A conundrum for Labour strategists

My next set of posts will concentrate on how the Tories should be thinking about the next election (still using current boundaries and systems for now), but before leaving Labour behind for now, I have a conundrum that Labour will need to wrestle with.

Some left-of-centre Lib Dem voters at the last election are without doubt unhappy at the turn of events. It appears that many of these are now supporting Labour. But should Labour regard this as an unalloyed positive? Should it be encouraging more to desert the Lib Dems for it? Counterintuitively, this is not a no-brainer.

The easiest group of Lib Dem voters by far for Labour to peel off must surely be those Lib Dem voters who voted for them tactically to try to keep the Conservatives out. I refer to these as Braggites, in honour of Billy Bragg who very actively encouraged just such voting behaviour. In contrast with both Conservatives tactical voters and Lib Dem tactical voters, these have been out in huge numbers. Indeed, at the last election the residual Labour vote in Lib Dem/Conservative marginals was squeezed to the limit (and beyond what I thought before the election was possible). Take a look at this table:

https://docs.google.com/document/edit?id=1RDQ4wbLZZuCg_0Nc6gQRS9wjL7BuYDualkvWJ2orvrc&hl=en&authkey=CJ6b3NkH

It sets out the Lib Dem seats which are on the Conservatives' top 100 targets, together with the Labour vote and percentage share. You will note that in most constituencies (and every English constituency in the top 100 Conservative targets where Labour were not in fact second), Labour is below 10% of the vote share and sometimes a long way below 10%. When the Conservatives or the Lib Dems are third, 15-20% is a more normal level of vote share. We can, I think, ascribe this to Braggite voting.

Now the Braggites are the group that have most to be cheesed off about. The Lib Dems have not behaved as they expected (whether they were right to expect something else is a different matter, but not relevant for present purposes). Many of them will probably vote untactically next time. But how many?

The impact of this is important: for the Tories. Without getting another vote, they could pick up 13 more seats if Labour's natural level of support is 15% and the Braggites return home. And a further 7 if Labour's natural level of support turns out to be 20%. And a further 5 seats if Labour's natural level of support turns out to be 25% and all the Braggites return home (including four seats not even within the Tories' top 100 targets). So the Conservatives could have an overall majority of 12 without picking up another vote.

Labour seems to have spent the last three months doing its utmost to regain Braggites. And obviously it does want to gain voters. But how does it gain voters in the areas that it wants to gain them without giving the Tories a leg-up? It's not an easy dilemma to resolve.

antifrank

The golden eggs: Lib Dem voters in Labour/Conservative marginals

One subject that is sure to remain the subject of controversy for the next few years is the extent to which Lib Dem voters can be harvested by the Labour party in opposition to the coalition. It is critical to the Labour party's strategy for the next few years. There is certainly a good deal of anecdotal evidence that many leftwing Lib Dems are profoundly uneasy about the coalition and Lib Dem polling is down, while Labour polling is up (though most pollsters have seen a rise in Conservative polling also). How might this translate into a strategy in marginal seats on the ground?

Where (politically) are the Lib Dem voters?

The first (slightly surprising) thing to note is that Lib Dem voters are to be found in substantial numbers almost everywhere. I attach a table (Asterisks against the name of the constituency indicate that Labour is third. Obelisks against the vote share indicate that the Lib Dem vote share was static or fell.):

http://docs.google.com/document/edit?id=1un5JXh9AKRaGcAhcWbOTkneGsD5wQJ3PCEhVTk2PEEk&hl=en&authkey=CNmdzOoN

On the list of Labour's top 150 targets, 125 of those seats are Conservative-held. The Lib Dems polled more than 10% in each and every one of those seats. It increased its vote share in 92 of these seats. This is not the performance of a party being relentlessly squeezed in a marginal.

By way of contrast, Labour polled under 10% in 12 out of the first 13 Lib Dem-held Conservative targets. Indeed, in the 20 English Lib Dem seats in the top 100 Conservative targets, Labour polled under 10% in all bar three: Norwich South and Bradford East (in both of which Labour finished second) and Berwick-upon-Tweed. In several cases, Labour lost its deposit. In every one of these 20 seats, Labour's vote share was down and usually substantially. Labour's vote share was also down in all of the first 20 Conservative-held Lib Dem targets. While it polled above 10% in 9 of these, this is less good than it looks: it had held four of these seats before the last election and been second in two more).

