Britain’s Constitutional Crisis: Nobody Is Driving the United Kingdom

On Monday in Cardiff, the leaders of the SNP, Plaid Cymru and Sinn Féin signed an agreement asking the British government to “prepare, plan and facilitate constitutional change” across Scotland, Wales and Northern Ireland.

They were careful about the capacity in which they met. They came as party leaders, not as representatives of their governments. The technicality kept the gathering one step short of official intergovernmental action. It did not make the picture less strange: for the first time, the First Ministers of all three devolved nations lead parties committed to constitutional futures outside the United Kingdom. (The Guardian)

There is no immediate secession wave. Recent polling puts support for independence at roughly 47% in Scotland, 36% in Northern Ireland and 32% in Wales, and Plaid Cymru has ruled out an independence referendum during this Senedd term. (The Guardian) The Union is not collapsing on schedule.

Cardiff exposed a slower problem. Britain has no settled answer to who owns the decision when one of its nations wants to reconsider membership. That ambiguity was manageable while independence was hypothetical, distant or confined to one part of the country. It gets harder to sustain when three devolved governments are led by parties that regard the present constitutional settlement as provisional.

The same design flaw runs through British government more broadly. Westminster keeps the sovereignty. Responsibility is scattered. Governments inherit promises that last for decades, make decisions whose costs arrive years later and are judged on political horizons that can shrink to months.

Britain has become a temp-agency state.

The temp-agency state

When Andy Burnham entered Downing Street in July, he noted that he was Britain's seventh prime minister since 2016. He then promised a ten-year plan for the country. (GOV.UK) The arithmetic should make any ten-year promise sound provisional.

Before Burnham arrived, Britain had reached its tenth home secretary in ten years. The Institute for Government found that more than half of ministers changed jobs or left government in Keir Starmer's first major reshuffle. Whitehall itself is hardly motionless: civil-service turnover was 8.9% in 2024-25, after 12.7% the year before. (Institute for Government)

The permanence sits less in the people than in the commitments. Departments carry forward budgets, pension promises, legal settlements, infrastructure plans, treaties and debt long after the office-holders who inherited or approved them have gone.

David Cameron called the Brexit referendum and resigned after losing it. (GOV.UK) Liz Truss lasted 49 days. Her premiership ended; the market consequences became somebody else's problem. The pattern is institutional. Decisions can outlive decision-makers by decades, while the political price of postponement often lands somewhere else.

That changes behaviour. A minister who imposes pain now so that a successor inherits a more stable state may be doing good public policy and terrible electoral politics. A minister who pushes the pain beyond the next election often faces the opposite incentives.

Britain has arranged too much of government around that mismatch.

The promise that outlives the politician

The state pension shows the mechanism in unusually clean form.

The basic and new state pensions rise each year by whichever is highest: earnings growth, inflation or 2.5%. A large increase becomes the base for the next one. The ratchet rarely runs backwards.

The Institute for Fiscal Studies estimates that state-pension spending is now about £16 billion a year higher than it would have been without the triple lock. State pensions consume roughly 4.9% of GDP, up from 3.6% twenty years ago. The Office for Budget Responsibility's central long-term projection puts the figure at 7.7% by the early 2070s, with the triple lock accounting for about 1.6 percentage points of that increase relative to earnings uprating.

There is a defensible social case for protecting pensioners from inflation and from falling behind wages. The political problem is the timing. A government that changes the rule takes the punishment now. Much of the fiscal reward arrives later, possibly under another party.

So every government can find a reason to wait.

The OBR's July assessment says almost all of its scenarios eventually put public debt on an unsustainable path, with the baseline turning sharply upwards in the 2040s. It also says earlier action would probably cost less than delay. The warning and the incentive point in opposite directions.

Bond markets have started charging more for long duration. On 8 September Britain sold 30-year debt at a yield of 5.8168%, the highest at a gilt auction or syndication since comparable Debt Management Office records began in 1998. Demand was strong, so the auction was no warning of imminent funding failure. Britain simply paid more for thirty-year money. (Investing.com)

Political tenure is shortening while public liabilities lengthen. Gradual costs are easy to postpone and hard to reverse.

One union, three answers

The constitution contains the same mismatch.

Northern Ireland has a legal route for testing whether it wants to remain in the United Kingdom. Under the 1998 settlement, the Secretary of State must order a border poll if it appears likely that a majority would vote for Irish reunification. The trigger leaves room for judgement, but the mechanism exists in law. (House of Commons Library)

Scotland has no equivalent standing right. In 2022 the Supreme Court ruled unanimously that Holyrood could not legislate unilaterally for an independence referendum because the Union is reserved to Westminster. The 2014 vote happened after London temporarily transferred the necessary power through a Section 30 order. A renewed Scottish request after the May 2026 election was rejected. (UK Supreme Court)

Wales sits in much the same position. The union of England and Wales is reserved, so transferring referendum competence would require Westminster's participation. Plaid Cymru is asking for something more basic than an immediate vote: the right for Wales to determine the timing and terms of a future one. (House of Commons Library)

One state therefore has three different constitutional answers to the same problem of consent.

