[Footnote 3: If in a four-member constituency the number of voters is 21,000 and the parties are in the ratio of 12,000 to 9000, the larger party would, if each elector had three votes, have 36,000 votes in all and the smaller party would have 27,000. No candidate of the smaller party could obtain more than 9000 votes, whilst the 36,000 votes of the larger party carefully divided among four candidates would also allow each candidate to receive 9000 votes. If then the larger party had slightly more than 12,000 supporters out of a total of 21,000, the larger party would obtain all four seats, as each of its candidates would, if the votes were carefully distributed, receive more than 9000 votes each.]

[Footnote 4: "If it is desired that the body should not be a representation of a single interest and a single class of opinions, some means must be adopted to guard against its falling entirely into the hands of the dominant party. With this view we would recommend that, in the election of the council, each elector should have as many votes as there might be members to be chosen, and should be entitled to give all these votes to a single candidate, or to distribute them among several. By this arrangement a monopoly of power in the Legislative Council by any one party, or any one district of the Colony, would be prevented, since a minority of the electors, by giving all their votes to a single candidate, would be enabled to secure his return."—Earl Grey, The Colonial Policy of the Administration of Lord John Russell, vol. ii., Appendix, p. 362.]

[Footnote 5: East India (Advisory and Legislative Councils, &c.) (Cd. 4426), p. 14.]

[Footnote 6: East India (Advisory and Legislative Councils, &c.) (Cd. 4426), p. 45.]

[Footnote 7: Port of London Act, 1908, Schedule I., Part IV. (1).]



CHAPTER V

THE SECOND BALLOT AND THE TRANSFERABLE VOTE IN SINGLE-MEMBER CONSTITUENCIES

"Le depute, au lieu de representer la majorite des electeurs, devient prisonnier de la minorite qui lui a donne l'appoint necessaire pour son election."

—YVES GUYOT

" ... every fool knows that a man represents Not the fellers that sent him, but them on the fence."

—J. RUSSELL LOWELL

Three-cornered contests.

It was stated in the first chapter that the rise of the Labour Party as a political force, with an organization wholly independent of those of the older parties, would make a change in our voting system imperative. Both prior and subsequent to the appointment of the Royal Commission on Electoral Systems political organizations have shown themselves keenly alive to the necessity of such a change. At the meeting of the General Committee of the National Liberal Federation at Leicester, on 21 February 1908, a resolution in favour of the early adoption of the second ballot was carried unanimously. The Trades Union Congress, at its meeting in September 1908, less eager to pronounce in favour of a reform of such doubtful value, passed a resolution in favour of an authoritative "inquiry into proportional representation, preference or second ballots, so that the most effective means of securing the true representation of the electors may be embodied in the new Reform Bill." The spokesman of a deputation from the Manchester Liberal Federation, which waited upon Mr. Winston Churchill on 22 May 1909, said: "The point on which we wish to speak to you to-day is the reform of the present system of voting, which we hold to be out of date, archaic, and in great need of reform." Mr. Churchill's reply was a significant reinforcement of Mr. Asquith's previous declaration, that "it was impossible to defend the present rough and ready methods." "I think," said Mr. Churchill, "the present system has clearly broken down. The results produced are not fair to any party, nor to any section of the community. In many cases they do not secure majority representation, nor do they secure an intelligent representation of minorities. All they secure is fluke representation, freak representation, capricious representation." The figures of two bye-elections—those of the Jarrow Division of Durham and the Attercliffe Division of Sheffield—will show how completely Mr. Churchill's language is justified. The figures are as follows:—

JARROW ELECTION, 4 July 1907

Curran (Labour) 4,698 Rose-Innes (Conservative) 3,930 Hughes (Liberal) 3,474 O'Hanlon (Nationalist) 2,124 _ 14,226

ATTERCLIFFE ELECTION, 4 May 1909

Pointer (Labour) . . . . 3,531 King-Farrow (Unionist) . . . 3,380 Lambert (Liberal) . . . . 3,175 Wilson (Ind. Unionist) . . . 2,803 _ 12,889

In the case of Jarrow the successful candidate obtained just less than one-third of the votes polled, and in the case of Attercliffe the member returned represented a little more than a quarter of the electors. The representation which results from elections of this kind is without doubt most capricious and uncertain in character. A House of Commons so built up could have no claim to be representative of the nation, and its composition would be so unstable as seriously to impair its efficiency. Nor can we afford to regard such elections as being a mere temporary feature of our parliamentary system. The General Election of 1906 showed a notable increase in the number of three-cornered fights over previous general elections, and the bye-elections during the four years 1906—1909 were marked by a still further increase. The Report submitted by the Executive Committee of the Labour Party to the Portsmouth Conference in January 1909 foreshadowed a very large addition to the number of Labour candidates. Some thirty-eight candidates, in addition to the then existing Labour members in Parliament, had been formally approved by the Executive Committee of the Labour Party after due election by the Labour organizations to which the candidates belonged, and although constituencies were not found for all of these new candidates, the number of three-cornered contests in the election of Jan. 1910, in which Liberal, Unionist, Labour (or Socialist) took part, was no less than forty-one, and this number would have been greater had not several Liberal candidates withdrawn. Owing to the desire on the part of the Liberal and Labour parties to avoid the risk of losing seats there were in the elections of December 1910 fewer three-cornered fights. But the Labour party, the permanence of which is no longer open to question, will not be content to remain with its present share of representation. It can however gain additional seats only at the expense of the older parties, and although the Liberal party, as in the Mid-Derby bye-election of May 1908, may sometimes yield seats to Labour nominees, it is not to be expected that the Liberal organizations will always be willing to give way. At the Mid-Glamorgan bye-election in May 1910 the local organization, against the advice of the chief Liberal Whip, nominated a Liberal candidate, and succeeded in retaining the seat although it had been "ear-marked" by the Labour Party. In Scotland, where Liberalism is less complaisant than in England, no seat has been surrendered to the Labour Party without a fight, and when a Labour candidature was threatened in December 1910, in the Bridgeton division of Glasgow, the Liberals retaliated by threatening to place a Liberal candidate in the Blackfriars division where Mr. Barnes, the Labour representative was again standing. These facts should dispel any illusion, if such still exist, that the problem of three-cornered fights is a transitory phenomenon which can safely be ignored. The political organizations, with a true instinct, have realized the importance and urgency of this problem, and increasing pressure will doubtless be brought to bear upon the Government to introduce a system of second ballots, or some other electoral method, that will give effect to what Mr. Churchill has described as "the broad democratic principle, that a majority of voters in any electoral unit, acting together, shall be able to return their man." The advocates of the second ballot and cognate methods of reform seek a solution of this one problem only. They desire to maintain the essential characteristic of the present system—the exclusive representation of the majority in each constituency—and make no attempt to remedy any of the other evils associated with single-member constituencies. But the question at once arises whether the problem of three-cornered contests can be solved by attempts to preserve the distinctive feature of the present system—the representation of the majority only. A little reflection must convince the reader that such a solution deals with the form of the problem rather than with its essence. For the new problem arises from the fact that three parties instead of two are now seeking representation in Parliament, and no remedy can be regarded as effective which does not provide for the realization of the legitimate aspirations of all three parties. This the system of second ballots has completely failed to do; indeed its results only reinforce the arguments of previous chapters, that so long as we compel the electors of any one district, whatever their divisions of opinion, to be all represented by one man, their real representation will be impossible. An examination of the effects of the second ballot in those countries in which the system has been tried fully justifies these statements, and fortunately the body of experience now available is so considerable that the conclusions to be drawn therefrom have an authoritative character.

The second ballot.

