Proportional Representation: A Study in Methods of Election

by John H. Humphreys

CHAPTER VIII

LIST SYSTEMS OF PROPORTIONAL REPRESENTATION

"'One man, one vote; one party, one candidate'—thus runs the cry."—COUNT GOBLET D'ALVIELLA

List systems of proportional representation are based upon the block vote or scrutin de liste—the method of election generally used on the Continent of Europe and in the United States of America when several members are to be elected for the same constituency. With the scrutin de liste, lists of candidates are nominated by the various political organizations or groups of electors; each elector has as many votes as there are members to be elected, but he may not give more than one vote to any one candidate. The party which can obtain the support of a majority of the electors can carry its list to the exclusion of all others; minorities are crushed even more completely than with the system of single-member constituencies. But as constituencies returning several members are an essential requirement of any scheme of proportional representation, the scrutin de liste facilitates the introduction of a proportional system, for the only great change involved is the allotment of seats to the respective lists in proportion to the totals of votes obtained by each. But this change brings in its train a change in the nature of the vote. It remains no longer a vote only for candidates as individuals; it obtains a twofold significance, and becomes what is termed the double simultaneous vote (le double vote simultanée). In the first place it is a vote for the party list as such, and is used for determining the proportion of seats to be allotted to the lists; and, in the second place, it is a vote for a particular candidate or order of candidates for the purpose of ascertaining which of the candidates included in a list shall be declared successful. This double function of the vote is characteristic of all list systems of proportional representation. Other changes of a subsidiary character, which experience has shown to be advisable, have been adopted in different countries so that the various systems differ in detail in the methods both by which seats are apportioned among the competing lists and by which the successful candidates are chosen.

The Belgian electoral system.]

List systems are in operation for parliamentary purposes in Switzerland, Belgium, Würtemberg, Sweden, and Finland. The simplest of these is that adopted by Belgium, and the description of a Belgian election may serve as an introduction to the study of other systems. Through the courtesy of M. Steyeart, the President of the Tribunal of First Instance and Chief Electoral Officer for the constituency of Ghent-Eecloo, the author was enabled to watch the elections in May 1908 in that constituency. Proportional representation is, however, only one of the points in which the Belgian and English electoral systems differ, and in order to obtain a true estimate of the working of the Belgian law it is necessary to distinguish between results which are due to the franchise qualifications and those which are due to the system of proportional representation. The effects arising from these two separate features of the electoral system have sometimes been confused, and it is therefore desirable to give a brief outline of the conditions which govern a Belgian election.

In the first place, Belgium has manhood suffrage modified by a system of graduated voting. Secondly, each elector is compelled to vote or, at least, to present himself at the polling place. Thirdly, both the Chambers are elective, and, although provision exists for the dissolution and the election of Parliament as a whole, only one-half of each Chamber is, in the ordinary course, elected at a time, each Senator being elected for a fixed period of eight years, and each member of the House of Representatives for a period of four years.

The franchise.

The unique franchise system embodied in the Belgian constitution in 1893 was adopted only after months had been spent in discussing the schemes of rival parties. All attempts at compromise failed until attention was seriously directed to the suggestions of M. Albert Nyssens, Professor of the University of Louvain, contained in his pamphlet Le Suffrage Universel Tempéré. His proposals had the merit of recognizing the validity of the arguments advanced by all the political parties. Conservatives desired the introduction of a system based on occupation coupled with the payment of taxes; many Liberals were anxious to secure special recognition for electors of admitted capacity—in short, an educational qualification; the Radicals inside and Socialists outside Parliament demonstrated continually in favour of universal, direct and equal suffrage. The claim for universal suffrage was recognized by granting to every male Belgian who had attained the age of twenty-five years the right to vote, but a counterpoise to so democratic a suffrage was sought in the granting of additional votes to electors possessing specified qualifications. A supplementary vote was awarded to every married man who had attained the age of thirty-five years and paid five francs in taxes on his dwelling. An additional vote was given to every owner of land or house property of the value of two thousand francs, or to the possessor of an income of a hundred francs derived from Belgian public funds. Thus were met the demands of the Catholics for the representation of property, whilst the Liberal advocacy of the claims of the educated voter were met in a similar way. Two additional votes were awarded to those who had obtained a diploma of higher education; to those who filled, or had filled, a public position; or to those engaged in a profession which implied the possession of a good education. The highest number of votes awarded to any elector, for parliamentary purposes, whatever qualifications he might possess, was three.

Compulsory voting.

