How AV could make a difference: A thought experiment

Alternative Vote (AV) is, in my opinion, a better electoral system than First-Past-The-Post (FPTP), because it takes full account of the wishes of the whole electorate, and not just those who vote for the two leading candidates. It makes tactical voting unnecessary, allowing the individual to vote for the candidate she really likes best, without worrying about wasting her vote. I have conjured up an example in my imagination, to show how things might be different in my home constituency under AV. As far as the voter is concerned, when he enters the polling station, the procedure is very simple indeed. She numbers the candidates on the ballot paper in order of preference, like so:

Constituency: Wyre Forest

Mark Garnier	Conservative
Richard Taylor	Kidderminster Community and Health Concern (ICHC)	2
Nigel Knowles	Labour	1
Neville Farmer	Liberal Democrat	3
Michael Wrench	UK Independence Party (UKIP)
Gordon Howells	British National Party (BNP)

Note that if at any point she finds all of the remaining candidates equally abhorrent, she can leave some of the spaces blank. In practice it will be rare that fourth and fifth preferences are used in the counting of the votes anyway. Making some assumptions about voter behaviour, we can take a wild and probably inaccurate stab in the dark at what the outcome might have been in 2010 under AV. Firstly, here is the actual election result for the above constituency:

Mark Garnier	Conservative	18,793
Richard Taylor	ICHC	16,150
Nigel Knowles	Labour	7,298
Neville Farmer	Liberal Democrat	6,040
Michael Wrench	UKIP	1,498
Gordon Howells	BNP	1,120

Now, let us assume that if the election had been carried out under AV, the first preference votes would have corresponded exactly to the result under FPTP. This is extremely unlikely to be the case in reality, since voters are likely to have voted ‘tactically’ under FPTP. For instance, a UKIP supporter may have felt that his candidate was unlikely to win under FPTP, and chosen to vote for the Conservative candidate instead, in order to prevent KHHC or Labour winning the seat. As we shall see, under AV, he does not need to do so, since if UKIP are eliminated, his second preference vote will be counted.

Nevertheless, for the purpose of this demonstration, I will use these results as the first preference votes, which, when expressed as percentages, look like this:

Mark Garnier	Conservative	36.9%
Richard Taylor	ICHC	31.7%
Nigel Knowles	Labour	14.3%
Neville Farmer	Liberal Democrat	11.9%
Michael Wrench	UKIP	2.9%
Gordon Howells	BNP	2.2%

So what happens to these results under AV? Well, as we can see, no candidate has won an outright majority, so we must proceed to round two. The candidate with the lowest proportion of votes cast, the BNP’s Gordon Howells in this case, is eliminated, and the second preference votes of his supporters are added on to the remaining candidates’ scores. If we imagine that the BNP base’s second preference vote was split more or less evenly between UKIP, the Conservatives and Labour, we still do not arrive at a majority, so we proceed quickly to round three. We shall assume for simplicity’s sake that all of the UKIP voters’ next preferences will be for the Conservative candidate, and the table looks something like this:

Mark Garnier	Conservative	41%
Richard Taylor	ICHC	31.7%
Nigel Knowles	Labour	15.4%
Neville Farmer	Liberal Democrat	11.9%

Already we can see that over 5% of the population, who, under FPTP, threw away their votes on radical parties, have been given a second chance to determine the result. If we now assume that the Lib Dem second preference votes went overwhelmingly to the Tories, they would finally achieve the approval of over 50% of the constituency and win the seat. However, this seems unlikely, so let’s imagine instead that the Lib Dem support was divided more or less equally between Garnier, Taylor and Knowles:

Mark Garnier	Conservative	45%
Richard Taylor	ICHC	35.7%
Nigel Knowles	Labour	19.3%

The Conservatives in this scenario are still five percent shy of winning the endorsement of the majority in Wyre Forest, and since it is inconceivable that the hardcore of Labour voters would prefer a Tory to Dr Taylor, the latter would have romped home with 55% in the final round of counting.

Of course, this example is merely an illustrative thought-experiment. It is impossible to predict how the electorate would actually behave when given the opportunity to rank candidates in order of preference. But it does show that the outcome of an AV vote can be very different from the outcome of a FPTP vote, and, crucially, that the outcome is that which satisfies (or fails to disappoint) the greatest proportion of the constituency.

Originally published in Perspectives May 2011, No. 12.