A Proportional Representation System for the Election of the Australian Parliament

Proposed by Stephen Lesslie


Download PDF version

Image marginal

Reproduced with the kind permission of Alan Moir, Sydney Morning Herald


1 Rationale

The electoral reform proposed here is a proportional representation system designed to re-engage the Australian people by ensuring that they are given genuine choice in the election of their parliamentary representatives.

2 Preamble

The proportional representation model proposed here for the House of Representatives, and concurrent reform of the proportional representation system for the Senate, will ensure:

3 What is Proportional Representation?

Proportional representation is an electoral system where, instead of returning only one member per electorate, two, and preferably more members are elected from the same electorate. Each successful member is elected, not by a majority, but by a quota. A candidate is elected when he or she achieves enough votes to reach a quota. For every quota a political party receives it will elect a candidate. The result is that members are returned in the same proportion as votes cast for their party.

A quota is determined by the formula:

\textit{quota} = \frac{\textit{total formal vote}}{\textit{c...
...s to be elected}+1} +1\qquad\textrm{(Disregard any fractions)}

Proportional representation allows:

3.1 Benefits of Proportional Representation

4 Outline of Proposal

These two points are the key to this proposal and, with the refinements outlined below, allow a electoral system under which fine distinctions in voting support will be accurately reflected in the composition of the parliament, but micro parties and opportunistic groupings will not be elected.

4.1 House of Representatives

For the House of Representatives the electorates should be as large as practicable.

Under the Australian Constitution (S.24) each State is entitled, in accordance with its population, to a set number of Members in the House of Representatives. Seats can not be shared between States. The current division of seats between the States is New South Wales 49; Victoria 37; Queensland 29; Western Australia 15; South Australia 11; Tasmania 5. The Australian Capital Territory and the Northern Territory have two seats each.

Under this proposal Western Australia, South Australia and Tasmania would each be one electorate returning fifteen, eleven and five members respectively.

New South Wales would be divided into one thirteen-member and three twelve-member electorates.

Victoria would be divided into two twelve-member and one thirteen-member electorates.

Queensland would be divided into one fourteen-member and one fifteen-member electorate.

The Australian Capital Territory would be one electorate returning four members.

The Northern Territory would be one electorate returning three members.

Senators representing the territories would be abolished.

The proposed model returns, as far as practicable, similar numbers of members from each electorate. States that are divided into electorates should be divided so that each electorate reflects as far as possible the characteristics and idiosyncrasies of the State as a whole; it is therefore essential that all electorates should include both regional and capital city voters.

In States that are to be divided into electorates it is not necessary to have electorates return an uneven number of members. In two party terms, the proposed quotas are small enough to enable one of the major parties to gain a majority of the seats. In multi-party terms it is likely that a minor party or independent candidate may obtain a quota, leaving the balance as an uneven number of members.

The model proposes fourteen (14) separate electorates for the House of Representatives.

New South Wales Eastern NSW (EN) 13 members
  Northern NSW (NN) 12
  Western NSW (WN) 12
  Southern NSW (SN) 12
Victoria Western Victoria (WV) 12
  Central Victoria (CV) 13
  Eastern Victoria (EV) 12
Queensland Northern Queensland NQ (14)
  Southern Queensland SQ (15)
Western Australia 15
South Australia 11
Tasmania 5
Australian Capital Territory 4
Northern Territory 3

Figure 1: Proposed Electorates.

The model proposes fourteen separate electorates for the House of Representatives.

State (Seats)     Electorates (members)     Quotas
NSW (49) 3 (12) & 1 (13) 7.70%, 7.15%
VIC (37) 2 (12) & 1 (13) 7.70%, 7.15%
QLD (29) 1 (14) & 1 (15) 6.67%, 6.25 %
WA (15) 1 (15) 6.25%
SA (11) 1 (11) 8.34%
TAS (5) 1 (5) 16.67%
ACT1 (4) 1 (4) 20.00%
NT (3) 1 (3) 25.00%

With the exception of Tasmania and the two Territories2 all electorates will have quotas between 6.25% and 8.34%.

Because Tasmania is only entitled to five members in the House of Representatives, it is not possible to avoid the relatively high quota of 16.7%. This proposal still ensures better representation than the current five, single-member electorates.

A state should be subdivided into electorates when the quota for election falls below 5%. The next state to do this will be Western Australia but it will be some time, if ever, before the population increases sufficiently to entitle it to the 20 members needed to trigger this requirement. The populations of New South Wales, Victoria and Queensland are unlikely to change sufficiently to require a change in the number of their electorates. As a result, the system once established is very stable and gerrymander proof. When a State gains or loses a seat any redistribution is insignificant or unnecessary.

The following histogram shows the various quotas in percentages for electorates returning up to 20 members.

Figure 2: Relationship between electorate size and quota.
{ \psset{xunit=0.6cm,yunit=0.15cm}
\end{pspicture}} % end psset\end{center}\end{figure}

In an electorate such as South Australia, returning 11 members, a candidate will be elected after receiving 8.34% of the vote. A party receiving 16.68% will elect two candidates. If a party receives 50% of the vote it will elect 6 members.

The sizes of the proposed electorates are as large as practicable ensuring that the quotas are small enough to enable fine distinctions in voting support to be reflected in the composition of the parliament.

These quotas are also large enough to ensure that micro parties and opportunistic groupings are not elected.

Any party that fails to obtain at least 5% of the vote would find it very difficult to have a candidate elected, thus eliminating the need to introduce any arbitrary threshold.

It is recommended that as most voters are aware of which council area they live in, these new electorates be based on local government boundaries. For discussion and analysis only, this paper has used existing Federal electorates.

4.2 Senate

Each State is represented in the Australian Senate by twelve senators. Six senators are chosen each half-senate election and all twelve in double dissolution elections. The mode of election is proportional representation but because of above-the-line voting coupled with compulsory preferential voting senators are elected in the same order as chosen by the major parties. Minor and micro parties are often able to secure a position despite being able to attract only a small percentage of the vote. The changes proposed for the House of Representatives would be mirrored in the election for the Senate. The position of Territory Senators would be abolished. (see below)

4.3 Australian Capital Territory and Northern Territory

The Australian Constitution (S.122) allows the Commonwealth Parliament to determine the representation of the territories in either house as it thinks fit. Currently there are eight parliamentarians representing the two Territories: two Senators and two Members from each Territory. All, including Senators, are elected for the life of a single term of the House of Representatives. Proportional representation is the mode of election for the Senators. With only two Senators to be elected it has inevitably resulted in one Government and one Opposition member being elected from each Territory. The less popular party needs only to secure a third of the vote to gain a quota. A proportional representation system for the Territory Members of the House of Representatives would also result in a similar ``pairing'' of parliamentary representation.

The Northern Territory is, on a population basis, only entitled to one member of the House of Representatives. This was increased to two by the last Parliament. Both sides of Parliament supported this increase. Labor would probably have won a single Northern Territory seat. The Government hoped to neutralise Labor's advantage. Labor hoped to win both seats. At the 2007 Federal election Labor did win both seats. This extra seat should be abolished.

In this model it is proposed that the position of Territory Senator be abolished, but that the number of House of Representative members be increased by two for each territory. The Australian Capital Territory would thus have four Members and the Northern Territory three Members. This keeps the parliamentary representation of the Territories at what they currently have or should have.