The Reports furnished by His Majesty's representatives abroad show that the second ballot, in one form or another, is, or has been, in force in the majority of continental countries. The forms differ in detail, but reference need only be made to the three chief types. In Germany the two candidates highest at the first poll proceed to a second election. It was this form of the second ballot that was introduced into New Zealand in 1908. In France all candidates in the original election and even fresh candidates may stand at the second election. At this second poll a relative—not an absolute—majority of votes is sufficient to secure the election of a candidate. As a rule only the two candidates highest at the first election take part in the second ballot, and therefore in practice the German and French methods closely approximate to one another. The third type concerns the application of the second ballot to the scrutin de liste or block vote in multi-member constituencies. It was formerly used in the Belgian parliamentary elections, and is still employed in the election for the Belgian Provincial Councils. The candidates who receive the support of an absolute majority of the electors voting at the first ballot are at once declared elected; the candidates next highest on the poll, but only so many as are equal to double the number of vacancies remaining to be filled, take part in a second ballot.

The object of the second ballot—to ensure that every elected candidate should finally have obtained the support of a majority of the electors voting in the constituency for which he has been returned—has, generally speaking, been achieved. But that does not solve the problem of the representation of three parties; a general election based on such a system yields results which are far from satisfactory. The party which is unsuccessful in one constituency may suffer the same fate in the majority of the constituencies, and this is the fatal flaw in all forms of the second ballot. Moreover experience has shown, and it is evident a priori, that with this system the representation of any section of political opinion depends not upon the number of its supporters, but very largely upon the attitude taken towards it by other parties. For, at a second ballot, the result is determined by the action of those smaller minorities which were at the bottom of the poll at the first ballot. No party can be certain of securing representation unless in its own strength it can obtain an absolute majority in at least some of the constituencies. The largest party in the State, if its voting strength is evenly distributed, may be at the mercy of hostile combinations at the second ballots, unless it is so large as to command a majority of votes throughout the country, and when three parties have entered the political arena it rarely happens that any one of them is in this favourable position. That being so, the new element of uncertainty associated with the system of second ballots may yield results which are further removed from the true representation of the whole electorate than the results of the first ballots.

Experience in Germany.

Continental experience has shown that the coalitions at the second ballots are of two types. One party may incur the hostility of all other parties, and if so, the second ballots will tend uniformly to the suppression of that party. The combination of parties whose aims and purposes are to some degree allied may be regarded as legitimate, but the cumulative effect of such combinations over a large area is most unfair to the party adversely affected. No defence at all can be urged in palliation of the evils of certain other coalitions also characteristic of second ballots—the coalitions of extreme and opposed parties which temporarily combine for the purpose of wrecking a third party in the hope of snatching some advantage from the resulting political situation. Sometimes such coalitions are merely the expression of resentment by an advanced party at the action of a party somewhat less advanced than itself. But, whatever the cause, the coalitions at the second ballots do not result in the creation of a fully representative legislative chamber; on the contrary, they tend to take away all sincerity from the parliamentary system. Illustrations of the first type of coalitions abound. The German general elections afford numerous examples, but as a special note on the working of the second ballots in Germany is to be found in Appendix II., it will suffice to quote some of the results of the election of 1907. The Social Democrats were engaged at the second ballots in ninety constituencies. At the first ballots they were at the head of the poll in forty-four of these constituencies, but at the second ballots they only succeeded in retaining that position in eleven. In the forty-six constituencies in which they were second at the poll they were only able to improve their condition in three cases. These figures show how the German Social Democrats suffered from hostile combinations. It was with the utmost difficulty that they obtained representation in constituencies other than those in which at the first elections they were in an absolute majority. No wonder that one of the planks of the platform of the Social Democratic party is proportional representation.

Austria.

The Social Democrats of Austria suffered in the General Election of 1907 in the same way. Professor Kedlich,[1] in an article entitled "The Working of Universal Suffrage in Austria," wrote as follows: "The Christian Socialists have ninety-six seats in the new House, the Social Democrats eighty-six ... The number of seats won by them weighs still heavier in the balance when we reflect that in many second ballots the majority of the opponents of social democracy joined their forces against them. Not less instructive are the relative numbers of the votes recorded for each of the parties. Over a million votes were given to the Social Democrats as against 531,000 for the Christian Socialists." Such results destroy the representative character of legislative bodies. The same lesson on a smaller scale is to be gathered from the Italian elections. Speaking of the General Election of 1904, the Rome correspondent of The Morning Post pointed out that, in not a few constituencies, like the second division of Rome, a rally of Clericals at the second ballots enabled the Conservative Monarchists to triumph over the Socialists.

Belgium.

The combinations of allied parties against a third party, as in the examples already given, may be defended, but the coalitions at second ballots, as has been pointed out, are not always of this character. Should parties, angered and embittered by being deprived of representation, use their power at the second ballots to render a stable Government impossible, then the results are disastrous. Such were the conditions which obtained in Belgium before the abandonment of second ballots. "The system," says Sir Arthur Hardinge, "answered well enough so long as only two parties contested an election; but the moment the Socialist Party formed a distinct third party, after the establishment of universal suffrage in 1894, it began to act in a manner which produced unsatisfactory results.... The overwhelming victory of the Clerical party in 1894 was largely due to the fact that in every second ballot between Catholics and Socialists the Liberals voted for the former, whilst in every second ballot between Catholics and Liberals, with the single exception of the Thuin Division, the Socialists preferred the Catholics as the creators of universal suffrage and as, in some respects, a more genuinely democratic party, to the Liberals, whom the Labour leaders regarded with peculiar hatred as the apostles of free competition and individualism. In 1896 the Socialists were in their turn the victims, as the Liberals had been in 1894, of the working of the system of second ballots. Liberal electors at these elections voted everywhere at the second ballots for Clerical against Labour candidates, with the result that the Clericals won every one of the eighteen seats for Brussels, although the total number of Clerical electors in a total electorate of 202,000 was only 89,000, as against 40,000 Liberals and 73,000 ultra-Radicals and Labour men. Two years later the Liberals swung round to an alliance with the Socialists against the Clericals, and in several constituencies, owing to the system of second ballots, the Socialists, although actually in a minority, won all the seats with the help of the Liberals, who on the first ballot had voted unsuccessfully for Liberal as against both Catholic and Labour candidates. It was the practical experience of conditions such as these which gradually convinced all the Belgian parties that, given a three-cornered fight in every, or nearly every, constituency, the only way of preventing a minority from turning the scales and excluding from all representation the views of nearly half the electorate was to adopt the system of proportional representation."[2]

Count Goblet d'Alviella furnishes an excellent example of the working of the second ballots at Verviers in the General Election of 1898, the last parliamentary election in Belgium, at which second ballots were used. In the election for Senators the Socialists spoiled the chances of the Liberals by voting for the Clericals, whilst, in the election for the Chamber, the Liberals, not to be outdone, spoiled the chances of the Socialists by also supporting the Clericals. The Clericals thus obtained all the seats both in the Senate and in the Chamber with the assistance of the Socialists and of the Liberals in turn. The absurdities of the General Election of 1898 were so flagrant that on the day after the election so determined an opponent of proportional representation as La Chronique exclaimed, "Can anything be more absurd than the working of the second ballots in this country? ... What becomes of the moral force of an election in which parties are obliged, if they wish to win, to implore the support of electors who yesterday were their enemies? Such support is never obtained without conditions, and these conditions are either promises which it is not intended to keep or a surrender of principles—in either case a proceeding utterly immoral."[3]

France.]

