Follow the evidence

TNS have published a new poll today (1st March) showing that the SNP vote is holding up and probably growing.

The raw data, with don’t knows being excluded, show

Party

Constituency

Regional List

SNP

60%

55%

Labour

21%

21%

Conservative

13%

13%

Liberal Democrat

4%

4%

Others (Green, Rise. Solidarity etc)

3%

6%

Putting this into Scotland Votes (http://www.scotlandvotes.com/holyrood) forecasts the following outcome

Party

Seats

SNP

80

Labour

27

Conservative

15

Liberal Democrat

2

Others (Green, Rise. Solidarity etc)

5

There are a few immediately noteworthy aspects to the findings of this poll.

  1. That the Labour, Conservative and Liberal Democrat votes are the same for constituency and regional list (i.e. both votes). The SNP vote is however slightly down between constituency and regional list – from 60% to 55% – suggesting some, though limited, traction to the proposition to “lend” your second vote to another party other than the SNP.
  2. Of the 56 regional list seats, the SNP would be forecast, on the basis of this poll, to win 9, but with 55% of the vote, while Labour with 21% of the vote would win 27, which emphasises that the “lending” of votes by SNP voters has not, so far, proceeded very far.
  3. If we treat “others” as independence parties (as the Greens, RISE and Solidarity are) – ignoring UKIP – then of the 129 MSPs, 85 could be said to support independence – that is 66% of the Parliament. Perhaps not quite Jim Sillars’ “independence Parliament” but certainly heading in that direction.

What is not apparent from the above figures is the degree to which respondents have excluded themselves by saying that they are undecided, or not being willing to divulge their intention. The raw data (http://www2.tnsglobal.com/l/36112/2016-02-26/j3jl1b/36112/156035/TNS___Holyrood_Voting_Intention_Poll___1_March_2016.pdf) shows that 31% were undecided and 5% were unwilling to say how they would vote. In other words, we don’t know how 36% of this sample of 1034 are going to vote (they seem likely to vote as there was an option to say “probably won’t vote”, but this was selected by only 7% of the sample.

So, it seems legitimate to conjecture how this group – larger than the vote for any party other than the SNP – are likely to vote.

There are of course any number of possibilities. Each party would only be expected to claim that they will take more of this available vote than any others. However, if we look at the evidence, three things seem to be becoming apparent

  • The Labour vote seems pretty constant, with limited variance, suggesting that the hypothesis that Unionist but Labour voters, given some independence-friendly comments by Kezia Dugdale (to the effect that its ok to vote Yes but vote Labour in elections) has not encouraged their core, but strongly Unionist, vote to defect to the Conservatives. But despite constant arguments by such as Lord Foulkes and Duncan Hothersall that the electorate will “wake up” and the SNP will be ‘found out’, there seems to be little forward movement in the Labour vote. Indeed it might be argued that Labour’s aim in this election is to continue as the opposition to the SNP.
  • In turn the Conservative vote, certainly on the basis of this poll is slipping back to normal levels, and certainly shows limited sign of forward movement
  • The Liberal Democrats remain in the doldrums, despite the best efforts of Willie Rennie. Or perhaps because of them?
  • What forward movement there has been for the last 18 months is with the SNP. It might be argued then that the ‘undecideds’ and ‘wont says’ are more likely to vote SNP. Indeed, the tables above assume that 60% of the ‘undecideds’ and ‘wont says’ will vote SNP, that 21% will vote Labour and so on. The hypothesis informing what follows is that the SNP are able to win these ‘undecideds’ and ‘wont says’ at a faster rate.

We already pointed out that the Labour, Conservative and Liberal Democrat votes were identical in constituency and regional list, suggesting their voters using both votes in the same way. The exception was the SNP whose constituency vote was 5% more than its regional list, suggesting a – albeit small – tendency to switch the second vote. Let’s suppose that the SNP, by winning more of the ‘undecideds’ and ‘wont says’ votes than other parties are able to correct that, giving the following distribution of votes.

Party

Constituency

Regional List

SNP

60%

60%

Labour

21%

19.5%

Conservative

13%

11.5%

Liberal Democrat

4%

3%

Others (Green, Rise. Solidarity etc)

3%

6%

This requires the Labour, Conservative and Liberal Democrat votes to shrink at least proportionately – the first two by 1.5% and the Lib Dems by 1%. The “others” vote remains the same.

Putting this into Scotland Votes (http://www.scotlandvotes.com/holyrood) forecasts the following outcome

Party

Seats

SNP

86

Labour

25

Conservative

12

Liberal Democrat

1

Others (Green, Rise. Solidarity etc)

5

This raises the “independence MSP” number to 91, or 70% of the Parliament, so at least more of an “independence Parliament”.

The above forecast for the constituency vote suggests that Labour would be completely wiped out, but that the Conservatives would win two seats. However, if their constituency vote were to decline by 1.5% they too would be wiped out – a figure well within margin of error.

This would leave as the only seat the SNP would be unlikely to win is Shetland, but we might ask whether the transfer of votes there would be sufficiently similar for their candidate to hang on, given the shenanigans of their current MP.

In other words, we seem to be in the foothills of Jim Sillars’ “independence Parliament” – 65% – 70% would not be a bad outcome, though 75% would be better. On the basis of this poll the SNPx2 argument does appear to beginning to stand up. However, even in this poll, the great majority of SNP members would be constituency MSPs, with only 16 being Regional List, even if they did get 60% of the regional list vote.

If the following, very optimistic vote were to come about

Party

Constituency

Regional List

SNP

65%

65%

Labour

17%

15.5%

Conservative

11%

10.5%

Liberal Democrat

4%

3%

Others (Green, Rise. Solidarity etc)

3%

6%

seats would be distributed as follows

Party

Seats

SNP

93

Labour

19

Conservative

11

Liberal Democrat

1

Others (Green, Rise. Solidarity etc)

5

and I think it would have to be said that 98 of 129 seats being held by “independence MSPs”, or 76% would be an independence Parliament.

What would this require? The following table repeats that immediately above, but with the addition (in red) of the figures from the most recent TNS poll

Party

Constituency

Regional List

SNP

65% 60% +5%

65% 55% +10%

Labour

17% 21% -4%

15.5% 21% -5.5%

Conservative

11% 13% -2%

10.5% 13% -2.5%

Liberal Democrat

4% 4%

3% 4% -1%

Others (Green, Rise. Solidarity etc)

3% 3%

6% 6%

First of all the bigger changes are in the Regional List, particularly the 10% added to the SNP vote. However, half of that could be achieved by encouraging SNPx2 among those who already say they are going to vote SNP for their constituency vote. The addition to the SNP vote is therefore 5%. This might actually be achieved by the Labour, Conservative and Liberal Democrat being so demoralised that they fail to turn out to vote. In that case, without a single additional vote, the SNP could increase their proportion of the vote, because the vote for other parties has simply fallen. Otherwise it seems likely that they would have to persuade what we might call “soft” unionists to vote SNP, which might achievable, but would not be easy.

Therefore, what this data, and making a few, we would argue not unrealistic assumptions, suggests is that an “independence Parliament” is becoming a possibility. But ……………..