PROPORTIONAL SYSTEM OF ELECTION IN TASMANIA (REPORT OF MR. MANSFIELD, CHIEF ELECTORAL OFFICER, ON THE).

Appendix to the Journals of the House of Representatives, 1913 Session I, H-39

H.—39

1913. NEW ZEALAND.

PROPORTIONAL SYSTEM OF ELECTION IN TASMANIA (REPORT OF MR. MANSFIELD, CHIEF ELECTORAL OFFICER, ON THE).

Return to an Order of the House of Representatives dated the 12th November, 1913. Ordered, " That there be laid before this House the report of Mr. Mansfield, Chief Electoral Officer, on the proportional system of election in Tasmania, together with the newspaper extracts attached thereto, as supplied to the Legislative Council."—(Mr. Russell.)

TASMANIAN SYSTEM OF PROPORTIONAL REPRESENTATION. The Hon. Minister in Charge of Electoral Department. I have the honour to report that in accordance with your instructions I visited Tasmania, arriving in Hobart on the evening of the elections, 23rd January. I was present that night at the count of the votes for Denison, and later witnessed the Franklin count. The Premier (Hon. A. E. Solomon), the Chief Electoral Officer (Mr. H. E. Packer), the Assistant Returning Officer (Mr. E. L. Piesse, LL.B., expert and mathematician), and other responsible officers kindly extended to me every courtesy, and provided all possible facilities for my obtaining the fullest information in regard to the proportional system. The process of the count under the Tasmanian system is complicated and tedious, but I followed it without much difficulty and obtained a good insight into the method, and brought back sufficient data to enable me to demonstrate the system here. Tasmania is divided into five electorates, each returning six members, and at the elections in April last the Government (Liberal) secured the return of 16 members and Labour 14. A deadlock in Parliament ensued, followed by a dissolution, when the country was appealed to again. The result of last month's general election, however, is, so far as the strength of the respective parties is concerned, precisely the same as at the previous election, the distribution being—Liberals, 16 members ; Labour, 14. In their official report on the general election of April, 1912, the Chief Electoral Officer, Assistant Returning Officer, and Clerk in charge of the Electoral Department pointed out that the smallness of the Liberal majority in the House " is mainly to be attributed to the fact that the districts each return an even and not an odd number of members," and that with seven-member constituencies the Liberals would probably have had 20 members to Labour's 15. As the representation in proportion to the actual strength of the parties should only have been Liberal 191 members, Labour 159, this result would come about by the larger party receiving the benefit of the fractions. By adopting seven-member constituencies, therefore, the larger party would probably have obtained a majority of 5 members, although only entitled, according to the actual strength of the parties, to a majority of 3-2 members. No alteration was made, however, in the number of members to be returned for each constituency, with the result (as already mentioned) that the parties were again returned with exactly the same respective strength as at the previous election. Judging from comments in the Press, and from the remarks of informed persons with whom I had conversation, there is a good deal of dissatisfaction with the system of elections as at present constituted, and it was indicated that an effort would probably be made to have the law amended. I am under the impression that some, while not wishing to see the fundamental principle of the proportional method interfered with, would welcome some modification of the present method by which the results of the poll are arrived at, while others would support a change back to single-member constituencies with preferential Voting so as to secure the return of a candidate by an absolute majority. I was greatly struck to find that many electors did not appear to understand the principles of their proportional representation system, although in operation there for the past three general elections. Even amongst those who possessed a general idea of the working of the system there was but a most hazy conception of the method by which the results of the election were arrived at; and this is not surprising—the process is bewildering to most people. That it works out with almost mathematical accuracy so far as the strength of the parties in the electorate is concerned is quite apparent, but the transfer of the remoter preferences is apt to produce most unexpected changes at the last moment. It may sometimes happen that preferences three or four times removed will finally decide the election of a candidate —>a complex phase of the system that is not fully realized by the average voter when marking his ballot-paper. The delay in arriving at the final results of the poll, which may extend over a period of a week, is a cause of much dissatisfaction.

I—H. 39.

H.—39

2

I had a long conversation with the Hon. Mr. Solomon, the Premier, who was of opinion that singlemember constituencies with the preferential vote would be found to give greater satisfaction to us here in New Zealand than the proportional system. Mr. Solomon admitted that from a theoretical point of view the Hare-Clark system was perfect, but from a politician's point of view it was not perfect, inasmuch as you could not in actual practice divide a candidate up fractionally in the way it was done on paper. ■ Even under the present electoral law in Tasmania it is not found practicable to carry out by-elections under their proportional system. The whole electorate is polled for the return of a member, and if there are more than two candidates standing the preferential method of voting is applied so as to ensure the return of a candidate with an absolute majority. I attach a table showing the result of a by-election for the district of Wiimot held in June, 1910, which the Premier, Mr. Solomon, handed to me to support his favouring single-member constituencies with the preferential vote as a system of election which would ensure the return of a member who represented a majority of the electors. I am constrained to say that, generally speaking, the objections to the Hare-Clark system which I noted in Tasmania were not against proportional representation, but rather against the complex method by which the results of the election were arrived at—brought r bout by the use of the transferable Vote. It appeared to me that the complicated process of the transference of surpluses was regarded more as a necessary evil than as a satisfactory method of arriving at results, the average elector regarding the whole procedure as quite beyond his mental grasp. The question of substituting the party-list system of proportional representation for the transferable vote is engaging serious attention in Tasmania. This system has much to recommend it. On the ballot-paper the candidates are arranged in lists according to party, and the elector votes for as many candidates as there are seats to be filled-—-not by numbering the candidates in the order of his preference, but by giving a vote to each of them. Instead of having to decide upon his various preferences the elector may vote straight out for his party by giving a vote to each candidate on the list (up to the number of candidates to be elected), just as he may do at an ordinary municipal election. He is not, however, prevented from voting for any candidates irrespective of parties. There being no " preferences " to deal with, the duty of the Returning Officer in ascertaining the result of the poll is exceedingly simple. He ascertains the number of votes secured by each candidate, as well as the total of all the votes obtained by all the candidates in each party, and apportions the number of seats to each party accordingly, selecting the particular candidates who are highest in the list. In the event of a seat becoming Vacant through the death or resignation of a member the next highest candidate in the party's list at the general election is selected. Thus the proportional strengths of the parties in the House remain undisturbed, and the trouble and expense of a by-election are obviated. It will be seen that the party-list system subordinates the candidate to the party, while the use of the transferable vote subordinates the party to the candidate. The use of the party-list does away entirely with the complicated and tedious method of arriving at results such as we have with the transferable vote, while it is claimed for it that it returns each party in exact proportion to their respective strengths. The objections to the party-list system do not appear to me to be very serious, and if it is intended to adopt proportional representation in some form or other this system should be given careful consideration. Mr. Piesse, LL.B., who is Assistant Returning Officer in Hobart, and an acknowledged authority ■on the subject of proportional representation, has recently contributed to the Press of Tasmania articles on the party-list system. I attach reprints of same for your perusal. Regarding the application of proportional representation to New Zealand there would be no difficulty so far as the parliamentary poll was concerned, but I see very great difficulties in the way of carrying out a poll under the proportional system (which requires large constituencies).simultaneously with the licensing and national prohibition polls. The whole position would become exceedingly complicated, and even if the. process could be made at all workable the risks of irregularities in connection with the conduct of the polls would be very great. In a memorandum dated the sth December last, which is still with you, I outlined some of the difficulties in the way of carrying out a parliamentary poll under the proportional method simultaneously with the licensing polls. If it is decided to retain single-member constituencies in part or throughout the whole of the Dominion, I think that the use of the preferential vote (where there are more than two candidates) would prove much more satisfactory than the second ballot. The method provides, in reality, for first and second ballots being conducted at one operation, and its adoption would not involve any disturbance in present electoral arrangements. Three Australian States —-viz., Queensland, Victoria, and Western Australia—have adopted preferential voting with single-member constituencies, and the system has, so far as I can learn, given satisfaction. In New South Wales the second ballot is in vogue, while in South Australia they have the old relative-majority system such as we had prior to , the adoption of the second ballot in 1908. I attach the following : Clippings from Hobart Press in connection with the elections, containing comments upon the Hare-Clark system ; table showing the results of the counts of first choices at the recent Tasmanian elections, and final results ; table showing result of by-election for the district of Wiimot, held in June, 1910 (mentioned above) ; and printed pamphlets on party-list system of proportional representation (mentioned above). F. W. Mansfield, 19th February, 1913. Chief Electoral Officer.