This change confers a number of advantages:

In the Australian Capital Territory, the most likely result will be that Labor will win three seats and the Liberals one seat. This 3 : 1 representation in the House of Representatives has the same effect on the relativities of party support in the House of Representatives as the current representation of two Labor seats and no Liberal seats, but now approximately 85% of the electorate is represented in the House of Representatives by a member of a party they supported. A quota for election would be 20%. If the Greens, or any minor party, achieved this they would win a seat and over 95% of the voters would be represented.

A single Northern Territory seat is marginal and has, in recent years, been won by both the Government and Opposition. Therefore, whichever party wins the majority of the vote will win two of the three proposed seats.

4.4 Robson Rotation

Parties would be arranged on the ballot paper in groups as for current Senate elections, and the positions on the paper would be chosen as at present by lot, starting at A and going to Z or beyond. Parties would be able to advise their supporters to vote for Group K: National Party, or Group E: New Age Aquarius Party.

Internally, within the party groups, the Robson rotation currently used in Tasmanian and ACT elections should be used. This rotation mixes up the candidates within a party grouping and ensures that no individual candidate is unfairly disadvantaged by having an unfavourable draw on the ballot paper. Each candidate has an equal share of the top and bottom positions on the ballot paper. The order of candidates also is randomised so no candidate is unfairly advantaged by being the next candidate after a popular candidate and thereby gaining an advantage with a flow-on of preferences. The Robson rotation means that bitter fights over the order of candidates on the ballot can be avoided.

The major parties should favour this Robson rotation because, as well as being democratic in eliminating the donkey vote, it also maximises the chances that a major party will be able to increase its representation above that which, on a superficial examination, appears to be its mathematical entitlement (see Spreading the Vote).

In undivided States the major parties would know the number of seats their percentage vote would entitle them to and, hoping for a favourable outcome, would therefore nominate a few more than that number so that they can appeal to all sections and regions of the state. This also ensures that there are replacements available in case of casual vacancies.

A similar pattern would emerge in the more populous states. In a thirteen-member electorate such as proposed for Central Victoria, the major parties would each be confident of 5 quotas and hopeful of 8 quotas. The parties would therefore nominate approximately 9 to ensure that the party appealed to all sections and regions of the electorate.

4.5 Above-the-Line Voting

Above-the-line voting should not be used.

Above-the-line voting thwarts and undermines the integrity of any proportional representation system. The option of using the designated party boxes or voting below the line, whilst giving lip service to voter choice, is in reality a party list system which allows preference harvesting and backroom deals.

Currently, in Senate elections, the likelihood of a candidate's being elected is directly related to their place on the list. Since the introduction of proportional representation to the Senate in 1949, no mainland Senate candidate has ever defeated a candidate from higher up on the same party ticket. Tasmanians, with their better understanding of proportional representation, have on very rare occasions achieved this, but the introduction of above-the-line voting has now made this virtually impossible.

Above-the-line voting would allow a repeat of the 1995 and 1999 NSW Legislative Council experience when deals involving a plethora of micro parties, designed to harvest preferences, allowed one or more of them to be elected despite having minimal support.

In the proposed model it would not be in the interests of either the minor or major parties to attempt to organise such deals because of the difficulty that parties would have to advise supporters how to swap preferences outside their own party grouping. Micro party supporters would either exhaust their votes or move directly to one of the major players in the elections, but would not travel through a dozen different parties before finally stopping at the ``I Picked the Cleverest Name'' party.

Above-the-line voting denies a party's supporters a genuine choice of candidates. It also denies those same supporters the opportunity to have extra members of their party elected. (see Spreading the Vote)

4.6 Spreading the Vote

Because the proposed model offers voters a genuine choice between candidates, the vote for a party will be distributed over a number of candidates. Any party that obtains more than a quota has the real prospect of electing more candidates than may seem mathematically possible. The more quotas a party gains the more likely this is to happen.

For example 5 to be elected; 3 teams; 8 candidates; total votes 5100; quota 850 (16.67%)

ALP 2256 2.66 quotas 44.2%
Liberal 2244 2.64 quotas 44.0%
Third Party 600 0.71 quotas 11.8%

With a list system, including above-the-line voting, the third Liberal candidate would be eliminated. This occurs because the remaining 0.64 of a quota is less than the 0.71 of a quota obtained by the Third Party candidate. The most likely result would be: ALP 2, Liberal 2 and Third Party 1


ALP 2244 2.64 quotas 44.0%
Liberal 2256 2.66 quotas 44.2%
Third Party 600 0.71 quotas 11.8%

A small change has the third ALP candidate eliminated but gives the same result:

ALP 2, Liberal 2 and Third Party 1

But a genuine proportional representation system, that eliminates the donkey vote, and political parties that allow voters to choose their favourite candidate, may result in an election where the candidates share their party's votes equally.

ALP 1 752 Liberal 1 748 Third Party 1 301
ALP 2 752 Liberal 2 748 Third Party 2 299
ALP 3 752 Liberal 3 748  

The number of votes for each party remains the same but a Third Party candidate can not be elected. Even if every vote from the second Third Party candidate passes to the other, that candidate still does not have enough votes to avoid also being eliminated. It is now the Third Party voters who decide who is the final candidate elected; as they would have done had there been no Third Party candidates contesting the ballot.

Result ALP 3 and Liberal 2 or ALP 2 and Liberal 3.

This is the perfect split for the major parties but the same result would occur even if the split were not so perfect.

ALP 1 849 Liberal 1 849 Third Party 1 600
ALP 2 800 Liberal 2 800 Third Party 2 0
ALP 3 601 Liberal 3 601  

Here the third candidate for the major parties only received just over a quarter of their parties' votes (26.71%) but still remains in the ballot and the last position is determined by the preferences of the Third Party candidate.

A major party candidate can still be elected despite having less than the combined vote of the Third Party candidates.

ALP 1 845 Liberal 1 845 Third Party 1 301
ALP 2 824 Liberal 2 824 Third Party 2 299
ALP 3 581 Liberal 3 581  

A thirteen-vote leakage from the second Third Party candidate would see either the third ALP or Liberal candidate elected.

Only parties with sufficient support to obtain a quota can be guaranteed election. Conversely, any party that does not gain at least half a quota is very unlikely to win a seat, and no party can reasonably hope to be elected until they gain at least 85% of a quota. Third parties won't be able to assume that they will pick up the remainder of the major parties' quotas, as currently happens in Senate elections.

Can the major parties expect results such as described above? Yes! The more quotas a party receives the more likely it is.

Spreading of the vote in the 1951 double dissolution in Tasmania gave the Liberals four of the five long term Senate seats, helping to give the Menzies government control of the Senate for six years. Figures for the 1951 Tasmanian Senate Election can be found in the Appendices.

4.7 Informal Votes

The requirement to number every square in a ballot is either based on a cynical and undemocratic belief that informal votes will help a party's chances or based on the belief that it is necessary to prevent exhausted votes so as to maximise the number of votes involved in the final two party preferred outcome. The fear that exhausted votes will distort a result is overstated and insulting to the intelligence of the Australian voter. In a proportional representation ballot exhausted votes become insignificant.

