PROPORTIONAL REPRESENTATION SOCIETY OF AUSTRALIA

Tel +613 9589 1802, +61429176725                                                        info@prsa.org.au

www.prsa.org.au



2016-09-26

 

The Gross Defects of Plurality (First-Past-the Post) Electoral Systems

 

INTRODUCTION:

The Plurality Concept: The term plurality is synonymous with the electoral terms first-past-the-post, relative majority, and simple majority. It refers to the basis on which votes are counted in order to determine who is elected by those votes. Such systems can, as shown below, be used for the filling of a single vacancy, or the filling of a number of vacancies as a group. Confusingly, the term simple majority - when used to describe voting on motions - traditionally means a vote just above a bare 50%, which is an absolute majority, as opposed to a larger majority that might be prescribed before certain motions can be carried. However, when 'simple majority' is used - often disingenuously - to describe vote-counting systems, it means the largest single group, which can be, depending on how many groups are contesting the election, well below 50% of all votes cast, easily as low as below 25% of the overall vote if there were as many as just five different groups.


A plurality electoral system is one where ballots are normally not valid unless they have been marked by the voter [traditionally with a cross (X) marked against the name of each candidate voted for] to indicate the candidate or candidates that the voter wishes to have elected, but no more candidates can be indicated in this way than the number of vacancies to be filled. Ballots are counted as described under “Counting of the Votes” below.

 

The American term plurality for this crude and primitive counting system is preferable to the British synonyms above, as first-past-the-post, which uses a racing term to beg the question as to what the “post” is, describes a quite different operation from judging which single candidate, in the simple case of a single position to be filled, should most appropriately represent a large number of voters. The racing jargon begs the question as to what the “post” is, by assuming that it is necessarily the largest number of first preference votes.

 

Proponents of "first-past-the-post" voting choose to ignore the fact that a much sounder and much more absolute “post” is the target of determining, as Australia’s majority-preferential system used in single-member electoral districts has reliably done, since 1918, which of the two remaining candidates polling highest - after the candidates receiving the fewest first preferences have been successively excluded from the count and their ballot-papers transferred to the remaining candidates - has received an absolute majority of votes. An absolute majority of votes is gained when one of the two remaining candidates gains more ballot-papers than the other. Relative majority uses the ambiguous term, majority, to mean what can be a minority of the total first preference vote if that is only 50%, or less, of the total vote.

 

The system for marking and counting ballot-papers in polls in Queensland for those of its municipalities, mainly rural shires, that have only a single electoral district - where that multiple plurality counting used for Senate polls in1903-17 still applies - now differs by requiring the voter to mark a sequence of numerals rather than crosses. Many voters there might be misled by that and not realize that their marking of such numerals is not given effect to as a transferable vote, as elsewhere in Australian public polls, but has the effect of allowing the Returning Officer to treat the number of numerals marked on a ballot-paper - up to the number of positions to be filled – each as a separate valid vote, and to disregard further such numerals that might be marked beyond the required number. That change conceals the crude plurality system actually used, by giving it a preferential appearance, but it has removed a longstanding difficulty with multiple plurality systems of voters marking more crosses than there are positions to be filled, which can lead to high rates of invalid ballots.

 

The English common law electoral system is a plurality system, but it allows ballots to indicate votes for fewer candidates than the number to be elected if the voter so chooses, which allows for the traditional plumping by voters although, as is explained below, that is not - unlike transferable vote systems - a fully reliable system for translating voters' intentions into electoral outcomes. More recent variants of that system are:

(a) the single non-transferable vote, which was used for Japan’s Lower House until it was replaced by a hybrid party-list PR and single-member system (not the same as New Zealand’s Mixed Member Proportional system) at the end of the 20th Century,

(b) the limited vote, and

(c) the now discredited multiple plurality system used for Australian Senate elections from 1903-17, which is referred to below and arbitrarily required that a valid ballot had to show exactly the same number of votes for candidates as there are vacancies to fill, which thus prevented plumping.

 

Form of Ballot-papers: Traditionally, ballot-papers for plurality elections invite the voter to indicate his or her votes by means of marking a cross (X) alongside the name of each of the candidates voted for, but more recent practice in some jurisdictions, such as certain Queensland local government polls and elections in certain Australian unions, has allowed the voter to mark his or her preferences among the candidates in order by marking the numbers 1, 2, 3, 4, 5 etc. against their respective names. In a country like Australia, where the predominant voting system is preferential voting, a benign interpretation of that practice would be that it was instituted to facilitate voting, although a less benign, but possibly more accurate interpretation, would be that it conceals from the less perceptive voters the nature of the system that applies.

