The penultimate Guardian/ICM poll of the 2017 General Election campaign continues where the previous few ICM polls left off. The Conservatives retain an 11-point lead which they saw from ICM through the weekend, following up on the 12-point lead published in The Guardian last Monday.
Headline vote intention figures are (compared to the last Guardian poll on 30th May):
Conservative 45% (nc)
Labour 34% (+1)
Lib Dem 8% (nc)
UKIP 5% (nc)
Green 3% (nc)
SNP 4% (nc)
Plaid Cymru 1% (nc)
Other 1% (nc)
Compared to other polls over the weekend ours slots in the higher end of the Tory lead range, one point off ComRes 12-pointer. At the other end, Survation revealed a Tory lead of only 1-point. A moderately significant dividing line has emerged between sets of pollsters, largely pivoting on how we treat turnout. This has been widely discussed in polling circles since the Labour ‘surge’, which is at least partially based on younger people and 2015 non-voters saying they will now turnout, and vote for Labour.
Those pollsters, like us, who show higher Tory leads are implicitly sceptical about the extent of this self-reported turnout. Those with lower Labour leads largely take it at face value. But whichever turnout weighting scheme is applied, the impact is clear – as Sturgis & Jennings of the University of Southampton established in their paper, which was published yesterday.
https://sotonpolitics.org/2017/06/04/will-turnout-weighting-prove-to-be-the-pollsters-achilles-heel-in-ge2017/.
|
|
Vote estimates with turnout weight
|
Vote estimates without turnout weight
|
Pollster |
Fieldwork End Date |
CON
|
LAB |
CON |
CON |
LAB |
CON
|
(%)
|
(%) |
lead |
(%) |
(%) |
lead
|
ORB/Sunday Telegraph |
4th June |
46
|
37 |
9 |
44 |
38 |
6 |
IpsosMORI/Standard |
1st June |
45
|
40 |
5 |
40 |
43 |
-3
|
Panelbase |
1st June |
44
|
36 |
8 |
40 |
39 |
1
|
YouGov/Times |
31st May |
42
|
39 |
3 |
41 |
39 |
2
|
Kantar |
30th May |
43
|
33 |
10 |
40 |
34 |
6
|
ICM/Guardian |
29th May |
45
|
33 |
12 |
41 |
38 |
3
|
Survation (phone) |
27th May
|
43 |
37 |
6 |
43 |
37 |
6
|
ComRes/Independent |
26th May |
46 |
34 |
12 |
43 |
38 |
5
|
Opinium |
24th May |
45
|
35 |
10 |
42 |
36 |
6
|
Survation (internet) |
20th May |
46
|
34 |
12 |
43 |
33 |
10
|
GfK |
14th May |
48
|
28 |
20 |
45 |
29 |
16
|
|
|
|
Mean |
= 10 |
|
Mean |
= 5 |
|
|
|
S.D. |
= 4.5 |
|
S.D. |
= 4.9 |
The imposition of historical-based turnout probabilities (i.e assuming that behaviours will tend toward the historical pattern) drives down the Labour share and upweights the Tories. Self-reported turnout scales on the other hand – largely employed by those pollsters showing the smallest Labour leads – hardly impact on the headline numbers. June 9th will show which was the better scheme, but the age profile of voters from all General Elections since 1964, courtesy of the House of Commons Library, suggests over-statement of self-reported turnout is likely. For example, it’s hard to reconcile between 38%-54% estimated turnout among 18-24s at General Elections since 1997 with one recent poll, which suggested that 82% of them would turn out to vote.
But who knows? Jeremy Corbyn has, to common agreement, run a good campaign and has motivated sections of the society who have tended to disengagement. On Friday, we will have the answer on whether he has bucked the trend, or not.