If the break-up of the United Kingdom doesn’t bother you, read no further. You are unlikely to be concerned by a clear majority for independence in Scotland’s new parliament - even if most voters choose parties that want to keep the union.
But those of us who want the UK to stay together should prepare for a bitter battle between Westminster and Holyrood. It will not just decide the future of these islands. It could raise big questions about the way Scottish democracy works.
The election for a new Holyrood parliament takes place on May 7. If a clear majority of MSPs support independence, the SNP will claim a mandate to demand a new referendum on independence. They might succeed beyond their wildest dreams.
Two things have come together to make possible a wide divergence between the views of voters and the ambitions of those they elect. Both are caused by a voting system that was designed to prevent such divergence but this time may encourage it.
Holyrood’s parliament has 129 members. 73 are elected for local constituencies by first-past-the-post. The other 56 are chosen from lists in each of Scotland’s eight regions. This is a top-up process designed to ensure a roughly proportional outcome overall. Each elector has two votes – one for their local MSP, the other for a regional party list.
Here are the two ways in which this May’s election could produce a perverse outcome – and the greater the number of pro-independence voters who act rationally, the more perverse the result.
The first concerns the election of constituency MSPs. Labour’s success two years ago across Britan – almost two-thirds of MPs with barely one-third of the popular vote – could be matched, and even exceeded, by the SNP in six weeks’ time.
The latest Nortstat poll for the Sunday Times – and it’s fairly typical of all recent surveys – shows the SNP winning 34 per cent of the constituency vote. This is likely to give it around 57 seats, with just 16 divided between Reform, Labour, Conservative and Liberal Democrat. That’s first past the post for you.
This would leave the SNP eight seats short of the 65 it needs for a majority in Holyrood. Would they get them from the 56 regional MSPs that will also be elected?
The short answer is no. It arises from the second feature of Scotland’s voting system: its regional list seats. Their purpose is to correct distortions in the constituency section. Scots cast a second vote that decide how many of these 56 seats (seven in each of eight regions) each party wins.
In 2021, the SNP won 40 per cent of the regional vote but just two seats. This was because it had done so well in the constituency section; the allocation of regional seats took account of this. The SNP ended up with 64 seats overall, just short of a majority. In broad terms, democracy worked: just over half of all Scots voted for a pro-independence party (SNP, Green, Alba). The SNP (64 MSPs) and Greens (8) together secured a modest pro-independence majority. (Alba won no seats and has now withdrawn from the fray.)
This year could be very different. This is because the SNP is almost certain to win no regional seats at all. According to Norstat just 30 per cent will cast their regional vote for the SNP, ten points down on five years ago. An average of all recent polls gives the same number. On that level of support – or anything like it – and given how many constituency seats the SNP are likely to win, they have no chance of winning ANY regional seats.
For those who think, “so what, that’s just numbers; what about real politics?”, here’s my answer. Let me put it in Trump-style capitals.
EVERY REGIONAL VOTE FOR THE SNP WILL BE A WASTED VOTE. IT CANNOT ELECT A SINGLE EXTRA SNP PARLIAMENTARIAN. EVERY VOTER WHO WANTS TO GUARANTEE A MAJORITY FOR INDEPENDENCE AT HOLYROOD SHOULD CAST THEIR REGIONAL VOTE FOR THE GREENS, NOT THE SNP.
We can go further. SNP supporters who simply want to punish Keir Starmer, Kemi Badenoch and Nigel Farage (and who don’t care either for their Scottish counterparts) should also vote Green in the regional section, in order to minimise the number of Labour, Tory and Reform MSPs. Anyone voting SNP in their regional list will hurt the Greens and help the cause of the unionists they dislike.
Yes, the voting system is that absurd. However, there’s nothing new about voters switching to their second choice to achieve a larger objective. Two years ago, tactical voting punished the Tories across Britain. The equivalent for pro-independence Scots next month can be regarded as strategic voting: using people power to ensure the greatest number of pro-independence MSPs in Holyrood. It’s not just that Green is a rational choice for SNP supporters in the regional section: it’s their ONLY rational choice.