French elections also furnish examples of the use of the second ballots for the purpose of fostering dissension between opponents. At the General Election in 1906 it was stated that the Conservatives in the South of France, despairing of obtaining representation themselves, intended to support the Socialists at the second ballot in the hope of obtaining an advantage by accentuating the difference between the Socialists and the Radicals. M. Jaures indignantly denied that there was any understanding between the Socialists and the Conservatives, and took advantage of the accusation to write in L'Humanite a powerful plea for proportional representation. "This reform," he declared, "would make such unnatural alliances impossible. Each party would be induced and, indeed, it would be to each party's advantage to fight its own battle, for every group would have an opportunity of obtaining its full share of representation. There would no longer be any question of doubtful manoeuvres, of confused issues; Socialism would have its advocates, Radicalism its exponents, Conservatism its leaders, and there would be a magnificent propaganda of principles which would inevitably result in the political education of the electorate. Every movement would be assured of representation in proportion to its real strength in the country; every party, freed from the necessity of entering into alliances which invariably beget suspicion, would be able to formulate quite clearly its essential principles; governmental and administrative corruption would be reduced to a minimum; the real wishes of the people would find expression; and if parties still continued to dispute for power, it would be to enable them to promote the more effectually the measures for which they stood." In spite, however, of this eloquent disclaimer on the part of M. Jaures, the Conservatives have at the bye-elections continued their policy of supporting the Socialists. The bye election of Charolles in December 1908 is a case in point. At the first ballot the figures were as follows:—

M. Sarrien fils (Radical) 5,770 votes M. Duoarouge (Socialist) 4,367 " M. Magnien (Conservative) 3,968 "

At the second ballot—

M. Ducarouge (Socialist) 6,841 " Elected M. Sarrien fils (Radical) 5,339 " M. Magnien (Conservative) 301 "

It should be explained that the Conservative candidate, although his name still appeared upon the ballot paper, retired before the second election, and it is evident that the votes of many of his supporters were given to the Socialist candidate. In the following April (1909) several further instances occurred. At Uzes a vacancy was caused by the death of a Radical Socialist member who, at the General Election of 1906, had beaten the Duc d'Uzes, a Reactionary, the Socialist candidate on that occasion being at the bottom of the poll. In the bye-election the Socialist was returned at the head of the poll, but so obvious was the fact that the Socialist owed his victory to Conservative support, that he was received in the Chamber by the Radicals with the cry of "M. le duc d'Uzes." Uzes was typical of other elections and, as the Paris correspondent of The Morning Post remarked, "the successes of the Unified Socialists in the recent series of bye-elections are in part to be attributed to the votes of the Reactionaries, who voted for the Unified candidates as being enemies of the Republic." This abuse of the purpose of second ballots—an abuse engendered by the failure of the minority to obtain direct representation—destroys the last semblance of sincerity in the representation of a constituency, and must hasten the abolition of the second ballots in France in the same way as combinations of a similar nature rendered imperative the introduction of a more rational system of election in Belgium.

The foregoing facts are sufficient to show that a system of second ballots does not necessarily result in the formation of a legislative chamber fully representative of the electorate. In Germany the largest party has had its representation ruthlessly cut down by the operation of the second ballots. Indeed, were it not for the overwhelming predominance of this party in certain areas it might not have obtained any representation whatever. In Belgium the effect of the second ballots was to deprive the middle party, the Liberals, of their fair share of representation. In 1896, owing to the coalitions of Socialists and Catholics at the polls, the Liberals had only eleven representatives in the popular chamber. All their leaders had been driven from Parliament, their electoral associations had become completely disorganized save in some large towns, and in many constituencies they had ceased to take part in elections. Yet the results of the very first elections (1900) after the establishment of proportional representation, showed that the Liberals were the second largest party in the State, and that it was a party which still responded to the needs and still gave voice to the views of large numbers of citizens.

The bargainings at the second ballots in France.]

The system of second ballots not only deprives large sections of the electorate of representation, but the very coalitions which produce this result bring parliamentary institutions into still further disrepute. These coalitions are condemned in unequivocal terms by Continental writers and statesmen of widely differing schools of thought. The scathing language of M. Jaures has already been quoted, and we find his views endorsed by politicians of the type of M. Deschanel, an ex-President of the Chamber of Deputies, who declared that these coalitions entirely falsify the character of the popular verdict. Again, M. Yves Guyot, an ex-Minister, asserts that "the second ballots give rise to detestable bargainings which obliterate all political sense in the electors." M. Raymond Poincare, a Senator and a former Minister, condemns the system of second ballots in equally forcible language. "It will be of no use," he says, "to replace one kind of constituency by another if we do not, at the same time, suppress the gamble of the majority system and the jobbery of the second ballots." These expressions of opinion on the part of individual French politicians could be multiplied, but it will be sufficient to add to them the more formal and official declaration of the Commission du Suffrage Universel, a Parliamentary Committee appointed by the Chamber of Deputies. In the Report issued by this Committee in 1907, it is declared that "the abolition of the second ballots with the bargainings to which they give rise will not be the least of the advantages of the new system [proportional representation]."

The "Kuh-Handel" in Germany.

It would appear that the German second ballots are also characterized by this same evil of bargaining. Karl Blind, writing in The Nineteenth Century, March 1907, stated that "in this last election the oddest combinations have taken place for the ballots in the various parts of the Empire and within different States. There was no uniformity of action as to coming to a compromise between Conservative and Liberal, or Liberal and Social Democrat, or Centre and any other party, as against some supposed common enemy who was to be ousted from his insufficient majority by a subsequent alliance between otherwise discordant groups, or who wanted to have his insufficient majority increased to an absolute one by the addition of the vote of one of the defeated candidates whose friends finally choose the 'lesser evil'....

"To some extent these necessary, but sometimes rather sordid, transactions are made all the more difficult through the very existence of separate States with 'Home Rule' legislatures of their own. Political development has in them gone so far in a centrifugal sense that the nation has been sadly split up and the public mind too much divided into merely local concerns and issues....

"Irrespective of this baneful influence of a so-called 'Home Rule' state of things on the life of the nation at large, I must confess that the huckstering at the second ballots does not strike me as an ideal institution. It generally goes, in Germany, under the name of Kuh-Handel (cow-bargain). It often brings out the worst symptoms of intrigue and political immorality.... Those who dabble in the Kuh-Handel either lead their own contingent as allies into an enemy's camp from spite against another adversary, or they induce their own men to desist from voting at all at a second ballot, so as to give a chance to another candidate, whom they really detest with all their heart, but whom they wish to use as a means of spiting one still more deeply hated."

The position of a deputy elected at a second ballot.]

The separate experiences, therefore, of France, Belgium, and Germany all yield convincing and corroborative testimony to the demoralizing influence on political life which results from the coalitions at the second ballots. Insufficient attention, however, has been directed to one aspect of this influence, its pernicious effect upon the inner working of parliamentary institutions. The deputy who is elected as the result of a coalition of forces at the second ballot finds himself in an extremely difficult and unstable position. Instead of being the representative of the majority of the electors he too often becomes, in the apt phrase of M. Yves Guyot, "the prisoner of the minority," and, whilst in Parliament, he is being continually reminded of the power of that minority to make or unmake him at the next election. The persistent pressure of that minority explains those contradictory votes in the French Chamber which, to a foreigner, are often incomprehensible. The deputy will usually act in accordance with the opinion of the group to which he belongs and vote accordingly, but at a subsequent sitting he will find it necessary to vote in such a way as will give satisfaction to that minority whose support assured his success at the previous election, and without whose support he cannot hope for re-election when the time comes for a fresh appeal to the country. The pressure which such a minority can exert must often be intolerable, and must, in any case, render it impossible for any deputy either to do justice to himself or to the legislative chamber to which he belongs.[3]

The alternative vote.

The shortcomings of the system of the second ballot are so pronounced and are so generally recognized that there now exists but little, if any, demand for its introduction into this country, and more attention has therefore been given to the mechanism of the alternative vote as affording a means of securing the object of the second ballot whilst avoiding many of its inconveniences. Under this suggested plan the voter is invited to mark his preferences against the names of the candidates on the voting paper by putting the figure "1" against his first favourite; the figure "2" against the man he next prefers, and so on through as many names as he may choose to mark. At the end of the poll the number of papers in which each candidate's name is marked "1" is ascertained, and if one of them is found to have secured the first preferences of an absolute majority of all the persons voting, he is declared elected; but if no candidate has obtained such a majority the papers of the candidate who has obtained the least number of first preferences are examined and transferred one by one to the candidate marked "2" upon them. In this transfer, the papers on which only one preference had been marked would be ignored, the preferences, to use the current phrase, being "exhausted." If, as the result of this transfer, any candidate has secured the support of an absolute majority of the number of effective preferences he is declared duly elected; but if there is still no candidate with an absolute majority the process is repeated by distributing the papers of the candidate who is left with the lowest number of votes, and so on until some candidate has got an absolute majority of effective preferences.

