Recent Posts

 Which figure(s) are you reading?

Table V003 shows the figure of 'Don't Knows' as 15% (or 240 of 1647)

She was never afraid to state the obvious.

By coincidence, YouGov have very recently gone to the trouble of asking actual Londoners what the Met Police should do.

www.standard.co.uk/news/politics/yougov-poll-labour-lead-london-boris-johnson-party-downing-st-plan-b-b977489.html

"Four out of five think Mr Johnson has not been honest in responding to the allegations of Downing Street parties during lockdown, including 69 per cent of 2019 Conservative voters.

As he fought for his political life, the PM told MPs last week that he believed a “bring-your-own-booze” party on May 20, 2020, in the garden of No10 when Britain was in lockdown was a “work event”. He said he went into the garden to thank staff before going back into his office 25 minutes later.

However, nearly three quarters of Londoners believe his assertion that he thought it was a work event is “not an honest account”.

Around the same proportion of Londoners say the Metropolitan Police should investigate the string of parties in No10 when the nation was under Covid restrictions."

(YouGov interviewed 1,166 adults in London between January 13 and 17.)

There's some useful historic data on UK voting patterns from 1974-2010 here:

www.ipsos.com/ipsos-mori/en-uk/how-britain-voted-october-1974

What this shows is that the large differences in voting pattern by age are a recent thing. The Conservatives have generally only fared a little less well with younger voters, and had a commanding lead of 18-24s in their 1983 landslide.  Plenty of other useful data here - such as the decline of social class, and historic voting patterns by gender. The Tories did better with women voters in the 70s and early 80s, after which male and female voting pattersn were near identical, and Labour have been faring better with women from 2005 onwards. 

colin

I think it's worth setting out those Ipsos Mori findings on the 2016 referendum by age and social class

Age18-34
AB 71/29 Remain
C1 71/29 Remain
C2 54/46 Remain
DE 56/44 Remain

Age 35/54
AB 61/39 Remain
C1 53/47 Remain
C2 35/65 Leave
DE 36/64 Leave

Age 55+
AB 48/52 Leave
C1 37/63 Leave
C2 32/68 Leave
DE 30/70 Leave

It is clear from these figures how age was a significantly greater factor than social class in voting patterns.

"ABs-over-55" were closer to "DEs-over-55" than to "ABs-under-35" ( by 18 points diffenece to 23)
And "DEs-under-35" were closer to "ABs-under-35" than to "DEs-over-55" ( by 15 point difference to 26)

colin

" I think many 66 year olds would have an opinion about being described as "old"."

Please have a look at my actual words which were comparative "...older voters (over65)...".

As for your point about sample sizes. Surely you know that the size of the sample is not the only factor in evaluating polling? YouGov's findings from 2016 regarding the over-65 cohort are borne out in subsequent polling relating to brexit- see for example further polling on 'hindsight' from July-Dec 2016. I think I'm on strong ground in saying that YouGov have a far stronger polling record than Ashcroft (remember his 2015 constituency polls) and that on this specific area there is far greater corroboration for their findings than for his. See for example, Ipsos Mori, whose figures from 2016 are 34/66 for age 64-74 and 37/63 for 75+.

www.ipsos.com/ipsos-mori/en-uk/how-britain-voted-2016-eu-referendum

Ipsos Mori's figures are particularly useful in debunking Ashcroft's contention about 'social class' as they publish figures by sub-set of age and class. In thier 55+ age bracket, sub-divided by 'class', ALL classes voted Leave, although the ABs only by 52/48. Conversely, their 18-34 age bracket ALL social classes voted Remain, with their DE social class doing so by 56/44.

And for further evidence of the age-effect, look at the age-cohort figures from August 2016, which was the first of nearly 200 polls where YG have asked whether respondents regard Brexit as having been the right or wrong decision. The age cohorts range from 22/63 for 'Wrong' (u25) to 64/27 for 'right' (o65). These are far greater margins than by social class, which are 37/52 for 'right' (ABC1) to 58/30 for 'wrong' (C2DE)


d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/dwpdjgfxsp/TimesResults_160802_Trackers&VI_W.pdf

The Tories have been lower than 30% only two and a half years ago, but with a large share of their (former) voters switching to BXP towards the end of May's leadership, and then back again.

