CHAPTER IV.

THE REFORM: TRUE PROPORTIONAL REPRESENTATION.

How to give effect to the principles of organization and leadership in an electoral reform--that is the problem which we shall now attempt to solve. We have already laid down the theoretical requirements, which are (1) proportional representation to the two parties--the majority and the minority, and (2) the election by each party separately of its most popular leaders; and we shall now have to consider also how these requirements are modified by practical considerations.

+Proportional Representation to the Two Parties, the Majority and the Minority.+--It will be as well to illustrate the method proposed by reference to the conditions imposed by an actual election, such as that for the Federal Senate. The Commonwealth Bill provides that each State shall be polled as a single electorate, returning six senators. Suppose that 120,000 electors vote on party lines in any State. It is clear that a party which has the support of 20,000 electors is entitled to one senator; also, that a party which has the support of 40,000 electors is entitled to two senators; of 60,000 electors to three senators, and so on. Now, suppose that one party has the support of 50,000 electors, and the other of 70,000 electors, then the minority is entitled to two and a half senators, and the majority to three and a half senators. But senators are living units, and cannot be divided into fractions. The question therefore arises, Which is entitled to the odd senator, the majority or the minority? And the answer is that they are both equally entitled to him; for it is as much a tie as if each party has the support of 10,000 electors in a single-seat electorate. But if the minority had the support of 49,999 electors, or one elector less, it would be entitled to only two senators, and if it had the support of 50,001 electors, or one elector more, it would be entitled to three senators.

From the above simple facts can be deduced general rules applicable to any particular case. It is evident that the result is not affected by the number of votes allowed to each elector, providing only that each elector has the same number of votes. It is also quite irrespective of the number of candidates nominated in the interests of each party. But it would never do to allow party organizations to control nominations. How are we to combine individual candidature with party nomination? The only way to do this is to require that each candidate shall declare, either when nominating or a few days before the election, on which side of the House he intends to sit, and be classified accordingly as Ministerialist or Oppositionist. To decide the relative strengths of the two parties, it is then only necessary to take the aggregate votes polled by all the candidates nominated for each party as a measure of the amount of support which it receives.

The great advantages of this provision are at once apparent. There is no incentive to limit the number of candidates so as to prevent splitting the votes. On the contrary, it is to the interest of each party to get as many strong candidates as possible to stand in its interests. There will be no necessity to ask any candidate to retire for fear of losing a seat to the party. Thus the control of nominations, which leads to the worst abuses of the present system, will be entirely obviated.

Now, suppose that in the instance we have already given each elector is allowed to vote for one candidate only, the total number of votes recorded will be 120,000. Then the _unit of representation_ or number of votes which entitle a party to one senator will be 20,000 votes; each party will be entitled to one senator for every whole unit of representation, and the odd senator will go to the party having the larger remainder. For instance, if the aggregate votes polled by all the Ministerialist candidates be 72,000, and by the Oppositionist candidates 48,000, the Ministerialists, having three units plus 12,000 remainder, are entitled to four senators, and the Opposition, having two units plus 8,000, to two senators.

Similarly, if each elector be allowed to vote for a number of candidates, all these figures will be increased in proportion. For example, if each elector has three votes, the unit of representation would be 60,000 votes. The following general rules may therefore be stated:--

1. The unit of representation is equal to the total number of valid votes cast at the election, divided by the number of seats.

2. Each party is entitled to one seat for every whole unit of representation contained in the aggregate votes polled by all its candidates, and the odd seat goes to the party which has the larger remainder.

The fact that the last seat has to be assigned to the party which has the larger remainder is sometimes advanced as an objection, but it is evidently the fairest possible division that the size of the electorate will permit. Of course, the larger the electorate the more accurately proportioned will be the representation. Hence the representation would be most accurate if the whole assembly were elected in one large electorate. But if, for the sake of convenience, the assembly be elected in a large number of electorates in which the relative proportions of two parties vary the gains which a party makes in some electorates will be balanced by losses in others, so that the final result would be almost as accurate as if the whole country were polled as one electorate. It must be remembered that the result in any electorate cannot be foreseen, and that it is a matter of chance which party gains the advantage. Now, if the limits of variation comprise even a single unit of representation, each party will stand an equal chance of gaining, and therefore the laws of chance will ensure that the gains balance the losses in the different electorates. Supposing a party which averages 40 per cent. in the whole country to vary between 30 per cent. and 50 per cent, in the different electorates (which may be taken as a fair assumption), the unit of representation should equal 20 per cent., or one-fifth. Under these conditions the laws of chance will ensure correct representation, so long as the electorates do not contain less than five seats.

