II.

A TEST ELECTION.

AMONG the assertions which do duty as arguments against the adoption of proportional representation, none is more plausible, or is more frequently used, than that which declares that the inevitable complication of the system must be a fatal bar to its use in popular elections. Being anxious to test the real value of this contention, I recently made an experiment, the result of which may, I think, be of interest to readers of this Review, and which certainly goes some way to prove that the argument referred to is about as conclusive as most others put forward by the advocates of pure majority representation. It occurred to me that if it could be shown that the system of proportional voting might be easily understood and made use of by children under fourteen years of age, and if the process of counting the votes recorded could be successfully and rapidly conducted by persons of ordinary intelligence, unaided by previous practice or mechanical appliances, a great step would have been made towards proving that the plan might be attempted on a large scale with a fair hope of success. I admit, at the outset, that the average intelligence of children in the upper standards of a good elementary school is probably higher than that of a large number of voters at the present time. Still, inasmuch as the scholars of to-day must be the voters of to-morrow, the comparison cannot be said to be unfair. The following is a brief account of the experiment I made:-By the kind permission of the Rev. William Sinclair, of St. Stephen's, Westminster, I was permitted to conduct a test election in the elementary school attached to his church. I selected seven candidates whose names were likely to be familiar to the children. The following is the order in which they were placed upon the voting paper:-King Charles I., Queen Elizabeth, King Henry VIII., Mary Queen of Scots, Oliver Cromwell, the Duke of Wellington, and William the Conqueror. The electors numbered 184, of whom 131 were boys, and 53 girls.

Three members were to be elected. It was plainly necessary to

supply a certain amount of information to take the place of common knowledge. The work of newspaper articles, political agencies, and current conversation had to be taken into account. In order, therefore, to put the children upon a level with the ordinary voter, Mr. Sinclair in a few words explained to them the following facts:

1. That they were supposed to be voting for members of Parliament.

2. That each voter had only one vote, which, however, might be transferred according to the numbers marked upon the voting papers.

3. That there were two parties, Liberal and Conservative. The boys were to be Liberals, with the following candidates, Henry VIII., Oliver Cromwell, the Duke of Wellington, and William the Conqueror. The girls were to be Conservatives, and their candidates were to be Charles I., Elizabeth, and Mary Queen of Scots.

4. Anybody might vote for one of the other party if he or she very much wished it.

As a supplement to the verbal explanation, a placard to the following effect was posted in the room :—

Each voter has one vote.

Instructions to Voters.

That vote will be given first to the candidate against whose name you put 1. If that candidate has enough votes to secure his election without your vote, it will be given to the candidate against whose name you put 2.

If the candidate against whose name you put 2 has enough votes without your vote, it will be given to the candidate against whose name you put 3.

And so on.

It is not necessary to put numbers against more names than you wish.

This form of instruction, for which I am indebted to a friend, appears to me an almost ideally concise and complete formula for the purpose. The voting was conducted by Mr. Blennerhasset, M.P., and myself, first in the boys' schoolroom, then in that of the girls. There were two polling stations, and the votes were recorded with perfect order and in a very short time.

At first the boys' votes were received at one polling station only; about halfway through the process a second was added. The operation took about thirty-five minutes. The fifty-three girls voted in twelve minutes. There seemed no hesitation nor difficulty on the part of any of the voters. No questions were asked, and no help was given. The children belonged to the three upper standards, and varied in age from ten to fourteen years.

For the sake of perfect clearness I here reproduce four specimens of the actual voting papers, as filled up in various ways:

The next process was the examination and counting of the votes. This was done by Mr. Bompas, Q.C., and myself, the results being recorded by Mr. White, secretary to the Proportional Representation Society.

Two methods of counting were adopted-the first that suggested by Mr. Bompas; the second, I believe, by Mr. Parker Smith. Mr. Bompas' plan is as follows:-The votes are kept in their registered order as received from the polling stations, and the first votes of each candidate are then sorted and placed in separate heaps or files. This done, and the spoiled votes rejected, the quota is calculated, and any candidate who has already more first votes than the quota is declared elected. The exact number of votes required is then deducted from the file of the successful candidates, the lowest registered numbers being first removed. The surplus votes are then distributed according to the preferences marked upon them. When these votes are exhausted the candidate lowest on the list is declared 'not elected,' and his votes are in turn distributed among any candidates still requiring them. This process of elimination is carried on until all the vacancies have been filled. Mr. Parker Smith's plan of counting differs from that just explained merely in this one particular, that no attention is paid to the registered numbers, but the votes forming the quota of an elected candidate are deducted

merely in the order in which they happen to have been collected after the papers have been mixed and sorted. It is contended that the first plan has some advantages in case of a scrutiny being necessary, though probably by a very simple method the second plan might be made to afford the same facilities. This, however, is a question of detail.

As a matter of fact we counted our votes in both ways. The first counting gave the following results. First votes-Charles the First, 32; Elizabeth, 16; Henry the Eighth, 6; Mary Queen of Scots, 1; Oliver Cromwell, 15; the Duke of Wellington, 112; and William the Conqueror, 0; making a total of 182. Two votes were rejected, the names of the candidates having been written upon them as well as the numbers. All the other papers were clearly and correctly filled up. One or two votes were plumpers. Several of the girls had given a second vote to a boys' candidate-Henry the Eighth and the Duke both receiving support in this way. One or two boys had also gone outside party lines to vote for Charles the First. One girl had numbered her vote up to seven, but nearly all the other voters had been contented with three or four transfers.

The quota was now calculated according to the rule, dividing the number of good votes by the number of seats plus one, and taking the next integer above the quotient obtained. Thus 182-4=45+1=46. Forty-six was thus the quota required to insure election. The Duke having more than the required amount was declared elected; and the forty-six votes received by him bearing the lowest registered numbers were removed. His remaining votes were then distributed according to the second preferences marked on

them.

મ

not elected,' and her Henry the Eighth having votes were in turn trans

On the second counting William the Conqueror came to the front, receiving no less than twenty-six votes. Mary Queen of Scots still having only one vote was then declared vote was transferred to Charles the First. only fifteen was the next to go, and his ferred. Queen Elizabeth now went out of the competition, and Oliver Cromwell having obtained the quota and been declared elected, the struggle lay between William the Conqueror and Charles the First. The latter, a Conservative candidate, was evidently most

1 This calculation looks complicated. It is not so in fact. The quota is simply the number which, if obtained by any single candidate, will leave a remainder which, however divided, will not admit more additional members than there are vacancies. A simple example will show this. Suppose there are 12,000 voters and three seats. Apply the rule given above: 12000÷43000+1=3001. Brown obtains 3,001 votes; Smith, Jones, and Robinson, the remaining three candidates, have 8,999 votes to divide between them. There are only vacancies for two of them, and it will be seen that only two of them can get the quota. For it is impossible to divide 8,999 into three equal parts, each part to be equal to or more than the quota given, namely 3,001. Thus 8999÷3=2999. It is true that Smith and Jones may get more than 3,001, but then Robinson must get less; and accordingly it is true to say that a candidate who gets 3,001 votes must be elected. Q. E. D.