It has to be said that it is the Labour party's amazingly low vote shares in Conservative/Lib Dem marginals that is exceptional, not the Lib Dems' higher vote shares. The Conservatives have not been squeezed anything like as hard in Labour/Lib Dem marginals. In only one of these seats on either the Labour or Lib Dem target list is the Tory tally below 10%: Dunfermline and West Fife.
The logical conclusion, therefore, is that historically Labour voters have been unusually willing to vote tactically in Lib Dem/Conservative marginals (and this is borne out by anecdotal experience). If this is so and these are many of the voters who are thinking about no longer voting tactically, the main beneficiaries will be the Conservatives.

Where (geographically) are the Lib Dem voters?

I attach a list of the Conservative-held Labour target seats broken down into categories of Lib Dem vote share:

http://docs.google.com/document/edit?id=1FV31Yh2N1rtdWk8UBVU9nbLrfAl3OpDh5b2YNzVESFg&hl=en&authkey=CJmwzL4K

Many of the Labour/Conservative marginal constituencies where the Lib Dems have polled particularly well fall in two areas of relative Lib Dem strength. In an arc stretching from Warrington South in the south west, through Altrincham & Sale West, through three Lib Dem-held constituencies, Wythenshawe and Sale West, Stockport, through High Peak, Colne Valley, Calder Valley, Lib Dem-held Burnley and Pendle, the Lib Dems have an arc of seats curving around Manchester in which they scored at least 20% and more usually 25% or more.

There is of course an even hotter spot in the south west. 49 out of 55 seats have at least a 20% Lib Dem vote share and more often 25% or higher. More than 30% is the norm. The Lib Dems hold 15 of these seats.

Set against this, there are areas of relative Lib Dem weakness. There is something in the Thames estuary water, because there isn't a single seat abutting the water from Bethnal Green & Bow to Castle Point on the north side inclusive and from Erith & Thamesmead to Gravesham on the south side inclusive where the Lib Dems tally more than 15%. Perhaps the Lib Dems don't do unglamorous very well.

All of which is leading into the fundamental question...

Who votes Lib Dem anyway?

Many Labour supporters are very fond of the notion that Lib Dems are pseud0-Labour supporters: people too well brought up to vote Labour. Certainly, the Lib Dems have been very happy to take advantage of this notion in Conservative-held seats where Labour cannot win. But in seats that the Lib Dems could not win and were not even second, the Lib Dem vote share and absolute numbers of votes generally went up at the last election. If these voters were fundamentally left-leaning rather than right-leaning, why didn't the Labour party's message not to let the Tories in work this time in Labour/Conservative marginals? Come to that, if these voters hated Gordon Brown so much, why didn't they vote Tory? The answers to these questions should determine Labour's strategy to win the next election.

I don't have definitive answers, but here are a few suggestions as to why voters might vote Lib Dem in a Labour/Conservative marginal:

Not voting tactically, but for the preferred party

This, I suggest, will make up a plurality of Lib Dem voters in such seats. Lots of voters don't vote tactically, some because they haven't considered it and some because they won't consider it. That then leads onto the question of what the Lib Dems offered that neither Labour nor the Conservatives offered. I suggest that the answer is a reliability on civil liberties, a wish to redistribute but not to soak the rich, a genuine interest in localism and an open inclusive tone on immigration and the EU, coupled with the most pacific internationalist tone. They offered economic soundness via Vince Cable and an exciting, charming fresh faced leader in Nick Clegg.

There is a definite regional range in Lib Dem support. In the southwest, the Lib Dems have largely supplanted Labour as the opposition to the Conservatives. Around Manchester, the Lib Dems have made major advances in former Conservative territory. Labour should not assume that a single message will work equally well with all who supported the Lib Dems positively at the last election.

