Proportional Representation: A Study in Methods of Election
Chapter 24
THE KEY TO ELECTORAL AND CONSTITUTIONAL REFORM
"De la maniere de regler le suffrage depend la ruine ou le salut des Etats."--MONTESQUIEU
_Electoral problems awaiting solution._
The Liberal, Conservative, and Labour parties are all agreed that a large measure of electoral reform is long overdue, but hitherto the various parties have contended only for such reforms as would strengthen their own parliamentary position. Liberal and Labour politicians, looking at the inequality in the voting power of electors, have demanded a reform of the franchise; they urge that every man should have one vote and no more. The Conservative party, looking at the inequalities in the size of constituencies, have demanded a redistribution of seats on the ground that all votes should be of equal value. Liberals, again, feeling the difficulties which have attended the emergence of third-party candidatures in the constituencies, ask for a reform in the method of voting so as to ensure that the member elected for any constituency shall represent a majority of the citizens. Apart from the question of the enfranchisement of women, which involves considerations of a different order, these are the three electoral problems with which public opinion has been chiefly concerned.
The efforts of parties to give effect to the reforms in which they have been more particularly interested have so far ended in failure. In 1905 Mr. Balfour introduced a Bill for the redistribution of seats, unaccompanied by any reform of the franchise. The measure was met with the cry of "gerrymander!" and its disappearance with the fall of the Government was regretted by few. In 1907 the Liberal Government attempted to deal with the franchise problem, apart from any scheme of redistribution. It endeavoured in Mr. Harcourt's Plural Voting Bill, a highly complex measure, to give effect to the principle of "one man, one vote." This Bill was strongly opposed on the ground that the reform was partial in character. If, said the opponents of the measure, it is unfair that one elector should have twelve votes whilst another elector has but one, it is equally unfair that the vote of an elector in one constituency should be twelve times as valuable as the vote of an elector in another constituency. The justice of the argument must be admitted, and explains why the rejection of the Plural Voting Bill by the House of Lords aroused comparatively little public feeling. Yet the rejection of this Bill has focussed attention upon the deficiencies of our franchise laws, and the eyes of all politicians are turning towards that more comprehensive measure of electoral reform which cannot be indefinitely postponed. Such a measure has been categorically promised by Mr. Asquith on more than one occasion. So far back as 1908, soon after his accession to the Premiership,[1] he made the following public declaration: "I regard it as a duty, and indeed as a binding obligation on the part of the Government, that before this Parliament comes to an end they should submit a really effective scheme for the reform of our electoral system."
_The simplification of the franchise._
What are the lines on which a really effective scheme can be framed? The fate of the partial measures already referred to is at least an indication of the difficulties which will attend any attempt to carry an incomplete scheme. It may be assumed that an effective scheme must deal with the three problems named: franchise (including registration), redistribution, and three-cornered contests. Each of these factors must be dealt with as simply as a due recognition of the problem to be solved will allow. The complexity of Mr. Harcourt's Plural Voting Bill was due to the fact that we possess no less than twenty[2] different franchises. But the remedy is easy. "If," said the late Sir Charles Dilke, "they wanted to cheapen the cost, to remove the disgrace from this country of having registration more full of fraud and error than anywhere else, they could only do so by some simple franchise. All registration reform was condemned to failure until they made up their minds on a simple and easy basis for the franchise, sufficiently wide to enable them to absorb all existing franchises." Such a simple franchise is to be found in manhood suffrage, which would admit of the easy transfer of electors' names from the register of one electoral division to another. The chief objection to this solution, which arises from the fear that the most numerous class in the country may monopolise representation, may be met by linking the adoption of a simple franchise with a system of election which shall give due representation to minorities.
_Redistribution._
Redistribution must be treated with like boldness, but before considering the principle on which this reform must be based, it would be well to give some indication of its urgency. Here are the figures of four of the largest and four of the smallest English constituencies as given in the Parliamentary Return of 1911:-- *** Constituency. Electors. Constituency. Electors. Romford (Essex) 55,951 Durham. 2,698 Walthamstow (Essex) 42,029 Bury St. Edmunds 2,878 Wandsworth 39,821 Whitehaven 2,989 Harrow (Middlesex) 38,865 St. George's, Tower Hamlets 3,252
_Should be automatic._
It will be observed that an elector in Durham has twenty times the political power of an elector in the Romford Division of Essex. Nor are these discrepancies confined to England. There are great divergencies between the electorates of individual constituencies in Scotland and Ireland, and any measure of redistribution which attempted to deal effectively with these would necessarily have to be of a far-reaching character. Even were it possible to effect a readjustment by the creation of parliamentary areas containing an equal number of electors, so rapid are the changes in the electorate that the scheme would be out of date almost before it came into force. Mr. Ellis T. Powell has published a valuable table entitled "the process of electoral evolution,"[3] in which he has arranged the constituencies in the order of their size as measured by the number of electors who were on the registers in 1886, and again in 1906. The table shows how remarkable has been the change in their relative importance. The rapidity of the change is still further indicated by a comparison based upon the 1908 register. Any one who has the curiosity to count the number of constituencies which retained the same position on the list both in 1906 and 1908 will find this to be the case in nineteen constituencies only out of a total of 481. So great, indeed, has been the change since 1901, the date of the last census, that no satisfactory scheme of redistribution could be framed upon the population figures of that year. It would seem that the only satisfactory principle upon which the problem can be solved is that of an automatic redistribution of seats on the completion of every census, but the difficulties associated with such a solution, if the present system of single-member constituencies is retained, are so overwhelming as to render it almost inadmissible. True, the South African Constitution provides for the automatic redistribution of seats after every quinquennial census,[4] and the Canadian Constitution contains a similar provision, but the inconveniences attaching to a rearrangement of boundaries are not so great in new countries as those which obtain in an established country. Moreover, as time goes on, the inconveniences associated with rapid changes in boundaries will be felt more and more both in Canada and in South Africa. For local authorities[5] rightly complain of the difficulties which arise from the creation of different areas for different purposes and the consequent overlapping of boundaries, and these difficulties would increase were fresh parliamentary divisions created every ten years. The problem which would be involved in the creation of new parliamentary divisions for London is such as to render a satisfactory scheme almost impossible. Apart, however, from these considerations, the difficulties of another kind attendant upon the creation of new constituencies are so great that it is quite easy to understand the unwillingness of the leaders of both political parties to embark upon schemes of redistribution. The influence of boundaries upon the political fortunes of parties is so well known that any rearrangement, whether in the metropolis or in the large towns, would probably be looked upon with very grave suspicion, and the more so that in several towns party organizations have already endeavoured to obtain the maximum of party advantage under existing conditions.
_Secures neither one vote, one value nor true representation._ Further, it has been proved beyond question that a redistribution of seats will, if single-member constituencies are retained, fail to accomplish the end which its advocates have in view, namely, one vote one value. For redistribution can only secure equality in the size of electoral districts, and this is not the same as equality in the value of votes. With equal electoral districts it would still be possible in two adjoining constituencies for one member to be returned by a large majority and the other by a small majority. In Wales it might still happen that a Conservative vote would be valueless for the purpose of obtaining representation. Equality in vote value is only secured when the votes of electors of all parties are equally effective. This can only happen when the representation of parties is brought into agreement with their voting strength.
The Royal Commission on Electoral Systems entered very carefully into the probable effect of redistribution upon the representation of parties within the House of Commons, and came to the conclusion that, so far "as facts can be adduced to test it, the theory that the varying size of constituencies accounts for the exaggeration of majorities falls to the ground." This conclusion--and the Commission could hardly have come to any other--is in agreement with the opinions expressed both by Mr. S. Rosenbaum, of the Royal Statistical Society,[6] and by Mr. J. Rooke Corbett, of the Manchester Statistical Society.[7] The following summary of the results of Mr. Corbett's analyses of the eight General Elections 1885-1910 shows conclusively that redistribution would fail to remedy the inequalities in representation arising from a system of single-member constituencies:
GENERAL ELECTIONS, 1885-1910
Majority Majority Year of of seats under system Majority under Election Party. actually of equal a proportional gained. electorates. system. 1885 Liberal 158 178 86 Liberal 1886 Conservative 104 102 8 Liberal 1892 Liberal 44 46 34 Liberal 1895 Conservative 150 172 12 Conservative 1900 Conservative 134 150 16 Conservative 1906 Liberal 356 362 104 Liberal 1910(Jan.) Liberal 124 136 66 Liberal 1910(Dec.) Liberal 126 122 38 Liberal
"It is sometimes said," states Mr. Corbett, "that if the single-member constituencies were made equal in size these inequalities of representation would disappear. It is difficult to understand how any one with even the most elementary knowledge of the facts can support such a proposition. An examination of the foregoing summary will show that no readjustment of the electoral constituencies would do much to remedy the enormous inequalities which occur at present. In fact strict equalization of the constituencies would be as likely to make matters worse as to make them better. Thus, in the year 1885 the Liberal majority of 158, which under a proportional system would have been 86, by a system of equal electorates would have been transformed into a majority of 178; in the following year a Conservative majority of 104, which, with a proportional system, would have been a Liberal majority of 8, would under a system of equal electorates have been transformed into a Conservative majority of 102." Mr. Rosenbaum states: "I am firmly persuaded that it is not possible for redistribution alone to effect those particular reforms which the advocates of proportional representation urge.... Proportional representation would secure in the House of Commons a representation of each party in strict arithmetical proportion to the number of its supporters in the country. Redistribution can remove anomalies due to over-representation in one part and under-representation in another part of the country. So far as the over-representation in one area is accompanied by an excessive proportion of members of one party, and the under-representation in another area is accompanied by a deficiency of members of the opposite party, redistribution might have some counterbalancing results. There is, however, no real security that redistribution by itself might not aggravate rather than mitigate this particular trouble."
_The problem simplified by proportional representation._
It will have been observed that the difficulties of redistribution arise from the system of single-member constituencies, and it is this which also renders all schemes useless for the purpose of securing equality in the value of votes. An effective and simple solution of all difficulties is available. Abandon the system of single-member constituencies with their ever-changing boundaries, and treat the natural divisions of the country (its counties, large towns, &c.) as permanent constituencies with representation varying with the rise or fall of their population. This is the scheme of redistribution required by a system of proportional representation, and its adoption would simplify the most difficult of all the problems of electoral reform. It would make possible that automatic redistribution of seats, which must be an essential feature of any satisfactory scheme of redistribution, without involving these alterations of boundaries which, in addition to their other disadvantages and even dangers, interfere so seriously with administrative efficiency. With such a system the areas for local or parliamentary purposes might easily be brought into agreement. Already "we have strong county patriotism fostered by tradition, by ecclesiastical and judicial affairs, county council government, county territorial organization, and even county cricket and football; to have, therefore, county electoral areas would be at once popular and intelligible to all; besides, it would be a reversion to an old tradition";[8] and if the large towns were made parliamentary constituencies this also would be a reversion to the conditions which existed before 1885. It would be infinitely easier to add representatives to or take them away from such electoral areas than it would be to redivide the boroughs and counties for the purpose of creating new constituencies.
Commenting on the work of the Delimitation Commission, to which was entrusted the duty of creating the new constituencies for the South African Assembly and Provincial Councils, the Secretary, in a letter to the author, says: "The task set the Commission proved exceedingly difficult. While it was, so to speak, imperative to give due consideration to all the principles enjoined by the Act, the great object naturally was the framing of constituencies both for the Union Assembly and for the Provincial Councils which would be able to send representatives who, in turn, would reflect the will of the various sections of the people. The conditions enjoined by the Act made it very difficult to produce schemes which could on all hands be considered entirely satisfactory.... Good as the result is, there is no question that had the first recommendation of the South African Convention in favour of proportional representation been adopted, the work of the Commission would not only have been much simplified, but the chances of framing constituencies with representatives forming a true mirror of the various sections of the people would have been increased by more than fifty per cent.... If there had been any doubt in my mind my work on this Commission has removed that doubt, and proved to me that the only remedy for our various electoral ills is a system of proportional representation." This considered testimony, from one who has been immersed in the practical details of redistribution, is of great value, but it can occasion no surprise, for proportional representation admits of automatic redistribution of seats, provides for the permanence of boundaries, renders gerrymandering impossible, and, above all, secures equality in the value of votes.
_The case of Ireland._ There is one special difficulty,[9] however, which must be faced in the consideration of any scheme of redistribution for the United Kingdom--the number of representatives to be allotted to Ireland. The permanent over-representation of any one part of a kingdom united for common purposes cannot easily be defended, but the South African Constitution furnishes an example of a larger representation being accorded temporarily to the smaller states for the purpose of facilitating the union of all; whilst in South Africa, Australia, and the United States the separate states or provinces have equal representation, irrespective of size, in the Senate. If the continued over-representation of Ireland would in any way facilitate the process of the unification of the United Kingdom, that in itself would be a very powerful and sufficient reason for maintaining the number of Irish members at its present level. A system of proportional representation might simplify the solution of this particular difficulty, for the over-representation of Ireland would not have the same disturbing effect upon the composition of the House of Commons if the different divisions of political opinions within Ireland obtained their fair share of representation. For proportional representation would produce a very important modification of the electoral conditions within Ireland. According to Mr. J. Rooke Corbett, the Irish Unionists who, at the General Election of 1906, obtained 18 representatives, were entitled to 34. But that is not the only change that would take place. There would result a softening of those racial divisions which are now the chief characteristic of Irish representation. Moderate opinion would be encouraged to take a more active part in elections and to seek representation. Nor can it be said that the political conditions of Ireland are such as to render proportional representation within Ireland either impracticable or nugatory in its effect. Mr. Archibald E. Dobbs, High Sheriff of county Antrim, has framed a scheme with special reference to Irish conditions[10], and Lord MacDonnell, who was intimately associated with the details of the Irish Council Bill of 1907, has said: "He made the subject the matter of as close a study as he could at the time, and everything he read more fully satisfied him of the great desirability of the system. He felt that it was more needed in Ireland than in any other part of the British Empire, because, although for the purpose of general politics the division into Nationalists and Unionists could be defended, for the purpose he had in view--the internal administration of Ireland--it was essential that all views, not only the Nationalists and the Unionists, but the great political school of thought under the name of the old Whigs should also be represented. The results of his labours perhaps it would not be discreet for him to disclose, but he was quite satisfied of the practicability in Ireland of a scheme of proportional representation[11]."
_Three-cornered contests._
But even if the Electoral Reform Bill provided for a simplification of the franchise and a redistribution of seats, yet such a measure could not be described as a complete and effective scheme of reform. The Bill must provide a solution for the further problem arising from three-cornered contests, which have greatly increased in number in recent elections. On what principle is this difficulty to be solved? Formerly there was a strong demand for the second ballot, but its defects have been so constantly exposed that the remedy more generally advocated is the one recommended by the Royal Commission on Electoral Systems, viz., the adoption of the alternative vote (the transferable vote in single-member constituencies). This proposal, however, ignores the real difficulty, which is found in the fact that three parties, and not two, are now seeking representation. Three-cornered contests have, so far, affected adversely the fortunes of the Liberal Party; and the alternative vote, whilst tending, at least temporarily, to redress the situation, does so without providing any adequate guarantee for the fair representation of other parties. Were this remedy adopted it may be assumed that Liberal candidates would be nominated in those constituencies which are now represented by members of the Labour Party, and at least there would be a cessation of the process of withdrawing Liberal candidates from other constituencies ear-marked by the Labour Party. Were all these constituencies contested by the three parties it might easily happen that the smallest party would obtain no representation whatever. Conservative electors might record their second choice for the Liberal candidate, and in this way secure in each case the defeat of the Labour candidates. On the other hand, an alliance between Labour and Conservatives might procure the defeat of the Liberal candidates. The representation of any one party would depend upon the action taken by members of other parties.
As the probable effects of the alternative vote becomes more fully understood its inadequacy as a remedy will be more clearly realized, and this proposal, instead of facilitating, may hinder the passage of a comprehensive measure of reform. On the contrary, the wider reform of proportional representation, providing as it would for the just and fair representation of three parties (and this is the problem for which a solution has to be found), has far greater claims to the consideration of practical politicians. It simplifies the problem of redistribution; it is the way by which equality in the value of votes can be secured; it provides for the fair representation of three parties, and, in guaranteeing the adequate representation of minorities, facilitates the adoption of a simple franchise. Proportional representation is, as it were, the master key which unlocks the difficulties associated with a comprehensive measure of electoral reform. Based on a broad simple principle, the justice of which is apparent to all, it provides the means by which each of the separate parts of such a measure can be most easily and effectively dealt with. Indeed, it is difficult to conceive on what other principle any permanent solution of the electoral problem can be based, or by what other means the difficulties inherent in a comprehensive measure of reform can be successfully overcome.
_Partial adoption of proportional representation not desirable_
Some who recognize the great merits of proportional representation have suggested its application to urban constituencies by way of experiment. Thus, Mr. Winston Churchill has expressed the opinion that "the proportional representation of great cities was a point upon which electoral reformers ought to concentrate their minds."[12] A partial application of the reform might be of value as further evidence of its practicability, but there is no need for this further evidence. The full benefits of the system cannot be expected from such experiments, and although a partial measure is apparently working satisfactorily in Wuertemberg, the history of the movement shows that such schemes usually arouse fierce opposition. An attempt to introduce a partial scheme in Belgium provoked a storm of indignation and had to be withdrawn, and the amendment to the original draft of the South African Constitution, carried in the Cape Parliament, limiting the proposed application of proportional representation to the towns, resulted in its complete abandonment for the elections for the House of Assembly. All partial applications of proportional representation are apt to work unfairly. In Belgium, the Catholics were stronger in the rural districts than in the towns and the proportional representation of the towns alone would have strengthened the political position of the Catholics. Similarly the limitation of proportional representation to the towns in South Africa would have strengthened the political position of the Dutch in those constituencies without giving a corresponding advantage to the minorities in the country areas. Were a partial application attempted in Great Britain it would be necessary to overcome the initial difficulty of selecting the constituencies to which the experiment should be applied, and in the absence of an agreement between the parties, it would be difficult, if not impossible, to escape the fatal charge of partisan selection.
_Proportional representation and democratic principles._
What hinders the adoption of a complete scheme of proportional representation? Is it not primarily a lack of courage and of trust in the principle of democracy? But does it need a greater courage, a greater belief in the value of the democratic principle than the grant of self-government to the Transvaal and to the Orange River Colony within a few years of the Boer War? The courage and faith in the latter case have been abundantly justified, and were statesmen actuated by a similar courage and belief in democracy to propose a system of proportional representation there would undoubtedly be a public response which would astonish them; for reforms which are obviously based upon justice are quickly and gladly accepted. Democracy cannot be carried to its highest pitch of perfection if the electoral methods by which representative institutions are brought into being are fundamentally defective. "By proportional representation," said Mr. James Gibb, "if electors were enabled to put more intelligence and conscience into their votes, the nation would be the gainer. The character of the electorate is of paramount importance, one outcome of it being the character of the House of Commons. The electors have not yet had a fair chance of showing what they can do in the making of a House of Commons. The question put to them is in such a form that they can hardly give an intelligible reply. The single-member system seems to imply a belief that the elector's liberty of choice must be narrow. We have now arrived at a point when another step is due in the evolution of the people's liberties, when an individual elector should obtain a greater freedom of choice and therefore a more intimate relation to national affairs.[13] Further, the smooth working of democratic institutions requires that no section of the electors should be permanently divorced from the governing body. Such separation begets a feeling of hostility towards the institutions of the country. Thus, Lord Dunraven has referred to Ireland as a country in the government of which some of its best citizens are not allowed to take part. Similarly, many British settlers in the Orange Free State, although resident for several years, never had any representative in the State Assembly. The natural feeling arose that the government of the country was a matter which did not concern them, and they never attended the meetings addressed by the member of the Assembly for the district. It may be true that minorities must suffer, but there is no reason why they should suffer needlessly. Here justice and expediency go hand in hand. It is to the advantage of the country that all should be associated with the representative body which speaks in the name of the whole, whether that body be a town council, a county council, or a House of Commons.
_Constitutional reform._
As pointed out in the opening chapter, the question of electoral reform is intimately associated with the constitutional problem which has occupied Parliament since 1906. This problem contains two factors--the relation between the two Houses of Parliament, and the constitution of the House of Lords. The House of Commons claims greater power in legislation on the ground that it is the expression of the national will. This demand has called forth a movement for reforming the House of Lords in order that it may fulfil more adequately its duties as a Second Chamber. The Unionist leaders have proposed that the peers should delegate their powers to a small number and that the House should be strengthened by the introduction of nominated and elected elements. With regard to the suggestion that a certain number of Lords of Parliament should be nominated by the Crown, all evidence points to the fact that such nominations invariably become party in character. No Government can afford to ignore the claims of the party which supports it, or to miss the opportunity of strengthening its position in one of the Houses of Parliament. The Canadian Senate, which is a nominated body, fails to give satisfaction, and there is a strong demand for its reform. At the conclusion of Sir John Macdonald's long lease of power the Senate consisted nearly wholly of Conservatives. Now that the Liberal Government has been in office for a good many years, the Senate is nearly wholly Liberal. Obviously, the introduction of a nominated element will not provide a Second Chamber that will command public confidence.
The elected element might be chosen indirectly by the County Councils or by the House of Commons, or the much bolder course of direct popular election, advocated by Sir Edward Grey, might be adopted. Direct election is distinctly preferable to indirect election by bodies created for other purposes. The experience of the United States, France, Sweden, and all other countries where the Upper House is elected by local legislatures, provincial councils, or municipalities, show that elections to the local authorities are fought on questions of national politics. But whether indirect or direct election is determined upon, it is already clear that the only possible method of election is that of proportional representation. The Royal Commission on Electoral Systems has reported that there is much to be said in favour of the transferable vote as a method of election for a Second Chamber, and this verdict has since been endorsed in numerous articles in the press. Thus a writer in the _Quarterly Review_ says that: "If an elected element is thought to be necessary for the popularity and effectiveness of a reformed Upper House, then let a certain number of members be elected in large constituencies by means of proportional representation."[14] Were the minimum age qualifying for a vote in such elections raised to twenty-five or more there would naturally be provided the conservative tendency to which that House is intended to give expression, and were peers eligible as candidates doubtless such peers as were interested in politics would experience little difficulty in securing election.[15]
The principle of election has been adopted for the Senates of Australia and of South Africa. In the former the majority system with direct election is used; in the latter, a proportional system with indirect election. The difference in the results is most striking. In Australia each of the States is polled as a separate constituency, each elector having three votes. The result of the election of 1910 was as follows:--
AUSTRALIA: SENATE ELECTIONS, 1910
State. Votes Polled. Labour Non-Labour Seats Obtained. Votes. Votes. Labour. Non-Labour. Victoria 648,889 692,474 3 -- New South Wales 736,666 735,566 3 -- Queensland 244,292 124,048 3 -- South Australia 171,858 148,626 3 -- Western Australia 128,452 109,565 3 -- Tasmania 92,033 75,115 3 -- --------- --------- -- -- 2,021,090 1,997,029[16] 18 --
It will be seen that the Labour Party polled 2,021,090 votes and obtained 18 seats, whilst their opponents, with a poll of no less than 1,997,029 votes, obtained none. So effectively does the majority system in the form of the block vote blot out minorities. The Hon. W. Pember Reeves, in commenting upon these figures,[17] said that: "Such results give rise to revolutions."
In South Africa each State is represented by eight Senators chosen by the local Parliaments by means of the single transferable vote. The first elections gave the following result:--
SOUTH AFRICA: SENATE ELECTIONS, 1910
Seats Obtained. States. Dutch Parties[18] British Parties[18]
Cape Colony South African 6 Progressive 2 Transvaal Het Volk and Progressive and Nationalist 5 Labour 3 Natal Dutch 1 British 7 Orange Free State Orangia Unie 6 Constitutionalist 2 -- -- Total 18 Total 14
In the one case minorities are completely suppressed; in the other the minority in each State obtains representation.
These two illustrations show that if the House of Lords is to be strengthened by the infusion of an elected element chosen by large constituencies, a true system of election must be adopted. This is the conclusion arrived at by Professor Ramsay Muir[19] after a careful examination of the different methods by which a Second Chamber can be constituted. All suggestions as to the selection of peers by hereditary peers, of peers qualified by service, by nomination, by indirect election, by direct election on a limited franchise, are ruled out and the direct election of a new Second Chamber by the single transferable vote is advocated in order that the new House may contain those elements which fail to secure representation with a system of single-member constituencies. But if, by the adoption of direct popular election and proportional representation, the Upper House were made more truly representative than the Lower, then whatever resolutions were passed defining the relations between the two Houses there is not much doubt that power would tend to pass into the hands of the more representative House. In commenting upon the Royal Commission's report _The Nation_[20] said: "Perhaps the most pregnant sentence in this whole report is that in which the Commission suggests that proportional representation might be a suitable basis for an elective Senate. We have our liberty of choice, and democracy may find its account in either alternative. We may prefer to retain an imperfectly representative Lower House. But if we place above it a really representative Senate the whole balance of the Constitution might be altered, and the Senate become the more venerable, the more democratic, and in the end, the more powerful Chamber. We may, on the other hand, reform the House of Commons, and render any Senate superfluous. In either event, proportional representation may become the ultimate key to our constitutional problem."
