There is a lot of absolute nonsense out there about the Scottish Parliament’s electoral system, especially concerning list MSPs. This post attempts to neatly rebut that nonsense. It’s quite long so unless you’re super keen, I wouldn’t read it in one go, it’s more a reference to point to for when you see a specific nonsense claim; hence why each claim is numbered, for easy-peasy reference!
AMS – Additional Member System, the form of mixed-member proportional representation used for elections to the Scottish Parliament, combining FPTP constituencies with proportional regions.
FPTP – First Past the Post, the UK’s traditional voting system using constituencies represented by a single member.
List – In AMS terms, a list of candidates submitted by a party for a region. Lists in Scotland are “closed”, meaning the order of candidates is determined by the nominating party. In other countries, lists are “open”, meaning voters can select between party candidates.
STV – Single Transferrable Vote, the semi-proportional system used for elections to Scottish Councils, where voters rank candidates in order of preference.
1. AMS was specifically designed to stop the SNP Winning a Majority
This is untrue. AMS was the recommendation of the broad Scottish Constitutional Convention, not of any one party. The Labour Party, who established the Scottish Parliament, accepted AMS as a necessary compromise with the Liberal Democrats during that process. It’s undoubtedly the case that it was an acceptable compromise to Labour as the balance in favour of constituencies (73 vs 56 list) would have seemed to favour them, but that doesn’t equate to trying to lock out an SNP majority.
It’s true that AMS has made it difficult for any one party to win a majority; but that difficulty has applied equally to all parties, and is a feature of proportional systems in general. Given Labour’s long-standing strength in Scotland by the 90’s, the most obvious route to keeping the SNP out of government would have been First Past the Post.
Had the 2007 election been purely FPTP, Labour would have won 37 of the 73 constituencies, giving them their third successive majority government. Without the SNP minority AMS delivered in 2007, it’s highly unlikely the party would have won at all in 2011, never mind the majority they did manage to win.
Apart from anything else, the SNP as a party support proportional representation. They are opposed to FPTP, stating; “and we will continue to call for the first past the post voting system to be replaced with proportional representation, so that every vote and every part of the country counts.”
2. We should use FPTP because List seats mostly benefit Unionists
List seats have made up the majority of the SNP’s parliamentary representation for most of the time since the Scottish Parliament was founded. In the 1999, 2003 and 2007 elections, more than half of the SNP’s MSPs were elected via the list. As noted in the previous section, without list MSPs the SNP would not have been the largest party in 2007.
Indeed, for the first two terms, the Lib Dems won more constituencies than the SNP despite winning substantially less votes. In 1999, they actually won almost twice as many constituencies on half the vote, driving home the message that FPTP does not fairly represent voters.
List seats have also worked in the favour of the Greens and, in 1999 and 2003, the SSP, as well as the late Margo MacDonald’s three successful runs as an independent. That means the majority of list seats went to pro-independence MSPs in 1999, 2003 and 2007.
The current balance of list seats in favour of pro-Union parties is because the SNP are over-represented in constituencies. This shows AMS is working as intended to ensure broadly proportional representation, not that the list inherently benefits them.
3. Constituency MSPs have a personal mandate that List MSPs do not so we should use FPTP
Overall, this is not the case. The overwhelming majority of voters vote along party lines. It’s clearly absurd to suggest that the large increase in SNP vote in every UK Parliament constituency in 2015 had nothing to do with the party, and likewise that the SNP decline and Tory increase in 2017 was purely down to the candidates selected.
The presence of a person’s name on the ballot does not necessarily indicate personal support. In fact, constituencies are effectively closed party lists of one. Just as with the List vote, there is no ability to express disapproval of a specific candidate without voting against the party as a whole.
A good example would be the Glasgow Shettleston constituency, held by the SNP’s John Mason. Mason has a record of voting against expanding LGBTI+ rights, is opposed to abortion, and supportive of teaching creationism in schools. A habitual SNP voter who lives in Shettleston had no choice but to vote for John Mason if they wanted an SNP constituency MSP, even though many voters would find his views on these issues distasteful.
That’s not to say candidates are completely irrelevant to the result; good or bad candidates can shift things by a few points. But only in rare cases (such as the election of Dennis Canavan as an independent in 1999 and 2003) do the majority of voters cast their ballot for an individual rather than for the party they are representing.
4. Candidates should only be allowed to stand for either Constituency or List, not Both
Banning what are known as “dual candidacies” is a bad idea, with no sensible basis.
It would be a barrier to smaller parties contesting constituencies at all, as it would be necessary to find up to 73 constituency candidates plus up to 96 list candidates. They would also have to gamble with their limited parliamentary representation and potentially lose experienced MSPs if they didn’t win their most desired constituency or list seats.
Even larger parties can be badly damaged by a ban. Labour had such a ban in place for the 2011 election, except for MSPs who had notionally lost their constituencies on boundary changes. They then went on to lose most of their constituencies. As a result, a number of poorly prepared and inexperienced candidates the party didn’t actually want to be MSPs were elected.
