Friday, 16 May 2014

Core Cities: Labour's reservoir

Up to this point, I have largely looked at constituencies by geographical region. This is how everyone else does it, and for good reason. It is undoubtedly the case that different regions have a different sense of place. The South West is very different from Yorkshire, the West Midlands are very different from the North East.

But there are different ways of bunching constituencies and some of these may prove more fruitful than mere contiguity. The eight largest cities in England after London have banded together to form a group which calls itself the Core Cities. These are Birmingham, Bristol, Leeds, Liverpool, Manchester, Newcastle, Nottingham and Sheffield. I have noticed that constituencies in the Core Cities often behave differently from others in the same region but behave similarly to each other. So I decided to have a proper look at this.

Here is the Core Cities website:

As you will see, the Core Cities boast that together with their surrounding urban areas they are home to 16 million people, generate 27% of England’s wealth (more than London), are home to half of the country’s leading research universities and contain 28% of highly skilled workers (graduate level or above). But with the exception of Bristol, all of them have below average GDP per capita, Sheffield and Liverpool especially so.

These cities look quite different from much of the rest of the country.  They have much higher levels of public sector workers (25.7% of all employees, as compared with a GB average of 21.4% in 2011), so would have been much worse hit by government cuts.  The Core Cities also had greater private sector job losses in the recession than the GB average. Gross disposable household income of residents in the Core Cities is 20% lower than the GB average.  There is evidence that the Core Cities in aggregate have been growing more slowly than medium sized cities.  It is important to recognise that the different Core Cities have had different experiences.  For example, Nottingham had very severe public sector job losses in the recession, but equally grew much more rapidly than other Core Cities in 2012.  While the grouping is convenient and meaningful, it is not the be all and end all.

When looking at the seats, we face is a definitional problem. Where city boundaries are drawn is often a matter of historical accident. What most people would regard as Manchester is far bigger than Manchester City Council: Manchester United play home matches within Trafford Council. I have therefore decided to look at this first using narrow definitions of the extent of cities and then looking at looser definitions (so for this I've also included seats in Greater Manchester, Merseyside, Tyne & Wear and the old West Midlands county). This will not fully satisfy anyone, but the underlying picture is, I think, clear, either way.  It's worth bearing in mind that some other seats may be formally outside the metropolitan area but may behave in a similar manner to those listed below.

Here are the constituencies in these seats:

Under the strict definition of Core Cities, there are 43 seats.  Of these, 35 are Labour held, 4 are Conservative and 4 are Lib Dem.  There are a further 59 seats in the rest of Greater Manchester, Merseyside, Tyne & Wear and the West Midlands.  Of these, 9 are Conservative, 5 are Lib Dem and 45 are Labour held.  Out of 102 seats in the core cities, Labour hold 80.  Labour is nearly as dominant here as the Conservatives are in the East and South East (neither of which region contains a Core City).  Labour also holds all seven Glasgow seats and four out of five Edinburgh seats (both would be of Core City stature, had they been English).  If we treat these as Core Cities also, over a third of all Labour seats are to be found in Core Cities.

This is not just a north/south thing.  It's a big city thing.  Of the 33 seats in the North West that are not Core City seats, 19 are held by the Conservatives, two by the Lib Dems and 12 by Labour.   Of the 40 seats in Yorkshire & The Humber that are not Core City seats, 17 are held by the Conservatives and 23 by Labour.  The Conservatives hold 26 of the 31 seats in the West Midlands that are not Core City seats and 31 out of 43 of the East Midlands seats that are not Core City seats.  Only in the North East and Scotland do we see a different story.

Oh, and London.  London is a big city where Labour don't have it all their own way.  Once again, London is the exception.

It makes no sense to present these seats as a Conservative battleground, so I have presented them as a Labour battleground here, listed by odds:

Note how many seats have no markets as yet (and they aren't going to be interesting markets if and when they appear).  If the bookies and punters are correct, Labour are poised to gain another seven Core City seats.  They have respectable chances in at least another five seats too.

Given the difficult times that many of the Core Cities have gone through, we might expect to see gains ahead of national swing for Labour here (or losses restricted).  So the 1/4 on Labour in Walsall North looks sound.and I'd rather be on the 11/10 on Labour in Elmet & Rothwell than the 10/11 on the Conservatives.

Elsewhere, the expectation of Labour outperformance seems fully priced in.  But this is as much about not thinking that there are bargains on the other side of the fence.  The Conservatives are likely to find these seats heavy going.

Nothing much to see here: the North East in 2015

The North East is the smallest of the English regions.  In political terms, it is certainly the least interesting.  25 of the 29 seats are held by Labour, mostly with towering majorities.  Even the most enthusiastic aficionado of politics would struggle to find five interesting constituencies for the next election.

I'm a natural completist, however, so I shall give the North East due attention.

The North East has a very different feel from London and the south east.  It has far lower than average wages and house prices that are a third of those in London.  On the other hand, the region has the most equal distribution of salaries in the country, with fewer workers taking home salaries in the bottom ten per cent of national earnings than anywhere else in the country.  GVA per head is only a little ahead of Wales and behind Northern Ireland, at just three quarters of the UK's national average. Between 2007 and 2012, unemployment rose faster than in any other UK region, to more than 10%, the highest in the country. It carried on rising throughout 2013 as well. From 2006/08 to 2010/12, median household wealth fell by 10% in the North East. No other region saw a fall.

The cuts have hit the North East hard.  When the coalition came to power, 25.7% of all employees were in the public sector.  That percentage has now declined to 21.4%.  This remains the highest percentage outside Northern Ireland.

All this leads me to believe that even if the Conservatives or Lib Dems recover in the polls nationally, they are going to struggle to make any headway in an area that has suffered much of the pain and seen none of the gain.

Here is the ONS's regional profile:

I've presented the seats in the region in a slightly different format:

Because the Conservatives are interested in so few seats in the region, a Conservative battleground is pointless.  Instead I offer you the seats in a standard presentation, with best odds detailed on those few occasions where it is relevant.  We only need to consider four of the constituencies: Stockton South, Middlesbrough South & East Cleveland, Berwick-upon-Tweed and Redcar.

