(a) "large-enrolment multi-member electorates tend to distance elected representatives from their constituents,"(b) "the likelihood of a minor party holding the balance of power is increased, and this may lead to unstable government,"
(c) "complicated voting procedures may obscure voter intentions and produce delays in counting."
(d) A fourth criticism often made is that the Hare-Clark system in Tasmania encourages destructive intra-party competition.
[Hyperlinker's note: Objection (a) is responded to in the remaining sections of this file (sections 17-22). Objection (b) is responded to in the next part (sections 23-30). Objections (c) and (d) are responded to in the part after that - (c) particularly in section 41 and (d) particularly in section 31. The other sections of that part deal with other miscellaneous objections.]
Now, all other things being equal, a small electorate would ideally be preferable to a large one in that it would give each MP fewer constituents to service.
But no system of proportional representation will work without multi-member electorates in some form. Even the West German system, which retains single-member electorates, has some MPs chosen from lists whose only "electorate" is an entire country or state. In the Hansard Society's version, some electorates would each have what amounted to two MPs (the winner plus a "best loser") while other electorates, possibly with more voters, would only one MP each.
Having larger electorates, then, is the price of having proportional representation. Is it worth it?
The argument often heard against Queensland's zonal system is that the difficulties of serving large electorates can be compensated by providing greater travel and communications facilities, [66] and should not override the principle of fair representation. The same argument applies here.
The laws and policies decided by Parliament have a far greater effect on people's lives than all the constituency contact that individual MPs may have. [67] It is a fundamental principle of democracy that the government's laws and policies should not be imposed by a minority on the majority. It is a mere convenience that voters and MPs should be acquainted personally.
Democratic principle, then, would accept some diminution in personal contact to ensure an accurate electoral result. However, under Hare-Clark there is no conflict between the two goals. The system's capacity of intra-party choice will actually make MPs more accessible to their constituents, even in larger electorates; and in some ways, the use of multi-member electorates will make personal contact between voters and MPs more of a reality.
Moreover, the larger size of multiple electorates, as compared to the existing single-member electorates, should not be exaggerated.
"In the days when electorates might number only 300 voters each, it was possible to speak of the member as representing that part of the population that was enfranchised. But in Australia today, House of Representatives electorates average over 75,000 voters, so that real communication between the member and any significant part of the electorate is impracticable." [68]
Even in Queensland, the smallest State electorates contain eight thousand voters and twelve thousand people; it is hard to see, despite the traditional claims, how any one human being could really have a "close and informed contact" with so many people.
Even in countries like Britain and New Zealand, [69] which are tiny by comparison with Australia, the same complaint is made that rural electorates are too large in area for effective representation.
Further, even the smallest electorates may not produce the close relationship claimed. Local council wards are the smallest electorates existing in Queensland, yet local elections suffer more from apathy, lack of participation, and low voting turnout than do Federal and State elections. Many voters do not know their local councillors or candidates. [70]
By contrast, a Representative in the United States Congress has, on average, half a million constituents. A US Senator for a large State like California has over twenty million. A Senator or Representative for Alaska or Texas has an electorate larger in area than New South Wales. Yet legislators in the USA keep in closer touch with their constituents than do MPs in Australia [71] - for example, by employing staff and using direct mail. [72]
In Australia, the Northern Territory has 50,000 voters over an area comparable to Queensland's. Yet Paul Everingham, the former Chief Minister, was well-known by the Territory's voters.
Conservative voters in Brisbane Central would no doubt regard themselves as having a stronger attachment to Flo Bjelke-Petersen - even though, as a Senator, her electorate covers the whole State with 2.5 million people - even though with Peter Beattie, their local Labor MLA, might live next door.
It seems, then, that the important question is not size as such, but the chance the voter has of getting a sympathetic response from a MP s/he approaches.