_Federal Home Rule._
The same question, the method of election, must enter into the consideration of those larger schemes, Federal Home Rule and Imperial Federation, which have been mooted in the discussion of the constitutional relations between the two Houses of the Parliament of the United Kingdom. A writer in _The Times_,[21] whose series of letters attracted considerable attention, said that the "central idea of Federalism appears to be that our present single Imperial Parliament, which does, or makes an attempt at doing, all the complicated work--first of the Empire, and second of the United Kingdom of Great Britain and Ireland, and third of the various countries which together make up the United Kingdom--is no longer adequate to the purpose. The Federalists therefore propose that the Imperial Parliament, while maintaining its supremacy absolutely intact, shall delegate a large part of its functions to a number of subordinate national or provincial Parliaments, who shall manage the domestic affairs of England, Scotland, Ireland and Wales, or of such other territorial divisions as may be agreed upon. These national or provincial Parliaments will be entirely independent one of another, but all will acknowledge the full and absolute sovereignty of the Imperial Parliament." Mr. Birrell stated that "Federation beginning here at home, as it is called, is ripening for a speedy decision. Such a Federation once established would be able to find room for our Dominions overseas as and when they wished to come in. We should have then a truly Imperial Parliament, at the door of which any one of our Dominions could come in, and as it were hang up its hat and coat in his Mother's House and take part in common Imperial proceedings, and in the government of this great Empire."[22] These are great changes, and without entering too deeply into details of how these new bodies are to be brought into being, it is certain that one of the conditions of their successful working is that they must be fully representative. It is inconceivable that a national council can be set up for Wales, or for Scotland, or for Ireland, without provision for the adequate representation of minorities. Lord Morley, in instituting the new Councils in India, was compelled to make provision for the representation of Muhammedans. Mr. Birrell, in the Irish Council Bill of 1907, proposed that minorities should be represented by members nominated by the Crown. It is impossible to reconcile this reactionary proposal with democratic principles, and there can be no possible reason for its adoption when there is a method of election available which enables minorities to choose their own representatives.
_Imperial federation._
Mr. Birrell's vision of an Imperial Parliament for the British Empire raises once more the value of a true method of election. An Imperial Parliament will not accomplish its purpose--the consolidation of the Empire--if the basis of representation is such as to give undue emphasis to the separate interests of the constituent States. Further, it would seem desirable that the establishment of such a Parliament should be preceded by the more complete unification of the various States, for in no other Empire are there so many racial divisions, and it is from these that the greatest of political difficulties spring--in Ireland the division between north and south; in the United Kingdom between Ireland and Great Britain; in South Africa between the Dutch and British; in Canada between the French and British. The majority system of election brings out these differences in their acutest form. In Canada in 1910 no representative from the Province of Quebec attended the National Conference of Canadian Conservatives; of the four Provinces forming the South African Union it was in the Orange Free State, where in the local Parliament the minority was almost wholly deprived of representation, that racial differences gave rise to the keenest feeling. Proportional representation has proved itself to have been of the greatest value in bi-racial countries such as Belgium where the representation of political parties no longer coincides with racial divisions. The adoption of proportional representation in the United Kingdom, in Canada, and for all elections in South Africa would complete the consolidation of these various divisions of the Empire, and even where racial difficulties do not exist, as in Australia and New Zealand, the fair representation of all classes of citizens would free questions of Imperial politics from the dangers of exaggerated party majorities.
_Conclusion._
Whether it is a question of improving existing institutions, or the creation of further representative bodies, the method of election is all important. All other departments of human activity show continuous improvement, and the substitution of scientific for rule-of-thumb methods of election is an improvement long overdue. It may even be said that the continued successful working of representative institutions demand such an improvement. The accomplishment of other electoral reforms can be more easily attained by the adoption of a system which allows of the fair representation of all. The reform of the House of Lords, whether by the delegation of the powers of existing peers to a small number, or by the introduction of an elected element, or its establishment on a completely democratic basis, necessitates the adequate representation of minorities. Federal Home Rule is impracticable unless due provision is made for minority representation. But in the contemplation of newer legislative bodies it must not be forgotten that it is of the utmost importance that the prestige of the House of Commons--the mother of parliaments, and, as such, the glory of English-speaking peoples--should be maintained at the highest level. Yet its predominance in the Parliament of the United Kingdom can be permanently secured only if it is made fully and completely representative. The House of Commons must once more renew itself; it must establish itself on sounder foundations. Its privileges and powers have been won by the efforts of past generations. To the present generation falls the opportunity of perfecting its organization and of strengthening its foundations by making it in truth the expression of the national will.
[Footnote 1: Reply to Deputation of Liberal members at House of Commons, 20 May 1908.]
[Footnote 2: "This number might be reduced to eleven, if minor variations were grouped."--Sir Charles Dilke, National Liberal Club, 10 May 1909.]
[Footnote 3: _The Essentials of Self-Government,_ 1909, p. 62.]
[Footnote 4: Section 41 of the South Africa Act, 1909, reads thus: "As soon as may be after every quinquennial census the Governor-General-in-Council shall appoint a commission consisting of three Judges of the Supreme Court of South Africa to carry out any redivision which may have become necessary as between the different electoral divisions in each Province, and to provide for the allocation of the number of members to which such Province may have become entitled under the provisions of this Act."]
[Footnote 5: The Town Clerk of Edinburgh, Dr. Hunter, urges a rearrangement of the Parliamentary Divisions of the city, so as to assimilate them to the municipal wards. "Confusion and unnecessary expense are caused by the present arrangement.... The municipal area of the city is represented in Parliament partly by the four city members, partly by the member for Leith Burghs, and partly by the member for the County of Midlothian. The distinction thus existing between the Municipal and Parliamentary divisions of the city necessitates the annual making up of separate rolls of voters for municipal and for Parliamentary purposes respectively, involving heavy additional expense (amounting to upwards of L1100 per annum), which would be avoided if the areas for both purposes were assimilated." Assimilation is desirable "not merely in order to save needless expense, but in the interests of candidates and electors as well as of the electoral agencies. In the dual arrangement at present existing the usual organizations for electoral purposes of all kinds have to be duplicated. Not one of the Parliamentary wards correspond with any of the municipal wards."--_The Scotsman_, 9 August 1910.]
[Footnote 6: "The General Election of January 1910, and the Bearing of the Results on some Problems of Representation." Paper read before the Royal Statistical Society, 19 April 1910. Mr. Rosenbaum, however, rejects proportional representation on political grounds. These have been considered in the two previous chapters.]
[Footnote 7: "Electoral Statistics." Paper read before the Manchester Statistical Society, 12 December 1906.]
[Footnote 8: Joseph King, M.P., in evidence before the Royal Commission on Electoral Systems, 1909.]
[Footnote 9: This difficulty would disappear with the adoption of Home Rule.]
[Footnote 10: _Real Representation for Ireland_, 1908.]
[Footnote 11: Report of Annual Meeting of the Proportional Representation Society, 21 July 1909.--_Representation,_ vol. ii. p. 154.]
[Footnote 12: In reply to a deputation of the Manchester Liberal Federation, 22 May 1909.]
[Footnote 13: _Minutes of Evidence_, Royal Commission on Electoral Systems, 1910 (Cd. 6352), p. 104.]
[Footnote 14: _Cf._ "Two Chambers or One," _Quarterly Review_, July 1910.]
[Footnote 15: The indirect election of the United States Senate gives so little satisfaction that the House of Representatives on 14 April 1911 approved of the proposed amendment to the Constitution providing for popular election by 296 votes to 6.]
[Footnote 16: Of these, the Fusionists polled 1,830,353 votes.]
[Footnote 17: Address to the London School of Economics, 5 October 1910.]
[Footnote 18: These broad distinctive titles are here given, although the author recognizes that the Nationalist and Unionist parties in South Africa are not exclusively Dutch or British.]
[Footnote 19: _Peers and Bureaucrats_, by Ramsay Muir, Professor of Modern History at Liverpool University.]
[Footnote 20: 21 May 1910.]
[Footnote 21: "Pacificus," _The Times_, 31 October 1910.]
[Footnote 22: Address to the Eighty Club, 25 July 1910.]
APPENDIX I
THE JAPANESE ELECTORAL SYSTEM--THE SINGLE NON-TRANSFERABLE VOTE
The following memorandum has been written by Mr. Kametaro Hayasbida, the Chief Secretary of the Japanese House of Representatives, in reply to a series of questions, the particulars of which are set out in the memorandum.
_Failure of single member system._
The Original Election Law of our country was promulgated in 1889, the same year in which took place the promulgation of the Constitution. Under this law the system of small electoral districts was single-adopted, and each _Fu_ or _Ken_ (administrative district) was divided into several electoral districts each of which constituted a single-member constituency (with the exception of some large districts which, impossible of further division, had two seats allotted with the system of _scrutin de liste_). The system was, however, found in practice to be very unsatisfactory, as it often happened that a minority of the voters, instead of the majority, in certain _Fu_ or _Ken_ obtained the majority of the members returned, and, on the other hand, a party with a majority at the polls could not sometimes, as the result of the grouping of the voters in the small electoral districts, secure any representation at all. Under such circumstances it was utterly impossible for each political party to obtain representation in reasonable proportion to the strength of its voters; or, in other words, the electors of the country at large had never succeeded in being properly represented in their legislative body. As the inadequacy of the system was thus apparently shown I formulated in 1891, by somewhat what modifying Marshal's cumulative voting system, a system of large electoral districts combined with that of the single vote, and urged for a revision of the Election Law.
_Multi-member constituencies. Single vote adopted 1900._
Since then several elections had taken place; and the defects of the existing law were more strongly pronounced at each successive election. It was, however, not until the year 1898 that the Government at last introduced a Bill for a revision of the law with the view of adopting the system I had the honour of formulating. After heated discussion in three successive sessions, the Bill was passed in 1900 and sanctioned as a law. This is our present Election Law. In the revised system the _Fu, Ken_, and _Shi_ (the administrative districts) constitute at the same time the electoral districts, and a voter in each district has but one vote for one candidate, while several seats (according to the population) are allotted to the district.
The above is a brief historical sketch of our electoral system. I shall now try to answer your questions in order.
_Equitable results._
As to the first question whether our system secures the representation of each party in reasonable proportion to its voting strength, I cannot do better than answer it by pointing out a few instances in the General Election which took place on the 15 May 1908.
TABLE I
THE CITY OF TOKYO (11 seats)
Seats in Seats Parties. Votes. Proportion Obtained. to votes. Seiyu-Kwai (Liberals) 6,579 2.71 2 Konsei-honto (Progressives) 2,216 0.91 1 Daido-ha (Conservatives) 2,879 1.18 2 Yuko-Kwai (Radicals) 4,656 1.91 2 Churitsu (Independent) 10,414 4.29 4 ------ ----- -- Total 26,744 11.00 11
All parties except the Seiyu-kwai and Daido-ha succeeded in obtaining their representatives in reasonable proportion to their respective voting strength. The explanation given for the particular case of the Seiyu-kwai is that the party, unable for some reason or other to limit the number of candidates, had placed five candidates instead of three or four, and caused its own defeat by splitting the votes. I take at random, or rather in the order they come, a few more districts, and the results obtained are as follows:--
TABLE II
TOKYO-FU (5 seats)
Parties. Number of Seats in Seats Candidates. Votes. Proportion Obtained to Votes. Seiyu-kwai 5 12,794 4.02 4 Kensei-honto - - - - Daido-ha. 1 13,122 .98 1 Churitsu - - - - ------ ---- - Total 6 15,916 5.00 5
TABLE III
THE CITY OF KYOTO (3 seats)
Parties. Number of Seats in Seats Candidates. Votes. Proportion Obtained to Votes. Seiyu-kwai 1 1,284 0.45 - Kensei-honto - - - - Daido-ha - - - - Yuko-Kwai - - - - Churitsu 3 7,304 2.55 3 - ----- ---- - Total 4 8,588 3.00 3
TABLE IV
KYOTO-FU (5 seats)
Parties. Number of Seats in Seats Candidates. Votes. Proportion Obtained. to Votes. Seiyu-kwai 5 18,928 4.01 4 Kensei-honto -- -- -- -- Daido-ha -- -- -- -- Yuko-kwai -- -- -- -- Churitsu 1 4,701 0.99 1 -------------------------------------- Total.... 6 23,629 5.00 5
TABLE V
THE CITY OF OSAKA (6 seats)
Parties. Number of Seats in Seats Candidates. Votes. Proportion Obtained. to Votes. Seiyu-kwai 5 8,666 3.32 4 Kensei-honto -- -- -- -- Daido-ha -- -- -- -- Yuko-kwai 1 2,612 1.00 1 Churitsu 2 4,368 1.68 1 --------------------------------------------- Total.... 8 15,646 6.00 6
TABLE VI
OSAKU-FU (6 seats)
Parties. Number of Seats in Seats Candidates. Votes. Proportion Obtained. to Votes. Seiyu-kwai 5 15,137 3.57 5 Kensei-honto -- -- -- -- Daido-ha 1 2,199 0.52 -- Yuko-kwai 1 1,304 0.31 -- Churitsu 3 6,786 1.60 1 --------------------------------------------- Total.... 10 25,426 6.00 6
Throughout all electoral districts similar results were obtained. The Churitsu (_i.e._ those belonging to no party), considered as a group, had not everywhere been as successful as the other parties, as observe in Tables V. and VI. Each candidate of this group is quite independent of the other, and has no political views or propaganda in common, nor any organization whatever. Therefore, each case is totally different from the other. Although all independent candidates or voters are in these tables grouped as Churitsu, it is not proper to consider them in the same category with the other parties.
Now, judging from the results in the General Election, a few instances of which are given above, I may say that our present system, if not fully satisfactory, tolerably secures the representation of each political party in approximate proportion to its voting capacity.
_The new system and party organization._
As to the first part of your second question, whether, to obtain these results, the system involves a great deal of calculation on the part of political organizations as to the exact number of their supporters, I should say that, as the same system and method of election are uniformly adopted in the city, county, borough and village elections as well as in the elections of the Prefectural Assembly, it is not a very difficult task for all political parties to ascertain from the results of all these elections their relative strength, and to estimate the number of their supporters.
As to the second part of the question, whether it is necessary to issue precise instructions to the electors as to the candidates for whom they should vote, my answer is this: as every political organization through its branch in every _Fu_ and _Ken_ and the sub-branches in the cities, counties, towns and villages, is always in close touch with its constituents, and is constantly explaining its position and propaganda, with the view not only to instruct them but also to extend the sphere of its influence, it is not so difficult as it seems to decide the number of candidates. When it is once decided efforts are made on the part of the organization to distribute the votes among the candidates in such a way that not one of them receives a defeat at the hands of the other party. To attain this object the methods are not very complicated, for every elector has but one vote for one candidate; and, moreover, the stronger candidates, so long as their own position is secured, will endeavour to distribute a portion of their votes among the weaker candidates. This being the case, the member returned with the greatest number of votes may not be the most popular candidate, but the party as a whole is much more likely to succeed in getting representatives in proportion to the strength of its voters.
_The position of independents._
As to the third question, whether the system enables men of independent mind and character to maintain their position in Parliament, I should emphatically state that the revised system is much better than the old in this respect. Under the old system even such a prominent man as Mr. M. Matsuda (the Speaker of the House of Representatives some years ago, and the Minister of Finance in the present Government) suffered several defeats. But under the new system it has never happened that the leader of a party has lost his seat at any election, as he may seek his election at the safest district. To men of independent mind and character the new system offers the greater opportunity to maintain their position in the House, for in the election they may, in spite of the opposition of parties, draw their votes from all parts within a large electoral district. It may be said that the larger electoral district we have, the greater opportunity we afford to independent candidates. For instance, both Mr. Y. Ozaki, the Mayor of Tokyo, and Mr. S. Shimada, by being independent candidates, have never lost their seat in Parliament, and in the last General Election were returned for their native prefecture or town with a great number of votes.
This brings me to the end of my answers to your inquiries. In conclusion I may say a few words about the public opinions in our country as to the Election Laws.
_Public opinion and the new system._
Despite the fact that the new system enables the elector of the country to be more reasonably represented in the House, still there are some ambitious politicians urging for their own selfish purpose to restore the old system. But, as almost all prominent members in both Houses are fully cognizant of the relative merits and demerits of the two systems, there is not much chance of our returning to the old system.
APPENDIX II
THE SECOND BALLOT
A Note on the German General Elections of 1903 and 1907.
The German Reichstag, which consists of 397 members, is elected by a system of single-member constituencies. Every member, however, must have obtained a majority of the votes polled, either at a first or second ballot, in the constituency for which he has been returned. The German Official Returns furnish very complete details of the elections, including the figures for the first and second ballots, and the summaries at the end of the Returns disclose a very striking divergence between the proportions of seats obtained and votes polled by the various political parties. These discrepancies have attracted general attention, and have usually been attributed to the great variation in the size of German constituencies. As a matter of fact, the effect of redistribution on the proportionality between seats and votes is not nearly so large as is generally supposed. Apart from the consequences of neglecting the votes of the minority or minorities in each constituency, wherein lies the gravest defect of a single-member system, the second ballot is a disturbing factor of considerable importance. So far from diminishing the disproportion between seats and votes polled by the various parties, the second ballot frequently increases that disproportion. In order to appreciate the respective effects of unequal constituencies and of the second ballots it is necessary to consider these two factors separately. This will be facilitated by making a comparison between the results which would have been obtained without second ballots with the results actually obtained. The following tables, which are based upon the official returns, give the votes polled and the seats obtained by the five principal groups:--
GERMAN GENERAL ELECTION, 1903
Parties. Votes. Results without Results with Second Ballot. Second Ballot. Social Democrats 3,010,771 122 81 (31.7%) (30.7%) (20.4%) Centre Party 1,875,273 104 100 (19.7%) (26.2%) (25.2%) National Liberals 1,317,401 32 51 (13.9%) ( 8.1%) (12.8%) Conservatives 1,281,852 79 75 (13.6%) (19.9%) (18.9%) Radical Parties 872,653 11 36 ( 9.2%) ( 2.8%) ( 9.1%)
GERMAN GENERAL ELECTION, 1907
Parties. Votes. Results without Results with Second Ballot. Second Ballot. Social Democrats 3,259,029 73 43 (28.9%) (18.4%) (10.8%) Centre Party 2,179,743 101 105 (19.3%) (26.4%) (26.4%) National Liberals 1,630,681 47 54 (14.5%) (11.8%) (13.6%) Conservatives 1,632,072 91 84 (13.6%) (22.9%) (21.2%) Radical Parties 1,233,933 30 49 (10.9%) ( 7.6%) (12.3%)
_The effect of unequal constituencies on representation_.
The Social Democrats were affected to a greater extent than any other party by both the factors referred to. In 1903 the Socialists polled 31.7 per cent, of the votes, and, at the first ballots, were at the head of the poll in 122, or 30.7 per cent, of the constituencies. In other words, if the system of second ballots had not been in force, the Social Democrats would have obtained very nearly their fair share of representation. If, in addition, there had been a redistribution of seats by which the sizes of constituencies had been equalized, the Social Democrats would have obtained more than their share of representation. The strength of the party lay in the large towns, and if, for example, Berlin had the additional eight seats to which it was entitled nearly all of them would have fallen to the Social Democrats. Again the three divisions of the district of Hamburg returned Social Democrats with overwhelming majorities. Were the representation allotted to Hamburg doubled, as it should be, all six seats might possibly have fallen to the Social Democrats.[1] An equalization of the size of constituencies might have produced in 1903 the phenomenon which has occurred so often in England. The largest party would have secured a number of seats far in excess of that to which it was entitled by reason of its strength. In 1907 the Socialists polled 28.9 of the votes, but only succeeded in reaching the head of the poll at the first ballot in 73, or 18.4 per cent. of the constituencies. A redistribution of seats would have added to their representation in the large towns, and the first ballots would have yielded a result which would have corresponded more fairly with their polling strength.
_The effect of second ballots_.
In both years the system of second ballots has had the effect of reducing very considerably the representation of the Social Democrats. In the year 1903 the Social Democrats won 56 constituencies by absolute majorities, and were engaged in the second ballots in 118 constituencies. In 66 of these constituencies they were at the head of the poll, but succeeded in maintaining this position at the second ballots in 24 only. In the remaining 52 constituencies they were second on the poll, and at the second ballots they were able to win only _one_ of these seats. In these 118 constituencies the Socialists polled 1,170,000 votes at the first ballots, whilst the other parties polled 1,920,000. As a result of the second ballots the Socialists obtained 25 seats and the remaining parties obtained 93 seats.
The figures of the year 1907 tell a similar tale. At the first ballots the Social Democrats were at the head of the poll in 73 constituencies. The second ballots reduced this number to 43. They were engaged in the second ballots in 90 constituencies; they were at the head of the poll in the first ballot in 44 of these constituencies, but kept this position in 11 only; they were second on the poll in the remaining 46 constituencies and won in 3 cases only. In these 90 constituencies the Social Democrats polled at the first ballot 1,185,000 votes, whilst the other parties taken together polled 1,888,000 votes; the Socialists obtained 14 seats, the other parties obtained 76 seats.
In both these elections the second ballots affected very adversely the representation of the largest party. If this party, without the second ballot and with a fair distribution of seats, might have obtained more than its share of representation, then the second ballots would have acted as a corrective, but not necessarily so. There is no reason why the second ballots should not have added to the over-representation already obtained. This will be seen from the figures of the elections in the Kingdom of Saxony. This division of the German Empire is entitled to 23 representatives in the Reichstag. In 1903 the Socialists won 18 of these seats with absolute majorities; they were engaged in the second ballots in the remaining five constituencies; they won four (all those in which they were at the head of the poll at the first ballots) and only lost the one constituency in which they were second on the poll. The Social Democrats, who at the first ballots polled 58.8 per cent, of the votes, thus obtained 22 seats out of 23, and the second ballots in this case only confirmed the overwhelming preponderance which the system of single-member constituencies had conferred upon the larger party.
_Second ballots and the swing of the pendulum_.] It would, indeed, seem that a system of second ballots rather accentuates those great changes in representation which are the normal characteristic of a system of single-member constituencies. In the elections in Saxony in 1907 the Social Democrats were still by far the largest party, obtaining 48.5 per cent. of the votes. They succeeded in obtaining eight seats by absolute majorities and were engaged at the second ballots in eight other constituencies. They lost every one of these constituencies, although at the first ballots they had been at the head of the poll in five of them. The unfavourable swing of the pendulum reduced their representation at the first ballots, and the second ballots merely increased their misfortunes.
Nor would redistribution have lessened the violence of these changes in the constituencies in which second ballots were necessary. Thus, for example, Frankfort-On-Main, with an electorate of 77,164, should return two members instead of one. The constituency was won by the Socialists in the second ballots of 1903, but was lost at the second ballots in 1907. In both years the Socialist candidate was at the head of the poll at the first ballots. Similarly the constituency of Elberfeld-Barmen, with an electorate of 67,241, won by an absolute majority in 1903, was lost by the Socialists at the second ballots in 1907, although their candidate had been at the head of the poll at the first ballot. If these and other constituencies had received additional representatives, the violence of the changes in the composition of the legislative body would in all probability have been increased.
_The second ballot and the representation of minorities_.
A study of the statistics of the German General Elections shows that the representation obtained by the various parties depends very largely upon their supremacy in certain localities. In these districts the minorities have been unrepresented for many years, the second ballots having in no way saved them from practical disfranchisement. Thus the Centre Party is in the ascendant in the Rhenish Provinces. In the district of Cologne, Muenster, and Aix-la-Chapelle, the Centre Party monopolizes the representation, returning in 1907 every one of the 15 members to which the districts were entitled. In the adjoining districts of Dusseldorf, Coblentz and Treves they returned 16 out of 24. In Bavaria, the districts of Lower Bavaria, the Upper Palatinate, Lower Franconia and Schwabia, which are entitled to 23 members, were represented wholly by members of the Centre Party. Taking the kingdom of Bavaria as a whole, the Centre Party obtained 34 seats out of 48, although they polled only 44.7 per cent of the votes at the first ballots. There is therefore reproduced in Germany the conditions which obtain in certain parts of the United Kingdom--the permanent supremacy of one party which monopolizes, or nearly so, the representation of the district.