The alternative vote undoubtedly possesses many and valuable advantages as compared with the second ballot. In the first place, its introduction into the English electoral system would keep English voters in touch with Colonial rather than with Continental practice. Preferential voting[4] has been in use in Queensland since 1892; it was adopted in 1907 by the West Australian Parliament, and was proposed in a Bill submitted by Mr. Deakin to the Australian Commonwealth Parliament in 1906. Moreover, the alternative vote enables the election to be completed in a single ballot; and the fortnight that is wasted between the first and second ballots on the Continent would be saved. There has also been claimed for this method of voting this further advantage, that it would prepare the way (perhaps by rendering it inevitable) for the more complete reform—proportional representation.

The principle of the alternative vote is extremely simple. It is embodied in two Bills which were introduced into the House of Commons in 1908 by Mr. John M. Robertson and by Mr. Dundas White; and also in a modified form in a Bill introduced in 1907 by Mr. A.E. Dunn. Its purpose and mechanism is set forth in the memorandum of Mr. Robertson's Bill as follows:—

"The object is to ensure that in a parliamentary election effect shall be given as far as possible to the wishes of the majority of electors voting. Under the present system when there are more than two candidates for one seat it is possible that the member elected may be chosen by a minority of the voters.

"The Bill proposes to allow electors to indicate on their ballot papers to what candidate they would wish their votes to be transferred if the candidate of their first choice is third or lower on the poll and no candidate has an absolute majority. It thus seeks to accomplish by one operation the effect of a second ballot."

Mr. Robertson's Bill, as originally introduced in 1906, was applicable to single-member constituencies only; but the amended form in which the Bill was re-introduced provided for the use of the transferable vote in double-member constituencies as well, but, in doing so, still maintained the essential characteristic of the existing system of voting—that each member returned should have obtained the support of a majority of the electors voting. Mr. Dundas White, however, in applying the alternative vote to double-member constituencies, made a departure from this principle, and proposed to render it possible for a candidate to be returned who had obtained the support of less than one-half but more than one-third of the voters.[5] The effect of Mr. Robertson's Bill would have been that it would still be possible in double-member constituencies for the party finally victorious to secure both seats; whilst with Mr. Dundas White's provisions the two largest parties would in all probability have obtained one seat each.[6]

The difference between the two measures is, however, of no great consequence; the number of double-member constituencies is not very large, and their number may be still further reduced in any future scheme of redistribution of seats. It will, therefore, be sufficient to consider what effect the alternative vote would have in single-member areas. Let us take the Jarrow election, in which there were four candidates, and apply to that election the possible working of the alternative vote. The figures for the election may be repeated:—

Curran(Labour) . . . . 4,698 Rose-limes (Unionist). . . 3,930 Hughes (Liberal) . . . . 3,474 O'Hanlon (Nationalist) . . 2,122

The electors would, with the alternative vote, have numbered the candidates on the ballot papers in the order of their choice, and, as none of the candidates had obtained an absolute majority, the votes of the lowest candidate on the poll would be transferred to the second preferences marked by his supporters. If, for purposes of illustration, it is assumed that every one of the 2122 supporters of Mr. O'Hanlon had indicated a second preference, that 1000 had chosen Mr. Curran, 1000 had chosen Mr. Hughes, and 122 had chosen Mr. Rose-Innes, then the following table will show the effect of the transfer:—

Candidate. First Count. Transfer of O'Hanlou's Votes. Result.

Curran (Labour) 4,698 +1,000 5,698 Rose-Innes (Unionist) 3,930 + 122 4,052 Hughes (Liberal) 3,474 +1,000 4,474 O'Hanlon (Nationalist) 2,122 -2,112 —

Total 14,224 — 14,224

Only three candidates now remain for consideration, and their position on the poll as the result of the transfer is as follows:—

Curran . . . . . . 5,698 Hughes . . . . . . 4,474 Rose-Innes . . . . . 4,052

As neither has as yet obtained a majority of the total votes polled, it becomes necessary that the votes given for Mr. Rose-Innes, who is now lowest on the poll, should be transferred in accordance with the next preferences of his supporters. It is conceivable that the larger proportion of these preferences would have been given for the Liberal candidate, Mr. Hughes, rather than for Mr. Curran, and, if so, the final result might easily have been the election of Mr. Hughes as member for Jarrow.

The alternative or contingent vote in Queensland.]

Before considering the value of the transferable vote in single-member constituencies as a means of securing a true expression of the national will, it may perhaps be pointed out that the procedure prescribed by the Queensland Act differs from that contained in the English Bills. The regulations of the Queensland Act are as follows:—

"When one member only is to be returned at the election, if there is no candidate who receives an absolute majority of votes, all the candidates except those two who receive the greatest number of votes shall be deemed defeated candidates.

"When two members are to be returned, and there are more than four candidates, if there is no candidate who receives an absolute majority of votes, all the candidates except those four who receive the greatest number of votes shall be deemed defeated candidates."

It will be seen that the system here prescribed approximates to the German form of the second ballot, according to which only the two candidates highest on the poll may stand again. Were the Queensland form of preferential voting applied to the Jarrow election, both Mr. Hughes and Mr. O'Hanlon would be declared defeated candidates, and only the further preferences recorded by their supporters would be taken into account in determining the relative position of the two highest candidates, Curran and Rose-Innes. The provisions of the West Australian Act of 1907, and of Mr. Deakin's Bill of 1906, followed the more elastic and undoubtedly superior method embodied in the English proposals.

Sir J.G. Ward, in introducing the Second Ballot Bill into the New Zealand Parliament in 1908, defended the selection of this electoral method on the ground that the system of preferential voting introduced into Queensland had been a partial failure. He stated that the privilege of marking preferences had not been extensively used, and quoted the opinion of Mr. Kidston, a former Queensland Premier, that the marking of preferences should be made compulsory. As explained in the course of the New Zealand debates, part of the alleged failure of the Queensland system was due to the unnecessarily cumbrous nature of the regulations. The Queensland Electoral Acts still retain the old method of voting—that of striking out from the ballot paper the names of such candidates as the elector does not intend to vote for. The confusion produced in the mind of the elector may readily be imagined when he is instructed to strike out the names of candidates for whom he does not intend to vote in the first instance, and then to mark such candidates in the order of his choice. Moreover, the provisions, as detailed above, for giving effect to preferences are so defective that only a proportion of the preferences marked can be taken into account. Even so, preferential voting in Queensland sometimes has a decisive influence upon the result of the election, as the following example, taken from the elections of 1908, will show:—

WOOLLOONGABBA ELECTION

First Count.

Votes. 1st Candidate . . . 1,605 2nd " . . . 1,366 3rd " . . . 788 ——- Total . . . 3,759

The votes recorded for the third candidate were then distributed according to the preferences marked, which were as follows:—

1st Candidate . . . 15 2nd ,, . . . 379 No preferences . . . 394 —- 788

The result of the distribution brought the second candidate to the top of the poll, the final figures being as follows:—

2nd Candidate . . . 1,745 1st ,, . . . 1,620

West Australia

Where the more simple and straightforward instructions have been adopted, as in West Australia, it has been found that a larger percentage of the electors make use of the privilege of marking preferences. Here are the figures for the constituency of Claremont in the elections of 1908:—

First Count.