"The smaller the election district, the closer, in general the contact. Of course; but is mere contact a qualification for election? Would you like to be compelled to patronise the dentist who happens to be nearest to you? ... If [a voter] wants to vote for a neighbour, he can do so under proportional representation, and with a good chance of success. But if he wants to vote for someone farther away, why should he be prevented from doing so?". [73]
Under Hare-Clark, the average distance from voter to representative is not increased; and the average distance that a voter has to travel to find a sympathetic representative is reduced. [74]
Of course, electorates must retain some "acceptable level of local proximity of the MP to the voters" [75] - which is why we do not recommend that any electorate should have more than nine MPs. But this size limit enters the equation only once the electorate has enough MPs to give every voter a reasonable choice of representatives.
While under Hare-Clark electorates will be larger in area, the increase will not be as great as might be thought at first, because neighbouring population centres can be incorporated into the large rural electorates. [76]
All the electorate's MPs (or at least those of the same party) could work as a team to cover the electorate, with joint electorate offices and staff, coordinated media efforts, handling of constituency queries, and dividing up issues on which to concentrate. [77]
It would also make sense to provide long distance numbers by which rural constituents can ring their MPs toll-free. In any event, to post a letter to an MP, however far away s/he may live, costs only forty-one cents; if a matter is truly important, most constituents would prefer to communicate in writing anyway.
Indeed, if Hare-Clark does make candidates and MPs more responsive to voters' wishes, then the larger electorates could be a positive safeguard against such responsiveness degenerating into narrow parochialism and "subservience to small local interests." [78]
This is claimed to reduce the accountability of members to their constituents: "Personal accountability may be blurred, and constituency work made more a party-political affair, as several MPs in a region might claim, or deny, responsibility for developments affecting the constituency." [80]
Tasmanian MLAs have been criticised as "everybody's members, but nobody's members, with no defined area to represent and no defined responsibility." [81]
Such criticisms are unjustified. No one has complained that a Tasmanian electorate is not a "defined area" when it is represented by a single MHR at the Federal level. The description "everybody's members" may be accurate enough, considering that over 90% of the voters saw their vote elect an MLA. But it is hard to see how seven candidates elected by the voters could be called "nobody's members."
Moreover, these criticisms seriously underestimate the capability of the average voter and the flexibility of the multi-member system:
"There is sometimes advanced the rather childish objection that an elector with several MPs will not know which of them to approach with his problem. Usually, an elector will apply to the member he has helped to elect, but he is not obliged to do so; he may prefer to call on one who lives near, or on one known to have a special interest in the matter involved." [82]
MPs can easily arrange an informal territorial or functional division of labour; for example, to informally divide the electorate into single-member "bailiwicks." [83] Voters do not have to be restricted by this division; and no MP, depending on personal votes for survival, could afford to "pass the buck" by turning away any constituent seeking help, for "representatives cannot identify their real constituents - those who voted for them - under any form of secret ballot." [84]
Multi-member electorates are not completely alien to Australian traditions. The Upper Houses of South Australia, Western Australia, and Victoria have used multi-member electorates without proportional representation, instead electing members in rotation by majority vote.
Some MPs, such as Ms Ros Kelly, may like to believe that "under a system of single-member electorates, candidates would be judged on their own merits and on their responses to local issues [and their] standing in the local district". [85] This statement may be understandable, as in this context Ms Kelly was attacking the d'Hondt list system; but it is still quite wrong. Despite all the good work she has undoubtedly done as a local member, it is highly unlikely that her own personal standing among voters in the ACT would be enough to get her elected if she were not the endorsed Labor candidate.
Surveys have often shown that very few Australians even know the name of "their" local member. [86]
In a single-member electorate, as mentioned earlier, up to 49% of voters may find their votes wasted and themselves nominally "represented" by an MP whom they did not vote for and may, indeed, strongly oppose.