_Summary_
The system of second ballots has therefore had a considerable influence in creating that divergence between the votes polled and the seats obtained which has characterized German elections. The representation of any one party depends, to a very large degree, upon the attitude taken towards it by other parties. The system in no way acts as a corrective to the anomalies arising from single-member constituencies, and may even accentuate the violent changes associated with them. Moreover, the system does not provide representation for minorities, and therefore does not ensure a fully representative character to popularly elected legislative bodies. It may be mentioned that all the criticisms here directed against the second ballot apply with nearly equal force to the use of the alternative vote (_see_ p. 95), a thinly disguised form of the same principle which appears to be meeting with some acceptance in this country.
[Footnote 1: The minority would, of course, have had a better chance with six divisions. Dr. Ed. Bernstein, to whom the author submitted this memorandum, makes the following comment: "I am not so sure that the equalization of the size of the constituencies would in 1903 have secured to the Social Democratic party a number of seats far in excess of its voting strength. But this is a subordinate consideration. The possibility of an unproportional representation of parties, even if the seats are equally distributed, is undeniably there, and this ought to settle the question.]
APPENDIX III
THE SWEDISH SYSTEM OF PROPORTIONAL REPRESENTATION
The principle of proportional representation was first discussed in Sweden in 1867. The new Danish Constitution of that year provided for the use of the transferable vote (Andrae's scheme) in the election of the Upper House, and Herr S. G. Troil proposed in the Swedish Parliament that the three most important of its committees should be elected by means of the same system. The motion was not carried, and a similar motion, made by Professor H. L. Ryoen in 1878, was equally unsuccessful. It was not until 1896 that the next step was taken, when the Government, in view of the increasing demand for a more democratic franchise, proposed a proportional system of election. Nothing came of this proposal immediately, but from this date the agitation for an extension of the franchise gave rise to the demand for the proportional method of election in order to ensure the representation of minorities.
_The former constitution of the two chambers_.]
The story of the struggle for reform will best be understood if prefaced by a statement of the franchise conditions previously existing in Sweden. The Upper, or First, Chamber of the Riksdag, was elected by members of the provincial councils and of the councils of the five largest towns. The other towns sent members to their provincial councils. The members of provincial councils were elected in two stages; the primary electors chose electors of the second degree, who in turn chose the councillors. The primary electors in the country[1] had ten votes for every 100 kroner of rateable income, subject to a limit of 5000 votes. The electors of the second degree had only one vote in the election of councillors, and councillors had only one vote in the election of members of the First Chamber of the Riksdag. Owing to the great advantage conferred upon primary electors possessed of large incomes these electors largely controlled not only the composition of the town and provincial councils, but also the composition of the Upper Chamber. The election of members of the Lower Chamber of Parliament was direct; every person of not less than 800 kroner income was entitled to vote, but no one was entitled to more than one vote.
_The struggle for electoral reform_.
In 1899 M. Branting, the leader of the Socialist Party, proposed the adoption of proportional representation, coupled with universal and equal suffrage for the election of town councils. The main object of this proposal was to place town councils on a more democratic basis, but as the five largest councils elected representatives to the First Chamber the proposal would have had some influence upon the composition of that House. M. Branting's proposal was rejected, and when revived two years later met a similar fate. In 1902 two Liberals (MM. Hedlund and Carlsson) proposed that provincial councils should be elected by a proportional method on the basis of manhood suffrage, whilst a similar proposition was made in the same year in respect of the elections of the Lower House of Parliament. Both these motions were rejected, but in response to a demand from both Houses for an inquiry a Royal Commission was appointed to consider the problem of electoral reform. The Commission reported in the following year in favour of a list system of proportional representation with official ballot papers, and the Government proposed this system combined with manhood suffrage for the election of members for the Lower Chamber. This proposal was accepted in 1904 in the Upper Chamber, but rejected in the Lower Chamber by five votes. Next year it was again discussed, accepted by the Upper Chamber but rejected in the Lower by a majority of ten. A change of ministry took place, and in 1906 M. Staaff, the Liberal Prime Minister, proposed manhood suffrage with the "majority" system of election. But the Moderate Party insisted upon a proportional system, and the proposals of the Liberal ministry were rejected by the Upper Chamber. M. Alfred Petersson, of Paboda, then proposed manhood suffrage with a proportional system for the Lower Chamber, and a proportional system for the Upper Chamber, which, however, was to be elected as before by the provincial councils. This proposal was rejected by the Lower Chamber but accepted by the Upper Chamber, and M, Staaff resigned. The Moderates, with M. Lindman as Prime Minister, then introduced a Bill incorporating M. Petersson's proposals with the addition of the direct election of provincial councils and a less plutocratic franchise. This measure, which was adopted by both Houses in 1907, was confirmed after a General Election in 1909.
_The Swedish law of 1909_.
Under this law the proportional system is applied to elections for both Houses of Parliament, all parliamentary committees, town councils and provincial councils. For the Lower Chamber there is manhood suffrage. The Upper Chamber is elected still by the provincial councils and by the town councils of the five largest towns, but the elections of provincial councils are now direct. But, in order to maintain as much continuity as possible in the composition of the Upper Chamber, only one-sixth of the House is renewed every year. The maximum number of votes in the elections of both provincial and town councils is forty. The first election under the new system took place in 1909, when the Stockholm Town Council and several provincial councils were called upon to elect their proportion of members of the Upper House. In March 1910 the first elections to the Stockholm Town Council were held, and in the following May there were elections under the new system for all the provincial councils. In 1911 the first elections to the Lower House of Parliament will take place.
In Sweden, even under the new law, there are no official ballot papers and no nominations of candidates. This arrangement is supposed to preserve to the electors the fullest possible liberty in voting. In practice the party organizations print ballot papers containing the names of the candidates whom they support, and these printed forms are accepted by the returning officers. Every elector, however, is at liberty to strike out any of the names on these papers, to substitute other names, to vary the order in which the names are printed, or to prepare his own ballot paper.[2]
_The Swedish system of proportional representation_.]
The mechanism of the proportional system adopted has had regard to the practice mentioned in the preceding paragraph. The first proposal, that of M. Petersson, of Paboda, was only a crude approximation towards a proportional system. His scheme, in brief, was (1) that the number of votes recorded for each candidate should be ascertained; (2) that the candidate with the highest number of votes should be declared elected; (3) that a further count should then take place, the papers on which the successful candidate's name appeared being treated as of the value of one-half. The remaining candidates whose names appeared on these papers would be credited with half a vote in respect of each such paper. The non-elected candidates would then be arranged according to the number of votes obtained, the highest being declared elected. As soon as any two names on any ballot paper had been declared successful a fresh count would take place, such papers being treated as of the value of one-third. This process of reducing the value of the paper as soon as a further candidate appearing thereon was elected was to be continued until all the seats were allotted. The principle underlying this distribution of seats is the same as that contained in the d'Hondt rule of the Belgian system. A group of electors which was more than twice as numerous as any other group would obtain two seats before any was allotted to a smaller group. If the group was more than three times as large as any other it would obtain three seats before the smaller group received one, and so on. It was at once recognized that this scheme would tell considerably in favour of well-organized parties--parties whose supporters would accept the ballot papers printed for them without question. An example will make this clear. If, taking an extreme case, in an election for three members 8000 voters placed the names of two candidates, P and Q, on each of their ballot papers, whilst a more loosely organized group of 13,000 voters spread its support over four candidates, T, S, V and W, different sections voting for these candidates independently, the following result might take place:--
P Q . . 8,000 | T . . . 4,000 | S . . . 3,500 | V . . . 3,000 | W . . . 2,500
Candidate P, being the first in order on the 8000 ballot papers of the first group, would be declared elected, and Q, the remaining name on these ballot papers, would be credited with 4000 votes--half the original value of the papers. Q and T, having 4000 votes each, would then be declared elected. Thus one group, with 8000 votes, would carry two seats, and the other, with 13,000 votes, would only obtain one--a result due to a lack of combination.
_The allotment of seats to parties_.
The plan finally adopted is based on M. Petersson's proposal, but provides, as in the Belgian scheme, for the official recognition of parties. Electors may write at the head of their ballot papers the name or motto of a party. The papers bearing the same name or motto are then grouped together, the numbers in each group ascertained, and the seats available are allotted to these groups in accordance with the d'Hondt rule, irrespective of the number of votes obtained by individual candidates. Thus, in the example given, if electors of the second group had all headed their ballot papers with the same party name or motto the particular way in which they had distributed their votes among the candidates would not have affected the number of seats obtained by the group as a whole. The first group would have obtained one, and the second two seats.
_The selection of the successful candidates_.
The position of the candidates on each list is determined in accordance with the original proposal of M. Petersson. The candidate receiving the highest number of votes is declared elected, the papers on which his name appears are then marked down to the value of one-half, the relative position of the remaining candidates ascertained afresh, and the highest of these declared elected, and so on. This procedure, called the reduction rule, is however subordinate to a further rule (the rule of the order of preference), which is as follows. If more than one-half of the supporters of a party list have placed the same candidate at the head of their ballot papers, the first seat apportioned to the list is allotted to this candidate; if more than two-thirds have placed the same two candidates in the same order at the head of the ballot papers, these two candidates have the first claim to the seats apportioned to the party; if more than three-fourths have placed the same three candidates in the same order at the head of the list, these are given the first, second, and third seats, and so on. The selection of the successful candidates is determined in accordance with this rule so far as possible, but as soon as the application of the rule breaks down the relative claims of the non-elected candidates on the list are determined in accordance with the reduction rule. But if, say, three candidates have been declared elected in accordance with the rule of the order of preference, and it is necessary to choose others by the reduction rule, the papers containing these three names are treated as of the value of one-fourth in determining the relative position of the remaining candidates of the group.
_Free voters and double candidatures._
In order to complete the description of the Swedish system two subsidiary features, which will seldom come into play in actual elections, must be mentioned. Provision is made for those electors who owe no party allegiance, and who therefore do not wish to place any party name or motto at the head of their list. Such voters are called "free voters," and the votes recorded for their candidates are ascertained. These candidates are placed in a group by themselves, called the free group, but the number of votes recorded for each individual candidate in this group, and not the total number of votes recorded for all the candidates, forms the basis of comparison with the totals of the party lists in the allotment of seats. The second feature provides for the improbable case of two groups of electors or parties having placed the same candidate upon their list. In the event of such candidate being so favourably placed in two lists as to be elected by both parties, then, for the purpose of ascertaining the new value of the papers on which his name appears, each list is debited with half a seat. When, as already explained, one seat has been allotted to a list, the list total is divided by two in accordance with the d'Hondt rule for the purpose of the fresh comparison of totals; but if this candidate has already been elected on another list the total would be divided by one and a half instead of by two. A fresh total would be ascertained for each of the lists containing the candidate's name.
_An election at Carlskrona._
The author was permitted by the courtesy of the Burgomaster of Carlskrona to watch the election of provincial councillors on 24 May 1910, to represent the city in the Bleking provincial council, and a description of this election will show how the system works in practice. Carlskrona is entitled to nine members. For the purpose of the election the town was divided into two parts, but the polling place in each division was at the town hall. The register was prepared fourteen days before the election, and stated in addition to the name, address, and occupation of the elector, the amount of his (or her) rateable income and the number of votes to which he (or she) was entitled. The conduct of the election was in the hands of the Burgomaster, assisted by the magistrates of the town. As already explained, there were no official ballot papers and no nominations of candidates. Each elector voted for such candidates as he pleased, provided they possessed the necessary qualifications--those of an ordinary elector. Three parties--the Moderate, Liberal, and Labour--contested the election. Each party printed ballot papers containing the names of the candidates adopted by the party organization and with the name of the party at the head of the ballot paper. The ballot paper issued by the Moderate party was in the following form:--
_De Moderata_
_Borgmaestaren_--O. Holmdahl. _Grosshandlaren_--N. P. Nordstroem. _Lasarettslaekaren_--R. Lundmark. _Disponenten_--H. Berggren. _Kommendoeren_--G. Lagercrantz. _Radmannen_--C. G. Ewerlof. _Chefsintendenten_--I. Neuendorff. _Kaptenen, friherre_--F. E. von Otter. _Underofficeren af 2: dra graden_--O. W. Stroemberg. _Folkskollaeraren_--H. E. Mattsson. _Byggmaestaren_--K. J. A. Johansson. _Handlanden_--Aug. Andren.
_The Poll._
The ballot papers could be obtained at the committee rooms on, or prior to, the day of election, and also on the day of election from party agents at the doors of the polling stations. Each elector took his ballot paper folded to the Burgomaster, or presiding magistrate, who endorsed the back with the number of votes to which the elector was entitled. The presiding magistrate was assisted by two others who checked the accuracy of the proceedings. The poll opened at 10 A.M., the proceedings were adjourned for lunch at 1 P.M., the poll was again opened during the afternoon and closed about 8 P.M. The counting took place next day when, as comparatively few electors took advantage of their right to vary the order of the names as printed on the ballot papers, the number of votes recorded for each candidate was easily ascertained. Nor did the varying values of the ballot papers present any great difficulty. A calculating machine made the necessary additions both quickly and accurately. In this election only one paper was spoiled,[3] and it was very obvious that the provision of printed ballot papers by the party organizations made the act of voting a very simple one. The votes recorded for the different parties were as follows:--
Moderate . . . . . 20,334 Liberal . . . . . 8,732 Labour . . . . . 3,617
_The allotment of seats to parties.
There were nine seats to be distributed among the three parties. The distribution was carried out in accordance the d'Hondt rule, but the method of applying this rule differed from that employed in Belgium. In Belgium the party totals would have been divided by the numerals 1, 2, 3, &c., and the quotients ranged in order of magnitude, the ninth in order being termed the "electoral quotient." Each party would have received as many seats as its total contained this quotient. The Swedish method provides for the allotment of one seat at a time, and it does so because of the possibility of the same candidate being elected by more than one party. Save in the rare case mentioned, the arithmetical operations, though differently presented, are identical with those of the Belgian system. Thus, at Carlskrona the first seat was given to the Moderates--that party having received the highest number of votes. Before the next seat was allotted the value of the Moderate total was reduced by one-half, and the new total was then compared with the original totals of the other parties. The totals to be considered in the allotment of the second seat were, therefore, as follows:--
Moderate. . . . . 10,167 Liberal . . . . . 8,732 Labour . . . . . 3,617
The Moderate party being still credited with the highest total received the second seat, and their original total, 20,334, was then divided by three in order to ascertain to whom the third seat should be allotted. The totals at this stage were as follows:--
Moderate . . . . . 6,778 Liberal . . . . . 8,732 Labour . . . . . 3,617
The Liberal total being now the highest, this party received the third seat, and in order to ascertain to whom the fourth seat should be given the Liberal total was reduced in value by one-half, the totals of the other parties remaining as at the previous allotment. The totals for comparison were now:--
Moderate . . . . . 6,778 Liberal . . . . . 4,366 Labour . . . . . 3,617
The Moderate total was again the highest, and the party received the fourth seat. The process of reducing the totals in succession according to the foregoing rule was continued until all the nine seats were allotted. In this election the Moderates obtained six seats, the Liberals two, and Labour one.
_The selection of the successful candidates._
The returning officer had then to determine which candidates on each list should be declared successful. In the Carlskrona election this task was extremely simple, for the large majority of the voters had accepted the ballot papers provided for them by their parties. No less than 19,756 votes out of a total of 20,334 had been received for the Moderate list as printed by the party organization. The totals for each candidate were quickly ascertained. Moreover, it was possible to select all the successful candidates by the rule of the order of preference. More than six-sevenths of the Moderate votes having been recorded for the list as printed, the first six names on the list were declared elected. Of the Liberal votes, 8118 out of a total of 8732 were recorded for the party list as printed, and as this number constituted more than two-thirds of the total, the first two names on the list were declared elected. With regard to the Labour party, 3580 out of a total of 3617 votes had been recorded for the party list, and the first candidate on the list was therefore declared elected.
_The election of suppleants.
In common with all continental systems, supplementary members (suppleants) were chosen for the purpose of taking the place of an elected member who might die or retire before the council had run its course. The method adopted in Sweden is peculiar to itself. In Belgium the same rules serve for the election of the suppleants as for the election of members, and they are called upon to serve in the order in which they stand at the declaration of the poll. In Sweden it is held that each elected member must have a suppleant, or deputy, special to himself. The method of selection may be illustrated from the Carlskrona election. The candidate who was to be regarded as suppleant to Burgomaster Holmdahl (the first on the Moderate list) was chosen as follows: Holmdahl had received 20,334 votes, his name having appeared on every ballot paper of the Moderate party; the votes recorded for the unelected candidates on these papers were ascertained, the result being:--
Neuendorfs . . . . . 20,334 von Otter . . . . . 20,242 Stroemberg . . . . . 19,913 Mattsson . . . . . 20,119 Johansson . . . . . 20,237 Andren . . . . . . 20,170
Neuendorff being the candidate who had received the highest number of votes on these papers, was declared elected as suppleant to Holmdahl. A suppleant for Nordstroem, the second elected member, was then chosen from among the remaining five non-elected members. Nordstroem's votes were 20,235, and the votes recorded for the non-elected members on the same papers were:--
von Otter 20,143 Stroemberg 19,913 Mattsson 20,055 Johansson 20,195 Andren 20,071
Johansson, being highest with 20,195 votes, was declared suppleant to Nordstroem.
This method of choosing the suppleant seems to be unsatisfactory. The party as such does not determine who shall be called upon to fill a vacancy in its ranks; whether a non-elected member succeeds to a vacancy as a suppleant depends very largely on accident. A good illustration occurred in the selection of a suppleant from the Labour list. The party's candidates were as follows:--
Kloo. Karlsson. Ostergren. Olsson. Ek. Johansson. Jensen. Fagerberg. Pettersson.
The first candidate on the list had been declared elected, and obviously, in the opinion of the party, the next favourite was Karlsson, and had there been a second seat awarded to the list Karlsson would have been declared elected. In determining, however, whether he should be declared elected as a suppleant, his position on the list did not count, and as the party list had been voted for without alteration by most of the Labour voters, five of the non-elected candidates were credited with the same number of votes. The choice of the suppleant was made by lot, and fell in this case upon Johansson, the sixth name on the list. It may be said that there is; considerable dissatisfaction with the method of electing suppleant candidates, and the Stockholm _Dagblad_, in its issue of the 29 May 1910, stated that the choice of suppleant, although there might have been many thousand votes given to every candidate, depended upon so small a difference in the totals received by each that even one ballot paper might determine the result. This is a detail in the system that can easily be remedied, and steps are already being taken to bring the election of suppleants into agreement with the election of ordinary members.
_Comparison with Belgian system._
It will be of interest to compare the Swedish with the Belgian system. It has been shown that the method of allotting seats to different groups is identical in principle in both countries. This method, the d'Hondt rule, favours the largest parties, and this explains why, in the smaller Belgian constituencies, cartels or combinations of parties take place. The Swedish system enables such combined action to take place with greater facility. It enables two parties to make use of the same motto without presenting a common list of candidates. No inter-party negotiations are required, as in Belgium, with reference to the order in which the names of candidates shall appear upon the list. In Sweden each group can put forward its own list of candidates, and so long as the electors make use of the same motto at the head of the ballot paper the combination gains the additional representation which may fall to it as a result of being treated as one party, whilst the share falling to each section is determined by the number of votes recorded for their respective candidates.
The Swedish method of choosing the successful candidates from the various lists differs materially from that used in Belgium. In Sweden the d'Hondt rule is used not only for the allotment of seats to parties, but also in the selection of the successful candidates. In Belgium the use of the d'Hondt rule is restricted to the former purpose, and when once the electoral quotient is ascertained the rule is discarded. The difference in the two methods can be illustrated from the Stockholm municipal election of 1910. In the fifth ward the ballot paper of the Moderate party was as follows:--
Welin. Norstrom. Boalt. Roberg. Palmgren. Bohman. Ringholm. Herlitz. ------------------ Hafstrom. Svensson. von Rosen. Freden.
The line in the ballot paper divides the eight candidates for election as members from those who were standing for election as suppleants only. The votes recorded for the Moderate party numbered 118,483, of which 86,851 were given for the party ticket as printed. The number of votes accepting the party order of the first three candidates was about 93,000. This latter number was more than three-fourths, but less than four-fifths of the total, and therefore only the first three candidates on the ballot paper could be declared elected in accordance with the rule of the order of preference. The remaining four members had to be chosen by the reduction rule; the votes recorded for the five non-elected candidates were ascertained, the papers containing the names of the three elected candidates being treated for this purpose as of the value of one-fourth.
Some of the supporters of the eighth and sixth candidates had struck out the names of the fourth and other candidates. This manoeuvre had the result of placing these two candidates in the order named at the head of the poll at the fourth and fifth counts, and they were accordingly elected. Other candidates had received exclusive support, and it should be pointed out that it is the total amount of exclusive support recorded for all candidates which determines how soon the application of the rule of the order of preference breaks down. As soon as this takes place the election of any one candidate may depend, as in the election of the suppleants, upon the action of a comparatively small number of voters. Thus, some supporters of the fifth candidate, a Miss Palmgren, had struck out the names of all candidates save hers. Those papers which contained her name alone were treated as of full value, and although the votes of these supporters only numbered 1100, or less than 1 per cent. of the whole, they were sufficient to turn the scale in her favour. As, however, 86,851 votes out of a total of 118,453, had been recorded for the list as printed, showing that this proportion of voters preferred the fourth candidate to those that succeeded him, it would certainly seem that the result was not fair to this candidate. In Belgium if seven seats were won by a party which polled 118,453 votes, the electoral quotient would not be more than one-seventh of this total, and the election of the first candidate, instead of absorbing one-half the value of the votes, would consume only one-seventh. The election of the first two candidates would absorb two-sevenths instead of two-thirds, the election of three candidates would consume three-sevenths instead of three-fourths, and the election of four candidates would consume four-sevenths instead of four-fifths. In the Stockholm election more than five-sevenths of the voters had supported the party list as it was printed, and according to the Belgian system the first five candidates would have been declared elected.
_The system and party organization_.
The Swedish rule of selecting successful candidates is defended on the ground that it confers great power upon the electors. These can if necessary more effectively express their disapproval of the list put forward by the party organization, and as it is thought that a large number of voters too readily accept the party lead, a counterpoise is considered desirable. Recent experience in Belgium, however, would tend to show that a greater knowledge of their power has induced more and more electors to make use of the opportunity which that system allows of expressing individual preferences. If we regard a party as consisting of two groups--those that follow the party lead, and those which, whilst supporting the party, desire to assert their own preferences--then as between these two groups the Belgian system is strictly fair. If a party wins seven seats and four-sevenths of the party support the official list, this group would obtain four out of the seven seats; but in Sweden, as has been shown, at least four-fifths must support the official list before the first four candidates can be sure of election. The Swedish system discriminates in favour of the dissentients within a party, and this discrimination may have unexpected effects on party organization. The Belgian method has induced parties to welcome the support of all sections, knowing that such sections will not obtain more than their fair share of influence. In Sweden the tendency may be for party organizers to regard the support of various sections with suspicion, because, whilst these sections will obtain the full advantage of the party vote, their independent action may result in the gain of the section at the expense of the party as a whole. As a result of the Stockholm election referred to, the opinion was expressed by party organizers that it would be necessary to limit the number of candidates on a list to the number which the party knew it could carry. This would be an undesirable outcome of a rule designed to secure greater freedom for the elector, for it would tend to make party discipline more strict and parties exclusive rather than inclusive, as is the case in Belgium. It should, however, be added that in the large majority of the provincial council elections the selection of candidates was made in accordance with the rule of the order of preference. It would, therefore, seem that party organizers, as a rule, took care to present lists of candidates acceptable to the party as a whole.
_The great improvement effected by the Swedish system_.
The new Swedish electoral system, like all proportional systems, constitutes a striking advance upon the previous electoral conditions. The extent of the improvement will, of course, be seen from a comparison of some of its results with those of former years. For example, Stockholm used to be represented in the Lower Chamber by twenty-two members chosen by the "block" system, or _scrutin de liste_. The party in the majority monopolized the representation, and the absurdity of the system was well illustrated by an incident in the election of 1882, which was preceded by a severe struggle between the advocates of free trade and protection. At this election Stockholm returned twenty-two free traders, but as one of the elected members had not paid his taxes, all the voting papers containing his name were declared to be invalid. In consequence the twenty-two free traders were unseated and the twenty-two protectionist candidates were declared elected in their place. An attempt was made to ameliorate the evils of this system by dividing the town into five parliamentary districts, but, although so divided, Stockholm in 1908 returned twenty-one members, all of whom were either Liberals or Socialists, the large minority of Moderates being unrepresented. When the proportional system was applied in March 1910 to the election of the municipal council, each party obtained its fair share of representation in each of the six wards of the city, and the total result shows how large an improvement is effected by the new method:--
Parties. Votes Seats Seats in Obtained. Obtained. Proportion to Votes. Moderate 281,743 22 24 Liberal 142,639 12 12 Socialist 160,607 16 14 ----------------------------------- 584,989 50 50
In the election of the provincial council of Bleking the result was as follows:--
Parties. Votes Seats Seats in Obtained. Obtained. Proportion to Votes. ------------------------ ----------------------- Moderate 54,465 22 22.4 Liberal 36,595 10 15.1 Socialist 3,617 1 1.5 ---------------------------------- 94,677 39 39
The general fairness of these results is all the more remarkable, because in Stockholm there was a very considerable variation in the value of a vote in the different wards, whilst many of the constituencies in the province of Bleking returned only a few members, and these did not give full play to the proportional system. The figures confirm the experience of all other countries, that a proportional system, even when applied to comparatively small constituencies, yields results which approximate very closely to the ideal aimed at, the true representation of the electors.
