The (almost really) Complete Works of Lewis Carroll

The question, how to arrange our Constituencies and conduct our Parliamentary Elections, so as to make the House of Commons, as far as possible, a true index of the state of opinion in the nation it professes to represent, is surely equal in importance to any that the present generation has had to settle. And the leap in the dark, which we seem about to take, in a sudden and vast extension of the Franchise, would be robbed of half its terrors could we feel assured that each political party will be duly represented in the next Parliament, so that every side of a question will get a fair hearing.

The method which, after much thought, I venture to propose, will best be explained by showing how it carries out those general principles which ought to guide us in the matter.

(1) That each Member of Parliament should represent, approximately, the same number of Electors is almost an axiom. The monstrous injustice of letting a Member who speaks for a few hundreds cancel the vote of one who speaks for thousands needs no proof. Equal electoral districts, each returning one Member, would seem at first sight to secure this. But they do not really do so, since the Member only represents the majority of the district. Suppose the Electors to consists of 5 millions of one party and 3 millions of the other (call them “blue” and “red”). Now, if these 8 millions were scattered broadcast over the land, the most probable proportion (i. e. the one more probable than any other one), in which they would fall in any district, would be 5 to 3. Such a district would return a “blue” member, as indeed any district would, even with less than 5-8ths “blue,” so long as they were more than one-half. Hence a possible result of an Election would be a House—not containing, as it ought to do, 3 “red” for every 5 “blue,” but wholly “blue.” This extreme case is of course unlikely: but it is mathematically certain that the House would contain much too large a proportion of “blue” to fairly represent the Electors. It seems clear that each district should return several Members, so that Minorities may have a chance of returning some. But, if this be so, there is no reason why the districts should be equal, provided only that the number of Members returned be proportioned to the number of Electors in the district.

(2) That the Minority of the two parties into which, broadly speaking, each district may be divided, should be adequately represented, is another axiom. The common plan for electing several persons at once is to give each voter as many votes as there are vacancies to fill—and this fallacy still holds its ground, in spite of the obvious absurdity that it enables a bare majority of the Electors to fill all the vacancies. The plan, of giving each voter several votes, but fewer than there are vacancies to fill, is only a partial remedy of this injustice: e. g., with 3 Members to return, and each Elector having 2 votes, it is possible for just over 3-5ths of the Electors to fill all the vacancies. The plan, of letting a voter give 2 or more votes to one man, simply increases the “specific gravity”—so to speak—of a vote. Give each voter 6 votes, with permission to lump them if he pleases—and in the end you will find most of the votes given in lumps of 6, and the result much the same as if each had had one vote only. So we are brought to the plan advocated by the “Proportionate Representation” Society—quite the best, I believe, that has yet been suggested—to give each Elector one vote only, and to fix, for each district, the “quota” which shall suffice to return a Member. (The rule is easily seen to be to divide the total number of votes by the number of Members to be returned, increased by one, and to take the whole number next greater than the quotien.)

(3) That the waste of votes, caused by accidentally giving one candidate more than he needs, and leaving another of the same party with less than he needs, should be if possible avoided, is much more easily seen than is any practicable method for effecting this. The packing of votes—needing the constant supervision of a “caucus,” and also a very docile body of Electors, each willing to vote for any man on the “right” side—is a way, but a very clumsy one, for doing this. A much better way would be to let each man vote as he likes, and find some means of utilising surplus-votes, in order to bring in other Members of the same party. But how is this to be done? It is quite the most difficult question we have yet had to face. The P. R. Society says, “let each voter hand in an arranged list, and let his vote, if not required by his No. 1, be transferred to his No. 2, and so on.” But this involves a great difficulty, often pointed out, and never yet successfully grappled with. Which of the surplus-votes are we to transfer? The first thing to settle is, where the answer is to come from. Are we to leave it to chance? Or is some fixed rule to be made, which shall meet all such cases? Or is it to be settled arbitrarily? I must ask your readers’ patience while I discuss these three questions separately.

