonwealth’s powers only to the clearly expressed limitations set out in the Constitution. The reality is that it is impossible in a constitution to set out in express terms all the rules necessary to maintain the essential features of the Constitution – such as the federal arrangement, separation of powers, the representative principle, responsible government, due process, and free speech. Hence, it was inevitable that the High Court would depart from the Engineers’ doctrine in substance, if not in form, when faced with serious threats to the Constitution it has sworn to uphold. The presently discussed landmark cases lead us to expect that when the Constitution is imperilled, the Court will read down empowering clauses by subordinating them not only to express limitations but also to limitations implied from the basic features of the Constitution.

[25] The most direct derogation from the Engineers’ doctrine occurred in the Melbourne Corporation Case, which considered the constitutionality of Commonwealth legislation to compel the States and their instrumentalities to bank their funds exclusively with the fully Commonwealth owned Commonwealth Bank, and prohibit all other banks from conducting any business with the States. The Court determined that a law which is otherwise within a head of power granted to the Commonwealth by s 51 would yet be unconstitutional if it singled out the States for discriminatory treatment. Much was made of the fact that a law that had as its object the discriminatory treatment of the States was tenuously connected to the subject of the enabling paragraph, in this case paragraph (xiii), relating to ‘banking other than State banking’. However, the law would have been valid under the Engineers’ doctrine as the law was unequivocally one with respect to banking other than State banking and the plain meaning of the paragraph did not preclude the regulation of banking business in relation to the States and their instrumentalities and the law. The High Court was compelled to rely on the nature of federalism to establish the rule that the Commonwealth cannot legislate to discriminate against a State. Justice Dixon maintained that despite the ‘complete overthrow of the general doctrine of reciprocal immunity of government agencies’ it has never been countenanced that ‘the legislative powers of one government in the system can be used in order directly to deprive another government of powers or authority committed to it or restrict that government in their exercise’.[25] The majority judgment in the Melbourne Corporation Case contained the tacit admission that the Engineers’ doctrine had the potential to damage the Constitution if it did not yield to the fundamental features of the Constitution. The case revived in substance the idea of a province of State jurisdiction that cannot be transgressed without express constitutional authority. The fact that the judges will not call it reciprocal immunity or reserved powers is of little moment.

[26] In Queensland Electricity Commission v Commonwealth,[26] the High Court derived two prohibitions from the reasoning in the Melbourne Corporation Case. Justice Mason articulated them as:

[27] Justice Deane added that the Court would look behind ‘ingenious expression or outward form’ to determine whether ‘as a matter of substance, the actual operation of the law is to discriminate against the States’.[28]

[28] Some Justices of the High Court have been reluctant to formally concede the demise of the Engineers’ doctrine as an interpretive model. In Victoria v Commonwealth (‘Payroll Tax Case’), the decision that upheld the imposition of payroll tax on employers including State instrumentalities, Barwick CJ sought to explain the rule in the Melbourne Corporation Case not on the basis of federal implications but on the ground that the ‘topics of legislation allotted to the Commonwealth do not include the States themselves nor their governmental powers or functions as a subject matter of legislative power’.[29] This rationale was adopted by Windeyer and Rich JJ.[30] This is at best a spurious distinction. The Commonwealth’s taxation power, granted in s 51(ii), is expressly qualified only by the prohibition against discrimination ‘between States or parts of States’ (emphasis added). Subject to this qualification, the plain meaning of s 51(ii) allows the Commonwealth to select any person or entity for taxation. There is no argument, for example, that the Commonwealth may aim a special tax at high income earners though that category is not mentioned in any one of the topics of Commonwealth legislative power. The reason why States and their instrumentalities cannot be targeted for burdens is not because such objects fall outside the plain meaning of the empowering clauses, but because of implied limitations drawn from the structure, objectives and values of the Constitution. This much was made clear by Mason J in the Tasmanian Dams Case, when he rejected Chief Justice Barwick’s attempt to cast the rule in the Melbourne Corporation Case in terms of characterisation of subject matter.[31]

[29] An important elucidation of the operation of the implied limitation doctrine appears in the judgments in Re State Public Services Federation; Ex parte Attorney-General (WA) (‘SPSF Case’).[32] The Justices rejected the suggestion that the determination of the scope of a power specified in s 51 proceeded in two stages: the ascertainment of the true meaning of the paragraph, followed by the inquiry as to whether an implied limitation curtails the power. Chief Justice Mason and Deane and Gaudron JJ stated that the position rather is that

