Chapter Three:
A Role for the States in Initiating Referendums

Dr Jeffrey Goldsworthy

1. The Dynamics of Constitutional Change

It is well known that, since 1920, the interpretation of the Constitution by the High Court has permitted a continuous and cumulatively massive expansion of the powers of the Commonwealth Parliament. In particular instances the change has been dramatic and, at least initially, a source of consternation in State circles: most notably, the use of the Commonwealth's taxation and financial grants powers to monopolise income taxation, and of its external affairs power to implement international treaties, whether or not their subject-matters would otherwise be within Commonwealth power. In the case of many other Commonwealth powers, such as the corporations and industrial relations powers, a similar story can be told.

During the same period, the people of Australia have almost always refused to grant the Commonwealth greater powers, when the Commonwealth has asked them to do so. Referendums to amend the Constitution are only held if the Commonwealth government advises the Governor-General to submit a proposed amendment to the people, and if the proposal has previously been passed by at least one House of the Commonwealth Parliament. Not surprisingly, therefore, of the forty-two proposals referred to the people to date, twenty-three - which is just over half of them - has sought to enlarge the powers of the Commonwealth. But only two of these have been approved. This has not been due to undiscriminating conservatism on the part of the people. Of the other nineteen proposals, six have been approved: in other words, they have been almost four times as successful. 1

In a number of instances, the Court by interpretation has permitted the Commonwealth to exercise the very same powers which the people, voting in constitutional referendums, previously refused to give it. Of the surprisingly large number of examples, the most striking concerns the corporations power. On five different occasions, the Commonwealth attempted by formal amendment to acquire power to control the trading activities of corporations, or at least monopolies, and on every occasion the people refused to agree; yet in 1971 the High Court adopted a new interpretation of the corporations power which gave to the Commonwealth essentially what it had previously sought in that area. 2 Late last year, the Court held that the external affairs power enables the Commonwealth to exceed the limits inherent in its industrial relations power, despite the failure of four previous referendums to remove those limits. 3

Many more attempts to expand Commonwealth powers by referendum were made before World War II than have been made since, and it has been conjectured that one reason for the decline in the number of requests put by the Commonwealth to the people is that the favourable decisions of the Court have made them largely redundant. 4

But while the High Court has often been able to overcome the unwillingness of the people to expand Commonwealth powers, the people have not been able to overcome the Court's willingness to do so. This is because the people can only be asked to do so by the Commonwealth - which is the beneficiary of the Court's decisions. No Commonwealth government has ever sponsored a constitutional amendment to reduce Commonwealth powers, and none is ever likely to do so. Of the nineteen proposals which have not sought to expand Commonwealth powers, none have sought to transfer them to the States or to limit them in any substantial way. 5

We can summarise all this by saying, first, that there have been in Australia two main sources of constitutional change, one involving the High Court's interpretation of the Constitution, and the other, formal amendment by way of popular referendum; secondly, that the Court has consistently permitted the expansion of Commonwealth powers, while the people have almost always opposed it; and thirdly, that in practice the Court has sometimes been able to overcome the verdict of the people, but not vice versa . The outcome has been a kind of ratchet effect: Commonwealth powers have been continually expanded, but never subsequently contracted.

I do not intend these opening remarks to constitute a blanket condemnation of the High Court. I say this because I believe that Court bashing has recently been taken too far. A decision of the Court is not necessarily wrong just because it results in an expansion of Commonwealth power - not even when the people have previously refused to endorse that expansion themselves. Sometimes an expansion of power is the legitimate result of applying the original words of the Constitution to the very different circumstances of modern times. Moreover, the Court has a duty to decide what those original words, correctly interpreted, mean, and it may be that a previous referendum proceeded on the erroneous assumption that the Commonwealth did not already have some power which, in reality, it did have.

So my intention has not been to attack the Court, but simply to point out that there has long been an imbalance in the dynamics of constitutional change, and one which has denied to the people an explicit opportunity to decide whether or not they approve of the direction in which change has, for a very long time, proceeded. 6 If the Court is partly responsible for this, so too is s.128 of the Constitution, which controls formal amendment by referendum.


