![]() |
Home
| Databases
| WorldLII
| Search
| Feedback
Upholding the Australian Constitution: The Samuel Griffith Society Proceedings |
Thoughts on the 1949 Reform of the Senate
Malcolm Mackerras
Towards the end of this year we will commemorate a fiftieth anniversary which has so far largely escaped notice. To be precise, on Friday, 10 December, 1999 we will commemorate the fiftieth anniversary of the general election held on Saturday, 10 December, 1949. That was the election which began the Menzies years. It was, just as importantly, the first election with a new Senate electoral system.
The system in question is commonly called “proportional representation”. Until quite recently even a pedant like me used to call it that. However, further thought I have given to this subject causes me to revise my language. I should have been describing the system in two sentences, not one. The two sentences are: (1) When the most recent Senate election was double dissolution, the system might be described as one of PR (proportional representation). (2) However, when the most recent election was half-Senate, the system should be described as semi-proportional.
It was the result in 1998 which finally brought me to that description. There are now nine Senators for the Australian Democrats party. By contrast, there is only a single Senator for Pauline Hanson’s One Nation party. (That sentence assumes there will be such a Senator. At the time of writing Mrs Heather Hill has been thrown out by the High Court, and we are all assuming that the second One Nation candidate, Len Harris, will take her seat). If you were to tell a South African, an Israeli or a Dutchman (people accustomed to PR) that the Australian Senate is elected by a system of PR, he would assume that the 1998 election had given nine votes to the Democrats for every one given to One Nation. When you told him that One Nation had polled 1,007,436 Senate votes (9 per cent), while the Democrats had polled only 947,938 (8.5 per cent), he would quickly tell you that the system is not one of PR. It is, at best, semi-proportional.
Armed with the realisation above, I have gone through all the post-1949 statistics which produce this opinion. I now understand that my description in the second paragraph above should always have been the correct description. Over the entire period we should have been calling the system semi-proportional except when a double dissolution takes place. Only then could we rely on a proportional result.
The 1949 reform was the result of a piece of ordinary legislation which passed through the Commonwealth Parliament in 1948. At the time, debate was dominated by the short-term consequences of the new system. Looking back over 50 years, however, we can now give a verdict on the behaviour of the Senate. It is given in Table 1. Note that there have now been more parliaments under “PR” than there were under previous non-proportional Senate systems from 1901 to the election of the 18th Parliament in 1946.
Date of Electiona
|
Number of Parliament
|
Prime Minister
at Opening
|
Date of Dissolution
|
Hostile Senateb
|
|
|
|
December 1949
|
19th
|
Menzies (Liberal)
|
March 1951
|
December 1972
|
28th
|
Whitlam (Labor)
|
April 1974
|
May 1974
|
29th
|
Whitlam (Labor)
|
November 1975
|
October 1980
|
32nd
|
Fraser (Liberal)
|
February 1983
|
December 1984
|
34th
|
Hawke (Labor)
|
June 1987
|
Government Senate Majority
|
|
|
|
April 1951
|
20th
|
Menzies (Liberal)
|
April 1954
|
May 1954
|
21st
|
Menzies (Liberal)
|
November 1955
|
November 1958
|
23rd
|
Menzies (Liberal)
|
November 1961
|
December 1975
|
30th
|
Fraser (Liberal)
|
November 1977
|
December 1977
|
31st
|
Fraser (Liberal)
|
September 1980
|
Government de Facto Senate Control
|
|
|
|
December 1955
|
22nd
|
Menzies (Liberal)
|
October 1958
|
December 1961
|
24th
|
Menzies (Liberal)
|
November 1963
|
November 1963
|
25th
|
Menzies (Liberal)
|
October 1966
|
November 1966
|
26th
|
Holt (Liberal)
|
September 1969
|
October 1969
|
27th
|
Gorton (Liberal)
|
November 1972
|
March 1983
|
33rd
|
Hawke (Labor)
|
October 1984
|
July 1987
|
35th
|
Hawke (Labor)
|
February 1990
|
March 1990
|
36th
|
Hawke (Labor)
|
February 1993
|
March 1993
|
37th
|
Keating (Labor)
|
January 1996
|
March 1996
|
38th
|
Howard (Liberal)
|
August 1998
|
(a) The date of the election is that for the House of Representatives, which was also usually a Senate election date. The government of the day at all times enjoyed a House of Representatives majority and, therefore, total control of the Lower House.
