A paper presented to the eighth conference of this Society in 1997 advanced an exposition of the intention of the framers of the Constitution in establishing the Senate.1 That intention was expressed in quoted statements by leading framers to the effect that, by representing the people of each of the States equally, the Senate would require a double majority in the legislature to pass any laws: a law would be passed only with the support of a majority of the people and a majority of the people of a majority of the States, both speaking through their elected representatives. This applied to ordinary laws the same formula as is applied to changes to the Constitution by referendum under s.128. In short, the Senate was to ensure that the legislative majority would be geographically distributed, and that it would be impossible to form a majority in the legislature from only one or two regions. Without that safeguard, it would be possible for the legislative majority to rest on the votes of Sydney and Melbourne, to the detriment of other parts of the country.
This theory of the geographically distributed majority was contrasted with the current common and facile treatment of the Senate, which is along the following lines: the framers of the Constitution intended that Senators vote by States and according to the interests of their States, and as this has never happened, the intention of the framers has never been realised. This erroneous but commonplace pseudo-analysis is refuted by even the most casual reference to what the framers actually said.
On that basis, it was suggested in the 1997 paper that the Senate has fulfilled the framers' intention, in that it has been impossible to form a legislative majority from only one or two States, or from Sydney or Melbourne, and the formation of geographically distributed majorities has avoided the extreme alienation of the outlying parts of the country such as has been evident in Canada.
The emergence of the two-party system, with highly cohesive parties divided by class and ideology, did not altogether overthrow this effect of equal representation in the Senate, but, combined with first-past-the-post and later simple preferential voting, produced lopsided party results in the Senate, greatly at variance with the actual pattern of party voting by the electors.
The cure for this situation was proportional representation, introduced in 1948, which has now been in effect for more than half the life of the Constitution. It was suggested in the earlier paper that proportional representation enhanced the formation of a geographically distributed majority in the Senate by providing an ideologically distributed majority. Given that the electors vote for parties, proportional representation ensures that parties achieve representation nearly in proportion to their share of the electors' votes. In that sense, Senate elections produce a more representative result than elections for the House of Representatives. In that House, single-member constituencies result in the elimination of minority parties that may attract up to 20 percent of electors' votes, and in parties winning majorities, and thereby winning government, with less than 50 per cent of the electors' votes, and often with fewer votes than their major rivals, before and after the distribution of preferences.2
There is an historical difficulty with this thesis that proportional representation complements the geographically distributed majority. No one raised this difficulty in 1997. It could have been raised by the simple question: why did Richard Baker oppose proportional representation?
Richard Baker was a leading framer of the Constitution, who participated in the Conventions of 1891 and 1897-8, and who later served a six-year term as a Senator and as President of the Senate. He was the leading exponent of equal representation of the people of the States in the Senate, of the equality of power between the Senate and the House of Representatives, and of the theory of the geographically distributed majority. When the then government attempted, in 1902, to introduce proportional representation for Senate elections, however, Baker vigorously opposed this measure, taking the unusual step of speaking against it from the Chair in the Senate.3 He opposed proportional representation because it would produce "cranks or faddists", what we would call single-issue candidates, rather than representatives of a broader range of public concerns. He defended the then existing electoral system on the basis that it was better to have a two-party result, with representatives dividing basically between liberals and conservatives.
This defence of a two-party system by a leading exponent of equal representation of the States in the Senate surprises us. How could the Senate perform the role he envisaged for it with a two-party system? Was not the two-party system the very problem which led to the ultimate introduction of proportional representation? The answer to this conundrum is that the parties envisaged by Baker were not the parties of ten years later and for most of the Commonwealth's subsequent history. His parties were perfectly consistent with the constitutional role of the Senate. There was no inconsistency between Senators dividing into two parties and their representing the people of their States by ensuring that the legislative majority was geographically distributed.
