Tuesday, July 20, 2010

Ashcroft Research Shows Tories Could Do Better Under AV - And LibDems Worse

I think after reading THIS, our new friends in the Liberal Democrats might be reconsidering their blind support for the Alternative Vote system. Michael Ashcroft has commissioned some research which shows that a dramatic decline in support for the LibDems (as evidenced in recent opinion polls) could lead to a large rise in the number of seats won by the Conservatives.

There has been a widespread assumption that the Conservatives have nothing to gain from electoral reform, and the work that has been done so far – such as the YouGov poll for the Spectator earlier this month – has indeed suggested that the Tories would be the biggest net losers when comparing A.V. with First Past The Post (FPTP).

As ever, though, national polls can only tell us so much – it would be in the marginal seats that A.V. would make a decisive difference. Would voters in these seats behave differently under the two systems? And would the effect be different depending which parties were in contention? A newly commissioned 6,000-sample poll helps to shed some light on the debate.

1,500 people were interviewed in each of four clusters of target seats: the 50 most marginal Labour-held seats with the Conservatives in second place; the 50 most marginal Conservative-held seats with Labour in second place; the 25 most marginal Liberal Democrat-held seats where the Conservatives are second; and the 25 most marginal Conservative seats where the Lib Dems are second.

The findings are striking. Under a FPTP election, Labour would gain 28 of the seats in which it is currently in second place to the Conservatives. Although the Conservative vote in these seats was only fractionally down since the general election, a 4-point drop in the Liberal Democrat share exclusively benefited Labour. In the 25 Liberal Democrat-held seats, though, the collapse in the Lib Dem vote was much more dramatic: the party was down 15 points from its general election position. In this scenario, the Conservatives would win all of these seats plus a further five – more than compensating for its losses on the Labour battleground.

Under A.V. the swings were less dramatic but the effect was no less interesting. In Con-Lab marginals, while Labour voters were much more likely to give their second preference to the Lib Dems than to the Conservatives, Liberal Democrats were more likely to give their second preferences to the Tories than Labour – albeit by a smaller margin. Labour would gain 16 Conservative seats under this system. In the Liberal Democrat-Conservative battleground, although the Lib Dems were the significant net beneficiaries of second and third preferences, they were so far behind on first preferences that the effect of the transfers was to narrow the Tory lead, not eliminate it. In this scenario the Conservatives would gain 19 of Lib Dem seats in which they are in second place – leaving them three seats up on the deal, compared to only two under FPTP. This gives the rather counterintuitive result that in an election now, the Conservatives could do as well, and possibly better, under A.V. than under FPTP.

There will be further complicating factors, of course. We don’t yet know for sure the impact of the reduction in the number of constituencies – the Conservatives should be the net beneficiaries, but the parties will fight over every boundary line; until the new map is drawn we won’t know how many seats will change hands for a given swing. It is also possible that, under A.V., swings could start to vary between seats even more than was the case on 6 May, if voters start to set even more store by the merits of individual candidates when allocating their preferences. And fairly small shifts in headline voting intention (particularly an increase in support for the Liberal Democrats) could change the result under A.V. significantly, given their advantage in second and third preferences.

We took the opportunity to gauge the opinion of the marginals more widely. Not surprisingly, the coalition’s performance was more highly rated in the Conservative-Liberal Democrat battleground than where Labour are strong contenders, though a clear majority thought the new government was doing well overall – indeed nearly a quarter admitted to thinking it was doing better than they expected. However, only one in ten of those who did not vote Conservative or Liberal Democrat on 6 May said they were now more likely to vote for either party because of the way they have conducted themselves since the election.

But more alarm bells should ring for the LibDems over another finding...

Voters in these seats were evenly divided as to whether the coalition represents “the beginning of a new type of politics”, though small majorities thought so on the Conservative-Liberal Democrat battleground. Just under half thought the way David Cameron and the Conservatives had behaved since the election “shows that the Conservative Party really has changed for the better”.

Significantly, even in the Liberal Democrat-held seats, less than a quarter of voters thought the Lib Dems were having a significant impact on the coalition government’s agenda. Most thought the government’s agenda is very similar or the same as what they would see if the Conservatives were governing alone.

To read the full report (it's 8 pages), click HERE.


Nich Starling said...

Iain, fairness and fair votes is the reason for the Lib Dem support of PR, not self interest.

Iain Dale said...


AV is not PR though, is it?

Duncan Stott said...

Errr, this research doesn't show the Lib Dems would do worse.

The findings are that the Tories would gain 30 seats from the Lib Dems under FPTP, compared to 19 seats under AV.

So the Lib Dems lose less to the Tories under 11 less under AV.

What this research doesn't tell us is what would happen in Lab/LD marginals.

Not that this matters. The electoral system shouldn't be chosen based on which party does best out of it.

Anonymous said...

may I comment a piece over on the Guardian


Polling found that labour people -- " were all slightly strange people ... strange personally, I mean. They were people who really did want to spend their evenings sitting in church halls or community centres agonising over quite arcane points of detail.
And they weren't just doing it that night, but every night – "

Seems to me Brown was their perfect leader.

Anonymous said...

Get a life, Iain! It's only two and a half months since the election, and goodness knows what the political landscape will be looking like in five years' time, or even if the coalition will still be in existence. And we may not have AV, in any case, if the Lib Dems lose the referendum.

All these findings are hardly earth-shattering, and the main thing it demonstrates is that whether you have FPTP or AV makes little difference to the disproportionality of the result.

