A Labor Perspective on Senate Reform


John Faulkner

In calling this paper `A Labor Perspective on Senate Reform', I have deliberately left myself considerable latitude. As part of a more general case concerning the importance of our voting systems to the strength and stability of our institutions of government, I will both revisit the 1948 reform and canvass the need for further reform. In doing so I will discuss three specific proposals for future reform: a change to the procedures for allocating short- and long-term senators after a double dissolution election, the introduction of four-year terms, and a limitation on the Senate's power to block appropriation bills.

Political parties bring a partisan perspective to any analysis of voting systems and consideration of proposals for change. It is in the nature of a political party that it will seek whatever advantage it can gain from `electoral reform'. But this does not mean that our perspective is exclusively partisan. Far from it. The vast majority of politicians would agree with the observations of a recent Research Note of the Parliamentary Library:

Electoral systems have a number of functions which need to be held in balance for the effective operation of the democratic process. An electoral system needs to be representative of geographic regions, political beliefs and the societal and cultural aspects of the population. The system must be fair and must not discriminate in favour of one group against another … Other functions of and criteria for an electoral system include the promotion of stable government, the facilitation of viable and effective opposition, that seats won should as far as possible be in proportion to votes received, and that power exercised should bear some relationship to the vote received. [1]

Voting systems must be fair and have integrity. Whatever the defects of the systems we employ, both for the House of Representatives and the Senate, I am confident Australians would agree they meet these criteria. I also believe that the Australian Labor Party has a very strong record of promoting and defending the fairness and integrity of our electoral system.

In assessing the Senate voting system, we must also be mindful of the constitutional requirement that all states have equal representation. This was part of the price of the constitutional settlement that led to the establishment of the Commonwealth, and political realities are such that it is unlikely that this provision will ever be changed. This requirement does, of course, distort, or more accurately subvert, the `one vote, one value' principle. It is in the nature of a congenital defect if you like. It produces the current situation, for example, where 12 senators represent 330,000 electors in Tasmania and the same number represent more than four million electors in NSW. As I have argued before, whatever else this system might be, it is hardly a model of representativeness.

But the Senate system does produce a result that accurately reflects the voting strengths of parties within state and territory boundaries. This could not be said for the systems that preceded the introduction of proportional representation (PR) in 1948: the original first-past-the-post system and the post-1919 preferential voting system. The first system produced results such as that in 1910, when all 18 Senate seats were won by Andrew Fisher's Labor Party; 1914, when, after a double dissolution, the ALP won 31 of the 36 seats; and 1917, when Billy Hughes' Nationalists won all 18 seats. These results became even more bizarre under the preferential voting system. In preparing this paper, I found an interesting letter—on an archived file—from a proponent of proportional representation, Senator Burford Sampson (Lib, Tasmania). Writing to Ben Chifley in 1945, he declared, `The purpose of this letter is to emphasise the general opinion that the present method of election of members of the Senate is about as bad a voting system as such can be.' He went on to describe such `grotesque and unfair results' as in 1925, when Labor had 45 per cent of votes but did not have a single senator, while the Nationalists, with 55 per cent, had 22. And in 1943, when Labor under John Curtin won all 19 seats with 55 per cent of the vote. These winner-take-all voting systems are simply not appropriate for multi-member electoral constituencies such as we have in Senate elections.

The historical record shows that Sampson was not the only one who believed that something had to be done. In 1915, the Royal Commission on the Commonwealth Electoral Law and Administration recommended the adoption of a system of proportional representation for the Senate. Earle Page succeeded in having a motion to introduce PR in the Senate adopted in the House of Representatives in 1922. Joe Lyons committed his United Australia Party (UAP) government, in 1934 and again in 1937, to establish a Select Committee `to investigate and report on the method of election of the Senate'. The debate on this motion, which took place in the House of Representatives in May 1939, is instructive. [2] Were we to debate a similar motion now, in 1999, I am sure that very similar views would be expressed. It was the conservative parties that then argued the need for change. Thomas Paterson, Country Party Member for Gippsland, said:

Anyone who gives a moment's thought to the subject must realize that something should be done to improve the present system, under which a party which gains not more than 51 per cent of the total votes cast throughout Australia may win the whole eighteen seats at any normal election, whereas a party which may obtain 49 per cent of the total votes cast may be left entirely without representation.