This draws attention to the greatest change which overcame the Commonwealth after its founding, which would have disturbed the men of 1901, and did disturb those of them who survived to see it. The situation whereby electors vote for nationally-organised parties and for national leaders, often with little knowledge of local candidates, and the parties vote as blocs in Parliament, bound to the party line on every question, radically changes the system of government.
In the House of Representatives, this development put an end to "responsible government" as it was understood by the founders, in that governments and Prime Ministers came to control the House through their assurance of a controlled party majority. Governments became accustomed to using their control of the House to suppress all legislative activity in that forum.
We have come to regard this situation as normal, and no longer think it remarkable. Government legislation is pushed through without amendment, debate is curtailed, no serious inquiry into government activity is permitted if it would be even remotely politically embarrassing. Responsible government as it was understood in the 1890s survived only in the Senate, and only to the extent that the Senate, free of a government party majority, was able to hold the government accountable, however fitfully. (In other places a distinction has been drawn between governments being responsible (liable to be dismissed) and accountable (required to give account of their activities),4 but here the two closely-related concepts are merged.)
Proportional representation somewhat restored responsible government by making it more likely that governments would lack a Senate majority and therefore would be held accountable in the legislature. On the other hand, proportional representation greatly strengthened national party machines because it is essentially a system based on voting by party through party tickets. Thus, the statement that the Senate more accurately represents the voting pattern of the electors by States should be reformulated: the Senate represents more accurately the party choices of the people of the States.
In this connection it is necessary to make two observations about the nature of parliamentary scrutiny and control in the Australian parliamentary system as it now operates. Governments are accustomed to the total support of their party members, at least in public (and perhaps not only in public: it appears that government party meetings have in recent decades become somewhat like meetings of the House of Representatives so far as the government party is concerned, in that dissent tends not to be openly expressed in that forum, but is relegated to "private discussions".) Scrutiny of legislation, amendment or rejection of legislation, and vigorous inquiry into government activities usually comes only from other parties, and in the other House on which the government does not depend for its tenure of office. It is therefore easy for governments to fall into the habit of regarding all parliamentary scrutiny as simply the machinations of their political enemies.
In conducting themselves in this fashion, governments are their own worst enemies. The absence of parliamentary responsibility and the attribution of scrutiny to political opponents means that governments' mistakes go uncorrected and accumulate until their responsibility to the electorate finally hits them.
It should be emphasised that the manifestations of legislative scrutiny and control may be useful to the public, and also to the government itself, even where they are the work of the government's political opponents and rivals. Indeed, it is possible to make out a strong case that legislative scrutiny and control is all the more effective and desirable because it is the work of political opponents and rivals, and that governments benefit by these activities coming from that source, but it would be even more beneficial if government backbenchers would participate.
Secondly, in this situation the public relies on oppositions and minor parties always being champions of accountability. At least at the federal level and in the Senate, this is usually the case. There is, however, a sort of institutionalised hypocrisy involved. Major parties which aspire to gain majorities and form governments are champions of accountability when in Opposition, but resist accountability to the death when in government. Minor parties and independents are able to enjoy consistency by favouring accountability all the time. The public can only hope that this state of affairs continues. If an Opposition, intent on protecting itself in office in the future, combines with a government to suppress accountability, the public is in danger. If the major parties regularly coalesce in this way, bad government results.
Another point, foreshadowed at this stage, is in the form of a question: by conducting themselves in this way, are the major parties encouraging electors to vote for minor parties and independents, particularly in Upper Houses?
In another paper on the effects of proportional representation on the parliamentary system, a list was provided of accountability mechanisms established by the Senate over many years.5 The list ranged from the establishment of a committee to scrutinise delegated legislation in 1932 to ensuring that taxation legislation is not made retrospective to a date earlier than a public announcement of the government's legislative intention. It was pointed out that any reasonable assessment of these measures would lead to the conclusion that they are valuable adjuncts to the parliamentary system from the point of view of the public. It was also pointed out that most of these measures had been opposed by the government of the day on their introduction, and had been introduced only because of the absence of a government party majority in the Senate. The list continues to expand: for example, last June the Senate agreed to an Order requiring government departments and agencies to publish details of their contracts, including statements of reasons for any provisions regarded as confidential. This list underlines the point that a measure of parliamentary accountability of government exists only because of proportional representation in the Senate.