(b) A “hostile Senate” is defined as one which was dissolved. Thus the date of dissolution column records the date of the double dissolution. In respect of the rows for “Government Senate Majority” and “Government de Facto Senate Control” the dissolution date is for the House of Representatives only. It is important to note that there were only five double dissolutions but there were 15 dissolutions for only the Lower House.
An opinion is now conventional that it is a dreadful thing that only five of the 20 Parliaments in Table 1 saw the government of the day with a Senate majority. I dissent from this “dreadful thing” view. To justify my dissent I give two further tables.
Table 2 sets out the percentages at those elections when a “big party” received a primary vote in excess of 49 per cent. I regard a 49 per cent support as high enough to be called a “majority” of the vote.
Party
|
1949
|
1951
|
1953
|
1975
|
Coalition
|
50.4
|
49.7
|
44.4
|
51.7
|
Labor
|
44.9
|
45.9
|
50.6
|
40.9
|
It will be noticed that two of the four entries above failed to give a majority of Senators to the party with the majority of votes. Labor continued to have a majority in the Senate in the Menzies 19th Parliament. By contrast, Menzies had a Senate majority throughout the whole of the 20th and 21st Parliaments – even though Labor won 50.6 per cent of the vote at the May, 1953 separate election for half the Senate. In all three cases the reason has to do with the rotation of senators.
By contrast, the Coalition’s 49.7 per cent in 1951 and 51.7 per cent in 1975 did produce a Senate majority in the 20th, 21st, 30th and 31st Parliaments. The reason has to do with the proportionality of the result when the most recent election was double dissolution, and the semi-proportionality of the result when the most recent election was half-Senate.
The continuing refusal of the Australian people to vote for the Government of the day in the Senate election has produced this consequence. The normal situation is that the government has de facto control of the Senate without enjoying an actual majority. While we may disagree with the decision of the Senate to reject this or that piece of government legislation, it is difficult to object on democratic grounds. In defence of my view, I set out in Table 3 the percentage of the vote secured by the Government (which in 1983 and 1996 entered the election as the Opposition party) at the Senate election relevant to the main life of that Parliament. Note that in 1964, 1967 and 1970 the election in question was a separate election for half the Senate.
Table 3: Selected Government Senate Percentages (Parliaments where Government had de facto Senate Control)
Election Date
|
Party
|
Percentage
|
December 1955
|
Coalition
|
48.7
|
December 1961
|
Coalition
|
42.1
|
December 1964
|
Coalition
|
45.7
|
November 1967
|
Coalition
|
42.8
|
November 1970
|
Coalition
|
38.2
|
March 1983
|
Labor
|
45.5
|
July 1987
|
Labor
|
42.8
|
March 1990
|
Labor
|
38.4
|
March 1993
|
Labor
|
43.5
|
March 1996
|
Coalition
|
44.0
|
While a government may, perhaps, reasonably object to Senate rejection of legislation by a genuinely hostile Senate (of which there have been only five cases in 20 Parliaments after 1949) it is difficult to see how any government could reasonably object to Senate rejection of legislation during any of the Parliaments covered by Table 3. It is worthwhile noting that three of those Parliaments met the technical conditions of s.57 of the Constitution – as is set out in Table 4. It is also worth noting that none of Menzies, Hawke or Howard could justify to himself the pulling of the “trigger” he had established for himself in the 22nd, 33rd and 38th Parliaments. That is why the subsequent election was of the conventional type, that is to say, for the House of Representatives and half the Senate.
