AV: What impact will it have?

Posted on April 21, 2011 at 10:01 am,

On Tuesday, I explained what AV is and isn’t, and yesterday I looked at some of the myths and lies about AV. Today, I want to look at the impact AV would have on election results. I’ll start with some general points, before looking at how it will affect the individual parties (though I won’t be including the SNP, Plaid Cymru, or the Northern Irish parties because I don’t have enough local knowledge of those areas)

General Impacts

The first thing to point out is that AV discriminates against parties who are hated outside of their base. In 1997, AV would have massively increased the Labour landslide, because the only voters who didn’t hate the Tories after 17 years of Thatcherism were the ones who were still voting Conservative. Of course, in a year like 2010 – when Labour and the Tories were both deeply unpopular – this effect may well have been balanced out.

The second thing is that under AV there would be no need for parties to deliberately stand down in the occasional seat, as they could simply ask their voters to use their second and third preferences in a particular way instead.

Under the existing system, this sometimes happens to ensure that a hated candidate loses (in 1997, Labour and the Liberal Democrats stood down in Tatton so that independent Martin Bell could oust the scandal-ridden MP Neil Hamilton). Small parties might also do this to ensure that a similar party or candidate wins. So in 2010 the Greens didn’t stand a candidate in Birmingham Hall Green in order to help Respect’s Salma Yaqoob – who has worked closely with the Birmingham Greens. In the end she was 4000 votes short, but if the Greens had stood, their votes would likely have come from her rather than the winning Labour party.

The third thing to note is that it may well change the way election campaigns are run. Because there will be no benefit in encouraging tactical voting (as happens in a lot of seats under First Past the Post), election campaigns under AV will be about policy, rather than about whether a seat is a two horse race.

Impact on Political Parties

Before examining the parties (which I’m ordering from right-wing through to left-wing), I want to do a little bit of political geekery, which is basically the foundation of my arguments. If you just want the conclusions, skip down to the BNP heading.

In order to win a seat under first past the post, you just need to get more votes than anybody else. In Norwich South, this meant that the Lib Dems were able to win with just 29% of the vote. Under AV, there are a couple of ways to win a seat.

The first method is to win on first preference votes. As you’d expect, this means getting half the first preference votes. However, if you get 45% or more on first preferences, and the second party is some way behind you, then the chances of them overtaking you with second or third preference votes taken from other parties are pretty slim.

The second method is to win on transfer votes. To demonstrate this, I’ll use an example of the rare case when a third party can win.

Counting first preferences in our imaginary constituency we have:

Conservative 10,053 (30%)
Labour 9,345 (28%)
Green 7,789 (23%)
Other parties 6,389 (19%)
Turnout: 33,576

None of the other parties get enough transfer votes to overtake the Greens, so to make the example simple, we’ll skip ahead to the final three:

Conservative 11,488 (34%)
Green 11,167 (33%)
Labour 10,921 (33%)

Whilst the Greens have fewer first preferences, they are more popular than both Labour and the Conservatives amongst voters for parties who have been eliminated. Because they were fairly close behind Labour, this has taken them into second place. In order to win, they need to have 322 more transfers from Labour voters than the Tories do. After these preferences are reallocated, the result looks like this:

Green 20,167 (60%)
Conservative 13,409 (40%)

Because the vast majority of Labour voters hate the Tories, this gives the Greens an extremely comfortable win on the final round, and the majority of voters in our imaginary constituency have an MP they are reasonably happy with.

So, now we know how you win a seat under AV, how will the individual parties be affected?


The BNP are the clear losers from AV. The number of BNP supporters who vote tactically is almost certainly close to zero and, as their best result at the General Election was a distant third place with only 14.8% of the vote, they have no chance of winning on first preferences. It’s difficult to imagine them winning more than the tiniest handful of transfer votes either. AV will keep them out much more effectively than first past the post would.


Are also unlikely to be immediate winners. Whilst they will almost certainly win plenty of first preference votes from people who currently vote Conservative, their support is too widely spread for them to be in a strong third place anywhere. If you don’t believe me, look at their 2010 general election results – they came third in only four seats, and in all four seats the winner got 45% or more of the vote. The only seat where they got more than 10% was Buckingham, a seat in which none of the major parties was on the ballot paper. If UKIP were to develop a good targeting strategy, then they might well be able to break through into Westminster more easily than under first past the post, but at the moment that seems a long way off.

