Archive for the 'Coalition' Category


Marf on the LAB leadership

Friday, May 15th, 2015

bitesthedust (1)


The likely final polls with when we can expect them

Monday, May 4th, 2015

If I’ve made a mistake or you have better info then please email me

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble


David Herdson: The election remains far from a foregone conclusion

Saturday, May 2nd, 2015

Polling+station+cyclist (1)

Groupthink can be a most dangerous thing. To listen to many commentators, analysts and indeed many politicians, such is the consensus that you might believe that the election was all over and the result in; that result being Con and Lab roughly level, an SNP landslide in Scotland, the Lib Dems parliamentary party halved and UKIP and the Greens failing to register meaningfully. So it might be but we should at least entertain other possibilities.

Why has such a consensus developed? Much of it is surely due to the monotony of the YouGov and Populus poll results, the most frequent pollsters and also the most stable in output. The result is that that each additional poll reinforces confidence that they must be right and hence gives an impression of where the election’s heading. In a similar vein, if everyone interprets those polls in the same way then ‘most likely’ tends towards the apparently inevitable simply by reaffirmation.

Yet elections are not that predictable. Late swings can occur but in current circumstances, more likely is that opinion is being misread – or more accurately, the wrong polls are being dismissed. So what could go wrong?

The Tories do much better than anticipated

The assumption is that it’s currently very tight in the national polls with a point or so either way. Maybe so but there have been several polls in April that generated very different results. Mori, Opinium, ICM, ComRes (phone), Survation and Ashcroft have all reported Tory leads of 4% or more at some point. You would not expect that many polls so top-side for the Cons if the actual lead was only 1% just on natural fluctuation. Sampling might still be the explanation (there is always a slim chance of the improbable), or the methodologies may be out but there is a third option: that they’re right. Just because something appears to be rogue doesn’t mean it is – particularly if several pollsters find it.

Labour does much better than expected

Unlike the Tories, Labour hasn’t chalked up any six-point leads recently. Unlike the Tories, it doesn’t need to. Even allowing for a cataclysm reducing Scottish Labour to single figures (and everything over the last six months points to that being the very best they can hope for), the online polls suggest a swing to Labour in England of at least 4%. On top of that, as Mike noted yesterday, there’s some evidence that Labour is doing disproportionately well at picking up ex-LD votes in the marginals. While seat-specific polling is more patchy, single-constituency polls are a new discipline and have not always proven particularly accurate in the past. A Labour lead in England (which is what a 2% GB-wide lead would imply) should see Labour some way over 300 seats and probably capable of forming a government with the Lib Dems. Unless …

The Lib Dems do catastrophically

Everyone expects the Lib Dems to do badly. More than four years of election results and polling point to it and nothing’s changed to reassess that expectation. Yes, they did hold the one parliamentary by-election they defended but only by the quirk of a particularly split opposition. Their trump card has always been assumed to be incumbency; their cockroach-like survival abilities, as Tim Farron once put it. However, they look doomed in most of their Scottish seats, will struggle mightily where facing Labour and could be faring just as badly in the South West. If so, the ‘third force’ in British politics could easily be looking at their smallest Westminster presence since the SDP-Liberal Alliance formed in 1981 (as well as having been relegated to fourth).

Where does any of that leave us? What seems certain is that there will be no outright winner; the SNP sweep seems to make that all but certain given that there will likely be a minimum of 90 non-Con/Lab MPs and possibly quite a few more. What we shouldn’t assume however is that the SNP will necessarily have the power they’re anticipating (though if they find they don’t, another independence referendum becomes almost inevitable within five years and probably three).

David Herdson

p.s. Mike put up a thread earlier this week based on Sporting Index’s market on big-name casualties which got me thinking about this year’s Portillo Moment: the defining image of the election. Not all elections have them and they’re far from all the same. In 1992, it was the government holding Basildon that became iconic; in 1983 it was a member of the opposition – Tony Benn – losing that marked the failure of Labour in general and the far left in particular.

