With these elections, the practice of parties issuing 'how-to-vote' instructions is firmly established. The combination of this practice with the use of single-member districts has had many unfortunate results, probably the most serious being that candidates and Members of Parliament are strongly encouraged by the system to be concerned about the views of the few people who control party endorsements rather than those of the voters. 

The Provisions for elections of the Federal Senate are much more satisfactory than those for the House of Representatives or for State Lower Houses other than the Tasmanian House of Assembly. Being designed to give proportional representation, the method used in Senate elections at least ensures that parties are represented approximately in proportion to the voting support they receive. 

Unfortunately, the practice of parties issuing how-to-vote instructions is also followed in Senate elections. In States other than Tasmania, voters have come to regard this as a normal procedure and most of them follow the instructions of one party or other. The practice has therefore had the effect of limiting the capability of the method to translate the considered choice of the voters into effective representation, and, with some important exceptions, it has operated almost as a party-list method. 

Tasmanians and all of those elsewhere who are interested in strengthening and improving democracy are fortunate to have someone of Dr. Howatt's qualifications and experience to investigate problems in the machinery of elections and representation. Even In this interim report, he has made a significant observation that may be new even to most of those concerned with the operation of electoral systems, at least in the non-Tasmanian States. He has shown that the use of Senate-style voting tickets, in contrast to the free selection of candidates provided for voters under the Tasmanian-pioneered Hare-Clark system, can actually affect adversely the parties that use the tickets. 

Tasmanian voters are fortunate to have the freedom of choice and accurate representation that the Hare-Clark method ensures. Any suggestion of a major departure from existing practice should be scrutinized very carefully to avoid the possibility of loss of any of the special characteristics of the method that has served the voters of Tasmania so well in the past. 

J.F.H. Wright, B.Sc., President of the N.S.W. Branch of the Proportional Representation of Australia. 

C.G. Ball, Returning Officer for Tasmanian House of Assembly elections and Commonwealth Divisional Returning Officer for Denison 

K. F. Febey, State Chief Electoral Officer of Tasmania. 
 

The writer of course is solely responsible for the views in the report. 

The various alternatives to the Robson bill need to be examined before the bill itself can be assessed properly. One alternative which has been mentioned by some as a possibility is the method used by the major political parties and others for (a) determining the position of candidates' names within groups on the ballot-paper for Senate elections and (b) for directing electoral support to particular candidates. 

2. What is the Senate method? 

In effect, the method empowers the management of the contesting parties, firstly, to list the names of its candidates on the ballot-paper in whatever order it wishes (irrespective of specific criteria such as alphabetical order or drawing by lot) and then to so advise the electoral authorities. Secondly, party management then issues "how-to-vote cards with specific preference numbers printed beside the names of the various candidates. Party management then appeals to its electoral supporters (through the distribution of these cards at Polling booths and by advertising them in the press and elsewhere) to copy onto their ballot-papers the numbers exactly as they appear on the how-to-vote cards. The parties under this system are in effect saying to their potential supporters "Please take our voting card, but don't think for yourselves; merely copy as we have directed, without any selection on your part, the preference numbers onto your ballot-paper precisely as the numbers appear on our card". 

However, the political parties in Senate elections (in contrast to the model furnished by Assembly elections in Tasmania) do not offer a wide choice of candidates, but instead restrict the choice to a minimum, namely, to only three candidates even in the case of the major parties, where there are, nevertheless, five vacancies to fill. In effect, the controllers of these parties say to their supporters, "We'll offer you fewer candidates than there are vacancies to fill, thus reducing to the lowest limit the number of choices we make available to you, even though the electoral system readily enables a party to give its supporters a wide choice of candidates". In short, party management in Senate contests is saying to the voters "We don't want you to think for yourselves; instead we only want you to be ciphers for us, in order to copy down numbers according to our directions". ⁽¹⁾ 

Such a method is highly undesirable even for Senate elections as will be explained in the final report. However, since elections for the Tasmanian House of Assembly are conducted on a far higher plane and in a very different manner from Senate elections, these differences would make Senate-style tickets both impractical and most undesirable for electing Members to the Assembly, as will be detailed in the final report. This final report will further point out how attempts to apply Senate-style regimented voting to Assembly elections also would jeopardize (if not rupture from the start) the continuance of the two-party system in Tasmania, the stable functioning of Parliament, and the control - strong and healthy - (as compared with other States and nations) which the voting public in Tasmania now exercises over its Parliament - a control unique in its excellence. Further, the final report will explain how this precious heritage special to Tasmania - and which could serve, if known, as an urgently needed model for the rest of the world - would be severely damaged, if not destroyed, by the introduction of Senate-style regimented voting for Assembly elections. ⁽²⁾ 

Readers wishing a few words of assessment of the system, such as can be covered in a brief newspaper article, could refer to Appendix E of this paper. A longer, but still very summarized assessment, may be found in Tasmanian Parliamentary Paper No. 22 of 1958, by the writer entitled Democratic Representation under the Hare-Clark System.

