The Alternative Vote (AV), it is often argued, is a voting system that provides more data on the preferences of the voter as it requires the available candidates to be ranked in order of choice. However, though there is certainly more noise, I’m not convinced this equates to more information.
For example, I have four candidates, A, B, C and D. I like candidate A and I loathe candidate D; how should I vote? Bearing in mind that the form of AV proposed for the UK doesn’t require a full list of preferences, the answer (it would appear) is ABC… or ACB. It’s the “or” that interests me. Since there is no ‘equal ranking’ – I can’t rank B and C as joint second – I am forced to make an arbitrary choice solely for the purpose of voting against D; most likely it will be a donkey vote.
If candidate C fails to obtain a majority in the first round and candidate A is eliminated through having the fewest number of votes, my randomly chosen 2nd preference is now counted for those who remain. Whilst I achieve my aim of voting against D, is democracy best served if candidate C, with the most (1st choice) votes, loses out to candidate B through having an inferior position in the alphabet?
Further, it may seem counter-intuitive but a vote of ABC doesn’t necessarily mean a preference of A over B, and B over C; it might do, but the only inference we can make with certainty is that it is a vote against D. Somewhat perversely, in comparison to this “anyone but them” portrayal of AV, our current first-past-the-post (FPTP) voting system in lacking this ability could be seen as requiring the voter to concentrate on who they vote for, rather than who they vote against. FPTP at minimum requires a “least worst” decision from the voter whereas AV, in this example at least, abrogates the responsibility.