The range of positive Lib Dem voters will respond differently to the coalition depending on their competing concerns. Some assets have been downgraded: Nick Clegg is in the shadow of David Cameron, Vince Cable has been sidelined, the cuts have bitten more strongly than most Lib Dem voters probably would have liked. Set against that, the budget did include a raising of tax threshholds in line with Lib Dem wishes and the civil liberties and localism agendas have been honoured. The coalition has been winning the argument with the public about the need for cuts, even if they are disliked in concrete terms.

I suggest - I put in no stronger than that - that if the Lib Dems do lose ideological votes in these seats, then it is not immediately obvious that those ideological votes will go to the Labour party rather than the Conservatives.

Single issue voters

Some voters vote for a party on the strength of a single issue. Labour probably should not chase these voters.

Protest votes

There are plenty of voters out there who wanted to register their alienation at this election. The big two parties polled a lower combined vote share than has been seen in any of our lifetimes (excepting JackW). The Lib Dems were partial beneficiaries of this, though other parties benefited also.

If these voters still feel alienated at the next election, they will not vote for any of the main three parties. The Lib Dem vote share will go down, but Labour would not benefit. Labour needs to think about how it can get these voters to identify with it.

Anti-establishment voters

Closely allied to protest voters are those voters who wanted to stick two fingers up at the establishment. They genuinely did not mind which of Labour and the Conservatives got in.

These voters are possibly easier for Labour to win. They are not exclusive to the Lib Dems: the Greens, BNP and UKIP all got such votes (UKIP actively encouraged them). If after five years the establishment is Lib Dem and Conservative, Labour may benefit here.

Voting for the candidate

Some voters vote for an individual candidate. It seems doubtful that too many third placed Lib Dems benefited from this, but you never know. Anyway, Labour will have to address this through candidate selection.

The summation of all this is that Labour probably do have a bit of an advantage in winning over anti-establishment voters and have plenty to work with in winning over positive Lib Dem voters, but that they don't have all that great an advantage over the Conservatives when squeezing the Lib Dems, even if the Lib Dem vote flakes away. They should be planning carefully on a message that appeals to soft Conservative voters also.

Compromising with the electorate: how Labour should plan for 2015

In my post yesterday, I noted that while Labour do not face that difficult a job in becoming the largest party, they have an altogether more substantial challenge if they want to get an overall majority. Even on current boundaries, they need a swing that is bigger than all bar four of the 16 post-War elections, and things are likely to get worse if the coalition’s electoral reforms go through even in part. Labour will need a positive strategy to combat that.

For now, I intend to continue analysing Labour’s position in terms of the 2010 electoral system. We can always update it later and I doubt that the points for consideration will change fundamentally. If anything, they will be strengthened.

I don’t intend suggesting policies, but I do intend to identify the general thrust of who Labour should be trying to win over. So, to be constructive, here are a few key points:

It’s not really about the Lib Dems

There are 19 Lib Dem seats on the top 150 Labour target seats. In four of them, Labour is in third place or lower. In five of them, Labour need a swing of 8% or more. Labour would be doing well to take 11 of these seats from the Lib Dems, even if the Lib Dems' vote shares collapse badly at the next election, and that’s without allowing for the incumbency that many of these MPs will have built up. Labour need 67 seats, so obliterating the Lib Dems won’t be a complete answer.

Could hammering the Lib Dems particularly help Labour in Labour/Conservative marginals? It seems doubtful. In those cases, voters have chosen the Lib Dems for one or more of the following reasons: they’re ideological Lib Dems (whatever that might mean); they like the candidate; they wish to cast a protest vote; they have a particular hobby horse; they wish to express opposition to the establishment. Only the first and last of these groups would seem fertile ground for Labour, and the Conservatives may be able to reach out to previous Lib Dem voters as effectively as Labour can.

It is about the Conservatives

Labour have 125 Conservative seats on their top 150 target list. Labour need to get a message that appeals in these constituencies. This will be a lot easier if Labour can win over Conservative voters in these constituencies.

Geography is destiny

Labour need to break out of their homelands. Only 83 out of the 150 top Labour targets are adjacent to a Labour seat (I have drawn up a list of these if you’re interested). The 67th seat on this list is Bermondsey & Old Southwark and would require a swing of 9.55% to take it. That level of swing has been achieved at a general election only once since the second world war, in 1997. So Labour need to work hard to break out of its current territories and win back some of those smallish town and semi-rural constituencies that it lost so badly this time if it wants to win an overall majority at the next general election.