The writer is deeply proud to be regarded as a fervent admirer of the Tasmanian-pioneered Hare-Clark system, motivated by his belief that the inherent features of this special method could solve problems of representation in many lands, thereby enabling these States and nations to meet the challenges of change, to respond to the needs of their people, and solve problems which at present seem to be increasing in almost all countries. 

Nevertheless, the writer's admiration and support for the Hare-Clark system has not blinded him from seeing ways in which the application of Hare-Clark principles could be refined and strengthened. Evidence of his wish not only to commend the Hare-Clark system but also to seek improvements in applying its principles could be demonstrated in many ways. One is to note the appearance of the twenty-one articles listed below, published by "The Mercury", all of them written to advocate adoption of some refinement to improve the system; this and other evidence do not show complacent and uncritical admiration of all aspects of the way the Hare-Clark system is at present applied. 

The writer has undertaken various major research projects on electoral systems, including, for example, a Ph.D. thesis for the University of Tasmania focussed on the Hare-Clark system. Years of research and thought on electoral systems and practices leave the writer with no doubt that the use of numbered how-to-vote cards by political parties is unqualifiedly undesirable and would be a most retrograde step, for reasons which the final report will set out. 

In short, regimented voting, Senate-style can under-represent the parties supported by the greater number of voters and over-represent the parties or groups attracting less support. The fact that such regimentation does under-represent the major parties and over-represent minor ones can be seen clearly from examples cited here in Part II. More will be cited in the final report. 

Under the time-tested Hare-Clark feature of giving the voters a free selection of candidates, the larger parties gain their fair share of the seats while not depriving any eligible minority of their fair share also. If any party or group obtains a full quota, or close enough to a full quota, it will win a seat as it is justly entitled to, under Senate or Hare-Clark conditions. 

However, suppose no party has sufficient votes (in primaries plus possible accumulated preferences) to fill a quota for the last seat to be filled - a common enough situation: who is entitled to that seat when two or more have partial quotas but not full ones? Under usual Hare-Clark conditions this problem is answered easily and naturally enough (as regards the vacancies within an electorate) - namely, by the rule of the order of elimination of candidates. 

In practice, owing to the fact that candidates, in the absence of numbered how-to-vote cards, seek support as individual persons, there is a natural spread of support (in primary votes and preferences) over many candidates. When surplus votes are transferred, as well as the preferences of excluded candidates, the votes normally are distributed - owing to absence of numbered how-to-vote cards - over the range of the continuing candidates. 

Under usual Hare-Clark conditions the final seat in an electorate will be won by the party which shows the highest average no. of votes per candidate at the time of the last exclusion. In other words, when voters are given the opportunity, as provided by existing Hare-Clark conditions, to select candidates freely as they choose, a normal spreading of support for a party's candidates occurs, being reflected in a well-spread primary vote or, as the case may be, during the distribution of surplus votes. 

An equitable criterion for allocating Parliamentary seats is that the members of Parliament should be elected in proportion, as nearly as possible, to the number of votes of their supporters. The Hare-Clark system, which allows a natural spread of support as a result of the free, undirected marking of choices by the voters, normally meets this criterion when filling the vacancies of a given electorate. 

In contrast, the Senate practice of electing candidates, based upon regimenting the vote by means of party how-to-vote cards, has an in-built feature which distorts this criterion. In practice, and conspicuously so in the Mainland States, the No. 1 candidate of each party is elected with a huge, abnormal surplus, which is passed on to the No. 2 candidate who is also elected with a surplus. This means, in the usual 5-vacancy Senate election, that the No. 3 candidate of the party is left with many fewer votes than if his party's vote were more evenly spread over all its candidates. As a consequence, the No. 3 candidate of each or both parties is left in a weaker position. 

As regards major parties, spreading the vote (as occurs in Hare-Clark practice) conserves their voting strength by enabling their candidates to stay in the count longer, thereby increasing their chances of winning more seats, namely, winning -the number which their vote would, on a proportional basis, entitle them to. 

In these cases both major parties were more entitled, on a proportional basis, to the seat, won by the minor party or Independent. In other words, both major parties forfeited an opportunity to win an additional seat by wasting their strength, through regimented voting, with huge surpluses for ?rds of their candidates while the other ?rd was forced, unnecessarily by normal Hare-Clark practice, into elimination. 

Senate-style regimented voting therefore not only deprives, in practice, the elector of a free choice of candidates, it also can, and frequently does, deprive the supporters of major parties of the additional seats to which they would be entitled on a more fairly proportional basis. 