The exercise of the franchise is regarded in Belgium as a duty which each citizen owes to the State, and the obligatory vote is therefore universally accepted without demur. The elector must attend at the polling place, take his ballot paper and deposit it in the ballot box. If he places the ballot paper in the urn without voting there are no means of ascertaining the fact; but unless he forwards to the Electoral Officer an explanation, in due form, of his absence from the polling booth he is liable to prosecution. The percentage of abstentions is thus very low, but, in addition to this result, the obligatory vote has had a considerable indirect effect upon the character of electoral contests. Voting has become an official matter. Formerly, as here, it rested with the political organizations to persuade and exhort electors to vote; now, each elector receives from the Returning Officer an official command to attend at the polling place.

Partial renewal of chamber.

The third difference—the partial renewal of the Chambers—dates from the constitution of 1831, and the reason for its adoption was the same as that which underlies the partial renewal of English municipal councils—the desire to ensure continuity in the composition and proceedings of Parliament. There was some justification for this practice under the old voting methods, for then the result of elections largely depended, as is the case in England to-day, upon the chance distribution of party strength. The composition of the Chamber of Representatives was liable to violent oscillations and changes, and the partial renewal of the Chambers moderated the violence of these changes. But whilst the partial renewal may be defended on these grounds, it has two distinct disadvantages. When only one-half of the Chamber is to be elected (as in the renewal of only one-third of our municipal Councils) a considerable diminution takes place in the amount of public interest evoked by an election. There is, moreover, a further and even more serious drawback that, when the election turns upon a question of vital importance, such for instance as the annexation of the Congo, the verdict of only one-half the people is obtained. In 1908 elections took place in four provinces only—East Flanders, Hainaut, Liege, and Limbourg—and so, whilst the citizens of Ghent and Liège were expressing their opinion upon the policy of the Government, the citizens of Brussels were reduced to the position of spectators of a fight in which doubtless many would have liked to have taken a part. The introduction of proportional representation has rendered this particular feature of the Belgian electoral system quite unnecessary. Electors are not so fickle as an irrational method of voting made them appear to be.

The presentation of lists.

For the purpose of parliamentary elections each of the nine provinces of Belgium is divided into large constituencies returning several members; Brussels returns twenty-one members, Ghent eleven, but several of the smaller constituencies return as few as three representatives. Fifteen days before the date of the election lists of candidates which, before presentation, must have received the support of at least one hundred electors, are sent to the returning officer. After verification, each list is given an official number and the lists are then published, no official title other than the number being given to the lists. In the copy of the ballot paper used at Ghent, shown on the opposite page, list No. 1 was presented by the Catholics; No. 2 by the Liberals; No. 3 by those Socialists who were dissatisfied with their party's list; No. 4 by the small tradesmen; No. 5 by the official Socialists; whilst No. 6 contains the name of a candidate standing as an independent. It will be observed that each of the first five lists is divided into two parts separated by the word "Suppléants." The candidates so described are not taken into account in the actual election of representatives; they are, however, voted for in the same way and at the same time as the other candidates, and are called upon (in the order determined by the result of the election) to fill any vacancy occasioned by the retirement or death of a duly-elected representative belonging to the same list. This arrangement obviates the necessity for bye-elections, and the relative strength of parties remains the same from the time of one election to the next. The order in which the names of the candidates appear upon the lists is arranged by the organizations responsible for their presentation. It should, however, be stated that this provision, about which public opinion is much divided, is not an essential feature of a proportional system. It was not a part of the original proposals of M. Beernaert, and it certainly strengthens the hands of political organizations, although, as will be shown subsequently, proportional representation considerably modifies, if it does not altogether prevent, abuse of the power conceded to political bodies.

[Illustration: List Ballot paper]

The act of voting.

The work of the elector is simplicity itself. He can select one list or one candidate in a list but not more for each of the votes to which he may be entitled. His choice can be recorded in four different ways. In each case the act of voting consists in pencilling one or other of the white spots contained in the black squares at the head of the lists or against the names of individual candidates. In the first place, the elector may vote by blackening the spot at the head of the list. The significance of such a vote is that the elector votes for the list, and, at the same time, approves of the order in which the candidates have been arranged by the party organization. Naturally all the party organizations and journals advise their supporters to vote in this way.

Secondly, the elector may vote by blackening the white spot against the name of one of the "effective" candidates on one of the lists. Such a vote implies that the elector votes for the list on which the candidate's name appears, but that, instead of approving of the order in which the candidates have been arranged, he prefers the particular candidate he has marked. The third and fourth methods are but variations of the second. The elector can indicate a preference for one of the supplementary candidates, or he can indicate preferences for an effective and also for a supplementary candidate. In brief, the elector votes for one of the lists, and either approves of the list as arranged or indicates the change he desires.

