Social Value for Data Centres: The Deal the Sector Hasn’t Made

IN 2019 OFFSHORE WIND MADE A DEAL. Government kept the consenting pipeline moving and offered revenue stability. Industry, through the Offshore Wind Industry Council (OWIC), committed to measurable social value outcomes on skills, supply chain and community benefit. Competitors found common ground on the things that benefited everyone.

The deal wasn’t a moral gesture. It was pragmatic. Without shared commitments, those who do the minimum set the benchmark, hardening opposition, lowering expectations, and slowing consenting for everyone.

Data centres haven’t made their equivalent deal. But they probably need to.

Not because the industry is irresponsible. Data centres are critical infrastructure. They enable productivity. Government agrees: data infrastructure was designated Critical National Infrastructure in September 2024, and the revised National Planning Policy Framework directs local authorities to consider data centre needs in plan-making and decision-taking.

But national recognition doesn’t resolve local politics. Benefits are national and digital; costs are local and physical. Communities experience land take, construction traffic, noise, and competition for constrained power and water. When the inconvenience is vivid and the benefit abstract, people want their concerns heard and their interests considered, not just managed. And unlike housing or retail, they see little return through the planning system. Most local authorities charge nothing in Community Infrastructure Levy on data centres, inheriting nil rates designed for marginal warehousing with entirely different economics. S106 terms focus narrowly on transport, with none of the education, healthcare or community contributions expected from other major schemes. The industry’s response has been better messaging. That helps. But it addresses perception, not coordination. The harder task is making local benefit a baseline, not an optional extra.

That task is becoming urgent. Grid constraints and economics are pushing development toward new locations – South Wales, the North East, Scotland – into communities where most people have never encountered a data centre, let alone formed a view on what one might mean for them. But many of these places carry the legacy of deindustrialisation. They know what it means to be valued for a resource, then abandoned when the economics shift. A data centre promising significant capex but modest employment on a twenty-year lease will be measured against that history.

The legitimacy gap won’t close by explaining the economics of cloud computing. It requires something harder: sustained presence, relationships that outlast construction, and accountability to place over time. That accountability has to be built, not promised.

Voluntary cooperation won’t be enough

The sector has already tested voluntary coordination on environmental performance. The Climate Neutral Data Centre Pact, launched in 2021, committed signatories to climate neutrality by 2030. Take-up was uneven; verification was limited; and the EU moved to mandatory reporting anyway on energy consumption, power utilisation, waste heat, and water use. The pattern is consistent: regulation follows when self-governance cannot deliver at scale. Social licence is on the same trajectory, just earlier in the cycle.

The reason voluntary coordination fails here is structural. Cooperation holds when people expect repeated interaction, when reputation travels with them, and when defection is costly. But projects are episodic. An operator can minimise commitments in one location and never face that community again. Reputational damage diffuses across the sector rather than sticking to the operator who caused it. There’s no repeated game, no travelling reputation, no shared identity strong enough to enforce norms.

The result is predictable. Without a shared baseline, a minority can free-ride on the credibility others have built, treating community impact as a local inconvenience rather than a sector-wide risk. Over time, this erodes permission to build. And the costs land on everyone, including the operators who did the right thing.

The zero-sum framing is wrong

Social licence is a commons. Every operator benefits when communities trust the industry. Every operator’s behaviour affects that trust. But no individual operator captures the full benefit of investing in that trust, and no individual operator bears the full cost of undermining it. The economics are collective, even when the economics of service delivery are individual.

Elinor Ostrom won a Nobel Prize for showing that commons don’t get managed well through top-down regulation or market mechanisms alone. They get managed well when users themselves build institutions for collective governance: shared norms, monitoring systems, and mechanisms for holding each other accountable. The sector can do this for technical standards: the ISO/IEC 30134 family provides mature efficiency metrics, and bodies like TechUK and the Data Centre Alliance coordinate effectively on policy and advocacy.

But it hasn’t done it for social license.

Responsible behaviour isn’t a cost to be minimised; a sector with credible social licence gets fewer contested applications, faster approvals, more predictable delivery. These are benefits no single operator can secure alone, but that all operators share when the sector acts collectively.

This is co-opetition: compete on technology, efficiency, reliability, service and price; cooperate on the conditions that let everyone operate. Legitimacy isn’t a differentiator in the way uptime is. It’s shared infrastructure.

Offshore wind understood this. Competitors who fought for CfD contracts found common ground on skills pipelines, supply chain development, and community benefit frameworks. They built OWIC as the vehicle for that collaboration, and used it to present government with a coherent proposition.

Data centres need to make the same distinction. Transparency frameworks, baseline community commitments, verification standards. No operator gains advantage by being the only one who reports honestly. Everyone gains when the sector can demonstrate collective credibility.

What a deal means in practice

Social licence isn’t just a reputational commons. It intersects with statutory regimes that already govern how local benefit is secured: Section 106 agreements, Community Infrastructure Levy, planning conditions, and potentially the NSIP regime for larger facilities. A sector-led baseline that ignores these hooks will float above the system that actually grants permission. One that connects to them becomes durable infrastructure rather than voluntary aspiration.

Offshore wind’s deal was explicit: government kept the consenting pipeline credible; industry, through OWIC, committed to measurable outcomes that protected its licence to operate. For data centres, the equivalent deal is straightforward: a place-based social value framework, agreed by the sector, that makes how local benefit is designed and verified consistent and comparable, while the commitments themselves respond to local need.[1]

The direction of travel is clear. The Social Value Commission, launched in February 2026 by major investors including Barratt Redrow, E.ON, Heathrow and Mitie, reflects a growing recognition across sectors that local communities must share in the benefits of private investment. But recognition isn’t resolution. A cross-sector commission can set the climate; only a sector-specific framework can set the baseline. Data centres are notably absent from that table.

QTS’s £110 million community commitment at Cambois[2] shows what’s achievable. It is the largest community investment package attached to any UK data centre scheme. But it emerged from a specific land transaction,  Northumberland County Council amending its buy-back option on the former Britishvolt site,  not from a sector baseline. The terms were negotiated around unique circumstances that won’t repeat elsewhere.

Government already has routes to approve data centres in the national interest. The harder problem is doing that repeatedly and credibly in places carrying the disruption and resource constraints. A place-based social value standard gives government a clearer backstop: a consistent baseline it can point to, enforce, and defend, so approvals aren’t justified by ad hoc promises and negotiated one site at a time. For industry, it creates expectations that travel across councils and protects responsible operators from being undercut by the sector’s worst behaviour.

Data centres don’t have years. The system is being unlocked and re-shaped now, across planning routes, grid access, water, cyber resilience and place-based incentives. The question is whether those emerging terms are shaped around a sector-led baseline, or assembled under pressure without one.

The risk isn’t regulation. It’s knee-jerk regulation without nuance, written in the absence of a credible sector baseline. One-size-fits-all conditions, crude thresholds, and criteria that can’t distinguish between materially different facility types and impacts. Offshore wind shaped the terms because it moved first. Data centres are running out of room to do the same.

As John Booth says “If you cite yourselves as the fourth utility, don’t be surprised when you’re regulated like one.”

The question isn’t whether governance is coming. It’s whether the sector arrives with a coherent framework – shared standards, verified commitments, genuine accountability to place – or waits until the framework is written for it.

Offshore wind shaped its own terms because it moved first. That option has an expiry date.

[1] Our Consumption to Contribution report sets out how the Four Capitals Framework can structure that assessment.

[2] https://www.northumberland.gov.uk/News/2025/Mar/Multi-billion-pound-data-centre-campus-gets-green.aspx