For a variety of reasons., some of them given in the appendices, the functioning of the Senate would be improved incomparably by the adoption of Hare-Clark practice in electing it. Conversely, the use of Senate-style regimented voting would produce disastrous consequences if used for electing the Tasmanian House of Assembly, for reasons noted in this interim report plus those to be covered in the final report.

Regimented voting would certainly reduce the vote-attracting ability and, consequently, the vote totals of any major party which attempted to apply Senate-style voting, which, bad enough for State-wide electorates and with restricted nominations, would prove disastrous for any major party which attempted to apply it for State elections - for reasons to be detailed in final report, as they are too numerous for including in this interim report. 

For example, there is no known realistic way (at least none known. to the writer) for determining workable criteria for ranking a preference order for candidates when the number of endorsements (as is the case under Hare-Clark conditions) is roughly twice the number of seats which a party can expect to win. A party ticket under these circumstances (which obviously are quite different from those now in practice for the Senate) would have to declare publicly its recommended order of preferences. If the preferences were intended to be heeded (if not why recommend any), they would, in effect, be saying which 3 of its candidates were assured of seats in Parliament and which 3 should be consigned to defeat. 

Criteria to be workable would have to win the general consent of the party and the community concerned. Hence, the criteria would not only have to be fair, but be seen publicly to be fair. In the absence of fair criteria, what basis for selection could be used - only some form of lobbying and making deals which certainly would rupture internally a party which attempted to put it into effect. Inevitably such internal disruption would spread disaffection among the voters, and cripple any major party experiencing such internal trauma, as will be explained in the final report. 

  2. 
 

  3. 
 

 4. 
 5. 

 6. 

8. 
 

9. 
 
 

10. 

11. 
 
 

12. 

As the reader will realize, the above is only a small sample of considerations yet to be analysed and weighed. These subjects and a further examination of some of the material in this interim report will be dealt with in the final report.

Hence, what is required is machinery for ascertaining the people's judgment, and already the Hare-Clark system does this better than any other method in the world, despite the need for certain refinements in applying the Hare-Clark principles. 

A key necessity for gaining a reliable indication of the judgment of the voters is to give them a choice both between political parties (to identify their broad policy preferences) and between individual candidates within that party (so that they, the voters, can select on the basis of the differing personal qualities between individuals of the same party philosophy). Of all electoral methods the Hare-Clark system provides these 2 vital requirements. 

Alone of all electoral systems, the Hare-Clark method allows the voter to make separate judgements both on the party policy and on a range of candidates within that party; in other words, the unique, Tasmanian-pioneered Hare-Clark system provides the means for enabling voters to judge effectively both measures and men, both policies and people. 

Under the single-member system, with only one candidate per electorate per party, an elector may be forced to vote for a candidate whom he does not favour personally because this candidate's opponent belongs to a party whose policies he finds unacceptable. Under most list systems of proportional representation the voter is given either a restricted choice or no choice between candidates of his preferred party (in any case, no range of choice comparable to the unique range offered by the Hare-Clark system); hence, the possibility which is an essential ingredient for expressing an adequate judgment, namely, an opportunity to choose between a range of candidates with the same general party philosophy, is denied under both the single-member system and, the list systems of proportional representation. 

Most fortunately, this opportunity does flourish under the Hare-Clark system as it now operates, by assuring the electors a free and unfettered choice between parties and, between individual candidates within parties. Senate-type regimented voting would (if used as applied federally) eliminate restrict, or lessen this choice - hence one of the most precious criteria for assessing the basis of representation and for determining the quality of government, would be impaired, if not destroyed, by use of Senate-style voting tickets. 

To recapitulate: before any party candidate can be elected to the Tasmanian House of Assembly, he must pass two tests, first approval by his party pre-selectors for nomination as a candidate of his party. Next, he must win the approval of the people expressing themselves through free choice. This double test is unique to Tasmania and is an invaluable heritage which needs to be passed on to other countries. But this heritage would be destroyed if the pre-selectors were given both the power to endorse and, thereby in effect, the power to elect through control of the placement of names on the party ticket and/or ballot papers.

APPENDIX A. - from "The Mercury", Hobart, December 17, 1956, entitled 
                            "Hare-Clark Remedy for Defects in Senate Voting". 
APPENDIX B - from "The Herald", Melbourne, December 17, 1958, entitled 
                           "Senate how-to-vote cards boomerang VOTE-TO-ORDER DOESN'T PAY" 
APPENDIX C - from Brian Fitzpatrick's LABOR NEWS LETTER, March, 1959, entitled 
                           "The Bad Luck of Mr. Little". 
APPENDIX D - from "The Mercury", Hobart, February 27, 1962, entitled 
                           "Weakness in Senate Poll" 
                            from "The Examiner", Launceston, March 6, 1962, entitled 
                            "Solution to Senate Voting Faults". 
APPENDIX E - from "The Mercury", Hobart, December 20, 1976, entitled 
                            "Hare-Clark Proved Again", is included to provide a resumé, 
                            however brief, of some of the merits unique to the Hare-Clark system. 

