On the Proportionality of the Scottish Election Result

[danieldwilliam]

Last night I had the good fortune to attend a briefing at Holyrood by Professor John Curtice** of work commissioned by the Electoral Reform System and hosted by Neil Bibby, former Scottish Regional Co-ordinator for Yes to Fairer Votes and now an MSP. I also got to meet Katie Ghose and Willie Sulivan from the ERS who I worked with a little on the AV Referendum.

Prof Curtice was talking to this report on the way the Scottish electoral system has operated and been used by electors. It’s not a report on why the SNP won the election.

http://www.electoral-reform.org.uk/images/dynamicImages/ScottishReport2011_FINAL.pdf

The interesting points from the report from me are

The D’Hondt system used in Scotland is much, much proportional than the Single Member Plurality system. *** No great surprise there, using proportional representation gives you a more proportional result. What was interesting was that the this result was no more or less proportional than other results in Scotland using the same system in previous elections.

The D’Hondt system using regional lists tends to favour larger parties. What makes the result look less proportional is that the previous government Party was in fact two parties, Labour and the Liberal Democrats. With the disproportionality concentrated more in one party’s favour than in two it looks starker. Also, the disproportionality has favoured the SNP when it wasn’t supposed to.

Curtice appeared to take some glee in pointing out to the Labour Party that they have reaped what they have sown when they set up the system thinking that they would always be the largest party. I think there is a word of caution there for the Tories with their support for Single Member Plurality. When you pick an electoral system because it favours you in the current structure then you shouldn’t be surprised when you suffer badly if the structure changes. I for one look forward with great joy to UKIP winning some seats at the next UK General Election. If I were a Labour supporter in a safe Tory seat I’d vote UKIP.

Due to the combination of the D’Hondt method and the use of regions there appears to be a de facto threshold of 5% of the vote in order to win any seats. Not quite but nearly. Running the result on the same votes using a national rather than a regional top up would give the Liberal Democrats 7 members rather than 5 with the Greens on nil rather than 2. Using a 4% threshold the Greens would be allocated 6 seats.

Curtice has usefully run the same votes through a different but similar method of seat allocation the Sainte Lague method. This gives a more proportional result and corrects some of the bias towards larger parties. Using Sainte Lague the Greens would have been allocated 7 seats. The same votes, in the same places counted slightly differently triple the representation of the Green Party George Galloway would have won a seat. Make of this what you will.

What you can’t see from running the votes cast through different system is how people would have voted using that different system. Would more people have voted Green if it looked likely that they would win 7 seats rather 2? Who knows?

Electors seem to be quite savvy about the system. Green voters are slightly more likely to cast a blank ballot in the constituency vote where Green candidates don’t stand. Voters generally seem happy to split their vote between the constituency and list ballots; often favouring incumbent constituency candidates.

The Labour Party’s refusal to allow candidates to stand in both the constituency and list ballots has resulted in a large turnover of Labour MSP’s. Proportionately they are very inexperienced. I fear that they could be overwhelmed as the largest non-government party and carry a record of ineffectiveness into the next election. Coupled with the advantage that incumbents have in the constituency vote I think Labour may struggle for a few terms with the effect of their “winning loser” policy.

The big take aways for me were; Labour have shot themselves in both feet and if you support a smaller party, or just favour greater proportionality, a small change in the voting system could make a big difference.


** http://www.crest.ox.ac.uk/curtice.htm

*** AKA First Past the Post

Comments

[f4f3.livejournal.com]

Fascinating stuff - I like the idea that we're willing to experiment, and might experiment further.

And yes, I know that tactical voting under FPtP is almost a moral imperative, but I just couldn't vote UKIP, especially if that meant I would help elect a UKIP MP.

[danieldwilliam.livejournal.com]

I would be prepared to play the long game when voting UKIP. The prize, the removal of the permanent propensity of our democracy to return a Conservative Goverment is worth my moral contamination.

It was a good session although the Q&A was focused on MSP's asking how they could better game the system *ahem* work with the system to make sure that the views of the people the representated were more clearly heard.

The spat between Joan MacAlpine, Curtice and one of the Labour MSP's about mandates was worth the admission price alone.

Blank ballots in the constituency vote

[jamie montgomerie (from livejournal.com)]

Electors seem to be quite savvy about the system. Green voters are slightly more likely to cast a blank ballot in the constituency vote where Green candidates don’t stand.

