Data centres sit at the heart of the modern economy. They keep the cloud running, support digital public services, enable real-time payments and are becoming the backbone of AI.
When the sector makes its case for new capacity, it often reaches for familiar metrics: employment and gross value added (GVA). These can help illustrate scale, and they matter locally – construction activity, operational spending and on-site roles do benefit communities.
But they do not capture what is actually happening inside the data centre: the provision of essential digital infrastructure that allows other organisations to operate more efficiently, innovate faster and become more resilient. That wider productivity effect is the real prize.
Why the usual impact assessment isn’t enough
Traditional ‘economic impact’ studies typically add up direct activity – building and operating a site – and then apply multipliers to estimate knock-on effects through supply chains and household spending. While useful, there are clear limitations to this approach for data centre developers, whether it is used locally or nationally.
First, data centres are not labour-intensive in operation. Once constructed, a large facility can have a relatively modest (if highly skilled) on-site workforce compared with the size of the plot and the scale of the power connection. For local planners weighing alternative uses of a constrained site, such as logistics hubs, a jobs-led story can make a data centre look like a poor trade-off.
Second, at national level, jobs and GVA are not the basis on which policy should be decided. The fact that a data centre requires people, capital and other resources is, if anything, a cost – those resources could be deployed elsewhere. The UK’s Green Book makes this point clearly. What matters is whether an investment raises the economy’s underlying efficiency.
That is the point at which data centres start to look much more like other strategic infrastructure. A hub airport such as Heathrow is valuable not primarily because of the payroll on site, but because it improves connectivity and expands markets, generating wider economic benefits. For data centres, the equivalent question is: how does domestic compute improve the performance, resilience and innovation capacity of the wider economy? This is particularly salient in the UK, given the persistent challenge of weak productivity growth over the past decade.
There are three main routes through which data centres can support productivity.
1. Location, location, location
It is easy to assume that the ‘cloud’ is detached from geography – that a data centre in Lisbon or Lagos could provide the same service to UK users as one in London. In practice, location often matters.
-
Latency-sensitive services: In parts of finance, gaming, industrial control and emerging real-time applications, milliseconds can matter. Lower latency can improve service quality, reduce error rates and enable new product features.
-
Physical access: Firms using colocation facilities may need to visit to install hardware or respond to incidents. Proximity makes this easier and quicker.
-
Resilience and failover: Hyperscale clouds are organised into ‘availability zones’ – groups of multiple data centres that can pass workloads between one another. Rapid data replication requires facilities to be relatively nearby.
-
Network effects: Internet exchanges and dense connectivity ecosystems derive value from efficient interconnection between participants.
2. Data centres as anchors
A data centre is a large and sophisticated customer for power and connectivity. In the right setting, that can help unlock investment that benefits others.
-
Digital connectivity: Major sites often require new fibre routes, diverse paths for resilience and high-capacity backhaul. With a coordinated local approach, those investments can support wider access for businesses and communities.
-
Heat reuse: Waste heat can be valuable where there is dense, stable demand and a viable heat network. It is not a universal solution – there are challenges around the grade of heat produced by a data centre, as well as the coordination required to ensure multiple sources of heat supply and demand in an area. But where it works, it can reduce energy costs and, in some cases, ease pressure on electricity networks.
-
Electricity upgrades: New connections can be challenging for the grid. However, it is possible that, given the right circumstances and incentives, the scale of a data centre investor can help bring forward investments that unlock capacity for other users. For example, several hyperscalers have made high-profile investments in new technologies like small modular reactors.
3. National sovereignty and resilience
Some public sector and regulated uses require tight control over data, systems and legal jurisdiction. Domestic capacity can make it easier to deploy secure and compliant compute for health, defence and other sensitive activities.
Resilience also has economic value. Disruption to digital services imposes real costs, and confidence in infrastructure supports faster digital adoption. In simple terms, if organisations trust the underlying systems, they are more willing to invest in digital transformation – and that is where productivity gains arise. In an island nation such as the UK, hosting more systems locally can also reduce exposure to risks associated with international connectivity, including undersea cables.
The credibility gap: plausible mechanisms aren’t the same as evidence
These mechanisms are often cited in submissions to planners and policymakers. However, they are rarely fully evidenced or quantified, which makes it difficult for decision makers to assess their materiality.
Clustering is sometimes presented as proof. Data centres concentrate around hubs such as Northern Virginia, Dublin and London. The argument is that this concentration demonstrates spillover effects. There are two problems with that claim:
First, data centres may cluster simply because underlying conditions are favourable – proximity to large markets, strong connectivity, available power or supportive policy frameworks.
What better measurement could look like
There is a clear evidence gap. Quantifying productivity benefits will likely require a combination of approaches.
The first is bottom-up valuation of specific mechanisms. This involves practical questions such as:
-
What is the value of reduced latency for particular services or sectors?
-
What is the value of avoided downtime from improved resilience? In the energy sector, service quality and security of supply are routinely valued – similar techniques could be adapted here.
-
Under what conditions does data centre-driven fibre investment translate into broader connectivity gains, and what are the effects on business performance?
-
Where heat reuse is viable, what are the quantifiable benefits in terms of costs and emissions?
The second is top-down econometric analysis to test whether the arrival of data centre capacity is associated with measurable changes in business outcomes – investment, innovation, digital adoption or productivity – once location choices and pre-existing trends are properly controlled for.
This is challenging, because sites are not randomly placed. But that is precisely why careful analysis is essential, distinguishing correlation from causation.
What developers can do now
Even before the research base catches up, there is one immediate priority: make the productivity story explicitly place-specific.
Not all data centre developments are the same, and nor are the places that host them. The productivity narrative will differ accordingly.
-
In a mature cluster such as West London, additional capacity may reinforce an existing ecosystem – dense connectivity, internet exchanges, hyperscale availability zones and specialist skills – that underpins sectors such as fintech, media and professional services. The productivity effect here is about deepening a globally competitive hub.
-
By contrast, a site in a less connected region may act as an anchor tenant for new fibre routes, a catalyst for grid upgrades or a foundation for a regional digital or AI strategy. In such contexts, employment and GVA may also play a more visible role, particularly where traditional industries have declined.
Articulating clearly how a specific project changes outcomes in its local context will make the argument more persuasive and more aligned with the increasing policy focus on place-based analysis.
Authors
James, a Director in the Telecommunications Practice, has over 20 years’ experience advising world leading operators, regulators and policy makers across the globe.
View profile
Alex advises on energy market issues with a focus on quantitative modelling and the transition to a zero-carbon system.
View profile