I keep an eye on the Con + RefUK share, which has not yet gone lower than 32%. I think it's fair to assume that when a General Election comes, the Tories wil be able to recover a large share of those of their former VI who have moved off to other parties on the right. 

robbiealive

Thanks for that.

But it's not really right to say that Labour had a '5% lead among women' in 2017, although they did get 4% more of the female vote (than male).

Per YouGov (again), there was a 6 point difference between male and female voters in the 2017 Election. Their figures were a 43/43 tie for women, and a 45/39 Tory win for men. It's also been a feature of both the last two GEs that the difference is mainly due to younger women, while voting patterns of older women and men are more similar -  per YouGov's analysis.

yougov.co.uk/topics/politics/articles-reports/2017/06/13/how-britain-voted-2017-general-election

So there's been a 6-point 'gender gap' in each of the last 2 general elections. In 2015, it was smaller, at 3 points (Men 37/29, Women 38/33) but this would probably have been larger but for the way men were disproporionately likely to support UKIP then.

yougov.co.uk/topics/politics/articles-reports/2015/06/08/general-election-2015-how-britain-really-voted

Survation 's figures show unusually low levels of Don't Know's (or 'undecided' as they prefer to call them). It's just 6% of Con2019, and 5% of Lab 2019, so my usual calculation of re-allocating half of DKs back to the party they voted for has no effect at all here. And their figures show just 4% of Con 2019 going to Ref.

I do wonder whether Johnson's actions may even appeal to some who were wavering between the Tories and Refuk. 'Lockdown sceptics' are common in these circles, and many in his own party were never keen on rules and restrictions throughout the pandemic. 

For some bizarre reason, Survation have provided three links, which take you to their Nov, Dec, and Jan tables. Like you, I opened up the November ones first. 

It states on their write up: "Full details and data tables including perceptions of the political parties here, here and here."

If you want the January 2022 tables, you need to click on the 3rd "Here".

A few nuggets from their figures:

The Scotland sample of 138 says SNP 45%, Lab 22%, Con 18%, LD 9%

Labour lead by 10 points in England, by 46% to 36%. The North of England sample  (of 431) has Lab ahead by 59/25.

Another huge 'gender gap' in their figures with level pegging of 37/37 for male voters, and a 50/28 Labour lead with females.

I really can't see the point of their sampling 32 people in N Ireland, especially as they clearly don't weight this on a represenatative Unionist/Nationalist basis. The DUP has the support of 16 men and zero women.

33% is the equal-highest that the Conservative VI has been in the dozen or so polls, since 'Partygate2' hit the news around 10 Jan. The norm for them seems about 31%, with Labour on 41-42%.

To me, the most surprising effect of all of this is the decline in the RefUK VI; for some time, they were the natural refuge for any dienchanted Tory voters, and as recently as mid-december they were getting 6% and 7% in some polls. The different pollsters have hem at different levels (mostly due to prompting, I think) but there has been a clear downward trend over the past 5-6 weeks; see the pale blue line in the graph here:


en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election

There has been a similar decline in the Green VI, but this is more understandable, with Labour getting more favourable publicity and/or a narrative they can exploit. 

That would be interesting in the light of recent polling, for example the most recent GB poll (R&W 17 jan):

Lab 50%
Con 24%
LD 9%
SNP 4%
Green 8%
Ref 3%

(Sub-Sample of 752 women)

For comparison, the Tories had a 9 point lead with women voters in GE2019 - per YouGov's dec 2019 analysis, as compared to 15 points in the male vote. . Other pollsters show less dramatic figures, but all recent polls that I've checked havea 'gender gap' somewhat wider than the 6-points we had two years ago. 

As I said, let's look at some polling data. 

I think you're probably right that very wealthy people are much more likely to vote Tory than Labour. Where would we be without such insights? But we were discussing the voting habit of the entire over 65 group, weren't we?

There is a fair amount of Scottish, Welsh and regional polling which could be used to support or disprove your first statement. Maybe take a look? 

Bury South is just about the easiest con-held seat for Lab to pick up. The longer-term trend has been towards Labour rather than the Tories. It was Tory in 1992, but Labour from 1997-2019.And the majority in 2019 was just 402 votes. Given Jeremy Corbyn's unpopularity with Jewish voters, which must have cost Labour votes here, I could imagine Lab regaining the seat even without any sort of national swing. 

He'd be very lucky to be acepted as the candidate for the same seat.


en.wikipedia.org/wiki/Bury_South_(UK_Parliament_constituency)