Regional Predictions for Holyrood 2016 as at August 2015

Original post 13/10/15.  Editted 14/10/14 using August and September data combined.
Editted again 13/11/15 - H&I have 8 constituency MSPs and NE 10. 


Now in my previous posts, I have argued that its important that we vote accordance to our beliefs in policies - see Vote for what you believe in - and also that with the SNP riding high and likely to do a clean sweep Constituency seats there is an opportunity to 'game ' the List vote to minimise the unionist and conservative (small C) old parties.

Some people get very excited about this: I dont think we are going to get mass switching of list votes from SNP in the List as return for Greens voting SNP in the constituencies.  But that's life.  But if we look at the detail, we see that small shifts can deny unionist seats with no risk to SNP List seats in four of the eight regions (You can't risk a seat you aren't going to get!).

But those predictions made so far look at the national picture and what we have is a Regional series of situations.  I've always said we need to look at the detail.  

The problem is that the data on such isn't very robust.  When a poll of 1029 people is split into the 8 regions, there aren't many people asked in each region.  A rule of thumb is that a poll of 1000 will be accurate +-3% - better for big parties and much worse for small parties.  Once the sample gets split between 8 regions the figures are much less reliable.  However having done that health warning, let's have a look at the regional situation - but I'm tempted to say 'Just a bit of fun - these are just predictions'.The data used can be found at http://www2.tnsglobal.com/l/36112/2015-08-07/7wshq7/36112/114148/TNS___Holyrood_Voting_Intention_Poll___10_August_2015.pdf and http://www2.tnsglobal.com/l/36112/2015-10-07/9tbc5m/36112/124920/TNS___Holyrood_Voting_Intention_Poll___8_October_2015.pdf


The sample sizes are very small but the overall results seem to be fairly stable: there isn't a great trend showing.   So what I am going to do is add the actual numbers sampled between the two months together to get a more reliance number.  It is still a small sample and all results must therefore be treated with caution. 

 Constituency predictions

The SNP vote is holding up very well in the constituencies - very much in line with the General Election results.  Of the 73 constituency seats, the SNP are predicted to win 71. 

Overall result

Now the model for allocating the regional seats I will use can be found at http://www.dhondt.eu/php/

For each region, we need to know the number of seats for each party and the regional list vote.  Plugging in these two sets of figures gives us the numbers of additional member seats allocated. 

But we have a problem.  ScotlandVotes doesn't tell us which two constituency seats are not won by the SNP.   Well, I'm going to assume that the LibDems hang onto one seat in the Highlands (let's not get into which one!) and that the Tories hang onto a seat in the South of Scotland.  I hope the SNP do better than this but let's go with the prediction and my assumption. 

Highlands and Islands

There is an extra constituency seat in Highland and Islands as Orkney and Shetland each have a constituency MSP.  And I am going to assume that the LibDems hang onto a seat in this, their traditional last redoubt in Scotland.  


Tiny sample warning! Number of Additional seat for SNP = zero - the level of support for the SNP in the Highlands simply isnt high enough to give a Regional Seat.    With 9 UKIP supporters found in the September sample,, they get allocated a seat.   You all know what to do. 

In practice with the well known and well respected John Finnie topping the Highland list, I'm pretty confident that the Greens will have enough votes to have a Highland Regional List seat.  The question is whether the Tory and UKIP voices can be overcome to quash UKIP.  

North East

Tiny sample warning! This is an SNP stronghold with in addition to the 10 constituency seats, the SNP predicted to win 2 List seats.   It's very much up to Greens to improve their own performance to win a List seat here at at present they are not close.  

Mid Scotland and Fife

Mid Scotland and Fife
Tiny sample warning! Again a strong performance by the SNP means that theygain a Regional List seat after the LibDems.  Greens need to target suurplus LibDem votes to deny Tories the last seat.  

Lothians

Lothians
Tiny sample warning! The SNP don't pick up any List seats and the Greens pick up two. A better than expected performance by Labour or Tory would put that at risk.   Greens should be appealling to SNP voters to help them.  

Central

Central
Tiny sample warning! A strong performance from the SNP means they pick up a List seat as do the Greens. It would take a lot for either to be able to deny the Tories the last seat. 

Glasgow

Glasgow
Tiny sample warning!  Labour pick up lots of List seats, the SNP none and the Greens one.  Greens would need an extra 4% on their vote to gain a second seat to the detriment of Labour.  So Glasgow is a place where SNP supporters could switch to Green to reduce the unionist seats.   I think that is more likely than Greens switching to SNP and risking their one list seat - the SNP have nothing to lose by switching.

It's embarassing having UKIP showing on the poll and 4% for Tories is VERY low.  

West

West of Scotland
A strong showing by the SNP means they pick up two list seats.  The Greens fails to achieve one.  Each must present their own cases to reduce the unionist parties further.

South


South of Scotland
Here I assume a Tory hangs onto a constituency seat.  I know.  Miserable, isn't it?

