One of the more interesting predictable aspects of the Oldham East and Saddleworth post-mortems was the discussion of what effect holding the election under the Alternative Vote system would have had. VoteNoToAV, an unofficial but enthusisatic member of the “No” twittersphere, got the ball rolling with this soon after the result was announced:
Anyone like to argue that this by-election held under FPTP wasn’t fair or democratic and that we didn’t get a clear winner? #NO2AV
There’s also this PDF distributed by the No campaign about the effects AV would have had on the by-election. As usual with much of the No campaign literature, it spends most of its time bashing the Lib Dems and never once gives any reason why we should keep First Past the Post. I was, however, intrigued by its first paragraph:
Had the 2010 election been held under AV, Liberal Democrats (sic) would have comfortably won Oldham East and Saddleworth last May…
I don’t know how they know this for certain, as they cannot know the second preference intentions of the Oldham East voters.
However, what the No camp seem to be implying is that, whilst under FPTP Phil Woolas was rewarded for his racist and illegal leaflets by winning the seat in May, under AV he would have lost.
It’s good to see the No camp sticking up for honest, fair campaining in this fashion.
The real lesson to take from Oldham East is that you cannot accurately predict who would win an AV election based on the election results of a FPTP one. It’s the same with the predictions of number-crunchers who say things like “In 1997 New Labour would have had a majority of 7million under AV”: you don’t know because people vote differently under different electoral systems. Also, the whole campaign would have been different, with candidates also campaigning for the second preference votes of voters for other parties, rather than merely relying on their core voters.
Under AV, the whole dynamic would have changed. Thus, you would see more votes for smaller parties (such as the Greens, the Pirates and, maybe, the Monster Raving Looney Party) because people could vote for a smaller party with their first preference whilst still being able to influence the outcome of the vote with their second. You would also have seen an increase in Tory first-preference voting, instead of some voting Lib Dem tactically. That’s because AV virtually ends tactical voting.
It’s true that the result would have been closer under AV, but, as John Rentoul says, “In the dull grey light of the waking world, Labour would probably have still won in Oldham under AV, just.” Not only would the result have been closer, but Labour would also have had to work very hard to ensure it retained the seat in the next election. That’s because AV makes more seats more competitive. It would have meant political parties having to knock on more doors and engage with more members of the public, and not rely on a core vote of 35-40%. And that can only be a good thing.
The upshot of all this is that when someone confidently says what the result of this by-election, or that election, would have been under AV, you should treat it with a serious pinch of salt.
Paperback Rioter 
I agree with your general point but I do think you’ve contradicted yourself here.
Yes to this: “The upshot of all this is that when someone confidently says what the result of this by-election, or that election, would have been under AV, you should treat it with a serious pinch of salt.”
Which contradicts this: “Under AV, the whole dynamic would have changed. Thus, you would see more votes for smaller parties (such as the Greens, the Pirates and, maybe, the Monster Raving Looney Party) because people could vote for a smaller party with their first preference whilst still being able to influence the outcome of the vote with their second.”
There’s also no evidence for the second point, no matter how logical we think the point is. In Australia the Green AV vote is still much les than the Green PR vote for the second chamber, as is the Green London Mayoral first pref vote compared to the PR list vote in the same election.
Thanks for your comment Jim, and the Australian example is a good one I hadn’t thought to consider. But…
You’re right that the Greens in Australia receive less votes under the AV system than when voting for the Senate under a proportional system, but it’s not that big a difference. In the elections this year the Greens won just under 1.5m votes, or 11.76% of the vote. In the Senate Elections, They won just over 1.6m votes, or 13.1% of the vote. So yes, less people vote Green under the AV system in Australia, but not significantly less.
Compare that to the Greens in the UK. In the European elections last year, held under a Proportional System, the Greens gained 8.7% of the vote with 1.2m votes. In May, the Greens only gained 285,000 votes, which isn’t even 1% of the vote. That’s a massive difference, and probably represents the fact that Green voters are more likely to vote tactically under FPTP. I reckon that you would see more Green voters voting Green under AV than under the present system.
My own analysis concurs with the No Campaign’s view that the Liberal Democrats would have won Oldham East and Saddleworth comfortably at the last General Election. I presume they are basing their analysis on some research done by the LSE conducted at the time of the last general election on 2nd preferences in a theoretical AV election. I use it when running my own analysis.
http://blogs.lse.ac.uk/politicsandpolicy/?p=3915.
I think you’re right to say that it’s very difficult to predict how people would cast 2nd preferences, or even 1st preferences in an AV election based on First Past the Post results (unless you’ve asked them.)
What I get out of the number crunching of the two Oldham East and Saddleworth elections is although the Liberal Democrats go on to win the seat by what looks a comfortable margin in the final round, in the penultimate round, between Labour, Liberal Democrat and Conservative candidates there are only a thousand or so votes between second and third place. I’m not certain but I think the Conservative party if they put the same effort they put into a marginal seat could persuade an additional few thousand voters to turn out for them.
I think it supports the Yes Campaign’s view that AV would create more marginal seats and more interest in voting thus increasing voter power and turnout.
My own write up is here.
Thanks for commenting and your analysis!
I think there’s nothing wrong with your number crunching, it’s just that, as you say, it’s difficult to know how much that would help predict who would voted how in an AV election. As I argue in the post, I’m not sure it’s terribly helpful.
I do agree with you though in that it shows AV would create more marginal seats, voter interest etc
Pingback: An A-Z of rubbish arguments from No2AV « Paperback Rioter
Pingback: An A-Z of rubbish arguments from No2AV | Yes to Fairer Votes – Birmingham
Pingback: An A-Z of rubbish arguments from No2AV (2nd edition) | Yes to Fairer Votes – Birmingham