The above facts furnish a complete answer to the arguments advanced by Mr. J.W. M'Cay, ex-M.L.A., in a series of articles in the _Age_ against the application of proportional representation to the Federal Senate. While apparently recognizing that it is utterly impossible for the minority to secure a majority of the representation, he based his objection solely on the fact that a minority is able with electorates containing an even number of seats to secure one-half of the representation, and thus lead to what he terms "the minority block."

The force of the objection will entirely depend on the size of the minority which is able thus to thwart the will of the majority. The Federal Senate will consist of 36 senators, each of the original States contributing six. No reasonable man would complain if the minority, being only entitled to 17 senators, actually returned 18, but Mr. M'Cay points out that it is possible for a minority entitled to 15 senators to return 18. To bring about this result he makes the absurd assumption that in each of the six States the minority polls exactly two whole units of representation, and a bare majority of a third unit. It is safe to say that this would not happen once in a thousand years. If the relative proportions of the two parties vary in the slightest in the different States some must be under and some over the assumed proportion. It is most probable that it will be under it in three States and over it in the other three States; and, under these circumstances, the party will return 15 senators, the exact number to which it is entitled. It may happen to be under the assumed proportion in only two of the States and over in the other four, and that the party will get one more senator than it is entitled to; but it is extremely improbable that it will get two more, and virtually impossible that it will get three more senators than its just proportion. Mr. M'Cay's conclusion that proportional representation can only be used in electorates returning an odd number of representatives is shown to be entirely unwarranted. Equally fallacious is Professor Nanson's rebutting statement that "scientific proportionalists recommend odd electorates." While the number of States remains even, the mathematical chance of a minority securing one-half of the representation is precisely the same whether the States return an odd or an even number of senators. As a matter of fact, the danger of a minority securing one-half of the representation is much greater at the intermediate elections for the Senate, when each State returns three senators, the reason being the smaller field.

We have dwelt at some length on the preceding example, because it serves to refute another error into which some of the proportionalists have fallen. It is held that the unit of representation should be ascertained by dividing the total votes, not by the number of seats, but by the seats increased by one. This unit is generally known as the Droop quota, having been proposed in a work published by Mr. H.R. Droop in 1869. Since one vote more than one-half of the total votes is sufficient for election in a single-seat electorate, it is argued that one vote more than one-third suffices in a two-seat electorate, one vote more than one-fourth in a three-seat electorate, and so on. The unit in a six-seat electorate would be one-seventh of the votes instead of one-sixth, and it is pointed out that by this means the whole six seats would be filled by whole units, leaving an unrepresented residuum of one-seventh of the votes divided between the two parties.

The error lies precisely as before in concentrating attention on one of the electorates, and in neglecting the theory of probability. The Droop quota introduces the condition that each party must pay a certain minimum number of votes for each seat, and the real distinction is that, instead of the minority and the majority having an equal chance of securing any advantage, the chances are in the same proportion as their relative strengths. If the majority be twice as strong as the minority, it will have twice the chance of gaining the advantage. To prove this, consider the position of a one-third minority in a number of five-seat electorates. The Droop quota being one-sixth of the votes, the minority will secure two seats or 40 per cent. in those electorates where it is just over one-third, and one seat or 20 per cent. where it is just under. Since the mathematical chances are that it will be over in one half and under in the other half, it will, on the average, secure only 30 per cent., although entitled to 33 per cent. Again, if the 670 members of the House of Commons were elected in three to five-seat electorates, and the Droop quota used as proposed by Sir John Lubbock, and if the Ministerialists were twice as strong as the Oppositionists, they would, on the average, return 30 more members than the two-thirds to which they are entitled, and this would count 60 members on a division.

The following table illustrates the erroneous result obtained by applying the Droop quota when a number of grouped-electorates are concerned. It will be noticed that where parties are nearly equal it makes very little difference which unit is used:--

+-------------+--------------------------------------------+ | STRENGTH OF | AVERAGE REPRESENTATION. | | +----------------------+---------------------+ | PARTY |Five-Seat Electorates.|Ten-Seat Electorates.| +-------------+----------------------+---------------------+ | 10 per cent.| 2 per cent. | 6 per cent. | | 20 " " | 14 " " | 17 " " | | 30 " " | 26 " " | 28 " " | | 40 " " | 38 " " | 39 " " | | 50 " " | 50 " " | 50 " " | +-------------+----------------------+---------------------+

The Droop quota, therefore, gives, not proportional, but disproportional representation.