This was so notably bad that when an expert commission was reviewing New Zealand’s similar electoral system in 2012, they referenced it (p38) as part of their justification for recommending against such a ban there. One of the big criticisms levelled against many List MSPs (often completely baselessly) is their ineptitude. It’d be quite perverse to support measures that would increase the number of unprepared MSPs if you want competent representatives. Labour did not make the same mistake in 2016.
Notable MSPs like First Minister Nicola Sturgeon may now represent constituencies, but were elected on the list in earlier elections despite standing unsuccessfully for constituencies. Only a handful of the SNP’s 1999 MSPs did not stand on both ballots.
5. Candidates that haven’t won Constituencies have been rejected by the electorate
This statement is incompatible with Proportional Representation. If we accept that FPTP is an unfair system which does not accurately represent the diversity of the electorate, failure to win seats under that unfair system cannot be taken as active rejection.
The nature of FPTP also makes it possible to win constituencies with fewer votes than losing candidates in other constituencies achieved. In Edinburgh Central in 2011, Labour’s Sarah Boyack came a narrow second with 31.9% of the vote, whilst at the same time the SNP’s Jim Eadie won Edinburgh Southern with 29.4% of the vote.
Boyack was nonetheless elected on the list; is it credible to suggest she was “rejected” by winning more support than Eadie did? Although he was not then elected on the list, Edinburgh Eastern showed an even more stark example of this with the losing Labour candidate achieving 40.1% of the vote.
Likewise, there’s a weird impact on smaller parties here. When Patrick Harvie stood for the Greens in Glasgow Kelvin in 2016, his 24.3% of the vote represented the highest voter share any Green candidate anywhere in Scotland had ever achieved by that point. Can it be fairly argued that the best Green result ever constitutes a “rejection” on par with Labour dropping to their lowest result in that constituency?
Additionally, the “rejection” narrative claims an ownership over elected representatives that doesn’t exist. Conservative MSP Murdo Fraser is a particular target for SNP supporters ire, having failed to win a constituency every election but holding a List seat. But as a Conservative MSP, Fraser isn’t for SNP supporters – or Labour, or Green, or any other party. His mandate comes from Conservative supporters, and it should be for those voters to determine whether he is a fit representative for them.
The solution to a perceived lack of mandate for list MSPs is to use Open Lists, which allow voters to specify which candidate(s) from a party list they would like to see elected.
6. The Greens are only in parliament because SNP voters gave them their list vote
Untrue. Whilst it’s highly likely some SNP voters did choose to “lend” the Greens their list vote, this claim is rooted in misremembering the run up to the 2016 election and misreading the results.
The folk memory of the Green’s 2016 campaign is of desperately begging SNP supporters to lend Greens their list vote to maximise the number of pro-Independence MSPs. That campaign never actually happened. Although the Greens did pitch aspects of their campaign at those who liked the idea of a continuing SNP government, it was on the basis of supporting Green policies and wanting to drive the SNP to be “bolder”. A party asking people who support its policies to vote for it is pretty much the definition of election campaigning. Their social media output was also focused entirely on the Green’s vision – not a single tweet by the party in the month before the election calls for tactical voting to maximise pro-independence representation.
Ironically enough, the idea of a split-ticket vote was met with varying degrees of enthusiasm and tolerance by the most vocally pro-independence wing of social media just after the referendum. They’d initially clamoured for a “Yes Alliance” for the 2015 election, and when both the Greens and SNP dismissed that, these same people could be found creating and promoting graphics like these, calling for people to vote for the SNP with their “head” in 2015, but then freely with their “heart” for pro-independence parties in 2016 in order to “maximise Scotland”; (apologies for picking on one guy, but it neatly shows the evolution)
Of course, once the SNP made it (quite rightly) clear they weren’t just going to roll over and let split voting happen, those exact same people fell duly in line behind the #bothvotesSNP message. The graphics being put out then shifted; in order to “maximise Scotland” you absolutely had to vote SNP.
Effectively, over the course of 18 months, the most enthusiastically Yes activists on Twitter tolerated or outright called for a strategy, abandoned that strategy, and then attributed it to the Greens. How’s that for weird?
Besides all that, election results – both 2016 itself and others – don’t back up the idea that Greens have no voters of their own and rely on kindly loans of list votes from the SNP. In 2011, there wasn’t even a whiff of such a “here’s how to maximise Yes MSPs” campaign because no one expected the SNP would sweep constituencies the way they did, yet the Greens achieved 4.4% of the list vote anyway. Between then and 2016, the Greens had a significantly enhanced profile that led to membership vaulting from 1000 to 9000 people, and then 6.6% of the list vote in 2016. Given that context, it’s simply not credible to suggest that increase wasn’t at least in part due to voters changing their primarily political affiliation to the Greens.