The assertions above about the semi-proportionality of half-Senate elections appear to be based on disproportionality caused by the rotation of Senators. It is true that this was the point which was noticed in the early days of the system. As we shall see, the 1949 Senate election was not of itself disproportional. Rather, the combination of Senators elected in 1946 and 1949 produced the dreadful result that a government which had achieved over 50 per cent for both houses in 1949 was badly outnumbered by the official Opposition party in the Senate. A double dissolution thus became inevitable in 1951.
These points about proportionality may seem rather pedantic but they are worth making in the light of comments made below. They raise two questions. First, when does an election result cease to be proportional and become semi-proportional? Second, when does an election result cease to be semi-proportional and become non-proportional?
The answer lies in a device of modern psephology known as
Table 4: Parliaments Which Have Met Conditions of Section 57
Number of
Parliament
|
Date of Election
|
Prime
Minister
|
Term of
Office
|
Date Conditions of s.57 First Met
|
s.57 Bills
|
Date of Dissolution
|
Next Election
|
Length of Parliament
|
5th
|
31 May 1913
|
Cook (Liberal)
|
First
|
28 May, 1914
|
1
|
30 July, 1914
(double)
|
5 September 1914
|
1 year 21 days
|
19th
|
10 December 1949
|
Menzies
(Liberal)
|
First
|
14 March, 1951
|
1
|
19 March, 1951
(double)
|
28 April 1951
|
1 year 25 days
|
22nd
|
10 December 1955
|
Menzies
(Liberal)
|
Fourth
|
27 March, 1958
|
14
|
14 October, 1958
(single)
|
22 November 1958
|
2 years 8 months
|
28th
|
2 December 1972
|
Whitlam
(Labor)
|
First
|
29 August, 1973
|
6
|
11 April, 1974
(double)
|
18 May 1974
|
1 year 1 month 15 days
|
29th
|
18 May 1974
|
Whitlam
(Labor)
|
Second
|
11 December, 1974
|
21
|
11 November, 1975
(double)
|
13 December 1975
|
1 year 4 months 2 days
|
32nd
|
18 October 1980
|
Fraser
(Liberal)
|
Third
|
10 March, 1982
|
13
|
4 February, 1983
(double)
|
5 March 1983
|
2 years 2 months 10 days
|
33rd
|
5 March 1983
|
Hawke
(Labor)
|
First
|
14 June, 1984
|
2
|
26 October, 1984
(single)
|
1 December 1984
|
1 year 6 months 5 days
|
34th
|
1 December 1984
|
Hawke
(Labor)
|
Second
|
2 April, 1987
|
1
|
5 June, 1987
(double)
|
11 July 1987
|
2 years 3 months 16 days
|
38th
|
2 March 1996
|
Howard
(Liberal)
|
First
|
25 March 1998
|
4
|
31 August, 1998 (single)
|
3 October 1998
|
2 years 4 months
|
the Gallagher least squares index of disproportionality. Named after the Irish political scientist Michael Gallagher, it measures in a single statistic deviations from proportionality.
Election
|
House of Representatives
|
Senate
|
1949
|
7.50
|
3.42
|
1951
|
5.36
|
3.03 (full)
|
1953
|
–
|
3.29
|
1954
|
2.88
|
–
|
1955
|
6.84
|
6.52
|
1958
|
11.05
|
6.18
|
1961
|
7.12
|
9.75
|
1963
|
9.00
|
–
|
1964
|
–
|
2.06
|
1966
|
10.83
|
–
|
1967
|
–
|
3.80
|
1969
|
6.95
|
–
|
1970
|
–
|
3.16
|
1972
|
6.90
|
–
|
1974
|
5.96
|
3.72 (full)
|
1975
|
14.05
|
3.08 (full)
|
1977
|
15.02
|
7.30
|
1980
|
8.46
|
1.51
|
1983
|
10.41
|
3.37 (full)
|
1984
|
7.82
|
5.35
|
1987
|
10.41
|
2.60 (full)
|
1990
|
12.49
|
4.39
|
1993
|
8.06
|
3.33
|
1996
|
11.24
|
4.54
|
1998
|
11.85
|
7.48
|
The index was 17.69. That composite statistic was calculated from a National Party over-representation of 15.46 (50.51 per cent of seats for 35.05 per cent of votes), a Labour Party over-representation of 10.77, an Alliance under-representation of 16.19, a New Zealand First under-representation of 6.38, etcetera.
