andrewducker

I can't wait until danieldwilliam buys UKIP.

I don’t need to buy them, they are giving themselves away to me in bits and pieces.

Yep, perhaps my top most favourite realistic result.

The Lib Dems show they can hold a Tory / Lib Dem marginal in challenging circumstances.

UKIP do well enough to frighten the Tory horses.

I was really, really enjoying Grant Shapps comment that the Tories only came third by a thousand votes. “Oh yes, “ I thought to myself, “and if you come second is a load of seats by only a thousands votes will you have won the general election?”

I did a small amount of maths about what would have happened under AV below. I suspect that the UKIP would have taken it, but it would have been incredibly close either way.

Some thoughts:

1. It's ironic that the LibDems probably won because of FPTP.
2. Nigel Farage must be kicking himself for not standing personally (not sure if it would have made a difference or not though).
3. How's that whole gay marriage policy working out for you, Dave?
4. Did Labour deliberately run a joke candidate to increase the LibDem vote?
5. Voters on the whole don't know or don't care about the Lord Rennard affair.
6. Voters on the whole don't think the entire party is like Chris Huhne.
7. Lots of Conservatives saying "Vote UKIP, get LibDem". Since UKIP came second, they should be able to reply "Vote Conservative, get LibDem".
8. The biggest obstacle for UKIP in reproducing this result nationally is the perception that they don't have enough talent to run the country. Name three UKIP politicians. I can't, and I'm broadly sympathetic to their policies. That used to be the obstacle for the LibDems.
9. I can't remember the main opposition party coming fourth in an English by-election before. (I don't doubt that it's happened, but it must be pretty rare.) Again, it makes me think that Labour were quite happy for some of their voters to vote LibDem.
10. What would have been the result under something like STV or AV? Let's assume that the 'Other' votes are reallocated evenly, all the Labour votes went to the LibDems and 70% of Conservative votes went to UKIP, with the rest to the LibDems. That looks pretty close. I can't be bothered to get my calculator out, but that percentage of Conservative voters second-choicing UKIP as apposed to the LibDems is the deciding factor.

Yeah, under AV it would have been incredibly close, I suspect. Although possibly very different. For all I know 10,000 of those 13,000 Lib-Dem voters would have voted Labour first, but felt they couldn't, because of FPTP.

Assuming, for simplicity, that the first round of voting went that way, I suspect you'd see 20% of Labour votes go to UKIP - who are pulling in support from all parties. If we assume that _and_ 70/30 split of Conservative votes, then it comes in at UKIP winning by 200 votes. So those splits are vitally important.

The talent problem should solve itself if UKIP get any success. If nothing else, I can see the right-wing of the Conservatives decamping en-masse, if they think that they stand a chance. It would be the Conservative party's SDP moment.

"If nothing else, I can see the right-wing of the Conservatives decamping en-masse, if they think that they stand a chance. It would be the Conservative party's SDP moment."

Or it could be their Conservative Party of Canada moment, according to Dan Hannan: http://blogs.telegraph.co.uk/news/danielhannan/100204821/the-eurosceptic-right-gets-more-than-half-the-vote-the-europhile-left-wins-with-less-than-a-third/

I couldn't see the latter happening while David Cameron was Conservative leader. A lot of UKIP voters are UKIP voters because they despise Cameron I suspect. But if the Conservatives did badly in 2015 (eminently possible) and UKIP did well (also possible - 20% of the popular vote in England doesn't seem that unlikely), then you could imagine a scenario of Cameron resigning and someone who hasn't called UKIP voters "loonies, fruitcakes and closet racists" becoming leader. And then, yes, some sort of popular-front-of-the-right would be possible, with the Clarke / Heseltine types retiring or decamping to the LibDems.

Oh yes, splitting the vote under FPTP is disastrous. So you have to make a decision as to whether to play the long game. And a four-party system is ever-more likely to get voting reform - particularly as UKIP have been strongly in favour in the past.

If we get a Euro-sceptic Conservative leader, then I can see that possibly defusing things, at least temporarily. You'll either see UKIP reabsorbed, or you'll see on-going infighting balanced by wanting to succeed.

And Cameron put a lot of effort into detoxifying the Conservatives - but even with that and a frankly disastrous showing by Labour, they weren't able to get a majority. I don't know what they _can_ do to get a majority, to be honest.

I’m not sure Cameron actually did put a lot of effort into detoxifying the Conservative Party. I think he put a lot of effort into making it look like he was putting a lot of effort into detoxifying the Conservative Party whilst putting a lot of effort into winking at the more toxic areas of the party and letting them think he didn’t actually mean any of the equal marriage, greenest government ever stuff. He’s reaping the reward of that now as we see the right wing of the Conservative Party basically saying WTF!