+Election by Each Party of its Most Popular Candidates.+--Still keeping in mind the six-seat electorate for the Federal Senate, we may note that there are two rival systems in the field--the _scrutin de liste_ or Block Vote, in which each elector votes for any six of the candidates, and the Hare system, which allows each elector an effective vote for one candidate only. The adoption of either of these systems would be unfortunate. To force each elector to vote for six candidates is probably to require him to vote for more than he is inclined to support, and certainly for more than his party is entitled to return; and, also, to put it in the power of the majority to return all six senators. To allow him to vote for one candidate only, on the other hand, is to break up both parties into factions by allowing the favourites of sections within the parties to be elected, instead of those most in general favour with all sections composing each party. An intermediate position is therefore best. No elector should be required to vote for more than three candidates, and no elector should be allowed to vote for less. Because in the first place it is evident that each party will, on the average, return three senators, and, secondly, it may be taken for granted that even the minority will nominate at least three candidates. Two alternative proposals may be submitted as fulfilling these conditions:--

_1. Each elector should vote for any three candidates, or

2. Each elector should have six votes, and have the option of giving two votes to individual candidates._

The first plan is the simpler, but the second is probably the better, as it allows more discrimination without sacrificing any of the advantages. Either proposal is practically equivalent to applying the Block Vote to each party separately; and whatever may be the objections to applying the Block Vote to two or more parties it is the simplest and best system to elect the candidates most in general favour when one party only is concerned. It is true that the majority will return rather more than one-half of the representatives and the minority rather less than one-half, so that the minority will have more votes in proportion to its strength. But with two parties of fairly equal but fluctuating strength the fairest way is to require each elector to vote for at least one-half of the number of representatives. Besides, apart from the fact that it is not known before the election how many seats each party will obtain, it is absolutely necessary that each elector shall have the same number of votes in order that each party be allotted its just share of representation. Moreover it is not proposed to limit the elector's freedom of choice in the slightest by confining him to the candidates of one party. The great majority of electors will vote on party lines, because every vote given to a candidate of the opposing party tells against the representation of their own party. The reason of this is that every vote counts individually for the candidate and collectively for the whole party. Any elector, therefore, who divides his voting power equally between the two parties practically wastes it as far as the party representation is concerned. But it is neither necessary nor desirable to bring about such a rigid party division as prevails in America, for instance, where a man is born, lives, and dies Republican or Democrat. If electors were confined to the candidates of one party, an elector who wished to vote for an individual candidate of the opposing party would be placed in the dilemma of deserting either his favourite or his party. The division into parties is really required in the elected body, and not in the constituent body.

+Rules for the Reform.+--We are now in a position to draw up a list of rules for the proposed reform, applicable to all legislatures in which party government prevails:--

1. Electorates to be grouped so as to contain at least three seats, and preferably not less than five seats nor more than twenty seats.

2. Candidates to declare when nominating, or a few days before the election, whether they are in favour of or opposed to the party in power, and to be classified accordingly as Ministerialists or Oppositionists.

3. Ballot papers to contain the names of all candidates nominated, arranged in two parallel columns, one headed Ministerialists, and the other Oppositionists. The list of candidates under each heading to be arranged in alphabetical order.

4. Each elector to have as many votes as there are seats, and to be allowed to give either one or two votes to any candidate. The votes to be distributed as he pleases among all the candidates of both lists.

5. The total number of valid votes cast at the election to be divided by the number of seats; the quotient to be known as the "unit of representation."

6. Each party to be allowed one seat for every whole unit of representation contained in the aggregate votes polled by all its candidates, and the last seat to go to the party which has the larger remainder.

7. The candidates of each party having the highest number of votes to be declared elected to the number of seats to which each party is entitled in accordance with the preceding rule.

8. In case of a tie between candidates or parties the lot decides.

The alternative plan for rule 4, which is somewhat simpler, would read as follows:--

4. Each elector to vote for half the number of candidates that there are seats, _i.e._, three votes in a five or six-seat electorate, four votes in a seven or eight-seat electorate, &c. The votes to be distributed as he pleases among all the candidates of both lists.