The allotment of seats to parties.

The number of representatives awarded to each party is determined by the method formulated by M. Victor d'Hondt, a professor of the University of Ghent. Its working may best be shown by an illustration. Let it be assumed that three lists have been presented; that they have obtained 8000, 7500, and 4500 votes respectively, and that there are five vacancies to be filled. The total number of votes for each list is divided successively by the numbers 1, 2, 3, and so on, and the resulting numbers are arranged thus:—

List No. 1. List No. 2. List No. 3. 8,000 7,500 4,500 4,000 3,750 2,250 2,666 2,500 1,500

The five highest numbers (five being the number of vacancies to be filled) are then arranged in order of magnitude as follows:—

8,000 7,500 4,500 4,000 3,750

The lowest of these numbers, 3750, is called the "common divisor"[1] or the "electoral quotient," and forms the basis for the allotment of seats. The number of votes obtained by each of the lists is divided by the "common divisor" thus:—

8,000 divided by 3,750 = 2 with a remainder of 500. 7,500 " 3,750 = 2 4,500 " 3,750 = 1 with a remainder of 750.

The first list contains the "electoral quotient" twice, the second twice, and the third once, and the five seats are allotted accordingly. Each party obtains one representative for every quota of voters which it can rally to its support, all fractions of "quotas" being disregarded.

The method of determining the electoral quotient may appear at first sight rather empirical, but the rule is merely the arithmetical expression, in a form convenient for returning officers, of the following train of reasoning. The three lists with 8000, 7500, and 4500 supporters are competing for seats. The first seat has to be allotted; to which list is it to go? Plainly to the list with 8000 supporters. Then the second seat has to be disposed of; to which list is it to go? If it is given to the first list, then the supporters of the first list will have two members in all, or one member for each 4000 votes. This would be unfair while 7500 supporters of the second list are unrepresented, therefore the second seat is allotted to the list with 7500 supporters. Similar reasoning will give the third seat to the list with 4500 supporters, the fourth to the list with 8000 supporters, which now will rightly have one representative for each 4000, and the fifth to the list with 7500. The question in each case is to what list must the seat be allotted in such a way that no one group of unrepresented electors is larger than a represented group. The separate allotment of seats one by one in accordance with the foregoing reasoning may be shown thus:—

8,000 (List No. 1) 7,500 ( " No. 2) 4,500 ( " No. 3) 4,000 ( " No. 1) 3,750 ( " No. 2)

This result of course agrees with that obtained by the official process of dividing the total of each list by the electoral quotient.

The selection of successful candidates.

The seats having been apportioned to the respective lists it becomes necessary to ascertain which of the candidates on the respective lists are to be declared elected. In this second process it will be seen now great an advantage is obtained by the candidates at the top of each list.[2] A11 the votes marked in the space at the top of a list, i.e., list votes, form a pool from which the candidates of the list draw in succession as many votes as are necessary to make their individual total equal to the electoral quotient, the process continuing until the pool is exhausted. In the example already given, assume that List No. 1 consists of three candidates, A, B, and C, arranged in the order named, and that the 8000 supporters of the list have given their votes as follows:—

Votes at the head of the List 4,000
Preferential votes for A 600
" " B 500
" " C 3,000
——-
Total 8,000

Candidate A, being the first in order on the list, has the first claim on the votes recorded for the list. The electoral quotient is 3750, and A's total 500 is raised to this number by the addition of 3250 votes taken from those recorded for the list. This secures his election, and there remain 750 list votes which are attributed to candidate B, this candidate being the second in order on the list. B, however, also had 500 votes recorded against his name, and his total poll therefore amounts to 1250. But candidate C has obtained 3000 votes, all recorded for himself personally, and as this total exceeds B's total of 1250, C would be declared elected. The two candidates chosen from List No. 1 would, in this case, be A and C. The successful supplementary candidates are ascertained in the same way.

A Belgian election. Ghent, 1908: the poll.

In a Belgian election the polling proceeds very smoothly and quietly. This is largely due to the fact that the law for compulsory voting has relieved the party organizations of the necessity of whipping up their supporters to the poll. At the election of Ghent, which the author was privileged to witness, the candidates for the Chamber of Representatives were as given in the ballot paper on page 177. It will be seen that six lists of candidates were presented, but in the election of Senators only the three chief organizations took part. There were eleven members of the House of Representatives and five Senators to be elected.