(3. a.) Shall we leave it to chance? “Yes,” says the P. R. Society: “and you will find that the surplus lists, headed ‘A,’ will be divided among B, C, &c., in the same proportions as the entire set of such lists; and this is surely what the voters wish, and will give an equitable result.” But this is precisely what I have shown, in my letter of June 5, will in many cases not give an equitable result.

(3. b.) Shall we have a fixed rule for transferring the surplus-lists? Yes, if you can find one that will, in all cases, work satisfactorily. None such has yet been suggested.

(3. c.) To this course we seem inevitable driven; namely, that somebody shall decide how to use the surplus votes. But who? The voter? That is what the P. R. Society wants: and it necessitates the arranged lists, which we have already seen cause to abandon. This power cannot be given to the voter: it is impracticable: the only practicable plan seems to be to let him name his one favourite, and leave to other hands the further disposal of his vote, in case it is not used for that candidate. But, what hands? The next idea that suggests itself is, the committee of the candidate. But surely the Electors would not have so much confidence in the committee as in the man himself. And to him I would refer the question, who is to have the surplus-votes. The Elector should be made to understand that, in giving his vote to A, he gives it him as his absolute property, to treat as he will—either by using it to secure his own return, or to help another candidate of whom he approves, or (if there be none such in the field) by leaving it unused. If he cannot trust the man, for whom he votes, so far as to believe that he will use the vote for the best, how comes it that he can trust him so far as to wish to return him as Member? Your readers may, no doubt, find objections to this scheme: but let them remember that what we are in search of is not a scheme free from objections (the quest would be hopeless) but the scheme which has the fewest.

(4) That the process of marking a ballot-paper should be reduced to the utmost possible simplicity, to meet the case of voters of the very narrowest mental calibre, I should have put as an axiom, but that the above-named Society appears to ignore it. No doubt they have found many school-children able to tick off, with great readiness, lists of kings and conquerors in a supposed descending scale of merit—but try it on Hodge, fresh from the plough! Give him a list of half-a-dozen of the neighbouring farmers, to be arranged in their order of merit, and see if he will ever be able to make up his mind! “I knows who’s best of two,” he might tell you, “but blessed if I can say who’s first, and second, and third, and fourth!” This simplicity of process is secured by my method.

(5) That the process of counting votes should be as simple as possible will be admitted by all who agree (as who does not?) that the sooner the result can be announced, and the less liable it is to be set aside owing to errors of calculation, the better. This also is secured by my method.

I proceed to give a summary of rules for the method I propose. Form districts which shall return 3, 4, or more Members, in proportion to their size. Let each Elector vote for one candidate only. When the poll is closed, divide the total number of votes by the number of Members to be returned plus one, and take the next greater integer as “quota.” Let the returning-officer publish the list of candidates, with the votes given for each, and declare as “returned” each that has obtained the quota. If there are still Members to return, let him name a time when all the candidates shall appear before him: and each returned Member may then formally assign his surplus-votes to whomsoever of the other candidates he will; while the other candidates may in like manner assign their votes. If, by this process, any fresh candidate obtains the quota, let him be declared “returned.”

This method would enable each of the two parties in a district to return as many Members as it could muster “quotas,” no matter how the votes were distributed. If, for example, 10,000 were the quota, and the “reds” mustered 30,000 votes, they could return 3 Members: for suppose they had 4 candidates, and that A had 22,000 votes, B 4,000, C 3,000, D 1.000: A would simply have to assing 6,000 votes to B, and 6,000 to C, while D, being hopeless of success, would naturally let C have his 1,000 also. There would be no risk of a seat being left vacant through two candidates of the same party sharing a quota between them: an unwritten law would soon come to be recognised—that the one with fewest votes should give place to the other. And, with candidates of two opposite parties, this difficulty could not arise at all: one or the other could always be returned by the surplus-votes of his party.