3

H.—39

Party Government and the Voting System. [Mercury (Hobart), 27th January, 1913.] The Chief Electoral Officer for New Zealand, who is studying here the working of the proportional system of representation, has been struck by the fact that a great many of our electors do not seem to understand the principle of the system, and do not realize that they have but one transferable vote. We may add that even some of those who do realize this find the system a little too much for them. It is for most people bewildering alike in its results and in the method of arriving at them. Matters are in suspense for a tediously long time, and the transfer of the " remoter preferences " is apt to produce such startling changes at the last moment that the electors may be pardoned for " ha'ing their doots "as to the process they have to take to so large an extent on trust. No doubt the result is beautifully perfect as a mathematical picture of the electors' minds, so far as they used their minds and so far as they represented the total electorate, but we are inclined to think that the admitted uncertainty of the system, by which no man can feel that his vote will go to elect the particular candidate he prefers and not his choice four or five times removed, has a good deal to do with the large number of abstentions. Another objection to the system, as made apparent by experience, is its tendency to make candidates of the same party display too much of the spirit of Hal o' the Wynd, who in the old Border battles " fought for his am hand." And since the question is on the anvil again, and Tasmania is, as it were, performing a test experiment for the benefit of many other States throughout the Empire, we may say that the chief point in which the proportional system seems to us open to serious criticism is in its association with party government. Party government may or may not be the.best conceivable system for managing the affairs of a country, but no one will deny that it has stood the test of time, that it is essentially the British way, and that for the present it is what the countries desire rather than to elect an executive Administration without party considerations. But if there is one thing certain about the proportional-Vote system it is that its logical outcome is the appointment of an Administration direct, and not government byparty. Thus we might have Mr. Solomon as Premier and Mr. Earle as Treasurer, and all manner of arrangements. The country does not want such a method of government, but wants representative democratic government on party lines, with the majority in a position to get a fair chance of good work, and the minority in a position to contribute to progress without obstructing it. For practical purposes, in a world where we cannot have mathematical or any other perfection, there is still much to be said for the single-member constituency, wliich has been practised in England from the very earliest days of representative government, when the " knight of the shire " went up to Parliament to sit with the barons and the bishops, the " lords temporal and spiritual." The single-m,ember constituency returns a man representing the majority of the constituency and familiar with its needs and wishes, while himself known to the constituents. Members may not be the best men in the land, but they are the best men for local needs, and each more closely reflects his own constituency's particular idiosyncrasies than any other man can. The defect of the single-member system is that minorities are not represented by it unless a party in a minority in some constituencies is in a majority elsewhere. Every one admits that strong minorities are entitled to representation in the national meeting-house. That is the reason, therefore, of the invention of the proportional system, with its essential feature of the " quota," so that any candidate securing the " quota " of votes is elected, no matter what the Voting strength of other parties. Now, what we wish to inquire, apropos of our own electoral system, is whether it may not be possible to modify the proportional method of voting so as to render it at once simpler and better from the point of view of efficient party government ? The valuable report on last year's election, prepared by Messrs. Packer (Chief Electoral Officer), Piesse, and Daly, pointed out that the smallness of the Liberal majority in the House " is mainly to be attributed to the fact that the districts each return an even and not an odd number of members," and that with seven-member districts the Liberals would have had probably 20 to Labour's 15. As the proportion, however, was strictly 19-1 to 15-9, this result is only arrived at by giving the Liberals the benefit of the fractions—-a method perfectly justified from the point of view of securing the efficiency of the parliamentary machine. Since it is so necessary to ensure that a Government shall have a working majority, and that a House shall not be in the comic-opera predicament of Bombardos and Pataquez—liable to be pulled from one side to the other every five minutes —-we think it might be worth while discussing not only the change from six-member to seven-member constituencies, but a further modification. , Why not at once recognize parties in an election ? At present parties are only recognized officially after an election, when the head of the State calls upon this member or that to form a Government, and necessarily must be guided by the strength of the respective " parties " as revealed by the pollingresults. What we suggest is that each candidate nominated should specify—-and that it should appear on the ballot-paper—his party. It might be Liberal, Labour, Independent, Conservative, Socialist, or what not, but it would be his official designation, and the Returning Officer would take official cognizance of it. Then, supposing the papers are numbered, just as now, in the order of the voters' preferences, the first thing to do at the close of the poll would be to count up, as now, the total of the valid papers to show the number of voters, and next to ascertain the quota. Then let the No. 1 votes be treated as indicating the electors' party preferences, and so ascertain the number of votes cast for each party. If any party has not received the quota to carry a seat no seat is allotted to it. The seats are then allotted to the parties in proportion to the No. 1 votes —Liberals so-many seats,. Labour so-many, or what-not. If there is a fraction, so that the last seat is compounded of two or three or more parties, let it go to the party having the largest fraction, or if thought better to the party having the majority of returned, members, in order to strengthen its working majority. Having thus arrived at the number of seats gained by the respective parties all that is required is to allot the