Labour should also look carefully at its target list by geographical region. It has maxed out in the north east – there are only three targets in that area and two of them look likely gains if there is any kind of revival in Labour’s fortunes. There are 9 targets in Scotland, but only one of these is Tory. Scotland is not going to be central to Labour’s election effort.

By way of contrast, there are 16 targets in the northwest that would be taken on a 6% swing. Labour’s next leader should make sure that the shadow minister for the Duchy of Lancaster pays close attention to that county’s politics.

The west midlands has 20 Labour target seats, but is particularly noteworthy because 8 of these need a swing to Labour of between 5 and 8% (and four more that need bigger swings still). Labour need a concerted push here. The east midlands has 17 target seats, but 10 of these fall on a swing under 5%. These look easier targets.

Perhaps the hardest work needs to be done in the east and the south east – with 16 and 18 target seats respectively, but only 6 Labour MPs between the two regions, Labour need to show that they can once again speak to a substantial cross-section of these regions’ constituents.
London also has 18 targets for Labour. Cutting across these three regions, there are roughly 30 target seats that are either in London suburbia or the wider London economic orbit. Labour cannot afford to let these seats become secure in the Tories’ grip. So Labour need to think carefully about how they can put together a message that appeals to the residents of the wider London economy. It’s going to be a tough job.

antifrank

A look back at the races

Ah, after the surprising delight of seeing I was pretty good at qualifying, the less pleasant task of looking at my race performance and trying to discern why I’m not very good at it.

First off, a list of profit/loss per race [not including qualifying, obviously], assuming a £10 stake per tip and no laying/backing as applicable to hedge bets.

Bahrain – 3 tips, one right – red £10.50
Australia – 2 tips, both wrong – red £20
Malaysia – 3 tips, all wrong – red £30
China – 4 tips, 3 right – green £38.81
Spain – 3 tips, all wrong – red £30
Monaco – 1 tip, wrong – red £10
Turkey – 1 tip, right – green £17
Canada – 3 tips, all wrong – red £30
Valencia – 1 tip, wrong – red £10
Silverstone – 1 tip, wrong – red £10
Germany – 1 tip, wrong – red £10
Hungary – 1 tip, wrong – red £10

So from £240 total stake money that’s a net loss of £114.69. Which is, er, terrible. Green from just 2 of 12 races.

When I was reading through the old articles, something struck me: I need to stop dicking about on minor markets, classifieds, points, safety cars and first lap leaders.

All (four…) of the winning tips were on the major markets (podium and winner). I’m going to try and predict a winner (or tip laying someone for the win) at every race from now on, and generally avoid the minor markets.

The winner market matters most. This is because it has most liquidity which affords a helpful combination of allowing larger stakes and the best opportunity for hedging a bet. At the last race I made a piffling extra sum on bets not tipped because there were only a few pounds available (Hulkenberg for Q3). So, getting the win right matters most.

Incidentally, I don’t think this can be attributed to the lack of refuelling. Whilst that did throw me at first, it’s no excuse for months of inaccuracy. I do think I lack a certain intuition. Others correctly predicted Button’s wins, at long odds, but I never saw them coming at all. I think I work best by analysis of numbers, and tried to make up for the lack of refuelling weight information by going for daft markets, to dramatically bad effect.

There is no P3 qualifying simulation equivalent for the race. However, we do know certain things (the McLarens have the edge on straights due to best top speed, Red Bull and Ferrari have superior downforce etc). We also now have a good history of the season, and certain patterns have emerged (excepting changeable conditions, Hamilton tends to perform better than Button, Schumacher can struggle to make Q3 as Button can occasionally, Vettel often cocks up a promising position etc).

It’s fortunate for me that I decided to have a go at qualifying betting. Despite my woeful race record it’s roughly evens overall. But, there are still many races to go, and opportunities abound for me to make amends and improve my stats.

In addition, I’ve been keeping note for the last few races of how my own results (with laying/backing) compare with a straightforward £10 bet with no hedging for every stake, and so far hedging has proven £25.08 more successful (from Silverstone to Hungary). Only Hungary had its result weakened by hedging, and that by £1.60 [admittedly people hedge at different times, so this is somewhat imprecise].