South Africa has the most proportional system in the world today. At the June, 1999 general election the Gallagher least squares index of disproportionality was only 0.16. For example, the African National Congress won 266 of the 400 seats in the National Assembly, which is 66.50 per cent of the seats. It had 66.36 per cent of the votes. The New National Party won 28 seats, exactly seven per cent of the 400 seats. It won 6.87 per cent of the votes; and so on.
Table 5 gives the equivalent Australian figures. Double dissolution elections are shown with the word (full) by their side indicating an election for the whole Senate. It should be noted that only in 1961 did the Senate election result turn out to be less proportional than that for the House of Representatives. It was a half-Senate election. My judgment is that an index of less than four constitutes a proportional result, while an index of more than ten constitutes a non-proportional result. Anything in between is semi-proportional. Thus I say that only when the whole Senate is elected can the system be said to be one of PR. Notice that the index was 3.03 in 1951, 3.72 in 1974, 3.08 in 1975, 3.37 in 1983 and 2.60 in 1987.
The title of this paper is Thoughts on the 1949 Reform of the Senate. It is to be noted that the reform was implemented by a Labor government and that the Liberal Party was initially critical of significant aspects of the new system. However, from 1951 onwards the Menzies Government became quite comfortable with the new arrangements. Indeed, why would it not be comfortable? From 1951 onwards Menzies had a majority in both Chambers.
Reference to “the 1949 reform” may, perhaps, give the impression that I am unaware of two subsequent changes. Not so. I am very much aware of them. It is just that I do not regard either the changes of 1975 or those of 1984 as being of any significance. Having never argued for the reversal of the 1949 reform, I have also never argued for the reversal of the changes of either 1975 or 1984, both of which were also implemented by Labor governments, and opposed by the Liberal Party.
The 1975 change simply added two Senators from each of the Australian Capital Territory and the Northern Territory to the existing ten Senators from each State. That had the effect of increasing the total size of the Senate from 60 to 64. Territory Senators are peculiar in that they do not rotate. Consequently they always have three year terms because their election is always tied to that of the House of Representatives. It could be argued that the addition of these “lower class” Senators dilutes the federal nature of the Senate. The Senators from the States serve terms of six years and do rotate. They are the “upper class” Senators.
I can see the argument that Territory Senators should be removed. Their presence undermines the idea of the Senate as a States’ house. However, it seems to me that the Senate has evolved into a multi-party House – and that is where its value lies. Since the 1975 change did not alter that, I oppose its reversal. In any event I am an elector of the ACT and do not want to lose my Senate vote! Finally, since the High Court has now twice upheld the constitutional validity of Territory Senators, I cannot see any reason why a paper for a conference titled Upholding the Australian Constitution should recommend the abandonment of Senators who have sat quite constitutionally for nearly a quarter of a century.
Where the Whitlam Government had implemented the 1975 change, it was the Hawke Government which changed the rules from the 1984 general election onwards. Henceforth there would be 12 Senators from each State, increasing the total to 76. Future half-Senate elections would be for six Senators, where previously they had been for five. The Liberal Party opposed that change also. One of the arguments it used was that elections should always be for an odd number (say, five or seven), never for an even number (say, six).
There is some logic in the argument. In an odd-numbered election, a party’s vote of 50.1 per cent would give it three out of five or four out of seven. A majority of votes would give a majority of seats. By contrast, if six are elected, one party with 56 per cent of the vote may win three seats while another party with 44 per cent of the vote may also win three. Thus a clear majority of the votes only translates into half the seats.