Foulkes . . . . 1,427 Briggs . . . . 825 Stuart . . . . 630 ——- Total . . . 2,888

When the votes recorded for the candidate lowest on the poll were distributed it was found that nearly 75 per cent, of his papers were marked with additional preferences. The numbers were as follows:—

Briggs . . . . . 297 Foulkes . . . . 174 No preferences . . . 165 —- Total . . . 636

The final figures were as follows:—

Foulkes . . . . 1,601 Briggs . . . . 1,122

These figures doubtless show that even in West Australia, when the transferable vote is applied to single-member constituencies, a considerable number of the electors will not indicate a preference for any candidate other than for that of their own party, but similar abstentions occur at the second ballots in France, where it is found that a considerable percentage of the electors usually refrain from going to the poll on the second occasion. The Labour Party in Queensland has sometimes issued instructions to its supporters to abstain from marking preferences for the purpose of keeping the party solid and absolutely separate from other parties. Such action necessarily increases the percentage of abstentions. Nor can any remedy for action of this kind be found in making the marking of preferences compulsory. Even in Belgium, where "compulsory voting" is in force, the compulsion only extends to an enforced attendance at the polling place. The act of voting is not compulsory, for a blank unmarked ballot paper may be dropped into the voting urn. The compulsory marking of preferences when the elector has none may still further vitiate the results of elections in a most undesirable way, whilst abstention from preference marking merely deprives those abstaining of a privilege which they might exercise if they chose. It is quite conceivable that an elector after voting for the candidate of his choice may be indifferent to the fate of the remaining candidates and, if so, an enforced expression of opinion on his part would not be of any real value, and should not be counted in determining the result of an election.

Mr. Deakin's failure to carry the alternative vote.

Does then the alternative, or contingent vote, as used in West Australia, solve the problem of three-cornered fights—the problem of three distinct parties seeking representation in Parliament? When a single seat is being contested it is doubtless sufficient if the member elected represents the average views of his constituents, but a General Election based on such a system would yield results no more satisfactory than those of the second ballots. Neither the second ballot nor the contingent vote are acceptable after their true effects are understood, a fact which explains the failure of Mr. Deakin's Government to carry their Preferential Ballot Bill in 1906. Several of the seats held by the Australian Labour Party—as in the elections of Jarrow, Colne Valley, and Attercliffe—were won by a minority vote; the Melbourne Age published the following list of seven constituencies in Victoria where Labour members represented only a minority of the voters:—

Non-Labour Labour Constituencies. Votes. Votes

Geelong . . . . 1,704 1,153 Ballarat West . . . 2,038 1,034 Jika Jika . . . . 1,366 1,183 Williamstown . . . 1,931 1,494 Bendigo West . . . 1,654 1,248 Grenville . . . . 1,457 1,268 Maryborough . . . 1,929 1,263

Totals . . . 12,079 8,643

Preferential voting would have placed these seats at the mercy of a combination of the other parties, and, somewhat alarmed by the too eager advocacy of the measure on the part of the Age, the Labour Party, which had voted for the second reading of the Bill, procured its defeat on the first division in committee. It is impossible to defend the present system by which the Labour Party, which numbered two-fifths of the voters in these seven constituencies, obtained all seven seats, but, on the other hand, it cannot be alleged that a system of preferential voting, which would have enabled the other parties to have deprived these electors of all representation, was a satisfactory solution of the difficulty. In neither case would justice be done to the claims of three parties to representation.

Probable effect of the alternative vote in England.

A consideration of the possible results of the introduction into the English electoral system of second ballots or the transferable vote in single-member constituencies will show that neither reform will solve the problem presented by the rise of a new party. It is obvious that the Labour Party could by a combination of Conservative and Liberal voters be deprived of representation in all constituencies save those in which they had the support of an absolute majority of the electorate. Nor would the conditions remain the same as they are to-day. In many constituencies in which the Liberals have allowed a straight fight to take place between Tariff Reform and Labour candidates, the Liberal Party would intervene; and should combinations at the polls result in the defeat of Labour candidates, what would be the effect upon the temper and spirit of Labour voters who found themselves under an "improved" voting system less able than before to secure representation in Parliament? Would there not possibly arise a disposition on the part of the disfranchised minority to pursue on the next occasion a wrecking policy such as has distinguished the second ballots both in Belgium and in France? Even apart from precipitate action which might arise as the result of ill-feeling, the alternative vote would afford an opportunity for a predetermined policy on the part of a minority to create dissension between the opponents. The manipulation of the alternative vote would be easily understood. An angry minority of electors could be instructed beforehand to use it, as we know from experience they have used the second ballot on the Continent. Would politicians, following an exclusive electoral policy, hesitate to avail themselves of the weapon which the alternative vote would place in their hands for the purpose of annihilating any section they especially disliked, in the same way as the Liberal Party in Belgium was destroyed by Catholic and Socialist combinations at the second ballots? We cannot escape the conclusion which all experience yields, that both these electoral methods place the representation of any party at the mercy of either temporary or permanent coalitions of other parties. To an even greater degree than under the existing regime, the result of a General Election would fail to reflect public opinion.

The advocates of the alternative vote assume, with but little justification, that this method will be free from the bargainings that have distinguished the second ballots on the Continent. The bargainings naturally take place between the first and second ballots, because that is the most suitable time for the striking of bargains, for the strength of parties is definitely known. With the alternative vote such transactions would take place before the election, upon the basis of the probable position of parties as ascertained by the party agents. Even if experience should show that the transferable vote did not lend itself so easily as the second ballot to the perpetration of those bargains which are detested by all Continental statesmen, yet it is probable that the successful candidate would, like the deputy elected under the system of second ballots, become "the prisoner of the minority." The figures of the election would disclose to what extent the member returned had owed his success to the smallest minority. This minority would be only too conscious that it held the key of the situation, and the member would doubtless be exposed to the same intolerable pressure as has been brought to bear upon members of the French Chamber of Deputies. In any case the position of the elected member would be most unsatisfactory. Were a Labour member returned with the assistance of Tariff Reform votes, would not the parliamentary relations between the various parties become as embittered as when the Unified Socialist candidate at Uzes was enabled by Reactionary votes to capture a Radical seat? What recriminations would accompany the election of a Conservative candidate whose victory was due to Labour votes given to him as an expression of resentment at the action of Liberals in other constituencies? What would be the relations between the Liberal and Labour parties if in a constituency now represented by a Labour member, a Liberal candidate, with the aid of Conservative votes, displaced him? These strained relations would not only exist within the House of Commons itself, but also and perhaps in a more pronounced form in the constituencies themselves. Such conditions would not only invite the sarcasm of all critics of democracy, they would produce the much more serious effect of crippling the successful working of parliamentary institutions.

The alternative vote not a solution of the problem of three-cornered contests.]

Neither second ballots nor preferential voting can solve the problem of three parties seeking representation. They may preserve the outward form of the distinguishing characteristic of the present system—that each successful candidate should secure the support of the majority of the electors voting—but this apparent conformity to the requirements of majority representation is only secured at the cost of destroying the sincerity of the parliamentary system and of rendering the composition of the House of Commons still more unstable than it is to-day. In England the competition of the three parties is most pronounced in the industrial areas, and Mr. Winston Churchill, apparently recognizing the futility of the alternative vote as a solution of the new difficulty, had good grounds for his suggestion that electoral reformers should concentrate their minds upon the proportional representation of the great cities.[7] For proportional representation attacks the new problem on entirely different lines. It provides for the realization of the essentially democratic principle, that the various sections of political' opinion are entitled to representation in proportion to their respective strengths, and that such representation should be independent of the action of other parties. Once this democratic principle is admitted we are in view of the only effective solution of the problem of three-cornered fights—a solution which not only solves this particular difficulty, but meets those serious defects of our electoral system to which attention has been directed in the two preceding chapters. "The theory of Government by party," says Professor Nanson of Melbourne, "is to find the popular mind by the issue of a number of contests between the 'ins' and the 'outs.' But owing to the multiplicity of political issues, this theory is now no more tenable than is the theory that every question can be answered by a plain 'yes' or 'no.' ... We require a system capable of finding the mind of the people on more than one issue. With such a system all the difficulties caused at present by the existence of three parties disappear. Instead of being a hindrance three parties will be a help. For each will help to organize public opinion, and so enable the mind of the public on important issues to be more definitely and clearly ascertained."