 

Counting of the Votes: Where there are multiple vacancies to be filled, the non-transferable votes on each ballot-paper are counted as being of equal value to each other, even though a voter might have a distinct order of preference among the candidates, as there is no mechanism for such preferences to be given effect to.

 

Candidates are elected consecutively according to who receives the largest number of votes. There is no pre-determined percentage of the overall vote required to be gained before a candidate is elected (hence the description "relative majority"), so a candidate can be elected with a very much smaller percentage of the vote than under any other electoral system if the votes are spread very evenly among the candidates. In elections to fill a single vacancy or elections by the plurality bloc vote referred to below, all the successful candidates can be elected by the votes of very much less than 50% of the voters, with well over 50% of the voters having no effect on the outcome whatsoever, unlike preferential voting, which ensures that a majority, or quota, as the case may be, of voters decides the outcome for each vacancy.

 

PLURALITY VOTING SYSTEMS FOR FILLING A SINGLE VACANCY: These systems, for parliamentary and congressional elections, are now largely confined to parts of the former British Empire such as the United Kingdom, United States of America, Canada, India and Pakistan, although Australia, New Zealand, Eire, South Africa, Malta and Sri Lanka have changed to other systems. These plurality systems are badly flawed as they allow members to be elected in each electorate with well below 50% of the vote, and allow splitting of the vote among candidates of similar views to the benefit of more isolated candidates. Such systems are also used for electing single office-bearers, such as the President of France. The non-transferable nature of the votes cast is the main fault there.


A 2011 referendum in the United Kingdom allowed voters to choose between that country’s longstanding plurality system in single-member electoral districts used to elect the members of the House of Commons, and the “Alternative Vote" (AV), which is the system of optional preferential voting used to elect the Legislative Assembly of New South Wales. Australia's House of Representatives abandoned plurality voting in 1918 and has used the Alternative Vote ever since, but with full marking of all preferences being required for a valid vote. Just over two-thirds of voters agreed with the Conservative Party Prime Minister, David Cameron, who campaigned, with his party, to keep the UK’s existing system, although his Coalition partners, the Liberal Democrats favoured AV, as did the Labor leader, and many in his party.

 

MULTIPLE PLURALITY VOTING SYSTEMS: The PRSA History page refers to the quite unfair and unreasonable electoral system used for the election of the Australian Senate from 1903 to 1917. That system also applied, by default, to the first election in South Australia for members of the House of Representatives in 1901. It is variously called a multiple "X", a multiple relative majority, a multiple first-past-the-post, a plurality at large, a block vote, or a multiple plurality vote. If an electoral system is unspecified in a constitution, the multiple plurality system is unfortunately the common law default system. See the example of the ACF, a major Australian body that replaced its multiple plurality system with PR in 1974.

 

Under the common law multiple plurality system, a voter traditionally marked an "X" against the name of each candidate voted for, with the consequential proviso - given the very rudimentary provision for the expression of a voter's wishes that plurality systems allow - that the ballot would be invalid if an "X" appeared against the names of more candidates than there were vacancies to be filled. A later, even less democratic variant, used for Australia's 1903-17 Senate elections, arbitrarily specified that a ballot would be invalid if the number of candidates voted did not equal the number of vacancies, thus forestalling the traditional option in plurality elections of plumping. The term plurality refers to the largest vote gained by a candidate or candidate, even if that candidate or candidates have failed to gain an absolute majority of votes.


The unfair consequence of that arbitrary prohibition of plumping is illustrated by the examples in the following table:


HOW PROHIBITING PLUMPING IN MULTIPLE PLURALITY POLLS CAN BENEFIT INCUMBENTS
Number of vacancies
Number of incumbent candidates, which cannot exceed the number of vacancies
Number of non-incumbent candidates, which may be any number from zero upwards
Explanation of how the balance between the number of
incumbent and non-incumbent candidates has an undesirable
effect on the outcome of the poll, which is not the case with
preferential voting with the single transferable vote.