The constituency was divided into 350 polling districts, the maximum number of electors for a district being 500. To each district was assigned a polling place in charge of a presiding officer, appointed by the returning officer of the district; the presiding officer was assisted by four citizens, each of whom was required to be in possession of the maximum number of votes, and to be at least forty years of age. In addition, the party organizations sent duly accredited witnesses to watch against possible fraud, and to assure themselves of the absolute regularity of the proceedings. The poll opened at 8 A.M. Each elector had to present his official "summons" to vote, and received from the presiding officer one, two, or three ballot papers according to the number of votes to which he was entitled. The elector took the papers to a private compartment, as in an English election, marked them, placed them in the ballot box and received back his official letter, now stamped—evidence, if need be, that he had carried out the obligation imposed upon him by law. At 1 P.M. the poll was closed; the ballot boxes were opened and the ballot papers counted in the presence of the assessors and party witnesses for the purpose of ascertaining that all papers in the possession of the presiding officer at the opening of the poll had been duly accounted for.

The counting of the votes.]

In order to maintain as far as possible, not only the secrecy of the individual vote, but the secrecy of the vote of any locality, the votes of three polling places were counted together, the grouping of polling places for this purpose having been previously determined by lot. Thus the votes counted at the town hall (polling district No. 1) were those recorded in the districts Nos. 1, 112, and 94. The proceedings were directed by the presiding officer of the first polling place, assisted by the presiding officers of the other two. The only other persons present were witnesses representing the three chief parties. The counting commenced soon after 3 P.M., and was completed, both for the Senate and Chamber, by 7 P.M. The papers were sorted according to the votes given for each list, subsidiary heaps being made for those candidates who had received individual votes of preference. A separate heap was made of spoiled and blank voting papers, but it was evident from the very commencement of the proceedings that the method of voting had presented no difficulty to the elector. Of the 1370 votes recorded in this division for candidates for the Chamber there were but twenty-six spoiled papers; of these thirteen were blank, indicating that the voters, although attending the poll, did not wish to record any opinion. The thirteen other papers showed in nearly every case some confusion in the mind of the elector with the elections for the communal councils, when the elector can give several votes of preference. The official returns, after endorsement, were forwarded by post to the returning officer, whose duty it was to prepare the returns for the whole constituency. The figures for each district were given to the press at the close of the count, and special editions of the journals, containing the probable result of the election, were issued the same evening.

The final process.

The compilation of the returns for the whole constituency took place on the following day. The returning officer presided, and was assisted by four assessors, a secretary and three witnesses, who attended on behalf of the chief parties. In addition there were two professional calculators, who were responsible for the accuracy of the arithmetical processes. The proceedings, in brief, consisted in extracting the details of the returns furnished by the 120 counting places. The final sheet for each list showed not only the total number of votes obtained by the party, but the number of votes of preference recorded for each candidate. The votes for each list were as follows:—

List No.1. List No.2. List No.3. List No.4. List No.5. List No.6. 78,868 39,788 913 1,094 23,118 271

The process of allotting the seats to the respective parties then commenced. The totals for each list were divided by the numbers 1, 2, 3, and so on, and arranged thus:—

List List List List List List
No. 1. No. 2. No. 3. No. 4. No. 5. No. 6
78,865 39,788 913 1,094 23,118 271
39,432 19,894 11,559
26,288 13,262
19,716 9,947
15,773
13,144
11,266

The eleven highest figures thus obtained were then arranged in order of magnitude, and the seats allotted accordingly:—

1st Seat 78,865 (List No. 1—Catholic)
2nd " 39,783 ( " No. 2—Liberal)
3rd " 39,432 ( " No. 1—Catholic)
4th " 26,288 ( " No. 1—Catholic)
5th " 23,118 ( " No. 5—Socialist)
6th " 19,894 ( " No. 2—Liberal)
7th " 19,716 ( " No. 1—Catholic)
8th " 15,773 ( " No. 1—Catholic)
9th " 13,262 ( " No. 2—Liberal)
10th " 13,144 ( " No. 1—Catholic)
11th " 11,559 ( " No. 5—Socialist)

Thus the Catholics, Liberals, and Socialists obtained six, three, and ten seats respectively. It will be noticed that the eleventh figure, 11,559, which is the "common divisor," or "electoral quotient," is contained six times in the Catholic total, with a remainder of 9511; three times in the Liberal total, with a remainder of 5000; and twice in the Socialist total.