[Footnote 1: The town councils were elected in one stage; each elector had one vote for every 100 kroner income, subject to a limit of 100 votes. The members of the town council, when electing members of the provincial councils, had only one vote each.]
[Footnote 2: A ballot paper is not declared invalid even if it contains the names of more candidates than there are members to be elected (except at the elections of parliamentary committees). The names in excess are regarded as suppleant candidates (see _Election of Suppleants_) to the number of two in the elections for the Riksdag and the town councils, and to a number equal to the number of members at the election for the provincial councils. Any additional names on a ballot paper are regarded as non-existent.]
[Footnote 3: This paper bore the signature of the elector.]
APPENDIX IV
THE FINLAND SYSTEM OF PROPORTIONAL REPRESENTATION
_The influence of the Belgian system._
The system of proportional representation introduced into Finland by the electoral law of 1906, while it presents little or no difficulty to the voter, is, in its method of counting the votes, perhaps the most complicated of the systems at present in force. It has for its basis the Belgian List system and the d'Hondt rule, but the variations which were introduced with the object of safeguarding the rights of the electors against the possible tyranny of party managers are so important that at the first glance its resemblance to the parent system is not easily recognized. The Belgian model is followed more closely in the method of distributing the seats to the various parties than in the manner in which the successful candidates are chosen from the party lists. In its internal party arrangement the Finnish system shows boldness, originality, and, it must be added, no little complexity of procedure.
_Schedules and "compacts" in place of lists._
Finland is divided into sixteen electoral districts returning from six to twenty-three members, with the one exception of Lapland, which is a single-member constituency. In each constituency any group of not less than fifty electors can put forward a schedule of not more than three candidates, however many may be the total number of members to be elected. Each of these schedules may be headed with the name of a party or some political motto. The persons responsible for these schedules may, and commonly do, combine them in groups known as "compacts," and it is these compacts, and not the original schedules, which correspond roughly to the party "lists" of the Belgian system, the only limit to this power of combination being that the combined schedules must not contain the names of more candidates than there are vacancies to be filled. But as the names of the same candidates may, and constantly do, occur in many different schedules within a single compact, a first glance at a Finnish polling paper would seem to show in each combination the names of more candidates than there are vacancies. The compact bears the name of the political party to which it belongs. Combination into compacts is, of course, optional, and a certain number of schedules are put forward independently. A vacant corner is reserved on the ballot paper where any elector who is not content with any of the schedules submitted may make his own schedule.
_An election in Nyland_.
The system may be more fully understood from some details of the election of 1907 in the Nyland division. In this division, the largest in Finland, returning twenty-three members, no less than seventy-two schedules were presented, or which all except five were combined into compacts. The five remained isolated. Of the combined schedules seventeen were included in the compact of the Swedish party, but the individual candidates in these seventeen schedules numbered only twenty-three, the legal limit, the same names being repeated in several schedules. The old Finnish compact contained thirteen schedules, the Young Finns seventeen, the Social Democrats eight, the "Christian" compact seven, the "Free Christian" compact three, and the Radicals two.
As already stated, the voter's task is not difficult. He, or she, simply marks the schedule of his, or her, choice. The voter can also, if he wishes, alter the order of the names in a schedule. The effect of doing this will be apparent in a moment. That the task is simple is conclusively shown by the fact that the percentage of spoilt votes was in the Nyland division only 0.58 per cent. For the whole country the percentage was only 0.93, and this with universal adult suffrage and a poll of 899,347, or 70.7 per cent, of the electorate.
_The returning officer's task_.
The task of the returning officer is twofold. He has to ascertain (1) the relative positions of candidates within each compact (or independent schedule), and (2) their position relatively to the candidates of other compacts in the final allotment of seats. He proceeds as follows. He first counts the votes on each schedule, reckoning a full vote to the first name, a half vote to the second, and a third of a vote to the third (the effect of an alteration of the order of names in a schedule by the voter is now apparent). Thus if schedule No. 1 (in the specimen ballot paper on page 323), containing the names Schybergson, Neovius, and Soderholm, receives the support of 6000 voters in all, of whom 3000 have placed Schybergson as No. 1, 2000 as No. 2, and 1000 as No. 3, Schybergson will have a total of 3000 + 2000/2 + 1000/3 = 4333. Similarly, if Neovius obtains the support of 2000 as No. 1, 2000 as No. 2, and 2000 as No. 3, his total will be 2000 + 2000/2 + 2000/3 = 3666; Soderholm, the third candidate, would receive 1000 votes as No. 1, 2000 as No. 2, and 3000 as No. 3, and his total would be 1000 + 2000/2 + 3000/3 = 3000. But these individual totals of 4333, 3666, and 3000 are used merely to determine the order of the candidates within the schedule itself, and having performed that function, they are not taken further into account. In the example given (as would usually be the case in practice) the order within the schedule has not been disturbed, and the candidates are credited, the first (Schybergson) with the full number of the voters who supported the schedule--6000; the second (Neovius) with one-half that number--3000; the third (Soderholm) with one-third of that number--2000. These last figures are called "numbers of comparison," a phrase intended to throw light upon their function. The same process is gone through with all the other schedules in the same compact. The returning officer then adds up all the numbers of comparison which each candidate has obtained in all the schedules within the compact where his name appears, and arranges candidates within the compact in the order of these totals. Thus, in the actual election of 1907, in the Nyland division, Schybergson headed the Swedish party compact with 9192 as the total of his "numbers of comparison," Soderholm coming next with 6837.
_The allotment of seats_.
When the candidates in each compact have thus been arranged in order (and the votes given in writing by independent voters have also been counted), the returning officer proceeds to the second stage of his duties--the determination of the position of candidates with reference to their competitors in other compacts; and it is on this position that the actual allotment of seats depends. For this purpose he primarily takes into account, not the "numbers of comparison" of individual candidates, but the total number of voters who have supported each compact; he credits this total to the candidate who has the highest "number of comparison" within the compact; credits the next candidate with one-half this total, the third candidate with one-third, and so on, finally arranging the whole of the candidates in order. Thus far this stage of the process is identical in substance with the Belgian method, though the appearance is different. For, obviously, if List (or compact) A, of which the candidates are G, H, I, in that order receives 12,000 votes, while List B, with candidates P, Q, R, receives 10,000, and List C, with candidates X, Y, Z, receives 8000, it is all one whether the returning officer applies the d'Hondt rule and assigns two seats to List A (thus seating G and H), two seats to List B (thus seating P and Q), and one seat to List C (thus seating X), or whether he tabulates the result of the polling thus:
G 12,000 \ P 10,000 | X 8,000 > Elected. H 12,000/2 i.e. 6,000 | Q 10,000/2 i.e. 5,000 / Y 8,000/2 i.e. 4,000 Not elected, and so on.
But at this point a characteristic feature of the Finnish system comes into play. Candidates' names may occur in more than one compact, and may be found in isolated schedules, or on the written papers of independent voters as well. Consequently their final order cannot be determined by this simple application of the Belgian method. The returning officer must[1] add to the number of votes credited to a candidate of any one compact such additional votes as he may have obtained either as a member of another compact or from independent voters. Thus, in the Nyland elections, Miss Sohlberg, whose name will be found at the head of Schedule 48 within the Swedish compact, obtained the eleventh place within that compact. The total number of voters supporting this compact was 44,544, and Miss Sohlberg was therefore credited with an eleventh of this total, or 4049 votes. But Miss Sohlberg's name also occurred in Schedules 62 and 63 in the "Free Christian" compact and Schedule 21 in the "Christian" compact, and as her share of the votes of these compacts she received 153 and 325 respectively. She also received four votes in writing. Thus her final total was 4049 + 153 + 325 + 4, or 4531 in all, and it was this number which determined her position on the poll.
_Successful candidates in the Nyland election._ This explanation will perhaps be more comprehensible if the actual result of the polling in the Nyland division, so far as the first 25 candidates are concerned, is given in a tabular form:--
Final Names of Party. Number of Additional Final Order Candidates. Votes resulting Votes. Total. of from Place of Poll. Candidates on Compact. 1 Schybergson Swedish 44,544 2.33 44,546.33 2 Haeninan Social Dem. 40,951 6.5 40,957.5 3 Soderholm Swedish 22,272 0.33 22,272.33 4 Sillanpaeae Social Dem. 20,475.5 8.83 20,484.33 5 Kaekikoski Old Finn 20,402 9.33 20,411.33 6 Oljemark Swedish 14,848 -- 14,848 7 Siren Social Dem. 16,650.33 2.33 16,652.66 8 Rosenquist (G.) Swedish 8,908.8 2,932.83[2] 11,841.63 9 Rosenquist (V.) Swedish 11,136 4.33 11,140.33 10 Helle Social Dem. 10,237.75 3 10,240.75 11 Palmen Old Finn 10,201 8.83 10,209.83 12 Pertillae (E.) Social Dem. 8,190.2 4.67 8,194.87 13 Ahlroos Swedish 7,424 1 7,425 14 Pertillae (V.) Social Dem. 6,725.17 1.5 6,726.67 15 Reima Old Finn 6,800.67 5.67 6,806.34 16 Erkko Young Finn 6,521 6.32 6,527.32 17 Ehrnrooth Swedish 6,363.43 75.83 6,439.26 18 Laine (M.) Social Dem. 5,850.14 4 5,854.14 19 Wasastjerna Swedish 5,568 -- 5,568 20 Ingman Social Dem. 5,118.88 3.5 5,122.38 21 Laine (O.) Old Finn 5,100.5 -- 5,100.5 22 von Alfthan Swedish 4,949.33 -- 4,949.33 23 Johansson Social Dem. 4,550.11 1.33 4,551.44 (All the above were elected.) 24 Sohlberg Swedish 4,049.45 482.45[3] 4,531.9 25 Gustaffsson Swedish 4,454.4 4.5 4,458.9 &c. &c.
_Equitable results._
It will to some extent be gathered from the foregoing table that the total number of the supporters of the various compacts or parties in the Nyland division and the number of seats won were as follows:
Seats Seats in Parties. Votes. Actually Proportion Won. to Votes. Swedish 44,544 9 8.7 Social Democrat 40,951 9 8.0 Old Finn 20,402 4 4.0 Young Finn 6,521 1 1.3 "Christian" compact 2,932 - .6 "Free Christian" 458 - .1 Radical 168 - - Isolated schedules 1,356 - .3
Total 117,332 23 23.0
The result is thus in reasonable correspondence with the demands of a strictly proportionate allotment of seats; this statement is also true of the results for the whole of Finland, as the following table will show:--
Seats Seats in Parties. Votes. Actually Proportion Won. to Votes. Social Democrat 329,946 80 74.1 Old Finn. 243,573 59 54.7 Young Finn 121,604 26 27.3 Swedish 112,267 24 25.2 Agrarian 51,242 9 11.5 Christian Labourer 13,790 2 3.1 Minor groups 18,568 - 4.1
Total 890,990 200 200.0
An exactly mathematical distribution is, of course, not to be expected from this, any more than from any other method which does not adopt the system of treating a whole country as a single constituency. As to the mechanism of the system it only remains to add that the process of counting was found to be very lengthy. In the Nyland division, where the results were ascertained sooner than in any other case, the elections were held on 15 and 16 March, but the result was not announced until the 2 April. To people accustomed to the greater rapidity of ordinary electoral methods this will seem a serious drawback. Possibly improved arrangements may shorten this long interval between the elections and the announcement of the result.
It would obviously be premature to attempt to estimate the political effects of the Finnish system as compared with other systems of proportional representation.
_Elector's freedom of choice._
The Finnish system has been in operation since 1907, and the whole political circumstances of Finland have undergone so many striking changes, and so many new factors are at work that to disentangle particular causes and effects is an impossibility. But plainly the Finnish machinery gives a greater freedom to the elector than the Belgian system. The Finnish system in fact encourages the electors to arrange the candidates of a party in the order preferred by the electors themselves, and not in the order dictated by the party managers. There is no "party ticket" for which the elector can vote blindfold. He must choose the schedule that he prefers; he can even rearrange that schedule, or, if he chooses, can make one of his own. No doubt the schedule itself is ready made for him, but it contains three names only, and is not the equivalent of the Belgian "list." On the other hand, the elector who chooses to vote for a schedule within a compact adds, whether he likes it or not, to the total votes of the compact, and so may help to return not the candidate of his choice, but the candidates preferred by the majority of the party with which he is in sympathy. An illustration of this fact may be taken from the Nyland poll. The old Finnish party were alive to the possibilities of the situation, and combined their lists with great skill so as to attract votes. They placed their favourite candidates in nearly every schedule, but not at the head of the schedule. At the head of the schedule they placed some man of local popularity, usually a peasant proprietor, whose name was not repeated in many, if any, other schedules. Thus the local favourite attracted votes to the schedule, but in the race for the highest numbers of comparison the candidates whose names appeared on few schedules were left behind those whose names appeared on many schedules even in the lower places.
A portion of the official ballot paper showing the compact put forward by the Swedish People's Party is printed on the opposite page. In one corner of the ballot paper was a blank schedule in the following form.
THE ELECTOR who does not approve of any of the preceding lists should write here the names of his candidates in the order in which he wishes them to be elected.
CANDIDATES
_Name_....................................................
_Profession or Occupation_................................
_Address_.................................................
_Name_....................................................
_Profession or Occupation_................................
_Address_.................................................
_Name_....................................................
_Profession or Occupation_................................
_Address_.................................................
FINLAND GENERAL ELECTION, 1907
Part of Ballot Paper--Nyland Division.
The Voters' Compact of the Swedish People's Party.
1 HELSINGFORS. Experienced Members of the Diet:-- --Schybergson, E. K. --Neovius, A. W. --Soderholm, K. G.
33 EAST NYLAND-LOUISA. Justice and Progress:-- --Rosenquist, G. G. --Stromberg, J. --Ehrnrooth, L.
34 MID-NYLAND-NIOKBY. The Welfare of the Rural Population;-- --Topelius, G. L. --Alfthau, K. von --Rosenquist, B. T.
35 MID-NYLAND-ESBO. The Welfare of the Rural Population:-- --Wasastjerna, O. --Schybergson, E. --Soderholin, K.
36 WEST NYLAND-KYRK-SLATT. The Welfare ol the Rural Population:-- --Nordberg, G. --Ehrnrooth, L. --Oljemark, K. T.
37 WEST NYLANB-EKENAS. The Welfare of the Rural Population. Law and Justice:-- --Oljemark, K. T. --Schybergson, E. --Soderholm, K.
38 BORGA. Knowledge and Experience:-- --Runeberg, J. W. --Bjorkenheim, G. --Rosenquist, G. G.
39 HELSINGFORS. Sound Development of the Community;-- --Westermarck, Helena. --Rosenquist, B. T. --Bjorkenheim, G.
40 HELSINGFORS. Law and Justice:-- --Sorterholm, K. --Alfthan, K. von --Westermarck, Helena,
41 HELSINGFORS. Legality and Progress:-- --Westermarck, Helena. --Neovius, A. --Ehrnrooth, L.
42 HELLSINGFORS. Swedish Culture:-- --Rosenqnist, B. T. --Gustafsson, F. prof. --Soderholm, K.
43 HELSINGFORS. Friends of Labour and of the People:-- --Alfthan, K. von --Gustafsson, F. prof. --Gronroos, F.
44 HELSINGFORS. Experience and Practical Knowledge:-- --Runeberg, J. W. --Schybergson, E. --Neovius, A.
45 HELSINGFORS. The Labourers' Welfare:-- --Ahlroos, F. --Holmberg, W. --Ehrnrooth, L.
46 HELSINGFORS. Commerce and Industry: --Heimburger, W. F. --Bjorkenheim, G. --Schybergson, E.
47 THE SKERRIES OF NYLAND: Navigation and Fisheries:-- --Hjelt, Th. --Renter, O. --Alfthan, K.
48 THE PROVINCE OF NYLAND: HELSINGFORS. Temperance, Morality and Popular Education:-- --Sohlberg, H. --Ahlroos, F. --Rosenquist, G. G.
[Footnote 1: This right of addition is subject to a limit. The reinforcements must not raise a candidate's total above what he might obtain if the votes given to all compacts or lists, where his name occurs, were divided by the figure which indicates his order within the compact from which he derives his principal strength.]
[Footnote 2: This large reinforcement of votes came from the Christian compact, where this candidate's name appeared as well as in the Swedish compact.]
[Footnote 3: See reference to Miss Sohlberg in preceding paragraph.]
APPENDIX V
THE STATISTICS OF THE GENERAL ELECTIONS, 1885-1910
The following tables are taken, with permission, from a paper read on 12 December 1906, by Mr. J. Rooke Corbett, M.A., before the Manchester Statistical Society, of which a second and revised edition was published in April 1910 by the Proportional Representation Society.
In these tables the totals for England, Wales, and Monmouth, Scotland and Ireland are shown separately, and the figures for England have been further subdivided according to the ten divisions into which the kingdom is divided by the Registrar General for the purpose of his work.
These ten subdivisions are as follows:
Metropolitan-- London. South East-- Surrey. Kent. Sussex. Hampshire. Berkshire. South Midland-- Middlesex. Hertfordshire. Buckinghamshire. Oxfordshire. Northamptonshire. Huntingdonshire. Bedfordshire. Cambridgeshire. East-- Essex. Suffolk. Norfolk. South-West-- Wiltshire. Dorsetshire. Devonshire. Cornwall. Somersetshire. West Midland-- Gloucestershire. Herefordshire. Shropshire. Staffordshire. Worcestershire. Warwickshire. North Midland-- Leicestershire. Rutlandshire. Lincolnshire. Nottinghamshire. Derbyshire. North-West-- Cheshire. Lancashire. Yorkshire-- West Riding. East Riding (with York). North Riding. Northern Division-- Durham. Northumberland. Cumberland. Westmorland.
The first three columns, A, B and C, show the number of members allotted to these several divisions, the number of registered electors, and the number of members to which each division would be entitled if the 670 members of which the House of Commons is composed were divided among the several divisions in proportion to their electorates.
In taking the electorate as the basis of a proportionate redistribution of seats it is not intended to prejudge the question whether population or electorate is the better standard. The electorate has been taken because the figures are available for the very year in which the election takes place, whereas the population is only enumerated once in ten years.
The columns D and E show in two groups the number of members elected for these divisions, Liberal, Labour, and Irish members being gathered together in one column, Conservatives alone occupying the other.
It is one of the disadvantages of our present system of representation that it makes it quite impossible to ascertain the relative strength of the several parties into which the voters are divided. In the great majority of contests there is a Liberal, Labour, or Irish Nationalist candidate on one side, and a Unionist candidate on the other, and there is practically no evidence as to how many of the supporters of either candidate belong to each of the parties concerned. Any estimate of the relative strength of the Liberal and Labour parties or of the Unionist Free Traders, and Tariff Reformers must be largely a matter of guesswork. All that is possible, therefore, is to divide the voters into two groups, as has been done in these tables.
The columns F and G show the total electorate of the constituencies held respectively by the two groups of members shown in columns D and E.
The figures in these two columns are of value in showing the probable result of a scheme of redistribution. The South-Eastern counties may be taken as an example. These are at present represented by 48 members. The Liberals held three constituencies in January 1910 containing an electorate of 31,221 (columns D and F); the Conservatives held 45 constituencies containing an electorate of 604,887 (columns E and G). If a redistribution of seats was made on the basis of equal electorates, the South-Eastern counties would be entitled to 55 members (column C). It may be assumed that in any rearrangement of constituencies the parties would retain their predominance in the areas which they now represent, and if so the result of a rearrangement of constituencies on the basis of equal electorates would be that in January 1910 the Conservatives would have obtained 52 seats and the Liberals 3 (column K). Similarly in the General Election of 1906 the Liberals in Wales and Monmouth held 34 seats, the Conservatives none. If the constituencies had been rearranged, the Liberals would have held 35 seats, the Conservatives none. The majorities throughout the United Kingdom which would be obtained under a scheme of equal electorates are shown in column K.
The columns H and I show the number of electors who voted for the candidates of the two groups; Liberal, Labour, and Irish Nationalist voters in one group, Conservative voters in the other.
In computing the figures in these columns an allowance has been made for uncontested constituencies on the following basis. It has been supposed that the changes of public opinion which affect the contested constituencies affect uncontested constituencies also, and in estimating the number of voters in an uncontested constituency it has therefore been assumed that the strength of each party varies from one election to another in the same ratio as in the contested constituencies in the same county.
The three columns J, K and L show respectively the actual majorities obtained, the majorities which would have been obtained if the country had been divided into single-member constituencies of equal size, and the majorities under a system of proportional representation.
The figures in the last two columns have been calculated with reference to the totals in column C, which gives the number of members to which each division would be entitled on a proportional basis.
In order to ascertain the figures given in column K _(i.e._ the probable results with equal single-member constituencies) it has been assumed, as already explained, that the two groups would, after the redistribution of seats, be predominant in the same areas as before the rearrangement.
_The representation of minorities._
The tables give abundant evidence of the anomalies associated with our electoral system. One of the most striking is the great difference in the amount of representation secured by minorities in different parts of the country. The amount of representation secured by a minority has not depended upon its size, but upon the way in which it has been distributed. The following table shows the amount of representation obtained by important minorities in the General Election of January 1910:--
THE REPRESENTATION OF MINORITIES, ELECTION JAN. 1910
Size of Seats Total Seats Area. Minority. Obtained. for Whole Area Ireland . . . . . . . 145,437 21 103 Scotland . . . . . . . 265,770 11 72 S. East: Counties. . . 220,995 3 48 Wales and Monmouth . . 116,696 2 34 Northern Counties . . 75,897 9 32
The figures show that in Ireland a minority of 145,437 obtained twenty-one representatives, whilst a minority of 116,696 in Wales and Monmouth obtained only two. The good fortune which befel the minority in Ireland, not only in the elections of 1910 but in all the elections since the Redistribution Bill of 1885, has been due to the fact that this minority is concentrated in one corner of Ireland and can transform itself into local majorities. The larger minority in Scotland, owing to its distribution throughout the country, obtains much less representation; the minorities in the south-eastern counties of England and Wales are also distributed throughout these two areas and likewise suffer. The minority of 75,879 in the northern counties being less evenly diffused was more fortunate, and obtained nine representatives. The figures for the election of December 1910 disclose similar anomalies.
GENERAL ELECTION, 1885
Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 489,396 57 LLI 22 165,345 162,228 Con 38 324,051 188,067 16 19 3 England South-East 48 406,955 47 LLI 4 34,883 144,659 Con 44 372,072 187,831 40 39 7 S.Midland 38 312,477 36 LLI 14 123,665 124,717 Con 24 188,811 129,544 10 8 East 29 257,022 29 LLI 18 173,521 107,710 7 11 1 Con 11 83,501 98,137 South-West 40 314,603 36 LLI 27 229,612 144,273 14 16 4 Con 13 84,991 117,442 W.Midland 58 544,415 63 LLI 45 427,549 248,825 32 36 8 Con 13 116,866 198,212 N.Midland 34 328,844 38 LLI 26 255,836 55,503 18 22 4 Con 8 73,008 120,933 North-West 70 654,751 76 LLI 24 231,123 263,670 Con 46 423,628 292,942 22 22 4 Yorkshire 52 536,553 62 LLI 36 398,426 248,078 20 30 8 Con 16 138,127 189,930 20 30 8 North 32 305,015 35 LLI 25 262,287 144,803 18 25 5 Con 7 42,728 96,708 ENGLAND 461 4,150,031 480 LLI 241 2,302,248 1,740,466 21 52 16 Con 220 1,847,783 1,619,746 Wales and Monmouth 34 286,145 33 LLI 30 263,199 149,782 26 27 11 Con 4 22,946 79,006 Scotland 72 576,828 67 LLI 58 485,116 289,032 44 45 15 Con 14 91,712 181,706
Britain 567 5,013,004 580 LLI 329 3,050,563 2,179,230 91 124 42 Con 238 1,962,441 1,880,458 Ireland 103 777,954 90 LLI 85 624,760 404,892 67 54 44 Con 18 153,194 139,273
Total 670 5,790,958 670 LLI 414 3,675,323 2,584,122 158 178 86 Con 256 2,115,635 2,019,731
Majority 158 1,559,638 564,391
NOTE.--The figures in columns K and L are calculated with reference to the totals in column C. Thus the figure L 54 for Ireland in column K of the last section of the table indicates that under a system of equal single-member constituencies Ireland's 90 members would be Liberal etc. 72, Unionist 18, a Liberal majority of 54, and the corresponding figure L 44 in column L indicates that under proportional representation the 90 members which Ireland would return would be Liberal etc. 67, and Unionist 23. a Liberal majority of 44.