ST.—39

4

seats to the candidates as indicated by the voters' preferences, reckoning only the preferences of those voters belonging to the party concerned. There need be no " cutting up," or transference of votes, or elimination of candidates, all of which the plain man finds unsatisfying devices. Let us say the Liberals are entitled to five seats, then those five Liberal candidates which have the largest number of votes are elected, the " quota " for this purpose being neglected. Thus, it seems to us, parties would be represented according to the number of their respective adherents, and individuals returned according to the electors' preferences. The Hare-Clark System : Comments by Liberal League Secretary. Launceston, 24th January. The organizing secretary of the Liberal League (Mr. A. T. Marshall) made a few remarks this evening concerning the election. He said, "As far as lam concerned lam well satisfied with the result. Of course, the figures are not entirely conclusive, and if they turn out as it appears they will the decision will be 16 straight-out Liberals and 14 straight-out Labourites. The fact of getting rid of the Wilmot obstructionist is ample justification for the appeal to the country. Naturally one would have welcomed an extra gain to provide for a comfortable working majority, but in view of the sanguine hopes of prominent Labourites that Labour would probably gain a seat in Bass, Denison, and Darwin, we should feel satisfied at the manner in which their prophecy has ended. If our Labour friends are the democrats that they profess to be they should recognize the fact that a majority of the people are in favour of Liberal rule, and instead of putting all their energy into the unsatisfactory work of obstruction should give what ability they have in assisting the work of developing Tasmania." Mr. Marshall added that he regarded the Darwin figures with a good deal of satisfaction. They had a minority at the last election of 700-odd votes, and this had been reduced by half. When the King Island returns were in he thought the minority would be something less than 200. Asked if he considered an alteration of the Electoral Act necessary to ensure a working majority for either party, Mr. Marshall said it would be almost impossible for a long time to come for either party to obtain a satisfactory working majority under the present electoral system. It was unsatisfactory, for either one or the other might obtain a large majority of primary votes in an electorate and yet have equal representation. He had long held the opinion that Tasmania would be well advised to adopt the Victorian system or something approaching it, with single-seat contests and preferential voting. [Prom " Passings Notes," by " Jacques," in Mercury (Hobart), 25th January, 1913.] Is this the final trial of the Hare-Clark system ? That is a question which people are asking all over the State, and I believe that a great majority take the same view as I do—that having been given a very fair trial it must be written down as sadly wanting. lam prepared to stand up on a public platform and to prove to any audience of intelligent people that the theory is absolutely without flaw. Equally, and with much better heart, I am prepared to show that in practice it is as bad as bad can be. It does not give good results to either side, and that ought to be sufficient. It does not represent the opinions of the electors, because in an electorate one party may have a majority of 2,000 votes and yet equal numbers be returned. It destroys party cohesion. It practically compels each candidate to be entirely selfish, and to do everything he can to secure No. 1 votes for himself, even at the expense of his colleagues and his party. And, worst of all, people do not understand what they are doing. If adult suffrage is to be the rule things ought to be so arranged that the ordinary person, even though he may not be highly educated, can know what his vote means. Let us come back to the single-electorate system is my advice, which is worth noting. lam a plain man, and I like to have a ballot-paper containing the names of Smith and Jones only. Then I know that if I put a cross opposite the name of Jones lam voting for him. With the Hare-Clark system lam blessed if one voter out of five knows what the effect of his vote will be. The Election Returns. [Mercury (Hobart), 27th January, 1913.] The counting of the votes under the Hare-Clark system involves so much delay in arriving at a definite result that elections lose much of their dramatic effect. The election was held on Thursday, and the result of the Denison poll was made known on Friday afternoon, that of Bass on Saturday morning, and that of Wilmot on Saturday evening. The results in Darwin and Franklin divisions will not be known until to-morrow. In the first the probabilities favour an equal division, though there may be a change in the personnel of the Labour representation. There is doubt about what is likely to happen in Franklin. Mr. Hean may yet overtake Mr. Ewing, and cross-voting may enable the Liberals to make a bid for a fourth seat. If they do not win it, it will be a close finish. Hare System and Women. [Daily Post (Hobart), 27th January, 1913.] The Hare system of voting is, to the feminine mind, a complex problem. In quite a number of instances at the recent election votes were cast by women, who came to the booth to vote Labour, for Liberal candidates. The preferences provided for in the system puzzled numbers of the fair sex, some of whom, in the Denison Electorate;, in sheer despair gave their primary votes to the first name on the paper and wrote the other numerals down to seven in sequence. There were at least fifty women in the Denison Electorate who went to the polling-booth and, though qualified to vote, returned

