Tuesday 23 December 2014

Testing the boundaries (3): the Lib Dems vs all comers

My first post in April was about the Lib Dems' prospects and I have returned to the case regularly. It's time to take another look, this time considering where value bets might lie if the Lib Dems overachieve or underachieve against any particular batch of opponents.

Let's have a look at the Lib Dem seats by reference to different opponents: first, against the nationalists, then Labour and finally the Conservatives.  The three battles are different, but the contours of the battle in each case are similar.  In each case the Lib Dems will be fighting off a threat based on national politics by appealing to the public using the virtues of the local candidate.  This means that if the Lib Dems do adequately, results will vary considerably depending on the perceptions in each seat of the virtues of the local candidate, but if the Lib Dems crash and burn, the national picture will be a much more reliable predictor of events.

Incumbency is critical.  Here is a slightly out-of-date list of the Lib Dem MPs and their plans for 2015:


Jeremy Browne subsequently decided to stand down in October 2014 and Mike Hancock is not being reselected for Portsmouth South.

The nationalist challenge

Here are the seats where the Lib Dems are facing a serious nationalist challenge:


This table, as with the tables that follow below, is organised by the Lib Dem price of taking the constituency.  There is some duplication between these tables because the Lib Dems feature in a few three and four way marginals.   Every seat where the Lib Dems are priced at 16/1 or less is included.

Far more than Labour, the betting public are convinced that the Lib Dems are going to be hammered by the surge in SNP support.  They are favourites to retain only four of the twelve seats that they hold where the SNP or Plaid Cymru are in the running, and shorter than 2/1 in only seven of these seats.  On current polling, this is actually kind to the Lib Dems: in the most recent YouGov poll in Scotland, the Lib Dems, who took 19% of the vote in 2010, are polling 4% (this is not particularly out of line with other Scottish polls) and the SNP have taken 49% of the 2010 Lib Dem vote: 


The most recent Welsh opinion poll has the Lib Dems on 5% (down from 20% in 2010):


We have so far seen no constituency polls in any of these seats since the Scottish referendum, so we must use such information as we have.  If the polls do not improve, the Lib Dems should actually be pleased to take as many as four seats from this list. 

If the polls are broadly right, is there potential value betting against the Lib Dems?  Earlier in the year I pooh-poohed the idea that the SNP might be as short as 10/1 to oust Charles Kennedy.  The same bet is now available at 3/1.  I still find this hard to credit (though I accept that it is not a 10/1 shot now).  The evens on the SNP in Caithness, Sutherland & Easter Ross is, however, one that I can't look past. 

While Plaid Cymru are not soaring in the polls like the SNP, the Lib Dems are apparently cratering in Wales as well as Scotland.  The 7/4 on Plaid Cymru may be worth a flutter.  Mark Williams built up a big majority in 2010 and clearly has a strong personal following, but Plaid Cymru held the seat from 1992 to 2005 and have held the Welsh Assembly constituency since the Assembly's inception.  On the assumption that the Lib Dems will fall back in Ceredigion to some extent as elsewhere in Wales, Plaid Cymru will be the natural repository for those votes.  I've put a little on this, but I'm not betting the house.

If, conversely, the Lib Dems are going to outperform, where might that be?  As I noted previously, their good results will be very seat-specific.  Incumbency is going to be hugely important, so I would not be looking for outperformance in either Gordon or Fife North East, where there will be new candidates.  Could Michael Crockart hold Edinburgh West?  He's a first termer, so he should get some form of incumbency boost, and he may act as the rallying point for Conservatives or unionist Labour supporters who want to stop the SNP.  It's a seat that the Lib Dems have held since 1997 and in 2005 they held it with 50% of the vote.  It seems unlikely that they can hold on but it's a 6/1 shot.  I've put a small amount on this too.

In general though, I'd rather be backing the SNP than the Lib Dems in these seats.

The Labour challenge

Here are the seats where Labour are challenging the Lib Dems:


The Lib Dems' battle against Labour doesn't look a happy story either.  They are favourites to win only four of these seats and priced below 2/1 in only seven seats.  They are expected to get a hammering from Labour. 

If you take the view that the national picture is of most relevance in the Lib Dem battle against Labour, there isn't a seat on this list which is safe for the Lib Dems.  Personally, I expect Nick Clegg to hold Sheffield Hallam and Greg Mulholland to hold Leeds North West, possibly with some tactical Conservative support in each case. 

There have been constituency polls in some of these seats.  Lord Ashcroft has conducted polls in Sheffield Hallam, Birmingham Yardley, Burnley, Bermondsey & Old Southwark, Cambridge, Cardiff Central, Hornsey & Wood Green, Redcar, Bradford East, Brent Central, Manchester Withington and Norwich South.  ICM have also conducted constituency polls in Cambridge, Redcar and Sheffield Hallam - though the methodology of these has been the subject of much public debate, given that they were apparently commissioned with a view to pushing Vince Cable's chances of taking over from Nick Clegg as party leader.

By and large, these constituency polls back up the markets' perceptions of this battleground.  The Lib Dems were recorded as leading only in Sheffield Hallam, Birmingham Yardley and Bermondsey & Old Southwark, and in none of these by any great margin.  In many of the seats, Labour were apparently out of sight of the Lib Dems.

Given the uncertainties of constituency polls, if you are looking for Labour bets, you might take the view that the 7/4 on Labour in Bermondsey & Old Southwark is a decent bet - Labour were only behind by 1% in this poll.  Certainly it looks better value than the evens on Labour in Cambridge, where Labour were only ahead by 1%.   