I have an intuition to why this 'works', but I'd never considered it in the past. Could you elaborate on the theory behind it?

Re: Blank ballots in the constituency vote

[danieldwilliam.livejournal.com]

I don’t think it’s any more or less complicated than :-

It is very difficult for a Green candidate to be elected in a Single Member Plurality constituency. Therefore the Green party saves its time and money and doesn’t stand a candidate.

Committed Green voters (some) are indifferent between the choice of Labour, SNP, Lib Dem and Conservatives. They don’t want to vote for any of them and it makes no (or little) difference to the outcome of the election they are interested in.

I wonder if a really canny Green would be better off picking the constituency candidate from a party who was most likely to be one seat short in the proportional allocation of seats and therefore save an extra seat for Greens.

Prof C noted that in a few seats, notably the Northern Islands seats, there was the converse effect with people voting for the constituency candidate they favoured but not bothering with a list vote.

Re: Blank ballots in the constituency vote

[jamie montgomerie (from livejournal.com)]

Ah - okay. I thought there might be something more tactical behind it - i.e. a blank ballot in the constituency vote would have the effect of reducing the percentages of the other candidates without adding to any, thus making the green vote in the list vote count for more. I admit I was woefully unknowledgable about how the list votes interact with the results of the constituency votes - having read up on it now, I understand that it's just the list votes that determine the proportionality, so my intuition on why this 'worked' tactically was incorrect.

It does sound like voting for a party that would be most likely to be one seat short in the proportional allocation would be the canny thing to do - although that presumes they'd be near enough winning the constituency under FPTP that a few extra Green votes could push them over.

Presuming you're voting for a party in the constituency vote, not using your the list vote seems like a particularly strange thing to do if you understand the system (even limitedly, as I did before I asked my question...). Aren't you basically handing the equivalent of 1/(number of parties) worth of a vote to all the other parties?

Re: Blank ballots in the constituency vote

[danieldwilliam.livejournal.com]

I think situation where the additional advantage from tactically voting for the one seat short party is going to be rare. We might see it in Edinburgh Southern for example.

Re the Constituency vote it does seem strange. There were a independents in both seats who weren't standing in the List so this might be the explaination for most of the observed oddity.

I think you have it right, handing over 1/xth of a vote to everyone else and therefore pretty much nullifying your constituency vote on average(?)

[ken o'neill (from livejournal.com)]

Listening to John Curtice is always a pleasure, one of the few things I miss from my previolus job was attending his seminars. He ahs long said that the Scottish electorate are canny enough to use the system to their advantage, voting one way on the constituency ballot - say to vote for the least objectional party who is likely to win the seat - and another on the regional ballot - for the party they actually support. I would have loved to hear what he had to say and his prediction two years ago that we would see less independent MSPs this term held true.

Prof Curtice has also previously pointed out that we vote differently depending on which election it is - Westminster, Holyrood or Local Authority. Just because somebody votes for the SNP at Holyrood it doesn't follow that they will for their council or for the Commons. Again, we are savy enough to use the system to make it represent what we see as our best interests.

I would love to hear more about the stooshie twixt himself, Joan MacAlpine and the anonymous (and possibly new?) Labour MSP. As for the mandate issue, Labour never had an overall majority in terms of the electorate support them in 1997 or since, so it is ironic and possibly down to bitterness that they are raising it now. They didn't seem to mind before. However, they do have a point. After the election I calculated less than half the adult population of Scotland voted in May and of those around 40% voted for the SNP. The result was staggering but it hardly constitues an overwhelming mandate in traditional voting terms.

[danieldwilliam.livejournal.com]

Prof C was talking about the turnout and exploring the differential turnout between Westminster and Holyrood (low 60’s compared to 50.2* (ish)).

JM started to ask a question / make a statement and Prof C said (and I paraphrase) – I see where you’re going with this J but it’s difficult to call polling less than half the votes cast on a turn out of only just above half a mandate for any thing as big as Independence.

Haha – says unknown Labour MSP – no mandate, no mandate, no mandate (waggles fingers from nose)**.

Prof C – nor of course did the previous Labour / Lib Dem governments have much of a mandate or indeed the UK Labour Government on somewhat less of the vote than the SNP achieved.