Let’s put numbers to this. The following table shows
a) John Curtice’s estimates of how Norstat’s vote shares translate into the total seats likely to be won by each party; and
b) My estimates of the seats won on the alternative assumptions that either half or all SNP supporters switch to the Greens to maximise the number of pro-independence MSPs.
The first point is that on each scenario, total support for unionist parties is 52 per cent, and for pro-independence parties 42 per cent. Yet on conventional assumptions, there would be an 11-seat majority for leaving the UK. But this majority widens substantially if SNP supporters cast their regional vote for the Greens: to 31 (80-49) if half of them do, and to 49 (89-40) if they all do.
Those figures are estimates. All polls are subject to margins of error. Translating votes into seats can also be an uncertain business. We cannot be certain that conventional assumptions will produce a pro-independence majority at Holyrood. Together, the SNP and Greens need to fall just six seats short of Curtice’s projections for the unionist parties to win a collective majority. This adds to the incentive for SNP supporters to give their regional vote to the Greens. They have the power to guarantee a majority for independence – and if enough of them make the strategic switch, they will ensure a landslide.
How should London respond if there is a clear majority at Holyrood for a fresh referendum on independence, however it is achieved? One obvious retort is to look at votes, not just seats. As we have seen, and unlike five years ago, most Scots seem likely to reject the pro-independence parties.
However, Labour in particular should be wary of deploying that argument. It would undermine its claim to have a mandate to take big decisions about the UK. Two years ago. Its 34 per cent vote share across the UK gave it 63 per cent the MPs elected. What’s sauce for the Westminster goose is sauce for the Holyrood gander.
While that argument heats up, one conclusion stands out. Scotland’s parliament has a voting system that, like that for Westminster, is no longer fit for purpose.
What should replace it? The case for a truly proportional system is not as clear-cut as its advocates claim. But if that is the aim, the way to avoid today’s perverse incentives is to adopt the new system that comes into force next month for electing the Welsh Senedd. (Single-member constituencies have been scrapped. Six members will be elected from party lists in each of 16 regions.)
That’s a debate for when Scotland’s votes have been counted and the skew in the results becomes clear. It’s bound to spill over into the arguments about how we elect MPs to the House of Commons. Prepare for a bumpy ride.
NOTE For those who want to explore alternative assumptions for strategic voting, the key thing is the formula for allocating the seven list seats for each of Scotland’s eight regions. This is how the Scottish Parliament website sets it out.
The d’Hondt system allocates additional seats to political parties or independent candidates according to the number of regional votes cast for that party or independent candidate divided by the number of seats (constituency and regional) already gained in that region, plus one.
The party with the highest total after this calculation gains one additional member. The divisor for that party, or individual, is then increased by one (because of the seat won in the first round) and the calculation is repeated. The starting point for the calculation is again the total number of regional votes. The party, or individual, with the highest total wins a seat. This process is repeated until all seven regional seats are allocated.
Funny thing is, when you ask a Unionist what is the purpose of the union for the greater good, they can never answer. Independence sentiment has grown because Westminster has ruled badly: this was most obvious when the Tories were in power. Huge questions such as Scotland’s & N Ireland’s exclusion from the hardline Brexit stance taken from Westminster.
Peter is spot on in exposing the way the AMS system can be gamed. It would be ironic to see pro-union pro first past the post politicians drawing attention to vote share rather than seat share!
One comment though which is that the list PR system, to be used in Wales next month, while less ‘gameable’ has drawbacks of its own, most notably the lack of local, personal representation- a point Peter has raised in the recent past. List PR leaves you only able to vote for a party not a person and in a multi member (six in the case of Wales) semi-regional constituency.
Far better to go for the single transferable vote, (stv) which while also working in multi member constituencies expressly gives voters the right to vote for individuals - and also choosing between party candidates. Much more choice and connection for the voter.
And - it so happens! - stv is already in use in Scotland, for local elections. So Scottish voters are familiar with the system and it would also bring consistency to the voting systems Scotland uses.