Buckets, partitions, and a quiet crisis in UK fundamental physics

The new architecture

In December 2025, before the STFC crisis became public, the Secretary of State set out a new funding framework. The UKRI envelope of around £38.6 billion is now divided into four ‘buckets’: roughly £14 billion for ‘blue sky’ curiosity-led research, £7 billion for growth, £8 billion for sovereign missions, and £7 billion for talent and infrastructure. The intent is to align research spending with declared priorities – life sciences, AI and quantum. Stated like that, it is hard to object.

But two structural problems are baked into the implementation, and they fall almost entirely on the fundamental end of physics. The first is that the headline cash settlement is only marginally larger than the last; set against inflation and the rising costs of international facilities, it is a real-terms cut to discretionary research.

The second is more subtle. The large projects that STFC oversees – accelerators, telescopes, neutron sources, satellite missions – have time horizons of twenty to thirty years and fixed running costs that lie largely outside the UK’s control. After the previous STFC crisis, in 2009, Lord Drayson introduced ring-fences known as the ‘Drayson partitions’ between three budget streams – international subscriptions, large UK facilities and the core research programme – to prevent inflation or exchange-rate movements in the first two from quietly turning the lights off on the third. Those protections have been allowed to erode without parliamentary discussion. As Michele Dougherty, the STFC Executive Chair, told the Commons select committee in March: ‘They have been eroded over time – that is clear.’

What has been cut, and how

In December 2025, UKRI removed four projects from its Infrastructure Fund pipeline: the LHCb Upgrade II at CERN, the UK contribution to the Electron-Ion Collider at Brookhaven, Critical Mass UK, and the RUEDI chemical-imaging project. Each principal investigator was informed by letter on 19 December that their project ‘has not been prioritised by UKRI at this time’. All four had already passed UKRI’s own multi-year prioritisation process and received preliminary funding: the UK’s LHCb Upgrade II commitment had been £49.4 million over ten years, of which around £5 million was already spent, and the contribution to the EIC was to have been £58.8 million over nine years, supporting seven UK universities and two national laboratories.

Three months later, UKRI announced four new Infrastructure Fund awards, headlined by £51.2 million for the National Cryogenic Facility at STFC’s Daresbury Laboratory. None has been through the same length of peer review and prioritisation. The deprioritised projects were quietly removed from the UKRI Infrastructure Fund web page; the original portfolio description, which spoke of ‘global leadership’ and the UK as ‘a partner of choice for collaboration’, was replaced with wording emphasising growth. When Parliament’s Science, Innovation and Technology Committee formally requested the minutes of the Infrastructure Advisory Committee meetings at which the four were deprioritised, the request was declined. The decision sits uncomfortably with the Haldane principle that individual project choices should rest with the funders, not with government.

Capability lost

Treaty-level agreements with CERN, the European Southern Observatory and others buy the UK a seat at facilities no single nation could build alone. But a seat at the table is not the same as a research programme. To extract any actual science from these facilities – to build the detectors and instruments, train the postdocs, run the data pipelines and write the papers – requires a properly funded domestic programme. It is precisely that domestic programme that the new framework treats as discretionary, and that STFC’s prioritisation exercise – aimed at a 30% reduction – now puts on the chopping block.

The numbers already tell the story. The Astronomy Grants Panel’s community report, published in February 2026, records that new postdoc-years funded nationally fell from around 80 in 2024/25 to 22.5 in 2026/27 – a fall of nearly 75%. Ten of the 26 research organisations that submitted four or more Small Award proposals received nothing at all. Particle physics is on a parallel trajectory. Unlike the 2010 crisis, from which the community recovered within a year, no scenario on the table restores funding to previous levels. The LHCb spokesperson, Vincenzo Vagnoni of INFN Bologna, has described the UK’s withdrawal from LHCb Upgrade II as ‘a fundamental betrayal of more than two decades of scientific leadership and a blow to the foundational stability of European big science.’

The talent pipeline is the most fragile asset in the system. A discipline that loses two consecutive postdoc cohorts loses the supervisors of the next one and the lecturers of the one after that. Capability built over decades can be dismantled in a single spending review.

Why this matters for the growth the government wants

There is a temptation to argue that fundamental physics deserves protection on its own terms – the cathedral-of-knowledge case – and to leave it there. That argument is correct, but it cedes the ground the government has chosen to fi ght on.

The Secretary of State has been explicit that the tests of success are jobs, turnover and venture capital. Fundamental physics scores on every one of those metrics, just not on a parliamentary-term timescale. The World Wide Web came out of a particle physics laboratory. MRI came out of nuclear magnetic resonance. The trapped-ion qubits that IonQ has just paid $1.1 billion to acquire (via Oxford Ionics) came out of decades of atomic and particle physics whose grant applications would not have survived a ‘bucket 2’ growth filter at the time. Quantum is the obvious example today because we can finally see the commercial endpoint. We cannot yet see the next one – which is precisely why curiosity-led research must be sustained at scale, not preserved as a rhetorical fig leaf.