[30] Justice Brennan added that in considering the text of the paragraph, its subject matter and the constitutional context, there is no ‘sequence to be followed in considering one factor before another’.[34] This approach was endorsed by the joint majority judgment in Re Australian Education Union; Ex parte State of Victoria (‘AEU Case’).[35] The significance of these opinions lie in the fact that they drive more nails into the coffin of Engineers’ literalism. They reinforce the view that the meanings of the paragraphs of s 51 cannot be ascertained in isolation but that they are informed and controlled by other parts of the Constitution as well as its overall structure.

[31] The Communist Party Dissolution Act 1950 (Cth) was enacted with the specific purpose of proscribing the Communist Party of Australia (‘CPA’) and other organisations of similar persuasion and of liquidating their assets. The Act gave extraordinary discretionary powers to the Governor-General to proscribe parties other than the CPA and to determine particular persons as ‘communists’, thereby statutorily disqualifying them from holding public office. The key feature of the Act was its failure to lay down any rules the violation of which would lead to judicially imposed penalties. Instead, the task of determining organisations and persons for the deprivations of liberty and property was left to the executive branch in the form of the Governor-General. The law was challenged in The Communist Party of Australia v Commonwealth (‘Communist Party Case’).[36]

[32] The question before the Court was whether this extraordinary legislative derogation from due process was authorised by the defence power (s 51(vi)) or the power to make law incidental to the maintenance of the Constitution and the laws of the Commonwealth (s 51(xxxix) read with s 61). These powers are themselves regarded as ‘purposive’ powers, or powers granted in terms of the achievement of specified purposes. In this case, the purpose was patent. Yet, the Court determined that these measures were not authorised by the powers in s 51(vi) and (xxxix) during peacetime though they may be permissible in times of actual war. The critical factor in this conclusion was the impact of the legislation on the rule of law, a notion that finds no express articulation in the Constitution. In other words, the Court truncated the scope of the defence power by reference to ideas that the Engineers’ doctrine sought to eliminate from judicial consideration. In his judgment, Dixon J emphasised the fact that the rule of law is an assumption upon which the Constitution has been framed. Hence, a law which violates that assumption cannot be incidental to the maintenance of the Constitution.[37] The assumption was inarticulate and implied in the Constitution.

[33] The High Court has consistently resisted the executive’s efforts to make federal courts perform non-judicial functions on its behalf although there is no express prohibition against such vesting. In R v Kirby; Ex parte Boilermakers’ Society of Australia (‘Boilermakers’ Case’),[38] the High Court decisively rejected the Commonwealth’s attempt to use the Australian Conciliation and Arbitration Commission as a body that would make industrial awards and also enforce them in the event of violation. The impugned law was, without any doubt, a law with respect to the subject specified in s 51(xxxv); namely, ‘conciliation and arbitration for the prevention of industrial disputes extending beyond the limits of any one State’. The Court’s primary reason for limiting the Commonwealth’s power under s 51(xxxv) was to uphold a fundamental constitutional value; that is, maintaining a separate and independent judicature. The High Court, for this reason, has limited Parliament’s power to entrust executive functions to federal judges even in their personal capacity where the performance of such functions is incompatible with their exercise of judicial power.[39] In general, most of the rules concerning the separation of judicial and non-judicial powers are implications from the structure of the Constitution. Yet, they have proved powerful enough to override the Engineers’ policy of giving expansive effect to Commonwealth powers in the absence of express limiting provisions.

[34] Recent decisions of the High Court raise real doubts concerning Parliament’s capacity to enact Bills of Attainder or deny citizens the rudiments of fair justice. In Polyukhovich v Commonwealth (‘War Crimes Case’),[40] the Court rejected a constitutional challenge to s 9 of the War Crimes Act 1945 (Cth) that enabled persons who had committed specified war crimes in the European theatre during World War II to be tried and punished in Australia although their crimes did not technically violate Australian law at the time they were committed. Six of the seven judges (Brennan J dissenting) had no trouble relating the law to the external affairs power that s 51(xxix) bestowed upon the Commonwealth. Yet they took the view that had the law been a Bill of Attainder, it would have been unconstitutional. In a direct contradiction of the Engineers’ doctrine, Mason CJ held that despite the absence of an express prohibition against Bills of Attainder and ex post facto laws, ‘the separation of powers effected by our Constitution, in particular the vesting of judicial power in Chapter III courts, imports a restraint on Parliament’s powers to enact such laws’.[41]