2. Section 128 and its History

Many of the delegates to the various Constitutional Conventions in the 1890s worried that the powers of a national government, no matter how narrowly they might initially be drawn, might gradually expand at the expense of the States, and they attempted to prevent this. As John Cockburn of South Australia said in 1891:

"We know that the tendency is always to the centre, that the central authority constitutes a vortex which draws power to itself. Therefore, all the buttresses and all the ties should be the other way, to enable the States to withstand the destruction of their powers by such absorption." 7

In 1901, when the Commonwealth Constitution came into effect, there were many reasons for confidence that it adequately protected the interests of all six States. One important reason was the design of the Senate, in which each original State was guaranteed equal representation. The Senators were expected to represent their respective States, and to vote on State lines whenever necessary to protect their State's interests. In this way, the Senate was expected to protect not only the interests of the States in general, but those of the less populous States in particular, given that the House of Representatives would be dominated by the representatives of New South Wales and Victoria. 8

The Senate was expected to play this role with respect to proposed amendments of the Constitution as well as ordinary laws. Section 128 of the Constitution provides two methods for initiating referendums to amend the Constitution. 9 The first, ordinary method allows both Houses of the federal Parliament to do so. If both Houses pass a proposed amendment, by absolute majorities, s.128 says that the proposal "shall be submitted" to the people in a referendum. Since a proposed amendment can be introduced in either House, the original intention was that the representatives of the States in the Senate would be able to initiate proposals, as well as to block any initiated in the House of Representatives which threatened State interests. On the other hand, the House of Representatives would also be able to block proposals originating in the States' House.

This method of initiating amendments was the only one that s.128 originally included, when it was put to voters in the first round of referendums in 1898. But that draft had to be renegotiated when the required majority of voters in New South Wales failed to endorse it. 10 At a Conference held in January, 1899 the Premier of New South Wales, Sir George Reid, obtained the agreement of the other Premiers to various changes, which he hoped would enable the required majority to be obtained in a second referendum (as indeed they did). One of these changes was the inclusion in s.128 of an alternative method of initiating referendums.

This alternative method seems to allow either House of the federal Parliament to initiate a referendum. If, on two separate occasions, either House of the federal Parliament passes a proposed amendment, by an absolute majority, and the other House rejects or fails to pass it, "the Governor-General may submit" the proposal to the people. On its face, in 1900 this would have appeared to enable the States' House - the Senate - to initiate a referendum even if a majority in the House of Representatives, and therefore the Government, were opposed to it.

This alternative method, inserted in 1899 at the insistence of New South Wales, had previously been discussed and rejected in the Constitutional Convention held in Melbourne in 1898. At that time, it was proposed by representatives from Victoria. Many other delegates at the Convention opposed it, because they saw it as an attempt to over-ride the Senate, to prevent Senators representing the less populous States from blocking proposed amendments passed in the House of Representatives. Victorian and New South Wales delegates replied that the alternative method, which they supported, was even-handed - it would enable either House to initiate a referendum without the assent of the other. At one point, when Alfred Deakin suggested that power might be given to the State Parliaments to initiate referendums, Isaac Isaacs replied that the Victorian proposal "will give the States power to do that through their accredited representatives, the Senators." 11

Sir George Reid, the New South Wales Premier, added his eloquent support to the Victorian proposal, emphasising the opportunities it offered to the States:

"The view I take is that, instead of degrading the Senate, it puts it in a position of absolute equality with the House of Representatives . . . The Senate will be a popular body springing from the people of the States; and surely the representatives of all the States, if they agree that a certain amendment of the Constitution should be proposed to the people, should not be blocked by the representatives of the nation in the House below.
"... Surely the States, as represented by the Senate, have a right to take such a verdict with regard to the proposed change... [T]he States' House ... may well have occasion to ask for an amendment of this Constitution." 12

But notwithstanding this argument, a majority of delegates rejected the proposal, and they did so partly because they suspected that it would enhance the power of the House of Representatives and diminish that of the Senate. 13 They were right to do so. The Premiers of the less populous States may have been duped, when Reid persuaded them to adopt much the same proposal at the Premiers' Conference in 1899. 14

