A route is still required.

The Scottish Government has published a Draft Independence Referendum Bill. The question is the 2014 question. The ballot is a Yes/No paper. The date is blank. The franchise is the Scottish local-government franchise, with “qualifying foreign national” substituted for “relevant citizen of the European Union”. The Bill applies the Referendums (Scotland) Act 2020, subject to exceptions that have not yet been written. It cannot lawfully be introduced, passed or used unless Westminster first transfers competence.

That last sentence is the whole of the legal position. Everything else is politics.

What the Bill actually does

The draft is short. That is not a criticism by itself. The 2014 Act was also a conduct statute sitting on top of a prior transfer of power. This draft does four useful things and leaves the decisive work untouched.

1. The question is the right question.

“Should Scotland be an independent country?” is clear, binary and already accepted by both governments in 2014. People’s Future Scotland’s published position is that a mandate for independence is a simple majority of votes cast, in a legally held referendum, on that question. A Holyrood majority is a mandate to seek a referendum. It is not a substitute. The draft does not invent a new formulation, a multi-option paper, or a post-hoc “clarity” test. That is the correct starting point.

2. The machinery is the existing Scottish referendum statute.

Using the Referendums (Scotland) Act 2020 is the continuity option. It is the statute already on the books for devolved referendums. Campaign rules, counting, the Electoral Commission’s role and the conduct of the poll should sit there, not in a one-off political text. The draft’s weakness is that the “exceptions and modifications” are still in square brackets. A conduct Bill that does not yet say what it changes is not finished work.

3. The date is reserved to Parliament, with a ministerial power to postpone.

Leaving the date blank until introduction is honest. A date written now, before competence exists, would be a slogan. The regulation-making power to move the date later is conventional. It should remain tightly constrained and affirmative. A government that can move a constitutional poll by regulation, without a clear impossibility test, invites the charge that it is managing the calendar for party convenience.

4. The franchise is the 2020 Scottish franchise, updated for post-Brexit nationality law.

“Qualifying foreign national” is not a new invention in this draft. It is the category already used for Holyrood and local elections: a non-British, non-Irish, non-Commonwealth national who has, or is treated as having, leave to remain. The 2014 referendum used a residence-based franchise including 16- and 17-year-olds and legally resident EU citizens. This draft keeps residence and 16+ and replaces the EU-citizen category with the wider resident-foreign-national category now used in devolved elections. That is internally consistent with current Scottish electoral law. It is not, however, a small or uncontested choice for a vote on the existence of the state. Residence-based franchises for secession votes are used; they must be justified publicly in advance and held still. Changing the electorate after the campaign has begun would destroy loser consent. The Bill is right to fix the rule in statute. The Government still has to defend the rule on its merits.

None of that makes the Bill a referendum. It makes it a template for a referendum, but it still has no legal basis.

The law has not moved

The UK Supreme Court held unanimously in November 2022 that a Bill providing for a referendum on “Should Scotland be an independent country?” relates to the reserved matters of the Union and the UK Parliament. Purpose and effect both count. Calling the poll advisory does not take it outside competence. International self-determination arguments do not give Holyrood domestic legislative power. Scotland is not a colony. It elects MPs to the Parliament that holds the reserved power.

The only proven legal route remains a Section 30 Order under the Scotland Act 1998, or equivalent primary legislation at Westminster, approved by both Parliaments, as in 2012–13. That is People’s Future Scotland’s published position. It has not changed because a draft has been printed. Unilateral legislation, a “consultative” poll without competence, a “de facto” general election, or a unilateral declaration of independence would not produce a result that markets, the rest of the UK, or a large part of the Scottish public would treat as decisive. A contested mandate cannot carry currency continuity, nuclear basing, debt allocation, pensions transfer or recognition. The 2014 process worked because the rules were agreed before the votes were counted.

The May 2026 Holyrood result produced the largest pro-independence bloc the Parliament has had — 72 MSPs — and on 26 May that Parliament voted 72–55 to request a Section 30 Order. That is a political mandate to press for a lawful vote. It is not a mandate for independence, and it did not alter the Scotland Act. Prime Minister Andy Burnham has said another referendum is “off limits”. That is a political refusal, not a new legal doctrine. This draft also does not dissolve the constraint.

Publishing the template is legitimate preparation. Treating publication as delivery is not.

What the reactions tell you

The SNP presents the draft as the next step after the May motion: question set, franchise set, Westminster now to transfer the power. That is the correct sequence if the power is in fact transferred. It is incomplete if the only remaining verb is “must”.