For genuine democratic expression, it is necessary that the current high levels of informal voting be reduced. It is essential that optional preferential voting be adopted. Any consequent increase in exhausted votes will be more than compensated for by the reduction in the informal vote.

Currently in House of Representatives elections, voters are required to number every square. However, even if forcing voters to make a decision by choosing between candidates they do not like or have never heard of is considered democratic, it is paradoxically counter productive because the number of voters that are now involved in making the decision has been reduced by the consequent increase in informal votes. For every increase in voting complexity there is an even greater increase in the number of informal votes.

Under the proposed model, informal voting would be less than 3%. Informal voting in the Irish Dail, elected by an optional preferential proportional representation system, is normally less than 1%.

The following graph demonstrates the relationship between informal votes and the number of candidates contesting each electorate using results from the 2004 Federal election.

Candidates % Informal No of electorates
4 4.35% 3
5 4.40% 18
6 4.70% 29
7 4.89% 39
8 5.74% 30
9 5.70% 18
10 5.83% 7
11 6.83% 4
12 7.41% 1
14 11.83% 1

Figure 3: Informal voting in the 2004 Australian Federal election.
{ \psset{yunit=0.5cm}
...){\rotatebox{90}{Informal vote (\%)}}

In the 2004 Federal election for the electorate of Greenway, 84% voted for either the Labor or Liberal candidate. The 16% who voted for another candidate then determined who won the seat. Or was the result determined by the 11.83% of voters who voted informally because of the complexities of voting for 14 candidates?

The average informal vote for the House of Representatives in 2004 was 5.18% amounting to almost 640,000 voters; the informal vote in the Senate was better but still unsatisfactory at 3.75%. A large percentage of the Senate's informal votes would have been the result of voters trying to make an informed vote below the line but failing because of the rigid requirements.

The three multi-member electorates in the 2004 ACT elections did much better:

Electorate Candidates Members to be elected Informal Vote
Brindabella 21 5 2.7%
Ginninderra 23 5 2.7%
Molonglo 33 7 2.5%

The comparison with the informal votes in the ACT's two House of Representatives seats; Canberra (3.40%) and Fraser (3.48%) is stark. Both of these electorates only had five candidates each.

In fact, no House of Representatives' seat in the entire country had a lower informal vote than the three ACT House of Assembly seats and only five seats in the country had an informal vote lower than 3% (Higgins 2.76%, (6 candidates); New England 2.77%, (7); Bendigo 2.87%, (5); Indi 2.88%, (5); Kooyong 2.90%, (6)).

4.8 Exhausted Votes

There are three types of votes that are liable to become exhausted.

The first are votes that plump for only one candidate and in electoral systems that require voters to give preferences would be declared informal. In the proposed model these votes remain viable and add to the quota needed for the election of the candidate. As the major parties are likely to elect a number of candidates these votes are particularly valuable; with the spreading of the vote they help keep more candidates in the count. If a candidate is subsequently elected with preferences from another candidate then these, so called exhausted votes, remain with that candidate and are never distributed.

The second are votes that are surplus to the quota needed to elect a candidate, and because all the other candidates that the voter would have supported have either already been elected or eliminated from the ballot, have nowhere to go. These votes are natural and inevitable in any ballot, being equivalent to the votes in excess of 50% that a candidate receives in a single-member electorate.

The third are votes that give a preference or preferences to minor parties or independent candidates and stop before reaching a serious contender. These votes have no influence on the result and are equivalent to informal votes. These voters, however, are still part of the democratic process; had the electorate as a whole voted differently they might have played a part. They are an insignificant proportion of the total count and any attempt to remove or reduce them results in an even greater proportion of informal votes.

In the proposed model, the number of candidates to be elected from each electorate is as large as practicable, increasing the likelihood that the candidate to whom a voter gives a first preference will be elected. With optional preferential voting any candidate obtaining a quota of first preference votes will have any exhausted votes transferred at a reduced value. Any candidate who does not have a quota but is subsequently elected from another candidate's surplus or preferences will never have their second preferences counted and consequently no exhausted votes. In both cases, the genuine democratic expression of the voter will have maximum effect.

When the parties and the voting system allow voters to freely choose their favourite candidates, it is rare for a vote to travel much beyond its third preference.

Voters should be advised how many candidates are to be elected, but for a truly democratic and representative parliament, it is essential that voters be allowed to vote for as many or as few candidates as they choose.

Any vote that has a clear first preference should be considered formal.

4.9 Voter Understanding

Voters instinctively understand the value of their preferences and can utilise them intelligently without the need of how-to-vote tickets.

In the 2004 ACT House of Assembly election, 57.7% of voters in the seven-member electorate of Molonglo gave their first preference to a winning candidate. A further 14% and probably as many as 20% gave their second preference to a winning candidate. And 89.4% gave their first preference vote to a party that had a candidate elected.

The more candidates to be elected from a single electorate the greater these percentages become.

Despite the option of being able to stop after voting 1, voters in Molonglo understood the value of their preferences. At the stage in the count when all the minor parties and ungrouped candidates had been eliminated, leaving only those groups that elected a candidate, the exhausted vote was only 1.06% of the total vote (904 votes out of 85017). As stated above, these voters are still part of the democratic process, because had their fellow voters voted differently then those exhausted votes may have remained in the count helping to elect candidates of their choice. Any attempt to reduce this exhausted vote would lead to an even greater informal vote.

8941 voters in Molonglo chose minor parties and ungrouped candidates and 8037 of them (89.9%), realised that their first choice or choices had little chance of election and, despite having to find a different column, gave a continuing preference to a different party or candidate.

Voters clearly liked the option of being able to choose their favourite candidate within a group. The seven ALP candidates in Molonglo polled between 3.7% and 26.2% of the ALP vote and when the lowest, Cirson, was eliminated from the count 90.5% of her preferences went to other ALP candidates; the balance went Greens 4.2%, Liberals 2.6%, others 2.0%, and only 0.7% were exhausted. When Satler, the next ALP candidate was eliminated her votes were even tighter; ALP 92.2%, Greens 3.1%, and Liberals 2.7%, exhausted 2.0%

The preference exchange within the other groups was also tight. The following histogram shows the preferences to other members of the group when the votes of the first candidate from a group were distributed.

Figure 4: Retention rate of preferences to other members of a group in Molonglo (7 member electorate).
{ \psset{xunit=0.68cm,yunit=0.068cm,runit=0.68cm}
... \put(-2,1.06){.}
\end{pspicture}} % end psset\end{center}\end{figure}

Examination of the actual voting papers would, most likely, show an even smaller leakage of preferences from the major parties as the totals for their candidates have grown with the elimination of earlier candidates.

The major parties should note that their supporters are far more loyal than those of their minor party opponents and in fact they gain more from the leakages of the minor parties than they lose; this enhances the possibility of gaining extra seats.

The voters of the ACT have shown clearly that there is very little need for the major parties to issue how-to-vote cards.

4.10 Star Recruits

Political parties have trouble recruiting high profile candidates. Successful people do not want to give up a career outside politics on the chance that they might win a marginal seat and then lose it three years later. Safe seats are generally occupied by loyal party servants who are unwilling to give up their own political career for someone else.