[Footnote 1: The Albany Review, October 1907.]

[Footnote 2: Reports on the Second Ballot at Elections in Foreign Countries. Miscellaneous. No. 2. 1908. (Cd. 3875.)]

[Footnote 2: La Representation Proportionnelle en Belgique, p. 7.]

[Footnote 3: An illuminating passage occurs in M. Guyot's article on "The French Senate and Chamber of Deputies," in The Contemporary Review, February 1910:—

"A deputy is only elected for four years, and almost on the morrow he becomes again a candidate. If he has been elected at the second ballot, with a rallying of the minority of electors, who have only voted for him as better than nothing, and who can desert him at the next elections, his position is very uncertain. Universal suffrage results in many constituencies in great instability, and it is threatening especially for the men who having had power have been obliged to act, and in acting have dispersed certain illusions which they had perhaps entertained when candidates, and have thus given offence.... Though one be an ex-Minister one is none the less a man. The greater number of men—not only ex-Ministers but men who have any reputation in Parliament—have sought to migrate from the Palais Bourbon to the Luxemburg. The result is that the Chamber of Deputies has not ceased to suffer from a species of inverse selection. No body could retain its vigour under such a system. The most experienced men have left; the composition of the Chamber of Deputies has grown steadily weaker and weaker."]

[Footnote 4: In Australia the system is known as the contingent or preferentinal vote. In recent years the phrase "alternative vote" has been employed in England, and was adopted by the Royal Commission on Electoral Systems as a means of distinguishing the use of the transferable vote in single-member constituencies from its use in multi-member constituencies for the purpose of securing proportional representation.]

[Footnote 5: The regulations as to counting the votes contained in the Schedule to the Bill were based upon those in Lord Courtney's Municipal Representation Bill (see Appendix VI.), the practical application of which is described in Chapter VII.]

[Footnote 6: Mr. Crawshay-Williams introduced a further Bill (based on that of Mr. Robertson) in 1910. This Bill, in its final form, was made applicable, in accordance with the recommendation of the Royal Commissions on Electoral Systems, to single-member constituencies only.]

[Footnote 7: Reply to deputation of Manchester Liberal Federation, 23 May 1909.]



CHAPTER VI

PROPORTIONAL REPRESENTATION

"Celui-ci tuera celui-la. Voila la formula du scrutin d'arrondissement.

"Ceux-ci tueront ceux-la. Voila la formule du serutin de liste sans la representation proportionnelle.

"Ceux-ci et ceux-la auront leur juste part. Voila la formule du scrutin de liste avec la representation proportionnelle."—J. JAURES

It cannot be a matter for surprise that the methods of election adopted in the early stages of representative institutions fail to respond to the needs of the more complex political conditions of highly civilized communities. The movement in favour of improved electoral methods is in keeping with the advances made in all other human institutions. We no longer travel by stage-coach nor read by rush-light. We cross the Atlantic with a certainty and an ease unknown and undreamt of a little while ago. Means of intercommunication, the press, the mail, the telegraph, the telephone have developed marvellously in response to modern requirements. This continuous adaptation is the law of existence and, in view of modern political conditions we cannot permanently refuse to adapt our electoral methods to the more perfect organization of a progressive democracy. By cumulative pressure the evils set forth in the preceding chapters can have but one result; they will compel English statesmen, as they have compelled or are compelling Continental statesmen, to devise an effective remedy; and although individual politicians may resist and retard the advent of reformed methods, the demand for better representative institutions will in the end overcome all such resistance.

The essential features of a sound electoral method.]

What then are the requirements of a satisfactory electoral method? The evils to be remedied must yield the clue. Our present system—exclusive majority representation—has often, as we have seen, resulted in a gross exaggeration of the majority, sometimes in the total suppression of the minority; and, on other occasions, in the return of a majority of representatives by a minority of the electors. These evils have happened when only two parties have been seeking representation; when a third party enters the political arena the system completely breaks down, and all efforts to restore "majority" representation by a system of second ballots have proved an absolute failure. The attempts made in the past to secure the special representation of minorities, though most successful in many respects, have been of an empirical character, and have dealt with the problem in a very partial way. Yet it is not difficult to find a solution for all these problems which is at the same time satisfactory and effective. It is only necessary to return to the first principles of democracy, to keep steadily in view the meaning of that self-government which we desire to achieve through representative institutions. Self-government can only be realized when every section of the community through its own representatives can give expression to its needs in the assembly which is representative of the nation and which derives all its authority from the fact that it is so representative. This assembly acts in the name of the nation; its decisions are said to embody the national will. But if any considerable section of the nation is deprived, from whatever cause, of representation in the House of Commons, in what sense can it be said that its decisions give expression to the national will? The new electoral conditions force us, willingly or unwillingly, to the conclusion that no satisfactory solution can be reached until effect is given to Mill's fundamental principle of democracy—that the various sections of political opinion should be represented in the legislative chamber in proportion to their strength. Only in the fulfilment of that condition can we escape from the evils of the existing system and at the same time do justice to the claims of three organized parties to representation within the House of Commons.

Constituencies returning several members.

It is now no longer possible to accept Mill's declaration as theoretically perfect and then to dismiss it as wholly impracticable. If the political conditions are such that the proportionate representation of parties is the only satisfactory solution of our electoral difficulties, it becomes the duty of statesmen to find some way by which practical effect can be given to Mill's formula. There was doubtless some excuse for the cry of impracticability when, in launching in 1857 his proposals for proportional representation, Thomas Hare suggested that the whole kingdom should form a single constituency. This suggestion raised a barrier of prejudice against all proposals for proportional representation, which only to-day is being broken down, and led to a refusal to consider seriously any attempt to secure an amelioration of existing methods along more modest lines. Nevertheless, it must be admitted that the first step in the direction of realizing true representation must be the enlargement of our present electoral areas. So long as single-member constituencies are retained elections must necessarily take the form of a struggle for the whole of the representation allotted to the constituency. There is but one prize—a prize which is indivisible—and the proportional distribution of that prize is impossible. For a system of proportional representation the first requirement is the formation of constituencies returning several members. These electoral areas need not be formed in an arbitrary manner. Familiar divisions of the country, such as large towns, counties or parts of counties, may be treated as single constituencies. Glasgow, Manchester, Birmingham, Sheffield, Leeds would form constituencies in themselves. Counties which are large enough to return at least five members might also be treated as electoral areas, whilst the smallest counties would be grouped and the larger counties, if necessary, subdivided.

The proportional representation of the electors.

With such constituencies it would be possible to approximate to a true representation of the electors. Birmingham, which may be taken for purposes of illustration, returns seven members to the House of Commons, one for each of its seven divisions. The Unionists being in a majority in each of these seven divisions, are enabled to secure the whole of the representation allotted to the city, although there is a large minority of non-Unionists. If Birmingham were treated as a single constituency, and if the electors were divided as follows: Unionists, 40,000; Liberals, 20,000; Labour, 10,000, then it is obvious that any just system of representation would enable the Unionists, Liberal and Labour electors to obtain four, two, and one members respectively. Birmingham would then be represented accurately and fairly within the House of Commons; and if each large area was so represented we should, in this way, be able to build up a House of Commons which would reflect in true proportions the political opinions of the country. The undoubted fairness of such a system of representation will appeal with even more force if consideration is given to the grounds on which seven representatives are now allotted to a town of the size of Birmingham. Did Birmingham contain only 40,000 electors, all of whom were Unionists, it would only be entitled to four representatives in Parliament. The presence of a large number of electors who are not Unionists brings, however, the total electorate to 70,000, and Birmingham is granted representation on the basis of this total. Thus the additional representation, granted because of the presence of a large minority of non-Unionist electors, takes the form of additional Unionist members. The minority under the present system is not only disfranchised but penalised; the representation which is due to them is given to their opponents.