The highest number of preferences recorded for any individual candidate (although placards had been posted inviting votes of preference for M. Buysse, the candidate fourth on the Liberal list, and for M. Cambier, the candidate third on the Socialist list) were 1914 and 1635, much too small to effect any change in the order of the candidates as arranged by the associations. It remains to add that the task was accomplished with perfect regularity and despatch; the figures were checked at each stage, but as the number of votes polled in the double election (for the Senate and for the Chamber) amounted to no less than 270,892, it is not surprising that the compilation of the final figures was not completed until midnight.

Public opinion favorable to the system.

This was the fifth parliamentary election[3] in which the system of proportional representation has been put to the test; its practicability, both from the point of view of the elector and of the returning officer, is now no longer open to question. Interviews on the effects of the system with Catholic leaders like M. Beernaert or M. Van den Heuvel, with Liberals like Count Goblet d'Alviella, or M. Gustave Abel, the editor of La Flandre Liberale, or with Socialists like M. Anseele, revealed the fact that there is no party in Belgium which desires to return to the former electoral system. The Liberals and Socialists are hostile to plural voting, but their attitude to proportional representation may be summed up in the desire to make the system more perfect.[4] Constituencies returning three or four members are not sufficiently large to do complete justice to a system of proportional representation, and many, among whom must be included M. Vandervelde, desire the grouping of these smaller constituencies into larger ones. The general trend of public opinion is in complete agreement with the views of party leaders, and found forcible expression in the press comments on the elections in 1908 for the provincial councils.

The relation of the Belgian to other list systems.

The Belgian list method, although simple in form, is based upon a very careful examination of earlier list systems, and represents an attempt to avoid the defects and inconveniences of those systems. As already stated, the vote in a "list" system has two aspects. Indeed, in the canton of Solothurn in Switzerland each elector is invited, first, to record his vote for a list as a separate act, and secondly, to vote for the particular candidate he prefers.

In tracing the growth of the Belgian system it will be best to consider these two aspects separately, and, in the first place, the vote in so far as it affects the fortunes of the list. The object in view—the allotment of the seats in proportion to the total number of votes obtained by the respective lists—would seem quite simple of attainment, and would be so were the totals obtained by each list such that it was possible to divide the seats among them in exact proportion. Voters do not, however, group themselves in exact proportion, and it becomes necessary therefore to devise a rule of distribution that shall approximate to the desired end as closely as possible.

The different methods of apportioning seats to lists.

The first rule—a very simple one—was adopted because, in the words of Ernest Naville, "it seemed most intelligible to the general public." The grand total of votes polled by the different lists was divided by the total number of seats, and the distribution of seats was based upon the quotient, or "quota" thus obtained. The total of each list was divided by the quota for the purpose of ascertaining the number of seats to which it was entitled. The answers, as will be seen from the following example, usually contained fractions. Assume that seven seats are to be distributed among three lists, A, B, C; that the grand total of votes is 7000, and that the respective lists have polled as follows:—

List A 2,850 votes
" B 2,650 "
" C 1,500 "
——-
Total 7,000

The quotient in this case is 1000. The totals of the lists A, B, and C contain the quotient twice, twice and once respectively, but in each case with a remainder, and it is the remainder that constitutes the difficulty. According to the earliest list schemes the remaining seats were allotted to the lists having the largest remainders, and, in the example given, lists A and B would each receive an additional seat. Party organizers were not slow to perceive that it was advisable to obtain as many of the largest remainders as they could, and considerable dissatisfaction arose in Ticino from the action of the Conservatives, who very skilfully divided their forces into two groups, thereby obtaining additional seats. A simple example will explain. Assume that three deputies are to be elected, that the grand total of votes is 3000, and that the party votes are as follows:—

Party A 1,600 votes
" B 1,400 "
——-
Total 3,000

The quota would be 1000 votes. Party A, having the larger remainder, would obtain two seats, and party B only one seat; but if party B should present two lists and arrange for the division of its voting force, the following result might ensue:—

Party A 1,600 votes
" B1 700 "
" B2 700 "
——-
Total 3,000

The quota would still be 1000 votes, but party A would only obtain one seat, whereas party B would obtain two, because each of its two lists would show a remainder larger than A's remainder. This possibility led to a modification of the rule, and the seats remaining after the first distribution were allotted to the largest parties. But this was also far from satisfactory, as will be seen from the following example taken from a Ticino election:[5]—

Conservatives 614 votes
Radicals 399 "
——-
Total 1,013

The constituency to which the figures refer returned five members; the quotient therefore was 202, and the Conservatives obtained three seats on the first distribution, and the Radicals one. As, under the rule, the remaining seat was allotted to the largest party, the Conservatives obtained four seats out of the five when, obviously, the true proportion was three to two.