The proposed model allows star recruits to take their chance like the party's other candidates, to be able to shine and enhance the party's appeal without being locked up in a safe electorate where their vote drawing capacity is wasted (for example, Peter Garrett). If they do not shine then they were not a star candidate!

Quotas for women, country candidates and ethnic communities are not required because each party will, of necessity, include them to ensure the party has a broad appeal.

Members of Parliament are also able to take sabbaticals. A member may be able to represent his or her country in diplomatic posts or step down to recover from illness knowing that there is a good chance that they will be able to return at the next general election. Indeed their public service or recovery from illness may enhance their appeal with the electorate.

4.11 Three Cornered Contests

In the 2007 Federal election neither the Liberal nor the National Party attempted to win a seat from their coalition partner. The only three cornered contests were in safe seats or against sitting Independents. These inter-party agreements deny the opportunity for Liberal supporters to vote Liberal and National supporters to vote National. These agreements also freeze the respective ratios of the two parties in the parliament. The proposed model allows all voters whether they live in the city or country to be able to vote for the party of their choice.

4.12 Bulwark against Landslides

In landslide elections single-member electorates exaggerate the result. Proportional representation ensures that parties are represented in the same proportion as the percentage of votes received at the election.

In single-member electorates when a landslide is on, no matter how hard working a local member might be, if they are in a marginal or semi-marginal electorate they will be defeated. The proposed model ensures that when governments or oppositions are decisively rejected at the polls, the effective and hard working members are more likely to retain their seats.

The following tables show the distortion between votes cast and seats gained in landslide election wins.

4.12.1 1943 Federal Election

  ALP UAP CP Others
Vote 50.2% 21.9% 8.5% 19.4%
Seats 49 (66.2%) 14 (18.9%) 9 (12.2%) 2 (2.7%)

The two party preferred vote was ALP 59.1% (66.2%); UAP/CP 40.9% (31.1%).

4.12.2 1966 Federal Election

Vote 40.0% 40.1% 9.8% 7.3% 2.8%
Seats 41 (33.1%) 61 (49.2%) 21 (16.9%) - 1 (0.8%)

The two party preferred vote was ALP 43.1% (33.1%); LIB/CP 56.9% (66.1%).

4.12.3 1975 Federal Election

Vote 42.8% 41.8% 11.3% 1.3% 2.8%
Seats 36 (28.4%) 68 (53.5%) 23 (18.1%) - -

The two party preferred vote was ALP 42.8% (28.4%); LIB/CP 53.1% (71.6%).

4.12.4 Queensland State Election 2001 & 2004

  2001   2004  
  Vote % Seats Vote % Seats
ALP 48.9 66 (74.2%) 47.0 63 (70.8%)
Liberals 14.3 3 (3.4%) 18.5 5 (5.6%)
National 14.2 12 (13.5%) 17.0 15 (16.9%)
One Nation 8.7 3 (3.4%) 4.9 1 (1.1%)
Greens 2.5 - 6.8 -
Others 11.4 5 (5.6%) 5.8 5 (5.6%)

Here the ALP results are greatly distorted. 48.9% & 47% of the vote gaining 74.2% & 70.8% of the seats respectively.

4.13 Casual Vacancies

By-elections are inimical to the principle of proportional representation. A by-election held because of the death or resignation of a member from a minor party, or a major party in an electorate where the other party dominates, generally results in that party losing the election. This compromises the party representation as decided at the general election. In the proposed model by-elections should not take place. Instead, the Tasmanian system of re-examining the original ballot papers to determine who would have been elected if the dead or retiring member had not been on the ballot paper in the first instance is recommended.

This has the advantage of being very cost effective. If various former Howard government ministers choose to resign from Parliament then the resulting by-elections will cost the government hundreds of thousands of dollars and similar amounts will need to be spent by the parties contesting the by-elections.

For the House of Representatives this change will require a constitutional amendment of S.33 of the Australian Constitution. It is a minor change, not affecting the proposal as a whole, and could be left until other constitutional amendments are put before the Australian people.

The Australian Constitution (S.15) regulates the replacement of Senators and does require a member of the same party as the vacating Senator to be appointed. Whilst legal, it would be politically difficult for a party to choose as a replacement anyone other than the person next in line for election at the previous election.

Parties must stand sufficient candidates to ensure that there are replacements in case of casual vacancies. If they do not do so they forfeit the right to replace a candidate if it becomes necessary.

4.14 Electoral Deposits

The election of the Australian Parliament is an important responsibility of the Australian people. They should be allowed to take this responsibility seriously without having to consider the merits of frivolous candidates who have the potential to distort the results of the election and undermine the integrity of the process.

Only candidates with genuine support should stand for election. Candidates who receive over 4% are entitled to public funding at approximately $2.10 per vote, which can amount to over $8000. Electoral deposits should be substantial, at least $20,000 per candidate.

A substantial deposit will help ensure that only genuine candidates contest the election and that the results are not distorted through unnecessarily high informal and exhausted votes. This will help prevent the deliberate confusion of voters caused by nominating multiple candidates; by nominating candidates with the same or similar name as a prominent opponent; by registering emotive party names and by candidates nominating, not with the intention of getting elected, but to push a cause or promote a business or just to massage their ego.

4.15 Thresholds

The imposition of thresholds, to prevent candidates with less than a certain percentage of the vote being elected, undermines the principle that no vote shall assist the return of any candidate or party unless the voter so chooses. Thresholds can also, in some instances, force voters into ``tactical voting'', which undermines the integrity of the election result.

In the last New Zealand election, Winston Peters' New Zealand First fell below the 5% threshold. This shortfall of a few thousand votes caused a major change in the composition of the parliament and possibly even the change of government.

In the proposed model thresholds are unnecessary because quotas in excess of 5%, coupled with the effect of the Robson rotation, will ensure that candidates with minimal support are unlikely to be elected. If however, sufficient voters are determined to vote for all the minor candidates before supporting any of the major parties then those voters should be allowed to have their say.

Thresholds are arbitary, undemocratic and unnecessary. Administrative procedures should not be used to deny the will of the people as expressed at the ballot box.

5 Analysis of the 2007 Federal Election

Using the proposed fourteen electorates the author has analysed the results of the 2007 Federal election for both the House of Representatives and, using state wide figures, for the Senate.

5.1 House of Representatives

  Electorates Liberal National ALP Green Indep
NSW 4 (49) 18 (15) 4 (5) 24 (28) 3 (-) - (1)
VIC 3 (37) 16 (14) - (2) 18 (21) 3 (-) - (-)
QLD 2 (29) 12 (10) 2 (3) 14 (15) 1 (-) - (1)
WA 1 (15) 8 (11) - (-) 6 (4) 1 (-) - (-)
SA 1 (11) 5 (5) - (-) 5 (6) 1 (-) - (-)
TAS 1 (5) 2 (-) - (-) 3 (5) - (-) - (-)
ACT3 1 (2) 1 (-) - (-) 3 (2) - (-) - (-)
NT 1 (2) 1 (-) - (-) 2 (2) - (-) - (-)
Total 14 (150) 63 (55) 6 (10) 75 (83) 9 (-) - (2)

For a detailed breakdown of these figures see the Appendices.

