Henry Richmond Droop On Methods of Electing Representatives 1881
why we vote historically
only two candidates ran for a seat.
but by 1881 Henry Droop saw that voting was no longer for the best (most honest/knowledgeable) versus worst (less honest/knowledgeable) [a binary choice] but had changed, although still being between two main parties --
thus single-seat contests or in district with just two seats may result in disproportional result.
for that reason PR is needed.
(from On methods of electing representatives, p. 143-)
the present day, at any rate in electing representatives for parliamentary or municipal assemblies, electors do not seek exclusively or mainly to select the most honest, intelligent, and competent of the candidates. On the contrary, with but few exceptions, the electors pay very little attention to the personal qualifications of the candidates, and look only at the views they hold and the measures they promise to support. What they aim at securing is that their views and their measures should prevail in and be carried out by the assembly. I do not blame the electors for thus looking to principles and measures rather than to personal. qualifications; but it makes a great difference in the working of majority voting.
Whenever the majority of the electors in a constituenicy have discovered that they are agreed in supporting certain views and measures, they will naturally use the power which, under majority voting, they possess, of only electing representatives who hold the same views and will support the same measures. An election thus naturally becomes a contest between two parties, each of them trying to secure the votes of the majority of the electors for its own views and measures, and for the representatives who will support them. Smaller sections of the constituency, knowing that they cannot elect any representatives of their own selection, will annex themselves to one or the other of the two principal parties.
...
equally between the two parties, the proportions depending however not upon the comparative strength of the two parties in the constituencies, but on the number of oonstituencies in which each party hqppens to have the majority, and the number of representatives returned by these constituencies. This will usually exaggerate the difference between the two parties, and give the stronger party a much larger majority in the assembly than it has in the constituencies; but sometimes on the contrary it assigns the majority in the assembly to the party, which is really in a minority in the constituencies.
To make my meaning clearer, I will assume that each constituency has a number of representatives in exact proportion to the number of electors it comprises, an assumption which will be very nearly correct in countries where representation is in proportion to population, e.g., in the United States and in France, and which is being more nearly realized in the United Kingdom by every successive Reform Bill. I will further assume that there are I,990,000 electors who have to elect 199 representatives, or one representative for each 10,000 electors.
Suppose now that 100 of these representatives are elected by the A party by narrow majorities of 5,000 to 4,900 in constituencies returning only one member, of 1,200 to 9,800 in constituencies returning two members, and of numbers in the same proportion of 51 to 49 for constituencies returning three or more members, while the other 99 members are elected by the B party, by unanimous constituencies of overall 110,000 votes. Then the A party that has elected 100 representatives, and therefore has a majority in the assembly, will have only received the votes of 510,000 electors, while the B party, which has only 99 representatives, will have received the votes of 490,000 + 990,000 = 1,480,000 electors, or more than 74 percent, i.e., very nearly three-fourths of the 1,990,000 electors.
This is, of course, an extreme and improbable case, imagined to illustrate what majority voting may possibly do in the way of putting the minority in the place of the majority, but many very much more probable distributions of votes might be suggested, which would produce substantially the same result, i.e., that the majority of representatives would correspond to the minority among the electors.
Moreover, such cases are known to have repeatedly occur. (p. 144 4/62 online)
...
p. 24/62 talks of numerous parties under PR
limited voting 29/62
cumulative 30/62
preferential voting 39/62 means STV DM-3 to 16
STV practicable 44/62
time needed for vote counting 45/62
=====
it is said that
Henry Richmond Droop proposed applying a form of IRV to single-winner byelections briefly in his book On Electing Representatives (1881), p. 194
But I am not sure that he actually did say that as much as fans of Instant-Runoff Voting seem to think.
p. 194:
by-elections
"If this mode of filling up vacancies [Baily's plan to use votes from previous election, with use of unrepresented voters - details are unclear to me] were adopted, the candidates under the limited transfer by lists method, or with preferential voting, the electors would usually add some additional names to guard against the possibility of their lists being found exhausted when a vacancy occurred."
goes on to say similar is possible in IRV if each party puts forward multiple candidates in main election as back-up for use in case of by-elections.
so in a way he is saying vote trnsfers can be used to elect single rep. but he was not directly saying IRV is fine IMO
============================
Droop was author of
1871 PR as applied to the election of local governing bodies
1881 On Methods of Electing Representatives
Comments