The Senate and representation
The framers of the Constitution determined that the Senate would best operate if it were directly elected by the people of the states. It was suggested at that time that the best method of election would be proportional representation, which is designed to ensure that representatives are elected in proportion to their support among the electors. This system was not written into the Constitution, however; instead it was left to the Commonwealth Parliament to determine the actual method of election. The system of proportional representation, which, as was suggested when the Constitution was drawn up, is the logical method for electing representatives of a large area such as a state, was not adopted until 1948, taking effect in the elections of 1949.
The Senate by its constitutional design enlarges the Parliament’s capacity to represent the diversity of the Australian people by providing a balance to the numerical preponderance of the more populous states in the House of Representatives. As a consequence of the 1948 proportional method of electing senators, it does so in a fashion which more accurately reflects the state of electoral opinion in the nation. It corrects dysfunctions of the single member electoral system used for choosing the House of Representatives and thereby provides parliamentary representation for individuals and parties with significant voter support, which would be otherwise unrecognised in parliamentary terms except where such support is geographically concentrated.
The important role which the method of electing senators has in enhancing the representative capacity of the Commonwealth Parliament may be seen in the information in Table 1, which demonstrates that the party composition of the Senate almost invariably reflects the party disposition of voting in the electorate more closely than does the House of Representatives. As already observed, one effect of the Senate method is to remedy explicit deficiencies in the single member electorate system used for electing members of the House of Representatives.
Table 1 sets out, in abridged form, information concerning the relationship of percentage of the vote to percentage of seats in the Senate and the House of Representatives respectively for elections since 1949. While a direct correspondence between percentage of the vote and percentage of seats is rare, it is clearly the case, for almost all elections, that the correspondence between percentages of votes and of seats is closer in the Senate than in the House of Representatives. Moreover, it is almost never the case that the correspondence in the House of Representatives is closer than in the Senate.
The electoral system of the House of Representatives regularly awards a majority of seats, and government, to parties which secure only a minority of electors’ votes, occasionally less than 40 percent, and on several occasions less than those of the major losing parties.
Table 1 suggests that, in a House of Representatives election, the imbalance between percentage of votes and seats is most marked in what is known as a “landslide” victory. In 1958, for instance, the Australian Labor Party (ALP) received 42.8 percent of the vote in the Senate election and 42.9 percent in the House election. In that election, the ALP secured 46.9 percent of the Senate places at issue, but only 37.9 percent in the House. Again, in 1975, 40.9 percent of the Senate vote secured 42.2 percent of the Senate places for the ALP; a higher percentage of the vote in the House of Representatives, 42.8 percent, brought the ALP only 28.4 percent of seats in the House. Confirming the propensity of the House of Representatives method of election to exaggerate majorities, in 1983 a 49.5 percent share of the House vote yielded 60 percent of the seats for the ALP; in the same election, 43.6 percent of the vote for the Liberal and National parties brought a 40 percent share of the seats in the House. In the Senate, an ALP share of 46.9 percent of places in the Senate reflected a 45.5 percent of the vote; in this case, the Liberal and National parties’ 39.9 percent of the vote brought 43.8 percent share of places in the Senate. In their “landslide” victory of 1996, the Liberal and National parties secured 63.6 percent of the seats in the House with 47.3 percent of the vote; in the Senate their 44 percent of the vote delivered 50 percent of seats. In 1998 the Liberal and National parties secured a majority in the House with less than 40 percent of the votes and fewer votes than the Labor Party; in the Senate their votes were more accurately reflected.
Complaints by governments that proportional representation makes it impossible for the winning party to secure a majority in the Senate were refuted by the 2004 election, in which the Liberal and National parties secured a Senate majority of one with 45 percent of votes, while their majority in the House was again exaggerated. Those majorities were lost in the 2007 election, when the Senate results again produced a more balanced outcome.