The SNP pick up one Regional List seats and the Greens gain one.   The Greens should be targetting the LibDems to try to deny Labour a seat.  

Summary and Conclusions

In three regions, the SNP are highly unlikely to gain any List seats but they have a good chance in the other five.  Where Greens are strong: Highland and Islands; Lothians; and Glasgow, small switches of SNP votes to Green could dent unionist voices.   

Worth doing? - well SNP supporters in these three regions need to ask themselves if they want an extra unionist or an extra Green.  


6 comments:

  1. To vote SNP on the 2nd vote is nothing short of madness, If people used the second vote for another indy party we could see as many as 112 indy seats in edinburgh. Why do SNP continue to snob all the other parties but ask for their vote?!?!?!

    ReplyDelete
  2. ajsdisc, you simply can't talk about 112 indy seats, without either a massive slump in the Tory and SLAB vote, or a total desertion of the SNP vote in the regional vote or both! Even if every single SNP voter were to vote for another indy party in the regional vote, without a total collapse of the unionist vote, you still wouldn't get 112 indy seats. Do the maths! The D'Hondt counting system CANNOT be circumvented, as long as the unionist vote remains stable.

    ReplyDelete
  3. Untrue Anon.
    If Party X wins constituency seats and their ground is taken by Party Z on the Regional List, then the unionist seats can only come from their vote share of the 56 top up seats.

    If Party X won all the constituency seats on 55% of the vote and those votes largely transferred to Party Z so they won 50% of the regional list seats, the we would have approve 73 X plus 28 Y : leaving the unionist parties to share the remaining 28 between them.

    ReplyDelete
  4. The D'hondt system is fairly bullet proof and it would require amazing discipline from all concerned to make any major dent in the outcome of the calculations. In addition to that, you need to know in advance how the constituency votes go, then you need to know how each party fared in the regional vote, then based on this, you need to rely on everyone shifting their voting preference to satisfy the optimum outcome.
    Any of these factors are practically impossible to guarantee - and we haven't factored in the appeal for voters to transfer from the SNP to RISE, SSP, etc (which obviously would fracture the transferred votes and disadvantage every indi party to the extent that the Unionists would pick up the available seats).
    In a nutshell - the system is difficult to manipulate and it has a serious backlash if the tactical voting falls short, ie the Unionists calmly absorb those badly doctored seats.
    Finally, don't fall into the 'we gave the SNP our constituency vote, so they should give the Greens their regional vote' argument. If you don't stand a candidate in the constituency votes, a Green voter has only got one real option left (ie vote SNP) although I'm sure there will be SSP or RISE candidates in some constituencies which would offer an alternative. This short fall of Green candidates is a problem for the Greens and doesn't lock the SNP voters into some sort of 'tit for tat' arrangement between the two votes.
    It's unrealistic to expect SNP voters to magically know all this rigging stuff. As far as the vast majority of these voters are concerned, you vote for who you want to see in Holyrood - and if you have a candidate in the constituency vote, you vote for them. In the regional vote, you also vote for your party.
    The internet/facebook bubble is extensive, but doesn't come close to reaching the people who exist outwith its sphere - and that's an awful lot of people. Bearing this in mind, this makes things even tougher to orchestrate with regards to tactical voting and puts even more pressure on a strong discipline from all concerned.
    In conclusion: Tactical voting might make a lot of sense, but in practice it is near impossible to pull off - and a botched attempt (most likely outcome) could easily skew the calculations in favour of a Unionist party which would gleefully hoover up the seats in question. The SNP are predicted to do well enough in the constituency votes to render the regional seats less crucial in their objective of securing a majority. However, a lot of stuff could come to the fore between now and the election. The Unionists are turning the screws already and they will keep this up with a relentless determination. Just like the referendum, it will be a 'win at all costs' final stint. This could dent the SNP chances of a clean sweep, so lets not go overboard with tactical plans so far from the final days of voting. The SNP might need every single regional seat they can get to make a majority. Ironically, that would mean the Greens et al might be better served voting SNP in the regional vote.

    ReplyDelete
  5. Do you think there will be more detailed regional polls nearer the time? I have played with the calculator and would love to get a Green opposition to an SNP gov. I have some national stats here: https://www.facebook.com/julie.tomlinson.984/posts/10208742346542442 but they will probably not equate perfectly to regional seats. Would be good to choose whichever party could best pick up the list seats for a region.

    ReplyDelete
    Replies
    1. TNS are the only published region by region poll I've found but the sample size is very small and therefore are of limited help.

      Other sources which indicate areas of longer term support for Greens and relative weakness of SNP are worth looking at. These tend to reinforce the SNP/Green logic in Highlands, Lothians and Glasgow.

      Delete

I am happy to address most contributions if they are coherent. Comments with a constructive contribution to make to the discussion, even if it is critical will be posted.

Please do not add comments in the name of real people unless you are that person.

The views expressed in comments are those of the poster, not me.