The next article will be in a fortnight, and focus on Spa’s pre-qualifying situation.

Morris Dancer

Getting to 67: the size of Labour's task at the next election

It is, of course, a long way off a general election. And at present, unusually, we have no clear idea of the number of seats on which it will be fought, the boundaries of those seats or even the electoral system on which it will be fought. All this would suggest that we should wait and see before thinking too much about an election that may very well be in 2015.

But there are markets open at the moment: on Sporting Index, you bet on whether Labour or the Conservatives will get most seats or whether either will get an overall majority. And of more general information, it is essential for Labour supporters to understand whether they have an easy task (in which case the aim should be to help the coalition to fray) or a difficult task (in which case they need to have a major rethink). After all, they have to vote on a new leader, and they need to understand the challenge ahead when deciding who is right for that job.

So let's have a look at the next election assuming that it is going to be fought on existing boundaries and on the existing system. After all, any changes are almost certainly going to make things harder for Labour, not easier.

At present, the Conservatives have 306 seats, Labour have 258 seats and the Lib Dems have 57 seats. By historical standards, the seat count is close. Labour would need only 67 extra seats to get an overall majority and fewer seats to be the largest single party - perhaps 30 (bearing in mind that if Labour picks up seats from the Lib Dems, the Conservatives will probably pick up more). Given that the Conservatives have only just increased their seat tally by 97, neither sounds a superhuman challenge.

But, I contend, this substantially underestimates the scale of the problem. Have a look at these two links, one to the table of Labour targets compiled by Anthony Wells on UK Polling Report and one to the historical uniform national swings at general elections since 1945 from Wikipedia:

http://ukpollingreport.co.uk/guide/labour-target-seats

http://en.wikipedia.org/wiki/Swing_(United_Kingdom)

You will notice that to get 67 seats on a uniform national swing against all other parties, they will need to get a swing of 4.58%. A swing between Conservative and Labour of that degree has been achieved in only four general elections out of the 16 that have been held since 1945: 1970, 1979, 1997 and 2010. Applying basic mathematics, that gives Labour a one in four chance of getting an overall majority, all other things being equal. And that assumes that the swing is to Labour, not against it, which is not automatic: in five out of these 16 elections, the swing was to the governing party. I do not think that many Labour supporters have appreciated how tough that challenge is. There are far fewer marginals now: 67 seats is a short but steep climb.

It is much easier for Labour to become the largest party. For that, they would need a swing of 1.91% against all other parties. That has been achieved in 10 of the 16 elections since 1945 (though again the swing could be against Labour as well as to it).

So Labour supporters could realistically have good hopes of becoming the largest single party at the next election if the current system is used, but to get an overall majority, they need to do something out of the ordinary if they are not going to rely on the coalition failing. If the system is changed, the chances are that both are going to get harder. They need to pick a leader who is going to take them out of their comfort zone.

In my next piece, I shall look at what the seat distribution has to say about what strategy Labour should follow to get back to power.

antifrank

A look back at qualifying

To interrupt the looong break until Spa, I thought it might be interesting and possibly even helpful to look back the season so far. This weekend, I’ll write about qualifying (which I rarely, if ever, bet on last season). Next weekend, I’ll take a look at race betting.

The figures below all assume a £10 stake with no laying (or backing, as the case may be). [I do advocate setting up in-game lays/backs to mitigate losses].

Bahrain – 2 tips, both wrong – red £20.
Australia – no tips
Malaysia – 2 tips, 1 right – green £45.58
China – 2 tips, both wrong – red £20
Spain – 1 tip, right – green £45
Monaco – 1 tip, wrong – red £10
Turkey – 1 tip, right – green £27.50
Canada - 1 tip, wrong – red £10
Valencia – 1 tip, right – green £28.50
Silverstone – 2 tips, 1 right – green £23.34
Hockenheim – 1 tip, wrong – red £10
Hungary – 1 tip, right – green £8.50

So, that’s 1 neutral, [Australia because I didn’t tip], 6 green and 5 red. The net result so far is a profit of £108.42 from a total stake of £150. That’s quite a bit better than I was expecting, but does make me a bit concerned about how woeful my race betting must be to have a net loss overall (so far) for the two combined.