This time, in line with the comments in the main thread, many eyes will be turned to Sheffield Hallam if the Lib Dems do have a desperate night, or Morley & Outwood if the Tories perform unexpectedly well. However the most extraordinary events are likely to take place in Scotland and while Jim Murphy’s own East Renfrewshire seat should be a contender for the Portillo Moment, as should the Kirkaldy & Cowdenbeath one of former PM Gordon Brown, neither feels quite right to me. If anything, the prominence of their former MPs detracts from the phenomenon their change of allegiance would represent. More appropriate might be Glasgow North East, anonymously symbolic of an utterly dominant Scottish Labour machine. Labour has never failed to win less than 53% in it or its predecessors since before WWII and has frequently polled two-thirds or more. That, and a 7% lead in a January constituency poll, explains why the SNP are still odds-against at 13/8. But if Sturgeon’s army can take it, nothing would symbolize their national triumph more.


Both Farage and Clegg set to lose according to new Ashcroft polling

Wednesday, April 29th, 2015

Clegg trailing in Hallam even though a third of 2010 CON voters not supporting him

Farage’s battle remains tight

South Swindon – LAB CON target number 43


New Ashcroft Scottish seat polling finds SNP gains from LAB, LD & CON

Friday, April 17th, 2015

Charlie Kennedy’s seat goes to SNP

Ming’s seat an SNP gain


UKIP seats drop to their lowest level yet on the Commons seats spread markets

Tuesday, April 14th, 2015

With all the focus being on LAB and CON there’s not been that much attention to the decline of UKIP in the polls. Although this is not universal quite a number of surveys from firms that used to have the best numbers for Farage’s party now have big declines.

This is now being reflected on the spread markets where there have been a lot of UKIP sellers with very few buyers about.

Whether the party can start to stamp its imprint on this election is hard to say. So far, in spite of its guaranteed coverage status, it has yet to make an impact.

I’m being told that the reported poor polls for the party from Thanet South and Rochester were not the only ones. Apparently there are quite a few on their key targets. The fact that they have not been published speaks volumes.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble


Looks like LAB is preparing for LD coalition negotiations

Monday, April 13th, 2015


My favourite GE15 spread bet: That there’ll be a CON margin on seats of fewer than 12

Friday, April 10th, 2015

SPIN trend

Betting sentiment remains solidly with CON double digit seat margin

The chart shows the trend on the Sporting Index commons seats spread market over the past few months and as can be seen the Tories have opened a double digit lead which has remained for some weeks. Even yesterday’s poor polls for the Tories did not have much impact.

As well as the straight total seats betting as seen above there’s another market I like – CON supremacy over LAB. The terms are

    A prediction on the total number of seats won by one party versus another party at the UK General Election. Note: This market can have a negative result and is a supremacy market where the favourite is listed first and the underdog second.”

My bet is a sell at 12 seats which has now edged down to 9. What this means is that I’m a winner if the eventual margin is fewer than 12 seats – the actual level of my return dependent on the final gap. So if LAB ended up 20 ahead I’d make 20+12=32 multiplied by the stake level.

If my prediction is wrong and the Tories, shall we say, end up 20 ahead then my losses would be 20 minus the bet level (12) multiplied by the stake level. So at, say, £20 a unit (which is not what I’ve bet) I’d lose £160.

I’ve gone in to this market rather than the total seats because the gap is more interesting and a big unknown is how many seats the non-main two parties will end up with.

With spread betting the more you are right the more you win and the more that you are wrong the more you lose. So the most important first step is to work out the down-side risk. What’s the most you can lose and in this case I can’t envisage the circumstances in which the Tories are more than 30 seats ahead – in which case my loss would be 18 units.

There can be quite heavy potential losses with spread betting if you call it wrong as I found in 2001 when I made the assumption on election day that the reported very low turnout would favour CON more than LAB. That remains my biggest ever gambling loss.

My latest bet is hopefully more than covered by my debate bet on Sturgeon a week ago when I came away with a 35 unit win on which leader would poll best.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble