MORI’s monthly political monitor has topline figures, with changes from last month, of CON 42%(-6), LAB 32%(+4), LDEM 14%(-3).

On the surface of it, a huge swing back towards Labour. As a caveat though, the previous MORI poll seemed rather out of line with others – showing the Conservatives way up on 48% when other companies had them in the low 40s – while in terms of the Conservative and Labour position this one is very much in line with other companies figures.

(Incidentally, I don’t think the tables for this are up on MORI’s website yet, but there is an interesting presentation on voting in the London mayoral election last year, suggesting that voting behaviour correlated more strongly with race than it did social class). I expect this is actually a correction after an outlier, but we shall see – as readers will know, I do always urge caution with any poll showing a large shift of support.

UPDATE: Full results here. Mike Smithson over at politicalbetting argues that the difference is probably largely down to likelihood to vote. MORI’s figures without any filtering for likelihood to vote show a much smaller change: CON 39% (nc): LAB 33% (+2): LD 17% (-2). He’s probably right. Unlike Populus, ICM and ComRes who weight by people’s turnout to vote, so people who say they are 9/10 likely to vote count slightly less than people who are 10/10, but slightly more than people who are 8/10, MORI apply a harsh filter entirely exclusing everyone who doesn’t say 10/10. A small firming a party’s vote could therefore have a sharp effect on the topline figures – there’s a much more detailled explanation of how pollsters take account of likelihood to vote here.

Also interesting is that, like the Sunday Times YouGov poll at the weekend, David Cameron’s ratings have shop upwards, giving him a net approval rating of plus 22, equalling the highest MORI have ever found for him. As I said in reference to the YouGov poll, some of this will probably be an expression of sympathy given his recent family tragedy.

Finally, the proportion of people who are optimistic about the economy improving in the next 12 months has also gone back up again. To some extent I suppose it matches up with the post I made at the end of last year about Labour’s fortunes being linked with economic confidence (for the record, the TNS data I used in that post also showed people a little more optimistic last month, though not nearly to the extent MORI found). To be honest, I didn’t actually expect the correlation to persist, I expected Labour support to drop with economic confidence, but not that it would necessarily rise if confidence returned – government’s don’t always get credit for these things, and recovery can even make people less risk adverse and more open to alternative government. Still, worth keeping an eye on it.


27 Responses to “Ipsos MORI’s monthly poll”

  1. Last Mori, clearly an outlier. This one is on trend for Jan and 2 December polls.

  2. I find the Liberal figure too low, with a national average of 17% 14% is far too low to be accepted as representative.

    However the abour recovery was expected with Brown doing his global grand standing.

  3. Interesting to see how the ‘ethnic/minority’ vote was so skewed toward Ken during the mayoral race. I suspect much of the difference can be put down to the ‘racist’ smears aimed at Boris during the campaign.

    It shows how a well aimed scrap of bulls**t in the media can make a real difference toward the climax of a vote.

    A lesson that the Tories should heed over the coming year. They will need to stay squeaky clean to repel the barrage of spin and unsubstantiated innuendo that Labour will doubtless throw at them.

  4. looks like a correction in moris polling set up as the others seam to be +5% on the last poll so mostly in line with everyone ells.

  5. What are the breakdown for ‘others’, this should be interesting leading upto the Euros election, as it may give us an indication on how well the minor parties will do.

  6. Dean,

    14% does seem low, but then YouGov have had them on 14-15% twice in February.

    Personally I would put them on, 16% (+/-2%), so if they are out of the news then 14% doesn’t ring alarm bells for me.

    Peter.

  7. It is pretty hard to believe that the Lib Dems are only 2 points ahead of ‘others’ this month or indeed that the Tories were ever on 48% last month. These inexplicable fluctuations suggest to me that the adjustments MORI made last summer after the London mayor debacle have not quite eradicted all the problems.
    By comparison the stability in the polls conducted by YouGov continue to impress.

  8. Nick,

    “By comparison the stability in the polls conducted by YouGov continue to impress.”

    it’s only impressive if things are stable and yougov and net polling isn’t missing something. We just can’t tell.

    I am not saying your wrong, just that either could be right.

    Peter.

  9. Yes, Peter makes a point, perhaps the electorate are volitile in where their support goes. After all most people prob. still seek the best anti-labour party rather than be won over by the Tories. One thing I would say though is that the Liberals have a good opportunity to be the non-socialist Democratic radical alternative to labour and the Conservatives.

  10. We seem to have settled into a period of stability.

    Since the beginning of February every poll bar one has had Labour in a range of 28-32% and every poll bar one (not the same one) has had Cons in the range 40-44%.

    We can debate whether that gives a weighted average of 30% v 42%, or if the Con lead may be as low as 8% or as high as 16%, but it is clear that if an election were held now, then there would most likely be a Conservative government with a working majority, but not a landslide victory.

    Looking ahead, what is most likely to cause a significant or enduring shift in this position ?

    Barring major unforeseen events, the economy will remain the foremost issue until we have clear signs of recovery.

    Based on economic fundamentals, I still can’t see anything which could support a sustained Labour recovery, but can see a potential further drop in Labour support, either immediately after the budget, or else over the spring, as the full picture on the dismal fiscal position emerges.

  11. Anthony,

    had a look at the charts on the Mori site.

    The one on economic confidence seems to show a slight recover in the last month which oddly seems to coincide with at least a stabilising of labours position at around 30% when some were predicting it to dip towards 25%.

    So it looks like your theory is still holding up.

    Peter.

  12. There’s a very clear policy distinction now between “more for less”, or “tighten the belt” on one side and “loosen the purse strings” on the other.