Scottish Labour and the Scottish Conservatives have answered in the register they always use: empty gesture, day job, NHS, cost of living. Those criticisms have force when a government publishes a Bill it cannot pass and, in the same week, does not publish the fiscal, monetary and security work a Yes would immediately require. They do not answer the prior democratic question: whether there should be an agreed mechanism by which Scotland can be asked again. “Not now” is a timing argument. “Never, and no procedure” is a different claim, and it should be made plainly if that is the position.

On the independence side, the live temptation is the shortcut: treat the Holyrood majority as the decision, treat the next Westminster election as the referendum, or talk of UDI. Recent polling has found a minority willing to support a unilateral declaration; even an SNP MP has called that course “just not feasible” at present levels of support. People’s Future Scotland’s view is simpler. Shortcuts fail the legal test and the consent test at the same time. A result that half the country and the other government do not recognise cannot be the foundation of a state.

The UK Government’s line — referendums open rifts and distract from growth — is a preference about political energy. It does not explain how a voluntary union resolves a persistent, substantial demand for a vote. Preference is allowed. Pretending preference is constitutional law is not.

The test this draft does not meet

A serious independence process has two halves.

The first half is lawful process: agreed competence, a fixed question, a fixed franchise, a simple-majority decision rule, independent administration, and a prior commitment by both governments to respect the result. The draft is a partial sketch of that half. The missing piece is the transfer of power. Without it the sketch cannot be used.

The second half is the design of the state that would follow a Yes. That work is not in this Bill, and it is not optional. A Yes on this question, under this franchise, would still have to answer, in public and in numbers:

• Currency. Sterling from day one for wages, mortgages, savings and contracts. No formal currency union is on offer. Sterlingisation is legally available and operationally constrained. Lender-of-last-resort cover is not automatic.
• Fiscal starting position. The opening deficit, the treatment of UK debt, the tax base, and the rules that would bind a Scottish budget cannot be left as a pamphlet after the vote.
• Defence and NATO. An independent Scotland should remain in NATO and should align with the Alliance’s 5 per cent trajectory. Continuity of the UK nuclear deterrent at Faslane and Coulport under a long-term basing agreement is the arrangement that protects Alliance security, skilled work and a negotiated settlement. Unilateral removal is not a serious opening position.
• Borders and people. No hard border for people. Free movement with the rest of the UK under a Common Travel Area-style arrangement. Immigration policy for those arriving from outside that area is a Scottish choice and has to be specified, not implied.
• Pensions, deposits, payments. Accrued rights paid without a break. Deposit protection continuous. No gap in the systems people use on a Tuesday.

People’s Future Scotland has set those positions out at length in The Independence Debate. They are analyses, not slogans, and they are written to be wrong in public if they are wrong. A government that publishes a referendum template and does not, in the same political season, put comparable work on the table is asking the country to vote on a destination it has not described.

That is the standard. The 2014 campaign was fought on a White Paper that tried, however imperfectly, to answer the operational questions. Repeating the ballot paper without repeating the work is not progress.

What should happen next?

Four things follow from the law and from the draft.

First, keep the question and the 50-per-cent-plus-one rule. Do not reopen either as a tactic. The 2014 settlement on those points is the strongest available precedent.

Second, freeze the franchise in statute before any campaign, and defend it. If qualifying foreign nationals resident in Scotland are to vote on the existence of the state, say so plainly, publish the numbers, and do not alter the rule mid-contest. If a different franchise is proposed, propose it now.

Third, complete the conduct text. Fill in the square brackets in section 3 — “exceptions and modifications” to the 2020 Act — before anyone pretends this is a finished Bill. Campaign spending, designation of lead organisations, broadcast rules and the Electoral Commission’s remit are not details. They are how loser consent is earned.

Fourth, pursue the only route that produces a result everyone must respect. That is political agreement followed by a Section 30 Order or equivalent Westminster legislation. To make refusal expensive, you need a sustained, explicit public majority for holding the vote, plus a published design of independence that a No voter can read without concluding that Yes is a leap into fog. Drafting the ballot paper again does not substitute for either.

A lawful referendum is possible. This draft shows what the paper would look like. It does not show how the paper gets onto a legal ballot, and it does not show what the country would do the morning after a Yes. Those two questions still decide whether this is preparation or repetition.

People’s Future Scotland will test both questions in public. The Bill should be read. The law should be stated as it is. The design of independence should be argued at the level of institutions, money and security — not at the level of the next press release.

People’s Future Scotland is an independent analysis project. It is not a political party and is not affiliated to one. Reading is free. The Independence Debate and The 28 are published at peoplesfuture.scot