So, we can be pretty confident now that there is a genuine surge in Lib Dems support. Apart from how long it lasts, and the affect on the narrative of the election, the other unknown is exactly who has shifted, and perhaps more importantly where. We will get a better idea of how things are moving when we get some polling of marginal seats, and when we get an updated version of YouGov’s weighted regional breaks. In the meantime, we can draw some very tentative conclusions on the info we have so far.

If we compare the crossbreaks in the two post-debate YouGov polls so far with the averages of the crossbreaks in the three YouGov polls before the Lib Dem manifesto launch, and we compare the ComRes poll with the averages of the crossbreaks in their polls from before the debate, we can get some idea of which groups have swung the most. We need to remember that even combined together these are small sample sizes and they are not internally weighted, so at best this can only give us a broad brush idea of what is going on, but there are some consistent trends.

We can be fairly confident in saying that young people have swung the most strongly towards the Liberal Democrats. In both YouGov’s post-debate polls they have had the Lib Dems in the 40s amongst under-35s, compared to the mid-20s prior to the debate, an average increase of 17 points. ComRes have a similar pattern. The Lib Dem boost amongst under 35s seems to be at least twice that amongst all older voters.

Secondly women may have swung to the Lib Dems more than men (more specifically, there’s been a swing from Conservative to Lib Dem amongst both sexes, but Labour’s vote has fallen much more amongst women). The contrast here is not as large as is it with age though, so I’m less confident this is a genuine pattern.

On social class there is no obvious pattern, YouGov seem to be showing a larger swing amongst C2DEs, but it is the other way round with ComRes. On regional breaks the evidence is also unclear, though both ComRes and YouGov seem to have the Lib Dems making the most progress in the North, and YouGov’s figures are striking for the almost complete lack of a Lib Dem boost in Scotland. On other hand, late respondents to YouGov’s new Scottish poll in the Scotland on Sunday today (whose fieldwork straddled the debate) did show swing towards the Lib Dems.


310 Responses to “Who is switching to the Liberal Democrats?”

1 5 6 7
  1. @FIDO
    Plugging the figures into Electoral Calculus does show virtually the whole SW go LibDem.

    They give Con 251, Lab 213, Lib 153 Oth 15 + NI which is a bit different to here where we get 239-247-132-14-18

    Of course, as several people have said, the models may have to be reworked now.

    Report comment

  2. Fido – I live in East Devon. Lib Dem are very strong in the west country and could end up being the biggest party in the south west region ; however lots of rich retired / farmers in East Devon / Dorset so I expect they are likely to stay Conservative no matter whats going on elsewhere.

    However I would like someone to give approximate times of the first 10 key marginals between Lab/Con/Lib Dem where we can extrapolate the results of the election. I know Sunderland South normally returns first but this is a safe Labour seat. I think the Birmingham Edgbastion and Portsmouth North will be the first key ones. Is that right?

    Report comment

  3. @Colin
    @ All

    “The Lib Dems will probably get at least 30% at the GE.” Which might leave 62% for Lab + Con-in which case Cons need min. 33/29 for largest party.( UNS)
    The precise LibDem outcome is absolutely critical.”

    Assuming a UNS is applied across the UK and therefore discounting other factors such as marginal bonus/deficits, incumbancy and tactical voting.

    You are incorrect on all counts. Try using the Swingometer on this site for various combinations.

    First put in the size of the Libdem % (22 to 36 it doesn’t matter). Then put in your combinations for Con % and Lab %.

    There is ALWAYS a Lab largest party unless the Con % is a minimum of 7% greater than the Lab %. This occurs even if the Con% is down to as low as 27%, provided the Labour % is only 20%.

    Thus the critical factor on which has the largest party Lab or Con is that there must be a minimum 7 point lead by the Con over Lab.

    It is a fact that the inbalance in the distribution of the constituencies is currently equivalent to about a 7% points advantage to Labour. This occurs with PFTP and would probably STILL occur with other forms of PR, unless as well introducing PR you also corrected the constituencies inbalance in favour of Lab.

