Clegg’s ‘Stunning Victory’ would cost his party 30% of its MPs

by Jackie_South on March 3, 2013

eastleigh result iconsOn Friday, Nick Clegg hailed his party’s close victory in a constituency where they held every council seat as a “stunning victory”.

I thought I would examine the claim.  Of course, extrapolating a general election result from a by-election is a poor predictor of the likely outcome.  But given Clegg’s claim, such an extrapolation is a fair way of judging the veracity of the statement.

So, if you apply the changes in support for the Lib Dems, Conservatives, Labour and UKIP to every constituency, what is the outcome?

Well, clearly as the Lib Dems’ support fell by more than the other parties (by 14.4%) they would clearly not be winning any new seats.

However, as the Conservatives’ support fell by almost as much (13.9%) and they are second in most of the Lib Dem seats, there are not many of their seats that would turn blue.  Our calculations show that just one, the ultra-marginal Solihull (with a majority of 175), would go Conservative.

Although Labour’s vote stayed pretty much as it was, at a fairly pitiful 9.8%, this does mean that they make up ground on the Liberal Democrats, with a net swing towards them of 7.3%.  This would be enough to secure a victory in eleven seats: Birmingham Yardley, Bradford East, Brent Central, Burnley, Cardiff Central, East Dunbartonshire, Edinburgh West, Hornsey and Wood Green, Manchester Withington, Norwich South and Redcar.

Of course, the most spectacular change in support was for UKIP, gaining 24.2% of the vote.  Of course, UKIP’s vote is thinly spread and it is difficult for them to win anywhere.  But, the size of swing in Eastleigh, 19.2%, would be enough for UKIP to win five seats off the Liberal Democrats: Cambridge and St Ives in England, and Argyll & Bute, Gordon and West Aberdeenshire & Kincardine in Scotland.

The Scottish results are skewed by the four-party system (excluding UKIP) there – we have assumed that the SNP vote remains the same as in 2010 in the model (we have used the same assumption for Plaid Cymru in Wales).  In fact, the Liberal Democrats would fall from first to fourth place in Argyll and Bute, behind not only UKIP but also Labour and the SNP.  Similarly, they would be third, behnid both UKIP and the SNP, in Gordon.

Cambridge is an interesting potential British four-way contest, although this is probably a constituency where UKIP are unlikely to make as much headway as in Eastleigh.

Far more interesting is St Ives, where it is entirely possible that 2015 could be a three-way contest between UKIP, the Lib Dems and the Conservatives.

So, in total the Liberal Democrats would have lost seventeen seats on the Eastleigh result.  This is actually a lot better than the current national polls predict (which have them losing 30 seats).  This demonstrates that there is some resilience in Lib Dem held constituencies and that being in Coalition with their main competition nationally does perhaps protect them from total collapse in 2015.

But it is a strange form of ‘stunning victory’  that sees you lose 30 percent of your MPs.

Below, you can see the full results of the modelling, constituency by constituency.  For each party, the upper lighter bar is the 2010 result, the lower one the projection based on the Eastleigh result.  Any seats changing hands are indicated by the colour of the constituency title.

page 1 page 2 page 3 page 4 page 5 page 6 page 7 page 8 page 9 page 10 page 11 page 12 page 13 page 14 page 15


{ 8 comments… read them below or add one }

Mike Killingworth March 4, 2013 at 8:40 am

So, Jackie, which seat should Nigel the Perilous take on next time? I’d say St Ives…


john b March 4, 2013 at 8:42 am

I’m not sure about your modelling of the UKIP swing here. Might be more sensible to assume it’s a multiple of 2010 vote rather than an addition, so you don’t end up with impossibilities like UKIP MPs in Scotland, or inner-city seats like Southwark where drooling golf club bigots are rather at a low.

Interesting exercise, though.