For various reasons (some of them given in the article which follows) the writer considered the suggestion of following this Senate practice a very retrograde idea, and wrote this article in the hope of showing that instead of adopting Senate practice for Hare-Clark elections, the practice of regimenting the voters should be dropped and replaced, instead, by the Hare-Clark method of allowing the voters to select candidates as they pleased, without the use of numbered tickets. 

The writer showed the draft of the article to recently retired Premier Cosgrove, who read it and said he had no alterations to suggest. Re the idea of using regimented tickets for Hare-Clark elections [which of course are quite different from Senate contests in that considerably more candidates need to be endorsed by a party for Hare-Clark elections than it can hope to elect] Sir Robert Cosgrove rejected it emphatically and without hesitation said, 

"With Senate-style regimented voting the first three candidates on a ticket wouldn't need to work [since they would be elected anyway, if the ticket were followed] and the last three or four candidates (depending on whether seven or eight were endorsed by the party) wouldn't have any real incentive to work, since they would have no chance of being elected. Besides, the rivalry among candidates and their supporters for a winnable place on the ticket would result in such in-fighting within the party both before and after the nominations that the whole idea of using numbered cards was a sure way to put the cat among the pigeons". 

(The complete statement by former Premier Cosgrove will be included in the final report). 

The article entitled "HARE-CLARK REMEDY FOR DEFECTS IN SENATE VOTING" follows:- 

Although the method for electing the Senate has many points of similarity with the Hare-Clark system, it falls short of the Hare-Clark standard in several important ways 

These defects in the Senate system include the compulsory marking of preferences for all candidates. This, a carry-over from the former, non-proportional method of preferential block voting is not necessary. It is a senseless provision serving no good whatever. 

It results in an excessive amount of informal voting. Under Hare-Clark voting, informality averages around 4 per cent in contrast to an average of around 11 per cent for Senate elections in Tasmania. 

Many Tasmanians, who know that marking three preferences is sufficient for a valid vote in State elections, follow this habit when voting in Senate elections, thus making their votes informal. 

Marking every square is an unreasonable burden. 

The requirement of compulsory numbering of all candidates is offensive psychologically because it compels the elector to express preferences for opponents to his own party or for candidates whom he may despise. 

An elector should be free to vote for as few or as many candidates as wishes. To force an elector to vote for candidates he does not want, in order to support the candidates he prefers, is a serious infringement of voting freedom. This infringement may sometimes alter election results. 

Each elector is given, in effect, a single vote. Once this vote helps to elect a candidate, it is unnecessary thereafter to examine or consider it. Most ballot papers are never examined beyond the second preference. 

Roughly speaking, fewer than 15 per cent of the votes are examined beyond the third preference in most elections. Hence, marking a number in every square on the voting paper is a futile effort. 

Compare, for example, the experience revealed by the Hare-Clark system, which requires a minimum marking of only three preferences. In practice, most electors under Hare-Clark vote for the candidates of their own party, and then stop. 

One effect is that independent and minority party candidates in Hare-Clark elections must poll almost a quota of primary votes to win elections. 

If voters were required against their will, to number all candidates, the effect in past elections would have been to elect some independents who were not, in fact, able to reach quotas under existing conditions. 

If supporters of minority party candidates are numerous enough to elect a senator, they are justly entitled to the seat. But if the seat is won by preferences obtained compulsorily from major parties, the minority party obtains more representation than it should. 

A second defect in Senate voting, especially as seen in Mainland elections, is the practice of the political parties in advising their supporters to vote in a regimented 1-2-3, down-the-ticket fashion. 

Under Hare-Clark elections it is interesting to note in contrast that Tasmanian parties do not endeavour to dragoon their supporters into voting in a prescribed order. Consequently, Tasmanians exercise their judgment and choose between candidates in State elections. 

Candidates owe their election, therefore, to the support they receive from the electors, not to their position on the ballot paper. 

Elections under Hare-Clark conditions are consequently more healthy and real, resulting in better representation of the voters and higher prestige for elected members. In the party-regimented Senate elections of the Mainland, electors become conditioned to party tickets, and the individuality of Senate candidates is largely lost. 

Instead of permitting the electors to spread their primary votes naturally over several candidates, as under Hare-Clark elections, they do their best to concentrate votes on the No.1 candidate and thus give him an abnormal surplus. 

This practice of regimenting the vote would have resulted in the election of Mr. Little, DLP candidate for the Senate in Victoria, if there had been five vacancies to fill instead of six. 

A party's best interest is served by keeping as many of its candidates in the count as long as possible. Under unregimented conditions this occurs as a result of the natural spreading of support by the voters. 