JM – Haha, no mandate, no mandate…

Prof C – which still doesn’t mean the SNP have any better a mandate. You should all be a little worried about the low and falling turn out. (administers slaps to all and sundry).

That’s about the size of it.

*50.2% makes me weep.
** Again, I paraphrase.

[danieldwilliam.livejournal.com]

The result was staggering but it hardly constitues an overwhelming mandate in traditional voting terms.

Yeah - I was staggered by the result because it was so unexpected but I think Prof C's main point was that we should have been expecting it because the system was set up (designed perhaps) to ensure a low 50% majority of seats on a mid-40's vote.

Not a huge mandate for anything radical. I'm inclined to view the response of the people as "Aw go on then SNP, you have a wee go running the country and I suppose you can ask us your wee question whilst you're at it...

... and we'll think about it and let you know."

Prof Curtice has also previously pointed out that we vote differently depending on which election it is - Westminster, Holyrood or Local Authority

And he touched on that again here. One thing he did point out that I'm still chewing over was that in opinion polls most people see Holyrood as having power in the most important areas but turnout is more than 10% higher in Westminster. Lots of reasons for that I suppose but a strange non sequiter from the electorate.

[andrewducker]

I think that lots of people still see us as being subservient to Westminster, and so voting for Holyrood is just not as important.

[danieldwilliam.livejournal.com]

I think there is something in that.

I am coming to the view that I’d like to see the power / importance pyramid inverted with local authorities having more power / influence over people’s day to day lives than Westminster.

I’m quite attracted to the idea that I could bump in to my elected resprenstative is my local pub and give them the benefit of my wisdom in person.

The other factor affecting turn out that I’m considering is how much real difference there is about policy and style of government in Scotland. There seems to be a settled position of broadly a social democracy with a bit of puritan streak about it regardless of whether we have a Labour lead coalition or an SNP government.

There’s no danger of a Tory government or the Greens getting into government so I wonder if people are that fussed over the minor differences on offer.

[andrewducker]

Is there anywhere in the world where you elect your local rep, who then elects their regional rep, who then elects their national rep?

If I elected my counsellor, who elected my MSP, that might be interesting.

[danieldwilliam.livejournal.com]

I don’t think there is. I remember watching an episode of Yes Prime Minister where that scheme was proposed. You elected a street level representative who served on the suburb level council and elected a suburb level representative from their number for the city level council and so on. The idea being that you elected people from a population who you knew very well.

Given the programme I’m sure that the idea was mooted satirically although I was never sure if the idea was being mocked or the inadequacy of the current system.

I think I’d be willing to give the idea a go. I doubt it could be worse than our current system and it might be very much better.

One of the things I would hope for from the Brave New World where robots and thinking machines do all the dirty work is that we have more time to spend on running our communities.

Little things would make a difference. I notice that the council are busy cutting down many large trees on the Meadows (presumably ahead of the windy season). Fair enough, but I notice they aren’t replacing them on a one for one basis. There is probably a good reason for this but I’d like to live in a world where both the council and I had the time to have a bit of a chat about that.

[andrewducker]

Exactly.

Of course, remembering back to our conversations about Doctors, and who knows best, I suspect that most people _don't_ want to get involved on that level. But it would be nice if some of us could.

[danieldwilliam.livejournal.com]

Again, I think you are right, most people, currently wouldn't want to get that involved.

This is in the context of a world where those who work tend to work 30-40 hours a week, those who don't work are missing their loved ones who are at work and political participation is seen as a leisure activity.

I wonder if that changes if we move to a world where we worked 20 hours a week and many more people were paid to participate in political activity.

Could I be persuade, as part of my alloted 20 hours labour per week, to spend a few hours talking about the tree planting regime in the Meadows? Oh, yes I could.

[andrewducker]

Under those circumstances, definitely :->

[danieldwilliam.livejournal.com]

I hearby declare the People's Tree Hugging and Garden Meddling Front Of the Meadows open for business and declare myself President for Life of the Candidates and Resources Committee.

[andrewducker]

I shall look forward to seeing the fruits of your labours :->

[jamie montgomerie (from livejournal.com)]

That's basically how democracy in the Soviet Union worked, I believe (at least in theory - I think in practice it was rather corrupt, with candidates basically being dictated from the above, supposedly subservient, bodies to the lower ones).