It is true that most MPs are conscientious enough in attending to their constituents, regardless of their political affiliations, in relation to minor bureaucratic matters. But "very few [voters] choose the candidate they think most likely helpful about the nuisance next door or the disappearance of a bus service; most vote for the candidate of the party they most nearly agree with on the big questions of government policy." [87]
And it is precisely on these matters that a single MP simply cannot represent constituents with views opposed to his/her own: "It is a scandalous form of monopoly to deny troubled constituents, in need of help, an MP in whom they can confide." [88]
Some might claim that such voters could turn to a more sympathetic MP in another electorate. However, by convention, one MP will not encroach on another's constituency. Also, under the single-member system, the nearest MP of the voter's party could be hundreds of miles away.
This monopoly of the single member extends to other aspects of the MP's role, often distorting political discussion and debate: "In Parliamentary debates, the single member often contributes only one side of his area's problems - the side that suits his party's argument." [89]
And the boundaries of single-member electorates have become increasingly arbitrary, failing to reflect distinct communities of interest:
"... increasing numbers of constituents spend their working lives in neighbouring constituencies... MPs find themselves more and more working on local matters in a team of several Members, as constituency boundaries lose all meaning. It is the travel-to-work area, or the city, which is significant." [90]
So single-member electorates are not indispensable but, indeed, have serious flaws. The multi-member electorates of the Hare-Clark system are not only feasible; they also have several positive advantages.
The larger size of electorates allows a greater margin of tolerance in drawing boundaries, so it is easier to follow natural communities of interest.
Also, larger electorates are more marginal, because local concentrations of party support cancel out over a wider area. This makes the advantage from deliberate boundary-rigging harder to calculate (as with the "Tullymander" in Eire in 1977, which backfired on the party that devised it).
Under any electoral system, of course, boundaries should be drawn by an independent commission free of political control or interference.
Larger multi-member electorates would require much less frequent redistribution to take account of population trends. This would avoid public confusion, [93] the separation of communities from their former representatives, [94] the need for parties to adjust to new areas, [95] and unedifying disputes over whose seats should be abolished. [96] Further, under Hare-Clark "Good members survive swings against their party that would obliterate their counterparts in single-member districts." [97]
Moreover, "candidates need not have to shift house to "winnable" electorates, or live a long way from the area they represent." [98] At present, a Labor candidate in Toowoomba or a National candidate in Ipswich has to look far away from home to contest or represent a winnable seat. Under Hare-Clark, MPs would more likely be local people, well-acquainted with local problems.
One justification often offered for Queensland's zonal system is that, without such over- weighting of rural votes, the heavily-populated cities would dominate the whole state and centralised government would result.
This fear might seem unfounded, because how an MP votes in Parliament is determined by party, not locality. The Liberal member for Toowong votes with the Liberal member for Mt Isa, thousands of miles away, far more often than he votes with the Labor member for Archerfield next door.
However, this fear may have some basis given that the geographical division of Queensland largely coincides with the political division. Labor is stronger in the urban areas of Queensland, and the Nationals are stronger in rural areas. This may lead to suspicion that parties neglect certain areas in their election strategies: "campaign strategies and policy proposals will be geared to the areas in which they can win seats, not to areas they have no chance of winning" [100], and in government:
"Three-quarters of this State is not represented in the Cabinet of Queensland... If one travels up to the favoured area to the west of the Darling Downs, one finds back-to-back and wall-to-wall Ministers." [101]
The effect of the single-member system is to exaggerate the geographical/political division by producing "one-party regions" where one party monopolises the seats. In 1989, Labor won 31 of 36 seats (86%) in the Brisbane metropolitan area with only 57% of the votes. In the New South Wales Lower House, northern Sydney's representation is virtually one hundred percent Liberal, while the Western Suburbs are nearly one hundred percent ALP.
Other countries suffer even more from this same divisiveness engendered by the single- member system; an extreme example is Canada. [102] Britain now [1990] has "a sea of Tories across the whole south"; [103] while Labour is "a party confined to barracks far from Westminster," [104] but enjoys, on the other hand, an "unhealthy dominance of Scottish politics." [105]
If one party could achieve a clean sweep of the urban electorates under the single- member system, it could still dominate the rural areas, despite the zonal system. Even though the rural areas have more seats in proportion to their voter numbers, the urban areas still have far more seats (64 out of 89) overall.