Unlike Blair Cameron didn’t take on the outer edges of his own party and persude them he was a) correct, b) the only electable options, or c) so powerful he could do what he wanted without them.



I think the electoral position of the Conservative Party is very difficult over the next few electoral cycles. They seem to have the same problem with recent immigrant voters that the Republican Party has with Latino and Black voters in the US. Even where they agree on economic and social policy the voters still perceive that the Tories / GOP perceive them as the enemy.

Barring the Labour Party going fruitloopers insane and trying to get Len McClusky in to Number Ten I think we get a series of small-majority Labour governments and Lib Dem & X coalitions ending in a coin toss followed by electoral reform.

"Barring the Labour Party going fruitloopers insane and trying to get Len McClusky in to Number Ten I think we get a series of small-majority Labour governments and Lib Dem & X coalitions ending in a coin toss followed by electoral reform."

That's what I think we're going to see, but with the caveat that if we haven't had electoral reform within at least two elections, we're going to see a Labour/UKIP coalition form. Labour have always had an anti-EU rump, and they're essentially turning themselves into a populist party of the right anyway, much like UKIP.

A Labour / UKIP coalition (in the national interest) is a sobering (or perhaps drunkenness inducing) thought.

I was interested to hear from the strategy director of the ERS about how New Zealand ended up with electoral reform.

Basically,says he, they voted in centre-right government on a middle of the road fiscal conservative manifesto and got the Tea Party. Threw them out, elected a centre-left government on a fiscal conservative manifesto and got the Tea Party and then decide to throw out First Past the Post.

Chatting to a couple of NZ friends, what I found fascinating was the way the vote was carried out.

First a referendum saying "Should we just stick with FPTP? Yes/No."

And then, when the result was "No, something else please." there was another referendum asking "Do you want AV, AMS, or STV?" - and AMS won.

And then _another_ referendum, asking "Do you want AMS or FPTP" and finally AMS won that.

http://en.wikipedia.org/wiki/Electoral_reform_in_New_Zealand

Aye – which I think dismayed some of the STV purists (puritans) back in the UK as people seemed to have voted for AMS over STV.

It depends on whether you prefer proportionality over selection.

Most people, so far as I can tell, vote based on party, not individuals, and I like that AMS allows the Greens to have two seats on 4% of the vote, which STV doesn't.

Some years ago the University of Essex (IIRC) did some detailed fieldwork about second preferences under AV. The work’s a little out of date now, as we’ve had a general election since. It’s still the best current data we have.

They have UKIP picking up 9% of second preferences from Labour. 29% of Conservative votes go to UKIP next. The Lib Dems pick up 66% of Labour second preferences, and 54% from the Conservatives. No idea how this has been affected by the coalition.

When I simulate an election under AV using the UoE second preference I get the following results.

2nd Round – assume all top votes for non top four parties split equally between the top four parties. There are not enough non top four votes to shift Labour into 3rd place.

Liberal Democrats 13,856
UKIP 12,085
Conservative 11,073
Labour 4,602

Labour still last by a long, long way.

Next up is the crucial fight between UKIP and the Tories to not go out next. Just over a thousand votes in it.

The Labour vote splits about the same for UKIP and the Tories (9% plays 7%). This gives UKIP the win into the last round with the following figures

Liberal Democrats 16,893
UKIP 12,499
Conservative 11,395

What’s interesting is that (using the UoE transfers) there aren’t enough votes going to centre right parties from Labour to affect the placings here. Labour voters give 16% of their 2nd preference vote to centre-right parties. That is about 740 votes. Even if they all went to the Tories that doesn’t close the 1,000 vote gap between UKP and the Tories.

So assuming UKIP go into the second round what happens to all those Tory preferences? They break 5 to 3 to the Lib Dems and the Lib Dems romp home by more than 7,000 votes.

Liberal Democrats 23,047
UKIP 15,804


Assuming that counter-factual that the Tories managed to scrap into second place befor the last round what happens to the UKIP vote? The Tories pick up 49% of it and the Lib Dems pick up 14%.

Liberal Democrats 17,808
Conservative 18,008

The Tories win the seat by 200 votes.

But NB the method has a real warning attached beyond the currency or otherwise of the transfers. It assumes that a transferred voter behaves like a native voter. I.e. that a Labour voter who transferred to the Conservatives next preference is 49% for UKIP. So of the 700 or so Labour voters who transferred to the Tories under the counter-factual Tory Win 350 then go on to Vote UKIP, exhausting their ballots. One might expect Labour third preferences to be more heavily weighted towards the Lib Dems. Also, in Eastleigh, no Green candidate. Greens pick up about 16% of Labour second preferences; enough to mess about with the Tory – UKIP placings depending on how you treat exhausted ballots.