[35] In a series of cases decided in the 1990s, the High Court generated a theory of implied constitutional rights in dramatic contradiction to the Engineers’ doctrine. The two most notable of these cases are Nationwide News Ltd v Wills (‘Industrial Relations Commission Case’),[42] and Australian Capital Television v Commonwealth [No 2] (‘Electoral Advertising Bans Case’),[43] in which the Court implied, from constitutional provisions relating to the election of members of the two Houses of Parliament and the structure of the Constitution, the freedom of communication on matters of political concern. More importantly, the Court held that the powers conferred on the Commonwealth Parliament by s 51 were circumscribed by these implications. In the Industrial Relations Commission Case, a majority comprising Brennan J, Deane and Toohey JJ and Gaudron J concluded that the enactment of s 299(1)(d)(ii) of the Industrial Relations Act 1988 (Cth), which made it a criminal offence to publish statements calculated to bring the Industrial Relations Commission (‘IRC’) into disrepute, was within the power conferred by s 51(xxxv) read in isolation, but was nevertheless rendered unconstitutional by the free speech implications of the Constitution. In the Electoral Advertising Bans Case, the Court struck down amendments to the Broadcasting Act 1942 (Cth) that imposed a ban on broadcasting of political advertisements on electronic media during election campaign periods. There was no question that the law was authorised by s 51(v) read alone.

[36] It was observed previously that the Engineers’ doctrine denied the relevance of purpose and proportion to the process of determining whether a Commonwealth law is authorised with respect to a subject. Justice Kitto’s assertion in Herald and Weekly Times Ltd v Commonwealth that ‘how far they should go was a question of degree for the parliament to decide’[44] became a principal standard by which the Court was guided in determining the scope of Commonwealth power. However, as it turned out, this standard is easier stated than maintained. Categories that a text signifies are not based in some transcendent reality but acquire their characteristics from what is conventionally included in them. Though it is true that most people sharing a common culture and similar experiences will perceive a thing as falling either clearly within or clearly outside a described category, there will always be a shifting ‘grey area’ of uncertainty that literalism will not eliminate. The uncertainty is compounded when the category is defined, not in terms of a thing or state of things, but in terms of purposes. The inherent limitations of the doctrine in the Engineers’ Case is forcefully demonstrated by the doctrine of proportionality that the interpretation of empowering clauses unavoidably attracts. The doctrine is ubiquitous wherever written constitutions are taken seriously for it is grounded in logic.

[37] The doctrine has no place in a constitution that reposes all powers in a single authority. However, where legislative and executive powers are granted with respect to specified subjects, however loosely defined, the doctrine immediately springs to life. If the legislature is granted power to make laws with respect to subject ‘X’, two questions arise:

[38] The doctrine of proportionality is indispensable in answering the second question. At some point, the law loses touch with the subject and hence becomes unconstitutional. The concept of proportionality is central to the determination of the point at which the law is disconnected from the authorised subject.

[39] The Constitution bestows legislative power on Parliament with respect to specified subjects. Some of these subjects are signified in terms of purposes to be achieved. They include: naval and military defence of the Commonwealth and the States (s 51(vi)); conciliation and arbitration for the prevention and settlement of industrial disputes extending beyond State limits (s 51(xxxv)); and matters incidental to the execution of powers vested in Parliament, the federal judicature and the government ( s 51(xxxix)).

[40] The relevance of proportionality is most evident in relation to the so-called purposive powers. If a law may be made for a particular purpose, it is always pertinent to ask whether an enacted law is reasonably appropriate and adapted to, and hence proportionate to, the achievement of the specified purpose. As the Communist Party Case demonstrated, the attempt to extend the Engineers’ doctrine to subjects defined in terms of purposeful activity, such as defence of the States and the Commonwealth (s 51(vi)) and the maintaining the Constitution (s 61), gives rise to intractable problems. How does the Court determine whether a given law is one with respect to defence? The words ‘defence of the Commonwealth and of the several States’ are in themselves plain enough. However, beyond the core measures such as the establishment and maintenance of defence forces, the question whether a law is concerned with defence becomes increasingly subjective and incapable of resolution by reference to some notion of the plain meaning of text, as prescribed by the Engineers’ doctrine. While we may conceivably agree on some abstract definition of defence, the question of what constitutes defence beyond actual military action in the face of aggression will often be controversial. This difficulty is reflected in the High Court’s elastic view of the defence power that varies depending on whether the nation is in a state of peace, uneasy peace, actual war or transition from war to peace.