The result obtained using the proposed model more accurately reflects the national vote.

  % vote % seats (2007 election)
ALP 43.4 49.0 (55.3)
Liberal 36.3 41.2 (36.7)
National 5.5 3.9 (6.7)
Greens 7.8 5.9 (-)
Others 7.0 - (1.3)
No. of Seats   153 (150)

If the ``others'' are redistributed to the remaining parties in the same ratio as the original votes then the following result occurs.

  % vote % seats (2007 election)
ALP 46.7 49.0 (55.3)
Liberal 39.0 41.2 (36.7)
National 5.9 3.9 (6.7)
Greens 8.4 5.9 (-)
No. of Seats   153 (150)

This shows a high degree of correlation with the predicted theoretical results.

5.2 Senate

In the Senate the ALP picks up two extra seats at the expense of the Greens. In Western Australia the ALP comfortably takes the seat from the Greens, possibly even being elected on the Liberal party surplus. Remember it is the voters not the party that decides where to give, if at all, their preferences. In South Australia the Independent Xenophon is elected with a quota but has no surplus. The Liberals will win two seats and the ALP three. The ALP will be able to spread their votes sufficiently for all three candidates to remain above the Greens. The third ALP candidate will be elected on Greens preferences ahead of the third Liberal. This is a good demonstration of the power of the Robson rotation.

6 Conclusion

The proposed model gives the ALP the most seats in the slightly expanded House Of Representatives. With only 43.4% of the National vote they fell just short of an absolute majority of seats but are well placed to form either a minority government or coalition with the Greens.

The proposed model corrects the over representation of the National Party and the under representation of the Greens and ensures that there are both government and opposition members elected from every region in Australia. The model also demonstrates that the election of a government is not a matter of luck; a few hundred votes in a small number of seats can not determine the result.

The model demonstrates that thresholds are not necessary. No political party or Independent candidate with a vote of less than 5% is elected to either the House of Representatives or the Senate.

Advantages of the model also include:

Because the actual result was reasonably close to the predicted outcome from the proposed model, now is the ideal time to implement a proportional representation system. The proposal is constitutional and can be implemented by Act of Parliament.

It should be remembered that the 2007 election was fought in the marginal seats. Most Australian voters could have stayed at home and the result would not have changed.

A. 2007 House of Representatives Election

Here we analyse the results for the House of Representatives using votes cast at the 2007 Federal election.

House of Representatives: 14 electorates

ALP 6 6 6 6 7 6 5 7 7 6 5 3 3 2 75
Liberals 4 5 4 5 4 6 6 5 7 8 5 2 1 1 63
Nationals 2 1 1         2             6
Greens 1   1 1 1 1 1   1 1 1       9
Totals 13 12 12 12 12 13 12 14 15 15 11 5 4 3 153

Eastern NSW: 13 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Richmond, Page, New England, Cowper, Lyne, Paterson, Charlton, Shortland, Dobell, Robertson, Newcastle, Mackellar, Warringah.

Party Seats contested Total Votes % Quota
ALP 13 427606 39.2 5.48
Liberal 8 283448 26.0 3.63
National 5 167560 15.3 2.15
Greens 13 94818 8.7 1.22
Tony Windsor (New England) 1 52734 4.8 0.68
Christian Democratic 12 18765 1.7 0.24
Family First 8 8243 0.8 0.11
Democrats 5 4779 0.4 0.06
One Nation 3 2827 0.3 0.04
LDP 4 2455 0.2 0.03
CEC 8 1661 0.2 0.02
Socialist Equality Party 2 681 0.1 0.01
Comb Indep 13 22399 2.1 0.29
Comb Others4 4 3794 0.3 0.05
Total Votes: 1091770        
Quota: 77984        

The ALP will win six seats, the sixth on Greens' surplus and on the Independent Windsor's preferences. The Liberals will win four seats, the fourth on National Party surplus and Windsor preferences. The Nationals have two and the Greens have one quota in their own right. Tony Windsor with only 0.68 quotas cannot win a seat. No group not obtaining a quota will elect a candidate.

Result: ALP 6; Liberal 4; National 2; Greens 1

2007 Federal Election Result: ALP 7; Liberal 3; National 2; Windsor 1

Northern NSW: 12 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Parkes, Hunter, Greenway, Berowra, Mitchell, Lindsay, Chifley, Parramatta, Reid, Bennelong, Bradfield, North Sydney.

Party Seats contested Total Votes % Quota
ALP 12 428900 43.2 5.61
Liberal 10 368462 37.1 4.82
National 2 60902 6.1 0.8
Greens 12 60594 6.1 0.79
Christian Democ. 11 21699 2.2 0.28
Family First 10 8903 0.9 0.12
Climate Change Coalition 5 6542 0.7 0.09
CEC 7 2831 0.3 0.04
Democrats 3 2707 0.3 0.04
Socialist Equality Party 2 1330 0.1 0.02
LDP 3 1052 0.1 0.01
Horan (Parkes) 1 17098 1.7 0.22
Comb Indep 14 10018 1.0 0.13
Comb Others5 2 2459 0.2 0.03
Total Votes: 993497        
Quota: 76423        

The ALP will win six seats, the sixth on Greens preferences. The Liberals will win five seats, the fifth on Christian Democratic party preferences and the Nationals will win one on preferences from the Independent Horan and Family First preferences.

This is a good example of the premise that no candidate without a quota (the Greens in this case) can ever be confident about being elected.

Result: ALP 6; Liberal 5; National 1

2007 Federal Election Result: ALP 6; Liberal 5; National 1

Western NSW: 12 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Farrer, Calare, Riverina, Hume, Macquarie, Werriwa, Fowler, Prospect, Blaxland, Banks, Lowe, Sydney.

Party Seats contested Total Votes % Quota
ALP 12 455562 46.6 6.06
Liberal 10 293143 30.0 3.90
National 2 92720 9.5 1.23
Greens 12 69329 7.1 0.92
Christian Democratic 9 17018 1.7 0.23
Family First 6 10838 1.1 0.14
CEC 7 4416 0.5 0.06
One Nation 2 3039 0.3 0.04
LDP 2 847 0.1 0.01
Priestley (Calare) 1 19035 1.9 0.25
Comb Indep 4 8676 0.9 0.12
Comb Others6 3 2484 0.3 0.03
Total Votes: 977106        
Quota: 75163        

The ALP will win six seats. The Liberals four and the Nationals one. The twelfth seat will be much closer than it appears. The Robson rotation will allow the ALP to spread their vote and they will gain preferences from most other groups including the National Party. However it is more likely that a large number of votes from other groups will exhaust. On the balance of probability given to the Greens. Another example that a party without a quota is not guaranteed a seat regardless of how high their vote is.