What the Premiers agreed to, and what s.128 now provides, is that if on two separate occasions either House passes a proposed amendment, and the other House rejects or fails to pass it, "the Governor-General may submit" the proposal to the people. The use of the word "may" is curious. In the case of the first, ordinary method of initiating referendums, the word "shall" is used: if both Houses have passed it, the proposed amendment " shall be submitted . . . to the electors". This difference in wording suggests that when one House only has passed the proposed amendment, the Governor-General has a discretion - he may submit it to the people, but is not bound to do so. (In fact, it has been generally assumed that this discretion exists even when both houses have passed the proposal. 15 )

Given the constitutional convention that in exercising this kind of discretion, the Governor-General must act on the advice of Cabinet, and given that it is a majority in the lower House which determines the composition of Cabinet, it follows that Governors-General would almost automatically submit proposals passed only by the lower House to the people, but might well refuse to submit a proposal passed only by the Senate.

It follows from this that although s.128 appears on the surface to treat the two Houses even-handedly, "the Senate is in a position of distinct inferiority". 16 Although the Premiers in 1899 did not intend it, the effect of their compromise was to enable their precious States' House to be bypassed, without securing it the ability to initiate referendums itself. 17

That this was indeed the case was demonstrated in 1914, when the Senate twice passed six proposed amendments to which the Government was opposed. The Prime Minister, Joseph Cook, refused to recommend that the Governor-General refer them to the people, stating that he did not "so misapprehend my responsibilities that I should be willing to recommend the Governor-General to submit to the country a set of proposals in which I do not believe!" 18 When the Senate then requested that the Governor-General submit the proposals, the Governor-General replied that, following "the established usage of responsible government", he had consulted his ministers, who were "unable to advise me to comply with the request contained in the Address of the Senate. I accept their advice, and am unable to grant the request of the Senate." 19

It has been suggested that, notwithstanding this precedent, the Governor-General might be regarded as having a "reserve power" to act independently of ministerial advice in exercising this discretion. 20 But in the absence of an extraordinary constitutional crisis requiring urgent resolution, it is unlikely that any Governor-General would dare to do so. Moreover, a referendum is very expensive and no Governor-General is likely to set one in motion with no assurance that the Commonwealth Parliament will authorise the necessary appropriation of funds. As Reid and Forrest conclude:

"... the initiation of constitutional amendments is confined not just to the [Commonwealth] Parliament, but in practice to the Executive Government, without whose advice the Governor-General is unlikely to sanction the holding of a referendum. Consequently, no measure has ever been put to the electors except at the behest of the Government." 21

There is, moreover, a further problem. Even if s.128 did require the Governor-General, regardless of ministerial advice to the contrary, to submit to the people any amendments proposed by the Senate, it would not have greatly assisted the States. This is because the Senate has not fulfilled the Founders' expectation that it would represent the interests of the States. The Senate has functioned as a second party House, whose members' votes are determined by their parties' policies rather than Senate interests. 22 So even if the Senate were able to initiate referendums despite Government and lower House opposition, it would give the States no comfort.


2. Reforming Section 128

Many possible reforms favourable to the States might be proposed, for example, to limit some Commonwealth powers, to expand State powers, or to give the States a larger role in the appointment of High Court Justices. But at present none of these reforms, regardless of their merits, has any hope of being put to the people. Neither House of the Commonwealth Parliament is likely to co-operate, and in any event, the Commonwealth government is most unlikely to advise the Governor-General that a referendum should be held. For practical reasons, therefore, the initiating procedure is structurally biased against certain kinds of proposals being put to the people - even if they are meritorious. That bias can be cured only if s.128 is itself amended to include some other method of initiating referendums.

Empowering half or more of the State Parliaments to initiate referendums is a reform which was advocated at three sessions of the Australian Constitution Convention, attended by representatives of the Commonwealth and all six States, in 1973, 1975 and 1985, and was approved in 1985. Most recently, the Constitutional Commission which reported to the Commonwealth government in 1988 unanimously recommended it - even Gough Whitlam, a member of the Commission, endorsed it. 23 The Commission recommended that if, within a twelve month period, a proposed amendment were passed in identical terms by at least half the State Parliaments, representing a majority of Australians overall, it would have to be put to referendum within two to six months later. The Commonwealth government would be obliged to submit the proposal to the people, to provide funding for the referendum, and, if the proposal were approved, to recommend that the Royal Assent be given by the Governor-General.