Despite some recent signs of disarray in the ranks of the Middlesbrough Labour party, I firmly expect Labour to retain that seat for the reasons given above.  1/8 looks a very fair price on that prospect.  For the same reason, the Conservatives are going to have a huge struggle to retain Stockton South.  I don't like the look of the 7/2 on the Conservatives at all.  1/4 on Labour is probably decent value - I can easily imagine Labour taking this seat even if they make no overall progress next year.

So that leaves the two Lib Dem seats.  As I have previously noted, uniform national swing does not work where parties have suffered major drops in support. If as I believe the Lib Dems will tally moderate levels of support in untargeted seats, they are currently doing worse than uniform national swing in the seats that they are seriously contesting. The Lib Dems will have it all to do if they are to retain the seats they would hope to keep on a uniform national swing basis.

There's an big underround in Redcar, with Paddy Power and Ladbrokes taking radically different views of the seat.  Labour must be favourites, but Paddy Power's assessment that it's a 1/5 shot looks a bit too short to me.  1/3 looks more like it.  That means that bets on both the Lib Dems and Labour are value.  I prefer the Labour bet because it looks like a value winner rather than a value loser.

In Berwick-upon-Tweed, the incumbent Alan Beith is stepping down.  This should give the Conservatives a spring in their step in their attempts to take the seat.  But Labour unusually (for a Lib Dem/Conservative marginal) retain a vote share in excess of 10%, suggesting that the Lib Dems may have benefited less from tactical voting in the past than in some other seats.  In turn that suggests that there may be fewer 2010 Lib Dem voters who feel betrayed than in other Lib Dem seats.  That said, there was quite a sharp swing from the Lib Dems to the Conservatives at the last election and the Labour vote share has been declining over successive elections.  To my mind the odds are not far wrong.  I'm not betting on this seat at present.

Thursday, 15 May 2014

Deciding the next election (3): the North West in 2015

The third of the chief election-deciding regions is the North West.  Unlike the West Midlands and the East Midlands, the bulk of the seats in this region are held by Labour (mostly with sturdy majorities), but 13 of the top 100 marginals held by the Conservatives are also in this region.

The North West is far from the ripples of the London economy.  It has lower than average wages and higher than average unemployment. However, it has grown relative to the rest of the country over the period from 1997 to 2012, and in 2012 it had one of the fastest growing GVAs per head in the country.

The region is dominated by the two big cities in the south, Manchester and Liverpool. Both are core cities (the eight largest cities of England after London) and their experience of the last few years has been sharply different from the rest of the region. Manchester is a fast-growing city - a 19% population increase between 2001 and 2011, accounting by itself for 20% of the population growth in the region. In 2012, Greater Manchester was the third fastest-growing sub-region in the country. Liverpool also has reversed its longterm population decline in recent years.

Other parts of the region have falling population. Eight of the 17 local authorities with population declines from 2001 to 2011 were in the North West: Barrow-in-Furness; Knowlsey; Sefton; Burnley; Hyndburn; St Helens; Stockport and Blackpool. This is usually a proxy for relative economic decline also.  On the other hand, Warrington has the highest GVA per head in the region, well above the national average.

This dominance of the region extends to the seat count. There are 75 seats in the north west. Of those 75, 42 are in either Merseyside or Greater Manchester. Of those 42, all bar seven are held by Labour (four are Lib Dem and three are Conservative). Of the remaining 33 seats, 19 are held by the Conservatives, two by the Lib Dems and 12 by Labour. In the North West at least, the Conservatives don't so much have a northern problem as a core city problem. I shall look at this again in more detail in a later post. For now, it is enough to note that we should not expect the core city seats and the remainder of the North West constituencies necessarily to behave in the same way.

By way of contrast, the Lib Dems till now have prospered in Manchester and Liverpool: they polled above 20% in every constituency in an unbroken strip from Liverpool Riverside through the southern side of Manchester as far east as Sheffield South East, in many cases forming the main opposition to Labour.

Here's the ONS guide to the region:

And here are the constituency markets:

As usual, I have put together a Conservative battleground (largely because the constituency odds on the Conservative side generally compare a bit more favourably with the odds on the Conservatives getting an overall majority or being largest party, as compared with the Labour counterpart odds) with a selection of other seat markets at the end.  As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion). I have included Stan James's prices, though at the point of writing they have been suspended: no doubt they will return soon enough.

Once again, UKIP are nowhere really to be seen on the list. I find this surprising. I got pennies on with Stan James in Blackpool North & Cleveleys at 250/1. Given that Blackpool is a seaside town that has seen better days (its GVA per head is now just over 60% of the national average) and this constituency is a Conservative/Labour marginal, thus reducing the tally that UKIP would need to reach to win the seat, this seems far too long. We may be a long way from the east coast, but Blackpool shares more in common with Great Yarmouth than donkeys on the beach.  If UKIP do well notionally in this constituency in the European elections, as they very well might, this price should shorten a lot.

The only constituency market in the North West with an underround is Carlisle. I prefer the 4/1 on the Conservatives to the 2/7 on Labour, but both prices are very fair in comparison with the alternatives in other seats.

The standout bet in this list is on the Conservatives at 4/6 in Crewe & Nantwich. This price is way out of line with the price available in other similar constituencies. Unless you think that it's a 6/4 shot that Labour will get a very substantial overall majority, this bet seems marked. In my view the odds should be 2/7 or shorter.

The 6/5 on the Conservatives in Cheadle also looks a sound bet. The Lib Dems have efficiently squeezed the Labour vote here in the past, and enough of it will still be sulking next year to make this a very tough ask.  My only concern about this bet is that with so many Lib Dems in the wider area, they will be have resources to fight to save this seat, Manchester Withington and Hazel Grove, abandoning other seats for now.  But their chances of saving Manchester Withington look forlorn.  Can they save one or both of the seats that they are defending against the Conservatives?  They may need to choose.