a b The percentage in red below equals 100 times a divided by (a+b).
2
2 1 Voters opposed to the incumbent candidates are forced to lodge a vote of equal significance for one of those candidates also, and are unable to concentrate their vote solely on the only candidate they want elected, which is the sole non-incumbent candidate. The 2 incumbents could be re-elected with as low as 34% of voter support.
2 2 2 For the reasons given above and below, It is only when the number of incumbent candidates and non-incumbent candidates is equal that a bare 51% of voters can be sure of electing the combination of such candidates that they vote for under the crude multiple plurality system.
2 2 3 Voters for the incumbent candidates have, by definition, only 2 candidates to vote for, so their votes are concentrated on the 2 incumbents. Other voters could spread their 2 equal votes over 3 candidates, so an even splitting of that vote could mean that the 2 incumbents could be re-elected with as low as 41% of voter support.
2 2 4 Voters for the incumbent candidates have, by definition, only 2 candidates to vote for, so their votes are concentrated on the 2 incumbents. Other voters could spread their 2 equal votes over 4 candidates. An even splitting of that vote could mean that the 2 incumbents could be re-elected with as low as 34% of voter support.
2 2 5 Voters for the incumbent candidates have, by definition, only 2 candidates to vote for, so their votes are concentrated on the 2 incumbents. Other voters could spread their 2 equal votes over 5 candidates. An even splitting of that vote could mean that the 2 incumbents could be re-elected with as low as 29% of voter support.
3 3 9 Voters for the incumbent candidates have, by definition, only 3 candidates to vote for, so their votes are concentrated on the 3 incumbents. Other voters could spread their 3 equal votes over 9 candidates. An even splitting of that vote could mean that the 3 incumbents could be re-elected with as low as 26% of voter support.

  

The 3 systems used for Senate polls: For Australian Senate elections, once a uniform system applied after the first election of senators, the three systems that have been used - worked examples of which appear here - were:

·         from 1903 to 1917, that multiple plurality system described above, with plumping prohibited,

·         from 1919 to 1947, a multiple majority-preferential electoral system, and

·         from 1948 onwards, the quota-preferential (single transferable vote) system of proportional representation (PR).

 

The unfortunate 1919-47 system was used for certain municipal polls in New South Wales and the Northern Territory until 2012, and in Victoria until 2003. A routine outcome of the winner-take-all Senate elections before 1949 was that candidates of a single party, or point of view, won all of the vacancies for a State, and regularly, once in each of the first five decades of the Senate's existence, for the whole of Australia, as occurred in 1910, in 1917, in 1925, in 1934 and finally in 1943. The multiple plurality system existing before 1919 had similar effects, as the 1910 and 1917 examples show, but it also occasionally filled all vacancies for a State with candidates from the same party that had collectively failed to gain even 50% of the votes in that State, as happened when Vida Goldstein split the vote in Victoria in 1910. Bizzarely, and alone among Australian jurisdictions except for rural Queensland, Western Australia has abandoned its long usage of preferential voting in muncipal polls, and has instituted plurality voting.

 

Preferential systems predominating: The Commonwealth Electoral Bill 1902, as passed by the House of Representives, proposed the use of majority-preferential voting in the single-member divisions proposed for the House of Representatives, and quota-preferential voting in the States, each of which was to form a single multi-member electoral district, for the Senate, but the Senate successfully amended the Bill to provide for plurality voting in each House. Preferential voting was first applied for House of Representatives elections in 1918, and it now applies for elections to all Australian legislative chambers, for both single-member and multi-member electoral distrricts. Australia's most recent type of preferential voting in single-member electoral districts involves optional preferential voting, as is now prescribed in the NSW Constitution Act 1902.

 

Provision for the type of preferential system that generally replaced plurality systems in Australia, when the Commonwealth and all States except Queensland made that change around 1920, first appeared, in Tasmania, in two multi-member electorates for its House of Assembly in 1896, and was extended to all 5 Tasmanian Assembly multi-member electoral districts for the 1909 General Election. Preferential voting in Tasmania's single-member Legislative Council electoral districts first became law in 1907 but, over 100 years later, multiple plurality systems still persist in some Australian non-parliamentary elections. Preferential systems using the transferable vote allow voters to indicate their ranking for all the candidates standing, and for those indications to be brought into effect in the final result, whereas plurality systems limit voters to indicating their support for no more than the number of positions to be filled, and they do not alow voters to indicate their priority order among the candidates.

 

Prominent examples of plurality systems persisting are elections for:

·      WA municipal councils, where plurality systems have been restored for the second time,

·      local government in some rural parts of Queensland,

·      committees elected by all of Australia's Anglican Diocesan Synods, except Melbourne's, which now uses PR,

·      the National Committee of the Australian Republican Movement, and

·      the board of the Royal Automobile Club of Victoria (RACV).