While acknowledging the logic in the argument, I disagree with it. These are questions of the workability of institutions, not of logic. There is no reason why the Australian Senate should not work with six Senators being elected each time from each State. Indeed, I argue that the Australian Senate has worked just as well since 1984 as it did in the period 1949-84. Consequently I oppose the proposal to reverse the change of 1984. That is why I always revert to my pedantic point. If it be agreed that the Senate system is, in practice, semi-proportional there is no reason to keep quoting a theoretical PR point about the need for odd numbers. I always suspect that those who say this are either defective in their understanding of these things or have a political axe to grind.
Which brings me to the Liberal Party. The Menzies Government won a Senate majority in its second term, as a consequence of the Coalition receiving 49.7 per cent of the Senate vote in 1951. The Fraser Government had a Senate majority in both its first and second terms, when the Coalition had won 51.7 per cent and 45.6 per cent, respectively, of the Senate votes in 1975 and 1977. Consequently, the Liberal Party quickly got over its railing against the changes of 1949 and 1975. By contrast, the Howard Government has not had a Senate majority in either of its terms. Thus the Liberal Party is still railing against the changes of 1984. It has no basis to do so, as can be clearly seen from Tables 6 and 7.
Prime Minister
|
1st Win
|
Senate % at 1st Win
|
2nd Win
|
Senate % at 2nd Win
|
Decline
|
Robert Menzies (Liberal)
|
December 1949
|
50.4
|
April 1951
|
49.7
|
0.7
|
Gough Whitlam (Labor)
|
December 1972
|
(a)
|
May 1974
|
47.3
|
—
|
Malcolm Fraser (Liberal)
|
December 1975
|
51.7
|
December 1977
|
45.6
|
6.1
|
Robert Hawke (Labor)
|
March 1983
|
45.5
|
December 1984
|
42.2
|
3.3
|
John Howard (Liberal)
|
March 1996
|
44.0
|
October 1998
|
37.7
|
6.3
|
(a) The December 1972 election was for the House of Representatives only.
A “John Howard Prime Minister” is one who had been Leader of the Opposition but led his party out of Opposition to victory at a general election. The Senate record of the five post-War cases is set out in Table 6. One notices the decline in their Senate percentages at their second win. Most of all, however, I am impressed by the magnitude of the Howard decline and the low overall vote, a miserable 37.7 per cent in 1998.
Table 7: Non-Labor(a) Senate Percentages2
Election
|
Vote (per cent)
|
Election
|
Vote (per cent)
|
1910
|
45.6
|
1949
|
50.4
|
1913
|
49.4
|
1951
|
49.7
|
1914
|
47.8
|
1953
|
44.4
|
1917
|
55.3
|
1955
|
48.7
|
1919
|
55.2
|
1958
|
45.2
|
1922
|
52.0
|
1961
|
42.1
|
1925
|
54.9
|
1964
|
45.7
|
1928
|
50.5
|
1967
|
42.8
|
1931 (highest)
|
55.4
|
1970
|
38.2
|
1934
|
53.2
|
1974
|
43.9
|
1937
|
44.8
|
1975 (highest)
|
51.7
|
1940
|
50.4
|
1977
|
45.6
|
1943 (lowest)
|
38.2
|
1980
|
43.5
|
1946
|
43.3
|
1983
|
39.9
|
|
|
1984
|
39.5
|
|
|
1987
|
42.0
|
|
|
1990
|
41.9
|
|
|
1993
|
43.0
|
|
|
1996
|
44.0
|
|
|
1998 (lowest)
|
37.7
|
Average 1910–46
|
49.7
|
Average 1949–98
|
44.0
|
(a) The term Non-Labor means Liberal from 1910 to 1914, Nationalist in 1917, Nationalist-CP from 1919 to 1928, UAP-CP from 1931 to 1943, Liberal-CP from 1946 to 1980 and Liberal-National since 1983.