So far I have tipped entirely on the blue side of the fence. Relative to other seat prices, the 1/2 on Labour in Bury North looks good. Bury North is one of only two seats in Greater Manchester that the Conservatives hold and the other (Altrincham & Sale West) has a chunky majority with Labour in third. So Labour will be concentrating much of their undoubtedly potent firepower on this seat to try to win it back. As I note above, the Conservatives struggle in Manchester in a way that they have not struggled so much in the less urban parts of the North West. For all that, the 1/2 still looks way too short to me to be worth a bet, given the options available. If Labour take this seat, I expect them to have most seats. Far better to bet at 1.85 on Betfair on that prospect in my view.

Tuesday, 13 May 2014

Deciding the next election (2): the East Midlands in 2015

Moving from the West Midlands to the East Midlands, we stay in marginals-rich territory. 10 of the 100 most marginal Conservative seats can be found among the 46 seats in the East Midlands. Six of the 40 most marginal Labour seats are found here also. This is election-defining territory.

This region, however, possesses much less cultural unity than many other regions. While the centre of the East Midlands, based around the triangle of Nottingham, Derby and Leicester, has an identifiable local culture, the region does not have the same central pull as Yorkshire & The Humber, the East or the South West. This makes generalisations particularly dangerous. It might be safer to treat southern Lincolnshire (and perhaps Northamptonshire) as part of East Anglia and to look at north Nottinghamshire and north Derbyshire, which have Sheffield as their nearest big city, through the same psephological lenses as Yorkshire & The Humber.

In the last few years, economic performance has been mixed. Unemployment is pretty average for the UK but the region has lower than average wages. House price rises have been fairly flat.

Generally across Britain, the story of 1997 to 2011 was of England's GVA growth being led by London and the eight core cities (which include Nottingham), with medium sized cities lagging behind. In 2012, however, the trend reversed with medium sized cities outperforming the rest. Nottingham bucked the trend positively, recording 5.2% growth in 2012 while Leicester was an outlier in the other direction, recording only 0.9% growth.

There is a narrow point and a broader point here. First, seats around Nottingham look like more fertile territory for the Conservatives and more challenging for Labour, all other things being equal, than seats around Leicester. Secondly, it's incredibly dangerous reading too much into national statistics - what is being experienced on the ground in different locations may vary a lot. Leicester and Nottingham are less than 25 miles apart, but their recent economic experiences are quite different.

At the last census, the East Midlands had the highest proportion of cohabiting couple households of any of the regions in England and Wales (11 per cent). The East Midlands is also the only region where the average number of cars and vans per household has remained the same between 2001 and 2011 (it increased elsewhere). Oddly, the traffic increase on major roads over more or less the same period is higher than average. Perhaps East Midlanders have just taken to driving more.

Here's the ONS's handy guide to the region:

Anyway, let's have a look at the seats:

As usual, I have put together a Conservative battleground (largely because the constituency odds on the Conservative side generally compare a bit more favourably with the odds on the Conservatives getting an overall majority or being largest party, as compared with the Labour counterpart odds) with a selection of other seat markets at the end.  As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).

Observations? Well, first of all, the Lib Dems are an irrelevance. On paper they look competitive in two seats (Ashfield and Chesterfield). In practice, Labour should hold both of these seats very comfortably indeed - in this region at least the Lib Dems will be the un-credible shrinking party. The 2/9 on Labour in Ashfield is a sound if undynamic bet.  A 22% return in a year is not to be sneezed at.

I don't think it's yet fully appreciated how far the Lib Dems look set to be driven out of central England.  It's quite plausible that the Lib Dems will have no MPs inside an area bounded by Bermondsey & Old Southwark, Cambridge, Sheffield Hallam, Hazel Grove and Cheltenham.   On a bad night, the Lib Dem free zone could be substantially larger.

UKIP are short-priced in Louth & Horncastle and Boston & Skegness (both Saxon Shore seats), but elsewhere are no shorter than 16/1. There may be good UKIP bets among the longshots here, but if there are I haven't spotted them. I'd steer clear of South Northamptonshire and Daventry though - their local authority has the highest employment rates in the country. It's hard to imagine many residents in fear for their jobs as a result of immigration in those constituencies.

The bulk of the seats are, essentially, the type of Labour/Conservative battlegrounds that once upon a time spread across the entire country. Given Nottingham's good growth record, pocketbook concerns may mean that the Conservatives do relatively better here than elsewhere. The 9/1 on the blues in Nottingham South might be an adventurous longshot. Perhaps more sensibly, the 5/1 on the Conservatives in Sherwood is worth a flutter: a first time incumbent in a seat that adjoins Nottingham might be more resilient than one might expect. It certainly looks a better bet than the 7/2 on the Conservatives in Broxtowe, which is a seat with many of the same features and a substantially identical majority.

On the other side of the fence, Labour at 8/11 in Northampton North still looks good value. There's a huge Lib Dem vote to squeeze. By comparison, Erewash requires a bigger swing, has fewer Lib Dems to squeeze and Labour are 1/2 there.

Monday, 12 May 2014

Deciding the next election (1): the West Midlands in 2015

The West Midlands will be disproportionately important for deciding the next election. It holds 12 of the Conservatives' 100 most marginal seats and 13 of Labour's 100 most marginal seats. So it deserves particularly careful attention.

This is almost exclusively a Conservative/Labour battleground. The Lib Dems hold just two of these seats (both of which look very vulnerable) and are in second place in only two of the 25 seats that feature in the Conservatives' and Labours' 100 most marginal seats. The Lib Dems face the risk of being driven entirely out of the region at the next election. This in turn will make them focus exclusively on the few seats where they retain a chance of winning or remaining relevant. I therefore expect the swings against them in the marginal seats in this area where they are placed third already to be particularly large.