5

H.—39

to their homes without recording their votes. The ballot-paper was as Greek to them, and they were afraid to go through the ordeal of marking it. The Liberals made good use of the weakness of the women, with the result that scores of Labour supporters were induced to sign postal votes marked so as to add to the strength of the Liberal poll. The Final Election Count. [Mercury (Hobart), Ist February, 1913.] At long last the full results of the general election are known. It was, however, recognized a week ago that the two parties would number 16 and 14 respectively, the only doubt being as to whether Mr. Watkins or Mr. Hurst would gain the third Labour seat in Darwin. The former just managed in the last lap to get ahead of his competitor, and so there is no change in the representation of that electorate. The reduction of the Labour majority in Darwin is a good omen for the Federal election which is approaching, and if the Liberals work actively there is no reason why Mr. O'Malley should not give place to Mr. Marshall. So far as the State Parliament is concerned there is no reason to suppose that there will be any trouble in getting through the next three years. The Government has a clear majority, sufficient to carry its measures, all the Liberal members being dependable and solid. We"can still express the hope that the Labour party will realize that it has a duty to the State beyond that of trying to upset the Ministry, and that it will give reasonable assistance to the Government in carrying measures designed for the general welfare. The Bass Count —Principles and Method : A Simple Explanation. [By E. L. Trams, in Mercury (Hobart), Ist February, 1913.] Only a few months have passed since the general election, when I gave an explanation of the counting under the Hare-Clark system. Perhaps a few notes on the contest just passed will be of interest to those who wish to study the system. Out of a total of 20,329 names on the Bass roll 14,132 persons recorded their votes, being 1,001 short of the number who voted in April, 1912. Out of this total (14,132) 351 votes were declared informal, the usual mistakes being made by the electors—viz., putting the cross against the name, as per the Commonwealth system, or striking out the candidate's names, which is done at our municipal elections, while others marked two l's or two 2's, &c. This left 13,771 valid ballotpapers to be dealt with. First the quota had to be found, which is done by dividing 13,771 by 7 (one more than the number of candidates required to be elected), and to the quotient adding 1, therefore the quota was 1,968. The following was the result of the first count: Anderson, 1,974 ; Becker, 2,039 ; Guy, 1,131 ; Hayes, 1,643 ; Howroyd, 1,788; McKenzie, 1,157 ; Sadler, 1,130 ; Solomon, 2,909. It will be seen by these figures that there were three candidates who had reached the quota—viz., Solomon (with a surplus of 941), Becker (surplus 71), and Anderson (surplus 6). Solomon's surplus, being the largest, was first dealt with by re-sorting the whole of the papers he received in the first count according to the No. 2 choices, except when the 2 appeared against the names of Becker or Anderson, when the next available choice was taken. The totals thus ascertained were reduced to valid votes by multiplying them by the transfer value— i.e., the fraction formed by dividing the surplus (941) by the total number of papers (2,909), or a little more than one-third. Solomon's papers were distributed as follows : Guy, 9 papers or 2 effective votes, making his total 1,133 ; Hayes, 659 papers, or 213 effective votes, making his total 1,856 ; Howroyd, 18 papers, or 5 effective votes, making his total 1,793 ; McKenzie, 742 papers, or 240 effective votes, making his total 1,397 ; Sadler, 1,481 papers, or 479 effective votes, making his total 1,609. Fractional remainders are disregarded. Next Becker's surplus had to be distributed in the same manner. The transfer value was 71— 2,039, or about one twenty-eighth, and the re-sorting of his papers gave the following result : Guy, 1363 papers, or 47 effective votes, total 1,180 ; Hayes, 24 papers (of no value), total 1,856 ; Howroyd, 631 papers, or 21 effective votes, total 1,814 ; McKenzie, 11 papers (of no value), total 1,397 ; Sadler, 9 papers (of no value), total 1,609. In the same way Anderson's small surplus of 6 was distributed. The transfer value was 6 over 1,974, so that it required over 300 papers to make one effective vote. The result was—Guy, 1,429 papers, or 4 effective votes, total 1,184 ; Hayes, 22 papers (of no value), total 1,856 ; Howroyd, 518 papers, or 1 effective vote, total 1,815 ; McKenzie, 2 papers (of no value), total 1,397 ; Sadler, 3 papers (of no value), total 1,609. All the surpluses having been distributed, Guy (1,184), being the lowest on the list, was struck out, and his papers distributed. First the papers (1,131) he received at the first count were distributed according to the second choices at the transfer value of 1, since the first choices have been rendered valueless, and the result was —Hayes, 47, making his total at count s—-1,9035 —-1,903 ; Howroyd, 1,044, making his total at count 5—2,859 (elected) : McKenzie, 10, making his total at count 5—1,407 ; Sadler, 16, making his total at count 5—1,625; exhausted, 14. "Exhausted" means that there was no available number against an unelected candidate on the paper. For example, if the 1 was against Guy, 2 against Anderson, and 3 against Becker, and there were no more numbers, the papers became exhausted, thus illustrating the necessity for electors marking preferences for at least the full number of their party's candidates. Next the 9 papers (or 2 effective votes) Guy received from Solomon were distributed, the only candidate to receive enough papers to secure one effective vote being Sadler, making his total 1,626. The 1,363 papers, or 47 effective votes, Guy received at count 3 from Becker's surplus were distributed in the same manner, and again only one candidate received enough papers to secure an effective vote —viz., Hayes, 72 papers, or 2 effective votes, making his total 1,905.

H.—39

6

In the same manner the 1,429 papers, or 4 effective votes, Guy received from Anderson Were re-sorted, but no one secured enough papers to equal the Value of a vote. When the papers Guy received at count 1 were distributed Howroyd received enough to secure election, with a surplus of 891 above the quota. Before another candidate could be struck out this surplus had to be distributed at the transfer value of 891—-1,044. Hayes received 16 papers, or 13 effective votes, making his total at count 9—1,918 ; McKenzie received 8 papers, or 6 effective votes, making his total at count 9—-1,413 ; Sadler received 38 papers, or 32 effective votes, making his total at count 9—1,658 ; exhausted, 982. Now came the final action in this count. That was to exclude McKenzie, and declare the remaining candidates (Hayes and Sadler) elected. General Election, Tasmania, January, 1913 : Table showing the Results of the First Counts (First Preferences) in the various Electorates, and the Candidates finally ELECTED. Denison. (Quota, 2,122.) *Lewis (Liberal) .. .. 2,291 Valentine (Liberal) .. .. 1,011 *Davies (Liberal) .. .. 1,902 Edmonds (Labour) .. .. 764 *Giblin (Labour) . . . . 1,759 Rattle (Liberal) .. . . 620 * Woods (Labour) .. .. 1,393 Waterworth (Labour). .. .. 502 *Fullerton (Liberal) .. .. 1,321 O'Brien'(Labour) .. .. 392 ♦Barker (Labour) .. .. 1,294 Bottrill (Liberal) .. .. 369 Sheridan (Labour) .. .. 1,027 Paterson (Liberal) .. .. 202 Wilmot. (Quota, 1,756.) ♦Lyons (Liberal) .. .. 2,198 *Best (Liberal) .. .. 1,163 ♦Mulcahy (Liberal) .. .. 1,961 ♦O'Keefe (Labour) .. .. 1,008 *Hays (Liberal) .. .. 1,506 Cameron (Independent) .. 977 *Lee (Liberal) ... .. 1,456 Walduck .. .. 775 Shoobridge (Labour) . . .. 1,245 Bass. (Quota, 1,967.) ♦Solomon (Liberal) .. .. 2,908 ♦Hayes (Liberal) .. .. 1,643 ♦Becker (Labour) .. .. 2,039 McKenzie (Liberal) .. .. 1,157 ♦Anderson (Labour) .. .. 1,974 Guy (Labour) .. .. 1,131 ♦Howroyd (Labour) .. .. 1,788 ♦Sadler (Liberal) .. .. 1,130 Franklin. (Quota, 2,179.) ♦Earle (Labour) .. .. 3,240 Hean (Liberal) .. .. 1,423 ♦Cotton (Liberal) .. .. 2,881 Banks (Labour) .. .. 952 ♦Evans (Liberal) .. .. 2,019 ♦Dicker (Labour) .. .. 929 *Ewing (Liberal) .. .. 1,745 McDermott (Liberal) .. .. 484 ♦Martin (Labour) .. .. 1,580 Darwin. (Quota, 1,803.) ♦Ogden (Labour) .. .. 1,751 Hurst (Labour) .. .. 1,372 *Belton (Labour) .. .. 1,736 ♦Fallen (Liberal) .. .. 1,341 ♦Whitsitt (Liberal) .. .. 1,528 Laugh ton (Liberal) .. .. 1,079 ♦Payne (Liberal) .. .. 1,484 Lamerton (Liberal) .. .. 742 ♦Watkins (Labour) .. .. 1,426 Jones (Labour) .. .. 156 * Finally eleoted.