It is unnecessary to dwell on the absolute simplicity of these rules. They involve no radical departure from existing methods of voting or of counting votes. Once the totals are added up, the calculations necessary to decide the successful candidates are within the reach of a school child.

EXAMPLE.--Take as an example 13 candidates in a six-seat electorate who poll as follows:--

MINISTERIALISTS. OPPOSITIONISTS. BROWN 83,000 YOUNG 53,000 RYAN 74,000 BELL 51,000 COX 44,000 HUME 47,000 WHITE 42,000 JONES 45,000 PEEL 38,000 BLACK 34,000 ADAMS 35,000 ------- GREY 33,000 230,000 SWIFT 21,000 ------- 370,000

Total votes = 370,000 + 230,000 = 600,000.

Unit of representation = 600,000/6 = 100,000.

Ministerialists: 3 units + 70,000 remainder = 4 seats.

Oppositionists: 2 units + 30,000 remainder = 2 seats.

The Ministerialists, having the larger remainder, secure the last seat. The successful candidates are Brown, Ryan, Cox, and White (M.), Young and Bell (O.)

It will be noted that without the proportional principle the Ministerialists would have returned two members only, and the Oppositionists four.

It is to be distinctly understood that the simpler plan of voting for half the number of candidates that there are seats is practically as good as the other. In order to show, however, that the plan we have favoured may be simplified, we illustrate by a sample ballot paper a method which has been used in Belgium. Two white spots are printed opposite each candidate's name. An ink pad and stamp are then provided at each polling booth, and the elector stamps out a white spot for each vote he wishes to give. In the paper illustrated two votes are given to Brown, two to Jones, one to Grey, and one to Swift. This elector has, therefore, given two-thirds of his voting power to the Ministerial party, and one-third to the Opposition, and has thus directly influenced both policies. A further advantage of the proposal is the ease with which such a paper can be read by the returning officer.

BALLOT PAPER

* * * * *

Ministerialists. Oppositionists.

+---------------+ +---------------+ | | | | | | | o o | ADAMS | | o o | BELL | | | | | | | +---------------+ +---------------+ | | | | | | | x x | BROWN | | o o | BLACK | | | | | | | +-----+---------+ +---------------+ | | | | | | | o o | COX | | o o | HUME | | | | | | | +-----+---------+ +---------------+ | | | | | | | o x | GREY | | x x | JONES | | | | | | | +-----+---------+ +---------------+ | | | | | | | o o | PEEL | | o o | YOUNG | | | | | | | +-----+---------+ +---------------+ | | | | o o | RYAN | | | | +-----+---------+ | | | | x o | SWIFT | | | | +-----+---------+ | | | | o o | WHITE | | | | +-----+---------+

1. You are allowed Six votes, and can give either one or two votes to any candidate on either list.

2. Stamp out one of the white spots if you wish to give a candidate one vote.

3. Stamp out the two white spots if you wish to give a candidate two votes.

4. Your ballot paper will be invalid if you stamp out more or less than Six white spots.

+Character of Parties.+--We must now prove that the methods proposed will actually organize the people into two coherent parties. Let us suppose either party to be composed of three sections. The problem is to induce these three sections to work together, and to sink their petty differences in the general interest, in short to unite as a party, aiming at the control of administration with a definite policy on public questions. Let us further suppose the party entitled to three representatives. Now, it is quite conceivable that exactly the same three candidates would be elected if each elector had any number of votes from one to three, and this would actually tend to be the case the more united the party is. But herein lies the difference: that with one vote only any one section holding narrow and violent views can return an independent delegate, and therefore has a direct inducement to do so, while with three votes it is forced to work with the other two sections, for if it refuses to do so it is in their power to exclude its nominee. It is this power to exclude independent factions which is the first requisite to prevent the main parties degenerating into factions. Now, the advocates of the Hare system declare that each elector should have one effective vote only, no matter how many seats the party is entitled to. The elector would therefore only express his opinion as to the delegate of his own section, and not as to the constitution of the whole party, and there would be nothing whatever to prevent the election of the favourites of sections, instead of the representatives most in general favour with all sections.