In contrast, under the prevailing regimented conditions, a party wastes its voting strength by electing its No. 1 and No. 2 candidates with full quotas early in the count. This leaves fewer votes for their No. 3 candidate, who therefore could be forced into elimination unnecessarily. 

The basic features of the Senate PR system are splendid and yield representative results incomparably superior to the lopsided, "all or none" results of the old Senate system which preceded it. But despite its general excellence, the present method requires further reforms, for which fortunately, the Hare-Clark system can serve as a guide.

The first is due to the party regimentation of voters through how-to-vote cards. The big parties can actually hurt their chances by such cards, because in a normal five-vacancy poll, concentration on a No. 1 candidate can distort the total vote and help a minority party. 

The second fault is the compulsion on the voter to fill in a preference for every candidate. This is unnecessary and mainly futile. 

Basically, the proportional representation system is sound and fair. It has eliminated the grossly lop sided, violently-fluctuating "all or none'' of the former preferential block system. 

That system once produced a lone Labor Senator and 35 non-Labor members. And in 1946, 33 Labor Senators were returned with an Opposition of only three. 

In the last election the compulsory marking of all preferences made the informal vote shockingly high in all States. 

Once again, the parties regimented the electors with numbered how-to-vote cards. 

In Victoria, the election showed that such regimentation could distort the electors' wishes. 

For example, although the DLP did not win any of the six Senate seats, Mr. Little, DLP candidate, would have been elected if there had been only the usual five vacancies. 

It would be a major advance to abolish compulsory numbering of preferences for all candidates. This compulsion serves no good and produces a crop of evils. 

1. It causes excessive informal voting. Informality in Senate elections often runs to 10 per cent and sometimes higher. 

Ireland has used a similar PR system since 1923. But informality is around 1 per cent - mainly because an Irishman need mark only one preference. In Tasmania a system and ballot-paper largely similar to the Senate method are used for electing the State's House of Assembly. Voters have to mark a minimum of three preferences. Informality is gene rally around 4 per cent or 1ess. 

2. Compulsory numbering of all candidates is offensive psychologically. It compels the elector to express preferences even for candidates he may despise. 

To force an elector to vote for candidates he doesn't want in order to support those he favours is a deplorable infringement of voting freedom. When an elector can vote for as many or as few candidates as he pleases, voting becomes a meaningful selective process, not a mathematical exercise. 

In counting, most ballot-papers are never examined beyond the second preference, and fewer than 15 per cent are examined beyond the third preference in most 5-vacancy elections. Hence, fully marking a ballot-paper is largely futile. 

3. Compulsory numbering of all candidates means enforced cross-party voting and could produce results not desired by the electors. Most Tasmanians in their State elections vote for the candidates of their own party and then stop. 

So Independent and minority party candidates pick up very few preference from supporters of the major parties. If voters had had to number all candidates some Independents not able to reach quotas under existing conditions would have been elected. 

If supporters of minority party candidates are numerous enough to elect a Senator, he is justly entitled to the seat. 

Regimentation can operate against the interests of the party imposing it on its supporters. 

In the absence of regimentation, so long as the supporters of a party keep their votes within the party ticket, the party can expect to obtain maximum success. PROVIDED it does NOT advise its supporters to concentrate their No. 1 vote on a specified candidate and then number the preferences in a uniform order. 

Under PR elections a party's best interest is served by keeping as many as possible of its candidates in the count for as long as possible. 

This occurs as a result of the natural spreading of votes over several candidates as seen in State elections in Tasmania, where the party doesn't tell its supporters to vote 1-2-3 down the ticket. 

But if it does regiment a party wastes votes by electing its No. 1 and No. 2 candidates with full quotas early in the count. 

This leaves fewer votes for their No. 3, who could be forced into elimination. 

For example, in a 5-vacancy election with its larger quota, Senator Sandford, ALP, would have been lowest towards the end of the count and eliminated, thus electing Mr. Little. 

But if the main parties did not regiment their vote, but let it spread naturally their No. 3 candidate could have enough votes to stay ahead of Mr. Little, who consequently, would be eliminated, as can be seen by the figures applied to a 5-vacancy count. 

O'I'HER MAJOR REFORMS ARE STRONGLY NEEDED IN THE METHOD OF FILLING CASUAL VACANCIES AND IN CONDUCTING DOUBLE DISSOLUTION ELECTIONS.

He is an American student of Australian electoral methods, a former Fulbright Scholar working at present in the University of Tasmania. 

In the recent federal election there happened to be an extra Senate vacancy to be filled for Victoria. This meant that six places were to be filled, instead of the customary five. 

The upshot was that the ALP won three seats, the LCP the other three, and DLP candidate, Mr. Little, not running into a place. 

But had the situation been as usual, with only five candidates to be elected, Mr. Little, instead of running seventh and failing to get one of the six seats, could have run fifth, and won one of the five seats'. 