GENERAL ELECTION, 1886
Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 489,396 57 LLI 11 87,974 125,457 Con 49 401,422 185,072 38 37 11 England-- South-East 48 406,955 47 LLI 0 - 114,518 Con 48 406,955 184,221 48 47 11 S.Midland 38 312,477 36 LLI 9 73,292 94,213 Con 29 239,185 128,339 20 20 6 East 29 257,022 29 LLI 4 87,975 81,838 Con 25 219,047 102,732 21 21 3 South-West 40 314,603 36 LLI 7 63,063 96,753 Con 33 251,540 129,056 26 22 6 W.Midland 58 544,415 63 LLI 15 136,518 173,463 Con 43 407,897 218,753 28 32 8 N.Midland 34 328,844 38 LLI 14 147,138 125,078 Con 20 181,706 126,547 6 4 North-West 70 654,751 76 LLI 13 123,459 236,134 Con 57 531,292 282,187 44 48 6 Yorkshire 52 536,553 62 LLI 33 359,414 214,407 6 Con 19 177,139 180,728 14 22 North 32 305,015 35 LLI 23 247,275 123,901 5 Con 9 57,740 96,404 14 21 ENGLAND 461 4,150,031 480 LLI 129 1,276,108 1,385,762 Con 332 2,873,923 1,634,039 203 188 42
Wales and Monmouth 34 286,145 33 LLI 27 240,752 123,186 20 23 7 Con 7 45,393 82,179 Scotland 72 576,828 67 LLI 43 339,726 218,561 14 11 5 Con 29 237,102 188,164
Subtotal 567 5,013,004 580 LLI 199 1,856,586 1,727,509 Con 368 3,156,418 1,904,382 169 154 30
Ireland 103 777,954 90 LLI 84 616,735 376,445 Con 19 161,219 144,755 65 52 38
Total 670 5,790,958 670 LLI 283 2,473,321 2,103,954 8 Con 387 3,317,637 2,049,137 104 102
Majority 104 844,316 54,817
GENERAL ELECTION, 1892
Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 552,024 60 LLI 23 186,572 183,967 Con 37 365,452 214,275 14 20 4 England: South-East 48 463,073 50 LLI 4 38,534 147,136 Con 44 424,539 206,075 40 42 8 S.Midland 38 340,650 38 LLI 15 139,228 120,844 Con 23 210,422 147,347 8 8 4 East 29 276,491 30 LLI 13 134,632 108,866 Con 16 141,859 110,849 3 South-West 40 325,769 35 LLI 15 136,061 125,392 Con 25 189,708 136,449 10 5 1 W. Midland 58 577,397 63 LLI 16 143,567 204,453 Con 42 433,830 248,774 26 31 7 N. Midland 34 347,482 38 LLI 22 232,970 145,587 10 14 2 Con 12 114,512 130,380 North-West 70 707,392 77 LLI 26 284,970 282,139 Con 44 422,422 307,698 18 15 3 Yorkshire 52 571,864 62 LLI 35 418,414 244,099 18 28 6 Con 17 153,450 204,492 North 32 328,189 36 LLI 25 264,483 143,172 18 22 4 Con 7 63,706 115,626 ENGLAND 461 4,499,331 489 LLI 194 1,979,431 1,705,655 Con 267 2,519,900 1,821,985 73 57 15
Wales and Monmouth 34 314,063 34 LLI 31 294,395 152,326 28 30 10 Con 3 19,668 86,576 Scotland 72 606,203 66 LLI 52 449,994 267,631 32 32 8 Con 20 156,209 214,448
Subtotal 567 5,419,497 589 LLI 277 2,723,820 2,125,612 5 3 Con 290 2,695,777 2,123,009 13 Ireland 103 746,781 81 LLI 80 561,938 345,548 57 41 31 Con 23 184,843 157,181
Total 670 6,168,388 670 LLI 357 3,285,758 2,471,164 44 46 34 Con 313 2,880,620 2,280,190 Majority 44 405,138 190,974
GENERAL ELECTION, 1895
Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 573,141 61 LLI 8 70,056 161,328 Con 52 503,085 242,999 44 47 13 England: South-East 48 472,725 50 LLI 2 24,057 152,213 Con 46 448,668 217,096 44 44 8 S.Midland 38 358,501 38 LLI 3 30,569 116,143 Con 35 327,932 164,052 32 32 6 East 29 294,153 31 LLI 8 70,467 101,736 Con 21 223,686 122,999 13 15 3 South-West 40 330,670 35 LLI 10 76,141 124,852 Con 30 254,529 144,435 20 19 3 W.Midland 58 589,881 63 LLI 9 85,544 195,545 Con 49 504,337 259,382 40 45 9 N.Midland 34 351,792 37 LLI 16 186,167 143,142 1 Con 18 165,625 149,436 2 1 North-West 70 728,292 78 LLI 10 114,035 273,585 Con 60 614,257 332,101 50 54 8 Yorkshire 52 565,799 61 LLI 28 317,932 238,032 4 7 1 Con 24 247,867 225,871 North 32 339,289 36 LLI 20 222,202 145,085 8 12 2 Con 12 117,087 124,697
ENGLAND 461 4,604,243 490 LLI 114 1,197,170 1,652,261 Con 347 3,407,073 1,983,068 233 236 48 Wales and Monmouth 34 320,532 34 LLI 25 241,750 148,552 16 18 6 Con 9 78,782 108,036 Scotland 72 636,106 68 LLI 39 335,143 243,425 6 4 2 Con 33 300,963 234,138
Subtotal 567 5,560,881 592 LLI 178 1,774,068 2,044,238 Con 389 3,786,818 2,325,242 211 214 40
Ireland 103 727,562 78 LLI 82 549,467 317,910 61 42 28 Con 21 178,095 154,379
Total 670 6,292,443 670 LLI 260 2,323,530 2,362,148 Con 410 3,964,913 2,479,621 150 172 12 Majority 150 1,641,383 117,473
GENERAL ELECTION, 1900 Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 601,925 60 LLI 8 73,718 150,047 Con 52 528,207 247,777 44 46 14 England: South-East 48 512,408 51 LLI 3 23,362 140,277 Con 45 489,406 220,829 42 47 11 S. Midland 38 388,361 39 LLI 6 63,375 120,012 Con 32 324,986 164,148 26 27 7 East 29 319,997 32 LLI 9 80,447 101,785 Con 20 239,550 125,375 11 8 4 South-West 40 337,449 33 LLI 14 122,410 127,086 Con 26 215,039 142,269 12 9 1 W. Midland 58 630,931 63 LLI 10 96,089 200,113 Con 48 534,842 261,474 38 43 9 N. Midland 34 378,996 38 LLI 18 211,280 149,794 2 4 0 Con 16 167,716 153,294 North-West 70 794,142 79 LLI 14 176,183 281,634 Con 56 617,957 351,243 42 43 9 Yorkshire 52 612,892 61 LLI 26 326,841 239,045 5 1 Con 26 286,051 238,870 North 32 367,007 36 LLI 16 197,102 147,017 2 2 Con 16 169,905 135,459 ENGLAND 461 4,944,108 492 LLI 124 1,370,807 1,657,814 Con 337 3,573,301 2,040,508 213 212 52
Wales and Monmouth 34 342,209 34 LLI 28 286,628 161,190 22 24 8 Con 6 55,581 103,396 Scotland 72 683,840 68 LLI 34 312,781 254,112 Con 34 371,059 258,836 4 6
Britain 567 5,970,187 594 LLI 186 1,970,216 2,073,116 Con 381 3,999,941 2,402,740 195 194 44
Ireland 103 765,258 76 LLI 82 598,469 318,203 61 44 28 Con 21 166,757 145,906
Total 670 6,735,415 670 LLI 268 2,568,685 2,391,319 Con 402 4,166,698 2,548,736 134 150 16 Majority 134 1,598,013 157,417
GENERAL ELECTION, 1906
Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 626,011 57 LLI 40 385,762 251,937 Con 20 240,249 225,725 20 13 3 England South East 48 583,000 54 LLI 22 273,398 245,046 Con 26 309,602 241,097 4 4 S.Midlands 38 441,803 40 LLI 27 328,386 193,594 16 20 2 Con 11 113,417 172,159 East 29 368,662 34 LLI 25 333,564 170,039 21 28 4 Con 4 35,098 128,991 South-West 40 371,300 34 LLI 34 321,822 176,478 28 24 4 Con 6 49,478 144,342 W.Midland 58 679,903 63 LLI 35 402,148 288,832 12 11 1 Con 23 277,760 286,862 N.Midland 34 420,677 39 LLI 28 358,852 205,066 22 27 5 Con 6 61,825 151,924 North-West 70 869,792 80 LLI 55 680,843 420,969 40 46 12 Con 15 188,949 321,560 Yorkshire 52 667,863 62 LLI 41 556,233 340,865 30 42 14 Con 11 111,635 218,778 North 32 409,843 38 LLI 27 345,353 215,748 22 26 10 Con 5 64,490 123,003 England 461 5,438,859 501 LLI 334 3,986,356 2,508,574 207 233 53 Con 127 1,452,503 2,014,441
Wales and Monmouth 34 387,585 35 LLI 34 387,585 217,462 34 35 13 Con 0 -- 100,547 Scotland 72 750,401 70 LLI 60 629,360 367,942 48 48 16 Con 12 121,041 235,098
Britain 567 6,576,845 606 LLI 428 5,003,301 3,093,978 289 316 82 Con 139 1,573,544 2,350,086
Ireland 103 693,417 64 LLI 85 545,748 301,833 67 36 22 Con 18 147,669 144,708
TOTAL 670 7,270,262 670 LLI 513 5,549,049 3,395,811 356 352 104 Con 157 1,721,213 2,494,794 Majority 356 3,827,836 901,017
GENERAL ELECTION, JANUARY 1910
Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 658,795 57 LLI 26 246,838 254,154 Con 34 411,957 298,821 8 15 5 England: South-East 48 636,108 55 LLI 3 31,221 220,995 Con 45 604,887 334,022 42 49 11 S. Midland 38 490,592 43 LLI 11 146,312 197,717 Con 27 344,280 235,776 16 17 3 East 29 400,062 35 LLI 15 236,234 173,465 1 7 1 Con 14 163,828 170,027 South-West 40 386,514 34 LLI 18 201,726 172,692 2 Con 22 184,788 175,010 4 W. Midland 58 713,761 62 LLI 17 227,430 284,629 Con 41 486,331 334,874 24 22 6 N. Midland 34 446,752 39 LLI 23 334,766 216,469 12 19 3 Con 11 111,986 181,209 North-West 70 928,640 81 LLI 47 636,497 449,324 24 35 7 Con 23 292,143 382,796 Yorkshire 52 701,856 61 LLI 89 564,418 365,185 26 37 11 Con 13 137,438 248,507 North 32 430,594 38 LLI 23 354,697 216,760 14 24 6 Con 9 75,897 150,471 ENGLAND 461 5,793,674 505 LLI 222 2,980.139 2,551,390 21 3 Con 239 2,813,535 2,521,513 17 Wales and Monmouth 34 425,714 37 LLI 32 414,613 243,383 30 35 13 Con 2 11,101 116,696 Scotland 72 785,391 68 LLI 61 675,723 394,103 50 50 14 Con 11 109,668 265,770 Sub total 567 7,004,779 610 LLI 315 4,070,475 3,188,876 63 106 30 Con 252 3,188,876 2,903,979
Ireland 103 688,284 60 LLI 82 518,154 356,223 61 30 26 Con 21 170,130 145,437
Total 670 7,693,063 670 LLI 397 4,588,629 3,545,099 124 136 56 Con 270 3,104,434 3,049,416 Majority 124 1,484,195 495,683
GENERAL ELECTION, DECEMBER 1910
Table headings: Col A: Members Col B: Registered Electors Col C: Proportionate Number of Members Col D: Members - Liberal, Labour and Irish Col E: Members - Conservatives Col F: Electorate of Constituencies held by - Liberal, Labour, and Irish Nationalists Col G: Electorate of Constituencies held by - Conservative Col H: Voters - Liberal, Labour, and Irish Nationalist Col I: Voters - Conservative Col J: Majority - Actual Col K: Majority - With equal Single Member Constituencies Col L: Majority - Under Proportional Representation.
A B C DE FG HI J K L Prop Memb Electorate Voters Majority Memb Elect Memb Act Eq PR Metropolis 60 658,795 57 LLI 29 279,492 223,151 Con 31 379,303 264,281 2 9 5 England-- South-East 48 636,108 55 LLI 5 58,248 209,434 Con 43 577,860 311,888 38 45 11 S. Midland 38 490,592 43 LLI 14 170,762 190,120 Con 24 319,830 219,876 10 13 3 East 29 400,062 35 LLI 16 256,750 164,849 3 9 1 Con 13 143,312 154,529 South-West 40 386,514 34 LLI 14 159,494 164,698 Con 26 227,020 168,992 12 6 0 W. Midland 58 713,761 62 LLI 19 246,842 268,125 Con 39 466,919 316,574 20 20 6 N. Midland 34 446,752 39 LLI 21 298,037 202,351 8 13 3 Con 13 148,715 173,545 North-West 70 928,640 81 LLI 39 524,682 400,508 8 11 1 Con 31 403,958 386,045 Yorkshire 52 701,856 61 LLI 40 570,544 321,622 28 39 9 Con 12 131,312 239,067 North 32 430,594 38 LLI 25 375,574 200,583 18 28 6 Con 7 55,020 142,388 ENGLAND 461 5,793,674 505 LLI 222 2,940,425 2,345,441 7 Con 239 2,853,249 2,377,185 17 5 Wales and Monmouth 34 425,714 37 LLI 31 388,507 210,525 28 31 9 Con 3 37,207 121,013 Scotland 72 785,391 68 LLI 61 678,395 372,313 50 50 10 Con 11 106,996 277,183
Subtotal 567 7,004,779 610 LLI 314 4,007,327 2,928,279 61 88 14 Con 253 2,997,452 2,775,381
Ireland 103 688,284 60 LLI 84 536,675 350,029 65 34 24 Con 19 151,609 146,982
Total 670 7,693,063 670 LLI 398 4,544,002 3,278,308 126 122 38 Con 272 3,149,061 2,922,363 Majority 126 1,394,941 355,945
APPENDIX VI
PREFERENTIAL VOTING: THE TRANSFER OF SUPERFLUOUS VOTES
(A Memorandum by the Rt. Hon. J. Parker Smith)[1]
(1) _The Element of Chance Involved: Its Magnitude_
An objection, which occurs to every one who considers schemes of Preferential Voting, is that an element of chance is introduced into the result by the methods for the transfer of the superfluous votes of successful candidates. Supposing one part of the supporters of A, a successful candidate, have put down B as their second choice, and the remainder C, and that a certain number of A's votes are superfluous, and have to be transferred, how is it to be determined what number of AB votes, as they may be called, and what number of AC votes shall be transferred? If the question is settled by chance, as, by drawing the necessary number at random from A's heap, by declaring that voting papers shall be used in the order in which they were handed in at the polling booths, or by laying down any other set of arbitrary rules to determine the order in which they shall be counted, an element of uncertainty is introduced by which there seems to be serious danger that B and C will gain or lose unfairly.
Those who are accustomed to dealing with statistics will be prepared to find this danger less than might have been expected; but even they will be surprised to find of how small importance the arbitrary element is discovered, by actual calculation, to be.
The difficulty can be made clear by a numerical instance. Take the case of an election for several seats, where the necessary quota is 6000, and where a favourite candidate, whom we will call A, has received the first votes of 10,000 voters. Though all those voters have agreed in putting the same candidate first, they are divided as to who may wish to be returned next. Six thousand of them put B as their second choice, and the other 4000 C. If the 6000 votes which A requires are drawn wholly from the AB votes, the result of the transfer will be that C is credited with 4000 votes and B with none. This would be clearly unfair, for, in reality, B has received among A's voters much more support than C. To use up the 4000 AC votes and only 2000 AB votes, and to transfer 4000 votes to B and none to C would be equally unfair to C. The course which is exactly fair to both B and C is that the votes which are transferred should be divided between them in the same proportion as that in which the opinions of the whole number of A's supporters is divided. That is to say, strict justice will be done if every 1000 votes which are used or transferred are made up of 600 AB votes and 400 AC votes. Accordingly, A's quota of 6000 must be made up of 3600 AB votes and 2400 AC votes, and the 4000 papers left to be transferred will consequently consist of 2400 votes for B and 1600 votes for C.
This principle avoids all uncertainty, and is indisputably fair. It remains to consider how to carry it into effect. In most cases there would, in reality, be many more classes of votes than in the instance taken above. Even in such cases it is practicable, as will presently be shown, to divide the votes proportionately by an actual process of counting and separation. A certain amount of complication is, of course, introduced, but the extra labour involved does not seem impossible. The question whether this extra labour is necessary must be answered by examining the magnitude of the evil which it is sought to remedy.
If the votes are counted in a random order, it is clear there is a probability that the order in which they are drawn will correspond to the total numbers of each class in the ballot-box. It is reasonable to expect that when there are 10,000 ballot papers in an urn the composition of the first thousand drawn out will nearly be the same as that of any other thousand, or of the whole 10,000. The amount of this probability may be determined mathematically, and is very great.
This fact was clearly seen by Mr. Andrae, the statesman by whom the method of preferential voting was introduced into Denmark in 1855, and a mathematician of undisputed eminence. In answer to an objection of the kind now under discussion, he replied: "If this law of mine had already been in operation over the whole of Europe (including Turkey), for a period of 10,000 years, and if the elections in every part of Europe to which the law was applied were to take place, not every one, or three, or seven years, but every week in regular repetition, these elections throughout Europe, at the rate of a general European election per week, would still have to go on for more than a thousand times the period of years already stated; that is to say, for more than a thousand times ten thousand years, before the chances would be equal that the voting papers should come out of the urn in the order required to form the basis of this problem. Although, therefore, the supposed combination is, mathematically speaking, only an enormous improbability, yet, practically speaking, it is absolutely impossible."[2]
To state the matter more exactly, and as the result of an independent mathematical investigation, it appears that in the case we have stated, if 4000 voting papers were drawn out of A's heap at random, instead of the papers being carefully sorted and proportionately divided, the probability is that neither B nor C would gain or lose more than 11 votes. In other words, it is just even betting that the number of AB votes in the 4000 drawn would lie between 2411 and 2389 (inclusive), and consequently that the number of BC votes will lie between 1589 and 1611. The odds are more than 3 to 1 neither B nor C would gain or lose more than 20 votes, _i.e._ that the number of AB votes drawn will lie between 2420 and 2380; more than 10 to 1 that neither would gain or lose more than 30 votes; just 50 to 1 that neither would gain or lose more than 40 votes; and about 2000 to 1 that neither would gain or lose more than 60 votes. If the number of classes were larger or the number of votes to be drawn smaller, the effect would be much less. It will thus be seen that it is only in the case of very closely contested elections that the element of chance can affect the result. It will also be observed that the _element of chance will not be of importance as between the different parties,_ but only as _between different individual candidates of the same party_, since in almost all cases the electors who are agreed upon the candidate they most desire will also put for their second choice candidates of the same party.
In closely contested elections it must, of course, be admitted that as a result of this method, chance might decide which of two candidates of the same party should be elected. But in closely contested elections in large constituencies so many elements of chance are always and necessarily involved, that the introduction of a fresh one does not, in reality, make the result more arbitrary. Putting aside all the slight influences which at the last moment decide a score or two of featherweight votes, and assuming that every voter is profoundly convinced of the truth of his opinions, there remains the question of boundaries. A slight change in the line of the boundaries of the constituency might easily make a difference of fifty votes--a larger difference than what we are concerned with. To carry the dividing lines from North to South instead of from East to West, would, in many localities, completely alter the character of the representation.
These are, in reality, matters of chance, and more arbitrary in their nature than the order in which voting papers are drawn from an urn.
(2) _Method of Eliminating the Chance Element_
If, however, special precautions are still thought necessary, the following method of counting the votes appears to reduce, as far as practicable, the element of chance involved in the transfer of superfluous votes:--
The whole set of voting papers of the constituency being mixed, the papers, not yet unfolded, are drawn out one by one. Each is stamped, as it is drawn, with a corresponding number, 1, 2, ... in order. It is then unfolded, and sorted according to the names of the candidates marked first and second upon it. Suppose there are six candidates, A, B, C, X, Y, Z; the votes of any candidate, A, will be sorted into six heaps, viz., A votes (_i.e._ votes where A only is voted for), AB, AC, AX, AY, and AZ votes. If A is found to have received more votes than he requires, the order in which the votes will be counted to him will be as follows: Use first the A votes, then use up those heaps where the second name also is that of a candidate who has received more than the necessary minimum. If these heaps give A more than he requires, take the same proportion out of each of such heaps, taking out of each heap the last drawn votes first. If, however, these heaps are used up without giving A as many votes as he requires, take an equal proportion of the votes of each of the remaining heaps--taking out of each heap the last drawn votes first.
_Example_.--Take an election where 6000 is the necessary minimum, and suppose A has 8650 votes, composed as follows:
A 600 AB 2,700 AC 4,500 AX 50 AY 200 AZ 600 ----- 8,650
Using first the 600 A votes, we are left with 5400 to make up out of the remaining heaps.
1. Suppose B and C have received the quota. The 5400 can be taken from their heaps exclusively, for in their two heaps are 7200 votes; the proportion to be taken from each heap is therefore 5400 out of 7200, which is three quarters. Thus we make up A's number thus:--
A votes 600 Three-quarters of 2,700 AB " 2,025 Three-quarters of 4,500 AC " 3,375 ----- 6,000
And transfer the remainder (the AB and AC votes transferred being those stamped with the lowest numbers).
2. Suppose B and X have received the quota. Their two heaps amount to 2750 votes. Using these up, there remain 2650 votes to be made up out of the AC, AY, and AZ heaps. These three heaps together contain 5300 votes; and the proportion to be taken from each heap is 2650 out of 5300, or half. Thus A's number is made up as follows:--
A votes 600 AB " 2,700 AX " 50 Half of 4,500 AC " 2,250 Half of 200 AY " 100 Half of 600 AZ " 300 ----- 6,000
And the remaining votes of each of the three last classes--being those stamped with the lowest numbers--will be transferred.
It will be observed that the element of chance is not wholly excluded, since the question, which papers out of the AC heap are transferred, is left to depend upon the order of drawing. To exclude chance wholly, these would have to be sorted into heaps according to the third name upon them, and an equal proportion taken from each heap. The figures in the first half of this paper are sufficient to show that such trouble would be wholly superfluous.
[Footnote 1: This Memorandum is published by permission of the Rt. Hon. J. Parker Smith. Although written in 1884, the arguments still apply. The method described in the second part of the paper has been adopted in the Municipal Representation Bill (see Appendix VII.), but the method of application differs in detail.]
[Footnote 2: Quoted by Mr. (afterwards Earl) Lytton in his _Report on the Election of Representatives for the Rigsraad_.--House of Commons papers, 1864, vol. 61, p. 24 of No. 7.]
APPENDIX VII
THE SINGLE TRANSFERABLE VOTE
SCHEDULE TO MUNICIPAL REPRESENTATION BILL, 1910
THE FIRST SCHEDULE[1]
RULES FOB THE TRANSFER OF VOTES AND FOR ASCERTAINING THE RESULT OF THE POLL
_Arrangement of ballot papers._
1. After the ballot papers have been mixed, in accordance with the rules contained in the First Schedule to the Ballot Act, 1872, the returning officer shall draw out all ballot papers which he does not reject as invalid, and file in a separate parcel those on which the figure 1 is set opposite the name of the same candidate. The returning officer shall then count the number of papers in each parcel.
_Ascertainment of quota._
2. The returning officer shall then add together the numbers of the papers in all the parcels and divide the total by a number exceeding by one the number of vacancies to be filled, and the result increased by one, disregarding any fractional remainder, shall be the number of votes sufficient to secure the return of a candidate, herein called the "quota."