The Conservatives

Are likely to lose out. The good news for them is that, whilst some current Conservative voters will probably give their first preference vote to UKIP under AV, they will get those votes back on transfers. In marginal seats where UKIP get a good vote share, the Tories may well gain seats. The bad news for them is that there are a lot of seats where (presumably) left-wing voters form a majority, but where their votes have been split between Labour, the Lib Dems, Greens, and the various far-left parties. These are seats that the Tories are more likely to lose under AV,. Having said that, it’s worth noting that Tory safe seats tend to be safer than Labour ones, and they have more seats that they would win on first preferences.

The Liberal Democrats

Unlike a lot of commentators, I’m really unsure what AV will do to the Lib Dem vote. The problem with predicting it is that they’ve won an awful lot of their current seats by encouraging tactical voting – which makes it difficult to know just how many voters in winnable seats would actually put them as their first preference. They will probably hold on in seats like Colchester, where their MP has a lot of local popularity, but there are likely to be a lot of seats where they slip from their current second place to being in third place on first preferences. On the other hand, they are likely to gain seats from the Conservative in constituencies where they are able to keep Labour in a reasonably distant third. And, regardless of whether the referendum is won or lost, they’re going to lose most or all of the seats where it’s currently a straight contest between them and Labour.


Like the Tories and Lib Dems, Labour are likely to lose first preferences to smaller parties (most likely to the Greens and various small left-wing parties), but get them back on transfers. The chances are that the overall effect on Labour will be fairly neutral – although their seats are less safe, on average, than Tory seats, they’re only likely to lose from AV are in areas where Lib Dem voters are a large chunk of the electorate and where they prefer the Tories to Labour.

The Greens

We’re likely to do relatively well from AV. The evidence suggests that there are a lot of greens who vote tactically, so our first preferences should be a lot higher than our current general election vote. And with the current hatred of the Lib Dems, there are probably quite a few seats where we could ease into a comfortable third, which leaves serious potential for longer term gains. However, it’s less clear about our short-term prospects which, realistically, concern just two seats (unless there’s a lot of tactical voting going on in other areas where we are strong).

In Brighton Pavilion we won with 31.3%, and if first preferences were the same as under FPTP, the result would most likely have depended on redistributing the Conservative vote between us and Labour. Of course, next time Caroline Lucas will probably have an incumbent advantage, and we might get plenty of first preferences from Labour voters, but it’s far from certain that we would keep the seat under AV.

Norwich South is a lot more interesting. It’s quite possible that a lot of Lib Dem voters (who won the seat this time with 29% of the vote) would have voted first preference Green. We have consistently won more votes in that seat in local elections over the last few years, and had the highest vote in Norwich in the last Euro election, so it’s quite plausible that we would have been first, second, or third on first preferences under AV. With the inevitable collapse in the Lib Dem vote, it’s quite possible that we could take the seat under either AV or FPTP next time, and quite likely that AV would make it that little bit easier.

Far-left Parties

I’ve ignored most of the fairly small parties in this discussion because, whilst they will see more first preference votes, those votes will almost certainly be quickly redistributed to larger parties. For most of the socialist parties, the most they can realistically hope for will be to take enough first preferences from Labour that they keep a few deposits (for example, the TUSC’s best performance was Dave Nellist in Coventry North East, where he lost his deposit). However, there is one seat in which a leftist party came just 4000 votes behind Labour – Birmingham Hall Green

Whilst I don’t know the constituency very well, it’s not impossible to imagine that under AV Salma Yaqoob might have got the couple of thousand extra votes to beat Labour on first preferences. In that case, the Tory and UKIP vote would probably gone almost entirely to the Lib Dems (or been discarded), which would have put Labour down into third place. And then all RESPECT would have had to do to win would be to gain substantially more transfers from Labour than the Lib Dems did.

In Summary

The main impact on Parliament will be to redistribute support amongst the big three parties. However, AV does leave open a greater possibility for some small parties to break through in areas where they have a lot of local support.

One Trackback

  1. By Last thoughts on AV and Nick Griffin | on June 11, 2012 at 12:16 am

    […] notes from other places: Green Christian on how AV might effect politics, Stumbling and Mumbling on where AV could lead, Deborah Orr on how […]

Post a Comment

Your email is never shared. Required fields are marked *