However, as Mr. Howatt shows in his second illustration, a different distribution of ALP first preferences - spreading them over the three candidates, instead of concentrating them on Labor's top candidate - could still have kept Mr. Little out. 

At first sight, it is hard to swallow the likelihood of a candidate's vote standing him in better stead when there are fewer seats to be won. But in Tasmania, electoral experts have special knowledge of the vagaries of the Hare-Clark system of Proportional Representation, as used in federal elections for the Senate since the Chifley Labor Government introduced the system in 1949. 

In the examples it is sufficient to illustrate by showing only the last stage of the count, when the percentages of votes received by the various parties after the distribution of preferences would be approximately as follows: ALP, 43.0 per cent; DLP, 12.9 per cent; and LCP, 44.1 per cent. In the example, these percentages are expressed as whole numbers to facilitate understanding. The five-seat quota of 16.7 per cent, becomes 167 votes. 

In Illustration A, therefore, the ALP with 430 votes (i.e. 43.0 per cent would fill two quotas of 167 votes (16.7 per cent) and have a remainder of 96 votes resting with its third candidate, Senator Sandford. The LCP with 441 votes (44.1 per cent) would fill two quotas and have a remainder of 107 votes to the credit of Senator Hannan. As Mr. Little of the DLP would have 129 votes (12.9 per cent) he would be ahead of the ALP candidate and win election, as shown below. 

Illustration A: Mr. Little is elected - because the major parties regiment their supporters into concentrating their primaries by voting 1, 2, 3, down their respective party tickets. Thus:- 

ALP,  430 votes                                                                            DLP,  129 votes                            LCP,  441 votes  
(167)    +     (167)     +       96                                                        129                                               (167)     +     (167)      +        107  
quota          quota               votes                                                   votes                                             quota            quota                votes 
for              for                  for                                                       for                                                for                 for                   for  
Kennelly     Hendrickson    Sandford                                              Little                                             Gorton           Wedgwood       Hannan 

As both Sandford, ALP, and Hannan, LCP, would have fewer votes than Little, DLP, either Sandford or Hannan (in this case, Sandford) would be eliminated, and Mr. Little would receive the preferences and be elected. On the other hand, if the ALP and LCP do not waste their vote by concentrating it on their No. 1 and No. 2 candidates, Mr. Little becomes lowest on the poll and is eliminated, his preferences electing Senator Hannan, LCP, on the basis of illustrative figures shown. 

ALP,  430 votes                                                                            DLP,  129 votes                                        LCP,  441 votes  
150      +     143        +        137                                                     129                                                           152     +     147         +         142  
votes           votes               votes                                                   votes                                                        votes         votes                  votes  
for              for                   for                                                      for                                                            for             for                     for  
Kennelly     Hendrickson     Sandford                                             Little                                                         Gorton       Wedgwood        Hannan 

As spreading the vote conserves the strength of the major parties, Mr. Little becomes lowest on the poll, resulting in his elimination and the transfer of his preferences, which elect Hannan, LCP. In the absence of regimentation of voters, as seen in State elections in Tasmania and other places using a system similar to the Senate PR. method, a natural spreading of the No. 1 vote and preferences takes place. In Illustration B the variations in the spreading could be considerably greater than shown without the lowest major party candidate falling below Mr. Little. 

In this election (Nov. 22, 1958) both the ALP and LCP placed the DLP next in preference after their own candidates. 

First, the compulsory marking of all preferences again resulted in excessively high informality, averaging 10.6 per cent nationally. Informality was greatest in NSW with a State average of 12.8 per cent. In Tasmania informal votes accounted for 10 per cent in contrast to an average of around 4 per cent under Hare-Clark voting in State elections, for which a ballot paper is valid if marked correctly for a minimum of only three choices. 

Second, compulsory numbering of all candidates is offensive psychologically to many electors, who resent having to give preferences to candidates they disfavour or despise. 

Third, compelling an elector to extend preferences to all candidates not only delays greatly the counting process but also can misrepresent the intention of electors. For examples when Mr. Devitt (the third ALP Senate candidate was excluded, all his 15,507 ballot papers had to show a preference for either Mr. Orchard (Lib.) or Dr. Turnbull (Ind.). 

Studies show that most electors in past Hare-Clark elections express choices for their party ticket and stop - an understandable course. But in the recent Senate contest supporters of Mr. Devitt were compelled to show preference for either their chief traditional opponents or for a former Labor Minister now competing against their party. Forcing electors to choose between candidates, none of whom they may like, gravely infringes voting freedom and potentially can alter the results intended by the electors. 

This potential misrepresentation becomes more serious when combined with the unfortunate practice by political parties of endeavouring to regiment their supporters into voting, in sheep-like fashion, 1-2-3 down the paper. Though practised universally in Mainland Senate elections, regimented voting has scarcely occurred in Tasmanian Hare-Clark contests. 