One argument in favour of a reform along these lines is that it would rectify a mistake which has prevented the Constitution from fulfilling one of its original purposes. It would achieve something which the Premiers seem to have believed, albeit erroneously, that they had achieved: the inclusion in the Constitution of a mechanism enabling the States, in effect, to initiate referendums. They relied on the Senate, but it has failed them. A formal procedure permitting State-initiated referendums would simply remedy that failure, and restore a balance to the Constitution which it was originally intended to have.

A second argument is that the reform would be fairer to the States. The States created the Commonwealth, but now, like Dr Frankenstein, find themselves overpowered by their own creature. They are denied an advantage which the Commonwealth has: that of seeking a constitutional amendment to advance their distinctive perception of the national interest.

It has been argued that only the Commonwealth represents the nation as a whole, and that the States only represent parts of it. Since any amendment to the Constitution affects the nation as a whole, it is appropriate that it be initiated by the representatives of the nation. 24

But although each State may represent only part of the nation, they are all very important parts, and together they constitute almost all of it. Moreover, the whole point of a federation is that its parts have constitutional standing, and guaranteed rights and powers. The parts no less than the whole are legitimate stakeholders in any federal Constitution. If a majority of those parts believe that the Constitution could be improved, why should they not be able to put their case directly to the people?

But the strongest argument in favour of such a reform is that it would enhance the right of the people to determine the content of their Constitution. Some process for initiating referendums is required, and for reasons of economy, it should be one which screens out proposals which are frivolous or have no chance of success. But the process should not be one which allows the initiators to exclude serious proposals with a genuine chance of success, simply because they don't approve of them. The function of the initiating process should not be to prevent the people from making a decision which the initiators would not like. At least with respect to the Constitution, that kind of political guardianship is unacceptable. It is inconsistent with the rationale for s.128 in its current form, which is to enable the people to determine their own constitutional fate.

It would be naive to think that in initiating referendums, the Commonwealth necessarily acts as an impartial agent of the people. All organisations develop their own biases and interests, and seek to protect their own turf, and governments are no exception. What the Commonwealth believes is best for the nation is not necessarily best for the nation. If it were, the people wouldn't need to be consulted. The States as well as the Commonwealth make up the federal system, and have an equal stake in its proper functioning and an intimate knowledge of its day to day operations. They are as well placed as the Commonwealth to detect structural deficiencies which need reform, deficiencies which for reasons of its own the Commonwealth might not want to rectify. To prevent the people from rectifying such deficiencies is unfair to them even more than it is unfair to the States.

I have found it difficult to think of weighty objections to this proposal. It has been suggested that it might "increase strategic State manoeuvring", 25 but I can't see anything wrong with giving the States a bit more bargaining power in their negotiations with the Commonwealth. It might lead to an improvement in the success rate of referendums. Brian Galligan, a leading political scientist, has concluded that the failure rate so far is the fault not of the Australian people, but of "élites and federal Labor politicians who prefer a more centralised regime or change for its own sake", and often mount unnecessary and poorly argued campaigns. 26 Perhaps the threat of being bypassed by an alternative initiating process would galvanise the Commonwealth into paying more careful attention to the question of reform. Competition, after all, has its benefits.

Probably the weightiest objection is that the proposal would be an expensive waste of time, because just as Commonwealth-initiated referendums fail if they are strongly opposed by the States, so would State-initiated referendums fail if they were strongly opposed by the Commonwealth.

Conventional wisdom has it that proposed constitutional amendments have a genuine chance of being approved by the people only if they enjoy bipartisan support. All eight of the proposals which have met with success at referendums received bipartisan support. But bipartisan support at the federal level is only a necessary, and not a sufficient, condition for success, since five proposals which enjoyed it have failed.

A plausible hypothesis is that, in addition, it is necessary that State governments do not mount substantial opposition. 27 None of the eight successful proposals encountered any organised opposition at all, and in five of those cases, the Commonwealth and the States had agreed to them in advance. 28 As for rejected proposals, on the other hand, a 1983 analysis concluded that they "would all have wrought an appreciable change in the prevailing balance of power between the Commonwealth and the State Parliaments, either by extending Commonwealth power to affect an existing State jurisdiction or by altering the existing patterns of State influence." 29 (Of the six proposals which have failed since then, five either fell into the same categories or would have diminished the powers of State Parliaments.)