It must be said that the lack of a government party majority in the Senate has another consequence. The majority of the Senate may amend or reject legislation, not to improve it, but simply to substitute its own policy preferences for those of the government. It is often difficult to separate quality control over legislation from pure differences on policy, and certainly there is no way of avoiding the latter without removing the former and the desirable effects of parliamentary scrutiny and control. The earlier paper also exposed the fallacy of those who think that the Senate should be able to scrutinise without rejecting or delaying government legislation: the point was made that a scrutinising body without legislative power would simply be ignored and its scrutiny would be ineffective.
The major consequence of proportional representation has therefore been to disguise the death of responsible government. By depriving governments of their controlled party majorities in the Senate, it has made the Senate the sole legislative House, where legislation is the subject of something approaching real deliberation. It has also partially restored the responsibility of the executive government to the legislature. In the absence of the Senate, the central executive government could legislate by decree, with results which have been only too well demonstrated in other jurisdictions.
The full effects of proportional representation, however, are still not fully appreciated. It has produced another unintended consequence, also contrary to the intention of the framers of the Constitution. It has resulted in a reversal of the intended representational roles of the two Houses of the Parliament.
The House of Representatives was intended to represent most directly the voice of the electors as most recently expressed. Changes in the opinions and choices of the electors would be swiftly and directly conveyed through the House of Representatives. The Senate would reflect the opinions and choices of the electors with a built-in delay because of the six-year term of Senators and their election by rotation, with half the Senate turning over at the end of each parliamentary term. By representing whole States, the Senators would also give expression to the more considered opinions and choices of the people of the whole Commonwealth. This view of the respective roles of the Houses was in accordance with the classic expositions of the rationale of bicameralism and Upper Houses.
As a result of proportional representation, however, shifts in opinion in the electorate are reflected more quickly and more accurately in the Senate than in the House. The Senate has become the House representing the passion and turbulence in the electorate, while the House filters out that passion and turbulence. This has been demonstrated also in State Legislative Councils elected by proportional representation. Surprisingly, this effect is largely ignored by the commentators and pundits.
Some of those very commentators and pundits tell us that there has been a long-term loss of voter loyalty to the major parties and a drift to minor parties and independents. They concentrate largely on the effect this will have on the formation of governments in the House of Representatives. The point that this trend, if it is a trend, will be most starkly reflected in the Senate largely escapes notice, although the presence in the Senate of a One Nation Senator should draw attention to this phenomenon.
The imminent general election may provide an illustration of both the phenomenon and the neglect of it. Whether or not there is an abandonment by the electors of the major parties in favour of a spectrum of minor parties and independents, it is entirely predictable that the concentration of comment on election night will be on who has won government in the House of Representatives. It is also predictable that the winners will be winners in name only, in the sense that the winning party, if there is one, will achieve significantly less than 50 per cent of electors' votes, and quite possibly fewer votes than their major rivals, before and after the distribution of preferences.
The turbulence in the electorate, if it occurs, will be reflected more directly and urgently in the composition of the Senate. It is possible that no party will win a majority in the House, but it is far more likely that the electoral system for the House will continue to do what it has done for the past ninety years or so: produce a winner, even with a minority of votes. Some weeks after the election, attention will turn to the Senate, and the real position of the winners, in the Parliament and in the electorate, will be appreciated.