The rule subsequently devised aimed at reducing the importance of remainders in the allotment of seats. The total of each list was divided by the number of seats plus one. This method yielded a smaller quota than the original rule and enabled more seats to be allotted at the first distribution. The final improvement, however, took the form of devising a rule which should so allot the seats to different parties that after the first distribution there should be no seats remaining unallotted. This is the great merit of the Belgian or d'Hondt rule, which has already been fully described.

Criticism of d'Hondt Rule.

The d'Hondt rule certainly accomplishes its purpose; it furnishes a measuring rod by which to measure off the number of seats won by each list.[6] But the rule is not without its critics.[7] As in the earlier Swiss methods objection was taken to the undue favouring of certain remainders, so in Belgium objection is taken to the fact that remainders are not taken into account at all. The Belgian rule works to the advantage of the largest party, a fact that many may consider as a point in its favour.

A further simple example will explain how the larger parties gain. Assume that eleven seats are being contested by three parties, whose votes are as follows:—

Party A 6,000 votes
" B 4,800 "
" C 1,900 "
———
Total 12,700

Arrange these numbers in a line, and divide successively by 1, 2, 3, and so on, thus:—

Party A. Party B. Party C. 6,000 4,800 1,900 3,000 2,400 960 2,000 1,600 1,500 1,200 1,200 960 1,000

The eleventh highest number, which constitutes the measuring rod, will be found to be 1000; the largest party obtains six seats, the second party obtains four seats, with a remainder of 800 votes, and the third only one seat, with a remainder of 900 votes. The two smaller parties taken together poll 6700 votes but only obtain five seats, as compared with the six seats obtained by the larger party with 6000 votes; the two remainders of 800 and 900 votes, which together constitute more than a quota, having no influence on the result of the election. Even if, in the allotment of seats, the largest party has a remainder of votes not utilized, yet this remainder necessarily bears a smaller proportion to the total of the votes polled than is the case with a smaller party. Thus the system works steadily in favour of the larger party.

The question of remainders, or votes not utilized in the distribution of seats, is of minor importance when the constituencies return a large number of members. When, for example, as in the city of Brussels, there are twenty-one members to be elected, the votes not utilized bear a small proportion to those that have been taken into account in the allotment of seats. In Belgium, however, there are several constituencies returning as few as three members, and there is naturally a demand that these constituencies should be united so that the method of distribution should yield more accurate results.

If the d'Hondt rule, like every other method of distribution, is open to criticism from the point of view of theoretical perfection, it must be admitted that in practice it yields excellent results. The election at Ghent resulted in the return of six Catholics, three Liberals and two Socialists; it would have been impossible to have allotted the seats more fairly. Under the old non-proportional method the Catholics would have obtained eleven representatives and the Liberals and Socialists none. The immeasurable improvement effected by every true proportional method is apt to be overlooked in the critical examination of the working of these methods in those extreme cases which rarely occur in practice.

The formation of "cartels."

The steady working of the d'Hondt rule in favour of the larger parties has, however, not escaped the attention of advocates of proportional representation. The late Professor Hagenbach-Bischoff has formulated the proposal that parties should be allowed to put forward combined lists, and that in the first allotment of seats the totals of the combined lists should be taken as the basis of distribution. The need of some such provision may be shown by an example used in illustration of the d'Hondt system, at a meeting held under the auspices of the French Proportional Representation League.[8] A constituency with eleven members was taken; four lists, A, B, C, and D, received 6498, 2502, 1499, and 501 votes respectively; the d'Hondt rule made 834 the measuring rod, and gave A seven members, B three, C one, and D none. The question was asked why provision was not made for the transfer of the votes from list D to list C, so that if, for example, these lists were put forward by Radical-Socialists and by Socialists respectively, the parties might obtain the additional seat to which their combined totals entitled them. It will be seen that lists C and D, with a total of 2000 votes (more than twice 834), obtained but one representative, while list A, with 6498 votes, obtained seven representatives.[9]

Professor Hagenbach-Bischoffs proposal, which would meet this difficulty, has not been embodied in the Belgian law, but "cartels" (arrangements for the presentation of a common list) are formed between the Liberals and Socialists so as to lessen their loss of representation due to the working of the d'Hondt rule. The "cartels," however, do not give satisfaction, as experience shows that many Liberals who would vote for a Liberal list decline to vote for a "cartel" of Liberals and Socialists; whilst, on the other hand, extreme Socialists decline to support a Liberal-Socialist coalition. In the Finnish system, however, provision is made for the combination of lists in accordance with Professor Hagenbach-Bischoff's suggestion. Indeed, as the Finnish law forbids any list to contain more than three names, some such provision was necessary in order to allow each separate party to nominate a full list of candidates.

