There are many systems of PR (Proportional Representation); many suffer from flaws.  There is insufficient space in the margin to do more than outline a few flaws. In my opinion, the best PR is the d’Hondt formula (Thomas Jefferson’s method) which is mathematically sound.

The worst alternative to FPTP is the “Alternative Vote” (AV) system; AV is no more PR than is FPTP.

* AV hugely encourages tactical voting.

* AV encourages appeals to fringe groups, solely so as to pick up their 2nd (or lower) choice alternative votes.

* AV will in effect give many of those who vote for fringe groups a multiplicity of votes in successive “rounds” within the same election.

* AV will encourage parties perceived as being policy-wise “intermediate” to indulge in several legal but immoral practices to increase their chance of being the first (no sic) to reach 50%. I won’t elaborate or put ideas into the heads of the unprincipled.

* AV could have put Neil Kinnock in Number 10

FPTP is the worst form of voting there is – except, that is, for most everything else tried so far.  To remedy one obvious flaw of the UK’s FPTP: Redraw constituency boundaries to remove the built-in enormous bias to Labour, which has arisen through a combination of mass migration, sloth and socialist chicanery with LibDem collusion.

The status quo requires the Conservatives to obtain 10% more votes nationwide than Labour to get the same number of seats in Parliament (e.g. if Labour gets 35%, the Conservatives need to get at least 39% to be reasonably sure of getting more seats in the HoC). In some areas, the imbalance is up to three times higher.

Let Papua New Guinea retain the honour of being AV’s flagship. The rest of the world should vote “No!” to AV.


Consider a hypothetical constituency with 4 candidates, A, B, C and D.  We’ll look at two scenarios, CASES ONE and TWO.


Say the 100 votes cast (in order to “pass 50%”, at least 51 votes needed) were as follows:

A 37 (#2 irrelevant)
B 31 (A)
C 26 (B)
D 6 (C)

The bracketed letter shows the voter’s second choice. For simplicity, I assume that all the voters who chose a particular candidate made identical second choices (e.g., all 26 voters for C chose B as their #2). I could construct a more complex scenario with split (and absent) second choices but it would be hard to follow.  The flaw would still exist.

Under FPTP, A wins with a majority of 37-31=6.

Under AV, D is eliminated in the first round, and C gets the benefit of D’s 6 second choice votes, giving it 26+6=32<51. So now, B is eliminated (because 31<32, C’s amended vote total), and its second preference votes are transferred to A. As 37+31=68>50, A wins, like with FPTP.


Under FPTP: A wins
Under AV: A wins


A had campaigned harder, and, without any other voting consequences, convinced half the D voters to vote for himself instead (relegating D to their second choices), and the other half of the D voters to choose A instead of C as their second choice.

Under AV, A could be severely punished for getting more first and second choice votes!

Here is what the voting would have been:

A 37+3=40 (#2 irrelevant)
B 31 (A)
C 26 (B)
D 6-3=3 (A)

So, in TWO, A has 3 more first and 3 more second choice votes than in ONE. No other changes.

Sounds good for A?

Logically, A still wins under FPTP, with an increased majority of 9 (up from 6), a reward for his campaigning.

Under AV, as before, bottom-scorer D is eliminated, leaving A with 40+3=43<51, insufficient for an immediate win. In the next round, C is eliminated (26<31). So B picks up C’s second choice votes, giving B 31+26=57>50.

So B wins under AV…


In TWO, A got 3 more first and 3 more second choice votes than in ONE (at the cost of D and C respectively), and:
Under FPTP: A wins


Under AV: B wins!  Ridiculous.


In ONE, A won under AV.
In TWO, A worked harder and as a result got 3 more first and 3 more second choice votes than in ONE. There were no other changes and no shift of voting (first or second preferences) towards B.
But under AV, as a result of the extra first and second choices for A, B defeated A!


Let’s say the change from ONE to TWO happened not because of A’s campaigning, but because D supporters wanted to scupper A’s chances.  So, they changed their first or second choices (from D or C) to A, thereby (under AV) causing A to be eliminated… even crazier!
I acknowledge this relies on a high accuracy of knowledge of voting intentions.

Should, therefore, the question be, are all those who are lousy at maths AV-supporters, or are all AV-supporters lousy at maths?

Print Friendly, PDF & Email