Uncategorized

Polling Voting Intentions for Devolved Elections in Mid-Term

One aspect of last month’s YouGov poll for this blog that elicited particular comment was the results for Assembly regional list voting intentions. To remind you, the voting intention figures in the poll were:

 

Westminster

Labour: 48%

Conservative: 23%

LibDems: 8%

Plaid Cymru 9%

UKIP: 8%

 

National Assembly (Constituency)

Labour: 46%

Conservative: 19%

LibDems: 8%

Plaid Cymru 17%

UKIP: 6%

 

National Assembly (Regional List)

Labour: 25%

Conservative: 12%

LibDems: 9%

Plaid Cymru 23%

UKIP: 16%

Others: 14%

 

Some people have suggested that the List results are just implausible. Alternatively, Gareth Hughes has suggested that “It looks as if Labour supporters have now understood the nature of the regional list system and are not wasting their votes by voting Labour on the lists where they stand no chance of winning seats.“

While I’d like to believe that the poll results show increasing political sophistication on the part of Welsh voters, I suspect that the explanation may be a little more mundane. But to consider this it might help to look in detail at how the results were generated.

For Westminster, YouGov ask the following question:

“If there were a general election held tomorrow, which party would you vote for?”

Voting intentions for the National Assembly were asked by YouGov in the following way:

“In elections to the National Assembly for Wales you have two votes. One is for an individual member of the Assembly – or AM – for your constituency. The second is for a party list for your region. If there were a National Assembly for Wales election tomorrow, which party would you vote for in your constituency?”

“Now thinking about your second vote, for a party list in your region, which party would you vote for?”

This might seem simple enough. But actually it is not how YouGov have always asked about devolved voting intention in Wales. In some previous polls, they used this question wording:

“If there were an election to the National Assembly for Wales tomorrow, and thinking about the constituency vote, how would you vote?”

“And thinking about the regional or party vote for the National Assembly for Wales, which party list would you vote for?”

The following tables lists Assembly voting intention figures in all YouGov published polls in Wales (since they began polling in 2009).

Wales, NAW Polls (Constituency vote)

Poll

Lab

Con

Lib-Dem

Plaid

others

IWP-WGC/YouGov, July 2009

26

26

15

24

9

IWP-WGC/YouGov, October 2009

32

25

12

24

7

NAW/YouGov, March 2010

35

23

13

22

7

ITV-Wales/YouGov, 1-3 May 2010

32

21

20

22

5

IWP-WGC/YouGov, May (post-election) 2010

36

19

16

17

12

ITV-Wales/YouGov, June 2010

42

19

12

20

6

ITV-Wales/YouGov, July 2010

40

20

13

22

5

ITV-Wales/YouGov, August 2010

39

22

10

23

8

ITV-Wales/YouGov, September 2010

44

22

11

19

5

ITV-Wales/YouGov, October 2010

44

19

9

21

8

ITV-Wales/YouGov, November 2010

44

21

9

21

6

ITV-Wales/YouGov, December 2010

44

23

6

21

6

ITV-Wales/YouGov, January 2011

45

21

7

21

6

ITV-Wales/YouGov, early March 2011

48

20

7

19

7

ITV-Wales/YouGov, late March 2011

47

21

7

17

7

ITV-Wales/YouGov, mid April 2011

49

20

8

17

6

ITV-Wales/YouGov, late April 2011

45

21

8

18

8

ITV-Wales/YouGov, 2-4 May 2011

47

20

9

18

6

ITV-Wales/YouGov, end-Jan – Feb 2012

49

20

7

17

7

Cardiff Uni/YouGov, mid-April 2012

48

19

7

18

8

ITV-Wales/YouGov, July 2012

50

19

7

17

8

ITV-Wales/YouGov, August 2012

47

19

7

17

10

ITV-Wales/YouGov, February 2013

46

21

10

17

6

WGC/YouGov, July 2013

46

19

8

17

9

Wales, NAW Polls (Regional vote)