Table 7 illustrates my thinking best of all. Howard’s Senate vote was poorish in 1996, when it was exactly the long-term Coalition average of 44 per cent. His 1998 Senate percentage was the lowest ever for Non-Labor. The lowest in 34 elections! How dare the Liberal Party complain about the system! In any event, the Coalition presently has 46.1 per cent of the Senate seats when its most recent vote was 37.7 per cent. Under a so-called PR system the Coalition is over-represented by 8.4 per cent.
John Howard himself has repeatedly stated that he is not considering radical changes to the Senate electoral system. However, that has not deterred other Liberals (equally repeatedly) from railing against the present method of electing Senators. Since these people keep proposing so-called “reforms” I should like to consider whether these reforms are constitutional or not. If for no other reason, that duty is imposed upon me by the title of this conference which, let me remind you, is called Upholding the Australian Constitution.
Clearly, a reform which merely takes the Senate back to the pre-1949 period, or to the period 1949-75, or to the period 1975-84, would be perfectly constitutional. There is, in fact, a case for taking it back to the period 1975-84. During that period there were 64 Senators, while the numbers in the House of Representatives ranged from 127 to 124. Thus the argument would be the desirabilty of getting rid of 12 Senators and about 20 or so Members of the House of Representatives. Under the nexus provision of s.24 of the Constitution, the numbers in the lower House “shall be, as nearly as practicable, twice the number of the Senators”.
Given that New South Wales has reduced the size of its Legislative Assembly from electing 109 members in March, 1988 to electing 93 in March, 1999; and given that Tasmania has reduced the size of its House of Assembly from the 35 elected in February, 1996 to the 25 elected in August, 1998, there may well be a case for cutting 20 or so Members out of the House of Representatives. That could easily be done by reducing the size of the Senate from 76 to 64. Plainly that would be a constitutional thing to do.
The trouble is, it would neither make the Senate more workable, nor would it increase the likelihood that a government has a Senate majority. Subsequent to 1949 no Labor government has had a Senate majority. Menzies and Fraser did, Howard did not. Why? The first two Menzies Senate percentages were respectable, 50.4 and 49.7. Fraser’s first two were also respectable, 51.7 and 45.6. To understand Howard’s failure one need do no more than inspect his miserable percentages as set out in Tables 6 and 7.
Having ascertained that the reform consistent with the Constitution (reversion to the numbers of 1975-84) would not solve the Liberal Party’s problem, what about the other proposed reforms? Essentially both are radical and unconstitutional, as I shall explain. Consequently the Liberal Party’s problem is easily explained. It has nothing to do with nefarious changes introduced by the Chifley, Whitlam and Hawke Labor governments. The difference between Menzies and Fraser on the one hand, and Howard on the other, is Howard’s abject failure to persuade significant numbers of Australians to vote for his government at Senate elections. Where Menzies and Fraser succeeded, Howard has failed.
It must be very galling to supporters of the Howard Government to know that its Senate vote of 37.7 per cent has inflated into “only” 46.1 per cent of the Senators. How, then, do you get a decent size of inflation? Answer: Liberal Party luminaries concoct radical schemes of “Senate reform”, and assure gullible journalists that either scheme could be implemented by a simple act of the federal Parliament. When a difficult fellow like Malcolm Mackerras argues that the schemes are unconstitutional, you get your journalist friends to ignore him or put him down.
The first “reform” is being promoted by Senator Helen Coonan of New South Wales. It proposes that a threshold (or hurdle) be inserted at Senate elections. She is suggesting that the threshold be five per cent. Any party which gets less than five per cent would be cut out by the threshold. It could not get a Senator elected in any State in which the party’s vote is less than five per cent.
