Rev. 23 Jan 26
All preferential options (PV, STV, AMS) are currently constrained by what can be done within the constraints of an essentially manual count. But ==
We believe that PV, our proposed AV variant, is infinitely better than FPTP, and the very best replacement for it that can be offered
(a) in the short term
(b) while we are restrained to a manual count.
But it is clearly somewhat less than optimal,[1] and would merit being upgraded as soon as is technically possible. Two particular examples of its defects:
But this applies to all preferential systems: not just simple PV but also STV (multi-seat PV) and AMS (simple PV plus proportional top-up)
Head-to-head.[2] The established rules say that the candidate with the fewest top choices will be the first to be eliminated. However, they can be in fact be explicitly preferred to every one of the other candidates taken individually[3], which is clearly extremely unjust. But it’s the best that can be offered witha purely manual system.
Moreover, the candidate with the most top choices can in fact be explicitly less preferred than each of the other candidates taken individually, but will nevertheless be protected from immediate elimination.
So for the long term, when we are longer constrained by the capabilities of a manual count, we should consider an extension of the AV/PV algorithm. E.g.:
(1) Elect any candidate who has >50% of top choices, as in AV.
(2) Failing that, elect any candidate who is explicitly preferred to every one of the other candidates taken individually.[4] This involves the tedious assessment of every possible pairwise head-to-head contest, an excessive task for people to undertake, but easy meat for a computer.
(3) Failing that, eliminate any candidate to whom every one of the other candidates taken individually is explicitly preferred, and repeat from (1).
(4) Failing that, eliminate the weakest candidate and repeat from (1).
Elimination criterion. But which is the weakest candidate? Currently judged on which has fewest top choices. And since already counted in unsuccesful search for winner, no effort to then eliminate the one with fewest. Nor is any more serious option feasible while we have a purely manual count.
on some other criterion (but not necessarily AV’s crude “fewest top choices”),
highest average ranking, possibly weighted; <see earlier list>
But, having done a head-to-head assessment, an obvious choice is lowest number of wins, with lowest number of top choices as a tiebreaker.
NOTA. Vote for “None Of The Above”. Ranked like any other choice. Might have enough support to win a seat. Could be just advisory: just note the voters’ feelings and go on to their next choices, if any. But possibly binding, in which case the count terminates, and the election is rerun, with all unelected candidates barred from standing.
Equal preferences. Why in ranked choice voting should not two or more candidates be given equal preferences? Realistic, and convenient for the voter. Requires a small change to the evaluation rules, but unfortunately one which makes a manual count effecively impossible.
<MOVE TO STV>
Party ranking. A problem with multi-seat constituencies is a corresponding increase in the number of candidates. Not all voters will wish to give explicit preference to 30 candidates. Many would prefer to rank 6 parties instead, leaving it to each party to rank its candidates. So perhaps allow the voter to first rank individuals, if they wish, and then to rank parties. Requires no change to the evaluation rules, but does require a small change to the procedure, though not one which makes a manual count infeasible.
<SEPARATE PAGE?>
Non-geographic constituencies. <proportionality other than party proportionality> <e.g. ethnicity, religion, but possibly also age, sex, and even RSPB membership> <but see also HoL reform>
——————————————————————–
- See also voting single seat.pdf voting single seat.odt (wjw/20May2019) ↩︎
- We choose not to mention Condorcet. The word is not that well known, and attempts to define it seem to inevitably increase confusion. ↩︎
- Example: Vote is 40% ABCD, 10% BCDA, 20% CBDA, 30% DBCA. Head-to-head: AB, AC, AD all 40-60; BC 80-20, BD 70-30; CD 70-30. So head-to-head wins are B 3-0, C 2-1, D 1-2, A 0-3. ↩︎
- That would of course include the >50% winner if they had not already been selected by a more straightforward calculation. ↩︎