Photo of tools of measurement

Implications of varying mandate definitions for Indyref2 legitimacy at Westminster-level elections

Published: 6 July 2022
Author: Mark Shephard

Mark Shephard, Senior Lecturer at the University of Strathclyde, examines the scenario of the SNP campaigning solely on the issue of independence in the next General Election and variations in mandate definitions across the different political groups and academics.

If The UK Supreme Court rules that there is no legal basis for a second independence referendum, then the SNP have made it clear that they are going to campaign solely on the issue of independence in the next Westminster General Election (likely Spring 2024). Assuming we end up in this scenario (and we might not), this is quite a clever tactic as it puts the Unionist parties in a very tricky position.

Given the SNP’s majority seat performance at Westminster elections post-2014 (56/59 seats in 2015; 35/59 in 2017; and 48/59 in 2019), under current party competition arrangements it is going to be virtually impossible for the SNP not to gain a vast majority of seats. Under first-past-the-post elections, you win a seat if you get a plurality (and not necessarily a majority) of votes. This can be a majority of votes (SNP + Greens = 51% in 2015), but invariably it is a minority of votes (SNP + Greens = 37% in 2017 & 46% in 2019) that wins the day because the typically bigger unionist vote splits between three parties (Conservative, Labour, and Liberal Democrats) and not one main party.

Think of a cake. One party (SNP) gets under half the cake, the other 3 parties (Conservative, Labour, and Liberal Democrats) have to share the typically slightly larger half of the cake. Apart from some regional & local voting constituency outliers, the SNP ends up with virtually all the seats because their (often minority) slice of a constituency cake is invariably larger than the divided slices for the three unionist parties.

And this is why it is tricky for the unionist parties. If they carry on as they are, the SNP will win the vast majority of seats and likely claim a mandate for independence and/or a subsequent legal referendum. The unionist parties have several options including:

  1. Campaign as they normally do and lose most seats, BUT campaign on a referendum mandate equalling % of vote (and/or even % of total electorate voting SNP/Green) not % of seats.
  2. Strike deals with each other (e.g. the unionist party that came either first or second in a seat is the sole party on the ballot against the SNP).
  3. Combine into one big pro-union party.

In terms of what is politically doable, option 1 is risky if yes gain momentum like last time, but easier than option 2, and option 2 is easier than option 3. Indeed, option 3 seems near impossible given the fallout Labour had working with the Conservatives in the run-up to the independence campaign in 2014.

A further option is to not engage/play into this narrative, but that will be hard given the agenda-setting terms set by the largest party in Scotland and the main party of Scottish Government, not to mention public opinion being noticeably swayed by polling questions primed with majority of seats mandate logic.

However, we might also want to factor in slight costs for the SNP as well. We know that approximately one in ten SNP voters do not favour independence (see for example, Johns et al., 2020), so if independence is the rationale of the next General Election in Scotland, the SNP’s vote share could drop to the low 40% range. Given this, and the lack of majority votes even when paired with the Greens, my guess is the pro-independence side will settle on measuring mandate in terms of % seats if they don’t get a majority of votes, while the pro-union side will be more likely to measure mandate in terms of % vote if they get a majority, or % vote of total electorate if they don’t get a majority of the vote.

For academics, survey data will need to keep up with this tussle over mandate definition. We will need to be cautious that we are not priming and so skewing views on a referendum based on question wording preference for seats over votes. It will be interesting to see the different effects re: referendum legitimacy of wording “% seats” versus wording “% votes” (and even versus wording “% vote of the total electorate”). At the moment the emphasis in the Scottish Election Study (see 2021 post-questionnaire) is on discerning the differences in support for a referendum between an open-ended ‘mandate’ question versus number of SNP seats versus number of pro-independence party seats.

Above all else, whatever happens, the voters need to know how the parties are going to measure mandate, and ideally and hopefully they can all agree a consensus definition for the sake of the legitimacy of any outcome. At the moment, there seems to be a lot of inter and even intra-party confusion out there with some politicians talking about majority of seats (e.g. in the Scottish Parliament elections 2021) equating with the will of the people/democracy, while some talk about it being majority vote given the nature of how a referendum works.

Mark Shephard is Senior Lecturer at the University of Strathclyde. 

Photo by CHUTTERSNAP on Unsplash