The gap between the SNP’s constituency and list vote, and their distribution of votes around the country, also doesn’t allow for every Green vote to have come from an SNP constituency voter. With 46.5% of the constituency and 41.7% of the list, there’s a gap of 4.8%. Throwing in the 0.6% of the vote the Greens won in the three constituencies they contested still leaves 1.2% to be found; and then there’s Solidarity’s 0.6% and the 0.5% RISE won to throw into the mix, which might similarly be assumed to have come from the SNP. That’s 2.3%, or a third of the non-SNP pro-Indy list vote, you couldn’t account for from the SNP’s constituency voters even if you assumed that every “missing” SNP constituency vote went to a pro-Indy party, which wouldn’t have been the case.
Further, although the SNP’s list vote decline of 2.3% might look suspiciously close to the Green’s gain of 2.2%, results in individual regions show that can’t be the explanation. A sharp drop of 8% in the North East for the SNP only saw a 1% gain for the Greens, whilst in Glasgow both parties gained, up 4.9% for the SNP and 3.5% for the Greens. Votes churn in ways a lot more complex than are allowed purely by analysing national vote shares.
There also haven’t been any studies of who voted for the Greens on the list and why, and whether that vote represented;
- A primarily Green identified voter “loaning” their vote to another party for the constituency
- A primarily other party identified voter “loaning” their list vote to the Greens
- A genuinely split voter who didn’t primarily identify with either the Greens or the other party they voted for and considered both their votes to have equal weight
So it’s impossible to state how much of the Green vote was “loaned” and how much “belonged” to the party. But the council elections in 2017 do help give a sense of actual Green support where Greens have had consistent representation for years now. Although STV allows voters to rank parties by preference, an SNP identified voter who did a split SNP-Green vote in 2016 is not likely to have given Greens their first preference the next year, instead opting to e.g. vote SNP 1, 2 and put Greens 3. That kind of vote does not register in the headline, first preference figures.
In Glasgow City in 2016, the Greens achieved 10.2% of the list vote – the following year, it was 8.7% of first preferences, but that was up from 5.6% in 2012. In Edinburgh, it was 12.9% in 2016 to 12.4% in 2017, but up from 11.4% in 2012. At the same time, the SNP dropped 4% in Glasgow and 5.2% in Edinburgh, so no one could say with certainty that those declines represent “loaned” voters returning to the SNP, versus churn in support and differences in turnout for different elections.
Indeed, summed up across areas that were directly comparable between the two elections (i.e. had Green candidates at both), there’s only marginal difference between the relative performance of each party. The total SNP vote share in those areas amounted to 83% of what it had in 2016, whilst for the Greens their vote share was 79% of what it had been in 2016. What can be said with certainty is that the Green vote held up quite well, showing that the Greens do have both a core group of voters that votes primarily Green, and that group has grown since 2011.
As this is the longest section, a TL;DR might be appropriate;
- The Greens did not campaign for tactical split-voting in 2016 – they called for people who backed their policies to vote Green.
- Ironically enough, split-voting was actively promoted by Yes activists before they fell in line behind #bothvotesSNP in 2016.
- A massive surge in membership and profile must mean that some voters changed their primary political affiliation to the Greens between 2011 and 2016.
- It isn’t possible to account for all Green votes just by the difference between the SNP’s Constituency and List vote.
- Even if it was possible, there is no way to tell which voters were primarily SNP voters “loaning” list votes to Greens, and which were primarily Green voters “loaning” their constituency vote to the SNP.
- Strong Green performances in the Council elections the year later show both the existence of a core Green vote and prove that has grown since 2011.
7. The SNP would have a majority if their voters hadn’t given the Greens their list vote
The SNP have no one but themselves to blame. As noted in section 6, it’s dubious to claim loaned list votes represent a majority of Green votes anyway, but it’s not just that. There are four constituencies the SNP had won in 2011 that they lost in 2016, with no Green candidate to “split” the constituency vote;
- Edinburgh Southern
- Edinburgh Western
- North East Fife
- Aberdeenshire West
Additionally, the Labour-SNP margin in the Dumbarton constituency in 2011 was only 5.7%, which given Labour’s vote share nationally plummeted by 9.2% and the SNP’s went up 1.1%, should on the face of it have been an easy SNP win. That makes it a fifth constituency the SNP could and should have won but didn’t.
Had the SNP won any two of these five constituencies, they’d have won a majority. That they weren’t able to convince voters to back them in those constituencies is entirely down to the SNP and their campaign.
8. Okay, I don’t disagree with Proportional Representation, but AMS is pretty bad
Yes, it is. The use of small regions means it isn’t as proportional as similar systems in Germany or New Zealand. The closed nature of lists limits voter choice. The continuing use of FPTP constituencies places even tighter limitations on that choice. Having what are perceived to be two classes of MSP can be confusing. The ability to “split” your vote between ballots has led to election campaigns focused on tactical voting rather than policy.
However, there is insufficient parliamentary support for change. Changes to the electoral system require a two thirds majority, and either Labour or the Conservatives would need to back any proposal to make that number. Neither party actually supports proportional representation. Even if it only required a simple majority, the SNP are a minority, and they recently changed their policy to support a list based system, whilst the Greens and Lib Dems support STV.
However, improvements to the current system may be more acceptable to all parties, such as introducing open lists.