Electoral District of Wilmot: Election of One Member of the House of Assembly held on the 8th June, 1910. Abstract of Result of Poll.

Subdistricts. Kean (Labour). Mulcahy (Liberal). Walduck (Liberal). Informal. Beaconsfield Deloraine Devonport Bvandale Kentish Latrobe Longford Westbuiy Postal votes Totals first choices • ■ I 351 410 420 263 301 265 500 296 8 2,814 63 276 388 174 467 317 222 175 57 2,139 511 200 228 184 168 111 253 249 27 28 22 21 20 19 20 42 16 13 2,139 1,931 201

7

H.—39

Walduck struck out. Second count and distribution of Walduck's No. ] (1,931): — Kcan. Mulcahy. First count .. .. .. .. .. .. 2,814 2,139 No. 2 on Walduck's papers .. .. .. .. 172 1,759 Totals .. .. .. .. .. 2,986 3,898 (Informal papers, 201.) Majority in favour of Mulcahy .. .. .. .. .. 912 Proportional "Representation : The French Party-list System Applied to Tasmania. [By E. L. Piessb, B.Sc, LL.B., in the Mercury (Hobart), sth February, 1913.] In a leading article a few days ago the Mercury said, " Why not at once recognize parties in an election ? " and then gave the outline of a simple form of the party-list method of securing proportional representation. The method suggested is very similar to the system adopted last July by the French Chamber of Deputies, and a few remarks on this system and on the possibility of applying it in Tasmania may be of interest during the present discussion of electoral methods. Proportional representation is obtained in Tasmania, by using a single transferable vote. There are only a few countries which use the transferable vote—Denmark and South Africa are the best known —and, although it is proposed for the Irish Senate and for other elections in Great Britain, in most countries proportional representation is based on the party-list. Party-list systems subordinate the candidate to the party, as the report of the British Royal Commission on Electoral Systems points out, and. have for their primary object the return by each party of the proportion of members to which it is entitled, while the transferable vote subordinates the party to the persons, and in its origin is not a system of proportional representation at all. Party-list Systems. In 1899 Belgium adopted a party-list system. The example of Belgium produced a movement in its neighbour, France, in favour of proportional representation, or " la R.P.," as French writers usually call it for short. A permanent Committee of the French Chamber, the Commission dv Suffrage Universel, was appointed, and all projects for electoral reform were referred to it for report. In France, as in Tasmania, party advantage as well as political principle enters into the discussion of electoral questions, but nevertheless the Commission has produced a series of white-books in which the arguments for and against proportional representation, as well as the various systems which have been proposed, are discussed with admirable fairness and clearness. If any further discussion of electoral methods is necessary in Tasmania these white-books will provide a mine of information. The Secretary of the Belgian Senate, in his evidence before the British Royal Commission, said that there were over three hundred varieties of proportional representation. Most of these are partylist methods, and among them will be found one to suit any taste—the Belgian, in which an elector must vote only for one list; the French, in which he may vote for candidates from several lists ; the Finnish, in which the transferable vote is combined with the party-list; the system of Mr. Barford, of Melbourne, in which the mark system and the transferable vote are both combined with the partylist —-systems which give a result within an hour or two of the close of the poll, and systems which take several weeks. Of all these we choose the French for explanation because it is the simplest, because it has been adopted after discussion of all previous ones, and because it seems suitable for Tasmania. The French System. In France each candidate names his party when he nominates, and on the ballot-paper the candidates of each party appear on a list by themselves. It is the candidates who decide the names on the party-list, be it observed ; the Returning Officer has no dealings with league or caucus or boss. No candidate has to obtain the sanction of any party organization before he nominates ; but, of course, he must have the consent of other candidates before his name can be placed in the same list with theirs. Each elector has as many Votes as there are seats —not preferential votes, but equal votes, as in the Federal Senate election. With the present Tasmanian districts each elector would have six votes, and he could give these to any six candidates, either all of the same list or some of one list and some of others. An elector has not to decide the order of his preference, and a candidate has not to bid for first choices at the expense of his friends. The method of voting, then, is the ordinary " block " vote ; the only change is that candidates are arranged according to their parties. Apportionment of Seats. So far for the elector's part; nothing could well be simpler. The Returning Officer's part also is simple ; he ascertains (as is now done for the Federal Senate) the votes obtained by each candidate, and he then finds the total of the votes obtained by all the candidates in each party. Let us suppose that the following is the result: —■ LIBERAL. LABOUR. A 1,800 T 1.700 B .. .. .. 1,600 U .. .. .. 1.300 C .. .. .. 1,500 V .. .. .. 600 D 1,500 W 300 E .. .. .. 1,300 X .. .." .. 200 F .. .. .. 1,000 V .. .. .. 200 G 300 Z 200 Total .. .. 9,000 Total ~ .. 4,500