The seats on this list where Labour are short-priced to win are also well worth considering.  The constituency polling suggests that the 1/5 on Labour in Bradford East, the 1/5 on Labour in Redcar and the 1/6 on Labour in Burnley is safe even if the Lib Dems revive substantially.  The 1/3 on Labour in Hornsey & Wood Green looks very respectable, given that it recorded a double digit lead even after respondents were prompted to think about their constituency.

One thing worth noting about both Bermondsey & Old Southwark and Hornsey & Wood Green is that Labour seem to be doing disproportionately well in London against the Conservatives, and there seems no reason to believe that a strong Labour performance should stop at the constituency boundaries with Lib Dem held seats.

On the other side of the fence, I would take more notice of constituency polls as they relate to the Lib Dems than anywhere else, in particular paying attention to those occasions where the Lib Dems have a sharp increase in support when voters are prompted to think of the candidates in their own constituency.  The 10/11 on the Lib Dems in Birmingham Yardley looks good.  As I have previously noted, both constituency polls in Cambridge seem to have different problems and I'm on the 5/4 for the Lib Dems to take this seat.   While we haven't had a constituency poll in Bristol West, the 8/13 looks decent value for the Lib Dems to hold a seat with a large majority.

There is absolutely nothing in the constituency polling to suggest that the Lib Dems are going to salvage any of the seats on this list where they're longer priced. 

The Conservative challenge

Finally, the seats where they face the Conservatives:


At last the Lib Dems have grounds to hope.  They are favourites in 24 of these seats and joint favourites in a 25th.  They are shorter priced than 2/1 in 31 seats on this list.

These odds are broadly backed up by the numerous constituency polls in such seats that Lord Ashcroft has conducted.  They can be found here:




What is immediately apparent is that the Lib Dems' performance is a long way from uniform.  Sutton & Cheam and Carshalton & Wallington are swinging sharply to the Lib Dems.  Meanwhile, Somerton & Frome and Portsmouth South are showing large swings to the Conservatives.  The individual polls may err, but the overall picture of local variation is almost certainly correct.

I tentatively draw the conclusion that as far as their battles with the nationalists (and particularly the SNP) and Labour are concerned, the Lib Dems may well have crossed the event horizon and local reputations are probably not going to be enough to save MPs' skins in general, but that the Lib Dems are keeping their heads above water sufficiently in their national battle with the Conservatives that assiduous MPs can survive.

Once again, we should start by thinking about incumbency.  Where MPs are stepping down, the Lib Dems are unlikely to inherit the seat unless the majorities are huge.  It is no coincidence that both Somerton & Frome and Portsmouth South have retiring MPs.  We also have constituency polls for Berwick-upon-Tweed and Mid-Dorset & North Poole, where the Conservatives are also on course to win, though with less spectacular swings.

Incumbents are also standing down in Taunton Deane and Bath.  The majority in Bath would survive a swing of the kind being recorded in Somerton & Frome and Portsmouth South, so the 5/1 on offer there is probably not very good value.  The constituency poll in Taunton Deane (which already showed the Conservatives in the lead) predated Jeremy Browne's announcement that he was standing down, so the 8/13 on the Conservatives in that seat represents excellent value.  I'm already on the Conservatives in Portsmouth South and the 2/5 on the Conservatives in Somerton & Frome looks well worth it, given the size of the Conservative lead over the Lib Dems in the constituency poll.  Even allowing for the vagaries of constituency polls, this would need to be very wrong for 2/5 to be a poor price.

I pay particular regard to the constituencies where in Lord Ashcroft's polls he finds a particularly high increase in Lib Dem support when respondents are asked to focus on their own constituency as compared with the general voting question.  That rise is 15% or more in the following constituencies: Carshalton & Wallington, Cheltenham, Colchester, Eastbourne, Eastleigh, Kingston & Surbiton, Southport, Sutton & Cheam and Thornbury & Yate.  I infer that those constituencies have MPs who are particularly well-regarded locally.  I would not bet against them holding their seats and I would back them where the prices are attractive.  The 4/5 on the Lib Dems in Sutton & Cheam with William Hill is very good - take it.  Lord Ashcroft has conducted two polls in this seat and in both the Lib Dems have a double digit lead.  This should be somewhere around 2/5 in my book.  I have also backed the Lib Dems in Eastbourne at 4/7.

What of longer shots?  With the exception of Watford, polling in all Conservative-held seats looks dismal for the Lib Dems.  The 1/5 on the Conservatives in Newton Abbot is probably good value.

One seat which is worth keeping an eye on is Berwickshire, Roxburgh & Selkirk.  North of the border, the SNP have taken a big bite out of the Lib Dems' support.  The SNP would be doing sensationally well if they were to take a border seat, given how hostile to independence this region was, but that bite may be sufficient to let the Conservatives win even if they make no further progress.  Take the 11/10 on them to take this seat.

1 comment:

Dave B said...

The recent BES conference mentioned LD targeted spending. No idea where they found the data.

"Portsmouth South is one they are not hopeful of retaining. And they won’t fight Gordon very hard, he suggests.

But a seat like Hazel Grove will be interesting. Andrew Stunnell is standing down, he says. The Lib Dems are not spending much money there. But they are spending money in a next door seat, he says."

http://www.theguardian.com/politics/blog/live/2014/dec/09/british-election-study-conference-politics-live-blog#block-5487121ee4b02f79feb619d1