Josiah Francis, UAP Member for Moreton, also made the case in terms of democratic legitimacy:

I have heard nobody express satisfaction with the way in which the Senate elections are now conducted. After every election the press teems with criticisms of the present system, which is condemned by leader writers, newspaper correspondents, and public bodies alike. The present method does not give the community confidence in this Parliament, and without such confidence this Parliament cannot function as it should.

The then Labor Opposition roundly condemned the government for political opportunism. Eddie Ward, Labor Member for East Sydney led the charge:

I do not deny that for many years there has been dissatisfaction with the present method of election to the Senate ... what amazes me is the fact that for that long period the Government was unmindful of that dissatisfaction and that it has taken heed of it only after a general election which indicated that the fortunes of the Labor party are improving ... We believe that the Government is making these proposals at this stage in order to destroy the advantage which the Labor party now holds.

He was backed up by H.P. Lazzarini, Labor Member for Werriwa:

Whenever governments comprised of honourable members opposite in this or any other parliament of Australia have received a thrashing at the hands of the electors, they have always sought a way to thwart the popular will … Now that the United Australia Party and the Country party are fighting as independent bodies, each, I suppose, feels that it is entitled to proportional representation. Possible the Country party believes that it can win one or two more Senate seats.

While support for PR in the Senate waxed and waned from as far back as the constitutional conventions in the late 1890s, and moved back and forth across party lines, it was Ben Chifley's Labor Government which finally took action. As John Uhr has observed in his paper, the historical record of the background to Chifley's 1948 decision is sketchy. This prompted me to have a look at the Cabinet records of this period and some other relevant files at the National Archives. These show that the Chifley Cabinet considered a proposal to increase parliamentary representation on 3 July 1947, but deferred further consideration until a more thorough analysis of the results of the 1947 Census was undertaken.

The historical record makes it clear that dissatisfaction with the Senate voting system had been building for many years as had pressure for change to a system of proportional representation. Concern had also been growing about the gradual erosion of representativeness in the electoral system. At the time of the first federal elections in March 1901, each member represented approximately 20,000 electors. This number had grown to almost 60,000 by the time of the 1947 Census. My own reading of the primary sources suggests to me that it was the combination of these pressures coupled with the need for the Chifley government to respond to the results of the 1947 Census with redistributions in five of the six states that tipped the balance in favour of change.

Undoubtedly, the keen political instinct for self-preservation also played an important role. Uhr is correct to draw attention to Labor's interest in the `smaller, more stable and hence more secure seats for the Labor backbench' that the increase in the number of Members of the House of Representatives would bring. And there is no question that Arthur Calwell—who, according to Fred Daly, [3] was the `architect and enthusiastic sponsor of the enlarged parliament and proportional representation'—played this card successfully in persuading caucus to accept his proposed formula for change. [4] The interpretation that the Senate changes were designed to bolster Labor's majority in that House in the event of a victory by the Menzies Coalition is less convincing. After all, these proposals were formulated and approved a little over one year into Chifley's second term of office, when general elections were still some two years off.

While the system adopted in 1948 was a massive improvement, it was by no means perfect. There were serious defects. One was the treatment of casual vacancies, which was largely remedied by referendum in 1977. Not only could state governments appoint a senator who was not a nominee of the relevant party, but the filling of any casual vacancy in a subsequent election changed the quota for the election of a senator. Another defect, which can also lead to distortion of proportional representation, is the continuing possibility of non-simultaneous House of Representatives and half-Senate elections. Experience shows that separate half-Senate elections are treated as nation-wide by-elections, so as well as being inordinately expensive, they tend to favour a protest vote. This may explain why one has not occurred since 1970.