_Candidates with quota elected._
3. Any candidate whose parcel contains a number of papers equal to or greater than the quota shall be declared elected.
_Transfer of surplus votes_.] 4.--(1) If the number of candidates elected under the last rule shall not equal the number of vacancies, the returning officer shall as far as possible transfer from each elected candidate the votes (if any) in excess of the quota (herein called surplus votes) to the candidates indicated on the ballot papers as next in order of the voters' preference, excluding candidates already declared elected. The votes of the candidate having the largest number of votes shall first be dealt with, and the particular votes to be transferred shall be determined in accordance with the following regulations:--
(a) The returning officer shall arrange all the ballot papers in the parcel of the elected candidate on which votes capable of transfer are given by filing in a separate sub-parcel those on which a next preference is indicated for some one continuing candidate.
(b) The returning officer shall also make a separate sub-parcel of the ballot papers in the parcel on which the votes given are not capable of transfer.
(c) The returning officer shall count the ballot papers in each sub-parcel, and also the total of all the ballot papers containing votes capable of transfer.
(d) If the total number of votes capable of transfer is equal to or less than the surplus votes, the returning officer shall transfer all the votes capable of transfer.
(e) If the total number of votes capable of transfer is greater than the surplus votes, the returning officer shall transfer from each sub-parcel of votes capable of transfer the number of votes which bears the same proportion to the total of the sub-parcel as the number of surplus votes bears to the total of all the votes capable of transfer.
(f) The number of votes to be transferred from each sub-parcel under the preceding regulation shall be ascertained by multiplying the total of the sub-parcel by the number of surplus votes and dividing the result by the total number of votes capable of transfer. Fractional remainders shall be disregarded.
(g) The particular votes transferred from each sub-parcel shall be those last filed in the sub-parcel.
(2) The transfer of surplus votes shall be effected by making new sub-parcels of the ballot papers on which those votes are given, and adding those sub-parcels to the parcels (if any) of the candidates to whom the transfers are made, or, where any such candidate has as yet no parcel, a new parcel shall be formed for him from the papers transferred.
(3) All ballot papers in a parcel of an elected candidate not transferred under this rule shall be set aside as finally dealt with, and the votes given thereon shall thenceforth not be taken into account.
(4) If two or more parcels of elected candidates are equal in size, the returning officer shall decide which parcel he will first deal with under this rule.
(5) A transfer of votes under this rule shall not be made unless the surplus votes of the elected candidate, together with any other surplus votes not transferred, exceed the difference between the totals of the votes of the two continuing candidates lowest on the poll.
(6) This rule shall take effect subject to the provisions for filling the last vacancy herein-after contained, and if at any time it shall be possible to fill the last vacancy under those provisions, no further transfer under this rule shall be made.
_Result of transfer._
5. After the transfer of the surplus votes of an elected candidate, any candidate who shall, as a result of the transfer, obtain the quota of votes, shall be declared elected.
_Further transfer of surplus votes._
6.--(1) Unless and until the last vacancy shall have been filled under the provisions herein-after contained, if, after the transfers directed by Rule 4, there shall still remain a vacancy, and the votes of any elected candidate to whom a transfer has been made are in excess of the quota, the returning officer shall, as far as possible, take from the sub-parcel last transferred to that candidate a number of votes equal to the surplus.
(2) The particular votes to be taken shall be determined in accordance with the regulations given in Rule 4 hereof, in the same manner as if the votes included in the sub-parcel last transferred had been the only votes given to the candidate; the ballot papers so taken shall be added in separate sub-parcels to the parcels of the continuing candidates (if any) indicated thereon as next in order of the voters' preference, and the votes given thereon shall be transferred to those candidates accordingly. Where any such candidate has as yet no parcel, a new parcel shall be formed for him from the papers transferred.
(3) The remaining ballot papers in the parcel of the elected candidate (including the ballot papers taken from the parcel under Sub-Rule (1) on which the votes given are not capable of transfer) shall be set aside as finally dealt with, and the votes given thereon shall thenceforth not be taken into account.
(4) After any transfer of votes under this rule, any candidate who shall, as a result of the transfer, obtain the quota of votes shall be declared elected.
(5) The process directed by this rule shall be repeated until the last vacancy is filled, or until no candidate has any surplus votes, whichever shall first happen.
(6) If two or more parcels shall be equal in size, regard shall be had to the number of votes counted to each candidate under Rule 1, and the parcel of the candidate highest on that count shall first be dealt with, but if the numbers of votes on that count were equal, the returning officer shall decide which parcel he will first deal with under this rule.
(7) A transfer of votes under this rule shall not be made unless the surplus votes of the elected candidate, together with any other surplus votes not transferred, exceed the difference between the totals of the votes of the two continuing candidates lowest on the poll.
_Distribution of votes of lowest candidate_.
7.--(1) Unless and until the last vacancy shall have been filled under the provisions herein-after contained, if, after the transfers under the preceding rules, there shall still remain one or more vacancies, or, if no candidate shall have been declared elected under Rule 3, the returning officer shall exclude from the poll the candidate having the lowest number of votes, and shall distribute the votes capable of transfer on the ballot papers in his parcel among the continuing candidates next in order of the voters' preference. Any ballot papers in the parcel, on which votes not capable of transfer are given, shall be set aside as finally dealt with, and the votes given thereon shall thenceforth not be taken into account.
(2) If in any case the total of the votes of the two or more candidates lowest on the poll together with any surplus votes not transferred is less than the votes of the next highest candidate, the returning officer may in one operation exclude those candidates from the poll and distribute their votes in accordance with the foregoing provisions.
(3) After the distribution under this rule of votes capable of transfer, any candidate who has received the quota shall be declared elected.
(4) The surplus votes of any candidate elected under this rule who has received more than the quota shall be distributed in the manner directed by and subject to the conditions of the last preceding rule.
_Further distributions_.
8. The process directed by the last rule shall be repeated on the successive exclusions one after another of the candidates with the lowest numbers of votes until the last vacancy is filled either by the election of a candidate with the quota or under the next following rule.
_Filling the last vacancy_.
9.--(1) When the number of continuing candidates is reduced to the number of vacancies remaining unfilled, the continuing candidates shall be declared elected.
(2) When only one vacancy remains unfilled and the votes of some one continuing candidate exceed the total of all the votes of the other continuing candidates together with any surplus votes not transferred, that candidate shall be declared elected.
(3) When more than one vacancy remains unfilled and the votes of the candidate, who, if all the vacancies were filled by the successive elections of the continuing candidates with the largest numbers of votes, would be the last to be elected, exceed the total of all the votes of the continuing candidates with fewer votes than himself together with any surplus votes not transferred, that candidate and all the other continuing candidates who have not less votes than himself shall be declared elected.
(4) When only one vacancy remains unfilled and there are only two continuing candidates, and those two candidates have each the same number of votes and no surplus votes remain capable of transfer, one candidate shall be declared excluded under the next following rule and the other declared elected.
_Provisions for exclusion of candidates in special cases._
10. If at any time when a candidate has to be excluded under these rules two or more candidates have each the same number of votes, regard shall be had to the number of votes counted to each candidate under Rule 1, and the candidate lowest on that count shall be excluded, but, if the numbers of votes on that count were equal, the returning officer shall decide which candidate shall be excluded.
_Public notice of transfers._
11. The returning officer shall record and give public notice of any transfer of votes made under these rules and of the total number of votes counted to each candidate after any such transfer in addition to the particulars prescribed by Rule 45 to the First Schedule to the Ballot Act, 1872. Such public notice may be in accordance with the form given in the appendix to these rules.
_Recounts._
12.--(1) Any candidate or his agent may at any time during the counting of the votes, either before the commencement or after the completion of the transfer of the votes (whether surplus or otherwise) of any candidate, request the returning officer to recount the papers then comprised in the parcels of all or any candidates (not being papers set aside as finally dealt with) and the returning officer shall forthwith recount the same accordingly. The returning officer may also at his discretion recount votes either once or more often in any case in which he is not satisfied as to the accuracy of any previous count. Provided that nothing herein shall make it obligatory on the returning officer to recount the same votes more than once.
(2) If upon an election petition--
(i) any ballot papers counted by the returning officer are rejected as invalid,
or
(ii) any ballot papers rejected by the returning officer are declared valid,
the court may direct the whole or any part of the ballot papers to be recounted and the result of the election ascertained in accordance with these rules.
(3) Except as in this rule expressly provided, no recount shall be had whether on an election petition or otherwise.
_Determination of questions as to transfers.
13.--(1) If any question shall arise in relation to any transfer, the decision of the returning officer, whether expressed or implied by his acts, shall be final unless an objection is made by any candidate or his agent before the declaration of the poll, and in that event the decision of the returning officer may be reversed upon an election petition.
(2) If any decision of the returning officer is so reversed, the transfer in question and all operations subsequent thereto shall be void, and the court shall direct what transfer is to be made in place thereof, and shall cause the subsequent operations to be carried out and the result of the election to be ascertained in accordance with these rules.
_Definitions_.
14. In these rules--
(1) The expression "votes capable of transfer" means votes given on ballot papers on which a further preference is indicated for a continuing candidate. Provided that a vote shall be deemed not capable of transfer in any case in which--
(a) The names of two or more candidates (whether already excluded from the poll or declared elected or not) are marked with the same figure and are next in order of preference, or
(b) The name of the candidate to whom the transfer is to be made or of some candidate (whether continuing or not) higher in the order of the voters' preference is marked
(i) by a figure not following consecutively after some other figure on the ballot paper, or
(ii) by two or more figures.
(2) The expression "continuing candidates" means candidates not already declared elected or excluded from the poll.
APPENDIX TO SCHEDULE
EXAMPLE OF AN ELECTION CONDUCTED ON THE SYSTEM OF PROPORTIONAL REPRESENTATION SET OUT ABOVE
Let it be assumed that there are five members to be elected, and that there are ten candidates.
The valid papers are drawn from the general heap of ballot papers and arranged in separate parcels under the names of the candidates marked with the figure 1. (Rule 1.)
Each separate parcel is counted (Rule 1) and the total of all the valid votes is ascertained (Rule 2). It is found that the total of all the valid votes is 6000.
This total is divided by six (_i.e._ the number which exceeds by one the number of vacancies to be filled), and 1001 (_i.e._ the quotient 1000 increased by one) is the number of votes sufficient to elect a member, and is called the "quota" (Rule 2).
The result of the count may be supposed to be as follows:--
A 2,009 Elected B 952 C 939 D 746 E 493 F 341 G 157 H 152 I 118 K 93 ----- 6,000
A's votes exceed the quota and he is declared elected (Rule 3).
_First Transfer_.
It now becomes necessary to transfer A's surplus votes (Rule 4 (1)). A has in fact (2009 less 1001 or) 1008 surplus votes. All A's 2009 voting papers are examined and arranged in separate sub-parcels according to the second preferences indicated thereon (Rule 4 (1) (_a_)). A separate sub-parcel is also formed of those papers on which no second preference is shown, and which are therefore not capable of transfer. (Rule 4 (1) (_b_).) The result is found to be as follows. (Rule 4 (1) (_c_).)
A second preference is shown for G on 1,708 papers " " " D " 257 " " " " E " 11 " " " " F " 28 " ----- Total of votes capable of transfer 2,004 " No second preference is shown on 5 " ----- Total of A's votes 2,009
The total number of votes to be transferred is 1008, and it is necessary that they should be taken from the several sub-parcels in the proportions which the latter bear to all the votes capable of transfer; that is, there must be transferred, _e.g.,_ to G a number of votes bearing the same proportion to 1008, the total to be transferred, as 1708, the number of votes in G's sub-parcel, bears to 2004, the total of votes capable of transfer. In other words the number of the ballot papers on which each candidate is next preference must be multiplied by a fraction of which the surplus is the numerator and the total of votes capable of transfer the denominator, in order to ascertain the number of votes to be transferred to the candidate in question. In making the transfers fractions of votes are neglected (Rule 4 (1) ( e) and (f)).
The process is as follows:--
To G there are to be transferred 1,708 x 1,008 / 2,004 = 589 votes
" D " " " 257 x 1,008 / 2,004 = 129 "
" E " " " 11 x 1,008 / 2,004 = 5 "
" F " " " 28 x 1,008 / 2,004 = 14 " ------- 1,007
859, 129, 5 and 14 votes are now transferred to G, D, E, and F respectively, the particular voting papers taken being those last filed in their sub-parcels, and therefore at the top of the sub-parcels. These voting papers are added in separate sub-parcels to G, D, E, and E (Rule 4 (2)).
Their totals then become--
G . . . . . 157 + 859 = 1,016 D . . . . . 746 + 129 = 875 E . . . . . 493 + 5 = 498 F . . . . . 341 + 14 = 355
All the other voting papers in A's parcel (1002 in number) are set aside as finally dealt with (Rule 4 (3)), the figure 1002 being the quota 1001 with the addition of the one further vote of the surplus which, owing to the disregard of fractions, is not transferred. G having obtained more than the quota is now declared elected (Rule 5), and the poll stands as follows:--
A 1,002 Elected G 1,016 Elected B 952 C 939 D 875 E 498 F 355 H 152 I 118 K 93
_Second Transfer_
G has now more than the quota, and his surplus votes (1016 less 1001 or 15) would have to be transferred (Rule 6(1)) were it not for the provisions of Rule 6(7). But under that rule, the process of transferring a surplus is postponed in a case where the surplus is less than the difference between the two lowest candidates on the poll, and where, therefore, the transfer would produce no practical effect. In this case the difference between I and K, the two lowest candidates, is 118 - 93, or 25, and therefore it is not necessary to transfer G's surplus.
The returning officer proceeds to distribute the votes of the candidates with the smallest totals (Rules 7 and 8).
K's parcel is therefore examined and is found to contain 89 papers on which F is next preference, and 4 on which C is next preference.
Therefore 89 votes are transferred to F and 4 to C.
The poll now stands--
A 1,002 Elected G 1,016 Elected B 952 C 943 D 875 E 498 F 444 H 152 I 118
No further candidate has the quota.
_Third Transfer_
The difference between I and H exceeds G's surplus, which therefore is allowed to remain (Rule 6 (7)), and the votes of I as now lowest on the poll have now to be distributed in the same manner as K's (Rule 8). But as the combined votes of H and I, together with G's surplus (152 + 118 + 15 = 285), are less than 444, the total of F, the next highest candidate, the returning officer avails himself of Rule 7 (2), and distributes both H and I's votes at one operation.
I's parcel is found to contain 107 papers on which D and 11 on which B is next preference, and H's parcel is found to contain 108 papers on which B is next preference, and 44 on which there is no available preference marked. (In some cases, some or one of A, G, I, H, and K are marked as next in order of preference on the papers examined, but as all of them are already either elected or excluded they are left out of account.) Therefore, 107 votes are transferred to D, and 119 (108 + 11) to B, while 44 are set aside as finally dealt with (Rule 7 (1)). The result is to give B the quota, and he is declared elected.
The poll now stands--
A 1,002 Elected G 1,016 Elected B 1,071 Elected D 982 C 943 E 498 F 444
_Fourth Transfer_
B has now a surplus of 70 votes, and it is necessary to distribute this (Rules 7 (4), 6, and 4) as it exceeds the difference between E and F, which is 54 (Rule 6 (7)).
For this purpose only the 119 votes last transferred are taken into account (Rule 6 (2)).
These are examined and arranged in sub-parcels, in the same manner as A's votes were examined and arranged, with the following result: A next preference is shown for E on 84 papers. No further preference is shown on 35 papers. The total number of votes capable of transfer (84) is thus greater than the surplus (70), but, as there is only one possible transfer, the process is simple: 84 x 70/84 = 70; and so the 70 votes last filed in E's sub-parcel are transferred to E.
The poll now stands--
A 1,002 Elected G 1,016 Elected B 1,001 Elected D 982 C 943 E 568 F 444
_Fifth Transfer_
G's surplus is still not distributable (Rule 6(7)), but F is now lowest on the poll and his votes have to be distributed (Rule 8).
On examination it is found that of F's 444 papers, 353 show a next preference for C, and the remainder, 91, contain no further preference.
The 353 are transferred to C, who thus has more than the quota, and is declared elected, and the 91 are set aside as finally dealt with (Rule 7(1)).
The poll now stands--
A 1,002 Elected G 1,016 Elected B 1,001 Elected C 1,296 Elected D 982 E 568
This terminates the election; for, even if all C's surplus votes (295) and all G's surplus votes (15) were transferred to E, his poll would only amount to 878. But D's votes (982) exceed this total, D is therefore declared elected (Rule 9 (2)).
The final result is that A, G, B, C, and D are elected.
Public Notice of the Result of the Poll and of the Transfer of Votes
Number of valid votes ... 6,000 Number of members to be elected ... 5 Quota ... 1,001
[column names-- ] N: Names of Candidates V: Votes TA: Transfer of A's surplus RA: Result TK: Transfer of K's Votes RK: Result THI: Transfer of H and I's Votes RHI: Result TB: Transfer of B's surplus TB: Result TF: Transfer of F's Votes RF: Final Result
N: V: TA: RA: TK: RK: THI: RHI: TB: TB: TF: RF:
A 2,009 -1,007 1,002 -- 1,002 -- 1,002 -- 1,002 -- 1,002(E) B 952 -- 952 -- 952 +119 1,071 -70 1,001 -- 1,001(E) C 939 -- 939 + 4 943 -- 943 -- 943 +353 1,296(E) D 746 +129 875 -- 875 +107 982 -- 982 -- 982(E) E 493 + 5 498 -- 498 -- 498 +70 568 -- 568 F 341 + 14 355 +89 444 -- 444 -- 444 -444 -- G 157 +859 1,016 -- 1,016 -- 1,016 -- 1,016 -- 1,016(E) H 152 -- 152 -- 152 -152 -- -- -- -- -- I 118 -- 118 -- 118 -118 -- -- -- -- -- K 93 -- 93 -93 -- -- -- -- -- -- -- ___ ___ ___ ___ ___ ___ Effective votes 6,000 -- 6,000 -- 6,000 -- 5,956 -- 5,956 -- 5,865 Preferences exhausted -- -- -- +44 44 -- 44 +91 135 Total valid votes 6,000 -- 6,000 6,000 -- 6,000 -- 6,000 -- 6,000
[Candidates A, B, C, D, and G are elected.]
[Footnote 1: The rules contained in this schedule were examined and approved by the Select Committee of the House of Lords in 1907. They are substantially identical with those embodied in the Transvaal Municipal Act of 1909, and used in the municipal elections of Pretoria and Johannesburg in 1909, as well as in the model elections conducted by the Proportional Representation Society in 1906, 1908, and 1910.]
APPENDIX VIII
THE SINGLE TRANSFERABLE VOTE
SCHEDULE (4) OF TASMANIAN ELECTORAL ACT, 1907
In this Schedule, unless the contrary intention appears--
"Returning Officer" means the Returning Officer for the District:
"Quota" means the number of votes sufficient to elect a candidate:
"Surplus" means the number of votes which a candidate has obtained, at any stage of the scrutiny, over and above the quota:
"First choice recorded for a candidate" means a voting-paper on which the number 1 is placed in the square opposite the name:
"Second choice recorded for a candidate" means a voting paper on which the number 2 is placed in the square opposite his name:
"Transfer value" means that portion of a vote which is unused by--
(a) an elected candidate who has obtained a surplus,
(b) a candidate excluded on account of his being lowest on the poll, and which is therefore transferred to the candidate next in the order of the voter's preference. The transfer value of all votes is either 1 or some fraction of 1.
METHOD OF COUNTING VOTES
_First choice of each candidate to be counted_.]
1. The number of first choices recorded for each candidate shall be counted, and all informal voting papers shall be rejected.
_To find the quota_.
2. The aggregate number of such first choices shall be divided by one more than the number of candidates required to be elected, and the quotient increased by one, disregarding any remainder, shall be the quota, and (except as hereinafter provided in Rule 10) no candidate shall be elected until he obtains a number of votes equal to or greater than the quota.
_Candidates who have the quota to be declared elected._
3. Any candidate who has, upon the first choices being counted, a number of such votes equal to or greater than the quota shall be declared elected.
_If first choices exactly equal to quota, voting papers to be set aside_.
4. Where the number of such votes obtained by any candidate is equal to the quota, the whole of the voting papers on which a first choice is recorded for such elected candidate shall be set aside as finally dealt with.
_If a surplus, surplus to be transferred._
5. Where the number of such votes obtained by any candidate is in excess of the quota, the proportion of votes in excess of the quota shall be transferred to the other candidates not yet declared elected, next in the order of the voters' respective preferences, in the following manner:--
_Voting papers reexamined and second choices counted._
(i) All the voting papers on which a first choice is recorded for the elected candidate shall be re-examined, and the number of second choices, or (in the case provided for in Rule 12) third or next consecutive choices, recorded for each unelected candidate thereon shall be counted:
_Find the transfer value._ (ii) The surplus of the elected candidate shall be divided by the total number of votes obtained by him on the counting of the first choices, and the resulting fraction shall be the transfer value:
_Multiply second choices by transfer value._
(iii) The number of second or other choices, ascertained in paragraph i, to be recorded second for each unelected candidate, shall be multiplied by the transfer value:
_Add result on._
(iv) The resulting number, disregarding any fractional remainder, shall be credited to each unelected candidate, and added to the number of votes obtained by him on the counting of the first choices.
_If more than one surplus, largest to be first dealt with._
6.--(a) Where, on the counting of the first choices or on any transfer, more than one candidate has a surplus, the largest surplus shall be first dealt with. If then more than one candidate has a surplus, the then largest surplus shall be dealt with, and so on: Provided that, if one candidate has obtained a surplus at a count or transfer previous to that at which another candidate obtains a surplus, the surplus of the former shall be first dealt with.
_If surpluses equal, last difference to decide._
(b) Where two or more surpluses are equal, the surplus of the candidate who was the highest on the poll at the count or transfer at which they last had an unequal number of votes shall be first dealt with; and if they have had an equal number of votes at all preceding counts or transfers, the returning officer shall decide which candidate's surplus shall be first dealt with.
_If transfer raises candidate up to or above quota, he is to be declared elected._
7.--(a) Where the number of votes obtained by a candidate is raised up to or above the quota by a transfer as aforesaid, he shall thereupon be declared elected. And in such case, notwithstanding the fact that he may have reached the quota, such transfer shall be completed, and all the votes to which he is entitled there from shall be transferred to him, but no votes of any other candidate shall be transferred to him.
_If votes exactly equal quota, voting papers to be set aside._
(b) Where the number of votes obtained by a candidate is raised up to, but not above, the quota by a transfer as aforesaid, the whole of the voting papers on which such votes are recorded shall be set aside as finally dealt with.
_If surplus created, surplus to be transferred._
(c) Where the number of votes obtained by a candidate is raised above the quota by a transfer as aforesaid, his surplus shall be transferred to the candidates next in the order of the voters' respective preferences, in the following manner:--
_Voting paper of last transfer re-examined and third choices counted._
(i) The voting papers on which are recorded the votes obtained by the elected candidate in the last transfer shall be reexamined, and the number of third, or (in the case provided for in Rule 12) next consecutive choices recorded for each unelected candidate thereon counted:
_ Find the transfer value._
(ii) The surplus of the elected candidate shall be divided by the total number of voting papers mentioned in paragraph i, and the resulting fraction shall be the transfer value:
_Multiply third choices by transfer value._
(iii) The number of second (or other) choices, ascertained in paragraph i, to be recorded for each unelected candidate, shall be multiplied by the last-mentioned transfer value:
_Add result on._
(iv) The resulting number, disregarding any fractional remainder, shall be credited to each unelected candidate, and added to the number of votes previously obtained by him.
_When all surpluses dealt with candidate lowest on poll to be excluded, and his votes transferred._ 8.--(a) Where, after the first choices have been counted and all surpluses (if any) have been transferred as hereinbefore directed, no candidate, or less than the number of candidates required to be elected, has or have obtained the quota, the candidate who is lowest on the poll shall be excluded, and all the votes obtained by him shall be transferred to the candidates next in the order of the voters' respective preferences, in the same manner as is directed in Rule 5.
_First choices to be transferred first._
(b) The votes obtained by such excluded candidate as first choices shall first be transferred, the transfer value of each vote in this case being 1.
_Then other votes in order._
(c) The other votes of such excluded candidate shall then be dealt with in the order of the transfers in which, and at the transfer value at which, he obtained them.
_Each transfer deemed a separate transfer._
(d) Each of the transfers which takes place under the two previous clauses of this rule shall be deemed for all purposes to be a separate transfer.
_If transfer raises candidate up to quota, he is to be declared elected._
9.--(a) Where the number of votes obtained by a candidate is raised up to or above the by any such transfer as aforesaid, he shall thereupon be declared elected. And in such case, notwithstanding the fact that he may have reached the quota, such transfer shall be completed, and all the votes to which he is entitled therefrom shall be transferred to him, but no other votes shall be transferred to him.
_If votes exactly equal to quota, voting papers to be set aside._
(b) Where the number of votes obtained by a candidate is raised up to, but not above, the quota by any such transfer as aforesaid, the whole of the voting papers on which such votes are recorded shall be set aside as finally dealt with.