Thus, to completely guard against the perceived threat of urban domination, the over- representation of country electorates would need to be increased to an extent not seen since the "rotten boroughs" of 18th-century England - an extent that even most National Party supporters might consider too extreme, and which could well provoke Federal intervention.
On the other hand, Hare-Clark would ensure that both government and opposition parties won seats in every electorate and every region in the State. Thus, every electorate would have representatives on both the government and opposition benches. No government could win power without a broad geographical base of support throughout the State; and no party could afford to neglect any area, however sparsely-populated.
Under Hare-Clark both the large blocs, of urban seats and rural seats, would be shared by all parties, not just monopolised by the largest party in each region.
More voters will have a Minister as a local member. In Tasmania, every voter in each of the five electorates has access as a constituent to least one Minister.
Hare-Clark would also discourage governments from denying, or threatening to deny, development funds to electorates that failed to elect government MPs (as Joh Bjelke-Petersen did at Mt Isa in 1983). Any such denial or threat would be strongly opposed by the governing party's own MPs from the area, who would not wish their constituents to blame them for failing to keep the development funds coming in.
The overall pressure for "pork-barrelling" would stay at the same level, but its intensity would be equalised, and thus bear a more rational relationship to the needs of different areas. No longer would governments pour resources into a few marginal electorates while neglecting those it cannot win (or lose). In every electorate the final seat would be marginal, [106] so each area would be courted to about the same extent.
Yet the party's supporters may have little in common with that candidate beyond. the party label. A "wet" Liberal may wish to vote for Ian Macphee; but, because he lives in Wilson Tuckey's electorate, he has no choice but to vote for a "dry" or vote against his party.
This means that the members elected to Parliament can scarcely be regarded as those the voters have freely chosen:
"Personal popularity within an electorate may be worth a percentage point or so in a close election, but it counts for little against a major national swing. It is a sad fact that most of us neither know nor care much about our local member. We vote the party ticket, and elect a depressing number of boobies as a result... The Labor casualties on 13 December [1975] included many hard-working local members; the survivors included a number of weary hacks." [107]
Of course, in single-seat electorates with preferential voting there is nothing to stop parties running more than one candidate (provided they can count on a tight exchange of preferences). [108] But it is very rare in practice [109] (except, apparently, in Tasmania). [110] There would be little point in endorsing two or more candidates, since only one could win; they would run as rivals, not as allies. [111]
Indeed, far from allowing any greater choice within parties, the trend in Australia has been to reduce the voters' choice between parties, by avoiding three-cornered contests between coalition partners. [112]
Further, the people who actually preselect the candidates from each party are a tiny minority. As one MP commented:
"When I was first preselected in 1968, 120 people voted. Yet on 25 October 1969, 25,389 people in that electorate voted Labor. In other words, only 0.05% of the Labor voters in [that electorate] had any real say in who would represent them in Canberra for the next three years... At a rough guess, probably fewer than 5,000-10,000 party members in all three major parties have any say in who governs their country or State..." [113]
So rigid have been preselection practices that even some senior politicians have supported the introduction c/f US-style primary elections. [114] However, Hare-Clark would ensure the same choice without impairing party unity or favouring well-financed candidates as do primaries.
Under Hare-Clark, with five, seven, or nine seats per electorate, each party would put up about five, seven, or nine candidates accordingly. (Of course, no one party will win all of the seats; but a full list of candidates is still put forward to make a good showing, and to have some reserve for filling later casual vacancies). And because not all of the candidates from each party can be elected, the voters can pick and choose among them.