Under Hare-Clark voting, candidates therefore owe their election to the support they win from the electors, not to the order in which they are listed on the ballot-paper. 

Many evils come from regimented voting, but for the moment let us consider how it could, in fact, change what election results otherwise might be. The possibility of obtaining different election results from the same election figures arises in cases when no party (or independent candidate) has enough votes of its own to fill the last quota in a multi-member electorate. 

One of several such illustrations from the last Senate election occurred in Tasmania, where the fifth Senate seat (and 20 per cent of Tasmania's available Senate representation) was won by a candidate who polled only 11.6 per cent of the total primary vote. In this contest, after four seats were filled, Mr. Devitt ( ALP) had 15,507 votes to his credit, Mr. Orchard (Lib.) 16,326, and Dr. Turnbull (Ind.) 21,448. As Mr. Devitt was recorded as having the smallest vote, he was excluded and his papers divided between Orchard and Turnbull according to the preferences compulsorily indicated on them. 

Yet at the stage before Devitt's exclusion the Labor group had a total of 68,189 votes, the Liberals 69,608, and Turnbull 21,448. Suppose the party totals had been divided equally between the three candidates of each party then each Labor candidate would have had 22,929 votes and each Liberal candidate 23,202. Under these conditions, Dr. Turnbull, having only 21,448 votes would have been the lowest candidate and therefore eliminated, his preferences deciding the fifth seat. An exact spreading of the vote among the major party candidates would not have been needed in order to keep them ahead of Dr. Turnbull during the cut-up, provided the spread was sufficient to keep each of their candidates ahead of Turnbull's total. 

STRENGTH WASTED 

Under unregimented Hare-Clark voting the electors freely distribute both their primaries and preferences, over a wide choice of candidates. As a result, a natural spreading of primaries and preferences generally occurs. This also incidentally helps the major parties to retain their electoral strength during the cut-up, as can be seen from past Hare-Clark elections. 

In the recent Senate election, because of different voting and vote-counting conditions Dr. Turnbull was able to win a seat though obtaining only 69.4 per cent of a quota of first preferences (i.e. 11.6 per cent of the total vote). 

Under regimented voting, a party wastes its voting strength by electing its No. 1 and No. 2 candidates with full quotas early in the count. This leaves fewer votes for their No. 3 candidate, who therefore could be forced into elimination unnecessarily in instances when his party lacks three full- quotas. 

The basic principles of the Senate PR system are excellent and result in incomparably more representative Senates than the previous "one in, all in", system which not only denied minority representation but grossly distorted the representation of the major parties. Nevertheless some corrections in applying the basic principles are needed. 
 

He observed that Dr. Turnbull "made the grade into the Senate on his not very impressive primary poll" of 11.6 per cent, but that Senator McManus "knocking on the door of a theoretical quota" by obtaining 14.1 per cent of the primary vote in Victoria is defeated. 

In NSW, however, the D.L.P. with merely 7.9 per cent of the primary vote, almost won a seat, the correspondent reported. 

Two features of the Tasmanian Hare-Clark voting could remove these anomalies, as illustrated by the Senate election in Tasmania. 

First, under the Hare-Clark conditions the extension of preferences beyond a minimum of three is voluntary, not compulsory. 

The present requirement to mark all preferences in Senate elections can also misrepresent the intention of electors. 

OFFENSIVE 

For example, when Mr. Devitt (the third ALP candidate) was excluded, his supporters were compelled to show a preference either for their chief political rivals, the Liberals, or for a former Labor Minister now competing against their party. 

Forcing electors to choose between candidates, none of whom they may like, offends against voting freedom, and, in addition, can potentially alter the election results intended by the voters. 

This potential misrepresentation is sharply increased if the major parties try to regiment their supporters into voting sheep-like, 1-2-3 down the ballot-paper. 

This unfortunate practice in Senate elections is the rule on the mainland and is frequent in Tasmania. 

Since regimentation does not occur under Hare-Clark voting, the electors vote freely, picking and choosing candidates in the order they want. 

FAIRER 

This free selection by the voters in Hare-Clark elections results in a natural spreading of primaries and preferences over all of a party's candidates, as seen in past elections. 

Spreading the vote also incidentally helps the major parties to maintain their electoral strength. during the cut-up, thus giving a fairer reflection of public opinion and reducing the likelihood of the anomalies noted by the Canberra, correspondent. 

Let us consider how unregimented voting and the resultant spreading of the vote could produce different election results from the same election figures.

DIFFERENCE 

In this election, after four seats were filled, Mr. Devitt (ALP) had 15,507 votes to his credit, Mr. Orchard (Lib.) 16,326, and Dr. Turnbull (Ind.) 21,448. 

As Mr. Devitt was recorded as having the smallest vote, he was excluded, and his papers divided between Orchard and Turnbull, according to the preferences compulsorily indicated on them. 