Doug Daniel March 4, 2013 at 9:47 am

As an Aberdonian born and bred, I would just like to say that this is absolutely hilarious. UKIP winning Gordon and West Aberdeenshire? “Guid een!” as we say up here.

This article betrays a complete lack of awareness of the realities of Scottish politics. These are constituencies which contributed towards the SNP’s domination of the North-East in 2011, where their share of the vote was so high that they even managed to get a list MSP elected despite winning every single constituency seat. This is the area where the SNP broke the Additional Member System that is supposed to prevent a single-party majority government. And yet you think UKIP are going to come in and snatch a couple of Lib Dem seats while the SNP vote stagnates? That’s just nae happening.

Scottish voting habits have become quite sophisticated since we got a devolved parliament. People think nothing of voting Labour for Westminster (to keep the Tories out) and then SNP in Holyrood (to keep Labour out). The idea that a country which barely even elects ONE Tory MP is suddenly going to elect FIVE MPs who are the very embodiment of everything we hate about the Tory party is just daft.

You simply cannot extrapolate the results of a by-election in the bowels of South-East England and apply them to Scotland. If anything can be learned from Eastleigh from a Scottish perspective, it’s that Scotland and the South-East of England really are two very, very different places that are not suited to being governed by the same parliament. Whatever happens elsewhere in the UK, like the Romans, UKIP’s march will come to an abrupt halt at Hadrian’s Wall…


Cartoonist March 4, 2013 at 11:47 am

Whilst the author admits this is an unlikely scenario, it’s still an interesting read, I like it.

Of course it can only be hypothetical, less scientific without the polictical context of the suspicious constituent. Trying to predict an outcome of a general election requires more input from the electorate and something which has worked for America is the use of election stock markets. Here the electorate puts their money where their mouth is and the result is more accurate than any opinion poll. James Surowiecki writes about this stuff in his book, Wisdom of Crowds if you’re interested enough to part with six pounds.

Would it work here? Who knows? Hence this blog presents a delightful snapshot based on one constituency’s exposure to all the political posturing of Easter bunnies on heat.

Yet, if the question of a political party’s popularity mattered, I doubt David Cameron would make it past the guillotine without nervously scratching his neck whilst Mr Osborne hides behind locked doors rapidly stuffing his suitcases with freshly printed currency.
What Eastleigh has shown is what’s possible with a vacancy window of a few weeks. No doubt the Lib Dems knew this and bit off the hand of the constituent. It would have been more interesting to see how many more seats would have been lost in double the time.

BTW don’t you think someone should adopt the phrase, “You kip if you want to. This party’s not for sleeping.”


Charlie_East_West March 4, 2013 at 12:33 pm

Jackie – I have no idea how you can put together such brilliant research.
Borderline genius.

Obsequiously Yours,


Holebender March 4, 2013 at 12:41 pm

How can you make any assumptions about the nationalist vote in Scotland and Wales on the basis of a byelection in England? Your model may have some meaning for English constituencies, but it is crude and, frankly, insulting in its assumptions about not-England.

UKIP are absolute nobodies in Scotland, and that isn’t going to change as long as Scotland so plainly rejects all Tory-style politicians. Your caveat re Scotland is far too tame and shows a poor understanding of Scottish politics.


George_East March 4, 2013 at 12:52 pm

Er, its a model to test a hypothesis (‘stunning victory’), which is expressly said to be non-predictive. The UNS model is the same as the one used by UKpollingreport and Electoralcalculus and others. The assumptions are stated. In the old days had Peter Show been jumping up and down on a BBC by-election special (‘a little bit of fun, a little bit of fun…’), his results would have been the same.

You seem to be getting wound up over nothing in particular. No one is suggesting UKIP are about to sweep Scotland.


Nick Evans March 6, 2013 at 6:24 pm

Noted. Although the idea of UKIP catching up with Plaid in a constituency where nearly half the population speak Welsh is nice for a giggle too.


Leave a Comment

{ 2 trackbacks }

Previous post:

Next post: