ALTERNATIVE VOTE

Evening Post, Volume XCVIII, Issue 18, 22 July 1919, Page 2

ALTERNATIVE VOTE

TO THEiBDITOR. "*

Sir,—Chi 22nd May, 1909, in reply to a deputation that urged reform in the method of representation—which wa6 practically. identical with that obtaining in New Zealand to-day—Mr. Winston Churchill stated that "it was impossible to defend the present rough and ready methods." "I think," said he, "the present system has clearly broken down. The results produced are not fair to any party or to any section of the community. In many cases they do not secure majority representation, nor do ' they secure an intelligent representation of minorities. All they secure is flake representation, freak representation, capricious representation." Most, if not all, of those who have thought seriously '■ about the matter, will agree with the view expressed so forcibly by Mr. Churchill, and this, I take it, was the reason that prompted Mr. H. F. yon Haast, in his letter of 11th July, to suggest the adoption of the alternative vote as a cure for the evils of our present system. There are two ways known to British communities of applying the principle of the alternative (or contingent) vote—in one, the Queensland method, after the first count (if no one obtains an absolute majority of all the votes), all the candidates "except the first two are rejected and their rotes are transferred to one or other of the.first two, according to the nest choices of the electors. In the other, the West Australian method, only the lowest candidate is rejected, and at the second count his yotes are distributed as indicated by his supporters among the remaining candidates. If still no candidate has obtained an absolute majority, the votes of the then lowest candidate are distributed; and so on, Antil one candidate obtains an absolute majority. Obviously, the West Australian method (which is also the method proposed in various Bills; that have -come before the British Parliament) is less fluky and open to less' objection than the Queensland method, for it is quite possible that if there be, say, four candidates, the transferred votes of the fourth candidate may put the third candidate above the other two; yet the Queensland method would cut him out. I understand Ok: yon Haast to assume that the West' Australian method is to be used. But even by it a,.candidate may be rejected who is preferred by,* majority of the electors to the candidate finally chosen. In fact, the method does not secure the return of the best candidate out of three or more, all it does is to secure the defeat of the weakest candidate. • It is a palliative of the evils caused by the present system, not a cure. Let us take a simple example. Suppose there are three candidates, A, B, C, and' 400 electors, who vote thus:

In tb\ order of A, B, C (that is Al, 82, C 3), 20 votes.

In the order A, C, B (that is, Al, 83, C 2), 120 votes.

In' the order BAG, 40 votes. In the order BCA, 110 votes. In the order CBA, 40 votes. In the order CAB, 70, votes.

So that 140 «lectors put A first, 150 put B first, and 110 C first. By the "first past the post" method, B is elected ; but if we look- at the votes we see that 210 electors out of 400 prefer A to B (W0 plus CAB 70), and 230 electors prefer Cto B (110 plus ACB 120). With1 either method of using the alternative vote, C as. the lowest candidate on the firet count is.rejected, and his 110 votes are transferred, 40 to B, and 70 to A, so B has now 190 votes (150 plus 40) and A has 210 (140 plus 70). A is therefore declared elected. But C beats A on .220 papers (110,, plus BOA 110); and he :beats Bon 230 papers (see above); that. i 6, the electors prefer C to either .A or B; in short, Cis the true choice of the electors. We ought, therefore, to count the votes in such a manner that C is elected, if it can be done. There is such a method of counting, when, the riuiriber of candidates is three or more, which in every case gives the true result without a mistake. The method is simple, and could be easily applied by any returning officer, but to prove its truth a _ little rhathematics is necessary. (I will not inflict upon you the mathematical proof of the' absolute accuracy, of the method, which appears in volume 46 of the Transactions of the New Zealand Institute, but I enclose ia copy of the article containing ths proof.; > It is only fair to admit that even the present method in many cases gives the correct result, and that the West Australian method would probably give the correct result in the majority of instances—but in would fail in a considerable number of cases, possibly in quite a sufficient number to alter the relative strength of parties in Parliament.