Yet at the stage before Devitt's exclusion the Labor group had a total of 68,789 votes, the Liberals 69,608, and Turnbull 21,448. 

Suppose the party totals had been divided equally between the three candidates of each party -then each Labor candidate would have had 22,929 votes and each Liberal candidate 23,202. 

Under these conditions Dr. Turnbull, having only 21,448 would have been the lowest candidate and therefore eliminated, his preferences deciding the fifth seat. 

An exact spreading of the vote among the major party candidates would not have been needed to keep them ahead of Dr. Turnbull during the cut-up, provided the spread was sufficient to keep all other candidates ahead of Turnbull's total. 

Under regimented voting a party wastes its voting strength by electing its No. 1. and No. 2 candidates with full quotas early in the count. 

This leaves fewer votes for their No. 3 candidate, who therefore could be forced into elimination unnecessarily in cases when his party lacks three full quotas. 

An example illustrating the effect of spreading the vote can be seen in Denison in the State elections of 1928. Then Mr. Mahoney, Ind. Labor candidate, polled 75.3 per cent of a quota in primaries but was not elected because the votes for the major parties (40.9 and 40.4 per cent respectively) were widely spread and because Mr. Mahoney did not pick up many preferences. 

In the recent Senate election as a result of different voting and vote-counting conditions, Dr. Turnbull was able to win a seat through obtaining only 69.4 per cent of a quota of first preferences (11.6 per cent of the total vote). 

The basic principles of voting by proportional representation for the Senate are sound and desirable, and far superior to the previous "all or none" system of preferential, block voting. 

But if complaints about anomalies such as those cited by the Canberra correspondent are to be eliminated, some corrections in applying the basic principles are needed.

Comparing the Tasmanian model with other methods reveals its many superiorities. First, a uniquely wide choice of candidates: 

No other electoral system in the world provides the voter with such a broad selection of candidates. 

In the recent election each voter was offered eight or nine candidates to choose from in each major party, not to mention smaller parties or independents. 

This wide range thus gave the Tasmanian elector not only a choice of parties, but also a choice of candidates within parties. 

Unless given this selection within parties, voters may not be able to express their true judgment, as seen in the single-member electorates used generally in the other States, where normally only one candidate is allowed to stand for each competing party. 

There, in the absence of the many choices available to Tasmanians, the elector must accept, whether to his liking or not, this sole pre-selected choice of his party's organization, or else vote for a candidate of another party with whose policies he may not agree. 

With only one endorsement per electorate, a party can hardly expect to satisfy a wide range of opinion within it, especially where, in a two-party system, each party attempts to represent a broad spectrum of views. 

This invidious situation, inherent to single-member electorates, is further compounded if those empowered to pre-select for their party are not well representative of the whole of a party's supporters. 

By contrast, since multiple endorsements are essential to the Hare-Clark system, party pre-selection authorities have little excuse for not offering their supporters a balanced selection of candidates. 

The possibility under the Hare-Clark system for presenting balanced teams of candidates is likely to supply the electors with satisfying representation within a two-party framework. 

Nevertheless, for those voters who may want different representation, the Hare-Clark system provides ready opportunity for minority representation, by ensuring that any group of electors as large as a quota of votes, namely, a fraction as small as 12.5 per cent of the electorate, may win a seat. 

By contrast, the single-member system, by restricting the selection of candidates, creates the need for more satisfying representation yet prevents it from being achieved, since the system provides only one vacancy per electorate, not seven. 

Third, a strong Opposition: 

Election by quota guarantees that a party must receive one seat for each 12.5 per cent of the vote which it receives. 

Hence, a party polling as low as 37.5 per cent of the vote will obtain three seats in a seven-member electorate. 

In a House of 35 members an Opposition is therefore not likely to fall below about 15 seats in a two-party situation. 

By contrast, since the single-member system does not ensure any necessary relationship between votes received and seats won, an Opposition party may be reduced to far below its fair share of the seats on a proportionate basis, as seen at present in the Queensland Parliament and the Federal House of Representatives. 

Fourth, assurance of majority rule: 

The Hare-Clark system as now applied guarantees a close relationship between votes received and seats won, because members are elected on equal quotas from multi-member electorates. 

This high degree of accuracy is likely to result in a party with a majority at the polls receiving a majority in Parliament. 

This likelihood could be made a certainty by the adoption of a few refinements in the counting procedure. 

The single-member system, by contrast, may result in a party with a minority of votes receiving a majority of seats or in the governing party holding onto office by razor-thin majorities, as illustrated as present by the Parliaments of South Australia, New South Wales, and Great Britain. 

Fifth, other advantages: 

Too numerous to list here are a multitude of other advantages which the Hare-Clark system can provide, for instance, eliminating safe seats and uncontested elections and reducing greatly the dangers of gerrymandering.