So,conclusions. Very close seat under SMP remains very close seat under AV. The margin for error when modelling the outcome is greater than the margin of victory for both the second from last round and the last round. Based on quite old data. So who knows? What I think we can say is that under AV, with the actual votes cast this seat wouldn’t be a total banker for the Tories. Far from it, if they win, they squeak through two really tight rounds.

So, Condorcet then? :->

(STV, of course, solves this very well. Or AMS.)

Condorcet – surely the best voting system we’ll never be able to explain to anyone.

STV, AMS or AV+.

Although for a bye-election under STV you get AV. Unless you do what the ERS does and re-count the election as if the person dropping out had not stood for election.

Yeah, I love the idea behind Condorcet, but it is significantly harder to explain than STV, AMS or AV.

And yes, By-elections are a pain under any system. I can see the benefit of the ERS system, but it only works if each party has put forward at least more than one person than was elected.

Which reminds me - I was thinking about some of the problems with STV, and how to counterbalance them. And it occurred to me that one of the things it's supposed to do is allow you to vote for, say, the pro-war Labour candidate or the anti-war Labour candidate. But that only happens if they put forward more candidates than get elected. And I was wondering if it was feasible to enforce that each party could only have one less person elected than they had stand, so that people always had a choice. But that _also_ seems to have a variety of failure modes.

Any thoughts?

It’s one of the few problems with STV.

Certainly what we’re seeing in Scotland’s local council elections with the 3 and 4 seat constituencies is that parties are only putting up one candidate each and the bottom two placed of the 5 in-play parties and the strongest independent miss out on a seat. If you look at the Glasgow elections over the last two cycles you can draw the conclusion that in the first election the SNP missed out on lots of council seats because they only put up one candidate in seats where they could have won two.

I think requiring one more candidate than seats the party is allowed to win would work to solve the problem of having actual choice between factions.

But it’s not without it’s problems.

From the voter point of view, if I vote for someone I’d like them to be elected. I’d like to not have my prefered candidate win the election but not be allowed to take up their seat because the party they belong to didn’t put up enough candidates. Worse, what if I, as a centre-left voter don’t get my centre-left candidate because of the of the Ducker Restriction and the next candidate in the slot is from the populist right?

From a party point of view there are issues about deposits and resourcing. If the deposit is high then the party has to be able to stand to lose a chunk of money. If you imagine a city constituency with strong showings from Labour, the Greens and the Lib Dems, where a smallish shift in the vote might mean that the Greens pick up either half a dozen seats or just 1 seat with 5 lost deposits. If I’m the Greens’ election agent I have to gamble on either winning 5 more seats or losing 5 * £500 = £2,500. That sounds like a small sum but it is non-trivial to a small party. It’s probably the value of the all the leaflets we handed out during the AV campaign.

Further to that each candidate will want their own posters, leaflets etc.

An allied issue is that political parties often struggle to find enough candidates in hard to win seats.

This requirement also I think restricts parties from splitting. For example the Labour and Co-op Parties might decide to split if we moved to STV. Then I could decide to give high preferences to the Co-operatives on the grounds that I prefer the collective, self-help, small state approach to left wing politics. If both the Labour Party and the Co-op party were competing for all of the seats in inner city Manchester or Glasgow then they would have to find dozens of candidates each to fill every potential slot. Id be left with the problem of trying to work out which candidates were Co-op candidates and which were Blairites. This might incentivise them not to split.

There’s nothing to stop a particular party putting up cookie cutter candidates. In your example, all the candidates are pro-war. (And STV would make it easy for a dissenting Labour member or members to break away and form the Local Labour Anti-War Party.)

My final reservation is about accidentally exhausted ballots. You do get some voters who are in a hurry and strongly support a particular party who vote 1 – 2 for the first two candidates from that party they see and then dash off to pick up the kids or whatever and don’t give a third preference to the parties final candidate. This sort of thing seems to have cost the SNP one or two seats in Glasgow. I think if you’d been able to travel back in time and tell the voter – unless you mark a 3 in that box your guys are going to miss out on a seat here – the voter would have more completely filled in their ballot.

I think on balance the Ducker Restriction does more good than ill but I’d not relish arguing for it with Charlotte Vere.

Not sure I can think of an alternative that does any better.

I think the larger the constituencies the better the chances of having excess candidates. In a four member constituency you’d only put up one or two candidates. In a seventeen seat constituency even small parties would be more likely to put half a dozen candidates.