_If surplus created, surplus to be transferred._
(c) Where the number of votes obtained by a candidate is raised above the quota by any such transfer as aforesaid, his surplus shall be transferred to the candidates next in the order of the voters' respective preferences in the same manner as is directed in Rule 7, Clause (c): Provided that such surplus shall not be dealt with until all the votes of the excluded candidate have been transferred.
_Surpluses to be dealt with before further exclusion._
(d) Where any surplus exists it shall be dealt with before any other candidate is excluded.
_Process of exclusion to be repeated until there remain number of candidates required._
10. The same process of excluding the candidate lowest on the poll and transferring to other candidates his votes shall be repeated until all the candidates, except the number required to be elected, have been excluded, and the unexcluded candidates, who have not already been so declared, shall then be declared elected.
_If lowest candidates equal last, difference to decide._
11. Where at any time it becomes necessary to exclude a candidate, and two or more candidates have the same number of votes and are lowest on the poll, then whichever of such candidates was lowest on the poll at the last count or transfer at which they had an unequal number of votes shall be first excluded, and if such candidates have had an equal number of votes at all preceding counts or transfers, the returning officer shall decide which candidate shall be first excluded.
_If a candidate elected or excluded, his name not considered on voting paper._
12. In determining what candidate is next in the order of the voter's preference, any candidates who have been declared elected or who have been excluded shall not be considered, and the order of the voter's preference shall be determined as if the names of such candidates had not been on the voting paper.
_Exhausted votes._
13. Where on any transfer it is found that on any voting paper there is no candidate opposite whose name a number is placed, other than those who have been already either declared elected or excluded, such voting paper shall be set aside as exhausted.
APPENDIX IX
THE SINGLE TRANSFERABLE VOTE
REGULATIONS FOR THE ELECTION OF SENATORS UNDER THE SOUTH AFRICA ACT, 1909
I. In these Regulations:--
(1) "_Continuing Candidates_" mean candidates not elected or not excluded from the poll at any given time.
(2) "_First Preference_" means the figure 1 set opposite the name of any candidate; "second preference" similarly means the figure 2; "third preference" the figure 3, and so on.
(3) "_Unexhausted papers_" mean ballot papers on which a further preference is recorded for a continuing candidate.
(4) "_Exhausted papers_" mean ballot papers on which no further preference is recorded for a continuing candidate, provided that a paper shall also be deemed to be exhausted in any case in which--
(_a_) The names of two or more candidates, whether continuing or not, are marked with the same figure and are next in order of preference, or
(_b_) The name of the candidate next in order of preference, whether continuing or not, is marked
(i) By a figure not following consecutively after some other figure on the ballot paper, or
(ii) By two or more figures.[1] (5) "_Original Votes_" in regard to any candidate mean the votes derived from ballot papers on which a first preference is recorded for such candidate.
(6) "_Transferred Votes_" in regard to any candidate mean votes, the value or part of the value of which is credited to such candidate and which are derived from ballot papers on which a second or subsequent preference is recorded for such candidate.
(7) "_Surplus_" means the number by which the value of the votes of any candidate, original and transferred, exceeds the quota.
II. (1) The Governor in Council shall by Proclamation fix a date on or before which every candidate for election shall be nominated by two members of the Legislature in writing addressed to the Clerk of the Legislative Assembly. Such nomination shall contain the candidate's full name and address, shall be signed by two members of the Legislature, and shall be accepted in writing by the candidate.
A nomination paper may include any number of names not exceeding eight, but no member shall sign more than one nomination paper, and no candidate shall sign a nomination paper on which his name appears. The Clerk of the Legislative Assembly shall, after consultation with the Assessors hereinafter referred to, reject all nominations not made in accordance with these regulations.
(2) Immediately after the date fixed for receiving nominations the Clerk of the Legislative Assembly shall make a return to the Governor in Council showing the names and addresses of the candidates who have been duly nominated, together with the names of the members who have nominated them. He shall at the same time certify that such nominations have been duly made in accordance with these regulations, and forward to the Governor-in-Council the certificate by the Assessors mentioned in Regulation IV. (2).
In case of disagreement between the Clerk of the Legislative Assembly and the Assessors, the Speaker of the Legislative Assembly shall, at the request of the Governor-in-Council, inspect the nomination papers, and his decision on the point at issue shall be final.
(3) If the number of nominations received is less than the number of vacancies to be filled, the Governor-in-Council shall by Proclamation call for further nominations to be made on or before a date to be fixed therein. If the number of nominations received on the original date, or such further date as may be fixed, is equal to the number of vacancies to be filled, the Governor-in-Council shall by Proclamation declare the candidates so nominated to be duly elected.
(4) If the number of candidates nominated as aforesaid exceeds the number of vacancies to be filled, the Governor-in-Council shall by Proclamation summon a joint sitting of both Houses of the Legislature for the purpose of electing candidates to fill the vacancies in the manner prescribed in these regulations. Such sitting shall be continued for a period to be fixed in the Proclamation, not being less than two hours, and no member shall be allowed to vote except during the continuation of such sitting. Provided, however, that if all the members of the Legislature have voted before the expiration of the said period of two hours, the Speaker may close the sitting.
III. Each member of the Legislature present shall vote in person, and no voting by proxy shall be permitted.
IV. (1) The Clerk of the Legislative Assembly shall act as returning officer and shall, subject to these rules, do all things necessary for the conduct of the election.
(2) Two Assessors, not being Members of Parliament, shall be nominated, one by the President of the Legislative Council and one by the Speaker of the Legislative Assembly, who shall assist and advise the returning officer in his duties, both in respect, of the receiving of nominations and the conduct of the election. Immediately after the date fixed for the receipt of nominations the Assessors shall furnish the returning officer, for transmission to the Governor-in-Council, with a certificate stating whether or not they are satisfied that the nominations have been received in accordance with these regulations. Further, if either of the Assessors is for any reason dissatisfied with the conduct of the election he shall report his opinion, with the reasons therefor, in writing to the President of the Legislative Council and the Speaker of the Legislative Assembly, who, after consultation, may if they consider it necessary, order a recount to be made, and the returning officer shall act accordingly.
(3) Before entering on their duties the returning officer and the assessors shall be required to make oath or affirmation before the Speaker that they will faithfully and impartially discharge the duties of their offices according to the rules laid down herein, or such other rules as may be lawfully made.
(4) The returning officer shall furnish the Governor-in-Council with the names of the persons elected, and shall make to the President of the Legislative Council and the Speaker of the Legislative Assembly a complete return signed by himself showing the various steps of the election, and the result of the election. He shall also transmit to the Speaker of the Legislative Assembly a sealed packet containing the nominations, the actual ballot papers and the counterfoils, which shall be preserved for a period of at least twelve months. The Governor-in-Council shall notify by Proclamation the names of the persons duly elected.
V. (1) The voting shall be by ballot. The returning officer shall ascertain that the person desiring to vote is entitled to vote and shall enter his name upon the counterfoil in the ballot paper book, and shall then tear out the ballot paper corresponding to that counterfoil, and, having stamped the ballot paper with a perforating stamp provided for the purpose, shall hand it to the member. Every ballot paper shall contain the names and addresses of all the candidates duly nominated for election, printed in alphabetical order, in the form prescribed in the annexure hereto.
(2) When the member has received a ballot paper he shall take the paper to a compartment and desk provided for the purpose and signify in manner provided by the next succeeding section for whom he desires to vote. The member shall then fold the ballot paper so that the perforated mark may be visible, and having held up the ballot paper so that the returning officer can recognize the perforated mark, shall drop the ballot paper in the ballot box placed in front of the returning officer.
(3) If a member inadvertently spoils a ballot paper he may return it to the returning officer, who shall, if satisfied of such inadvertence, give him another paper and retain the spoiled paper, and this spoiled paper shall be immediately cancelled, and the fact of such cancellation shall be noted upon the counterfoil.
VI. Every member shall have one vote only. A member in giving his vote
(_a_) Must place on his ballot paper the figure 1 in the square opposite the name of the candidate, for whom he votes;
(_b_) May in addition place on his ballot paper the figure 2, or the figures 2 and 3, or 2, 3 and 4, and so on, in the squares opposite the names of other candidates in the order of his preference.
VII. A ballot paper shall be invalid
(_a_) Upon which a member signs his name or writes any word, or makes any mark by which it becomes recognizable; or
(_b_) Which does not bear the perforated mark; or
(_c_) On which the figure 1 is not marked; or
(_d_) On which the figure 1 is set opposite the name of more than one candidate; or
(_e_) On which the figure 1 and some other figure is set opposite the name of the same candidate; or
(_f_) Which is unmarked or void for uncertainty.
VIII. In carrying out these rules the returning officer shall
(_a_) Disregard all fractions;
(_b_) Ignore all preferences recorded for candidates already elected or excluded from the poll.
IX. The ballot papers shall be examined and the returning officer, after rejecting any invalid ballot papers, shall divide the remaining papers into parcels according to the first preferences recorded for each candidate. He shall then count the number of papers in each parcel.
X. For the purpose of facilitating the processes prescribed by these regulations, each valid ballot paper shall be deemed to be of the value of one hundred.[2]
XI. The returning officer shall then add together the values of the papers in all the parcels and divide the total by a number exceeding by one the number of vacancies to be filled, and the result increased by one shall be the number sufficient to secure the return of a candidate, herein called the "quota."
XII. If at any time under these regulations a number of candidates equal to the number of persons to be elected has obtained the quota, such candidates shall be treated as elected and no further steps shall be taken.
XIII. (1) Any candidate the value of whose parcel, on the first preferences being counted, is equal to or greater than the quota, shall be declared elected.
(2) If the value of the papers in any such parcel is equal to the quota, the papers shall be set aside as finally dealt with.
(3) If the value of the papers in any such parcel is greater than the quota, the surplus shall be transferred to the continuing candidates indicated on the ballot papers as next in the order of the voters' preference, in the manner prescribed in the following regulation.
XIV. (1) If and whenever as the result of any operation prescribed by these regulations a candidate has a surplus, that surplus shall be transferred in accordance with the provisions of this regulation.
(2) If more than one candidate has a surplus the largest surplus shall be dealt with first and the others in order of magnitude; provided that every surplus arising on the first count of votes shall be dealt with before those arising on the second count, and so on.
(3) Where two or more surpluses are equal the returning officer shall decide according to the terms of regulation XIX., which shall first be dealt with.
(4) _(a)_ If the surplus of any candidate to be transferred arises from original votes only, the returning officer shall examine all the papers in the parcel belonging to the candidate whose surplus is to be transferred, and divide the unexhausted papers into sub-parcels according to the next preferences recorded thereon. He shall also make a separate sub-parcel of the exhausted papers.
(_b_) He shall ascertain the value of the papers in each sub-parcel and of all the unexhausted papers.
(_c_) If the value of the unexhausted papers is equal to or less than the surplus, he shall transfer all the unexhausted papers at the value at which they were received by the candidate whose surplus is being transferred.
(_d_) If the value of the unexhausted papers is greater than the surplus, he shall transfer the sub-parcels of unexhausted papers, and the value at which each paper shall be transferred shall be ascertained by dividing the surplus by the total number of unexhausted papers.
(5) If the surplus of any candidate to be transferred arises from transferred as well as original votes, the returning officer shall re-examine all the papers in the sub-parcel last transferred to the candidate and divide the unexhausted papers into sub-parcels according to the next preferences recorded thereon. He shall thereupon deal with the sub-parcels in the same manner as is provided in the case of the sub-parcels referred to in the last preceding subsection.
(6) The papers transferred to each candidate shall be added in the form of a sub-parcel to the papers already belonging to such candidate.
(7) All papers in the parcel or sub-parcels of an elected candidate not transferred under this regulation shall be set aside as finally dealt with.
XV. (1) If after all surpluses have been transferred, as hereinbefore directed, less than the number of candidates required has been elected, the returning officer shall exclude from the poll the candidate lowest on the poll, and shall distribute his unexhausted papers among the continuing candidates according to the next preferences recorded thereon. Any exhausted papers shall be set aside as finally dealt with.
(2) The papers containing original votes of an excluded candidate shall first be transferred, the transfer value of each paper being one hundred.
(3) The papers containing transferred votes of an excluded candidate shall then be transferred in the order of the transfers in which, and at the value of which, he obtained them.
(4) Each of such transfers shall be deemed to be a separate transfer.
(5) The process directed by this regulation shall be repeated on the successive exclusions one after another of the candidates lowest on the poll, until the last vacancy is filled either by the election of a candidate with the quota, or as hereinafter provided.
XVI. If as the result of a transfer of papers under these regulations the value of the votes obtained by a candidate is equal to or greater than the quota, the transfer then proceeding shall be completed, but no further papers shall be transferred to him.
XVII. (1) If after the completion of any transfer under these regulations the value of the votes of any candidate shall be equal to or greater than the quota, he shall be declared elected.
(2) If the value of the votes of any such candidate shall be equal to the quota, the whole of the papers on which such votes are recorded shall be set aside as finally dealt with.
(3) If the value of the votes of any such candidate shall be greater than the quota, his surplus shall thereupon be distributed in the manner hereinbefore provided, before the exclusion of any other candidate.
XVIII. (1) When the number of continuing candidates is reduced to the number of vacancies remaining unfilled, the continuing candidates shall be declared elected.
(2) When only one vacancy remains unfilled and the value of the votes of some one continuing candidate exceeds the total value of all the votes of the other continuing candidates, together with any surplus not transferred, that candidate shall be declared elected.
(3) When only one vacancy remains unfilled and there are only two continuing candidates, and those two candidates have each the same value of votes and no surplus remains capable of transfer, one candidate shall be declared excluded under the next succeeding regulation, and the other declared elected.
XIX. If when there is more than one surplus to distribute, two or more surpluses are equal, or if at any time it become necessary to exclude a candidate and two or more candidates have the same value of votes and are lowest on the poll, regard shall be had to the original votes of each candidate, and the candidate for whom fewest original votes are recorded shall have his surplus first distributed or shall be first excluded as the case may be. If the values of their original votes are equal the returning officer shall decide by lot which candidate shall have his surplus distributed or be excluded.
ANNEXURE A
FORM OF FRONT OF BALLOT PAPER
___________________________________ | | _Counterfoil_ | Order of | Names of Candidates. _No._........ |Preference | | | _________________ |___________|________ | | | | JOHN BROWN | | | | Address............................ _The counterfoil_ |___________|______________ _must show_ | | _the number_ | | JAMES THOMSON _corresponding to_| | _that on the back_| | Address............................ _of the ballot_ |___________|______________ _paper. _ | | | | ALFRED JAMES | | | | Address............................ |___________|_____________ | | | | HENRY JONES | | | | Address............................ |___________|______________ | | | | ISAAC LEVY | | | | Address............................ |___________|______________ | | | | PAUL MAYNARD | | | | Address............................ |___________|_______________ | | | | JOHANNES OOSTHUIZEN | | | | Address............................ |___________|______________ | | | | HERBERT PAIN | | | | Address............................ |___________|_______________ | | | | GEORGE ROBINSON | | | | Address............................ |___________|_______________ | | | | JACOBUS SMIT | | | | Address............................ |___________|_______________ | | | | PETRUS VAN DER SPUY | | | | Address............................ |___________|______________
_Instructions to Members_
[_Printed below the List of Candidates on the Ballot Paper shown on opposite page_
A. Each member has one vote, and one vote only.
B. The member votes--
(_a_) By placing the figure "1" opposite the name of the candidate he likes best.
He is also invited to place
(_b_) The figure "2" opposite the name of his second choice.
(_c_) The figure "3" opposite the name of his third choice, and so on, numbering as many candidates as he pleases in order of his preference. The number of preferences is not necessarily restricted to the number of vacancies.
_N.B._--The vote will be spoilt if the figure "1" is placed opposite the name of more than one candidate.
[A number is printed on the back of the ballot paper corresponding with that on the counterfoil.]
ANNEXURE C
ILLUSTRATIVE ELECTION
_Example of an Election conducted on the system of the single transferable vote in accordance with the preceding regulations_
_Reg. IX._
Assuming that there are eight members to be elected, sixteen candidates, and eighty-four electors.
The valid ballot papers are arranged in separate parcels according to the first preference recorded for each candidate, and the papers in each parcel counted. Let it be assumed that the result is as follows:--
A 3 J 4 B 13 K 4 C 4 L 3 D 2 M 4 E 19 N 4 F 5 O 3 G 5 P 2 H 3 -- I 6 84
_Reg. X._
Each valid ballot paper is deemed to be of the value of one hundred, and the values of the votes obtained by the respective candidates are as shown in the first column of the result sheet.
_Reg. XI._
The value of all the papers are added together and the total, 8400, is divided by nine (_i.e._ the number which exceeds by one the number of vacancies to be filled), and 934 (_i.e._ the quotient, 933, increased by one) is the number sufficient to secure the return of a member, and is called the quota. The operation may be shown thus:-- Quota = 8400/9 + 1 = 933 + 1 = 934.
_Reg. XIII_. (1).]
The candidates B and E, the values of whose votes exceed the quota, are declared elected.
_Reg. XIII_. (3). _Transfer of surplus_.]
As the values of the papers in the parcels of B and E exceed the quota, the surplus of each candidate must be transferred. B's surplus is 366 (_i.e._ 1300 less 934), and E's surplus is 966 (_i.e._ 1900 less 934).
_Reg. XIV_. (2).]
The largest surplus, that of E, is dealt with first.
_Reg. XIV_. (4)(_a_).]
The surplus arises from original votes, and therefore the whole of E's papers are divided into sub-parcels according to the next preferences recorded thereon, a separate parcel of the exhausted papers being also made. Let it be assumed that the result is as follows:
G is marked as next available preference on 10 papers. H " " 5 " L " " 3 " -- Total of unexhausted papers 18 No. of exhausted papers 1 -- Total of papers 19
_Reg. XIV_. (4)(_b_).]
The values of the papers in the sub-parcels are as follows:--
G 1,000 H 500 L 300 ----- Total value of unexhausted papers 1,800 Value of exhausted papers 100 ----- Total value 1,900
_Reg. XIV_. (4)(_d_).]
The value of the unexhausted papers is 1800, and is greater than the surplus. This surplus is therefore transferred as follows:--All the papers unexhausted are transferred, but at a reduced value, which is ascertained by dividing the surplus by the number of unexhausted papers. The reduced value of all the unexhausted papers, when added together, with the addition of any value lost as the result of the neglect of fractions, equals the surplus. In this case the new value of each paper transferred is 966 (the surplus)/ 18 (the number of unexhausted papers) = 53, the residue of the value, 47, being required by E for the purpose of constituting his quota.
The values of the sub-parcels transferred are:--
G = 530 (_i.e._ 10 papers at the value of 53) H = 265 (_i.e._ 5 " " ) L = 159 (_i.e._ 3 " " )
These operations can be shown on a transfer sheet as follows:
TRANSFER SHEET
Value of surplus (E's) to be transferred 966 No. of papers in E's parcel 19 Value of each paper in parcel 100 No. of unexhausted papers 18 Value of unexhausted papers 1,800
New value of each paper transferred =
Surplus 966 / No. of unexhausted papers 18 = 53
Names of Candidates marked as the No. of Papers Value of Sub-parcel next available Preference. to be to be Transferred Transferred G 10 530 H 5 265 L 3 159
Totals 18 954
No. of exhausted papers 1 --- Loss of value owing to neglect of fractions -- 12
Totals 19 966
The values of the sub-parcels are added to the values of the votes already credited to the candidates G, H, L. This operation is shown on the result sheet.
As a result of this operation G's total is brought above the quota, and he is declared elected.
_Reg. XIV_. (2).]
The next largest surplus, that of B, viz. 366, is then transferred, the operations being similar to those described in the transfer of E's surplus. Assume that there are no unexhausted papers. The new value is therefore 366 / 13 or 28. The surplus is distributed according to next preferences, as follows:
A = (7 x 28) = 196 C = (6 x 28) = 168 Value lost owing to neglect of fractions 2 ---- Total ... 366
_Reg XIV. (5)._
G's surplus has now to be transferred, only the sub-parcel last transferred being re-examined. The details are as follows:--
Value of G's surplus 96 No. of papers in sub-parcel 10 Value of each paper therein 53 No. of unexhausted papers 10 Value of unexhausted papers 530
New value of each paper transferred = 96/10 = 9
The result of the distribution is shown on the result sheet, five papers of the value of nine each being transferred to A, and five of the same value to O.
_Reg. XV. (1)._
There being no further surplus, the candidate lowest on the poll has now to be excluded. D and P both have 200.
_Reg. XIX._
The returning officer casts lots, and P is chosen to be excluded.
_Reg. XV. (1)._
Being original votes the two papers are transferred at the value of 100 each, as shown in the result sheet, 100 going to L and 100 to N. D, now being lowest, is then excluded in the same way, 100 going to H and 100 to J, all transfers being made to the next preference as marked by the elector.
O now being lowest with 345, is next excluded.
_Reg. XV. (2)._
300 being the value of original votes, the three corresponding papers are transferred at the value of 100 each to K.
_Reg. XV. (3)._
45 being the value of transferred votes, the five corresponding papers are transferred at the value of 9 each to N.
M is then excluded; his papers represent original votes and are transferred to F. J is then excluded; of the 500 credited to him, 400 come from original and 100 from transferred papers, but the value of the latter being 100, all five papers are transferred at that value, 300 going to I and 200 to H.
A is then excluded, the value of his votes being as follows:--
Original 300 Transferred 196 " 45
The 300 original go to L.
The 196 transferred representing 7 papers of the value of 28 each, and the 45 representing 5 papers of the value of 9 each, all go to N.
C is then excluded, the value of his votes being as follows:--
Original 400 Transferred 168
The original go 300 to K and 100 to I, and the transferred go 84 to L and 84 to H.
H, I, K, and L now exceed the quota, and are declared elected. Seven seats are now filled.
_Reg. XIX._
I and K now both have a surplus of 66, which surpluses have to be transferred. I having had 600 from original votes, and K 400, K's surplus is first distributed.
_Reg. XIV. (5)._
The last sub-parcel of the value of 300 is dealt with, and the whole surplus 66 goes to F, he being the next preference on all three papers.
F then has the quota and is declared elected. The election is now completed, the full details being shown on the accompanying result sheet.
RESULT SHEET Number of Votes 84 Number of Members to Elect 8 8,400 Value of Votes 8,400 Quota ----- + 1 = 934 9
Column headings: 1: Names of Candidates 2: Value of Votes at 1st Count. 3: Distribution of E's Surplus. 4: Result. 5: Distribution of B's Surplus. 6: Result. 7: Distribution of G's Surplus. 8: Result. 9: Distribution of P's and D's Votes. 10: Result. 11: Distribution of O's and M's Votes. 12: Result. 13: Distribution of J's and A's Votes. 14: Result. 15: Distribution of C's Votes. 16: Result. 17: Distribution of K's Surplus. 18: Result. (E: Elected, NE: Not elected)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 A 300 300+196=496+45=451 541 541-541 -- -- -- B 1,300 1,300-366=934 934 934 934 934 934 934 E C 400 400+168=568 568 568 568 568-568 -- -- D 200 200 200 200-200 -- -- -- -- -- E 1,900-966=934 934 934 934 934 934 934 934 E F 500 500 500 500 500+400=900 900 900+66=966 E G 500+530=1,030 1,030-96=934 934 934 934 934 934 E H 300+265= 565 565 565+100=665 665+200=865 +84= 949 949 E I 600 600 600 600 600 600+300=900+100=1,000 1,000 E J 400 400 400 400+100=500 500-500 -- -- - K 400 400 400 400 400+300=700 700+300=1,000-66=934 E L 300+159= 459 459 459+100=559 -- 559+300=859 +84= 943 934 E M 400 400 400 400 400 400-400 -- -- -- N 400 400 400 400+100=500 +45=545+241=786 786 786NE O 300 300 300+45=345 345-345 -- -- -- -- P 200 200 200 200-200 -- -- -- -- -- Value of exhausted papers
Loss of value owing to neglect of fractions +12 = 12 +2= 14 +6= 20 -- 20 - 20 -- 20 -- 20 -- 20
Totals 8,400 8,400 8,400 8,400 8,400 8,400 8,400 8,400 8,400
[Footnote 1: The fact that a voter has not marked every preference correctly does not invalidate the whole of his preferences. His paper is only treated as exhausted when the wrongly marked preference is reached.
The following are examples:--
{ A 1 { A 1 { B 2 { B 2 (1) { C 3 (2) { C 3 { D 3 { D 5 { E 4 { E 6 { F -
In case (1) the preferences for A and B would be valid. If the third preference were reached the paper would be treated as exhausted, as it would be impossible to say for which candidate the voter really intended to give his third preference. In case (2) the preferences for A, B and C would be valid, but not the later ones, whether D had been elected or excluded or was still a continuing candidate. It is possible that the voter meant to give a fourth preference for some other candidate, _e.g._ F, but omitted to do so. It would not be possible to treat 5 as being meant to be 4.]