But it is not difficult to devise a scheme of proportional representation which should ensure that the electors of Birmingham and other large towns, and also of the various counties, should be truly represented within the House of Commons. Of this fact the recent history of electoral legislation on the Continent and in the Colonies furnishes incontrovertible proofs. Proportional representation has been embodied in the laws of several countries, and these laws work with perfect smoothness.

Experience in Denmark.

The first application of the principle took place in Denmark so long ago as 1855, two years before the publication of Mr. Hare's scheme, when M. Andrae, a Danish Minister of great eminence and ability, introduced it into the new Constitution promulgated in that year. The system of proportional representation was retained through the constitutional changes of 1863 and 1866, though, it should be added, the extent of its application was limited to the election of members of the Upper House. The citizens of each constituency, voting in two classes, choose by the ordinary method of voting an equal number of representatives. These representatives constitute an electoral college, the members of which proceed to the election of representatives of the constituency according to the method of proportional representation. This limited application of proportional representation still remains in force, and in recent years the principle has received further and increasing recognition. Parliamentary committees and committees of the municipalities of Copenhagen are chosen by a proportional method. The principle was applied in 1903 to the elections of the Congregational councils, but its most notable extension was effected in 1908, when the system was applied to all municipal elections, the first elections taking place in March 1909.

Switzerland

It will be seen that even in Denmark there was a considerable lapse of time between the limited application adopted in 1855 and its extension to elections of a more popular kind in recent years; and outside Denmark, although societies advocating the new principles were founded in England, France, Belgium, and Switzerland, proportional representation did not succeed in finding its way very readily to the statute book. It was not until 1890 that the first step was taken which has resulted in so rapid an extension of the system. The evils arising from the majority method of election had become so acute in the Swiss canton of Ticino[1] that proportional representation was adopted as a means of pacification. The elections in March 1889 resulted in the return of seventy-seven Conservative deputies by 12,783 votes, whilst the Liberals, with 12,166 votes, were only able to obtain thirty-five representatives. The Liberals alleged that this unfair result was due to a gerrymandering of the constituencies, and demanded a revision of the Constitution. The Conservative Government declining to take the necessary steps for this purpose, a revolution broke out in Bellinzona, in the course of which one of the members of the Government was killed and his colleagues arrested and imprisoned. The Federal Council intervened and sent its representative, Colonel Kuenzli, who recommended the adoption of proportional representation. After some hesitancy the party leaders agreed, and the Cantonal Council passed a law (5 December 1890), providing for the election by a system of proportional representation of a Constituent Assembly for the purpose of revising the Constitution. The suspicions of the Liberals were not, however, fully allayed and, thinking that they were again being duped, they decided on the eve of the election of the Assembly to abstain. This decision was adhered to, and as a result the first assembly in Ticino elected under the proportional system consisted of Conservatives only. The Conservatives remained faithful to the agreement entered into with the Liberals and voted the law of 9 February 1891, introducing proportional representation into the cantonal constitution and applying it to the elections for the Cantonal Council, Constituent Assemblies and municipalities. The law was approved by popular vote in the following March, and the system has since retained its place in the constitution of the canton[2]. The immediate object in view—the pacification of the canton—was completely attained and its success has led to its adoption in other cantons. It is now in force in Neuchatel, Geneva, Solothurn, Zug, Schwyz, Bale City, Lucerne and St. Gall, and also (for municipal elections) in Berne, Fribourg, and Valais, whilst there is an active and growing demand for its application to the Federal elections. The progress of public opinion in this respect has been tested by means of the Referendum in 1900 and 1910. On the first occasion 169,000 voters supported the extension to Federal elections, and 247,000 opposed it. In 1910 the number of voters in favour of the proposal had increased by 70,000, while the opposition had increased by only 15,000, and the adoption of proportional representation for Federal elections was defeated by the narrow margin of 23,000 votes in a total poll of half a million. At the same time twelve out of the twenty-two cantons approved of the extension, and it is generally agreed that the ultimate triumph of the proportional principle cannot long be delayed.

The need for proportional representation was particularly felt in the canton of Geneva, where religious differences often form the dividing line between parties. The canton is divided into three constituencies; one for the town of Geneva, one for that part of the canton on the right bank, and one for that on the left bank of the Lake and of the Rhone. With the scrutin de liste (the former method of election) the minority in each constituency was completely crushed. The Protestants of the right bank were deprived of all representation; the Catholics of the town obtained a few deputies as an act of grace on the part of the majority. In 1872, when the affairs of the Catholic church were being discussed, the Radicals and Independents succeeded in excluding from the Council all who were most directly affected by the question of the day. The proportional system was introduced in 1892, and as the election of members of the Federal Council was still conducted according to the old system the working of the two methods could be readily compared. "The elections for the cantonal councils in November 1892," wrote M. Naville, "were keenly fought, but calm; no recriminations followed, and political life pursued a normal course.... On the other hand, the Federal elections in October 1893 were riotous, blows being exchanged. Exclusive majority representation artificially creates disturbances.... Proportional representation introduces a pacifying element into all political struggles."

Belgium.

The introduction of a complete scheme of proportional representation into Belgium was also rendered necessary by the intolerable position arising from the former methods of election. The rapid growth of the Socialist Party with a distinct organization created a situation which, as already explained, was in no way relieved by the system of second ballots in force. Indeed, the coalitions at the second ballots not only discredited the system but greatly embittered the relations between the various parties. "In 1899," says Count Goblet d'Alviella, "Belgium was on the eve of a revolution—a revolution which was only avoided by the immediate and complete introduction of proportional representation into parliamentary elections." This, however, was not the first trial of proportional representation in Belgium, for Belgium, like Switzerland, affords an example of the gradual but certain extension of the new method of election. In 1894 proportional representation had been applied partially and tentatively to the larger municipal councils, and although this application was of a partial character it achieved a considerable measure of success. M. Braun, the Burgomaster of Ghent, speaking in May 1899, described its results in the following terms:—

"During the four years that proportional representation has been applied to the communal elections of Ghent, every one has been able to appreciate the happy effects of the reform. Everybody recognizes that, far from being endangered, the material prosperity of the city has increased, and that the ameliorating and pacifying effects of the altered electoral method have even exceeded the expectations and hopes of its advocates." [3]

The system of proportional representation adopted for the parliamentary elections was much more complete, and so great has been its success that there has arisen a strong demand for its introduction into the elections for the provincial councils in which the old majority system, with second ballots, is still used. The parliamentary elections in May 1908 were followed by the provincial elections in the ensuing month, and thus a favourable opportunity was presented of contrasting the working of the two systems. The grossly unfair results of the provincial elections drew forth from many journals most caustic criticism. Le Peuple expressed the hope that these provincial elections would be the last instance of the use of the majority system in Belgium. "Is it not," it proceeded, "absurd, stupid, detestable that the provincial councils are alone excluded from the system of proportional representation? Once for all we must have done with this jumble of confusion, dishonesty, and corruption." The Etoile Belge declared that "One thing is certain, the provincial electoral system can no longer be maintained without exposing us to the laughter of Europe. To apply one system of proportional representation to the parliamentary elections, another to municipal elections, and to maintain the majority system for the provincial elections, is really too absurd. For once we agree with Le Peuple and join our hopes and wishes to theirs." That these comments were fully justified a few examples will show. In the province of Limbourg the forty-eight seats on the provincial council were all obtained by the Catholics, whereas in the parliamentary elections of the previous month the Liberals, owing to the proportional system, were able to obtain two seats out of six. In the "Agglomeration Bruxelloise" no Catholic and only five Socialists were elected, although the Liberals numbered but a few more than a third of the voters. The provincial elections of former years afford further illustration. In 1898 at Ghent the Liberals of the first canton defeated the Socialists at the second ballots with the help of the Catholics, in the second canton they defeated the Catholics with the help of the Socialists, while in the third canton they were themselves defeated by the Catholics, who were assisted by the Socialists. In the same year at Brussels, where a second ballot took place in each of the five cantons, the Liberal minority captured every one of the forty-four seats. Sir Arthur Hardinge pointed out in his Report on the working of the Second Ballots in Belgium, that it was the failure of this electoral method that rendered a proportional system in parliamentary elections an absolute necessity; its failure in the provincial elections will result in its abolition from these also. No more convincing evidence of the satisfactory working of the proportional system can be given than this demand for its extension, the latest example of which in Belgium is its application by a new law passed in 1909 to the election of the Conseils de Prud'hommes.