But if there were only one party it would be impossible to make all the sections work together in this manner. Some of them would combine into a majority of the party, and would exclude the minority. With two great competing parties, however, the case is quite different. So far from either party wishing to exclude any small minority, both will compete for its support, providing only that it will fall into line with the other sections on the main questions of policy. Each section will therefore support the party which will consent to embody the most favourable compromise of its demands in its policy. If its demands are such that both parties refuse to entertain them, it will exercise no influence in the direction of furthering its own views. From this statement it is evident that no system of independent direct proportional representation within the party can be recognized as a right to which the different sections are entitled, as it would inevitably break up the party, and lead to sectional delegation. The sections would then change in character, and become violent factions. But, nevertheless, if the sections work together as described, every section will be proportionately represented in the party policy, and therefore by every representative of the party. Moreover, no section can dictate to either party, or obtain more than a fair compromise. For all the sections are interdependent, and any section which attempts to exert more than its just share of influence will sink in general favour, and will find those who are inclined to support its pretensions rejected at the election.

The difference between the two stages of representation may now be clearly appreciated. In the first stage we have seen that the fear of the aggression of the monarchy held all sections together in one party. In the second stage, however, it has been abundantly demonstrated by experience that the fear of each other will not hold the sections of the two parties together. The electoral machinery must, therefore, supply the deficiency.

+Party Lines.+--With the altered character of parties there is ground for hope that the basis of division will become questions of general public policy, and that all causes of factious dissension and of social disruption will tend to be repressed. This improvement is indeed urgently needed. For if in any country party lines are decided by geographical considerations, as town _v._ country; by class, as Capital _v._ Labour; by race as in South Africa; by religion as in Belgium; or by personal ambition for the spoils of office--in any of these cases the future of that country is open to the gravest doubt.

Perhaps the greatest danger which assails most democratic countries to-day is the risk of the working classes being persuaded by demagogues that equal political rights have been extended to them in order that they shall govern, instead of in order that they shall not be misgoverned. If the general welfare is to be advanced, all classes must influence the policies of both parties. This condition is indispensable to bring about the ideal condition of two parties differing only as to what is best for all.

Equally to be condemned is the narrow-minded and intolerant view of those who can see no virtue in an opposing party; who define, for instance, the distinction between parties as the party for things as they are, and the party for things as they ought to be; the latter being, of course, their own party. This is one of the objectionable features of Australian newspaper-made politics.

A more rational view of the distinction which often underlies party divisions is between those who desire change and those who oppose change. J.S. Mill points out how the latter may often be useful in preventing progress in a wrong direction. There are times when such attitude is called for, but generally speaking we may say that the fundamental distinction between parties should be a difference of opinion as to the direction of progress. Nor is it inconsistent for a party to change its opinion or alter its policy; on the contrary, it is essential to progress. The majority must often modify its policy in the light of the criticism of the minority, and the minority must often drop the unpopular proposals which have put it in a minority. These features are all essential to the working of the political machine.

+The Character of Representatives.+--Granting that all sections of each party can be induced to work together, the beneficial effect on the character of representatives would be incalculable. Instead of being forced to pander to every small section for support, they would appeal to all sections. The enlarged electorates which are contemplated would be arranged to embrace the widest diversity of interest, and a representative would then be free to follow his own independent judgment, unfettered by the dictation of small cliques. His actions might offend some sections who supported his election; but he has a wide field, and may gain the support of other sections by them. Therefore, he may actually improve his position by gaining more supporters than he loses. Contrast this with the present system, in which the representatives are cooped up in single-membered electorates to denned sets of supporters. The very principle of community of interest on which these electorates must be arranged in order to get a fair result is destructive of the idea of representation. It is no wonder, then, that the present system is tending towards delegation. Local delegation we have always had, more or less, but we are now threatened by class delegation also.

The conclusion of Mr. Kent in "The English Radicals" may be quoted on this point. He says:--

The question of the relationship of members to their constituents is at the present time perplexed and undetermined; for though the control of Parliament by the people is an indisputable fact, yet it is maintained by means of quite another kind from those which the early Radicals proposed. The result is somewhat paradoxical, for while the system of pledges has been contemptuously rejected, yet the theory that a member is a delegate tacitly prevails in English politics. That members of the House of Commons have tended and do tend to lose their independence it is impossible to doubt. A distinguished French publicist, M. Boutmy, for instance, has remarked the fact; and he thinks that in consequence a deterioration of the tone of politicians is likely to recur. Mr. E.L. Godkin, an American writer, whose judgments are entitled to respect, has expressed much the same opinion; "the delegate theory," he says, "has been gaining ground in England, and in America has almost completely succeeded in asserting its sway, so that we have seen many cases in which members of Congress have openly declared their dissent from the measures for which they voted in obedience to their constituents."

