The 2014 Scottish independence referendum

The Scottish independence referendum of September 18, 2014 asked “Should Scotland be an independent country?” and 2,001,926 voters (55.3%) answered No while 1,617,989 (44.7%) answered Yes, on turnout of 84.6% of an electorate that for the first time in a Scottish national poll included 16 and 17 year olds. It was held under a UK order, made with the UK government’s agreement, that temporarily gave the Scottish Parliament power to legislate for it. In 2022 the UK Supreme Court held that the Scottish Parliament cannot, without such an order, legislate for another independence referendum.

In 2014, people living in Scotland voted on whether Scotland should become a separate country. The two governments agreed the rules first. An independent body tested the question with voters and suggested clearer, more neutral wording, which was adopted. A very high share of voters took part, and the majority chose to stay in the United Kingdom. More powers were later given to the Scottish Parliament. Whether there should be another vote is still argued about, and the courts have said who has the power to call one.

See also the 2016 EU referendum, which raised the question again for many Scottish voters, and comparative cases.

Defined terms

The 2014 referendum involves several terms from the UK’s devolution arrangements.

• Devolution: the transfer of some legislative powers from the UK Parliament to the Scottish Parliament under the Scotland Act 1998, with other matters “reserved” to the UK.
• Reserved matters: subjects on which the Scottish Parliament cannot legislate. The Supreme Court held in 2022 that a referendum on independence “relates to reserved matters,” specifically the Union and the UK Parliament.⁠⁸
• Section 30 order: an Order in Council under section 30(2) of the Scotland Act 1998 that modifies the list of reserved matters. One made in February 2013 allowed the 2014 referendum.⁠^1,8
• Designated organization (“designated organisation” in the Act): the lead campaign group for one outcome, with a spending limit of £1,500,000 under the 2013 Act.⁠⁴
• The Vow: a pledge by the Conservative Prime Minister and the then Labour and Liberal Democrat leaders, signed “on the eve of the #indyref” and featured on the front page of the Daily Record, promising “permanent and extensive new powers” for the Scottish Parliament.⁠¹¹
• Smith Commission: the cross-party commission set up after the vote to agree further powers, whose recommendations the UK government says the Scotland Act 2016 implements.⁠¹⁰

What happened

The 2014 referendum followed a dated legal and political sequence.

• January 30, 2013. The Electoral Commission published question testing it had commissioned from Ipsos MORI, and recommended changing the Scottish Government’s proposed wording.⁠^5,6
• February 12, 2013. The Scotland Act 1998 (Modification of Schedule 5) Order 2013 was made, allowing an independence referendum held by December 31, 2014 with one ballot paper offering “only two responses”.⁠¹
• 2013. The Scottish Parliament passed the Scottish Independence Referendum (Franchise) Act 2013 and the Scottish Independence Referendum Act 2013, setting the date and question.⁠^2,3
• Eve of the vote, September 2014. The Vow was published on the front page of the Daily Record.⁠¹¹
• September 18, 2014. Polling day. No 55.3%, Yes 44.7%, turnout 84.6%.⁠⁷
• November 27, 2014. The Smith Commission reported, and its recommendations were taken forward in the Scotland Act 2016.⁠^10,12
• 2016. The Scotland Act 2016 declared the Scottish Parliament and Government “a permanent part of the United Kingdom’s constitutional arrangements”.⁠⁹
• November 23, 2022. The Supreme Court held that the Scottish Parliament does not have power to legislate for a referendum on independence without a modification of reserved matters, noting that the UK Government was “currently unwilling to agree to the making of another Order in Council”.⁠⁸

How it worked

The 2014 referendum’s mechanics combined an agreed legal basis, an independently tested question, a widened franchise and spending limits.

Legal basis. The 2013 order created an exception to the reservation of the constitution so that an independence referendum was within the Scottish Parliament’s power if it met conditions: a poll no later than December 31, 2014, not on the same day as another referendum held under Scottish Parliament legislation, with one ballot paper and only two responses.⁠¹ The Supreme Court later summarized that the Scottish Parliament’s powers “were not in issue during the 2014 referendum” because of this order.⁠⁸

The question. The Scottish Government proposed “Do you agree that Scotland should be an independent country?”.⁠⁵ The Electoral Commission commissioned testing through 203 depth interviews and 10 focus groups with 62 participants between November 17 and December 15, 2012.⁠⁵ Participants found the proposed question clear, but supporters and opponents of independence alike widely saw “Do you agree” as a phrase that “might lead people towards a ‘Yes’ vote”; the alternative “Should Scotland be an independent country?” was seen as more neutral.⁠⁵ The Commission found the original “easy to understand, to the point, not ambiguous, and did not mislead voters” but had “substantial concerns about the neutrality,” and recommended the minor rewording; the Scottish Government “promptly accepted” this and the Commission’s recommendation on spending limits.⁠⁶ The Act fixed the final wording and the date of September 18, 2014.⁠²

The franchise. A person could vote if aged 16 or over on polling day, registered on the Scottish local government register or the new young voters register, not legally barred, and a Commonwealth citizen, Irish citizen, or “relevant EU citizen”.⁠³ This was a residence-based franchise: Scots living elsewhere in the UK could not vote, and EU citizens living in Scotland could. Supporters argued that those living in Scotland bear the consequences; critics argued that people with strong ties to Scotland were excluded.