The prediction in 1948 by Attorney-General Herbert Evatt that proportional representation would enhance the status of the Senate has proved to be correct. PR has given the Senate a popular legitimacy it had previously lacked. It is not surprising that that legitimacy has in turn fortified the Senate's sense of independence, consciousness of its powers and preparedness to exercise them. PR created a situation where government control of the Senate became increasingly rare (governments only having majorities in the Senate after the elections of 1951, 1953, 1958, 1975 and 1977) and, since 1984 when Senate numbers were increased from 10 to 12 per state, almost impossible. Minor parties and independents have become a feature of the Senate, since the advent of the Democratic Labor Party in 1955, with their numbers gradually increasing from two in 1955 to the current 12. They have held the balance of power for 32 of those 44 years.

At no stage was the Senate's assertion of its sense of independence more apparent than in the years of the Whitlam government, from 1972 to 1975. During this period, the Opposition rejected a record 93 government bills, 25 more than the total number of rejections in the previous 71 years of the Senate's history. Having lost government for the first time in 23 years, the Opposition, displaying all the characteristics of a sore loser, announced in April 1974 that it would vote against the Supply bills in the Senate. Whitlam sought a double dissolution and was returned to government. Again, in October 1975, the Opposition announced that it would not pass the Budget bills in the Senate `until the Government agrees to submit itself to the judgment of the people.' The rest, as they say, is history. The country was catapulted into a constitutional crisis the like of which it had not seen before and has not seen since.

Malcolm Fraser and Doug Anthony Campaigning

National Times, October 27–November 1, 1975

Fred Chaney was the Liberal Opposition Whip at the time. He wrote last year that he

saw at close quarters the ultimate exercise of Senate authority when it denied the Whitlam government supply. It did this on the basis of a fiddled majority produced by the shenanigans of the state governments in NSW and Queensland. Nevertheless, this did not deter the conservative forces from the view that the Senate had a right not only to amend government legislation but to bring a government down. [5]

I do not believe the Senate should have that right. Nor does the Labor Party. Our platform supports `constitutional reform to prevent the Senate rejecting, deferring or blocking appropriation bills'. I will return to this issue shortly.

The flip side of the Senate's greater assertiveness and the growing influence of minor parties and independents has been the increasing frustration of governments. Generally they have displayed what I regard as the knee-jerk reaction, and I include both Labor and Coalition governments in this observation, and that is to propose amending the Electoral Act to alter the electoral system in such a way as to reduce the likelihood of non-major party representatives being elected. This reaction is at once an acknowledgment of the impossibility of achieving such an outcome by constitutional means and an appeal to the shared interests of the other major political party to join forces to enhance the prospects of both parties of enjoying at least occasional control the upper house. Paul Keating made this appeal quite explicit in 1994, during one of his periodic bouts of frustration with the Senate, when he told the front bench of the Liberal Party that it had to `think beyond your nose and the next election'. He noted that `the essential robust element of our democracy is the representative nature of the Australian ballot. The Senate is not a representative ballot; it is a proportional ballot.' [6]

Keating and Gareth Evans were both extensively reported in the early part of 1994 as canvassing a proposal to divide each state into 12 electorates, with senators being elected by preferential rather than proportional voting. The idea is certainly not a new one. A variant of this proposal was debated in the House of Representatives in 1939. Similar proposals have emerged regularly from the Coalition since it gained office in 1996 and developed a radically different view of the tactics it had employed in the Senate under Labor governments. We have had calls and proposals for `Senate reform' from Andrew Robb, Tony Staley, Wilson Tuckey, David MacGibbon, Tony Abbot, Peter Costello, Peter Reith and most recently, Helen Coonan.

What all these proposals have in common is that they involve an amendment to the Electoral Act, which would almost certainly require the support of both of the major parties. They are also designed to produce a particular outcome from Senate elections; that is, enhanced representation for the major parties at the expense of the minor parties. But it is highly unlikely that any of the proposals currently being put forward by Coalition representatives would have produced a Coalition majority in the Senate from the 37.7 per cent of the vote that it obtained in the last election, its lowest percentage in Australian federal history. As Malcolm Mackerras has pointed out [7], this is the reason why John Howard will not use the mechanism available to him under the Constitution to resolve a deadlock with the Senate. This is the reason why he will not contemplate a double dissolution and is instead seeking to manipulate the electoral system to reduce the prospect of such deadlocks. As the Proportional Representation Society of Australia has observed recently, `Deliberately distorting voters' wishes by electoral artifice simply cannot produce fairer electoral outcomes.' [8]

