Sorry for having to repeat myself

Today, the Rev Stuart Campbell, has taken it upon himself to rebut the seat predictors in a piece called “The Limits of Science” ( He has a lot of fun putting funny numbers into the ‘Scotland Votes’ website (, including giving every vote to the SNP and finding that somehow or other the Unionist parties still end up with five seats. Or that if you put the numbers in for how the votes stacked up in 2007, then Labour win rather than the SNP (though Labour win only narrowly), though the inaccuracy in the number of seats, Stuart does say, is “tiny” (a couple of seats changing hands that year would have made all the difference). All a jolly good jape though.

However, while Stuart does point to the limitation of sites, that votes are not uniformly distributed across the country, he does not mention their other limitation to the use he has put them to this morning. That is that they do not start from zero, but from the previous outcome and make their prediction based on the uniform swing across the country (cue the first limitation again). This is what gives us the strange outcomes that Stuart majors on this morning.

However, the Rev is not just having a laugh. He is repeating his no doubt sincerely held #bothvotesSNP view – viz. “Much campaigning by the various fringe parties for the Holyrood contest has been based on “seat predictors” like the one deployed to produce the figures in the piece, purporting to show that a tactical-voting strategy on the list can deliver a large gain in numbers of pro-independence MSPs compared to using both votes for the SNP.” (my emphasis)

Just for a bit of fun of my own, I put the most recent Survation poll into another site – Cutbot ( – which actually gives seat by seat predictions, though again based on there being a uniform swing. This suggests that the SNP overall would lose two seats (going from 69 in 2011 to 67), but win twelve more constituency seats than in 2011 (from 53 to 65). But they would have fourteen fewer Regional List members, winning only two Regional List seats.

The reason for this lies, as I explained in a previous blog, in the way the DeHondt voting system (which is used for allocating seats on the Regional List) works. In Central, Mid Scotland & Fife, Glasgow and the North East, the SNP win every constituency seat, but get not one Regional List seat. In fact, they only win Regional List seats in Highlands and Islands (where they have won all but two of the constituency seats) and in South Scotland (where they ‘only’ win four of the nine seats).

The problem, as I touched on in the earlier blog (, with the tactical voting strategy that is being urged on the electorate by such as RISE, Solidarity and the Greens, is that it needs a quantum state before it really begins to deliver results.

Its rationale lies in the fact that, for instance, if we take the Glasgow Region, because the SNP would have won all nine constituency seats, because of how the DeHondt system works in relation to the Regional List, the SNP in the first round of allocating List seats need ten votes (nine constituencies + 1) for every vote for a Unionist party. Cutbot suggests that this is too much of a disadvantage, and that even if their Regional List vote were to be what Survation forecast, which is more than the Labour and Conservative vote combined, the SNP get no Regional List seats at all, while Labour get four, Green get two and the Conservatives one.

What is required is a quantum shift from SNP to a single other independence minded party. Two things can go wrong with this. First of all, the shift from the SNP spreads itself over more than one party – purl anything you like from Green, RISE and Solidarity. Secondly, the shift does not go far enough to ensure that a single beneficiary party (any of the above three) gets past the Unionist parties.

To this end, and again just for fun, I reworked the Regional List forecast in only one way – to divide the SNP regional vote added to the Green regional vote (42% + 10% – so 26% each) which changes the outcome in two ways

  1. The Green Party would become the Opposition – bit of a blow not just to Ruth, but to Kezia – with 25 seats. I think that would be an excellent development (just for the avoidance of doubt, I am not a member of any party).
  2. The SNP would fall short of an overall majority by a single seat – they would have 64 seats, every single one a constituency member. This might be seen as a “bad thing”, though I have some sympathy for Patrick Harvie’s view to Nicola Sturgeon in one or other of the ‘Leader Debates’ that “a period in minority would do you the world of good”. Given that the SNP would need only one vote to secure a majority for any proposal, compared to 2007 when they were eighteen seats short of a majority, it should certainly be easier than then, particularly bearing in mind that the Unionist parties collectively would lose net seventeen seats – UKIP would gain two, Lib Dems lose one, Conservatives gain one and Labour lose nineteen.

Those who think as Stuart Campbell does will, of course continue to argue that the above is making use of an opinion poll, itself with a margin of error (and they weren’t very good at the last General Election), and a forecasting site which assumes a uniform swing over the country as a whole. Both of these are weaknesses, and there is no gainsaying that. However, weather forecasts are seldom precise – think “scattered showers” – but we continue to listen to and make use of them. I would not defend any of the forecasting sites to the nth degree (I doubt if they would either) – as with the polls on which they depend, there is a margin of error in the degree to which the average swing is replicated across Scotland.

However, it is equally true to say that if the SNP perform at Constituency level as the polls have consistently suggested for many months now, they give themselves a terrible handicap to win Regional List seats. Remember if the SNP take all the Glasgow seats their vote will be divided by ten for the first round in allocating Regional List seats in Glasgow – every Unionist vote will be worth ten SNP votes. That too would be a fact if the SNP constituency vote turns out as forecast at the election.

It is also true to say that a “first vote SNP, but regional list vote someone else” will only work if enough are convinced by it (i.e. there is enough transfer from the SNP) and if it goes to a single beneficiary. I am not a member of any of the three possible beneficiary parties – Green, RISE or Solidarity – so I have no axe to grind there. I can also see the difficulties for more ‘conservative’ (definitely lower case c) elements of the SNP vote having real difficulties in voting for RISE or Solidarity – indeed even for the Greens. This points me toward the value of Jim Sillars’ Yes Scotland Trust. I promised a blog on this a few weeks ago and haven’t yet got round to it. But in essence a Yes Trust would have Scottish independence as its primary aim, but act as a sort of citizens’ debating chamber about how best to achieve this, as well as its possibilities. There would also be the Electoral Commission to watch out for, lest they consider a Yes Trust to be nothing more than the SNP in another guise with the aim of undermining the electoral system. However, this seems to me to be the only viable alternative of asking people to vote for an alternative and hoping that in the main they chose just one, so giving us Sillars’ “independence Parliament”.

As a concluding aside my other ‘forecast’ based on the possibility of the Regional List vote for SNP and Green being evenly split, would yield these two parties 89 seats, or nearly 69% of the membership at Holyrood. In contrast the Survation poll would give SNP + Green only 79 seats between them, or 61%.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s