The experience of the Belgian "cartels" would seem to show that, even where party organization and discipline are highly developed, many electors resent the disposal of their votes by a bargain between the organizations concerned. The single transferable vote, by allowing each elector to indicate his second choice in the way in which he himself prefers, would enable smaller parties to obtain their share of representation without involving a preliminary compact between party organizations. A list system seems to establish a rigid division between parties, whilst there is no such corresponding rigid division in the minds of many electors. The model elections conducted by the Proportional Representation Society cannot perhaps be accepted as a conclusive guide to the action of voters at a real election, yet the number of Liberals who, in the last of these elections, gave an effective preference to a representative of the Independent Labour Party, in the person of Mr. Henderson, was very noteworthy. In the Belgian system no such fluidity is possible; the Liberal electors would be shut off from any relation with the supporters of Mr. Henderson, who could figure only upon the Labour Party's list.

The different methods of selecting successful candidates.

It will be seen that the problem of allotting seats to lists has been solved in several different ways. Similarly, different methods have been tried for the purpose of selecting the successful candidates from the respective lists. The instructions to voters vary accordingly. The earlier schemes (and the practice obtains in several Swiss cantons to-day) provided that each elector should have as many votes as there were members to be elected, and that he might distribute (without the privilege of cumulating) his votes over the whole of the candidates nominated, selecting, if he desired, some names from one list, some from another, and some from another. After the number of seats secured by each list had been ascertained those candidates were declared elected who, in the respective lists, had obtained the highest number of individual votes.

Panachage.

The practice of voting for candidates belonging to different lists—panachage, as it is called—has evoked considerable discussion, and still gives rise to differences of opinion among the advocates of proportional representation on the Continent. At first sight there would appear to be nothing to discuss, and that there was no possible reason why the elector should not be allowed to exercise his choice in the freest manner. It has, however, been found that this privilege can be used in an unfair way. When each elector has as many votes as there are candidates, and is not permitted to cumulate his votes on any one, it usually happens that the votes obtained by individual candidates in any given list vary but little in number. When in some elections it was realized that the party could only obtain a certain number of seats, but that it had a few hundred votes to spare, some extreme partisans used these votes for the purpose of voting for the least competent men of their opponents' list, and their action sometimes resulted in the election of those men in preference to the more competent men of the party. The danger from this cause would appear to be exaggerated, but although success has seldom attended the abuse of panachage, the fear of a successful attempt has a disturbing influence. The later Swiss laws allow electors to cumulate three votes, but not more, upon any one candidate, so that the success of popular candidates is assured.

The single vote and the case de tête.

The Belgian parliamentary system suppresses panachage, and that in a most effective way. In this system each elector has but one vote, and therefore can only vote for one candidate. In addition, the Belgian system confers upon the organization presenting a list the right to arrange the order in which the candidates shall appear upon the list, and, further, it provides that the voter may approve of this arrangement by voting at the head of the list in the space provided for that purpose and which is known as the case de tête. Party organizations naturally advise their supporters to vote in this way. Public opinion is divided on this feature of the Belgian system, but M. Van den Heuvel, formerly Minister of Justice, who took a responsible part in the passing of the law, and with whom the author discussed this provision, defended it most vigorously, on the ground that the party as a whole had a right to determine which of its members should be elected. In the absence of the provision referred to it might happen that some candidate would be elected in preference to one who was more generally approved of by the party. This may be made clear by an example given by M. Van den Heuvel himself. A, B, C and D are candidates. Suppose that the party is strong enough to return three candidates, but no more, and that five-sixths of the party are in favour of candidates A, B and C, whilst the minority, one-sixth, are ardently in favour of candidate D. It will be necessary that the majority of the party (the five-sixths) should cleverly divide their votes equally between the candidates A, B and C in order to prevent the possibility of candidate D being elected by a small minority of the party. A little reflection will show that in the absence of any such provision the popular candidate of the majority, say A, might attract too large a proportion of the votes, thereby allowing D to pass B or C. Each provision of the Belgian system has been most carefully thought out, and, if it strengthens the hands of party organizations, it does so in order to secure the representation of the party by the candidates most generally approved. It may, however, be pointed out that had the single transferable vote been used, the candidates A, B and C, who, in M. Van den Heuvel's example, were supported by five-sixths of the party, would have been sure of election; there would have been no need to have conferred a special privilege upon the party organizations.