H.—39

8

The Liberal list has two-thirds of the votes, the Labour list one-third. The Liberal party is thus entitled to two-thirds of the seats or four seats, and the Labour party to one-third or two. The next point is to decide which four of the Liberal candidates are to be returned. In the French system there is no surplus or transfer or exclusion ; all that is done is to take the highest candidates on the list. Thus the Liberal members would be A, B, C, D, and the Labour members would be T, U. If two candidates of a list are equal the Returning Officer can decide ; or, as in Prance, the seat goes to the older of the two candidates, or perhaps in Tasmania we might prefer that it should go to the younger. This is the French system, and it secures, in the words of the Mercury leader, that " parties would be represented according to the number of their respective adherents, and individuals returned according to the electors' preferences." Moreover, the proverbial schoolboy should be able to understand it, and it could not be said, as has been said in England of the transferable vote, that a man has to defend himself not against his enemies only but his friends. There remain one or two details and possible objections to consider. Preference to Majority. Seats cannot usually be allotted among the parties quite as simply as in the example chosen, for there are usually fractions or remainders to be considered. The question of remainders has been treated adequately by the French mathematicians and need not detain us now, but one rule of the French system deserves notice —namely, that when a party having a majority of votes has not as yet secured a majority of seats preference is to be given to it in allocating the remaining seats. Thus, at the recent election in Franklin, the Liberals, in proportion to their No. 1 choices, were entitled to 34 seats, and the Labour party to 2-6. Under the French system the Returning Officer would first give three seats to the Liberals and two to Labour, leaving one unallotted. Using the rule just referred to the unallotted seat would go to the Liberals (because, although they had a majority of votes, they had not yet obtained a majority of seats), and so the representation would have been four to two. Grouping of Lists. Provision is made in France for a small party or section of a party, which has sympathies with another list, to have a list of its own, and yet in the final allocation of seats to remainders to be grouped with the other party. This is called " apparentement "or grouping of lists. Let us take the following case, which may not improbably occur at the next Senate election ; the figures used arc based on the general election just held. Suppose there are three Liberal, one Independent Liberal, and three Labour candidates, and that they obtain Votes as follows : —■ LIBERAL. INDEPENDENT LIBERAL. LABOUR. A .. .. 31,000 D .. .. 19,000 X .. .. 32,000 B .. .. 31,000 V .. .. 32,000 C .. .. 31,000 Z .. .. 32,000 Total .. 93,000 Total .. 19.000 Total .. 96,000 With the " block " vote now used all the seats would be won by the Labour party, although in proportion to its strength this party would be entitled to only one out of three seats ; but if the party-list system were used the Returning Officer would ascertain in the first instance the shares of representation of the three lists in proportion to their totals. These shares are—Liberal, 1-3 members ; Independent Liberal, 0-3 ; Labour, 14. The Liberals, then, would be entitled to one member and the Labour party to one, leaving one still unallotted. Without provision for grouping of lists the remaining seat would go to the Labour party, since it has the largest remainder, and so Labour would obtain two seats, although in a minority. But with " apparentement " the remainder of the Independent Liberal, who has failed to secure a seat, is added to the Liberal remainder, making a total greater than the Labour remainder, and so the Liberal party gets the second seat, as is fair. The party-list system would thus enable the Independent Liberal to stand without danger of wrecking his party. By-elections. The party-list system can be used to provide in advance for vacancies. Whenever a vacancy occurs it can be filled by the next available candidate on the same list, and so the representation of the parties is not disturbed, as it is in our present system when there is a vacancy in the smaller party, and the seat is won at the by-election by the larger party. Objections. We mupt now mention some of the objections frequently taken to the party-list. A vote given to any candidate is not only a vote for him, but is also a vote for his party, since it is counted in the total of the party. This is what is meant when the party-list method is said to subordinate the candidate to the party. Now, the objection to this is that an elector may vote for a particular candidate who is not returned, and yet his vote may aid in the return of an obnoxious candidate of the same party. A similar objection is often taken erroneously to the transferable-vote system, but such an occurrence is impossible with that system unless the voter has given a later preference to the obnoxious candidate, With the party-list system the objection is valid.

9

H.—39

Cross-voting. The second objection relates to what is called in France " panachage," or cross-voting as we call it. In a strict party contest the great majority of the electors will vote for their party ticket. Thus, in the example of the Senate election, it will no doubt be found (apart from the disturbing influence of the Independent Liberal) that the three candidates of the two parties will obtain their votes mainly from strict party supporters, and so will be almost equal. Suppose, now, that in one of the parties there is a candidate who is conspicuously weaker than his fellows, and whose return in preference to his fellows would be an embarrassment to his party. Here, then, is an opportunity for the enemy ; let a few of them give votes to this weak candidate, and his return will be secured at the expense of his stronger fellows. That is what is called " panachage," intentional cross-voting, with a view to securing the return of the worst candidates of the other party. " Panachage "is much discussed by the French writers on proportional representation, but the best view seems to be that taken in one of the reports of the French Commission—'that " panachage " is unlikely, not only because it involves loss to one's own party, but because it may help the return of the best and not the worst candidates of the other party, and that it would never occur unless where parties were exceptionally well organized and party strengths known with unusual accuracy. The similar objection that the party-list puts it in the power of a small section within a party to determine, by refraining from voting for certain candidates, who shall be the members for the party seems unlikely for much the same reasons. New Parties. An important matter is whether the system allows of the rise of a new and, at first, small party. . It does, but in the form described it would be necessary for a small party which wished for no alliance to nominate as many candidates as seats. This objection can be overcome by allowing several votes to be given to one candidate, but this might reintroduce the struggle between candidates for the extra votes. Conclusion. In conclusion the present writer does not wish to express any opinion here on the validity of the objections taken to our present system, but he hopes that those supporters of proportional representation who are embarrassed by the defects of the present system will realize that there are other methods which avoid many of the difficulties of the transferable vote. The party-list system will not in itself produce a much larger majority with so small a House and so evenly divided a constituency. With thirty members no proportional system, even if the districts are split, can be relied on to give a decisive majority to a party which has only 53 per cent, of the voters, although the adoption of the French rule favouring the larger party would help to do so. If a majority of, say, four or five is required, not only should the number of members to a district be made odd, but the number of members in the House must be increased. To increase the number of members of Parliament might have other disadvantages besides that of increased expense, but. it would be a lesser evil than the single-member system, which might put a minority in power. Even with the Legislative Council method for securing that the member is returned by an absolute majority in the single-member district, and even in a contest between only two parties, it may easily happen that a party which is in a minority throughout the country obtains a majority of seats. Proportional representation, whatever its disadvantages, will at all events prevent a minority from ruling the majority. Proportional Representation : Party-list System. [By " Scrutator," in the Daily Post (Hobart), sth February, 1913.] The Hare system does not seem to be growing in favour. "It is too complicated," say some, and yet there are no more informal Votes than with simpler methods of voting, and very few electors lose votes for their parties by cross-voting or short-voting. "We do not understand what is done with our votes," others say, and yet it should be plain enough that a vote can never go to a candidate unless a preference has been given to him. Others dislike the scramble for No. l's. "A candidate has as much to fear from his friends as his enemies," they say, and an election causes ill-feeling among candidates of the same party. Others, again, do not understand that their second and third choices may never be used, and that the only sure way to get a candidate in is to give him their No. 1. To this misunderstanding, probably, Mr. Sheridan owes his defeat. " Why all these complications of quota and surplus and transfer and exclusion ? " others say. " Why cannot we have a system which any plain man can follow ? We can have such a system if we want it. Many countries have proportional representation now, but it is only in the British Empire and in Denmark that we find the Hare system. The object of the Hare system is that Parliament shall be a sort of map, on a small scale, of opinion throughout the country—a map in which each considerable body of opinion is represented, and represented to scale. Its object is to represent parties in proportion to their strength ; but in its working it ignores parties altogether, and compels voters, instead of Voting for parties, to arrange individuals in an order of preference. Other European countries take a more direct route to their object. "If we are going to provide for representing parties," they say in effect, " let us commence by recognizing parties. Let each candidate name his party when he nominates. Let the names of all candidates of the same party be arranged in a list by themselves, the ' party-list' which the system its name. Let the elector vote for as many candidates as are to be elected, not by numbering them in an order of preference, but by giving a vote to each of them as he does for the Federal Senate." If he is a party man, then, of course, he will give all his vote to candidates on his party's list, and he will have no difficulty in

2—H. 39.