All the polling evidence suggests that 2010 Lib Dems who have so far deserted that party have broken decisively for Labour (and I use a working assumption that a net of a quarter of the Lib Dem vote in Conservative/Labour marginals should be added to the Labour tally). If, however, the 2010 Lib Dems are squeezed further, it is not at all clear to me that they should be expected to break in the same way - after all, they have remained loyal to the Lib Dems till now and the remaining Lib Dems have very different views from those who have already stopped supporting that party. For now I would assume that in seats where the Lib Dems are squeezed beyond normal levels, they will break evenly between Labour and the Conservatives, leaving my working assumption untouched for now. But I will continue to look for straws in the wind as to whether this is correct.

How have the West Midlands been doing in recent years? Not well. The West Midlands saw its share of total UK gross value added (GVA) decline from 8.0% in 2001 to 7.3% in 2011. It is very much being dragged down by Birmingham, which is possibly the worst-performing area of the entire country. I was shocked to find out that Birmingham has the second lowest employment rate of any local authority in the country (57.5%), ahead of only Middlesbrough. No wonder Channel 4 decided to shoot Benefits Street in Birmingham. South Northamptonshire's employment rate is 88.9%. Birmingham is less than 40 miles from Daventry.

Birmingham's unemployment rate at the beginning of the year was 17.3%. Unemployment for the country as a whole is hovering around 7%. In the most recent English Indices of Deprivation (from 2010 and currently being updated), it is the most deprived local authority in England in terms of income deprivation. Overall, it is the ninth most deprived local authority in England, taking into account income, employment, health and disability, education, skills and training, barriers to housing and services, crime and the living environment.

Without much media comment, Birmingham has entered serious decline. I'm shocked both to find this out and at the way in which the media have failed to report this. On this occasion, accusations of metropolitan media bias are fully justified.

Sadly, Birmingham's decline will not directly affect the outcome of the next election. The Conservatives hold none of the nine Birmingham constituencies and only the most marginal of the eight that Labour hold, Birmingham Edgbaston, is potentially relevant to whether or not the Conservatives got an overall majority. If the bookies are correct, however, the Conservatives can get an overall majority without Birmingham Edgbaston.

What of the region as a whole? The West Midlands metropolitan area’s share of the region’s GVA has declined since 2001 compared with the rest of the region. So the rest must be doing relatively better.

Gross disposable household income (GDHI) of West Midlands’ residents was one of the lowest among the English regions, at £14,400 per head in 2011. It ranged from £12,470 per head in Stoke-on-Trent to £17,360 per head in Solihull. Notably, it is the region of England with the most people of working age with no qualifications. The West Midlands also has some of the highest levels of obesity in the country, with Tamworth having been named as the fat capital of Britain.

Here is the ONS's take on the region:

Here are the region's constituencies:

As usual, I have put together a Conservative battleground (largely because the constituency odds on the Conservative side generally compare a bit more favourably with the odds on the Conservatives getting an overall majority or being largest party, as compared with the Labour counterpart odds) with a selection of other seat markets at the end.

As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).  I haven't bothered completing it for the safest Labour seats in this table since this would be a waste of time and effort.

The Lib Dems are largely out of contention. I regard both of their seats in this area as very likely lost and am on Labour in Birmingham Yardley and the Conservatives in Solihull.

UKIP are not yet rated in this area, being perceived as a party of the rural poor rather than the urban poor. Their shortest price in the area is 16/1 in Walsall North and Dudley South. If UKIP do well in the European elections in the West Midlands, as I expect, some of those seat prices will come in sharply. Taking long prices now could make you money after the European elections.

There are three seats with an underround: Cannock Chase, Tamworth and Birmingham Edgbaston. Cannock Chase seems to be trending Conservative over time, and I would place the Tory chances at near evens, so I have topped up my earlier bets on the Conservatives here (compare the prices in Warwick & Leamington, a seat with a very similar battleground). I have already backed the Conservatives at 1/2 a month or so ago. I have now taken the 4/1 on Labour as well for a near-guaranteed 15% return in a year. I can always make my mind up later where I think the true price should lie.

It will not surprise you to find out that I don't fancy the Conservatives' chances in Birmingham. I expect them to do disproportionately badly here and find the 7/2 on them in Birmingham Edgbaston bafflingly short. 7/1 would be closer to the mark. It follows that the 1/3 on Labour is good value, and certainly better value than the 1/3 on Labour in Wolverhampton South West. If you must bet on a Conservative gain, keep an eye on Dudley North - but I expect that the 11/2 will lengthen after the European elections.

I expect the Conservatives to do relatively better outside Birmingham. But the prices at present don't particularly attract me. I expect them to move in my preferred directions, so I shall hold fire for now.

Friday, 9 May 2014

Fighting a separate battle: South West 2015

Unlike most of the country, the main action in the south west is between the Conservatives and the Lib Dems.  Proportionately fewer people voted for Labour in the south west in 2010 in a three horse race than voted for the Conservatives in Scotland in a four horse race. And the Lib Dems have not just supplanted Labour in many areas in second place, they have established local strongholds. They hold all bar one of the seats in Somerset. They hold half the seats in Cornwall (and held them all in 2005). They hold at least one seat in every county in the region.  Labour are irrelevant in large swathes of the region.

So the fight in 2015 will have an entirely different dynamic from that elsewhere in the country. The 5.3 million people in the south west who next year will mainly be choosing between the Conservatives and the Lib Dems might as well be taking part in an entirely different election from the 5.3 million people in Scotland who will mainly be choosing between Labour and the SNP.

Why are there so few Labour voters in the south west?  Are they voting tactically for the Lib Dems or are they simply not present?  I had a look at this in 2011:

As you can see, Labour have polled appallingly in Devon, Somerset and Dorset for at least a generation. There seems to have been some increase in tactical voting, but it's not overwhelming (and do note the correct comment by Mark Senior after the end of the post). As I noted in that post, there comes a point where a Labour tactical voter is indistinguishable from a lifelong Lib Dem voter.

