Westminster has spent a decade devolving economic decision-making away from London – to local mayors like Greater Manchester’s Andy Burnham, who, now as prime minister, wants to distribute that power across the nation.
Burnham has set out his stall. In September, he told the Commons he would not back a moratorium on new datacentres, insisting instead that communities should feel their benefits.
But championing datacentres is not the same as delivering datacentres. Grid connection queues, the price of electricity, planning portals clogged with applications that may never be built – all of these are national levers, and no amount of regional enthusiasm moves them. That’s the friction at the heart of the UK’s infrastructure problem, where regional ambition runs into a national system that’s stuck.
Two 2030 deadlines compound the urgency. These are the government’s target of 6GW of AI-capable datacentre capacity – which currently looks out of reach – and its pledge of a clean power grid.
In this article we look at how speculative applications and grid logjams are breaking planning, why the datacentre boom and the climate target now depend on the same clogged queue, and whether centralised reform – maybe even a “superministry” for critical infrastructure – is the only way out.
Mayoral playbook vs local reality
Greater Manchester’s devolution rests on a simple proposition. Namely, that local decisions are best made locally. Across Burnham’s mayoralty, from 2017 until he entered Number 10 this year, that meant transport, housing and skills. A fourth prize has since joined them – digital infrastructure. He courted it directly. As mayor, he visited the site of Kao Data’s £350m, 40MW datacentre in Stockport, and folded it into Greater Manchester’s digital blueprint and the region’s pitch as an AI hub.
That ambition makes sense. Mayoral authorities are well versed in steering investment, tying it to local skills and jobs, and giving the region one voice. For the things devolution was built for – co-ordination, convening, local economic strategy – the logic is sound.
But a huge datacentre is not a Bee Network tramline. It needs land, water and power at a scale far beyond what a mayor can control. A mayor can champion a site, but cannot deliver a grid connection, set the industrial electricity price a developer pays, or clear a queue backlogged with speculative bids. The true deciders of these levers are in Whitehall, and with the regulator, Ofgem – not Number 10 North.
That is the structural limit of the mayoral playbook. TechUK, a London-based non-profit organisation, reaches the same diagnosis. That is, the real determinants of datacentre investment are electricity prices and reliable grid connection timelines, which are national problems, not regional. Devolution can capture the benefits but it can’t remove the blocks on its own.
Burnham now sits on both sides of that line. Bill McCluggage, a former Cabinet Office director of IT strategy, said a Burnham government “must reconcile localism with national priorities by accepting that datacentres are now critical national infrastructure”.
The speculative planning trap
The deeper issue isn’t the number of datacentres being proposed – it’s that it’s hard to tell which ones are real. Planning applications and grid requests can be speculative, lodged to hold a position rather than to break ground. That leaves local authorities, water companies and energy providers trying to forecast with no way to tell which numbers are genuine.
This uncertainty does real damage. These speculative schemes distort the national picture and make it harder for genuine operators to win a connection, but they also carry a reputational cost. As Katie Davies, head of energy and infrastructure policy at TechUK, said, speculative applications “have branded the sector with a reputation for unreasonable demand”.
Ofgem’s answer is to make holding a place costly. The regulator has warned that “speculative schemes taking up space can delay viable projects and create uncertainty for network planning”, and has proposed an upfront fee to claim a queue position.
TechUK backs the principle. Davies argued developers should “hit real milestones, such as demonstrating financial capability, to keep their place in the queue”, but parts company with the regulator on price. She warned that “Ofgem’s proposed fee of up to £712,500 per megawatt is set too high”. It would deter speculators, she said, but would “alienate genuine investors”.
Instead, TechUK proposes three measures: a proportionate fee, a financial capability requirement, and a “look-back test” that penalises speculative behaviour with temporary bans on applications. Together, said Davies, they would “do the job without pushing investment overseas”. The principle is simple: raise the cost of holding a place you never mean to use.
The grid and water visibility crisis
It’s the speculation that can bring the most expensive damage. For TechUK, the crowded grid queue is the most pressing threat to investment, with energy prices the deeper, structural problem. Davies was blunt about the scale. “Some developers are facing waits of up to 10 years for a grid connection, which threatens project financing and renders some schemes impossible to invest in,” she said.
Speculative requests sit in the queue with real projects, inflating the load the grid is told to plan for. Forecasts are built around demand that may never materialise, while real schemes wait behind applications that will never be built. Clear the speculation, and the real queue would be far shorter than the headline figures suggest.
Water faces the same problem, if less publicly. Companies in this industry can’t size investment in supply when they can’t separate genuine schemes from speculative ones, so planned upgrades risk falling behind schedule or stalling.
This is where the climate goal part of the problem bites. The National Energy System Operator has said Britain can build a clean power system by 2030 even if datacentre demand quadruples, but only if regional networks expand four times faster than in the past decade and connection delays are cleared. The climate target and the datacentre boom depend on the same clogged queue.
Whether it’s electricity or water, the common thread is information – the planning system is being asked to commit real capital against numbers it has every reason to distrust.
National governance and the superministry debate
The question is who in Whitehall owns this problem. As it stands, nobody does – and that has become even less clear.
In July 2026, Burnham abolished the Department for Science, Innovation and Technology and scattered its remit. Here’s what happened to it:
- AI strategy – the Cabinet Office
- DSIT digital functions – Department for Culture, Media and Sport
- Energy and the grid – the Department for Energy Security and Net Zero (DESNZ)
- Responsibility for water – Department for Environment, Food and Rural Affairs
- Planning – local government.
- Connection decisions – Ofgem and the system operator.
- Clean power target and connections reform – DESNZ; the closest thing to a single owner, but it controls neither planning nor water.
A single datacentre touches all of them, yet no single department can see the whole picture. That fragmentation is the strongest case for a dedicated “superministry for critical infrastructure”.
Such a department could set a national view for demand, grid alignment and planning timelines, as well as halt speculative applications being judged in isolation by other departments that currently only see part of the picture.
There is an irony here: Burnham’s first reshuffle split responsibility further instead of pulling it together.
But a reshuffle alone won’t fix it. Planning is already moving, with datacentres falling under the Nationally Significant Infrastructure Projects regime. This provides the basis for decisions on development consent orders that replace local planning authority permissions for major infrastructure.
However, TechUK’s Davies said reform only goes so far. “This alone will not unlock investment,” she explained. “The real challenge is addressing electricity prices and providing quicker and more reliable grid connection dates.”
This is where the bigger cliff edge sits. The UK has the highest industrial electricity prices among International Energy Agency member nations. Davies noted that “a 100MW datacentre in the UK pays four times as much as an equivalent project in the US”.
This is increasingly due to levies that fund environmental and social schemes, plus the UK’s marginal-pricing structure that means gas sets the wholesale price for all electricity.
Davies’s solution is to shift the levies into general taxation, as the government has already done in part for households.
The bottleneck is national
Devolution, Burnham-style, can help govern the upside of a datacentre boom, such as local skills and economic capture for the overall environment. The mayoral model is an advance on what came before, but the bottleneck isn’t local. Grid transparency, energy costs and the tide of speculative applications are national problems, and they can’t be solved one city-region at a time.
Whether the answer is a formal superministry or coordinated national reform matters less than whether the fixes land. These include clearing the queue of speculative place-sitters and bringing industrial power costs back to within sight of competitors.
Do that, and regional ambition has something solid to build on. Fail, and the UK risks walking into systemic infrastructure gridlock by 2030, with digital demand and climate commitments both pulling against a system that can’t tell what’s real.