What should we make of this evidence? The first point to make is that, while it shows that Commonwealth proposals opposed by or perceived as detrimental to the States are unlikely to be approved, we do not yet have any evidence at all concerning State proposals opposed by or perceived as detrimental to the Commonwealth. So a possible conclusion is that, while it makes no sense to allow the Commonwealth to initiate proposals opposed by the States - because the expense of a referendum campaign will inevitable be wasted - there is as yet no good reason not to allow the States to initiate proposals opposed by the Commonwealth.

But the truth, I suspect, is that any substantial opposition - whether mounted by political parties, State governments, or the Commonwealth - would be enough to deter the people, who naturally prefer the status quo if they have any doubts. If so, then any proposal supported by the States but vigorously opposed by the Commonwealth would probably be defeated. It might then be concluded that if State and Commonwealth consensus is needed, there is no point in allowing State Parliaments alone to initiate referendums without Commonwealth support.

But if that conclusion were really justified, there would also be no point in allowing the Commonwealth to initiate referendums without State support. If our over-riding concern were to save the wasted expense of doomed referendum campaigns, we should change the current system to require that, before being put to the people, proposals be passed both by a majority of State Parliaments and by the Commonwealth Parliament. I doubt that anyone would support that proposal. But if it is reasonable for the Commonwealth to be able to initiate referendums without formal State approval, why not vice versa ?

Apart from questions of fairness, we should not be so confident in predictions based on past experiences that we completely rule out the possibility of the people approving a proposal supported by either the Commonwealth or the States, and opposed by the other. The proposal might turn out to be a surprisingly popular one.

It is sometimes argued that the Commonwealth must initiate any constitutional amendment to enable the States to initiate referendums, and therefore that, regardless of the merits of the reform, "it seems impossible to contemplate the Commonwealth government doing anything voluntarily to waive its advantage in this respect." 30 In other words, the reform is likely to be frustrated by the very structura

l bias which it is needed to rectify.

Rather than an argument against the reform, this is an argument in its favour. The main obstacle in its path demonstrates the need for it. Of course, the obstacle is a formidable one. The Commonwealth is very unlikely to accede to a request from the States that this reform be put to the people. But it might find it more difficult to ignore a request made by the People's Convention. A People's Convention might provide the only real opportunity to persuade the Commonwealth to allow the people to decide.



Endnotes :

Return to Table of Contents Copyright 1997 by The Samuel Griffith Society. All Rights Reserved
1 . E. Campbell, Changing the Constitution Ä Past and Future (1989), 17 Melbourne University Law Review 1, 3. The two successful proposals of this kind concerned legislative powers with respect to social services and Aborigines.
2 . This and other examples are discussed by M. Coper, The People and the Judges: Constitutional Referendums and Judicial Interpretation , in G. Lindell, ed., Future Directions in Australian Constitutional Law (1994) 73, 78Ä80.
3 . Victoria v. Commonwealth (1996) 138 ALR 129, at 210 per Dawson J.
4 . Campbell, op. cit. , 4.
5 . They have concerned "machinery of government" matters, relating to the internal structures or procedures of Commonwealth or State governments, local government, Commonwealth-State financial arrangements, the ability of the Commonwealth and the States voluntarily to transfer their powers, voters' rights, and other rights against State laws.
6 6. One could argue that the people have had many opportunities Ä namely, at every federal election. If the expansion of Commonwealth power was unpopular, surely some political party would have made an issue of it, and promised to submit suitable proposals to a referendum, in order to win office. But this ignores the common interest which all the major parties at the Commonwealth level have in maximising the powers which they all aspire to control. It also ignores the fact that federal elections are not decided by single issues, and especially not by constitutional issues, which are likely to be swamped by concerns perceived to be of more immediate practical importance.
7 . Quoted in L.F. Crisp, The Parliamentary Government of the Commonwealth of Australia (3rd ed., 1961), 25.
8 . Not all the founders entertained this expectation Ä some accurately predicted the future predominance of party allegiance: B. Galligan, A Federal Republic (1995), 81Ä4. But they did not persuade the majority: C. Howard and C. Saunders, The Blocking of the Budget and Dismissal of the Government , in G. Evans (ed), Labor and the Constitution 1972Ä75 (1977) 251, 254Ä55.
9 . Note that the word "referendums" rather than "referenda" is generally used in this context.
10 . A majority of voters in New South Wales did endorse the proposed Constitution, but it fell short of the number required by State legislation: see J.A. La Nauze, The Making of the Australian Constitution (1972), 240.
11 . Official Record of the Debate of the Australasian Federal Convention, Melbourne, 9 February, 1898, vol. I, 730.
12 . Ibid , 735; see also 761 (Reid) and 748 (Kingston).
13 . Ibid , 717 (Glynn), 725 (Downer), 733Ä34 (Symon), 747 (Dobson), 752Ä3 (Solomon) and 755 (Howe). The only reason for opposition to the proposal was the belief that it was inconsistent with the nature of representative government, which involves decisionÄmaking by representatives chosen in order to exercise wise and informed judgment on behalf of the people.
14 . But see Endnote 17, below.
15 . See Final Report of the Constitution Commission 1988 (1988), vol. 2, 883Ä85; E. Campbell, op. cit., 2n.7. For an argument that the assumption is wrong, see C. Howard, Australian Federal Constitutional Law (3rd ed., 1985), 570Ä71.
16 16. The Hon. Justice Evatt, Amending the Constitution (1937) 1 Res Gestae , 264.