"The second ballot has led in France to a. great traffic in votes, and Australian, experience, seems likely to give the same result \n the case of the alternative vote" . (Mr. Holman, Premier of New South Wales, reported in the Sydney Daily Telegraph of 20th August, 1913). "On either eystem the weakest party's vote is up at auction between the ' two larger parties. Both bid for its support.-

... In either case the member elected is, in the,, French political phra-se, 'the prisoner of the minority' ; he is- bound by pledges to voters who do not agree with'his own views" (J. Fischer Williams, "Proportional Bepreseiitation and British Politics."*

It has been urged, that the exercise By the electors of the' choices required for the alternative - vote wauld accustom than to the method of voting for proportional representation. This is, no doubt, true to some extent; but it would be just as easy to exercise the necessary preferences for P.R at once—no intermediate step, by the temporary adoption of an imperfect system, is required. Now, (1) The electors should clearly understand how to vote; I see no difficulty in getting them to express their order of preference to a reasonable extent (down, say, to five, or six candidates) by the numbers 1, 2, 3. (2} They should also be convinced that the method of counting always gives the true result according to the preferences expressed; that is, that, subjecT to the ordinary allowance for human error, the method' is fair.

(3) Will they expect to have this fact clearly proved? I think a considerable number of electors will expect' it, and will not accept a system where they do hot themselves understand why the method of counting is th.c correct one. Now the counting in the alternative voting system is on any method easy to follow, but the reason for the true method of ■counting is not quite bo easily understood. In fact, the counting under the Proportional Representation system, and its accuracy and fairness, are more easily made clear to the non-mathematical mind. So that, really, proportional representation would in the process itself' be more satisfactory to the average elector than the alternative vote in *■ single-member ■constituencies. Proportional representation has, of course, many other things in its favour.

Even if the alternative vote gives satisfactory results in the several constituencies, we liave to consider whether it secures a true representation of the whole electorate of the country—which depends largely upon the distribution of the party voting power among the constituencies. It is obvious that, if there are two parties, one with a- total electoral majority concentrated' in a few constituencies may be at a disadvantage compared with a party that has a y small majority in each of a. large number of single electorates, 4tbaugfe-Ofl..tiie. Kb.ole._:.it Js.in.

a minority in the country. For instance, let^ us suppose that the whites' have majorities of 2000 each in five places (10,000 in all), and that the blacks have majorities. of 200 each in fifteen places (3000 in all) ; then the blacks will have fifteen members, and the whites five members, although the latter have a total net majority of 7000 in the twenty places taken together.

That such a thing may happen is not merely a supposition; in the British General Election1 of 1886 the Liberal Home Rulers had' in all constituencies 2,103,954 votes, and the Unionists 2,049,137; the seats obtained were Liberal Home Rulers 283, Unionists 387. ''This election was regarded as a crushing defeat for Mr. Gladstone. He found himself in the Hous^ of Commons in a. minority of 104; but his supporters in the country were in a majority" (Humphreys, "Proportional Representation).

This is only one out 'of many instances that might be quoted from the experience of many countries." The circumstances were that nearly all the constituencies were single-member constituencies, and that it was a straight-out fight between two parties. .-' It is very easy to see that" the result would probably have been still more unsatisfactory if there had been three parties. Proportional representation gives a true reflection of the strength of nil parties in a community, not only in theory, but in practice. Experience has shown, in every election held under any system of proportional representation, each party obtains a number of representatives almost exactly in proportion to the-num-ber of its supporters. I cannot imagine a more truly democratic principle of representation than this.

The difficulties of proportional representation have been greatly exaggerated, and with your permission I propose in another letter to endeavour to explain the system as simply as possible.—■ 1 am, etc.,

GEORGE HOGBEN. Kha-ndallah, 17th July, 1919.