At that time, I concluded that to lose more than 10 seats, the Lib Dems would have to lose long term Lib Dem voters as well as reds in yellow clothing.  Right now it seems as though that is indeed what is happening. Their vote share is so low that uniform national swing doesn't work mathematically, as I have previously noted here:

If so, the Lib Dems are going to struggle to do as well as uniform national swing would imply in the seats that they hold.  One in seven votes for the Lib Dems was cast in the south west region and a quarter of their seats are here.  The Lib Dems are going to have to hunker down.

So what stands out about this region? Here is an ONS chart of the region:

As a region, it's older than the national average and the economy is doing well. It has a low crime rate, low unemployment and a highly educated workforce. But this masks quite a lot of intra-regional variation.

Cornwall is much poorer than the average in the rest of the south west.  Its GVA per head (which is closely related to GDP per head) was £13,848 in 2011, compared with a national average of £21,368 and an equivalent figure for the south west as a whole of £19,093. Swindon's GVA per head is more than twice that of Cornwall and the Scilly Isles.

Let's look at the seats in play:

 I have opted again to put together a Conservative battleground (largely because the constituency odds on the Conservative side generally compare a bit more favourably with the odds on the Conservatives getting an overall majority or being largest party, as compared with the Labour counterpart odds), with the one Lib Dem/Labour marginal at the end.

As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).

Surprisingly, UKIP don't really feature in the markets yet in this region.  In 2010, three out of the four seats where UKIP finished third were in the south west: North Devon, West Devon & Torridge and North Cornwall (the fourth was the Speaker's seat in Buckingham, which arguably doesn't count in the same way). But apart from Camborne & Redruth, they are not as short as 10/1 in any constituency in the region.  I don't list parties on my table at longer than 33/1 for reasons of time and space, but some of the longshots on UKIP in the southwest are worth backing.  Given that Cornwall in particular is one of the poorest areas of the country, I would have thought this was fertile ground for the purple team.  In an area rich with Conservative/Lib Dem marginals, UKIP's target in many constituencies is closer to 35% than to 50% for gaining a seat.  Five out of six of Cornwall's seats are marginals (and the sixth is not exactly safe).

If, as I expect, UKIP do well in the impending European elections and local elections, their prices are likely to come in wherever they perform strongly.  There's every reason to expect that one of the regions where they will perform strongly is the south west.  Even if (like me) you don't believe that UKIP will break the mould next year, there should be some form of trading opportunities in the wake of a good electoral performance later this month.  As always, buy the rumour and sell the fact.

As usual with seats where the Lib Dems feature strongly, no one really knows anything.  Out of 55 seats in the area, 16 have the favourite priced between 1/2 and 2/1.  That has not, however, led to many underrounds, mainly because so far Ladbrokes have not been followed into many of these seats by other bookies.  That will probably change in time, so these seats are worth watching closely.   There's a minuscule underround in Truro & Falmouth and a slightly bigger one in Bristol West.  If you ignore Labour (which seems safe enough to me), there's a minuscule underround in Torbay as well.  In these must-bet seats, my take is that the value is betting against the Lib Dems.  It's my point about uniform national swing not working.  That doesn't yet seem fully factored into the markets.

Elsewhere, the 1/3 on the Conservatives in Mid Dorset & North Poole and the 2/5 on the Conservatives in Somerton & Frome both look like good short-priced bets.  Quite apart from the ethereal majorities that the Lib Dems are defending, both have popular incumbents who are standing down.  These would be tough for the yellow peril to retain in current polling conditions in any case, but look a far greater challenge than those odds would suggest without the present incumbents.

If you want to bet on the Lib Dems, the 1/2 in North Devon is worth considering.  It has been held by Nick Harvey since 1992 and from 1959 to 1979 by Jeremy Thorpe, so it has a long Liberal tradition. Note, however, just how hard Labour has been squeezed in the past: it barely kept its deposit in 2010.  This may be a warning sign that the Lib Dem vote that year contained a higher than usual number of those left-leaning voters who have since been so dissatisfied.

Given Swindon's current prosperity, the 8/11 on the Conservatives retaining Swindon South looks good value.  I also like the look of the 8/11 on the Conservatives in Bristol North West.  Labour would need to come from third and get a swing of 6% to take the seat.  With the Lib Dems in clear second, that looks too challenging to me.

Thursday, 8 May 2014

Telling Yorkshiremen: Yorkshire and Humberside 2015

I'm moving further north now, but having looked first at areas of Labour strength and now at areas of Conservative strength, I'm turning to an area where both parties have considerable strength in different areas: Yorkshire and Humberside. Yorkshire has a particularly strong sense of local identity and this region is as populous as Scotland.

These last few years have not been good times for Yorkshire and Humberside in general. This predates the coalition's rise to power in 2010: between 2006 and 2011 it was the region with the lowest growth in gross disposable household income per head and average wages remain far below the national average. Its economy is not working as well as further south. Yorkshire and The Humber has the lowest labour productivity of all the English regions. It was hit hard by the recession: it had highest increase in unemployment rates and it still has the second highest regional unemployment rate, behind only the north east. Hull had an unemployment rate of 15.3%, second only to Birmingham. House prices in Yorkshire and Humberside were still falling last year.

All this leads me to suspect that the coalition message of recovery is likely to be less well-received in this region than further south.  Any swing away from the Conservatives and Lib Dems will probably be stronger here.

The region has its share of large towns such as Leeds, Bradford, Sheffield and Hull. Leeds and Sheffield are both in the self-selected core cities group that style themselves as England’s eight largest city economies outside London. The region also has vast wildernesses, where rural concerns over transport, petrol and agriculture are much more important.

Here is a handy regional summary from the ONS about the region (click on it to expand it):

Anyway, to the constituencies:

The observant will note that I have included seats in south Humberside that I have already listed when considering the eastern and south eastern coastline.  I am unrepentant: these are interesting seats.