The limited and cumulative vote.

The French Proportional Representation League, which, impressed with the simplicity of the Belgian system, desired to introduce it into France, refrained from advocating the adoption of the case de tête, and suggested that the order in which candidates should be declared elected on each list should be determined by the votes of the electors. The French League in its first proposal recommended that each elector should, as in Belgium, have but one vote. It was soon realized that the popular candidate of the party might attract a large majority of the votes, and that, in consequence, candidates might be elected who were the nominees of only a small section of the party. The League in its second proposal recommended the use of the limited vote, each elector having two votes when six deputies were to be elected, and three in larger constituencies. The League, however, followed the Belgian practice in confining the choice of the elector to candidates on one list. This proposition was examined in 1905 by the Commission du Suffrage Universel, which, in the Report, declared that it was impossible to approve of such a limitation of the elector's freedom. "Nous ne pouvons," runs the Report, "laisser si étroitment enchainer, garrotter, ligotter l'electeur proclamé souverain et qui doit en tout cas être libre." The Committee recommended the use of the limited vote without the restriction recommended by the League. In a further Report, issued in 1907, this Committee again emphasized the necessity of leaving the elector quite free in the choice of candidates, and a new Bill, drafted by the Committee, provided that each elector should have as many votes as there were deputies to be elected, and that he should be allowed to cumulate the whole, or several of his votes, upon any one candidate. Where, however, the cumulative vote has been introduced into recent Swiss laws, as in that of the Canton of Bâle City, the elector is not permitted to cumulate more than three votes upon any one candidate. It will thus be seen that the single vote, the multiple vote without the privilege of cumulating, the limited vote, and the cumulative vote, have all been proposed or adopted as methods of determining which candidates shall be declared elected.

Special characteristics of Swedish and Finnish systems.

This summary of the different methods used in solving the double problem of a list system—the allotment of seats to parties and the selection of successful candidates—is not fully complete.[10] Special features have been incorporated in the Swedish and Finnish systems for the purpose of securing as much freedom of action as possible to electors, and these systems are described in Appendices Nos. III. and IV. The differences between the various list systems are, however, not so great as those between a list system and the single transferable vote, but the consideration of these must be reserved for the next chapter.

[Footnote 1: The text of the Belgian law (Art. 263 of the Electoral Code) runs as follows: "Le bureau principal divise successivement par 1, 2, 3, 4, 5, &c. le chiftre électoral de chacune des listes et range les quotients dans l'ordre de leur importance jusqu'à concurrence d'un nombre total de quotients égal à celui des membres à élire. Le dernier quotient sert de diviseur électoral.

"La répartition entre les listes s'opère en attribuant à chacune d'elles autant de sièges que son chiffre électoral comprend de fois ce diviseur.">[

[Footnote 2: The order in which the names appear is arranged by the party presenting the lists.]

[Footnote 3: A further election (the sixth) took place in 1910.]

[Footnote 4: See La Representation Proportionnelle intégrale, 1910.
Felix Goblet d'Alviella (fils).]

[Footnote 5: Rapport de la Commission du Suffrage Universel, 1905, p. 45.]

[Footnote 6: Professor Hagenbach-Bischoff, of Bâle, formulated a different rule which is finding favour in Swiss cantons. The quota which will ensure the apportionment of all the seats among the lists without remainder is ascertained by trial. In practice the same results are obtained as with the d'Hondt rule. Full directions for applying the rule are contained in Clause XIII. of the law adopted for the canton of Bale Town.—Appendix IX.]

[Footnote 7: For recent French criticism, see page 202.]

[Footnote 8: At Lille, December 1906.]

[Footnote 9: The new French Bill (see Appendix X.) provides for the presentation of combined lists (apparentement).]

[Footnote 10: Cf. La Repésentation Proportionelle en France et en Belgique, M. Georges Lachapelle (1911) and the new report of the Commission du Suffrage Universel (No. 826, Chambre des Députés, 1911). M. Lachapelle recommends a new proposal, le système du nombre unique. The electoral quotient for all constituencies would be fixed by law at, say, 15,000 votes. The number of deputies chosen at each election would be allowed to vary. Each list in each constituency would receive as many seats as its total contained the quotient. The constituencies would be grouped into divisions. The votes remaining over after the allotment of seats in each constituency would be added together, and further seats would then be allotted to the respective lists.]