 

Flagging Incumbent Candidates: The Royal Automobile Club of Victoria - as did the National Roads and Motorists Association of New South Wales before it was privatized - compounds the defects of the multiple plurality electoral system it uses by having ballot rules that require the names of candidates that are retiring directors to be singled out and highlighted on the ballot papers, and on the candidate statements, with an asterisk (*), rather than letting such candidates include that fact in the candidate statements they lodge. That institutionalized "helping hand" has the effect of concentrating establishment votes on those candidates alone, and there are, by definition, never more of such candidates than the number of positions to be filled. In the 2014 RACV poll for service members, it was stated, on the ballot papers, that the asterisked candidates were not only recontesting Board members, but also that the Board supported their re-election.

This inequitable ballot rule can make it more likely that votes for other than incumbents will be dispersed, or split, among the field of non-incumbent candidates, which is often larger than the field of incumbent candidates, as the field of incumbent candidates is obviously limited to the number of places available, whereas the others are not. No means are provided for the voters to have their votes transferred, in accordance with preferences they indicate, from poorly-supported candidates to better-supported candidates of their choosing, in order for their votes not to be unnecessarily wasted.


An unsatisfactory and unnecessarily onerous alternative to marking an “X” against the name of each candidate, which is still used by some groups, was the rule, for the first election of senators for NSW, that voters had to - in order to cast a valid vote - strike out the names of all the candidates they were not voting for, so that exactly the number of candidates they were voting for would remain (this was the only Senate polling day where the State electoral rules still applied, so plumping was not prohibited). There were fifty candidates for those six positions, so it is not surprising that 17.5% of the ballots cast were ruled to be informal (invalid).


Single Non-transferable Vote systems: Another variant of plurality systems is the Single Non-transferable Vote, which is used to elect most of the members of the Upper House of the Japanese Diet. In this system of election to fill a number of vacancies simultaneously, each voter is allowed to mark a maximum of one vote against one of the candidates' names. Sometimes the result of a count using the Single Transferable Vote (quota-preferential proportional representation) can be the same as one using the Single Non-transferable Vote if the subsequent preferences marked using STV happen to be marked in a way that produces the same result, but it is by no means reliable or equitable, as voters for very popular candidates can find their votes to have been wasted if such candidates poll better than they need to secure election, as there is no provision for the transfer of surplus votes as there is with the preferential voting in STV. Similarly, large numbers of voters for unelected candidates can find their votes to have been totally ineffective, as there is no provision for their votes to be transferred to candidates that turn out to have better prospects of being elected. The last candidates elected in an SN-TV poll often have very little support.

 

DETAILS OF THE GROSS DEFECTS:

Failure to ensure representation of over 50%: The plurality, relative majority, or "first-past-the-post" idea ignores the fact that the candidate or candidates with the largest number of votes does not necessarily gain the support of any predetermined percentage of the votes. All that is required is that the winner's vote total be higher than the rival totals, yet that total might be so low as to leave an absolute majority of voters completely unrepresented by any of the multiple candidates elected. This objection applies even when only one vacancy is to be filled.

 

A majority-preferential voting system (Senate 1919-48) requires that the views of an absolute majority of voters, which is a number exceeding 50%, must be given effect to. A quota-preferential PR system (Senate 1948-) gives effect to well beyond 50% of the votes cast (all but a small percentage of unusable votes).

 

Splitting the votes of a group owing to a proliferation of similar extra candidates: All plurality systems suffer by leaving voters in a quandary if there is a relative abundance of candidates with a particular viewpoint compared with those of opposing viewpoints. A recent example was the first round of the 2002 French presidential election, which resulted in even the strongest-polling candidate, the outgoing President, Jacques Chirac, from the right, gaining less than 20% of the vote. The left was also split, but slightly more so, with the result that the second round, to choose between the two highest-polling candidates, pitted the right against Jean-Marie Le Pen, rather than its realistic opponents, the left. The result was a sham poll, where Chirac took nearly all the votes, owing to the lack of a realistic opponent.

 

Wasting the votes of many of those voting for very strongly supported candidates: Plurality systems being used to fill more than one vacancy as a group, whether by the multiple plurality system described above, by cumulative voting, or by a limited non-transferable vote suffer the flaw that very strongly supported candidates can gain far more votes than they need for election, thus depriving their supporters of the ability to have the magnitude of their vote translated into an appropriate number of seats won, leaving many seats won by very small numbers of votes left after a single, or a few winners have gained most of the votes, with no provision for transferring those votes as happens in the transfer of surplus votes in a quota-preferential PR election.