Presenting these is less straightforward because no one party has an interest in every quoted seat.  I have opted again to put together a Conservative battleground (largely because the constituency odds on the Conservative side generally compare a bit more favourably with the odds on the Conservatives getting an overall majority or being largest party, as compared with the Labour counterpart odds), with a selection of other seat markets at the end.  There are quite a lot of seats with no markets as yet.  None of them look interesting.

As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).  I haven't bothered completing it for the safest Labour seats in this table since this would be a waste of time and effort.  I should probably have stopped well before I did.

So, what conclusions do I draw?  First, UKIP are not a major presence in this region yet, if the bookies are to be believed.  The shortest odds available are the 16/1 in Scunthorpe.

Next, the seats aren't that competitive.  Just four seats are projected by the bookies to change hands (five if you include the expectation that Labour will retake Bradford West following George Galloway's by-election victory).  There are, however, a further five seats where the incumbent is favourite at odds of 1/2 or longer.  These need a close look.

There are a few seats where undynamic betting should win you money.  The price on the Lib Dems in Sheffield Hallam is absurdly long at 1/5, driven by animus against Nick Clegg.  A 20% return in a year is not to be sneezed at.  Not quite so absurd, but still long, is the 1/3 on Labour in Pudsey.  They might easily win this seat without even becoming the largest party.

There are three seats with an underround: Brigg & Goole, Bradford West and Cleethorpes.  Such seats are of course must-bet seats.

Bradford West is a seat where local knowledge would be invaluable.  If you don't have that, then by backing both Respect and Labour in appropriate stakes, you can get over a 9% return in a year.  If you must, take out an insurance policy on the Conservatives and the Lib Dems, but that would seem very cautious to me.

I looked at Brigg & Goole and Cleethorpes when I looked at the Saxon Shore seats.  The 7/4 on Labour in Cleethorpes is certainly better value than the 6/4 on that party in Calder Valley or for that matter the 11/8 on Labour in Colne Valley, where they would have to come through from third.

The 10/11 on the Conservatives in Colne Valley looks very fair for that reason. It certainly looks better value than the evens on the Conservatives in Keighley where Labour have a much smaller majority to overcome and are already in second place.

Conversely the Conservatives are bafflingly short at 7/2 in Halifax.  To overcome a very similar majority in Wakefield, the Conservatives are offered at 8/1.  And actually, this is an outside bet worth considering.  The seat has been trending away from Labour over many years.  Unlike most of Yorkshire and Humberside, Wakefield has been doing well in recent years.  Unemployment has been falling recently.  Long-term total GVA in Wakefield increased by 91% between 1997 and 2012, the strongest growth rate in the region and 8% above the UK growth rate. This outperformance has continued in recent years.   The Conservatives' message may be listened to more attentively here than in other marginals in God's own county.

Wednesday, 7 May 2014

Away from the coastal storm: inland eastern and south eastern England in 2015

Yesterday I looked at the coastal seats in the east and south east.  Now to look at the rest of these regions.

Away from the coast, the south east and the east are prosperous regions.  They are the second and fourth highest earning regions by GVA per head behind London (third highest is Scotland). Both regions have high employment, low crime and long life expectancy.  

Here are some basic facts about each region. Click on the pictures to enlarge them further:

Away from the coast, there are no very large towns.  Milton Keynes and Luton are the largest, each with about 250,000 people.  Both of these regions have London as their very large town, of course, and London dominates most of both of these regions economically and culturally.  Much of both of these regions is rural or semi-rural, so transport, fuel prices and agriculture are important subjects in these regions.  But these regions are populous.  Between them they have more than 14 million people, nearly a quarter of the population of the UK and nearly three times the population of Scotland.

Here are the remaining constituencies of interest in the east and south east England:

The Conservatives dominate here, holding all bar seven of these seats.  I have spared you the many Conservative seats which have not yet got markets - they all look utterly safe.

Once again, because these seats are primarily Conservative-held and all have some interest for the Conservatives (other than Oxford East, for which there is not yet a market), I have presented them as a Conservative battleground, ordered by odds on a Conservative victory.  As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the 2010 Lib Dem vote to Labour (the figure is derived by assuming that half will leave the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).

The first thing to note is proportionately how few of these 97 seats are expected by bookies and punters to be of betting interest.  Only five of these seats have odds which suggest that they will change hands, with a sixth a toss-up.  Just two more have favourites with odds of 1/2 or longer.  The coastal seats had far more seats in play out of a much smaller number of seats.

The next thing to note is that UKIP are much less relevant in these seats, if the bookies are to be believed.  In the coastal seats, UKIP were rated 10/1 or less in seven out of 51 seats.  Inland, they are as short as 10/1 in only one seat, Basingstoke, and that looks like an utterly lousy bet.

But there are seats to bet on.  There's an underround in Bedford, where you can get 5/2 on the Conservatives and 4/7 on Labour.  Both of these look like value to me - I see this as a 2/1 plays 1/2 contest.  Note that the Lib Dems remain strong in Bedford, where they retained the Mayoralty in 2011, so they may well retain more of their vote next year than we will see elsewhere in seats where they placed third last time around.

Labour look too long in Northampton North at 8/11.  The Lib Dems polled well here last time and that vote is likely to break substantially in Labour's favour.  Labour should be at least as short as the 4/7 in Bedford and Ipswich.  I'm on this bet.

There are two seats where Labour are looking to win the seat from third: Cambridge where they are evens and Watford where they are 4/5 odds on favourite.  Parties rarely win seats from third and Labour therefore do not look good value in these seats.  By a process of elimination that means that the Lib Dems are good value in Cambridge at evens.  I'm also drawn to the 9/4 on the Conservatives in Watford.  This is a prosperous part of the country that is on the edge of the London economic phenomenon with a first time incumbent and where the Lib Dems remain strong.

For a longshot, you might consider the Greens in Norwich South at 25/1.  There's no particular reason to assume that deserting 2010 Lib Dems will break for Labour rather than the Greens in this seat and the seat has been trending away from Labour for a generation.  While the Greens seem to have faded a bit in Norwich in the last few years, this isn't a 25/1 shot.  When the Greens won Brighton Pavilion in 2010, they had a no better starting position from the 2005 election result.  Note that Norwich South is a student constituency.