"Few women and immigrants get elected to Parliament. The occasional candidate from these disadvantaged groups who secures a party nomination will almost certainly be found in a hopeless contest where the symbolic gesture can be made because nobody else rants the job." [115]
The single-member system makes it hard for members of these groups to be preselected for winnable seats. To win the required majority, parties seek law-risk, lowest-common-denominator candidates, [116] most usually the identikit model of a middle-aged white male. Winning endorsement from the party machines is a "long and hard" battle for women and ethnic groups. [117]
As we have seen, then, multi-member electorates are essential for proportional representation, and need not impair personal links between voters and MPs. Further, under Hare-Clark, multi-member electorates would leave less scope for gerrymandering; would ensure [p 21/ p 22] greater regional balance; would allow voters a wider choice of candidates; and would make an opening for women and ethnic candidates.
66. Madonna King, "End zone voting, academics urge" (Courier-Mail, 25 April 1990). [Back to text]
67. Carstairs, p 221. [Back to text]
68. Geoffrey Walker, Initiative and Referendum: The People's Law (1987), p 32. [Back to text]
69. Keith Jackson, "Electoral Reform: An Academic Perspective" (Legislative Studies Quarterly, Autumn 1988), p 12. [Back to text]
70. C/f William H Riker, Democracy in the United States (New York, Macmillan, 1965), p 295. [Back to text]
71. Donald Horne, Elaine Thompson, Dean Jaensch, Ken Turner, Changing the System: Political and Constitutional Reform - Some Options and Difficulties (APSA Monograph #25, 1981), p 29. [Back to text]
72. Laurie Oakes, "Politicians' junk-mail junket" (The Bulletin, 29 August 1989), p 27. [Back to text]
73. Hoag and Hallett, p 142. [Back to text]
74. Sharman, p 98. [Back to text]
75. Holme, p 135. [Back to text]
76. New Zealand Royal Commission, p 53, para. 2.140. [Back to text]
77. Ihlein, p 20. [Back to text]
78. Viscount Bryce (House of Lords, 22 January 1918); quoted in Hoag and Hallett, p 142. [Back to text]
79. Herr and Hemmings, pp 216-220. [Back to text]
80. New Zealand Royal Commission, p 54, para. 2.145. [Back to text]
81. Alec Simpson (Executive Director, Institute of Public Affairs), letter to The Australian (23 November 1980). [Back to text]
82. Lakeman, p 167; c/f Wilson, p 19. [Back to text]
83. Buchanan, pp 20-21; Davis, p 187; Harrop and Miller, p 60; Herr and Hemmings, pp 216-220. [Back to text]
84. Hoag and Hallett, p 143. [Back to text]
85. Report of the Joint Standing Committee on Electoral Matters, p 102. [Back to text]
86. Donald Home, Elaine Thompson, Dean Jaensch, Ken Turner, Changing the System: Political and Constitutional Reform - Some Options and Difficulties (APSA Monograph #25, 1981), p 28. [Back to text]
87. Lakeman, p 166. [Back to text]
88. Richard Wainwright, "Single-Member Seats: Good Riddance", in The Case for PR by STV, pamphlet published by the Alliance Action Group for Electoral Reform, September 1986. [Back to text]
89. Wainwright; c/f Ihlein, p 20. [Back to text]
90. Wainwright. [Back to text]
91. As in South Australia; Peter Ward, "A new song sung to a very old tune" (Weekend Australian, 15-16 July 1989). Some argue that, in the USA, gerrymandering is actually easier now that the Supreme Court has outlawed malapportionment; the Court's insistence on near-exact equality means that electoral boundaries can no longer follow communities of interest such as geographical and local-government areas. [Back to text]
92. Despite Senator Ray's allegations of "unnatural geographical combinations of regional interests", (Ray, p 134) most of the controversy and acrimony surrounding redrawing of single-member electorate boundaries arises when the boundaries separate similar groups of voters with shared interests, rather than lump together dissimilar groups of voters.
