Communities Resist Forced Datacentres Over Energy Strain

The official account, offered with confidence by the companies concerned and echoed by governments this summer from Edinburgh to Delhi, is that datacentres are simply infrastructure - pipes and wires for a digital economy that will make everyone richer, the way railways once did, or the electric grid before them. Communities in Scotland and India who object are, on this telling, understandably nervous but ultimately misinformed; they will thank the industry later, once the AI dividend arrives.

But let us look at how this actually works. The efficient mechanism is not the delivery of a shared national benefit; it is the transfer of a highly specific cost - water, electricity, land, noise - onto a specific place, in exchange for a benefit that is diffuse, distant, and largely captured elsewhere. A datacentre near a Scottish town draws on the same constrained grid that heats that town’s houses in January. A campus outside an Indian city draws on the same aquifer that irrigates the fields around it. The company operating the facility, and the government that approved it, both experience the arrangement as a net gain - tax revenue, investment figures, a press release about “AI leadership.” The town experiences it as a net subtraction, and is then asked to regard this subtraction as evidence of national progress.

This is not new in kind, only in vocabulary. The Victorians built gasworks on the same principle: an unlovely, resource-hungry facility sited as close as planning would allow to the poor, in service of a light that shone brightest somewhere else. Nobody pretended the gasworks was decorative. What is peculiar about the datacentre is that it comes dressed in the language of destiny - not “we need somewhere to put this,” but “the future of the economy depends on it” - which makes objection to the siting feel, to the objector, like objection to the future itself. That is a considerable rhetorical achievement, and it is doing real work in Scotland and India this summer, where communities are discovering that the case against a specific facility, made in specific terms about specific strain on specific supplies, gets answered in the register of civilisational necessity.

The convention that actually governs these disputes is the same one that governed the siting of nuclear stations and reservoirs in the last century: national infrastructure decisions are made by an alliance of ministers and large capital, and local consent is solicited rather than required. The dignified process - consultation, environmental assessment, planning permission - exists and is not fraudulent. But it operates downstream of a decision already effectively taken, which is that the facility will be built somewhere, and the consultation’s real function is to determine where the friction will be absorbed with least political cost, not whether the friction is justified at all.

What should interest the constitutional observer is not the objection itself but its instrument. Communities cannot outvote a government-company alliance in a planning hearing, but they can erode the one thing the entire arrangement depends upon, which is confidence - confidence among investors that the pipeline of facilities will be delivered on schedule, confidence among governments that the growth story attached to AI will not become a story about blackouts and drought instead. Protest, litigation, and delay are not merely nuisances to Big Tech’s timetable; they are attacks on the credibility of the projection that justified the timetable in the first place. This is the same mechanism by which a run on a bank need not be rational to be devastating - it need only be persuasive to enough people that the institution’s promises will not be kept.

The gap between the dignified case for datacentres and the efficient reality of what they cost their neighbours is not, therefore, a scandal to be exposed once and then forgotten. It is a standing feature of how large infrastructure gets built under conditions of concentrated benefit and dispersed cost, and it will recur wherever the next hyperscale campus is proposed. The operational question is not whether the communities are right about the climate or the grid - reasonable people may differ. It is whether governments continue treating local consent as an inconvenience to be managed, right up until the year it is not, and a town somewhere discovers that its refusal, multiplied by enough other towns, is the only veto the system left it.