What is AV?
More details and data
Did you notice that in the count only C's second preferences were counted? The second choices of A's voters and B's voters were not. Now let's examine the fictitious data from Voting Central more closely. We know how the 2nd preferences of C voters were distributed - 6,000 votes to B and 600 to A. Now let's look at the 2nd preferences of the A voters:
- 6,000 votes to B; and 12,000 to C.
This adds up to 18,000 - the number of voters who put A first. Here's the 2nd preferences of the B voters:
- All 12,750 votes to A.
Now we get the final votes tally (remember 6,000 2nd preferences of C were non-transferable):
| Candidate | 1st Preferences | 2nd Preferences |
|---|---|---|
| A | 18,000 | 13,350 |
| B | 12,750 | 12,000 |
| C | 12,600 | 12,000 |
And now you can see why there may be many unhappy voters in Voting Central. Candidate A received the most 1st and 2nd preferences and still lost to candidate B. As Lakeman (author of the classic book on electoral systems Voting In Democracies) pointed out: "Though the Alternative Vote is commonest method of securing election by a clear majority [of those voting in the final round], it is not the method likely to give the most general satisfaction."
There are two points from this. First, if full results are published, it becomes more obvious with AV when the best overall candidate has lost and this may increase dissatisfaction rather than improving it. Dr John Cox, who submitted substantial evidence to the Richard Commission on using STV for the Welsh Assembly, says: "Far from bestowing legitimacy on those elected, AV creates even more post-election angst than does first-past-the-post - the transparency of AV leads to endless post-election recriminations if the leading candidate in the first round is defeated by the transference of votes from the least popular candidates." Regardless of the electoral logic, it is likely that AV will run counter to the psychology of voters familiar with FPTP.
Second, tactical voting may occur in Voting Central at the next election. Let's see how this might happen. If candidate B had been eliminated - and not C - the priceless 2nd preferences of B voters would have gone to candidate A, who would have romped home with a massive 30,000-vote victory. How many votes need to change hands to make this happen? Just 151 votes out of 43,350 - that's all candidate C needs to secure second place on the first preference count, thus eliminating candidate B.
But where are these additional votes for C going to come from? Party A has a surplus of first preferences and may attempt to get a proportion of its voters to put C down as first choice and thus win the seat! It's the old game of tactical voting with new rules. Where it occurs it will be more complex and more unpredictable than tactical voting today. The key difference being that political parties may start to orchestrate voting strategies more with their activists to secure the result they seek. We cover this issue again in Will AV end tactical voting?
We do not think the exact kind of scenario in Voting Central would be a common one in the UK but with 600 or more constituencies something like this may still happen in dozens of seats. If you don't mind this happening in your constituency then you probably won't see it as an issue. Others may feel very differently.