Colin Howard, in "The Constitution as a Legal Document", in Change the Rules! Towards a Democratic Constitution (Ringwood, Penguin, 1977] criticises the division of the Albury-Wodonga community into two separate electorates because of the state border. C/f submission by Federal MHR Elaine Darling (ALP, Lilley) to the Redistribution Commission for Queensland (18 April 1984). [Back to text]
93. Ihlein, p 20. [Back to text]
94. Shirley, p 18; Ihlein, p 18. [Back to text]
95. Lakeman, p 5. [Back to text]
96. As is currently happening in South Australia; "Federal seat in danger of being lost" (The Australian, 1 May 1990). [Back to text]
97. Bogey Musidlak, "Making the Most of Your Support" (Good Government, June 1984), p 6. [Back to text]
98. Ihlein, p 19. [Back to text]
99. Godfrey Hodgson, "Revolution that Rolled Right Over Britain" (The Independent, 30 December 1989). [Back to text]
100. Bogdanor, p 185. [Back to text]
101. Angus Innes, Hansard (Queensland Legislative Assembly, 24 August 1988, p 41). [Back to text]
102. Irvine, p 71. [Back to text]
103. E G Matthews, letter to New Statesman and Society, 13 October 1989, p 6. [Back to text]
104. Peter Kellner, "Labour's future: decline or democratic revolution?" (New Statesman, 19 June 1987), p 10. [Back to text]
105. David Spaven (Scottish Green Party), letter to New Statesman and Society (8 October 1989), p 7. [Back to text]
106. The swing which a party would need to surmount the quota next above, or to fall under the quota next below, would vary in different electorates, but this variation would be tiny compared to the current difference between safe and marginal electorates. An electorate with an even number of seats, if government and opposition parties were roughly equal in strength, could be considered "safe" because the result would inevitably be a tie.
[Note: Some members of the PRSA now favour 7- or 9-member electorates because, they suggest, a 5-member electorate can become a relatively non-swinging electorate. The quota in a 5-member district is 16.66% of the votes. Suppose the contest is between two parties who obtain 59% and 41% of the votes, winning 3 and 2 seats. Then the leading party is going to need a swing against it of 9%+ to lose a seat, and a swing of a further 7.7% towards it to win an extra one. In most elections, this is fairly unlikely. The electorate has therefore become "safe" and will be ignored by government and opposition. The required swings contract into the range of 6.25% or less once the number of members is 7 or greater. JP, Jan 2000] [Back to text]
107. Peter Cole-Adams, "Massacre at the polls a flaw in the system" (The Age, 22 December 1975). [Back to text]
108. As advocated by George Warwick Smith, "Superstar? What Makes a Prime Minister?" (Quadrant, October 1986, p 46), and occasionally practised by the National Party. [Back to text]
109. "Towards a democratic voting system" (Good Government, December 1968), p 3. [Back to text]
110. Davis, p 197. [Back to text]
111. Lakeman p 136. [Back to text]
112. "Coalition 'death knell' to Nats" (Weekend Australian, 5-6 May 1990). [Back to text]
113. Barry Cohen, "Non-democracy, Australian-style" (The Bulletin, 19 April 1988), pp 90-91. [Back to text]
114. Barry Cohen, "Non-democracy, Australian-style" (The Bulletin, 19 April 1988, pp 90-91); Sir James Goldsmith, "Gentrification or Growth; The Cultural Causes of Economic Failure" (Quadrant, October 1986, p 12); Stephen Loosley (Federal Labor Vice-President), "We should consider US primary voting model" (Weekend Australian, 8-9 October 1988); and the Liberal Party's Valder Report, cited in Patrick O'Brien, The Liberals; Factions, Feuds, and Fancies (Ringwood, Viking, 1985, p 168). [Back to text]
115. Colin A Hughes, "Machinery of Government", in Australian Politics: A Fourth Reader (ed Henry Mayer and Helen Nelson; Melbourne, Longman Cheshire, 1977), p 377. [Back to text]
116. Holme, p 135. [Back to text]
117. Jaensch, p 56. [Back to text]