Cities and the data centres investment boom

History warns us against allowing private sector interests to dictate the layout and pace of the development of critical networks, especially during an investment boom or hype cycle. Data centres are currently experiencing such a boom globally, so it is relevant for cities to consider these lessons to ensure the digital infrastructure of the future is built for the public good.

Data centres are both real estate and infrastructure. They are infrastructure as the sector has high barriers to entry (both financial and technical) and offers essential digital services predominantly via predictable, long-term contracts. They are also inherently real estate, given their need for land in specific locations, the specific nature of their architecture and their financial management (being typically leased to third-party tenants similarly to other warehousing and light industrial uses).

This dual nature of data centres requires the intervention of cities to ensure their development is beneficial to the public, not just tech companies. Investors have already started questioning the likes of Amazon, Google, and Microsoft on their data centres strategy, pointing out that transparency about the impact of new developments on local communities may help with building trust and support and how the lack of such transparency is making it hard to understand how the tech giants may meet their climate targets.

The UK stock market bubble of 1844-46, also known as Railway Mania, offers a helpful precedent for an infrastructure investment boom similar to the present data centres one. The Railway Mania investment bubble shaped the UK’s railway network—and as a consequence the country’s urban development—for the following 180 years, so it is an important cautionary tale.

Between 1845 and 1847, around 13,800 km (approx. 8,600 miles) of railway were authorised by the UK Parliament. Not all of the authorised railway was built out, as many developers did not survive the bursting of the market bubble. However, the authorised projects influenced the UK railway network until the modern day, when total railway length is circa 15,700 km (9,700 miles).

While the size and scale of the financial crash caused by Railway Mania is not comparable to the current risk posed by AI-related investment (at their peak, railway investment as a share of UK GDP reached 7%, more than what the present-day UK government spent on education and defence combined in 2023-24), there is still a risk of financial and economic fallout if this investment surge were to result in a burst bubble.

With the understandable focus on new data centres’ development on their electricity and water use, their operational vulnerability remains an underexplored risk. As these facilities become the backbone of public services and urban life, their susceptibility to natural disasters, power outages, or physical sabotage poses a direct threat to city governments and residents.

Crucially, the recognition that data centres are critical infrastructure (as the UK Government did in 2024) justifies active state and municipal intervention in their planned expansion. When a service is essential to the modern economy and public safety, its reliability should no longer be left solely to market forces.

The current geographic concentration of data centres around major cities benefits short-term corporate interests, like the inherent speed-to-market of more urban sites. However, the costs of the long-term systemic risks that accompany such concentration are unaccounted for: a natural disaster, an issue with the electricity grid, or a sabotage act in a single hub could affect the digital services of an entire region.

As shown in the chart below, in the UK data centre capacity is overwhelmingly concentrated in and around London, though this trend is common elsewhere too. In Japan, over 80% of data centre capacity is located around Tokyo and Osaka, while in India, over 40% of capacity is located in the Mumbai Metropolitan Area alone.

Cities’ planning policy should aim to spread data centre capacity across a wider geography—even if they remain relatively more concentrated around larger cities—to prevent a localised event from causing widespread disruption.

Data centres should not merely be located in a city; they must be integrated into its fabric and give back to it. Facilities should feed waste heat into district heating networks to lower residents’ energy bills and ensure their electricity demand does not drive up prices for local households, slow down other necessary development, or destabilise the local grid. Cities may also consider prioritising applications from data centres that provide support for essential local public services.

Like data centres, subsea telecommunication cables are a critical element for internet infrastructure, carrying as much as 99% of global inter-continental internet traffic.

Recent incidents of subsea cable damage have exposed a fragile global system developed primarily by private entities with minimal public oversight. The maintenance of this network is a significant pinch-point: there are only a few dozen active cable repair ships worldwide, controlled by a small number of private corporations and stationed in limited base ports.

If subsea telecommunications cables had been understood to be critical infrastructure from the start, the sector would now be less vulnerable to disruption and sabotage. This is now recognised and the UK’s Joint Committee on National Security Strategy recommends that the government adopt a clearer strategy to diversify cable routes at sea and on land (to avoid clusters of high-value targets) and fund a sovereign repair ship to guarantee speedier repairs.

Several C40 cities are already working to ensure data centre growth is sustainable and that their growth aligns with climate action planning. But many more cities will need to address the growth of digital infrastructure in the coming years. The historic parallels with railways and subsea cables are clear: instead of catching up with the security risk, cities and governments have the opportunity to shape the current investment boom to ensure the data centre infrastructure is less vulnerable and more resilient to future threats.

Listen to our latest Cities 1.5 Podcast episode: The hidden infrastructure of urban resilience. This week, host David Miller speaks to Ritwika Basu, an environmental social scientist and urbanist who researches this invisible labour in small and medium-sized cities in India. Ritwika is the guest editor of our forthcoming special journal issue Adaptation and Resilience in Cities of the Majority World: Advancing Equity and Justice in Practice.

Share

Chiara Morfeo: I recently read Experiments in Imagining Otherwise by Lola Olufemi. It mixes prose and poetry and invites the reader to literally write into the margins as well, encouraging us to invent a different future. It’s not Pollyanna in tone at all but conveyed with a refreshing dose of real talk about the challenges and frustrations that can arise when working in activist spaces. SO GOOD.

David Miller: Last Orders by Graham Swift. A novel about the journey four men make to Margate with the ashes of their friend, Jack Dodds. A booker prize winner, its a layered and nuanced story told from multiple perspectives about the relationships between men who find it hard to have conversations about difficult interpersonal issues.

Francesco Mellino: I’m reading Fantasia (Fantasy) by Italian artist and designer Bruno Munari. The book is both serious and playful, exploring the nuanced differences between fantasy, creativity, invention, and imagination. The layout and typography make it a pleasure to read, though the most fascinating thing is that it was written in the 1960s—well before computers, smartphones, and AI—yet it is perfectly modern in its style and message.