6/1 on Labour in Reading East would be a respectable bet.  With a large 2010 Lib Dem vote to capitalise on, the challenge is less daunting than the nominal majority would suggest.  I haven't made this bet though because Labour has seemed to struggle to have its message heard in these prosperous areas of southern England outside London.  

Tuesday, 6 May 2014

The Saxon Shore: the east and south east coast in 2015

In the third century AD, the Romans arranged coastal defences around the east coast from the Humber to the Solent to deal with unwelcome immigrants in a system called the Saxon Shore.  In 2015, many of the constituencies on the same coast will be dominated by the same concern at the ballot box.

The east of England and south eastern England are among the most prosperous parts of the country.  But this prosperity has very often not reached the coastline.  While there are some very attractive coastal towns in this area, there are others that have suffered decades of decline.  A 2007 Department for Communities and Local Government Select Committee report found that many coastal towns shared a number of common characteristics, including: physical isolation, high deprivation levels, inward migration of older people, large numbers of people passing through without settling, outward migration of young people and poor-quality housing.  A 2008 study by DCLG similarly concluded that, taking account of a range of socio-economic indicators, the larger English seaside towns are, on average, more disadvantaged than England as a whole, with high unemployment.

This sharp difference between the eastern and south eastern coast and the inland south east and east means that it is worth looking at these separately.  They seem to be behaving quite differently politically - at least, bookies and punters think so.  So let's take things littorally:

These are all the 51 coastal constituencies from the Humber to the Solent (including the Isle of Wight).

Because these seats are primarily Conservative-held and all have some interest for the Conservatives, I have presented them as a Conservative battleground, ordered by odds on a Conservative victory.  As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the Lib Dem vote to Labour (the figure is derived by assuming that half will desert the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).

The first thing to note is just how many seats are in play among these.  Five are expected by the bookies to change hands, with a sixth with the incumbent odds against (though still favourite) and a seventh rated a toss-up.  In a further seven seats, the incumbent is rated 1/2 or longer to hold their seat.  Without us really noticing, a whole new electoral front has opened up.

The next thing to note is how multi-party many of these seats are.  The Greens and the Lib Dems both hold seats on this coastline (indeed, the Lib Dems hold more seats than Labour do, with Labour's three seats being found at the opposite extremes of this coastline).  UKIP are quoted at best odds of 10/1 or less in seven of these seats.

This coastal stretch seems to be an area of particular UKIP strength.  Eight out of ten of the constituencies notionally won by UKIP last year were on the Saxon Shore:

I have already looked at the rise of UKIP in the eastern side of England:

As I noted on that thread, it seems that Labour has so far been unable to take all of the advantage you would expect them to take of conditions that would ordinarily favour them, and UKIP has stepped into the vacuum.

There is some support for that idea from the Survation constituency polls conducted for UKIP.  Six out of the eight conducted so far have been conducted in Saxon Shore constituencies: Eastleigh, Bognor Regis & Littlehampton, Folkestone & Hythe, South Thanet, Great Yarmouth and Great Grimsby:

UKIP were shown as winning Eastleigh and strong in the others.  When tested in South Thanet their vote was recorded as being not particularly likely to switch to the Conservatives and unimpressed at the idea of letting in Ed Miliband.  Whether or not you believe the vote shares recorded to be accurate, I see no reason to disbelieve these findings about the nature of UKIP's vote.  Their supporters aren't Tories on holiday but dissidents against both main parties.

For me the most striking of these constituency polls was Bognor Regis & Littlehampton.  It's historically been an uber-safe Conservative seat with the Lib Dems comfortably in second.  Yet UKIP were recorded as having more than a quarter of the polled support.  This is a huge turn-around.

We will no doubt see more of these constituency polls and I expect most of them will continue to concentrate on this coastline.

What's happening in the betting markets?  Well, the bookies seem confused.  That's great news for gamblers, because there are no fewer than four seats with underrounds: Brigg & Goole, South Basildon & East Thurrock, Cleethorpes and Brighton Kemptown.  There's guaranteed money to be made in these seats just by backing the different horses at appropriate stakes.  The underround in Brigg & Goole offers you a guaranteed 9% return if you so wish.

So these are four must-bet constituencies.  Where's the value?

No one knows for certain and anyone who claims otherwise is either a fool or lying.  So what follows is guesswork.

The first thing to note is that in all four seats the Lib Dems had a reasonable but not amazing showing (13.4% in South Basildon & East Thurrock, 15% in Brigg & Goole, 18% in the other two seats).  So Labour can hope for a bit of a lift there.

UKIP are as short as 20/1 in both Basildon & East Thurrock and in Cleethorpes.  So they may do better than average there.  If they draw their vote the same way in those constituencies that they draw their vote nationally (we don't know that, but in the absence of other evidence we should tentatively assume that), that also should help Labour.

I believe that the price on Labour is where most of the value probably lies in all four of these constituencies, but especially Basildon & East Thurrock and Cleethorpes.  Indeed, if the Survation polls are remotely correct, Labour may well benefit in many constituencies by the upheaval between the Conservatives and UKIP in these seats.  But I'm far from sure and we may get a lot more helpful data after the impending batch of elections.

The same effects, coupled with the apparent resilience of UKIP's vote as described in South Thanet, lead me to infer that Labour are value at 6/4 in Great Yarmouth.  In adjoining Waveney I tipped UKIP at 50/1.  That price is no longer available but you can still get 33/1.  This still looks like a good price, given the types of movements being described in the Alan Bown polls.

For a longshot, I've had a small stake on Labour in Rochford & Southend East at 20/1.  There's a large 2010 Lib Dem vote to squeeze here, and if UKIP is going to be as disruptive as the Alan Bown polling suggests, Labour could find itself getting home on a relatively small vote share.  I'm not sure what odds it should be, but it doesn't look like a 20/1 shot - maybe 8/1.