Result: ALP 6; Liberal 4; National 1; Green 1

2007 Federal Election Result ALP 8; Liberal 2; National 2

Southern NSW: 12 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Eden Monaro, Gilmore, Throsby, Cunningham, Macarthur, Cook, Hughes, Barton, Watson, Kingsford Smith, Grayndler, Wentworth,

Party Seats contested Total Votes % Quota
Liberal 12 379258 38.0 4.94
ALP 12 479102 48.0 6.25
Greens 12 95289 9.6 1.24
Christian Democratic 12 20420 2.0 0.27
Family First 8 7485 0.8 0.10
Democrats 3 2747 0.3 0.04
Socialist Alliance 2 2099 0.2 0.03
Socialist Equality Party 2 1425 0.1 0.02
Non Custodial Parents 2 795 0.1 0.01
Conservatives for Climate 3 370 - -
CEC 2 137 - -
Comb Indep 7 5160 0.5 0.07
Comb Others7 2 2272 0.2 0.03
Total Votes: 997111        
Quota: 76101        

ALP six seats, Liberals five seats and the Greens have one. This is a very clear cut result and the count in this election, would be faster than that for 11 separate single-member electorates.

Result: ALP 6; Liberal 5; Green 1

2007 Federal Election Result: ALP 8; Liberal 4

Western Victoria: 12 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Mallee, Wannon, Corangamite, Ballarat, Corio, Lalor, Gorton, Maribyrnong, Gellibrand, Wills, Melbourne, Melbourne Ports.

Party Seats contested Total Votes % Quota
Liberal 11 313591 30.1 3.91
National 1 53227 5.1 0.66
ALP 12 510012 48.9 6.36
Greens 12 96266 9.2 1.20
Family First 12 33339 3.2 0.42
Democrat 9 12459 1.2 0.16
O'Connor (Corio) 1 10530 1.0 0.13
DLP 2 4485 0.4 0.06
CEC 7 3201 0.3 0.04
Socialist Alliance 3 2290 0.2 0.03
LDP 2 960 0.1 0.01
Socialist Equality Party 1 418 0.0 0.01
Comb Indep 2 2489 0.2 0.16
Total Votes: 1043267        
Quota: 80252        

This result shows the power of the Robson rotation. The ALP can spread their votes sufficiently amongst seven candidates to remain ahead of either the fifth Liberal candidate or the National Party candidate. Eventually preferences from O’Connor, the former ALP member for Corio, the Greens' surplus and leakages from other minor party candidates will elect the seven ALP candidates. The Liberals will win four seats and the Greens one.

Result: ALP 7; Liberal 4; Greens 1

2007 Federal Election Result: ALP 10; Liberal 1; National 1

Central Victoria: 13 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Murray, Bendigo, McEwen, Calwell, Scullin, Batman, JagaJaga, Menzies, Kooyong, Chisholm, Deakin, Higgins, Bruce.

Party Seats contested Total Votes % Quota
ALP 13 485854 44.5 6.23
Liberal 13 451471 41.3 5.79
Greens 13 90200 8.3 1.16
Family First 13 30784 2.8 0.39
Democrat 13 12344 1.1 0.16
CEC 8 3655 0.3 0.05
LDP 2 1595 0.1 0.02
Christian Democratic 1 678 0.1 0.01
One Nation 1 433 0.0 0.01
Socialist Equality Party 1 273 0.0 0.00
Comb Indep 13 15240 1.39 0.20
Total Votes: 1092527        
Quota: 78038        

This is a straight forward result. The ALP will win six seats. The Liberals will win six seats, the sixth on Family First preferences, and the Greens have one seat.

Result: ALP 6; Liberal 6; Greens 1

2007 Federal Election Result: ALP 8; Liberal 5

Eastern Victoria: 12 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Gippsland, Indi, McMillan, Flinders, La Trobe, Casey, Aston, Holt, Dunkley, Isaacs, Hotham, Goldstein.

Party Seats contested Total Votes % Quota
ALP 12 420349 40.7 5.29
Liberal 11 441864 42.8 5.56
National 1 42632 4.1 0.54
Greens 12 72380 7.0 0.91
Family First 12 31540 3.1 0.40
Democrat 11 14008 1.4 0.18
What Women Want 1 1825 0.2 0.02
CEC 7 1799 0.2 0.02
D.L.P. 2 1533 0.1 0.02
LDP 3 718 0.1 0.01
Christian Democratic 1 533 0.1 0.01
Comb Indep 2 3924 0.4 0.05
Total Votes: 1033105        
Quota: 79470        

The ALP will win five seats. The Liberals will win six seats. The Greens will take the last seat on Democrat preferences.

Result: ALP 6; Liberal 5; Greens 5

2004 Federal Election Result: ALP 3; Liberal 8; National 1

Northern Queensland: 14 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Leichhardt, Kennedy, Herbert, Dawson, Capricornia, Flynn, Hinkler, Wide Bay, Fairfax, Fisher, Longman, Dickson, Petrie, Lilley

Party Seats contested Total Votes % Quota
ALP 14 491522 42.8 6.42
Liberal 10 314261 27.3 4.10
National 7 177484 15.4 2.32
Greens 14 60135 5.2 0.78
Bob Katter 1 32537 2.8 0.42
Family First 14 26840 2.3 0.35
Democrats 14 10274 0.9 0.13
One Nation 4 4046 0.4 0.05
LDP 6 2041 0.2 0.03
CEC 6 3626 0.35 0.05
Christian Democratic 2 753 0.1 0.01
Hutchinson (Fisher) 1 10596 0.9 0.14
Comb Indep 13 17491 1.5 0.23
Total Votes: 1149314        
Quota: 76621        

The ALP will win seven seats. The Liberals will win five seats and the Nationals two seats. Bob Katter will not be elected; his preferences will elect both the seventh ALP and the fifth Liberal. There is an outside chance that the Nationals could beat the Liberal for the fourteenth seat.

Result: ALP 7; Liberal 5; National 2;

2007 Federal Election Result: ALP 8; Liberal 3; National 2; Katter 1

Southern Queensland: 15 Members to be elected

Results are analysed from votes cast at the 2007 Federal election in the following existing federal electorates:

Maranoa, Groom, Blair, Forde, McPherson, Moncrieff, Fadden, Bowman, Rankin, Oxley, Ryan, Moreton, Bonner, Griffith, Brisbane

Party Seats contested Total Votes % Quota
ALP 15 529143 43.0 6.89
Liberal 14 504177 41.0 6.56
National 3 62020 5.0 0.81
Greens 15 73803 6.0 0.96
Family First 15 27218 2.2 0.35
Democrats 15 12153 1.0 0.16
One Nation 4 4573 0.4 0.06
CEC 10 2334 0.2 0.03
LDP 6 1490 0.1 0.02
Socialist Alliance 3 1174 0.1 0.02
Fishing 1 1010 0.1 0.01
Comb Indep 11 10444 0.8 0.14
Total Votes: 1229539        
Quota: 76847        

This is a very straight forward result. ALP will win seven seats. The Liberals will win seven seats, the Robson rotation keeps all Liberals in the count, ahead of the National, and they will benefit from Family First preferences. The Greens take the fifteenth seat on Democrat preferences.

Result: ALP 7; Liberal 7; Greens 1

2007 Federal Election Result: ALP 8; Liberal 6; National 1

Western Australia: 15 Members to be elected

Results are analysed from votes cast at the 2007 Federal election for all the Western Australian federal electorates.