Burnham has already tested the edge of this arrangement. He suggested that Scotland should be treated more like Northern Ireland once public opinion reaches sufficient clarity. John Swinney promptly wrote asking him to turn that idea into a legal mechanism. (Scottish Government)

The dispute has moved one level upstream. The immediate fight is over the trigger itself, over who gets to decide that Scotland may decide.

Devolution was built to absorb disagreement while leaving final sovereignty at Westminster. For a quarter-century, that ambiguity worked as constitutional shock absorption. Now all three devolved nations have First Ministers whose parties treat the present settlement as temporary. A constitution that avoided writing clear exit rules has reached the point at which those rules matter.

The invoice cuts both ways

Nationalists still have to face the arithmetic.

Scotland's latest official accounts put its notional fiscal deficit at £25.3 billion in 2025-26, or 10.9% of GDP, against 4.2% for the UK overall. Revenue per person was almost identical to the UK average. Spending benefiting Scotland was about 14% higher per person. The IFS concludes that a deficit on that scale would be unsustainable for a newly independent Scotland without higher taxes, lower spending or materially stronger growth.

Wales begins even further from fiscal autonomy. Around 79% of the Welsh government's day-to-day funding in 2025-26 came from the UK government, predominantly through the block grant. (IFS)

Irish reunification has its own bill, and nobody can price it with confidence. In July, Irish finance minister Simon Harris told the Dáil that serious estimates ranged from about €3 billion a year at the optimistic end to as much as €20 billion under more pessimistic assumptions. The outcome depends on growth, pensions, public-sector pay, UK contributions and the design of any transition. (The Irish Times)

Those numbers are powerful arguments for the Union. They make clear that independence is not a cost-free expression of identity. Scotland would face hard choices over tax, spending and growth; Wales relies heavily on transfers from Westminster; Irish reunification would force Dublin to make choices that opinion polls can postpone.

The same spreadsheet also creates a vulnerability for London. The Union increasingly sells itself as insurance: common borrowing capacity, transfers, pensions, welfare and defence. Insurance is valuable because the underwriter is trusted.

When Westminster warns that an independent Scotland's books do not add up, the OBR is simultaneously warning that Britain's own long-run debt path eventually becomes unsustainable without policy changes. The two problems are not equivalent. A newly independent Scotland would face an abrupt transition and much tighter financing constraints, while the UK has a larger tax base and established borrowing capacity. Still, the comparison is politically unavoidable.

If unionism becomes a spreadsheet argument, Westminster's spreadsheet becomes part of the product.

That is a harder problem than today's polling suggests. Support for independence can remain below 50% while support grows for the proposition that London should not have the final say over whether a vote is allowed. Cardiff was a claim about jurisdiction over a future choice. It did not need to be a secession timetable to matter.

When territory enters the bargain

The constitutional strain is beginning to register abroad.

Donald Trump spent this weekend in Ireland publicly endorsing Irish reunification, saying he would “love” to see it and brushing aside Burnham's objections to a border poll. An American president openly advocating an outcome on one of Britain's most sensitive constitutional questions is a break with the traditional White House posture of neutrality. (The Guardian)

Northern Ireland is only one pressure point. In April, Reuters reported that an internal Pentagon document floated reviewing the American position on Britain's Falkland Islands claim as a possible way to pressure NATO allies that Washington believed had failed to support US operations against Iran. By the end of August, Trump was publicly saying that he reviewed every position when asked about the Falklands. (Reuters via Investing.com)

Argentina has meanwhile stepped up action against companies involved in Falklands oil development. Israel's national security minister, Itamar Ben Gvir, urged Benjamin Netanyahu to recognise Argentina's claim and sanction Britain. That was a minister's proposal, not Israeli government policy. The proposal still tells us something about how Britain's territorial sensitivities are entering other people's political calculations. (Euronews)

These episodes are separate. There is no evidence of a coordinated strategy against Britain. Their common feature is a change in what outsiders seem willing to put on the table.

An American president can use Irish unity as public pressure. Pentagon planners can float Falklands sovereignty as pressure. Argentina can see room to intensify pressure. An Israeli minister can reach for the same territorial dispute in retaliation.

The borders have not moved. The apparent price of probing Britain's constitutional seams has fallen.

Geopolitical decline often begins as a change in other people's risk appetite. A state can keep every inch of territory while allies and adversaries start treating its red lines as negotiable.

Britain can survive all of this. It remains nuclear-armed, a permanent member of the UN Security Council, a major economy and a core NATO power. The nationalist parties still lack decisive referendum majorities. Fiscal ties across the Union are deep. Its courts, civil service, armed forces and financial institutions remain durable. Political turnover can reflect democratic adaptability as well as disorder.

Those facts make imminent disintegration unlikely. They do not solve the problem Cardiff exposed.

Westminster holds final constitutional sovereignty. Devolved governments hold democratic mandates of their own. The Treasury holds the arithmetic. Voters hold consent. Bond markets put a price on delay. Foreign governments are beginning to test the seams.

The driver's seat is crowded, and the controls do not converge in one accountable hand.

Cardiff made the missing rule visible. Britain can say who governs Scotland, Wales and Northern Ireland today. It still cannot say, in one coherent constitutional language, who gets to decide whether they remain tomorrow.

A country can improvise around that question for a very long time. It cannot make the question disappear.