Anyway, back to qualifying. Generally, I’ve done pretty well. Two losses at least could easily have been wins (Kubica for pole in Monaco when he came second and Alonso at Hockenheim where he lost by 0.002s). I seem to be reasonably good at both pole and the Q3 market, and was amused to read in an early season pre-qualifying post that I was considering stopping betting on it. If anything, I should ignore the race and bet only on qualifying, based on present stats.

I suspect I’m better at qualifying than the race because there is more information available. The P3 qualifying simulation offers a very good opportunity to assess prospective performance, and usually that’s enough to be either correct or close enough laying is possible.

Last season there was more pre-race information available because the fuel loads were effectively made public (weight was published, which indicates fuel on-board). This made it possible to forecast who would pit when, whether they’d be in traffic, the condition of their tyres, who would benefit from a safety car etc.

However, I’ll say more about this during the probably depressing race betting post I’ll write for next weekend.

Morris Dancer

Hungary: post-race analysis

Vettel had a decent start, Alonso got close but not close enough. So, one tip from two right but because of the woefully short odds if you bet £10 per tip and didn’t lay, you’d be down £1.50.

Congratulations to Jamei for his great call to lay Vettel at 1.5 or so. Infuriatingly, I did actually bet in-game to lay him at 1.28 when the safety car came out, but when he was second with Webber unpitted in first I closed that position down to end neutral. Then, a few laps later, Vettel got the drive-through. So, bloody annoying, but kudos to Jamei.

Alonso got second, and managed to whine like a bitch about it in the press conference. Kudos, El Grumpino, it takes a special sort of miserable git to complain when giving a press conference having achieved a bloody lucky second place.

The race, despite having few overtaking manoeuvres at the sharp end, was quite exciting. I’ve already mentioned the race-deciding incident which had me dreaming of ways to annihilate the ever more irksome Legard [BBC commentator]. Near the end of the race Schumacher pulled a disgraceful move, trying to defend his place by pushing Barrichello into the wall. Thankfully the Stigslayer survived and passed his nemesis to get 10th and a championship point.

Elsewhere, Hamilton was running a solid 4th, ahead of Massa, when his car suffered transmission failure. A rare reliability problem for the McLaren. Further down the grid his team mate netted a paltry 8th.

Kobayashi went banzai, as predicted by me, but did better than I expected, surging from 23rd to 9th [he was 5 to get points, a bet I considered but decided against].

Petrov put in a stunning performance, coming 5th, just ahead of fellow rookie Hulkenberg who got himself 6th. De la Rosa was 7th, making this a nice weekend for Sauber.

Mercedes had a shocker. Schumacher’s attempt to leave a Barrichello-flavoured bloody smear on the wall failed and he got no points, and Rosberg, who had looked good, ended up being rather unfortunately retired when a tyre was put on improperly and left as soon as he exited the pits.

Nice to get a tip right, even if it was at rubbish odds but a bit annoyed with myself due to the net loss. In addition, this was the first race since I started keeping more detailed records where laying would’ve put you in a worse position than not laying.

So, Vettel should’ve won this. He was fastest by miles, but a strange cock up over the safety car led him to get a drive-through, giving the grumpiest cheat in the world second place. Webber was fortunate, but would’ve got second at least due to some good lap times.

Here are the contenders for the Drivers’ title:
Webber 161
Hamilton 157
Vettel 151
Button 147
Alonso 141

I do think McLaren can pull this back. They’ve got a long break until Spa (last weekend in August) and last year showed they can develop the car throughout the season. In addition, it’s clear to see where they can improve (flexi-wing and blown diffuser). Furthermore, Hungary was a very bad track for McLaren and very good for Red Bull, and they were unlucky Hamilton didn’t get 4th, which would’ve had him still leading the title race.

Because it’s such a long time until the next race, I might write one or two inter-race articles, depending on what happens.

Disappointing to have a second consecutive losing weekend, but at least the margin was small. I think I’m reasonably good at qualifying but relatively poor on race day [like Vettel] so that’s something to try and work on.

Morris Dancer