Ryan said...

shocking news that as the election season has ended the polls show a drop in the LD vote share. then applying this against an AV election still shows a decrease. next you'll be running a report showing how LD would lose seats tomorrow under STV, dHont or flipping a coin systems.

Mark Thompson said...

Iain. AV is not PR, no but on a number of measures it is fairer certainly within individual seats.

And as for your ROFL comment to Nich's point, support for PR is one of the main reasons I joined the Lib Dems. I am a Lib Dem to a large extent because I support electoral reform. For me it is not the other way round. And I know a number of other Lib Dems who are in a similar position to this.

Oh, and extrapolating from polls two months after an election when the parliament will likely run for 5 years is a risky business, especially with something as potentially game changing as AV.

Morus said...

Iain - I've just posted this on ConHome. I can only assume the Ashcroft polling is either woefully misguided, or simply designed to scare LibDems on the fence from voting 'No' on the referendum.

"Even beyond my usual gripe (I don't think you can poll 'marginals' because they are not a discrete class which share any common characteristic allowing you to bundle them into a proper sample) this polling is silly.

The whole point of AV is that 'marginals' (as they are under FPTP) cease to be as important.

In FPTP there are maybe 125 marginals, highlighted here, that matter. But look at the 2010 General Election results.

Of the 632 seats in Great Britain (exc NI), there were only 212 where one party won more than 50% of the vote (75 Lab, 125 Tory, 12 LD).

Of the 212 where a party won over 50%, only 84 seats were won with over 55% of the vote.

So under AV, a small swing could put all but 84 seats (out of 632) in play - that's 548 out of 632 that will be tightly contested.

Polling the 125 seats considered most marginal under FPTP gives you no true indication of how AV would affect the next election."

James Burr said...

Yes, I've often thought this, too.

It seems to me the Tories would gain a lot of votes from UKIPers (1,500,000 votes in the last election) who would likely place them as a second choice, as well as from a fair few people who vote Lib Dem as a first choice and would have the Tories second.

Whereas, how would Labour gain? Much of Labour's support is in Wales and Scotland so it would seem that they may gain a few votes from people who have Nationalist/Socialist parties as their first choice but that wouldn't make much difference on a national level. Similarly, the Labour vote is likely to be diluted by numerous tiny parties (like the Real Labour Party, the Greens or the Socialist Workers or somesuch) which again would have no impact on the national scene.

And as far as the Lib Dems are concerned, now they have aligned themselves with the Tories, I can't see the Lib Dems gaining many second-choice votes from Labour voters (as well as losing a fair few "first choice" votes from Lib Dems who feel betrayed), so they will probably not benefit much at all from AV.

AV is a recipe for not having a "decisive Government", but I fail to see how the Tories would not benefit from it.

Dual Citizen said...

Actually Iain it doesn't show that the LibDems could do worse. The only apples-to-apples comparison is the one between AV and FPTP in Lord Ashcroft's poll, not with May 6.

On that poll the seat differences between AV and FPTP are

Conservative +1
Labour -14
LibDem +13

What this shows is that under AV it's the first and third parties that benefit, and the second party that loses out.

This is consistent with earlier analysis of 1997 (http://bit.ly/9Oqa8P) & 2005 (behind the Times paywall), which show Labour benefiting hugely from AV at the expense of the Tories.

And remember that those analyses, and indeed this poll were all based on constituency boundaries heavily biased towards Labour. Equalize the constituencies, and as long as the Conservatives stay ahead in the polls they would likely get a bigger majority (or actually a majority).

Dual Citizen said...

I saw your comment on ConHome too. I take your point and yes there's probably room for some extra analysis in the seats you mention. But just because more seats "come into play" doesn't mean that they are marginal. If a Tory MP with 53% on May 6 were to drop to 48% of 1st preferences next time, it would take around 90% of the others to put the Tory last in preferences for him/her to lose.

That said, looking at a couple of election results, I'm pretty sure Dr. Richard Taylor would have held Wyre Forest (instead of the Tories), and Peter Robinson would have not lost to the Alliance. Food for thought.

Will 883 said...

To back up the point made by Norfolk Blogger and others, Lib Dems support electoral reform because it will make for a fairer, more democratic country, not out of self interest.

To question this is very cynical and suggests politics should all be about self gain.

I actually think you would struggle to find many party members whose primary reason for backing electoral reform is because it will benefit the Lib Dems.

And no, AV is not PR, so well done on pointing that one out... I actually find it quite 'ROFL' that Tories argue against AV by saying it is not PR - whilst at the same time being against PR itself anyway! A move in the right direction is better than no move at all.

Anonymous said...

If the govt agenda is the same as if the tories were 'alone' - then why do we hear stories about the right being upset?

Given its the tories and LDs in a coalition its not unreasonable to expect them to put each other as second preference. LDs are a minority party thats a fact. The fact that they are working with the conservatives is bad news for labour, no matter what the outcome of an AV vote.

But the outcome for both parties in the next election really comes down to how they perform over the next five years. A recovering economy good schools and troops out of Afghanistan will count for more than predictive polls.

Libertarian said...

@Norfolk Blogger

If it was fairness and democracy you were after you would be campaigning for directly elected Executive

Libertarian said...

@ Mark Reckons

So why do you support a system that has party appointed candidates who are chosen on the back of the parties policies. Then the Queen decides who will run the country.

Personally I couldn't care less what system of voting you use, none of it is democratic, representative or fair.

Direct elections of the Executive are the only way to bring about fair democracy