Proposals for Senate reform that rely on radical change to the electoral system are no longer realistic in the Australian political context. Australians just will not buy them. Minor parties in the Senate are here to stay. They are a permanent consequence of the change to proportional representation and now account for some 25 per cent of the vote. Voters are not swayed by government complaints about Senate obstruction. In June 1997, after months of government protests about the Senate, a Bulletin Morgan poll showed 72 per cent of voters opposed any electoral change designed to make it easier for major parties to control the Senate. The poll found 18 per cent of voters preferred to see minor parties hold the balance of power in the Senate to keep a check on government policies. Seventeen per cent said they would vote for one party in the House of Representatives and another in the Senate. Only 10 per cent of voters thought it was a bad thing for the government not to control the Senate while most people (67 per cent), felt it could be good or bad, depending on the circumstances.

If there is a problem with Senate powers, and the behaviour of the Senate from 1972 to 1975 demonstrated conclusively that there is, then we should look for a remedy at the heart of the problem: the Constitution that sets down those powers. The Senate's power to reject, amend or fail to pass what might be termed ordinary legislation does not pose a threat to our system of responsible government. Yes, it often constitutes an irritant to governments, but there are remedies available. A government can either swallow its pride and set the bill aside or use the s. 57 deadlock process.

The real problem arises with regard to the Senate's power to deny financial sustenance to a government, particularly when such power is exercised not because of any objection to the content of the legislation appropriating the funds, but to bring down the government. This flies in the face of one of the basic principles of our system of government, that a government is responsible to the House of Representatives and continues in office only so long as it has the confidence of that House. This was very much a live issue at the constitutional conventions that led to the framing of the Constitution. Aficionados should consult the Final Report of the 1988 Constitutional Commission [9] and Brian Galligan's chapter in Responsible Government in Australia [10] for a detailed treatment of this issue. I have drawn on these sources extensively for the following historical account.

The problem of combining the traditional concept of responsible government centred on the people's house with a bicameral legislature comprising two almost equally powerful chambers dominated the Conventions and almost undid the whole Federation project. Interestingly, though, the battleground was the Senate's power to amend tax bills, rather than its power to block them. It was the larger states that argued that the traditional form and practices of responsible government would be placed in jeopardy and government become unworkable if both houses had the power of amending money bills. The small states wanted a powerful states' house in which to promote and defend their interests. They played down the risks to responsible government. The South Australian Chair of the Judiciary Committee at the 1897 Convention, Josiah Symon, argued that the Senate's right to amend tax bills would not threaten responsible government because they could `trust to the good judgment and conscience of the Senate', which would be comprised of `eminent men who will not readily or wantonly put difficulties in the way of the government of the country.'

The settlement of this argument in favour of the large states turned the attention of Convention delegates to the problem of deadlocks between the houses. Richard O'Connor proved to be particularly prescient. He distinguished between two types of deadlocks: ordinary ones that could be solved by compromise, and dangerous ones that required some special mechanism for resolution. He included all policy matters and even taxation bills in the first category and only appropriation bills for the ordinary annual services of government in the second. These latter deadlocks were dangerous because they would stop the whole machinery of government. For such deadlocks he proposed a joint sitting of the two houses if the Senate refused to pass a Supply Bill for the second time.

Ultimately, after extensive debate, the s. 57 double dissolution mechanism was adopted. It was acknowledged that this would only be suitable for handling ordinary deadlocks, given that the timing requirements were obviously too drawn out for supply deadlocks. But it seemed that only O'Connor saw this as a real problem. Most delegates considered his dangerous deadlocks as so serious as to be unthinkable. According to Patrick McMahon Glynn, a deadlock over an appropriation bill would `open up the way to a revolution' and the fear of such a thing occurring would `operate as a sanction to prevent it.' William McMillan suggested that the blocking of supply would throw the whole finances of the Commonwealth into confusion and `would mean revolution'. Another delegate—William Trenwith—agreed that this was `inconceivable'.