H.—39

10

discovering who these are. For he will receive a ballot-paper something like this, and if he is a good Labourite this is how he will vote :— Balloi-paj)er. LABOUR. LIBERAL. INDEPENDENT LIBERAL. Gray .. 1 Bakhap . . Cameron .. Hurst .. 1 Clemons . . —. — 1 Keatingj. . There is very little difference from the ballot-papers we shall use in May, only that good Labourites are put in a list by themselves and are labelled, and so are the Liberals ; and the Independent Liberal is left alone by himself. This little change would disturb no one ; any one who understands how to vote now would understand how to vote with the party-list ballot-paper. But notice what a change there is from the voting under the Hare system. With that system we should have a list of seven names, and the elector would have to arrange them, or at least three of them, in an order of preference. The average elector is a good party man, and when his party puts up three candidates he is content to vote for all of them on equal terms. He does not want to decide •between Mr. Bakhap and Senator Clemons, and he does not wish to place Senator Keating behind either of them ; he wants them all to get in, and he wants to help each of them just as much as the others. Also, the candidates can canvass for one another instead of outbidding one another for No. l's; Senator Keating can ask electors to vote No. 1 for Mr. Bakhap as well as for himself, and Senator Clemons can help both of them. The Labour party has already felt the bad effects of the struggle for No. l's : this is one of the reasons why each elector has several l's in the pre-elections. With the party-list he would have the same, but there would be no subsequent preferences : these are not necessary. When the poll is closed the presiding officer at each polling-place counts the l's for each candidate and telegraphs them to headquarters. There is no waiting for the ballot-papers to come in, as with the Hare system. It is just this feature which makes the Hare system almost impossible for a place like Queensland, where it takes weeks for mails to reach Brisbane. With, the party-list it is necessary only to get the l's by telegraph, and few polling-places are far from a line. The Eeturning Officer adds up the l's for each candidate at all the polling-places, and then the l's for all the candidates of each party. Suppose the following is the result; it is the result assumed in an article a few days ago in the Mercury, but probably the Liberals will not give their Independent as many votes as are shown here : —■ Totals for the Candidates. LABOUR. LIBERAL. INDEPENDENT LIBERAL. dray ..32,000 . Bakhap .. 31.000 Cameron.. 17,000 Hurst .. 32,000 Clemons.. 31,000 — .. 32,000 Keating .. 31,000 Total .. 96,000 Total .. 93,000 Total .. 17,000 The Labour party has obtained 96,000 —.that is, 32,000 electors have voted 'for it. The Liberals have obtained 110,000 votes —that is, between 36,000 and 37,000 electors have Voted for their candidates. This is likely to be something like the result of the election, unless one party gains very much on the other ; but probably the Liberal League will see that Senator Cameron does not get 17,000 votes, and if they succeed in keeping him down to 14,000 then there will be no Labour senators from Tasmania. With the " block " Vote that we shall use in May what will happen ? The three highest on the poll will be Gray, Hurst, and " Comrade X," yet to be selected, and these three will be elected. That is, a party with less than half of the votes—96,ooo out of 206,000 —-gets all the seats. No man can say this is fair ; no man can say that the Labour party deserves this reward simply because its discipline is better than the discipline of the Liberals. The fair result obviously would be that the party having just over half the votes should get two seats ; the Liberals should return two senators and the Labour party one. Now, what would happen with the party-list system ? The Returning Officer would find the total of all the votes polled ; this is 206,000. There are three members to be returned ; one-third of 206,000 —.that is, say, 68,667 votes —should return one member. Any list that has 68,667 votes is entitled to a member. The Labour party, then, in this first stage gets one member and has 27,333 votes over ; the Liberal party also gets one and has 24,333 votes over ; the Independent Liberal does not get in, and his 17,000 votes are not used. There is still one seat to be allotted ; which party shall get it ? If the Independent Liberal had said he had nothing to do with the real Liberals—-if he had labelled himself Tory, suppose—nothing could be done with his Votes ; they couki not be used to help either Liberal or Labour, and the remaining seat would go to the Labour party with this 27,333 votes, rather than to the Liberal party with only 24,333 votes ; and so the Labour party would get two seats and the Liberals only one, as would indeed be fair if there were no affinity between the Liberals and the Tory.