Actually, I've gone even further off Condorcet after reading about Condorcet Cycles, and how gameable the system is.

Yeah.

I don't think there is a perfect system out there. Each has its flaws and each its strengths. A lot depends on what you value.

But SMP is definately dominated by all other systems on offer unless what you value is a coin toss between two overly large and ill defined brands.

The bloody stupid d'Hondt system we use for Euro elections seems to me to be worse than SMP. Other than being (kinda-sorta) proportional (which I care about far less than preferentiality) it manages to have the defects of every system I can think of -- a single vote, incomprehensible counting system for most people, weird tactical voting effects *and* jobs for life for party placemen -- all rolled up in one.

We use the d'hondt system in Scotland for the proportional AMS part of our Holyrood electoral system.

I'm not a huge fan of closed list systems but I highly value proportionality. So I prefer it to SMP. </p>

As an aside I think the current Scottish system could be more easily converted to a better (IMO) system more easily than the UK system. We already have list top ups so could shift to AV+. We already have large constituencies so could shift to STV.

Come the day.

STV + topup would be fine by me.

Oh, I definitely value proportionality very highly, but I also value kick-the-bastards-out-ability. Get put on the top of a closed party list for a major party, and you never again have to worry about what the voters think. I do agree that the Scottish system has advantages over the UK one in terms of moving to an actually-good system, but I have a visceral loathing of closed lists.

Just so.

A few of the Scottish political parties have primaries where the members select or elect in some way who goes on the list and who doesn’t but that isn’t quite the same as me being able to say “Not that joker!”

Try the same exercise using these numbers:
http://lordashcroftpolls.com/2013/03/we-know-how-eastleigh-voted-heres-why/

Thank you, that was fantastic.

Also, it's great that the BBC is now reporting it as "Eastleigh by-election: Lib Dems hold on despite UKIP surge":
http://www.bbc.co.uk/news/uk-politics-21625726

Only UKIP can beat the Lib Dems here.

I can see the posters in my mind's eye.

If they do that the LDs should totally sue UKIP for stealing their stupid campaigning thing.

OMG I HATE THOSE GRAPHS.

Well, given recent events, maybe we'll get our first new campaigning idea for twenty years soon, and be able to ditch the old ones? Hope springs eternal...

I think their image problem goes beyond "not enough talent to run the country" - it's the general perception that all single issue parties (which they may or may not be, but are widely seen as) are hotbeds of infighting nutters. Any party that's been associated with Robert Kilroy Silk has got an image problem.

"Any party that's been associated with Robert Kilroy Silk has got an image problem."


Oh god, yes.

Re the Daily Mail that would be a risky strategy. Once the genie is out of the bottle...

My democratic feeling there is that there is a segment of the population who are on the right of the Conservatives, and that they need representatives. I'd be quite happy to have a saner Conservative party and the UKIP as separate entities, rather than the current mess.

All true.

I’m thinking of two risks.

One for us.

It’s the risk of tipping points under FPTP.

I think the risk is that once the DM backs UKIP at a general election and they pick up a few seats we get the sort of things that happened to the Liberals after the rise of the Labour Party (rather than the thing that happened to the Labour Party after the re-rise of the Liberal Democrats.) Everyone jumps ship to UKIP and the Tories are reduced to a rump.

This might just give the problem a new name but it might end up with a more right-wing party as the main right wing offer.

One for the Daily Mail. Once you’ve backed UKIP and it’s picked up a few seats and you’ve shown the Conservative Party that they can’t take you for granted how do you shift your support back again?

I suspect that the UKIP isn't going to ever get the mass support that the Conservatives have. Tacking that far to the right pushes far too many people away.

And many papers have switched supported parties before. You just declare that the one you were supporting are acting like idiots, and they can have your support back if they change.

You are probably correct.

However, the Labour Party in the 1920’s was considerably to the left of the Liberal Party. Although there are issues to do with the expansion of the franchise running parallel. This probably accounts for most of the transformation of the political landscape.

I think the risk for the DM is that having created the party that they want (populist on the right of the centre right) it might be difficult to switch back again to a Conservative Party that is less bits of UKIP.

There’s a line by I think Churchill about dictators being someone who rides a tiger up and down but can’t dismount for fear the beast will eat him.

Yeah, I mostly view the rise of The Labour Party as coming from the expansion of the franchise (and thus a good thing).

And yes, if all of the unpleasant bits of the Conservatives leave for the UKIP then the Daily Mail may find it no longer wants to support the Conservatives. But that's fine - the Conservatives have plenty of other newspapers.