As Galligan points out, the founders recognised that deadlocks that could occur between a government centred in the House of Representatives and the Senate `would be avoided primarily by the good sense of those who worked the system, or alternatively resolved by the cumbersome `mechanical' method of double dissolution'. However:

the founders did not anticipate the rapid rise of disciplined, class based parties soon after federation, nor did they envisage the degree of partisanship and reckless brinkmanship that characterised 1975 … The founders had provided no adequate mechanical means for breaking supply deadlocks because they never envisaged politicians would engineer such dangerous things for short run gains.

In a very real sense, then, the Senate's power to block supply and the delineation of this power in the Constitution is unfinished business. It is a circumstance not anticipated by our founders and therefore not provided for. There is, perhaps, a unique opportunity for us, as a country, to address this problem and take action to prevent a recurrence of the 1975 constitutional crisis.

As far as the principal parliamentary parties are concerned, Labor is committed to constitutional reform to prevent the Senate rejecting, deferring or blocking appropriation bills. The Democrats have given an undertaking not to use the Senate to block supply. And the Coalition is publicly casting around for ways to prevent the Senate from being `an obstructional competitor, frustrating or substantially delaying urgently required responses to national problems.' [11] Surely there is some convergence among these positions. I am not in any sense suggesting any form of collusion between the major parties to remove an irritant to both. This is not a matter to be settled between the two, or for that matter, the three main parliamentary parties. Ultimately it is a matter for the Australian people, who would have the final say via a constitutional referendum. But, as we all know, there is little use going to a referendum if the major parties take opposing positions on the question.

A useful starting point would be the Constitution Amendment (Legislative Council) Act which the NSW parliament passed in 1933. That Act amended the NSW Constitution by adding the following section:

5A (1) If the Legislative Assembly passes any Bill appropriating revenue or moneys for the ordinary annual services of the Government and the Legislative Council rejects or fails to pass it or returns the Bill to the Legislative Assembly with a message suggesting any amendment to which the Legislative Assembly does not agree, the Legislative Assembly may direct that the Bill with or without any amendment suggested by the Legislative Council, be presented [to the Governor for Royal Assent] notwithstanding that the Legislative Council has not consented to the Bill.

(2) The Legislative Council shall be taken to have failed to pass any such Bill, if the Bill is not returned to the Legislative Assembly within one month after its transmission to the Legislative Council and the Session continues during such period.

(3) If a Bill which appropriates revenue or moneys for the ordinary annual services of the Government becomes an Act under the provisions of this section, any provision in such Act dealing with any matter other than such appropriation shall be of no effect.

Senate reform via constitutional change has long been regarded as impossible given the poor success rate of referenda in our country. But the upcoming referendum on the Republic and the imminence of the millennium and the centenary of Federation will create a climate in which Australians will be more prepared to contemplate other possible changes to our Constitution. As Kim Beazley has said, it `may well crack open this century's constitutional conservatism.' [12] In such a climate, Labor would be prepared to discuss with other interested parties, other aspects of the Senate's constitutional powers and join in a public debate.

There are at least two other proposals Labor would advocate in such a context: changing the current system for rotation of senators and introducing four-year terms. I spoke about the first proposal in the Senate on 13 May last year [13] and subsequently moved a motion to give it effect. That motion was passed by the Senate on 29 June 1998. This may seem an obscure issue to many, but it is an important reform that will ensure that the allocation of short- and long-term places after a double dissolution election is conducted according to democratic principles.

The Constitution provides that senators should have six-year terms and that half the Senate should retire at each ordinary Senate election. In the event of a double dissolution under s. 57, the ordinary rotation of senators is disturbed and, following the election, the Senate is required to divide itself into short- and long-term senators. The method of division is left to the discretion of the Senate. Traditionally, the method that the Senate has settled on has been to divide senators on the basis of the order in which they were elected. But, while the proportional representation system accurately distributes representation, it is a poor method of ranking senators according to their voting strength. Any Senate candidate reaching a quota is ranked ahead of a senator elected by the distribution of a surplus from a candidate higher on the ticket. Thus, this arrangement favours minor parties. It ensures that they will receive a long-term senator as soon as they receive a quota. In a double dissolution election, this means that, provided they get over 7.69 per cent of the vote, they will receive a long-term senator, whereas a party that attracted 44 per cent of the vote may receive only two long-term senators. Such a result is inconsistent with the principle of proportional representation.