[41] In war time, the Court has permitted the Commonwealth to engage in food rationing and price controlling, labour regulation, alteration of private contractual arrangements and other measures that, in the opinion of the executive, ‘conduce to the successful prosecution of the war’.[45] In peacetime, the defence power authorises laws needed to maintain the armed forces, fortifications, supporting infrastructure and supplies, the conduct of courts martial and conscription. In periods of transition from war to peacetime, the Court would allow war time controls to remain in place for a reasonable time and will permit transitional laws such as those giving hiring preference to returned service personnel. In cases where the defence objective is not patent, the Court has been willing to look for textual as well as extrinsic evidence concerning attendant circumstances in order to determine whether the law is one with respect to defence. The elasticity of the categories signified by s 51(vi) and 51(xxxix) demonstrates the relative disutility of the Engineers’ doctrine to the interpretation of powers defined in terms of purposes.

[42] In more recent times, the test of proportionality has been cast in terms of the requirement that laws be ‘appropriate and adapted’ to the purpose for which the law is constitutionally authorised. The test of proportionality entered Australian constitutional law by way of Justice Deane’s judgment in the Tasmanian Dams Case, where his Honour stated:

[43] The requirement of proportionality occurs at two levels. First, it applies when the question is whether a law is authorised by the express provisions of the Constitution. Thus, in the Tasmanian Dams Case, the Court considered whether the challenged legislation was reasonably appropriate and adapted to the implementation of a treaty, which is one of the purposes for which laws may be made under s 51(xxix). Second, the test also applies when Parliament seeks to restrict implied freedoms guaranteed by the Constitution. The High Court has rightly held that the freedom of communication guaranteed by the Constitution, like all other freedoms, is subject to reasonable restrictions that may be imposed by Parliament in furtherance of legitimate objects. However, the Court has also ruled that the restrictions need to be appropriate and adapted to the ends to be achieved. A restriction would be appropriate if there are no alternative means of achieving the permitted end that would not limit the freedom. Thus, in the Electoral Advertising Bans Case, the majority treated as legitimate the Government’s declared object of keeping the electoral process clean but found the ban on political advertising to be inappropriate as there were alternative means of addressing that problem.

[44] The requirement of adaptation refers to the degree of curtailment. In Davis v Commonwealth (‘Davis’),[47] the High Court considered the provisions of ss 22 and 23 of the Australian Bicentennial Authority Act 1980 (Cth), which sought to prevent private commercial exploitation of the Bicentenary by prohibiting the use of any symbol or logo that carried the words ‘Bicentenary’, ‘Bicentennial’, ‘Sydney’, ‘Melbourne’, ‘First Settlement’ or ‘200 years’. The Court found that s 51(xxxix) authorised the Act setting up the Authority, but ruled that the prohibition in s 22 of the Act was excessive in its effect. Davis is particularly interesting as it was decided before the High Court formally declared the existence of an implied constitutional freedom of communication. Had the case arisen after the Industrial Relations Commission Case and the Electoral Advertising Bans Case, the question before the Court may have been whether the restriction of the freedom of communication was reasonably adapted to the legitimate object of celebrating Australia’s nationhood and the outcome almost certainly would have been the same. What is clear from Davis, the Industrial Relations Commission Case and the Electoral Advertising Bans Case is that in the eyes of the High Court, the requirement of proportionality becomes more stringent when the core values of the Constitution are at stake, and that freedom of communication is one such value.