German States.

Whilst the adoption of proportional representation in Switzerland and in Belgium was due to the pressure of particular circumstances, the marked success of the new method has not only resulted in its extension in those countries, it has also had a pronounced influence upon public opinion in neighbouring countries. The kingdoms of Southern Germany are following the example of the Swiss cantons. Wuertemberg, in the new constitution adopted in 1906, decided that the seats set free by the removal of the "privileged" members of the Lower House should be filled by proportional representation. Legislative proposals have since been discussed in Saxony, and in May 1910 a vigorous debate took place in the Bavarian Parliament, in the course of which Dr. Mueller declared that the advocates of the reform would not rest "until this unjust electoral system, this bulwark of short-sighted injustice and ill-omened party spirit, is set aside in the higher interests of justice and of civil and religious freedom." The principle has received a recognition even more general in character, for a ministerial decree issued in June 1901, relative to the associated committees of employers and workmen, enabled these bodies, if they so chose, to elect their members in accordance with the principle of proportional representation. Some sixteen towns, including Frankfort-On-Main, Munich, Carlsruhe, Fribourg, Mannheim, &c., availed themselves of the privilege, and the results have been most satisfactory. Much greater interest has been taken in the elections. In Carlsruhe, for instance, the number of voters increased from 1103 in 1897 to 3546 in 1903.

France

Similarly, the great success of the Belgian legislation gave birth to a fresh and more powerful movement in France. Founded in 1901, under the presidency of M. Yves Guyot, the Ligue pour la Representation Proportionnelle enlisted the support of deputies drawn from all political parties. The Electoral Reform group within the Chamber of Deputies during the Parliament 1906-10 consisted of over two hundred members, and, under the auspices of this group large and enthusiastic meetings were held in the great towns. The reform has the support of many leading newspapers, and the authoritative reports of the French Parliamentary Committee, la Commission du Suffrage Universel, contain strong recommendations in favour of the adoption of proportional representation. The first of these reports prepared in 1905 by M. Chas. Benoist[4] contains an admirable statement of the case for the reform, a plea which is powerfully reinforced in the report prepared two years later by M. Etienne Flandin.[5] The Bill recommended in this latter report was discussed in the French Chamber of Deputies in October 1909. The first clause of the Bill read as follows: "The members of the Chamber of Deputies shall be elected by the scrutin de liste according to the rules for proportional representation." The first portion of this clause—the members of the Chamber of Deputies shall be elected by scrutin de liste—was carried by 379 votes against 142, or a majority of 237. The second portion—according to the rules for proportional representation—was carried by 281 votes to 235, or a majority of 46. The Prime Minister, M. Briand, urged by many of his Radical supporters, who were unwilling to forego the advantages which they obtained from the existing system, then made the question one of confidence in the Government, and the whole clause, when put to the final vote, was defeated by 291 votes to 225. A noteworthy feature of these divisions was the size of the majority by which the system of single-member constituencies was condemned. At the General Election in April 1910 no fewer than 315 Deputies were returned pledged to the reform. M. Briand at once introduced a Bill which, however, did not fully meet the demands of the reformers, and the Commission du Suffrage Universel made important modifications in it with a view to securing more completely the proportional representation of all political parties within the country. On the fall of M. Briand in February 1911, the government of M. Monis announced its intention of supporting the amended scheme. The success of the movement, commenced in 1901 is now, after a decade of active effort, no longer open to doubt.

Holland

Holland, too, has felt the influence of the legislation of its neighbour. A constitutional commission, appointed by the Dutch Government, reported in favour of amending the fundamental law so as to render possible the adoption of proportional representation. The recommendations of this Commission were embodied by the Government in Bills presented to the States General in 1907, and although the proposals were subsequently withdrawn, the reform has the support of many of the leading statesmen, and a favourable report is anticipated from the new Commission to which the question of reform has been referred.

Finland.

In the North of Europe an equally successful and, in some respects, an independent movement in favour of true representation has taken place. In an excellent little pamphlet, published at Helsingfors,[6] it is stated that during those calamitous years between the coup d'etat of 1899 and the restoration of the constitution in 1906, there arose in Finland the conviction that only a democratic reform of its political institutions would afford a sufficient guarantee for the maintenance of its internal independence. The fruits of that conviction were seen in the draft of the new constitution for the Diet prepared by a committee appointed by the Finnish Government. Provision was made for the adoption both of universal suffrage and proportional representation. The report adds that the four Estates of the Diet, satisfied that proportional representation would ensure the just representation of all parties, willingly accepted the proposals for universal suffrage, and also agreed that henceforth the Diet should consist of but one chamber. Finland thus found herself, when the new constitution was granted, in the possession of an electoral system as democratic as any in the world.[7]

Sweden.

In Sweden a long and arduous struggle took place over the reform of the franchise. The Liberals and Socialists demanded that less weight should be given to the possession of property. The Conservatives resisted the demand. The adoption of proportional representation as a possible way out was proposed in 1902, and from that date the fight assumed another aspect. "The method of voting," wrote Major von Heidenstam, part author of the proposals embodied in the new law, "took from the beginning a very prominent place, strange to say the most prominent down to the last few months before the chief battle. We who went in for proportional representation had a very hard struggle for the first five years, but we won at last." The victory was complete; proportional representation was accepted for both Chambers of the Riksdag, for the committees selected by these Chambers, for County Councils and for Town Councils. When the final adoption of the reform Bills was voted in 1909 they were carried by very large majorities; in the first Chamber only 19 out of 141, and in the second Chamber only 53 out of 225, recorded an adverse vote.[8]

Australasia.

In this remarkable outburst in favour of proportional representation English-speaking countries are taking their part. Inspired by the late Catherine Helen Spence, an untiring advocate of the reform, the Effective Voting League has carried on an active campaign in Australasia. Legislative proposals for proportional representation have been discussed in recent years by the Commonwealth Parliament, and also by the Parliaments of Victoria, South Australia and West Australia. Although these measures have not become law, the work of Miss Spence and her colleagues has gained considerable support. Mr. Deakin has openly acknowledged his approval, whilst the results of recent elections, and more particularly that of the election in 1910 for the Commonwealth Senate, have increased the demand for reform. Proportional representation, too, is meeting with increasing sympathy in New Zealand where the system of second ballots, adopted in 1908, has failed to give satisfaction. In Tasmania the movement has made much greater headway. An Act was passed in 1896 applying proportional representation to the urban districts of Hobart and Launceston, but although this Act was an acknowledged success so far as the representation of these two towns were concerned, the differentiation between the voting methods applied to the town and country districts gave rise to dissatisfaction, and the measure was withdrawn in 1901. But when once the benefits of proportional representation had been felt its re-introduction in a more complete form was not long delayed. In 1907 a new Act was passed applying equally to town or country. The State is now divided into five electoral districts, and the six members allotted to each district are elected by the proportional method. The first elections under the new law took place in April 1909, and the result has met with general approval.