11

H.—39

But the Independent has not labelled himself Tory, and the official Liberals have not labelled themselves Tories ; they have all called. Jthemselves'Liberals. And[they have availed themselves (let us. assume) of a provision of the party-list system that lists may be grouped—that is, they have told the Eeturning Officer in advance remainder which either list has shall be used to help the other list. The Returning Officer now carries out this direction ; he adds the Independent Liberal's 17,000 votes to theVLiberals' spare 24,333, making 41,333, a total much larger than the 27,333 spare votes of the Labour party, and the Liberal party thus gets a second seat. And this, as we saw, is the fair result. The seats have now been allotted to the parties. All that remains is to ascertain which candidates shall be returned for each party. Here, again, the system is simple ; there is no quota or surplus, or exclusion, or anything of that sort to occupy the Returning Officer and his staff all night through. If a list is entitled to return two candidates the Returning Officer declares elected the two candidates on that list who have received most votes. If there are several candidates equal the Returning Officer must toss up or decide in any way he pleases. And that is all. The party-list, of course, has difficulties ; there is no electoral system which has not; but these can wait for a second article. In our first article we saw how simple the party-list system is. At the polling-booth it is simply the familiar " block " vote which we use for the Senate recorded on a ballot-paper on which candidates are arranged according to their parties. In the Returning Officer's room it requires only a simple rule-of-three sum and the result of the election can be ascertained within a few minutes after the totals for the candidates at the various polling-places are received. It gives proportional representation, and every one can understand how it gives it. Within a party it enables the return of the candidates most in general favour, and it allows a voter to give equal Votes to several candidates. But now we have to point out a peculiarity not altogether in favour of the party-list system, and to consider some of the objections to it. Let us have before us again the ballot-paper and the totals for the candidates in the example we used in our first article : — Ballot-paper. LABOUR. LIBERAL. INDEPENDENT LIBERAL. Gray . . 1 Bakhap . . Cameron .. Hurst .. 1 demons f. . —■ —■ .. 1 Keating .. Totals for the Candidates. Gray .. 32,000 Bakhap .. 31,000 Cameron.. 17.000 Hurst .. 32,000 demons.. 31,000 — .. 32,000 Keating .. 31,000 Total .. 96,000 Total .. 93,000 Total .. 17,000 The peculiarity of this system, in contrast to the single transferable vote, is that a vote is no longer a vote only for candidates as individuals—it is a vote also for the candidate's party. "In the first place," says an English writer, " it is a vote for the party-list as such, and is used for determining the proportion of seats to be allotted to the lists ; and in the second place it is a vote for a particular candidate for the purpose of ascertaining which of the candidates included in a list shall be declared successful." To take an example : Some of Mr. Bakhap's friends among the miners are Labourites, but they would like to see their neighbour in the Senate even if that means that there will be one Labour senator less. Accordingly they vote—Gray 1, Hurst 1, Bakhap 1. Now the 1 that Mr. Bakhap gets from such a paper is not only counted in his own total, but it is counted in the Liberal total; also, it is lost from the Labour total. If the election is close and many of Mr. Bakhap's friends do this it may mean that the Liberals will get a seat which otherwise would have gone to Labour. If this extra seat goes to Mr. Bakhap this is, of course, what his Labour friends intended, and they have secured what they wanted. But the point is that this seat, which the Liberals are to have solely because Mr. Bakhap's Labour friends voted for him, may not go to Mr. Bakhap. It may happen that, in spite of the Liberal League, many Liberals will vote for the three retiring senators, leaving out Mr. Bakhap ; and if they do so the candidates with most votes in the Liberal list will be Senator Clemons and Senator Keating, and these two and not Mr. Bakhap will be elected. So that Mr. Bakhap's friends have not only failed to get him in, but they have put in Senator Clemons, for whom they have no use, in place of the second Labour senator whom they might have returned. Cross-voting, then, is very dangerous with the party-list, or at all events with the form of it which has been described here. There is a party-list system in use in Finland in which the voter indicates his order of preference. Mr. Bakhap's friends would thus be able to say that if their votes were no use to him they were to go to the Labour senators, and were not on any account to be used for Senator Clemons ; but we are not.likely to adopt this, for the principal reasons for a change from the single vote to the party-list are the objections to the order of preference. If the party-list is adopted voters must be carefully warned of the danger of cross-voting, and no doubt most people will vote only in one list. But it is no great hardship to have to do this ; most of us want to do so now.

H.—39

12

There is a second danger often feared from the party-list. Let us again explain by an example, and this time we will take an actual case — the Senate election in Victoria in 1910. The votes obtained by the candidates were as follows : — Victoria — -Senate, 1910. LABOUR. LIBERAL. Findley .. .. 217,573 Best .. .. 213,976 Barker .. .. 216,199 Trenwith .. .. 211,058 Blakey .. .. 215,117 McCay .. .. 195,477 Goldstein .. .. 53,583 Eonald .. .. 18,380 _ _ ■ Total .. .. 648,889 Total .. .. 692,474 The result of the poll was that Labour won all the seats ; it ought to have had only one, as it had less than half of the votes. If the party-list method had been used the Liberals would have returned Best and Trenwith, and the Labour party Senator Findley. Now, the point is that the votes for the Labour candidates are almost equal, and that a few hundred persons who vote for one of them and not for the others will be able to decide who the Labour senator is to be. The danger that is feared is that the Liberal party, knowing itself to be in a majority and sure of two seats, and thinking that Mr. Barker and Mr. Blakey would be a much stronger senator and more dangerous opponent than Mr. Findley, will arrange for a few hundreds of its supporters to give votes to Mr. Findley, with the deliberate object of securing the return of the weakest Labour candidate. Thus it will be the Liberal party and not the Labour party which will decide who the Labour senator is to be. : But the danger does not seem to be a very real one. It will not often happen that a party is so sure of its strength that it will deliberately give votes to the other side. For the Liberal lead may be much less than is" supposed, and it may happen that these few votes will be just enough to bring the Labour total above the Liberal and so give two seats to Labour. However, if the danger is a real one it can easily be overcome. Taking now the case of the State electorates with six seats, it will be sufficient to provide either that an elector shall have five votes and not six, and that a list may contain more names or less names than five but not five, or else that an elector shall have six votes, and that a list may not contain six names, but either more or less than six. Either of these plans will result in the general favourites in the party getting a lead which will put them beyond the reach of any scheme of- their opponents or of any section of malcontents in their own party. The party-list system recognizes parties, but it is important to remember that it does not recognize party organization. It is not the party organizations which submit the lists—the candidates themselves do that—'and all that the Returning Officer has to know is that candidates consent to be placed in the same list. Of course, we shall not escape the influence of party organizations—the Workers' Political League and the Liberal League will still decide who are to be the official candidates —but it will be possible, as it is now, for independent candidates to run. Thus there will be no obstacle to the rise of a new party ; the only difficulty will be that it will have to run as many candidates as an elector has Votes. Of course, the party-list would not have suited Labour at the last Senate ; any proportional system would have reduced the number of Labour senators. But the block vote may not suit Labour at the next Senate election ; if the election is fought by three candidates on each side there may be Very few Labour senators. For the State Parliament the representation of the parties would not be affected, unless the number of members is altered or some device is introduced (as would not be unreasonable) to favour the larger party. In case of casual vacancies the party-list might suit Labour. In Belgium a casual vacancy is filled not at a new election, but by the next available candidate on the list from which the Vacancy occurred. Thus the balance of the parties is not disturbed by a casual vacancy. At present it is of the greatest importance to the Government that none of the Liberal members for Darwin or Bass should meet with a fatal accident, for if one does the vacancy will be filled by a Labour member, and out goes the Government. On the other hand, all Labour members for Denison and Franklin and Wilmot must take great care of themselves ; if one of them goes the Government majority is increased to four. With the party-list all these mischances can be provided against. The party-list is simple for the voter and simple for the Returning Officer ; any one can understand it. It recognizes that there are parties, but it does not add to the power of party organizations. It does away with the struggle for No. I's, and does not require electors to state a preference which they do not feel. It goes straight to its object—the representation of parties, not the election of persons. Switzerland, and Sweden, and Finland, and Belgium, and France, and other countries use it, and find that it does what it professes to do. Why should not Tasmania try it ? [Note. —Remainder of paper was not ordered to be printed.] Approximate Cost of Paper. —Preparation, not given; printing (1,400 copies), £0.

Authority : John Mackay, Government Printer, Wellington.—l9l3.

Price 6d.]