It is one of the greatest merits of the proposed reform that this vexed question of representation or delegation would be definitely settled. For, although the area of independent action is enlarged, definite limits are set to it.

+Possible Objections.+--We may now reply to some objections which have been or might be urged. At the outset we would point out that the critics nearly always base their objections on the conditions which have prevailed in the past or do exist in the present chaotic state of parties; and seldom appreciate the fact that they would lose force if a better condition could be brought about. Let us take the Melbourne _Argus_ report of Professor Nanson's objections:--

Professor Nanson pointed out that the scheme depended for its efficacy on the existence of party government, which the Professor was glad to say was being pushed more and more into the background. He took a practical illustration from the defeat of the O'Loghlen Government in 1883. In that case, after the election the Government came back with a following of one-tenth. The other combined party had nine-tenths, and of these a little more than half were Liberals and a little less than half were Conservatives. He pointed out that under Mr. Ashworth's system the Liberals would have got the whole of the Opposition seats and the Conservatives none, whereas under any intelligent modification of the Hare system the parties would have been returned in the proportion of five Liberals, four Conservatives, and one O'Loghlenite. The system contained the evils of the _scrutin de liste_ doubled by being applied to two parties, the evils of the Limited Vote, which had been condemned by all leading statesmen, and it played into the hands of these who were best able to organize.

Take the latter statements first. The evil of the Block Vote or _scrutin de liste_ is that it gives all the representation to the majority, and excludes the minority; its merit is that it prevents the formation of a number of minorities. How this evil will be doubled if it is entirely removed by allowing both majority and minority their just share of representation we leave the Professor to explain. The statement that the scheme would play into the hands of those who are best able to organize is absolutely without foundation. On the contrary, the organization is automatic. It would certainly encourage the formation of organizations to influence the policies of the parties, since every organization would be able to exert its proportionate influence, but that is an advantage, not an evil. We will leave the statement about party government alone, and now take the "practical illustration." The Professor here assumes three distinct parties, but it is quite evident there are only two. It is not usual for Liberal Unionists and Conservatives to fight one another at elections in Great Britain at present. In the same way, if a section of Liberals and a section of Conservatives unite to oppose a Government, they will work together and not try to exclude one another. Moreover, they will have a common policy, so that it matters little who are elected so long as they are the best men to carry out the policy. Is it likely the Conservatives would join the Liberals, if the latter were trying to get all the seats? Thus all the Professor's assumptions are incorrect. But even if they were correct the conclusion is still wrong. The Liberal section could not get all the seats if they tried. Imagine a ten-seat electorate, in which the combined party is entitled to nine members. The electors would not be required to vote for more than five candidates, whereas the Professor has assumed that they would be forced to vote for nine. He has forgotten that the Block Vote becomes the Limited Vote under the conditions named, and that the Limited Vote allows the minority a share of representation. Besides, in any case, these conditions would never arise in a country in a healthy state of political activity, because then parties would tend more nearly to equalize each other in strength.

It has also been objected that a Ministerialist candidate, say, might stand as an Oppositionist, if the votes of the Opposition candidates were more split up and it was likely to require less votes for election in that party. This is a rather fantastic suggestion. The candidate in question would have to declare himself in favour of a number of things which he would oppose immediately he was elected. If not, he would have to openly declare his intention, but that could easily be made illegal. In any case there would be very little gained, and there is further the risk that, if defeated, all his votes would count to the Opposition.

Another possible objection is that too many candidates might stand, since it is to the interest of each party to get all the support it can. But candidates are not likely to stand to oblige the party or when there is no chance of being elected. It is quite possible that, in a country already split up into numerous groups, the groups would refuse to act together, and that each group would nominate its own list. This is an extreme assumption, and certainly would not happen in British countries. And there would be a constant incentive to the groups to compromise, since a combination can return its candidates.

We hope now to have at least established the fact that the organization of a democracy into two coherent parties--a majority and a minority--is vitally connected with the electoral machinery. We do not claim that the method we have proposed will induce a people to vote on true party lines all at once, for human nature cannot be changed in a day; but we do confidently assert that it will greatly accelerate that desirable result, and will tend to give effect to the principles of organization and responsible leadership.