[45] The foregoing account of the implied limits on legislative power derived from the structural features of the Constitution and from the doctrine of proportionality demonstrate the ultimate failure of the Engineers’ doctrine to serve as a sustainable standard for interpreting the Commonwealth’s legislative power. The Engineers’ doctrine was flawed by two theoretical errors. Firstly, it failed to appreciate the nature of a constitution as distinguished from an ordinary statute. A statute usually sets out to make a limited change to the law or to establish an authority for some regulatory purpose. If Parliament disagrees with the way the courts construe and apply a statute, it can make its will clearer by another law. In contrast, a constitution sets up a system of government that limits and distributes powers and provides for the composition of various organs of government. It is a system of inter-locking and interacting components. A constitution has nothing to do with the will of a parliament, the latter being itself subject to the constitution. It is designed not to produce specific outcomes but to establish and maintain a system of government of a particular character. Not all the rules of such a system can be stated at the foundation, though it is possible to state many more than are stipulated in the Constitution. The first theoretical error of the Court in the Engineers’ Case was to regard the Constitution as not imposing any limitations on legislative power that are not specifically stated in the text. This error has been incrementally rectified by the High Court in the cases just discussed.

[46] The second theoretical error of the doctrine lay in its false assumption that the text of the Constitution is grounded in transcendental reality such that its true meaning can always be ascertained through a value free process of judicial inquiry. This attitude contributed significantly to the Court’s refusal to consider purpose and proportionality in determining whether laws were constitutionally authorised by express words. It is not suggested that words have no commonly accepted meanings. Without such conventional understanding of language, law is impossible. However, beyond such conventionally established core meanings, words cast penumbrae of uncertainty that are resolved through acts of construction by relevant epistemic authorities; in this case, judges. Thus, while terms such as ‘external affairs’, ‘defence’ and ‘trade’ have core socially constructed meanings, they also have penumbrae where meanings are constructed by positive judicial action rather than ‘found’ through passive inquiry. In this domain, literalism of the kind postulated in the Engineers’ Case becomes an objective disguise for what is in effect value laden judicial constructivism. In the heyday of the Engineers’ doctrine, judges refused to consider the question of proportionality on the basis that all that was constitutionally relevant was whether or not the law was connected to the authorised subject. The idea was that whereas ‘connection’ could be objectively determined, the assessment of ‘proportionality’ involved a political judgment that constitutionally resided in Parliament. The distinction between ‘connection’ and ‘proportionality’ was always problematic in theory. More seriously, the distinction allowed the Court to clothe its rulings with the guise of objectivity that they did not deserve. The High Court’s acceptance of the relevance of proportionality exposes to contest issues that were foreclosed by the Engineers’ doctrine and makes the Court’s rulings more informed and transparent.

[47] The rollback of the Engineers’ doctrine was inevitable given the fundamental features of the Constitution, namely: separation of powers, the representative principle, the rule of law and the federal structure. Its eclipse represents a significant retreat from the sovereignty model to the republican model.

[48] The great question for constitutional theorists is whether the pendulum will swing the way of sovereignty in the future. Neither the constitutional text nor its current interpretations provide any assurance against such a movement. As we have seen, texts have no transcendentally true meanings, but only meanings that human minds and actions allow them. The history of the federation shows that the same text may yield very different constitutional outcomes in different epochs. All that can be said from the theoretical perspective is that the High Court can return to the Engineers’ sovereignty model only at the expense of key underpinnings of constitutional government. Whether it will do so depends on the pressures of unfolding history.

[49] The High Court, as part of the political system, cannot be insulated from external pressures. Though courts can and indeed should resist momentary majoritarian pressures, they cannot deviate from strong and widespread shifts in public perceptions concerning law and the Constitution without seriously damaging its own authority. It is by no means inconceivable that some new orthodoxy of centralism and unrestrained power may emerge from the political arena. Constitutional texts have no magical properties and the shape of the living Constitution depends ultimately on the complex web of political, cultural and even economic constraints. However, given the internal contradictions of the sovereignty model of democracy and its own resilience, the republican model may be expected to endure in the foreseeable future in the absence of catastrophic shock.

[50] There is indeed some scope for further movement to the republican ideal in the form of greater recognition and entrenchment of basic rights, the eclipse of the hereditary monarchy, improvements in substantive and procedural due process and even a more thoroughgoing separation of powers based on a directly elected executive. However, the real and perceived threats such as those posed by the terrorist attacks in the United States on 11 September 2001 and by the penetration of Australia’s territorial waters by asylum seekers transported by organised people smugglers are likely to stymie further movement towards the republicanisation of the Constitution, at least in the short term.