Poll

Lab

Con

Lib-Dem

Plaid

others

IWP-WGC/YouGov, July 2009

26

26

14

20

15

IWP-WGC/YouGov, October 2009

30

27

11

21

12

NAW/YouGov, March 2010

32

22

12

23

12

ITV-Wales/YouGov, 1-3 May 2010

30

21

18

21

9

IWP-WGC/YouGov, May (post-election) 2010

28

16

21

18

17

ITV-Wales/YouGov, June 2010

40

20

12

19

9

ITV-Wales/YouGov, July 2010

37

20

14

20

8

ITV-Wales/YouGov, August 2010

39

21

9

23

8

ITV-Wales/YouGov, September 2010

41

20

12

19

8

ITV-Wales/YouGov, October 2010

40

18

9

23

11

ITV-Wales/YouGov, October 2010

41

20

9

20

11

ITV-Wales/YouGov, December 2010

42

22

5

21

10

ITV-Wales/YouGov, January 2011

41

20

8

21

10

ITV-Wales/YouGov, early March 2011

45

20

5

18

11

ITV-Wales/YouGov, late March 2011

45

20

8

16

10

ITV-Wales/YouGov, late March 2011

44

20

8

18

10

ITV-Wales/YouGov, late April 2011

41

20

7

18

16

ITV-Wales/YouGov, 2-4 May 2011

43

19

8

18

13

ITV-Wales/YouGov, end-Jan – Feb 2012

45

20

7

15

13

Cardiff Uni/YouGov, mid-April 2012

33

13

9

22

23

ITV-Wales/YouGov, July 2012

35

11

8

20

25

ITV-Wales/YouGov, August 2012

30

14

7

24

25

ITV-Wales/YouGov, February 2013

26

14

11

26

24

WGC/YouGov, July 2013

25

12

9

23

30

If you guessed that April 2012 was when the revised wording was introduced, well done. I hereby convey the Order of d’Hondt upon you.

Why did YouGov make this change? In 2011 their final polls over-estimated Labour support and under-stated support for ‘other’ parties (i.e. those outside the four currently represented in the Assembly). The new wording, with its longer introduction, was introduced to try to achieve greater accuracy by reminding survey respondents of the details of the voting system.

YouGov have run ‘split-sample’ tests on the two question wordings: where you give one randomly-selected sample one formulation and the another equivalent sample the alternative wording, to see if the different wordings elicit systematically different responses. This test has suggested that the new wording tends to produce slightly lower levels of reported Labour support on the list vote, and higher levels of support for Plaid and UKIP.

So have YouGov now got it right? I certainly do not criticise them for changing a question wording in pursuit of greater accuracy. But one problem they – and all other pollsters – face is that polling is attempting to use a sample to get close to the ‘true’ figure in the population as a whole. However, this assumes that there is such a ‘true’ figure to be discovered. In the case of National Assembly elections, this assumption may be problematic for some voters at least some of the time.

Given the relatively low media profile of devolved politics in Wales, and given also the relative unfamiliarity of the two-ballot AMS voting system used for devolved elections, it is very likely that plenty of voters forget about their ability to cast two votes under this system. Therefore, while they may well have some general voting preference for devolved elections, they may not have clear preferences for the two ballots. The new YouGov wording is intended to help them. But the reference to ‘your second vote’ may, I suspect, lead some of them to think of their second preference.

The plausibility of this interpretation is increased by looking at the details of the poll. Fully 24% of those giving a Conservative voting intention for Westminster indicated that they would vote UKIP on the list vote; the same percentage of Labour Westminster voters (and 25% of Labour Assembly constituency ballot supporters) also chose Plaid Cymru for the list vote.

It is worth saying that it is not only YouGov who have experienced problems in this area. Prior to the 2003 devolved election, in November 2002, NOP conducted a poll (for HTV) which found Plaid Cymru on 31% on the regional list vote, only 1% behind Labour, even though Plaid was well behind on the constituency ballot. In the actual election, of course, Plaid was well beaten by Labour on both ballots (and actually did worse on the list vote).

We should also remember that any problems gauging list vote intentions may well be largely self-correcting as we get closer to the Assembly election: the party campaigns, and media discussion, should help re-educate voters into how the system works.

Until then, we should perhaps continue to interpret regional list voting intention figures for the National Assembly with a degree of caution. And both Plaid and UKIP supporters should avoid get too excited. At least not yet.

Comments

  • BoiCymraeg

    This confirms my suspicion that there mus have been a shift in the way the polls were conducted to explain the seismic shift in Assembly polling patterns – there’s little to nothing in the actual election results of 1999-2011 that suggests people understand the voting system particularly well (as the poll seemed to indicate had suddenly happened).

Leave a Reply

Your email address will not be published. Required fields are marked *