The state basis of Senate elections does not significantly exaggerate representation in the Senate. While there are cases where election of a single senator brings a measure of exaggeration, it is usually the case that the share of places secured by minor parties is less than their share of the vote. In the case of the Australian Democrats, it was only in 1984 that the reverse was conspicuously the case (a 7.6 percent share of the vote brought a 10.9 percent share of seats). In 1975 a one percent share of the vote brought the Liberal Movement one seat, that is, 1.67 percent of the places. In the 1990, 1993 and 1996 elections for the Senate, Green shares of the vote, 2.8, 2.9 and 2.4 percent respectively, brought 2.5, 2.5 and 2.5 percent shares of the seats contested. In 1998, 2001, 2004, 2007 and 2010 the minor parties generally were underrepresented, but still more accurately represented in the Senate than in the House. It thus appears that even the divergence of the populations of the various states and territories does not have a significant effect on the national representivity of the Senate.
A very clear example of the capacity of the Senate system to improve representation in the Commonwealth Parliament is party representation of Tasmanians. In the period from the simultaneous dissolutions of 1975 to the general election for the House and the Senate in 1987, notwithstanding a party share of the vote of from 40.3 percent (1983) to 45.1 percent (1980), no candidate endorsed by the Australian Labor Party for a House seat was successful. In the same period there were 4 to 5 Labor senators from Tasmania. In 1998, 2001, 2007 and 2010, this situation was reversed, with Tasmanian Liberal Party voters unrepresented in the House.
More generally, the Senate has provided opportunity for parliamentary representation for parties, groups and individuals enjoying significant voter support which goes unrecognised in the single member electorate system by which members of the House of Representatives are chosen. These include the Democratic Labor Party from 1955 to 1974 and again from 2011, the Liberal Movement (1974-81), the Australian Democrats (1977-2008) and the Greens.
The effect of proportional representation on the representative character of the Senate is also illustrated by Table 2, which shows party affiliations in the Senate since 1901.
The representative character of the Senate has enabled it to uphold the responsibility of governments to Parliament. Much of the traditional doctrine on this question of responsibility derives from a period before the emergence of rigid parties and disciplined majorities within Parliament, most conspicuously in lower houses, the control of which is the condition of a ministry taking and maintaining office. In Australia this issue has added importance because there are few other national legislatures in which party voting is so disciplined as it is in the House of Representatives. This being so the need for alternative parliamentary avenues for holding a government to account is pronounced, and this need in Australia is supplied by its elected Senate. Since 1949 there have been only four relatively short periods (1951-56, 1959-62, 1976-81, 2005-07) in which a ministry has had a majority in the Senate. Conversely, the Opposition party in the House of Representatives, irrespective of its partisan complexion, has not had a majority in the Senate (with the exception of 1949-51, in unusual circumstances, following change of government, in 1974-75 and, in 2007-08). Accordingly, it does not follow that a ministry lacking a secure majority in the Senate is automatically confronted by a hostile Opposition majority. Any attempt by an Opposition to achieve its partisan ends by use of its numbers in the Senate must, to succeed, have the support of other non-government senators. The Senate when functioning as a repository of and forum for responsibility is thus more than a mere venue for a clash between government and Opposition working on the basis of pre-determined numbers. Governments have therefore been held to account in the Senate more effectively than in a house where they are always supported by a party majority.
A decline of accountability accompanying ministerial control of both Houses of the Parliament may well in the long run be adverse to governments themselves as well as to the country generally. This was the lesson that many drew from the fall of the then government in 2007 after its period of majority in the Senate gained in the 2004 elections.
All free systems of government need checks and balances against any excessive concentration of power and, so far as the Australian system is concerned, the Senate is the most important of the constitutional checks and balances, the more so because it is an elected institution. Lack of control of the Senate can no doubt be inconvenient to a government and at times frustrating, but such considerations are secondary to the greater good of responsible checks and balances exercised by a second chamber elected by universal adult franchise and closely reflecting the diversity of electoral opinion in the nation.