This kind of result, if it occurs, is certain to make legislative life much more difficult for the "winners" in the House of Representatives. We are accustomed to the situation of a government not being guaranteed the passage of its legislation, having to compromise with other parties, and having its activities more rigorously examined than would otherwise be the case. It is not generally appreciated that this occurs because proportional representation lends legitimacy to these activities in the Senate. They are legitimised by the composition of the Senate more accurately reflecting the votes of the electors. It is also not generally appreciated that this effect is in proportion to the disparity between the representativeness of the two Houses. If a government's majority in the House is clearly unrepresentative of the electors' votes, the more representative majority in the Senate is less likely to defer to that government's authority. Recent governments have been weakened by "winning" elections without even a plurality of votes.
If there is a desertion of the major parties in the forthcoming election, as some predict, a non-government majority in the Senate may be emboldened to make even greater use of its legislative powers. We could have a situation in which the loss of votes by the major parties is such that a government formed by either side is quite delegitimised, its authority reduced to nought. The deadlock-resolving provisions in section 57 of the Constitution are of little comfort to a government in this condition.
It is possible to take a Panglossian view of this evolved arrangement, and conclude that perhaps it works in the end. The House of Representatives may be seen simply as an electoral college, which produces a government in much the same rough and ready way as the American electoral college. The Senate produces a representative legislature and provides some measure of accountability of the executive government.6
There is much to be said for this interpretation. We are troubled, however, by the voice of President Baker speaking to us from our past. Is it healthy for political parties to become closed sects, even if there are many to choose from? Would it not be sounder for parties to represent the real diversity of people and opinions within their ranks? Should they not also allow their members to perform their representational role and speak for their constituents more freely? Would people prefer to be represented by parties of that kind? Baker's implied questions are still pertinent.
This leads to the question foreshadowed above: is the currently accepted modus operandi of major parties actually undermining their electoral support? By maintaining their show of monolithic unity, pretending that they have the only true answers to all questions, while those of their opponents are uniformly wrong, resisting or suppressing parliamentary accountability measures, and controlling their members so rigidly, the parties may be contributing to a drift of votes to minor parties and independents, and encouraging voters to "split" their votes between the two Houses.
The 1997 paper concluded with the view that there is nothing basically wrong with the institutions of government, and that what is required is reform of the political parties. There is a stubborn refusal to think this thought. If the kind of election result which has been described eventuates, it is also predictable that the result in the Senate will be seen as anomalous, and as demonstrating the need for change in the structure of the Parliament, rather than as an accurate reflection of the electors' opinions and choices and as a reflection on the party system. Some people, however, may draw the lesson that it is time for the parties to reform themselves, so that they can reflect the diversity of the country and the considered and stable underlying views of the people, as Baker thought that they ought to be able to do in 1902.
A first step towards that reform would be for governments to abandon the pretence of infallibility and total unanimity on all points great and small, and to allow their own members to examine and deliberate on legislation, to make amendments, and to inquire into government administration. Both Houses could then fulfil their intended role of making good laws, scrutinising government and holding ministers accountable. This could make governments more effective and the electors less likely to vent their frustrations at the ballot box.
1. Federalism and the Role of the Senate, in Upholding the Australian Constitution, Proceedings of The Samuel Griffith Society, Volume 8 (1997), pp.125-137.
2. Figures for percentages of votes and percentages of seats won by parties in elections since 1949 are in Odgers' Australian Senate Practice, 9th ed., 1999, pp.23-26. The winning party in the House of Representatives received fewer first preference votes than the other major party in 3 elections since 1949, in 1954, 1987 and 1998, and in 5 elections the winner had fewer votes than the losers after the distribution of preferences, in 1954, 1961, 1969, 1990 and 1998.
3. Senate Debates, 19 March, 1902, pp.11007-11010.
4. In the item cited in note 5.
5. Accountability Versus Government Control: the Effect of Proportional Representation, in Representation and Institutional Change: 50 Years of Proportional Representation in the Senate, ed. M Sawer & S Miskin, Papers on Parliament No. 34, December, 1999, pp.71-78.
6. Such an interpretation is offered by D Hamer, Can Responsible Government Survive in Australia?, 1994.