Party Seats contested Total Votes % Quota
Liberal 15 545365 46.3 7.41
ALP 15 433342 36.8 5.89
Greens 15 105106 8.9 1.43
Christian Democratic 15 24838 2.1 0.34
Family First 15 14215 1.2 0.19
One Nation 14 13529 1.2 0.18
Nationals 1 13459 1.1 0.18
CEC 13 3115 0.3 0.04
LDP 5 2730 0.2 0.04
Socialist Alliance 3 1141 0.1 0.02
What Women Want 1 590 0.1 0.01
Socialist Equality Party 1 157 0.0 0.00
Comb Indep 10 14861 1.7 0.20
Total Votes: 1177537        
Quota: 73597        

A very straight forward result. The Liberals will spread their votes sufficiently to win eight seats. The ALP will win six seats and the Greens one. Note: At the 2007 election there was a small two party preferred swing to the ALP however the Liberals made a net gain of one seat. In this proportional representation voting system the ALP gained one seat from the Liberals.

Result: Liberals 8; ALP 6; Green 1

2007 Federal Election Result: Liberals 11; ALP 4

South Australia: 11 Members to be elected

Results are analysed from votes cast at the 2007 Federal election for all the South Australian federal electorates.

Party Seats contested Total Votes % Quota
ALP 11 426639 43.2 5.18
Liberal 11 412621 41.7 5.01
Greens 11 68640 7.0 0.83
Family First 11 40031 4.1 0.49
Democrats 11 14957 1.5 0.18
Nationals 2 12420 1.3 0.15
One Nation 3 1643 0.2 0.02
LDP 6 1637 0.2 0.02
What Women Want 2 1455 0.1 0.02
Conservatives for Climate 1 1165 0.1 0.01
Comb Indep 5 6627 0.7 0.08
Total Votes: 988152        
Quota: 82347        

The ALP and the Liberals will each win five seats. The only seat in doubt is the eleventh. On the basis of probability given to the Greens who with Democrat preferences will stay ahead of the ALP and eventually be elected on the ALP surplus.

Result: ALP 5; Liberal 5, Green 1

2007 Federal Election: ALP 6; Liberal 5

Tasmania: 5 Members to be elected

Results are analysed from votes cast at the 2007 Federal election for all the Tasmanian federal electorates.

Party Seats contested Total Votes % Quota
ALP 5 139077 42.8 2.57
Liberal 5 124280 38.2 2.29
Greens 5 43893 13.5 0.81
Family First 5 7356 2.3 0.14
Quin (Lyons) 1 6092 1.9 0.11
CEC 5 1856 0.6 0.03
Socialist Alliance 2 859 0.3 0.02
LDP 2 606 0.2 0.01
Independent 1 1123 0.3 0.02
Total Votes: 325142        
Quota: 54191        

The Liberals and ALP have two quotas each. Because of vote spreading the fifth and final position will be between the ALP and the Greens. On the balance of probabilites given to the ALP, possibly even elected on the Liberals' surplus.

Result: ALP 3; Liberal 2;

2007 Federal Election: ALP 5

Australian Capital Territory: 4 Members to be elected

Results are analysed from votes cast at the 2007 Federal election for both of the ACT federal electorates.

Party Seats contested Total Votes % Quota
ALP 2 114244 51.1 2.55
Liberal 2 74295 33.2 1.66
Greens 2 29424 13.2 0.66
Democrats 1 2509 1.1 0.6
CEC 2 1295 0.6 0.03
Socialist Alliance 1 539 0.2 0.01
Independent 1 1275 0.6 0.03
Total Votes: 223581        
Quota: 44717        

The ALP has two quotas and the Liberals have one quota. Because of vote spreading the fourth and final position will be between the ALP and the Liberals. Green preferences will elect the third ALP candidate. Note: With above-the-line voting the Greens would have won the final seat despite polling only two thirds of a quota.

Result: ALP 3; Liberal 1

2007 Federal Election Result: ALP 2 (2 seats only)

Northern Territory: 3 Members to be elected

Results are analysed from votes cast at the 2007 Federal election for both of the Northern Territory federal electorates.

Party Seats contested Total Votes % Quota
ALP 2 46794 47.7 1.91
Country Liberal 2 40298 41.0 1.64
Greens 2 7903 8.1 0.32
LDP 1 358 0.4 0.01
CEC 1 235 0.3 0.01
Independent 3 2615 2.7 0.11
Total Votes: 98213        
Quota: 24554        

The ALP and Country Liberal Party each have one quota. Green preferences will elect the second ALP candidate.

Result: ALP 2; Country Liberal 1

2007 Federal Election: ALP 2 (2 seats only)

B. Common Misconceptions about Proportional Representation

B..1 That Proportional Representation will lead to unstable governments.

In the Westminster system, governments are formed by a party or coalition of parties having a majority in the House of Representatives. This ensures that the Government is able to pass legislation through, and survive votes of no confidence in, the House of Representatives. However, control of the House of Representatives does not ensure passage of legislation through the Parliament. Since the introduction of proportional representation for elections to the Senate in 1949 it is rare for a government to control both Houses of Parliament.

The Westminster system may give stable government but in the Australian context it does not guarantee effective government. Senators are under no pressure to pass legislation; their actions pass largely unnoticed by the Australian public and provided they keep their parties' endorsement, they are mostly guaranteed re-election. Even when they lose, they still complete their six year term. Senators from minor parties do not have any colleagues in the House of Representatives to consider when making decisions, and double dissolutions, with their subsequent reduced quota, are not a threat.

The failure of the Australian people, on a number of occasions, to support a constitutional amendment for simultaneous elections effectively limits any Australian Prime Minister's ability to call early elections.

Fixed terms for the Senate prevent early Senate elections. A House of Representatives election without a corresponding Senate election only gives the Opposition a free kick at the next half Senate election. Once the elections for the House of Representatives and the Senate are out of phase it is difficult to bring them back together again.

Double dissolution elections, because of the requirement to backdate Senators' terms to the previous 1 July, cause similar problems requiring either the House of Representatives election to be held a year early or an unsatisfactory half Senate election.

A Prime Minister's threat of an early election to force the Senate into passing legislation is largely bluff; Senators know this and have no fears about voting against government legislation.

In the Australian context, the traditional Westminster theory that a government must resign and call an election when defeated in the House of Representatives does not work. Neither a Government returned, nor a successful Opposition, gain an extended mandate and each is effectively limited to the timetable dictated by half Senate elections.

Until the Australian people decide to change the constitution the Governor General should accept the principle of three year parliaments and refuse any request from a Prime Minister for an election without a corresponding Senate election. Double dissolutions with their more stringent requirements and joint sitting provisions should still be allowed.

Under the proposed model, a government would still be formed by the party or parties having a majority in the House of Representatives. However, governments defeated on a particular item of legislation would not resign but continue on, submitting on that issue to the will of the parliament. This is not so different from current practice. Defeat of legislation in the Senate does not normally lead to an election.

If the House of Representatives takes the more extreme step of supporting a vote of No Confidence then the Governor General, in the absence of double dissolution triggers, should call on the Leader of the Opposition to form a government. Such an event happened in 1941 with the formation of the Curtin Labor government. The action of calling the Leader of the Opposition to form a government prevents such votes from being frivolous.

The current method of electing the Australian Parliament certainly results in stable governments being formed but it also results in governments that are often ineffective and powerless in parliamentary terms.

In the proposed model, where both houses are elected by proportional representation, passage of legislation through the House of Representatives may be harder, but passage through the Senate will be easier.