[Footnote 2: In small elections certain difficulties arise which are not present in the case of large elections.
(_a_) The quota becomes too large if calculated in the ordinary way. Assume that 27 electors are to elect 8 candidates. Then the quota is 27/(8+1) + 1 = 4. But 8 x 4 = 32.
There are not enough quotas to go round and difficulties would arise. The addition of 1 in the case of so small a number makes the quota disproportionately big. For this reason it is advisable to treat each paper as of the value of one hundred. In the case of the Transvaal the quota instead of being 84/(8+1) + 1 = 10 will be 8400/(8+1) + 1 = 934.
(_b_) The disregard of fractions in the case of small numbers may mean the waste of several votes. Take the following example:--
Seat to be filled, 8 Electors 25 Quota = 25/(8+1) + 1 = 3
First Count A 10 B 3 C 3 D E 2 F 1 G 1 H 1 I 1 J 1
A having 10 has a surplus of 7, which has to be distributed. According to the usual rule A's 10 votes are examined and the surplus is distributed in proportion to the next preferences. The preferences are as follows:--
For B....... 5 " C....... 2 " F....... 1 " G....... 1 " H....... 1
Each of these numbers must be multiplied by 7/10, _i.e._ the surplus over the number of unexhausted votes, and the following votes are transferred:--
To B.......3-1/2 " C.......1-2/5 " F.......7/10 " G.......7/10 " H.......7/10
The fractions which are ignored amount to 3 votes, which are consequently wasted. This difficulty is overcome by increasing the value of the papers to one hundred, or in other words by working out the results to two places of decimals.
(c) In a small election at the several stages there may be two or more candidates at the bottom with an equal number of votes. Resort has to be had to lot to decide which is to be eliminated. If the papers are raised to the value of one hundred this difficulty is much less likely to occur after the first count.]
APPENDIX X
LIST SYSTEM: BILL PRESENTED TO THE FRENCH CHAMBER OF DEPUTIES, 1907
The _Commission du Suffrage Universel_, a committee of the Chamber of Deputies, made a careful comparison of the various Bills which had been submitted to the Chamber for the purpose of securing the proportional representation of the electors. The Commission in their report,[1] which was issued in March 1907, recommended the adoption of the Bill, of which a free translation is given below.
The essential features of this measure, which has received the support of the leading advocates of proportional representation, are: (1) The allotment of seats to lists in accordance with the d'Hondt, or Belgian rule (Art. 8); (2) the use of the cumulative vote in determining the relative position of candidates (Art. 6). The elector is given as many votes as there are members to be elected, which he may cumulate upon any one or distribute among several candidates. The elector is not restricted in his choice of candidates to any one list.
_Text of the Bill_
(1) Members of the Chamber of Deputies shall be elected on the list system (_scrutin de liste_) in accordance with the scheme of proportional representation hereinafter stated. There shall be no second ballot.
(2) Each department shall elect one deputy for every 75,000 inhabitants. A remainder of 25,000, or more, inhabitants shall be reckoned as 75,000.
(3) A department shall form a single constituency, provided that where a department would elect more than ten deputies, it shall be divided into two or more constituencies, as determined by law hereafter.
(4) A "list" is constituted by a group of candidates who (after making the declaration prescribed by Article 2 of the Law of 17 July 1889) jointly appeal for the support of the electors.
A list shall not include a larger number of names than there are deputies to be elected in the constituency, but it may contain a smaller number. An independent candidate shall be reckoned as a distinct list.
(5) Each list shall be delivered at the prefecture at any time after the commencement of the electoral period, and at the latest ten clear days before polling day. It shall be registered and numbered at the prefecture, and a receipt for it shall be given to each candidate.
The name of a candidate shall not be registered unless he has signed the list. A list with more candidates than there are deputies to be elected shall not be accepted for registration.
A candidate whose name appears on one list shall not be entered on another unless he has notified the prefecture by writing under his hand, duly attested, that he retires from the former list, in which case his name shall be at once removed from the former list.
Twenty-four hours before the opening of the poll the prefect shall cause each registered list with the number thereto given to be posted on the doors of the polling station.
(6) An elector has as many votes as there are deputies to be elected in his constituency.
He may give all or any of his votes to the same candidate.
The reports of the local returning officer at each polling station shall state the number of votes obtained by each candidate. (7) A Central Board (_Commission de recensement_) shall collect the reports of the local returning officers, and ascertain the electoral total of each list, and allot the seats among the lists in proportion thereto.
The electoral total of a list is the sum of the votes given to the candidates whose names appear thereon.
(8) For the purpose of allotting the seats, each electoral total shall be divided by the figures 1, 2, 3, 4, and so on up to the number of vacancies, and as many of the resulting quotients as there are vacancies shall be arranged in order of size, beginning with the largest. The smallest of these quotients so arranged, corresponding to the last seat to be filled, shall be used as the common divisor, and to every list shall be allotted a number of deputies equal to the number of times which its electoral total contains the common divisor.
(9) Within each list the seats shall be assigned to the candidates who have the largest numbers of votes; in case of an equality of votes, the eldest candidate shall be elected.
(10) If two or more lists have an equal right to a seat, it shall be allotted as between the competing candidates to that one who has received the greater number of votes, and if those votes are equal the eldest candidate shall be elected.
(11) The unelected candidates of each list with the greatest number of votes shall be classed as first, second, and third substitutes (suppleants), and so on.
If any vacancy shall occur by death, resignation, or otherwise, the substitutes shall be summoned in their classified order to fill the places of the elected members of the list to which they are attached, provided that at the time of summons they are in the enjoyment of their political rights.
(12) If more than six months before the end of a Parliament, the representation of a constituency is diminished by one-fourth and there is no substitute who can be declared elected, bye-elections to fill the vacant seats shall be held in that constituency. (13) The present law shall extend to Algeria. Nothing in this law shall affect the representation of the Colonies.
NOTE.--Since the introduction of this Bill several other proposals have been considered by the _Commission du Suffrage Universel._ The draft Bill proposed in the last report (March 1911) is not based so strictly upon proportional principles as the measure given above.
The points of difference may be summarised as follows:--
(_a_) The use of the cumulative vote is retained (Art. 6), but there is a change in the method of allotting seats to various lists (Art. 8). The new method of allotment is as follows: an "electoral quotient" is found by dividing the number of voters by the number of vacancies, and as many seats are allotted to each list as the number of voters supporting a list contains this quotient. Since each voter has as many votes as there are seats to be filled, the number of voters supporting a list is determined arbitrarily by dividing the total number of votes cast for the list by the number of vacancies.
If there are any seats not allotted by this distribution they are awarded to any list which obtains an absolute majority of the votes. Should no party obtain an absolute majority, the remaining seats are allotted to the various lists in accordance with the method described in the succeeding Appendix. This method leads to the same distribution of seats as the d'Hondt rule.
(b) The Bill recognises an important new principle in permitting _apparentement des listes_. Parties may unite for the purpose of presenting lists in combination, and the lists so presented are treated for the purpose of the allotment of seats as if they emanated from one party. This is an elastic form of the Belgian "cartel," allowing parties to act together without loss of individuality. The seats won by any such cartel are allotted to the various lists composing the cartel in accordance with the second of the methods described in the previous paragraph.
[Footnote 1: _Chambre des Deputes, Neuvieme Legislature:_ 1907, No. 883. See note as to further report, March 1911, at end of Bill.]
APPENDIX XI
LIST SYSTEM: LAW ADOPTED BY THE CANTON OF BALE TOWN, 1905
The special features of the following law are as follows:--
(1) The partial use of the cumulative vote in determining the relative position of candidates (sec. 9).
(2) The allotment of seats to lists in accordance with the rule formulated by Professor Hagenbach-Bischoff (sec. 13).
The provisions for bye-elections are contained in sections 17 to 20.
(1) The elector is supplied three days before the election with copies of the various party lists; he is given as many votes as there are members to be elected; he may strike out any names and insert others in any of the lists supplied to him, or compose his own list; he may repeat the name of the same candidate three times, but no more; but in no case may the total number of names exceed the number of members to be elected.
(2) The Hagenbach-Bischoff rule, like the d'Hondt rule, aims at finding an electoral quotient which will allow all the seats to be allotted to the different parties without remainder. In the former rule this is found by trial. The following example explains its mechanism:--
Suppose, in an election for sixteen seats, five lists have obtained votes as follows:--
List. Votes. A 5,537 B 9,507 C 3,885 D 4,769 E 377 ------- Total 24,075
The first quota is ascertained as prescribed in section 11. The number of votes is divided by one more than the number of vacancies, and the result is increased by one, thus:--
24075/(16+1) + 1 = 1417
It will be observed that this quota is identical with the Droop quota of the single transferable vote system. The totals obtained by each list are divided by this quota, as many representatives being allotted to each list as the list contains the quota. Remainders are ignored.
Lists. Votes. Quota. Representatives. A 5,537 / 1,417 3 B 9,507 / 1,417 6 C 3,885 / 1,417 2 D 4,769 / 1,417 3 E 377 / 1,417 0 -- Total 14
Only fourteen out of sixteen seats have been allotted in this operation. It is obvious that the quota is too large, and a smaller quota is ascertained in the following way. The number of votes for each list is divided by one more than the number of members already assigned to such list, and the first seat still to be disposed of is allotted to that list which has the largest quotient. The following table shows the process:--
Lists. Votes. Quotient. Representatives. A 5,537 / 4 1,384 4 B 9,507 / 7 1,358 6 C 3,885 / 3 1,295 2 D 4,769 / 4 1,192 3 E 377 / 1 377 0 -- Total 15
The largest quotient is 1384, and this figure, which is taken as the new quota, allows of the allotment of fifteen seats. There still remains one seat to be disposed of, and the process just described is again repeated, as shown in the following table:--
Lists. Votes. Quotient. Representatives. A 5,537 / 5 1,107 4 B 9,507 / 7 1,358 7 C 3,885 / 3 1,295 2 D 4,769 / 4 1,192 3 E 377 / 1 377 0 -- 16
On this occasion all sixteen seats are allotted, the final quota being 1358.
The results obtained by the Hagenbach-Bischoff method are identical with those obtained by the d'Hondt rule. The operations required in the preceding example for the allotment of seats by the latter rule are as follows:--
List totals divided by A B C D E 1 5,537 9,507 3,885 4,769 377 2 2,768 4,753 1,942 2,384 -- 3 1,845 3,169 1,295 1,589 -- 4 1,384 2,376 971 1,192 -- 5 1,107 1,901 -- -- -- 6 -- 1,684 -- -- -- 7 -- 1,358 -- -- --
The sixteen highest quotients arranged in order of magnitude are:--
9,507 (List B) 2,376 (List B) 5,537 (List A) 1,942 (List C) 4,769 (List D) 1,901 (List B) 4,753 (List B) 1,845 (List A) 3,885 (List C) 1,589 (List D) 3,169 (List B) 1,584 (List B) 2,768 (List A) 1,384 (List A) 2,384 (List D) 1,358 (List B)
The lowest of these sixteen figures, viz. 1358, is the electoral quotient, and agrees with the final quota furnished by the Hagenbach-Bischoff rule. _Law for Elections to the Grand Council, on the principle of Proportional Representation, 26 January 1905_
1. Nomination papers for the various electoral districts must be handed in to the police department not later than three weeks before the day fixed for the re-election of the Grand Council.
They may contain the names of one or more persons eligible for election, provided that the total number of names in any nomination paper is not greater than the number of members which the electoral district in question is entitled to elect; any name may appear more than once, but not more than three times.
2. Nomination papers for town districts must be signed by at least ten qualified electors; those for country districts by at least three. An elector may sign one, and only one, nomination paper, on each occasion, in each electoral district.
When handing in the nomination paper the signatories thereto must designate one of their number to attend to any necessary formalities with the police department in connexion therewith.
3. The police department shall at once communicate with the candidates nominated, and call upon them to declare within two days whether they accept the candidature or not.
If the person nominated declines to stand for election his nomination shall be cancelled.
4. No candidate may appear on more than one nomination paper. If therefore any candidate be nominated in different electoral districts, or on several nomination papers in the same district, the police department shall, in informing him of the nominations, call upon him to declare, within two days, under which nomination he wishes to stand, and on receipt of his declaration shall strike his name off the other nomination papers.
If the candidate makes no declaration within the time fixed, the police department shall decide by lot under which nomination he shall stand.
5. The police department shall inform the representatives of the nominators of the cancellings due to the refusal of the nominees to accept nomination, or to the latter having been nominated more than once, and shall allow the former a period of two days in which to make further nominations. To these further nominations the declaration in writing of the person nominated, accepting the candidature, must be attached.
If this declaration is not attached, or if the proposed candidate already appears on another nomination, the supplementary nomination shall be rejected.
6. The final (definitive) nomination papers thus obtained shall be called lists, and no further alterations may be made in them. The lists shall each be printed on a separate sheet with the names of the candidates in the order in which they appear on the nomination papers. The lists shall also be provided with a number (in rotation) for each electoral district, and if the proposers have given them any titles these shall likewise be printed.
If more than one list have the same title the police department shall require the representatives of the nominators to make some distinction between them. If this is not done within two days, these lists shall be distinguished by further special numbers (in rotation).
The different lists shall be printed on paper of the same size and the same colour.
7. At least three days before the election these lists shall be delivered to each elector in an envelope, which shall at the same time serve as a voucher of the elector's right to vote. In addition to the printed lists, each voter shall receive a blank list containing no names, but as many numbered lines as there are members to be elected (free lists).
The voucher shall take the place of the present admittance card.
8. Electors must present themselves in person at the polling booth and deliver the voucher to the polling officers.
The latter shall retain the voucher, and in return give the elector an official stamp.
9. Each elector shall have as many votes as there are members of the Grand Council to be elected in his district, and shall for that purpose choose _one_ of the lists supplied to him. If he makes use of a printed list he may strike out any names and insert any others. Every vote is valid where the name of an eligible candidate is clearly given, and the only restrictions are that the same name may not appear more than three times, and that the total number of names may not exceed the number of members to be elected.
The voter may make the alterations he desires in the printed list selected by him, or fill in the free list either at the polling booth or before reaching it.
The voter shall affix the official stamp supplied to him to the list he has selected, and place the latter in the ballot box.
10. At the close of the poll the presiding officer shall open the ballot box and compare the number of voting papers therein with the number of vouchers received and the number of official stamps issued.
Only the official voting papers with stamps attached shall be valid.
11. The polling officers shall then examine the valid voting papers and ascertain by entering the votes on counting sheets how many votes each name has received.
If a voting paper contain more names than there are Councillors to be elected for the electoral district, then the votes in excess at the bottom of the list shall not be counted.
If a voting paper contain fewer names than there are Councillors to be elected in the district, then the number of votes not used shall be ascertained and shall be added (as list votes) to the list chosen by the elector, provided the latter has made use of a printed list.
The number of votes for each list shall then be ascertained by adding together the list votes and the vote given for individual candidates on the list.
If eligible persons not standing on any list receive votes, each of these names shall be treated as a separate list.
12. If no nominations have been handed in, those persons shall be elected who receive most votes.
In the event of equality of votes, the returning officer shall at once decide the matter by casting lots.
13. If one or more lists have been nominated, the vacancies on the Grand Council shall be divided among the several lists in proportion to the number of votes each list has received. The procedure shall be as follows:--
The total number of the valid votes shall be divided by the number of vacancies increased by one.
The quotient thus obtained increased by one (but disregarding fractions) shall be called the quota.
To each list there shall be allotted as many members as the number of times the quota is contained in the votes it receives. If the total number of members thus obtained is less than the number to be elected, the votes for each list shall be divided by one more than the number of members already assigned to such list, and the first seat still to be disposed of shall be allotted to that list which has the largest quotient.
The same procedure shall be repeated as long as any seats remain to be disposed of.
If two or more lists have the same claim to the last seat to be disposed of (equality of quotient), that list shall always take precedence in which the candidate who would be selected under the provisions of Clause 14 has received the largest number of votes. In case of equality of votes the returning officer (_Wahl-bureau_) shall immediately decide the question by casting lots.
14. From each list those candidates (to the number allotted to the list) shall be selected who have received the largest number of votes. Equality of votes is decided by lot, to be drawn immediately by the returning officer.
15. If to one or several lists are allotted more seats than there are names contained, all their candidates shall in the first place stand elected. The surplus seats shall be divided among the remaining lists by continuance of the procedure prescribed in Clause 13.
16. After ascertaining the result of the election, the electoral office shall draw up a report stating the number of the voting vouchers received, of the official stamps issued, and of the voting papers handed in, the number of the votes received for each name and for each list, arranged according to the lists, particulars of the allotment of seats and the names of the elected members.
Mention shall also be made of any irregularities which have occurred.
These reports shall be signed by all the electoral officers, and shall then be forwarded, together with the voting vouchers received, the unused official stamps, the voting papers and the unissued papers, to the Government Council.
The result of the election shall be affixed conspicuously outside the Chief Polling Booth.
The Polling Officers shall notify each elected candidate of his election in writing.
17. An elected candidate who did not appear on any of the nominations put in may refuse to accept his election within one week by giving written notice to the Government Council.
The Government Council shall then immediately order a bye-election.
18. Those elected candidates whose election is rendered void owing to their simultaneously having been elected as members of the Government Council shall be immediately replaced by the Government Council by the non-elected candidates on the same list who have received most votes.
If there are none, the vacant seats on the Great Council shall immediately be filled by supplementary elections, which shall also serve to fill any seats, if any rendered vacant under Clause 17.
19. Members retiring from the Great Council during their period of office shall be replaced immediately by the Government Council by the non-elected candidates on the same list who have received most votes. If there are none, supplementary elections shall take place in the first half of the next following month of May.
20. The same regulations shall serve for supplementary elections as for general elections.
21. The provisions of this law shall come into operation for the first time in the general election for the Grand Council which takes place in the year 1905.
The provisions of earlier laws and resolutions of the Grand Council referring to elections to the Grand Council are hereby repealed, in so far as they are contrary to this law.
INDEX
(The letter _f_ after a number signifies 'and page following.' The letter _n_ signifies "note.")
Accuracy of proportional systems, Acton, Lord, Acts-- Education (1867), Port of London (1908), Queensland Electoral (1905), Redistribution (1885), South Africa (1909), Tasmanian Electoral (1896), Tasmanian Electoral (1907), Transvaal Municipal (1909), Advantages of proportional representation, Advantages of single transferable vote. _See_ Single transferable vote Aldermen, election of, Allotment of seats to parties, Alternative vote, Andrae, M., Anson, Sir William R., Bart., Asquith, Rt. Hon. H. H., Australia, Austria, Avebury, Rt. Hon. Lord,
Bale, Balfour, Rt. Hon. A. J., Ballot papers, copies of, Ballots, second, Battersea Borough Council, Bavaria, Beale, Mr. W. Phipson, Belgian, or d'Hondt system, Belgium, Bernstein, Dr. Ed., Bills-- Alternative Vote (1908, 1910), Electoral Reform (1867), Electoral Reform (1884), Electoral Reform (France), Irish Council (1907), Municipal Representation Parliamentary Representation (1854), Plural Voting (1907), Redistribution (1905), Reform (1832), Representation of the People (1867), Birmingham, Birrell, Rt. Hon. Augustine, Blind, Karl, Block vote, Borough Councils, Boundaries, importance of, Bribery, Bright, John, Brown, Prof. Jethro, Brussels, Burke, Edmund, Bye-elections,
Cairns, Lord, Campbell-Bannerman, Sir Henry, Canada, Cape Colony Legislative Council, Carlskrona election, Cartel, _Case de tete_, Cecil, Lord Hugh, Cecil, Lord Robert, Chance, effect of, Churchill, Rt. Hon. Winston S., Clark, Justice, Commons, Prof. J. R., _Commission du Suffrage Universel_, Constituencies, size of, Constitutional reform, Corbett, Mr. J. Rooke, Cost of elections, Courtney of Penwith, Rt. Hon. Lord, Criticisms of single transferable vote, Cross voting, Cumulative vote,
d'Alviella, Count Goblet, Deakin, Mr. Alfred, Defects of majority systems, Denmark, d'Hondt system, Dicey, Prof. A. V., Dilke, Sir Charles, Dobbs, Mr. Archibald E., Droop, H. R.,
Edinburgh, Education Act (1867), "Effective voting," Elections, General, _See also_ Statistics Elections, procedure at, Elector, freedom of, Elector's task, _See also_ Voting Electoral Reform Bill (1884), Electoral Reform Bill (1867), Electoral systems, Royal Commission on. _See_ Royal Commission Elimination of lowest candidate, Executive under proportional representation,
Faddists, Fairness of proportional systems, Federal Home Rule, Finland, France, Franchise, Freedom of elector, French Electoral Reform Bill,
Geneva, Germany, Gerrymander, Ghent, Gladstone, W. E., Glasgow, Gove method, Gregory method, Grey, Earl, Group formation, Group representation, Gulland, Mr. J. W., Guyot, M. Yves,
Hagenbach-Bischoff, Prof., Hare, Thomas, Hare-Clark method, Hayashida, Mr. Kametaro, Holland, Home Rule, House of Commons, House of Commons committees, House of Lords, House of Lords, Select Committee on its Reform, House of Lords, Select Committee on Municipal Representation Bill,
Illinois, Imperial Parliament, Independents, the fate of, India, Ireland, Irish Council Bill (1907), Italy,
Japan, Jarrow, Jaures, M. Jean, Jenks, Prof. E., Johannesburg,
Labour Councils, Canadian, Labour Party, Lachapelle, M. Georges, Late preferences, effect of, Leeds, Limited vote, List systems, Localities, representation of, Lochee of Gowrie, Rt. Hon. Lord, London, London Borough Councils, London County Council, Lubbock, Sir John (Lord Avebury),
Macdonald, Mr. J. Ramsay, Majorities, exaggeration of. _See also_ Statistics, Majorities, small, Majorities, under-representation of, Majority systems, Manchester, Marshall, J. Garth, Mill, John Stuart, Milner, Lord, Miners' Association, Northumberland, Minorities, disfranchisement of, Minorities, representation of, Model elections, Monk, Mr. F. D., 122, 247 Morley of Blackburn, Rt. Hon. Lord, Muir, Prof. Ramsay, Municipal elections, Municipal Representation Bill (1907),
Nanson, Prof. E. J., Naville, Ernest, New Zealand, Nomination of public bodies, Northumberland Miners' Association,
Objections to proportional representation, Orange Free State, Oregon, Organisation of elections,
_Panachage_, Parliamentary Representation Bill (1854), Party exclusiveness, Party government, Party organisation, Peers, Scottish Representative, Plural Voting Bill (1907), Port of London Act (1908), Powell, Mr. Ellis T., Practicability of single transferable vote, Praed, Mackworth, Preferences, comparative efficiency of different, Present systems, defects of, Pretoria, Proportional Representation League (France), Proportional Representation Society, Provincial Councils, South Africa,
Queensland Electoral Act (1906), Quota, the,
Redistribution, Redistribution Act (1885), Redistribution Bill (1905), Referendum, Reform Bill (1832), Representation of the people (1867), Result sheet, Returning officer, duty of, Robertson, Mr. John M., Royal Commission on Electoral Systems, Russell, Lord John,
Saxony, School Board elections, Scotland, Scottish Grand Committee, _Scrutin de liste_, Seats, allotment to parties, Second ballot, Selection of successful candidate in a list, Senates-- Australia, Canada, South Africa, Sheffield, Single transferable vote-- Advantages, _See also_ Advantages of proportional representation Application, Criticisms, Mechanism, Single vote, Smith, Rt. Hon. J. Parker, Social Democratic Party (Germany), Solothurn, South Africa, South Africa Act (1909), Spence, Catherine Helen, Spoilt ballot papers, Statistics of elections-- America, Australia, Belgium, Finland, Germany, Japan, South Africa, Sweden, United Kingdom, _Suppleants,_ Surplus votes, transfer of, Sweden, Switzerland, Systems, majority. _See_ Present systems Systems, proportional. _See_ Bale, Belgium, Finland, France, Japan, Sweden, and Single transferable vote
Tasmania, Tasmanian Electoral Act (1896), Tasmanian Electoral Act (1907), Three-cornered contests, Ticino, Toronto, Trades Unions, Transfer of surplus votes, Transfer sheet, Transvaal, Transvaal Municipal Act (1909), Two-party system,
Ulster, United States,
Vandervelde, M., Vivian, Mr. Henry, Voting, modes of,
Wales, Wallas, Mr. Graham, Warwickshire, Whips in House of Commons, _See also_ Party organisation White, Mr. Dundas, Williams, Mr. Aneurin, Wuertemberg,
End of Project Gutenberg's Proportional Representation, by John H. Humphreys