Spending. Designated organizations could spend up to £1,500,000; registered parties £3,000,000 multiplied by their 2011 Scottish Parliament vote share, with a £150,000 minimum; other permitted participants £150,000.⁠⁴ The Commission had recommended limits substantially higher than the Scottish Government first proposed, to allow campaigning on a scale comparable to a Scottish Parliament election.⁠⁶

Evidence: measured results

The 2014 referendum’s official national result, as published by Scotland’s Electoral Management Board, was:⁠⁷

Young voters. Jan Eichhorn’s 2018 study in Parliamentary Affairs, using 2015 survey data comparing 16 and 17 year olds in Scotland with peers elsewhere in the UK after Scotland lowered the voting age, reports “potentially significant positive effects” of the lower voting age, clearest in electoral and non-electoral political behavior and smaller in general evaluations of politics, and stresses the role of parents and civic education in schools.⁠¹³ The study is about effects after enfranchisement, not about how 16 and 17 year olds voted in 2014.

Implementation of further powers. The UK government’s seventh annual report on the Scotland Act 2016 (May 2023) states its commitment to implementing the Smith Commission’s recommendations “in full” and that since February 2019 the Scottish Parliament has been able to legislate in every area where the 2016 Act gave it power; it also records that VAT assignment, part of the Smith agreement, was still to be settled through review of the fiscal framework.⁠¹⁰

Failures and criticisms

The 2014 referendum’s aftermath is contested by supporters and opponents of independence, and both positions are set out here.

From supporters of independence: promises and changed circumstances. Some on the Yes side argue that the Vow helped secure a No vote and that what followed fell short. In 2015, the then First Minister said the UK government was “failing to fully deliver on the vow,” citing welfare provisions, a UK veto over some Universal Credit changes, and the Sewel Convention not being given statutory force “as Smith proposed”.⁠¹¹ They also argue that the UK’s 2016 vote to leave the EU changed the basis on which Scotland had voted in 2014; see the Brexit referendum case. And they argue that a democratic mandate for a second vote should not depend on the consent of the UK government.⁠⁸

From supporters of the Union: the question was answered. Others argue that a clear majority on very high turnout settled the question, that the UK government delivered what was promised, and that repeated referendums on the same question create instability. The UK government said in 2015 that it was “honouring the Smith agreement in full” and the Prime Minister said the Vow “is being implemented in full”.⁠¹¹ Channel 4 News FactCheck concluded it was “hard to say that the the letter of the ‘vow’ has been broken,” adding that this was “hardly surprising, given how vague that promise was”.⁠¹¹

From constitutional analysts: no agreed path for a second vote. The 2014 arrangement worked because both governments agreed. The Supreme Court’s 2022 ruling made clear that, without a new order, the Scottish Parliament cannot legislate for another referendum, even an advisory one; the court accepted that such a referendum could “in that sense be described as advisory” but held it was still “not merely an exercise in public consultation”.⁠⁸ Those who want a second vote see this as a lock held by one side; those who oppose one see it as the constitution working as designed.

On the franchise. The residence rule,⁠³ was criticized both for excluding Scots elsewhere in the UK and, by others, for including non-UK citizens; defenders cited consistency with Scottish local and parliamentary elections.

Competing interpretations

Observers read the 2014 referendum in at least three ways.

A model of process. Agreed legal basis⁠¹ independent question testing,⁠^5,6 proportionate spending limits.⁠⁴ and 84.6% turnout⁠⁷ are often cited as features others should copy, regardless of one’s view of the outcome.

A decision that settled the question. On this reading, a referendum is a decision, and a clear majority should stand unless circumstances fundamentally change and both governments agree.

A decision overtaken by events. On this reading, the 2016 EU referendum and disputes over the Vow changed the terms of the 2014 choice, so a fresh vote is democratically justified.

The page does not choose among these.

What transferred and what did not

The 2014 referendum’s design influenced later practice in Scotland and the UK.

Transferred. The Electoral Commission’s approach of testing a question with voters and recommending neutral wording was applied again in 2015 to the EU referendum, where a yes/no question was replaced with Remain/Leave; see the Brexit referendum case. In Scotland, the lower voting age was extended beyond the referendum, which is the setting Eichhorn studied⁠¹³ The 2016 Act’s rule that the Scottish Parliament and Government may be abolished only “on the basis of a decision of the people of Scotland voting in a referendum”.⁠⁹ wrote a referendum requirement into statute.

Did not transfer. The cooperative route to a referendum through an agreed section 30 order has not been repeated. In 2022 the Supreme Court recorded that the UK Government was unwilling to agree to another⁠⁸

Application to this framework

The 2014 referendum informs several parts of My Vote Forward.

• Test wording with the people who will answer it. The Ipsos MORI findings⁠⁵ show bias that question authors did not intend and voters on both sides noticed. See question design.
• Agree authority before participation. The section 30 order⁠¹ settled who could hold the vote before anyone voted. The framework asks every proposal to name the body with authority; see proposals and jurisdictions.
• Define eligibility explicitly. The residence-based franchise⁠³ shows that “who counts” is a substantive choice. See identity and registration.
• Write commitments down. The dispute over the Vow⁠¹¹ shows how vague pre-vote promises produce post-vote conflict. The framework’s emphasis on recorded, checkable status changes addresses that kind of gap; see participation over time.
• Results are not the same as settlement. A high-turnout result can still leave a large minority and an open question. See representation.

Unresolved questions

• What conditions, if any, should justify a second referendum on the same constitutional question, and who should decide that they are met?
• How should pre-referendum commitments like the Vow be recorded so that their fulfillment can be checked?
• What franchise is most defensible for a decision about leaving a state: residence, birth, citizenship, or some combination?
• Did the lower voting age change how young people voted or only whether they participated? Eichhorn’s study addresses later participation, not 2014 vote choice.
• How much did the 2016 EU result change views on independence? That question was not addressed by sources verified for this page.