In 1984, parliament passed an amendment to the Commonwealth Electoral Act to provide for the allocation of short- and long-term senators after a double dissolution based on a recount with the quota that would have applied had a half Senate election been held. Labor's view is that this represents the fairest method of allocating places. Order of election and voting strength are different concepts. Order of election simply reflects the order in which successful candidates reach quota. Voting strength more closely resembles proportional support for parties contesting the election. Although the half Senate recount method is not perfect, it is a more accurate measure of voting strength and a more accurate measure of the will of the electorate. Hence it is more democratic.

At the time of the debate on the 1984 amendment, Senator Robert Ray, on behalf of the then government, stressed that, if this procedure were to enter into use, it should be agreed in advance of any double dissolution election. This remains Labor's view. If it is not agreed in advance, then parties will adopt a wait-and-see approach until after the election, when they will advocate the method that brings them most political advantage. It is my hope that the resolution of the Senate of 29 June last year will have sufficient force to ensure that the method that was approved by parliament in 1984 will be employed after the next double dissolution election, whenever that may occur. The Democrats should not interpret this proposal as simply a move by the major parties to disadvantage the minor parties. This proposal is supported by the Proportional Representation Society of Australia and Malcolm Mackerras has written, `The motion should be seen as a correct statement of democratic principle which may help major parties at this election, minor parties at that election.' [14]

The final issue I want to raise briefly is the evergreen of four-year terms. The logic in favour of such a change is overwhelming, coupled with a requirement for simultaneous House of Representatives and Senate elections. Three-year terms are the maximum under our Constitution. In fact, the average term of government since the introduction of proportional representation in 1948 has been precisely two years. This mitigates against stability and maximum productivity. It is wasteful, both of energy and dollars. We can ill afford such constant change. The vast majority of countries with democratic systems of government have four- or five-year terms. All state legislatures, with the exception of Queensland, have changed to a four-year system. The 1988 Constitutional Commission was just the latest expert body—the first being a Royal Commission in 1929—to recommend the extension of the three-year term to four years. It is time we revisited this issue.

I would submit that the Australian system of government, including the electoral system that supports it, is generally regarded as fair, effective and acceptable. The system of proportional representation introduced in 1948 largely reconciles the conflicts between democratic representation and the institutional design of the Senate, as a states' house with powers almost equal to those of the people's house. The important thing, in contemplating any change to these systems is, as I have noted, whether any adaptation of the system can preserve the benefits while alleviating problems. I am confident that all three proposals that I have put forward meet this test.


[1] Margaret Healy and Gerard Newman, `An Electoral Threshold for the Senate', Research Note Number 19, 1998–99, p. 1.

[2] CPD (Commonwealth Parliamentary Debates), 3 May 1939, pp. 57–70.

[3] Fred Daly, From Curtin to Hawke, Melbourne, Sun Books, 1984, pp. 51–52.

[4] Daly claims Calwell won over Caucus by convincing Labor senators they would be re-elected in 1949 `and that the new voting system favoured them in the future.'

[5] Australian, 29 December 1998.

[6] Laura Tingle, `Showdown at the Senate', Weekend Australian, 5 March 1994.

[7] Letters, Canberra Times, 13 July 1999.

[8] Quota Notes, March 1999.

[9] Constitutional Commission, Final Report, vol. 1, AGPS Canberra, 1988.

[10] `The Founders' Design and Intentions Regarding Responsible Government', in P. Weller and D.Jaensch, Responsible Government in Australia, Richmond, Vic., Drummond Publishing, 1980.

[11] Helen Coonan, `Dysfunctional Senate a handbrake on democracy', Sydney Morning Herald, 4 February 1999.

[12] Kim Beazley, Address to the Local Constitutional Conventions Forum, Canberra, 28 April 1999.

[13] CPD, 13 May 1998.

[14] Letters, Canberra Times, 25 May 1998.