 

Unfairly forcing some voters to support candidates they oppose: The non-preferential nature of the multiple plurality system used in the above examples is grossly unfair to voters that find the number of candidates they consider to be acceptable is fewer than the number of vacancies to be filled. Such voters are required, in order to cast a valid vote, to mark the balance of their vote for candidates they oppose. The multiple plurality system prevents voters indicating the ranking of their preferences for the individual candidates, and forces voters to give them all equal weight. The example in the table below shows how voters' power to give effect to their priorities by marking preferences is suppressed.

 

Plumping: The last flaw above does not apply when only one vacancy is to be filled, nor does it apply to the traditional, common law, form of plurality voting, which allows voters to vote for fewer than the number to be elected, a practice known as "plumping". Plumping occurs when voters  limit their votes to those candidates they support, rather than being constrained to also vote for those they oppose, in cases where those voters support fewer candidates than the number to be elected, or when they estimate that they are not supporters of the slate of candidates that are likely to be supported by the largest group of voters and that their only hope of any representation is to limit their vote to candidates that voters other than the largest single group of voters supporting a number of candidates equal to the full number to be elected are expected to support. Plumping was the subject of a rather confused Hobart City Council debate in 1916.

 

That traditional form was practised in British elections prior to the general abolition of two-member geographical electorates for the UK House of Commons in 1885, as it allowed a crude form of representation of minority opinion where such minorities were large enough. It is a crude form because there is no predictable quota as in PR counting so – depending on the size of the vote for the more strongly supported candidates – a minority vote can elect one or more candidates with an unduly low vote, or it can elect one or more candidates but have a wasted surplus vote that is not large enough to elect a further candidate. In that latter alternative those surplus votes are wasted, and the more strongly supported candidates can be elected with a lower vote than they otherwise would require, by the removal of the competition that surplus should give.

 

 

HOW AN 80% VOTE WINNER IS DEFEATED BY AN UNFAIR ELECTORAL SYSTEM

 

Multiple plurality systems that prohibit plumping prevent voters implementing

their priorities, by arbitrarily suppressing notional preference votes.

 

Even where plumping is allowed, voters often fail to understand its value, as they cannot be sure – as they could with

quota-preferential proportional representation – that their votes will not be over-concentrated on too few candidates.

 

 

51 VOTERS i-li

 

6 CANDIDATES A-F, FOR 5 VACANCIES

 

A

B

C

D

E

F

i

1

2

3

4

5

 

ii

5

2

3

4

 

1

iii

 

2

3

4

5

1

iv

5

1

2

3

4

 

v

4

5

1

2

3

 

vi

3

4

5

1

2

 

vii

2

3

4

5

1

 

viii

2

3

4

5

 

1

ix

3

5

2

4

 

1

x

1

5

3

2

4

 

xi

4

3

2

 

5

1

xii

5

 

3

4

2

1

xiii

3

2

4

 

5

1

xiv

4

3

 

5

2

1

xv

 

3

2

4

5

1

xvi

5

 

2

4

3

1

xvii

4

3

 

5

2

1

xviii

 

4

5

2

3

1

xix

2

 

4

5

3

1

xx

4

3

 

5

2

1

xxi

3

1

2

5

4

 

xxii

4

5

3

 

2

1

xxiii

 

4

3

2

5

1

xxiv

3

5

2

 

4

1

xxv

4

5

 

2

3

1

xxvi

4

 

3

5

2

1

xxvii

4

3

5

1

2

 

xxviii

4

5

3

2

 

1

xxix

2

4

3

5

 

1

xxx

2

3

4

5

 

1

xxxi

 

5

3

4

2

1

xxxii

4

 

5

2

3

1

xxxiii

5

2

 

3

4

1

xxxiv

4

3

2

 

5

1

xxxv

 

2

3

4

5

1

xxxvi

2

 

5

4

3

1

xxxvii

4

3

5

 

2

1

xxxviii

4

5

 

3

2

1

xxxix

2

3

4

5

1

 

xl

5

4

2

3

 

1

xli

 

2

3

4

5

1

xlii

3

 

4

2

5

1

xliii

4

5

1

3

2

 

xliv

2

3

4

 

5

1

xlv

5

4

3

2

 

1

xlvi

2

3

5

4

 

1

xlvii

 

4

5

3

2

1

xlviii

5

3

4

 

2

1

xlix

4

 

5

2

3

1

l

2

3

 

4

5

1

li

2

3

 

5

4

1

NO. OF "X" VOTES *

43

43

43

43

42

41

NO. OF NOTIONAL FIRST

PREFERENCE VOTES

2

2

2

2

2

41

 

* NOTE: To calculate this line, each of the notional preferences (1-5) above has to be treated as just an "X".