It is likely that coalitions formed in the House of Representatives will be reflected in the Senate. With the abolition of above-the-line voting individual Senators will not be as certain of re-election. They will have to consider the views of their House of Representative colleagues and may even be outnumbered by them in party caucuses.

The vote spreading capacity of the proposed model also makes it more likely that the major parties will control the Senate

B..2 That Proportional Representation will encourage the proliferation of minor parties.

In proportional representation systems where the order of election of candidates is predetermined by the parties utilising list systems this is often true. It is also partially true for election to the Australian Senate, where how-to-vote tickets and above-the-line voting ensures the election of major party candidates in the order that they appear on the ballot. This de facto list system generally supports the election of one minor party candidate at each Senate election.

In the proposed model minor parties, unless they represented a genuine and sizeable minority, would eventually find themselves squeezed out because of:

In Australia, the two parliaments where the lower house is elected by proportional representation, Tasmania and the Australian Capital Territory (ACT), also demonstrate the difficulty that minor parties have in gaining and maintaining any presence. Both of these parliaments have a majority government.

The Tasmanian House of Assembly has been elected by proportional representation since 1909. For most of the twentieth century only the two major parties were represented. In recent years a number of Greens have been elected. In 1998, in an attempt to cut or eliminate the number of Greens being elected, the then Liberal Party Premier supported by the Labor Party, reduced the number of members elected from each electorate from seven to five. At the subsequent election the Labor Party secured an absolute majority and the Greens were reduced to only one seat. An analysis of the votes seems to indicate that the same result would have been achieved even with seven member electorates. The 2002 election again saw the Labor Party achieve an absolute majority with the Greens picking up three seats at the expense of the Liberal Party.

The Liberals' tactic could hardly be described as a success. However, had they tried to eliminate the Greens by using single-member electorates the likely outcome is that the Labor Party, with 51.9%8, of the vote would probably have won all 25 seats in 2002, leaving 48% of the electorate without any representation.

The first election of the ACT House of Assembly saw the election of a large number of Independents and minor parties, most notably a group running on a ``no self government for Canberra'' platform. This group would have won more seats with a single-member electoral system. In subsequent elections the number of Independents and minor parties has steadily declined and the current House of Assembly has only one member who is not a member of either of the major parties. The Labor Party has majority control with nine of the seventeen members.

B..3 That minor parties will hold governments to ransom to get their own way.

This happens anyway and is the nature of politics.

The Country Party vetoed the appointment of Billy McMahon as Prime Minister after Harold Holt drowned.

The DLP held both the ALP and the Liberals to ransom through the long years of the Labor split without ever having a member in the House of Representatives. The Greens are attempting to dictate Labor Party policy on old growth logging and other environmental issues.

How much better for parliamentary democracy if the DLP and the Greens had been/were represented in the House of Representatives where their actions could be observed by supporters and opponents alike.

B..4 That governments formed will of necessity be coalition governments unable to make hard decisions.

Coalition governments are the norm in Australia. In the Federal sphere Liberal governments are invariably Liberal/National Party coalitions. Even when the Liberal Party had an absolute majority in its own right it maintained a coalition government. Labor Governments might appear to be monolithic but are in reality coalitions of their Left and Right wings with some Independents. These factions often caucus separately and always present separate slates of candidates for ministerial or front bench positions. The Liberal Party also has its own factions - the wets and the dries. Coalitions will work in Australia because once a coalition agreement is reached any falling out will be to the detriment of all parties to it.

B..5 That the Nazis came to power in Germany because of proportional representation.

Under proportional representation the Nazis made very little progress until the start of the Great Depression in 1929. Their largest vote in free elections was in July 1932 when they secured 37% of the popular vote and gained 38% of the seats. The rise of National Socialism in Germany needs to be assessed in its historical context of post war Germany and the Depression. Had the electoral system been single-member electorates, either first past the post, or the Australian system of preferential voting, this 1932 vote would have given the Nazis an absolute majority in the Reichstag.

In may be of interest to note that in 1948 the State of Israel established a proportional representation system for the Knesset using the entire country as one electorate. This system is very similar to the one used by the Weimar Republic.

B..6 That proportional representation ballots are hard to count and slow to count

For a detailed explanation of how to count a proportional representation ballot the reader should visit the Proportional Representation Society of Australia's website (http://www.prsa.org.au). The procedure, whilst detailed, is straightforward and any competent returning officer can conduct such a ballot. Computer programs have made these counts even easier and faster. On election night computer projections would give a very accurate final outcome.

In a close election the limiting factor in the time required for a result is the wait for postal and absentee votes to arrive. In the proposed model, because of the size of the electorates, the numbers of these votes will be greatly reduced. The count will also be much faster than a current Senate election because the spreading of votes over a number of candidates means that the number of preferences to be distributed will be smaller.

The Australian Senate and NSW Local Government elections have been conducted under proportional representation systems for many years.

C. House of Representatives election results since 1974

  1974 1975 1977 1980 1983 1984 1987 1990 1993 1996 1998 2001 2004
Labor 49.3 42.8 39.6 45.1 49.5 47.5 45.8 39.4 44.9 38.8 40.1 37.9 37.6
Liberal 34.9 41.8 38.1 37.4 34.4 34.4 34.3 34.8 37.1 39.0 34.2 37.4 40.8
National 10.8 11.3 10.0 8.9 9.2 10.6 11.5 8.4 7.2 8.2 5.3 5.6 5.9
Democrats - - 9.4 6.6 5.0 5.5 6.0 11.3 3.7 6.8 5.1 5.4 1.2
One Nation - - - - - - - - - - 8.4 4.3 1.2
Greens - - - - - - - - - - 2.6 5.0 7.2
Others 5.0 4.1 2.9 2.0 1.9 2.0 2.4 6.1 7.1 7.2 4.3 4.4 4.0

D. 1951 Tasmanian Senate Election

Ten senators to be elected: Quota for election 12,747.

John Chamberlain Lib 22,303 elected 6 year term
Hon James Guy Lib 8,948 elected 6 year term
Denham Henty Lib 6,343 elected 6 year term
John Marriott Lib 5,111  
Robert Wordsworth Lib 5,063 elected 3 year term
Reginald Wright Lib 25,272 elected 6 year term
William Lynch CPA 451  
Max Bound CPA 127  
William Morrow ALP 9,814 elected 3 year term
Hon Nicholas McKenna ALP 19,687 elected 6 year term
Charles Lamp ALP 3,253  
Justin O'Byrne ALP 8,530 elected 3 year term
William Aylett ALP 7,829 elected 3 year term
George Cole ALP 11,149 elected 3 year term
Reginald Murray ALP 6,328  
Total   158,568  


... ACT1
... Territories2
See section below which discusses the Territories.
... ACT3
...Comb Others4
Socialist Alliance, Fishing Party, Climate Change Coalition, Conservatives for Climate
...Comb Others5
One Nation and Socialist Alliance
...Comb Others6
Democrats, Socialist Alliance, Conservatives for Climate
...Comb Others7
One Nation, Climate Change Coalition
... 51.9\%8
ALP 51.9%; Liberals 25.4%; Greens 18.1%; Others 2.6%