To people with a smattering of knowledge (but no real understanding) about electoral systems the proposal sounds quite reasonable. It is actually outrageous, and unconstitutional. The plausibility of it comes from saying that successful countries with PR, such as Germany and New Zealand, have five per cent thresholds – so should Australia. We are invited by the argument to think like this: “New Zealand is very like Australia, let us copy New Zealand. Germany is a very successful federation, let us copy Germany”.
There is a simple reason why I do not accept the argument. I know something about electoral systems. Those advancing the argument obviously do not. Anyone who knows anything about electoral systems would know that Australia may be compared with Ireland and Malta, but not with Germany or New Zealand.
PR systems come in two types. There is the single transferable vote (STV) of Australia, Ireland and Malta. Then there are party list systems of PR which apply in a host of countries, including Germany and New Zealand. STV is a system of direct election. Under party list systems, by contrast, the voters merely distribute numbers of party machine appointments. Party list systems permit the concept of a threshold. A party getting below the threshold sees its entire vote thrown into the rubbish bin. STV does not permit a threshold. A candidate with a poor vote will be excluded at the appropriate stage of the count. His or her votes will then be transferred by preferential marking to other candidates still in the count. STV is a preferential system.
Would it be possible to change our Senate PR to a party list system? No, it would not. The first sentence of s.7 of the Constitution provides:
“The Senate shall be composed of senators for each State, directly chosen by the people of the State, voting, until the Parliament otherwise provides, as one electorate”.
Since STV is the only system of PR under which Senators are “directly chosen by the people”, it is the only form permitted by the Constitution.
Since Senator Coonan’s proposal for “Senate reform” is so clearly out of the question, what are we to say for the scheme being promoted by former Liberal Party federal director, Andrew Robb? I would say it deserves twice as much consideration as Coonan’s but should still be firmly rejected.
One of the merits of the Robb scheme is that it has, at least, been put into legislative form. The distinguished former Senator, Malcolm Colston, decided to make his last contribution to the Senate very recently. He tabled a 50-page Bill with the title Electoral Amendment (Senate Elections) Bill 1999. To quote from the longer title, the Bill is “to provide for the division of States into Wards for the purpose of choosing Senators, and for related purposes”.
The purpose of the Colston-Robb proposal is to turn the Senate into a federal version of the Victorian Legislative Council. That body presently has only Labor, Liberal and National members. Only they can get elected in a single-person preferential election. Every present Senator currently being an Independent, or currently belonging to the Democrats, Greens or One Nation, would be doomed to defeat when their present terms expire.
No doubt Andrew Robb is pleased that a very senior former Senator (the Father of the Senate, no less) should have given his prestige in support of what used to be regarded as Andrew Robb’s hobby horse. That is why I call it “the Colston-Robb Bill”. Unlike the Coonan scheme, a case can be mounted that the Bill is constitutional.
Essentially, the argument rests on the presence of the words “until the Parliament otherwise provides” in s.7 as quoted above. Whereas direct election is a permanent requirement, so the argument runs, the expression “voting . . . as one electorate” operates only “until the Parliament otherwise provides”. My reason for rejecting the argument is that I have studied the circumstances which led to the inclusion of those words which allegedly make the Colston-Robb Bill constitutional.
The Founding Fathers knew that they were making the Australian Senate to be a copy of the American Senate. At the time, the American Senate was partly elected by the people and partly chosen by State legislatures (according to the laws of the State in question). However, because there were only two American Senators per State but there would be six per State in Australia, our Founding Fathers were uncertain as to how it might work. As a statement of modesty they inserted those saving words for fear that the system may fail. They thought the Parliament should be given some flexibility.
If, during the first 20 years of Federation, the concept of Senators being elected from States voting as one electorate had proved demonstrably to be a failure, I have no doubt it would have been possible to enact the Colston-Robb Bill in, say, 1920. As far as I am concerned, the words “until the Parliament otherwise provides” expired in 1920, or thereabouts.
Would the High Court agree with that view? I think it would. It would decide that the Constitution effectively reads: “The Senate shall be composed of Senators for each State, directly chosen by the people of the State, voting as one electorate”. They would decide that such a reading is essential to the role of the Senate as the States’ House.