    Why else do you think that GB has indicated a willingness for PR, should he not have a majority. GB probably hopes to stave off any redress of the constituency inbalance until the next Boundary Commission is due (2015?). More than enough time to get an election on the current inbalanced boundaries in 1 – 5 years before then.

    Unless either Con get over 38% and at the same are 7% ahead of Labour, then on UNS they should not get a majority.

    For LibDems the task is even greater. Unless they can get 38% and a minimum lead over Lab of about 13% and at the same time a lead of 8% over Con, then LibDems will NOT even be the LARGEST party, let alone have a majority. (Although a greater than 8% lead over the Con can slightly affect that 13% required lead over Lab since the Con/Lab ratio will now be less than 7%.)

    If the LibDems hope to be a future govt it is essential that the constituencies inbalance is addressed at the same time as any change to a PR system. Otherwise the change to a PR system will bring the possibility of a LibDem govt or a LibDem lead govt any closer.

    Such a radical change to the constituencies could not be carried out quickly. Even if groups of constituencies were bundled together to be elected by PR, the inbalance gives Lab the same 7% advantage.

    Before the LibDems sell their souls to the devil, they had better be aware of the price! Remember in a PR type system, tactical voting may not be so critical in marginal seats.

    Report comment

  4. @All

    Para should read as below: Sorry forgot the NOT

    If the LibDems hope to be a future govt it is essential that the constituencies inbalance is addressed at the same time as any change to a PR system. Otherwise the change to a PR system will NOT bring the possibility of a LibDem govt or a LibDem lead govt any closer.

    Report comment

  5. FrankG

    With the greatest respect, that is a load of ill-informed (or perhaps wishful) codswallop.

    With the polls in the current state of flux it is impossible to make any sane projection of seats.

    Even after the polls settle (defined as several days with no movements outside MOE) – and that may not happen this side of 6th May – we cannot know what the new seat paradigm is for any given share of votes.

    It is as if we had a three dimensional house of cards with sepaarte stacks of Lab / Con seats, Lab / LD seats and Con / LD seats. They have all been tossed in the air and shuffled together, and wereally do not know where they have landed.

    I have never set great store by UNS or the assumed Lab bias in the boundaries. But I’ll now go one step further.

    All swing calculators are now worthless until AFTER the GE.

    Report comment

  6. I have been a lifelong labour supporter but the last 5 years have taken the biscuit, therefore there is no other alternative but to vote for the liberal democrats…lets see if these idiots can do any better, there nothing to lose but our dignity.

    Report comment

  7. FrankG – although they’ve dropped it as an official policy, the Lib Dems ultimate preference would surely be for multi-member STV voting. Since in most models, both the Conservatives and the Liberals do better under that system (based of analysis of the voting in the ’05 election, the Tories had roughly the same number of seats, but much better geographical distribution (Scottish and Cornish Tories are by far the most disenfranchised voters in the UK) and the Lib Dems take about 100 seats from Labour.

    Multi-member STV would have such large constituencies that the boundary bias largely disappears.

    If the Lib Dems have 30+ share of the vote, then they will have a lot of clout, and I expect a Lib-Con coalition with STV as part of the deal might well be on the cards. IMHO this would be a very good thing for the country, and we could move along from 20th century bipartisan politics.

    Report comment

  8. @eoin

    Eoin one of your four key findingds doesnt hold through (a stable lab Con deficit). The last four polls above show an ever increasing Labour deficit……

    from 4% to 5% to 6% to 8%…..
    very worrying for Labour

    Report comment

  9. Its funny how we forget so quickly the gap between Labour and Conservative…

    the key statistic for three years has become, well…kinda just another ordinary statistic in a sea of extraordnary numbers….

    Report comment

  10. @eoin

    Eoin one of your four key findings doesnt hold through (a stable lab Con deficit). The last four polls above show an ever increasing Labour deficit……

    from 4% to 5% to 6% to 8%…..
    very worrying for Labour

    Its funny how we forget so quickly the gap between Labour and Conservative…

    the key statistic for three years has become, well…kinda just another ordinary statistic in a sea of extraordnary numbers….

    Report comment

1 5 6 7