One further thought: so far the Conservatives have sought to appeal to would-be UKIP voters on a "stop Ed Miliband" basis, so far unsuccessfully.  Will they start to seek to appeal to more progressive voters on a "stop UKIP" basis in these constituencies?  Polling shows that UKIP is the party which has the largest number of voters dislike it.  So it's something they might like to try in Eastleigh, for example, if they're cheeky.  The recent Alan Bown poll in that constituency showing UKIP first and the Conservatives second makes that a plausible pitch.

Finally, something completely different: Brighton Pavilion.  You can back both the Greens and Labour at evens.  I haven't placed a bet on this constituency (local knowledge is vital in this case and I don't think I have enough), but despite the Green council's travails, I'd prefer to back the Greens if I were forced to choose.  There is no particular reason to assume that 2010 Lib Dems will break for Labour rather than the Greens and it is likely that Caroline Lucas will get some form of first term incumbency bounce.

Monday, 5 May 2014

How red our are valleys? Wales 2015

After London yesterday, I turn now to Wales.  And Wales's experience of the last few years has been very different from London's.

Wales is the poorest part of Britain.  Its GVA per head is £15,401, considerably less than half that of London and less than three quarters of the UK average.  That's in the same ballpark as Greece, Portugal and Estonia.  It saw the steepest rise in unemployment following the 2008 crash; however, last year unemployment dropped more steeply in Wales than almost anywhere else and the unemployment rate is now below the national average.  But prospects for Wales look poorer than elsewhere in the UK:

So the impact of the economic recovery on voting intentions may be less than in more prosperous areas.

Wales has older demographics than most of the UK:

In 2012 the median age was two years higher than the UK average, and people aged 65 and over made up 19.1% of the population, compared with 17.0% for the UK.  This may make voters more socially conservative - and more likely to vote.

Wales, of course, has devolved government.  This is, as it has been since its inception, Labour controlled.  It is responsible for, among other things, education, health and the environment, so many of the subjects that are the subject of political debate for the Westminster election in England are of less pressing importance in Wales.

Also, we should not forget the presence of a fourth party with Westminster MPs, Plaid Cymru.  This gives rise to special considerations in some constituencies, as considered below.

The most recent Welsh opinion poll dates from the middle of last month:

Labour: 45%
Conservative: 24%
LibDems: 7%
Plaid Cymru: 11%
UKIP: 10%
Others: 3%

This represents a 5.5% swing from the Conservatives to Labour.  The Lib Dems polled 20% at the last election, so have apparently lost two thirds of their vote in Wales, as in Scotland.

To the seats in play:

As previously, the third column reflects the effect on the majority in Labour/Conservative marginals that would be had by allocating a quarter of the Lib Dem vote to Labour (the figure is derived by assuming that half will desert the Lib Dems and that these will break for Labour rather than the Conservatives in a 3:1 proportion).

As you can see, I've taken a very broad view of what constitutes a seat in play, and still the competitive seats are few indeed. Only two seats are predicted by the bookies to change hands: Cardiff Central and Cardiff North.  Of the rest, only Arfon looks close to a toss-up. 

The one seat that's a mandatory bet is Cardiff Central.  There's an underround and simply by betting in properly weighted stakes on both Labour and the Lib Dems would yield a 7% return - not to be sneezed at.  But which do I prefer?

When I looked at this a couple of weeks ago, I liked the Lib Dems at 2/1. But I hadn't appreciated just how badly they've slumped in Wales as well as Scotland. Uniform national swing is meaningless at such levels. 

Unlike Scotland, the Lib Dem vote was very inefficiently spread in Wales. From 40 seats they picked up just three with over 20% of the vote. So they have a lot more wasted votes to lose in irrelevant constituencies. They can reasonably try to hold all three of their current seats through personal votes, at the expense of watching their vote crash almost everywhere else.  

But that vote won't crash completely. The Lib Dems tallied 18.2% in the 37 constituencies that they didn't take in 2010.  I doubt it will drop below 3% in almost any constituency and in some it will be considerably higher (especially as some areas of Wales which do not have a Lib Dem MP, such as Montgomeryshire, have strong Liberal traditions).   An average in these seats closer to 5% therefore looks more likely.

To hold their seats, the Lib Dem Welsh national vote share will therefore need to increase a fair bit.  2/1 on the Lib Dems in Cardiff Central looks right, which means that the 4/6 on Labour is the value. Good job there's an underround. 

If the Lib Dem vote is indeed going to crash in Wales, that means that the potential swing to Labour could be greater in those seats where the Lib Dems are not seriously campaigning, assuming that these 2010 Lib Dems continue to break in the way I have assumed.  I would therefore be very careful about betting on the Conservatives in any Welsh constituency.  But I don't see any clear bets on Labour here either.  The swings required do not make the odds offered attractive, with the possible exception of the 10/1 longshot on Labour in Clwyd West.

I suppose the 1/12 on Labour in Newport East beats what you could get in the bank.  It seems unlikely that the Lib Dems are going to be in genuine contention in seats like this in 2015, given their national polling.  It's not a bet I shall be making though.

How will Plaid Cymru fare?  Unlike their Scottish counterparts, they have languished in the Welsh national polls.  But Plaid Cymru has historically been a regional party and like the Lib Dems its national vote share will not help us predict its seat tally.  It is a left wing party, so no assumption can be made that 2010 Lib Dem voters who abandon that ship will break particularly for Labour rather than Plaid when both are in contention.  Indeed, in Ceredigion I would expect Plaid Cymru to benefit from former Lib Dem disaffection much as Labour will do in seats where Labour is second placed.  The 2/1 in that seat looks a bit mean on Plaid Cymru, however, given the size of the swing required.

Nor do I fancy a bet in either Arfon or Ynys Mon, where Labour and Plaid Cymru are tussling it out.  While Labour have benefited from a national swing in their favour against Plaid since 2010, these seats are very individual.  Without local knowledge, I won't be betting on either.