Anyone wanting to understand the likely reaction of the High Court would do well to study the judgments in the February, 1996 McGinty case which dealt with electoral boundaries in Western Australia. There is a train of thought which comes through all the judgments. It is that an electoral system of very long standing can acquire a legitimacy which makes constitutional a scheme which might otherwise be unconstitutional.
In the case of the Senate we can state as an historical fact that there are two fundamental characteristics of the system which have remained unchanged since 1901. They are “directly chosen by the people”, and States “voting as one electorate”. Any attempt to change either of those would have a character quite different from the several changes since 1901 which are well known (such as plurality changing to preferences, majoritarian characteristics changing to PR, the introduction of ticket preferences, etcetera).
Consequently my view is that the High Court would not permit either the Coonan or Robb schemes. However, I admit this against myself. Among the several legal experts I have consulted, opinion is divided 50-50 on whether the Colston-Robb Bill is constitutional or not. By contrast, opinion is unanimous that the Coonan scheme is unconstitutional.
It is my expectation that the Senate of the 39th Parliament (that is, the present Parliament) will be dissolved. My guess is that we shall have a double dissolution election in about April, 2001. The present would, in my way of classification, then be a “hostile Senate”. It is not an expectation which worries me. Nor does the prospect justify “Senate reform” of any kind.
From time to time the majorities in our two Chambers get out of harmony. In Table 8 I set out my assessment in that regard. It shows just 11 years out of 100 in which those majorities were out of harmony. Not a great number of years of discord. However, how does it happen? Some say it is because people vote differently between the two Chambers. I do not deny the fact of widespread vote-splitting. What I deny is that vote-splitting explains the discord.
Table 8: Years House of Representatives and Senate Majorities
Not In Harmony
Number of Parliament
|
Years
|
Prime Minister
|
Length of
Discord
|
5th
|
1913–14
|
Cook (Liberal)
|
One year
|
12th
|
1930–31
|
Scullin (Labor)
|
Two years
|
19th
|
1950–51
|
Menzies (Liberal)
|
One year
|
28th/29th
|
1973–75
|
Whitlam (Labor)
|
Three years
|
32nd
|
1982
|
Fraser (Liberal)
|
One year
|
34th/35th
|
1987
|
Hawke (Labor)
|
One year
|
39th?
|
1999-2001?
|
Howard (Liberal)
|
Two years?
|
In my opinion the explanation for the discord of Table 8 has, since 1949, always been found in the different electoral systems operating for the two Chambers. For example, in the present Senate we have a left-of-centre majority which reflects the votes in a broad sense. In the House of Representatives, by contrast, we have a comfortable Coalition majority when the Labor Party received 51 per cent of the two-party preferred vote while the Coalition received only 49 per cent.
I have deliberately put question marks in the bottom row of Table 8. The reason for those question marks is that we do not yet know whether the present will turn out to be a hostile Senate. If the present Parliament does turn out to be one of discord between the two Chambers it will be due entirely to the two different electoral systems. It will raise this response from supporters of the Senate:
“If Australia must continue to suffer from such a rotten electoral system for the House of Representatives, it must surely expect some discord between the Lower House with a rotten system and an Upper House with a good system”.
Mind you, that is not what I would say. I think it is a good thing to have those two different electoral systems. If occasionally their operation produces discord, then so be it. That is the consequence of having a system of checks and balances.
Endnotes:
1. Source: Ian McAllister, Malcolm Mackerras and Carolyn Brown Boldiston, Australian Political Facts, 2nd edition, Macmillan, 1997, pp.8, 78 and 80. Updated for the 38th Parliament by Malcolm Mackerras.
2. Source: Ian McAllister, Malcolm Mackerras and Carolyn Brown Boldiston, op.cit., pp.100–106.
Chapter Eleven
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/SGSocUphAUCon/1999/12.html