. That they did not intend it is indicated by the following paragraph from the joint statement released after their Conference:

"The Premiers agreed that, where there is a difference of opinion between the two Houses as to whether the people should have the opportunity of deciding if any alteration should be made in the provisions of the Constitution, one House should not have the power to prevent the question being decided by the people." See J. Quick and R. Garran, The Constitution of the Commonwealth of Australia Annotated (1901), 220.
The unintended result was probably the result of oversight rather than cunning on Reid's part. The wording of the relevant paragraph in s.128 was obviously copied, with necessary modifications, from the first paragraph of s.57, which says that if the Senate refuses to pass legislation passed on two occasions by the House of Representatives, "the Governor-General may dissolve" both Houses. The word "may" was carried over to s.128 along with much of the other wording of s.57.
18 . Commonwealth Parliamentary Debates, 4 June, 1914, 2191, quoted in G.S. Reid and M. Forrest, Australia's Commonwealth Parliament 1901Ä1988 (1989), 243.
19 . Journals of the Senate (1914), 98, quoted in Reid and Forrest, loc. cit. Geoffrey Sawer pointed out that on this occasion, the requirements of s.128 had not been complied with, because the House of Representatives had neither rejected nor failed to pass the Senate's proposals. But he concedes that the Governor-General did not offer that as his reason for rejecting the Senate's request: G. Sawer, Australian Federal Politics and Law 1901Ä1929 (1956), 124Ä25.
20 . R.D. Lumb and G.A. Moens, The Constitution of the Commonwealth of Australia Annotated (5th ed., 1995), 569.
21 . Reid and Forrest, op. cit., 244.
22 . There has been some debate on this score, but see Howard and Saunders, op. cit., n.9,260Ä61, and G. Winterton, Parliament, the Executive and the Governor-General (1983), 8n.77. John Nethercote pointed out during discussion that the Senate's attempt in 1914 to initiate a referendum was itself an early example of the Senate failing to act as a States' House : the proposals the Senate wished to put to the people involved expanding Commonwealth powers in accordance with Labor Party policies.
23 . Final Report of the Constitutional Commission 1988 (1988),vol. 2, 856Ä61 and 883Ä85. The Commission's two-volume Report is an invaluable resource for anyone interested in the Constitution and its reform.
24 . This view was expressed by Senator Tate, a Commonwealth representative at the 1985 session of the Australian Constitutional Convention : Official Record of Debates etc., Australian Constitutional Convention , Brisbane, 1985, p.288.
25 . B. Galligan, A Federal Republic, Australia's Constitutional System of Government (1995), 116.
26 . Ibid. , 131.
27 . For these figures, and the hypothesis, see Campbell, op. cit., 6.
28 . J. McMillan, G. Evans and H. Storey, Australia's Constitution, Time for Change? (1983), 29.
29 29. Ibid. , 28Ä9.
30 . Ibid. , 350.