    This could well lead to the Govt receiving more of the blame in the polling figures for worsening conditions, and more of the credit than normal if conditions are seen to be improving.

    I’m sure the economy will play the major part in deciding the next GE, but Labour still has much less of a fresh look about it than they should have if they were going to recover in time. Image counts for a lot too.

  13. WMA 42:30:16 so no real change. As you say, the previous Mori poll was egregious with a WMA error of 5.3 and a Retrospective error of 6.0. They now have a Retrospective Standard Deviation of 3.6 much the worst of any of the pollsters.

  14. @ Paul H. J

    “Looking ahead, what is most likely to cause a significant or enduring shift in this position ?”

    Well, if the Liberals can make inroads into the Labour vote, this might allow for a further collapse of labour voting into the depths of 25-28%, Liberals back on 20-23%.

    Its not just idle speculation surely- as this was pretty much what happened in a couple of poll findings in the last year (2008) and surely could be replicated in 2009:

    2009-02-05: 40:28:22 (ICM)
    2008-10-19 saw the Liberals clime back to 21 again.
    2008-06-08: 45:25:20 (Populus)
    2008-06-22: 45:25:20 (ICM)
    2008-06-05: 42:26:21 (ICM)
    2008-05-18: 41:27:22 (ICM)

    So you see the trend I speak of is well founded, and streaches back throughout last year and maybe into this year. Especially given that the latest finding was not just an ICM result, it was reciprocated in the 2009-02-12 ComRes finding: 41:25:22

    Its a a very real possibility that the Liberals could eat deep into the Labour vote, especially as their party gets increased coverage was the electioneering starts.

  15. if their were an election tomorrow the conservatives would have a majority of around 70 an 360 seats +150 and labour would fall back to around 229 -120 seats with the lib dems on 31 -31 others on 30 +1

  16. Once again Mori produces wild fluctuations which are clearly as Mike shows due to the very strict restriction tn the headlibe figures to those absoluely certain to vote fewer than 50% of the electorate . As I pointed out last month the tables also give a figure based on those 6-10 likely to vote which is rather more realistic
    This month that figure is
    Con 41 Lab 32 LD 17 Conservative lead 9 .

    The detailed data shows LibDems gaining 21 voters from Labour losing 6 to them and gaining 5 voters from Conservatives but losing 12 , a net gain , this net gain of voters just cannot be reconciled with a headline figure showing LibDems at 14% .

  17. Recent better polls for Labour may be due to a lesser focus on the economy by the media. The tabloid’s in particular have been obsessed with the story of Jade Goody’s cancer.

    I wonder what impact Brown’s partial apology will make on the next poll. I find it remarkable that it has taken so long for focus to fall on Brown’s part in the recession.

    The next time the unemployment figures are released may prove to be the end of Labour polling above the 30%, with the exception of a rogue poll.

  18. Unemplotmeny on it’s own has little efecct on support or lack of it for a government . There will always be many more people with many more votes in employment than those out of work . The Conservatives drove unemployment above 3 million twice in the early 80’s and early 90’s and were still reelected and the high unemployment of the 1930’s did not stop the government being elected in 1935 .

  19. New ICM/Guardian poll
    Con 42 Lab 30 LDem 20

  20. The LibDem position is quite interesting here as I’ve just looked at the ‘past vote’ numbers to try to determine the percentages of how people who voted for one party will go on to vote.

    1% of 2005 Con voters said they’d vote LibDem, 8% of 2005 Lab, 75% of 05 LD and 17% of 05 others. Applying those %ages to the actual 05 election gives the libdems 21%.

    I don’t know how applicable the method is, but it’s hard to believe the LibDems will be getting 14%.

  21. The MORI poll without filtering looks like a poll from 2 or 3 years ago. It seems to show the Labour core vote holding up very well, and that once again the Tories are doing well because of greater enthusiasm amongst their supporters; and it means that the Tories would continue to benefit from a lower turnout.

  22. “Yes, Peter makes a point, perhaps the electorate are volitile in where their support goes”

    I think that’s unlikely to be honest. What is much more likely to explain the relative volatility of MORI when compared to other polling companies is the fluctuation in people’s thoughts on whether they are certain to vote or not.

    In my opinion, allowing for margins of error, the polls have been effectively static over the last 2-3 months. And this is what you might expect given that the position of the government and the opposition have become so entrenched. There doesn’t seem much scope for the electorate switching support.

  23. John,

    I wasn’t actually saying I thought the electorate was volatile, rather that you just can’t tell which methodology is right you can only speculate.

    I am a big fan of YouGov but I don’t think it is the be all and end all.

    Equally I am increasingly of the opinion that different polling methodologies are better at picking up different things than others and that they can detect different trends.

    I have a feeling that random polls can detect a change in mood before a panel like YouGov but are equally more vulnerable to “chatter” that can produce rouges.

    Given that I tend to focus more on trend than headline (partly because the scarcity of Scottish polls mean that I have to try to deduce movement by averaging small samples anyway) that makes me more inclined to favour the YouGov approach.

    Peter.

  24. Is it right to say Lib Dems would take more votes from Labour than the Tories.
    If so does anyone have figures to suggest this?

  25. Andy Stidwell,

    I have to take issue with you on Tories benefitting from a low turnout.

    The evidence from 1997 / 2001 / 2005 is that as turnout fell, Labour managed to hold an ever greater share of seats relative to its number of actual votes, while it is pretty clear that Major managed to win in 1992 by achieving a record number of votes cast.

    It may be that a low turnout among Labour supporters will deliver more seats to the Conservatives than might otherwise be the case, but for Cameron to secure a clear majority it is far more important that there is a high turnout among actual / potential Conservative voters. This is evident from Mori’s filter.