What Should A City Receive For Hosting A Data Centre?

A data centre can bring billions in investment to a city while creating relatively few permanent jobs. It consumes land, electricity and grid capacity, may require large volumes of water, and can place new demands on local infrastructure. For municipalities, the central question is therefore no longer whether data centres are economically important. It is whether the value they create is shared fairly with the places that host them.

This question is becoming more urgent as governments compete for AI infrastructure. Data-centre capacity is now treated as a matter of digital competitiveness, economic resilience and technological sovereignty. Yet national ambitions are often delivered through local planning decisions. Municipal authorities must decide whether a project justifies the land it occupies, the energy it requires and the public infrastructure it may constrain.

The conventional argument for approving a large development rests on investment, employment and tax revenue. Data centres complicate that calculation. Construction can generate substantial short-term activity, but the completed facility may employ far fewer people than a factory, logistics hub or office campus of comparable size. Tax arrangements may also allow much of the economic value to be recorded outside the host municipality.

Cities need a more precise model. A data centre should be assessed as strategic urban infrastructure, with clear obligations attached to the resources it consumes.

The Problem: A Large Investment Does Not Guarantee Local Value

The headline value of a data-centre project can be misleading. A proposed facility may involve several billion euros in capital expenditure, but much of that amount pays for servers, cooling systems, electrical equipment and specialist construction. These assets support a wider digital economy, yet they do not necessarily translate into a broad local employment base.

Once operational, a highly automated data centre may require only a limited number of engineers, security staff, technicians and facility managers. The employment density can therefore be low relative to the amount of land and power reserved for the site.

This matters because cities make choices between competing uses. Industrial land allocated to a data centre cannot be used for housing, manufacturing, laboratories or businesses that might generate more jobs. Grid capacity assigned to one large facility may not be available for other employers, new residential districts or the electrification of public transport and heating.

The opportunity cost is particularly important in regions where electricity connections are already constrained. A data centre can consume as much power as a sizeable town, while its economic benefits may extend mainly to clients and shareholders located elsewhere.

Local government may still benefit through business taxes, property-related income and development fees. However, the distribution of tax revenue does not always reflect the physical burden carried by the host city. Corporate structures can place profits, intellectual property and commercial activity in other jurisdictions, leaving the municipality with less revenue than the scale of the site might suggest.

Residents may consequently see a development that uses substantial land and energy without delivering the employment or public revenue normally associated with a major investment. When local authorities reject such projects, their decisions are not necessarily hostile to technology. They may reflect a weak local proposition.

Data Centres Are Becoming Part Of The Urban System

Data centres are often treated as private industrial developments. Their relationship with the city is now much closer than that description suggests.

They depend on high-capacity electricity connections, redundant fibre networks, water or alternative cooling systems, secure transport access and emergency services. Their energy demand can influence local grid planning, while backup generators and cooling equipment create questions around noise, emissions and resilience.

At the same time, they can support services on which the city increasingly depends. Cloud platforms, healthcare systems, transport networks, financial services and public administration all rely on data infrastructure. AI will increase that dependence further.

This makes data centres strategically important, but strategic importance does not remove the need for scrutiny. Airports, power stations and transport hubs are also strategically important, and cities usually attach detailed conditions to their development and operation.

The same logic should apply to data centres. Planning consent should reflect the project’s role within the wider urban system rather than considering only the building and the land parcel on which it sits.

The Solution: A Local Value Agreement

Cities need a formal framework that defines what the municipality and its residents receive in return for hosting a data centre. This can be incorporated into planning agreements, development contracts or a broader local infrastructure compact.

The agreement should begin with tax transparency. Municipal authorities need a realistic estimate of the revenue the project will generate during construction and operation, including the assumptions behind it. Headline forecasts should distinguish between taxes paid locally, taxes distributed through national systems and income that depends on future profitability.

Where existing tax rules leave host municipalities with only a limited share, national governments may need to revise the allocation mechanism. A city that provides land, grid access and planning consent should not be almost entirely separated from the long-term fiscal value of the facility.

Direct financial contributions are only one part of the calculation. The strongest agreements connect the data centre to specific urban needs.

Make Waste Heat A Planning Requirement

Data centres produce large quantities of heat. In many facilities, that energy is released into the environment even when nearby buildings require heating.

Recovering waste heat can improve the local case for a project, particularly in cities with district-heating networks or large public buildings nearby. Residential developments, swimming pools, hospitals, schools and commercial properties may be able to use some of the recovered energy.

The concept is attractive, but cities should be careful not to treat every promise of heat reuse as an automatic benefit. Its viability depends on temperature, distance, seasonal demand, the design of the cooling system and the cost of connecting the site to users.

A credible proposal should therefore identify the intended heat recipients, the infrastructure required, the expected delivery volume and who will finance the connection. The responsibility should not be transferred entirely to the municipality after planning approval has been granted.

Where heat recovery is technically viable, it can become a condition of development rather than a voluntary sustainability feature. Where it is not viable, the operator should explain why and offer another measurable local benefit.

Require Investment In The Electricity Network

A large data centre may require a new substation, reinforced transmission links or upgrades to the local distribution grid. These investments can benefit the wider area, but only when they are planned accordingly.

Cities should determine whether new infrastructure will serve only the data centre or also provide additional capacity for homes, businesses, electric transport and future development. A private connection that consumes scarce grid capacity offers less public value than an upgrade that strengthens the wider network.

The project should also be assessed against the timing of other local electricity needs. A city planning new housing, heat pumps and electric buses may find that a data centre competes for the same connection capacity.

Approval should therefore be linked to a clear grid-impact assessment. This should cover peak demand, connection schedules, backup generation, demand flexibility and the consequences of delayed network upgrades.

Operators can also contribute by agreeing to flexible consumption arrangements. Some computing workloads can be shifted to periods when renewable electricity is abundant or local demand is lower. This will not be possible for every service, but it can reduce pressure on the grid and improve the project’s compatibility with local energy planning.

Put Water Use Under The Same Scrutiny As Energy

Water consumption receives less attention than electricity demand, although it can become a significant local issue. Some cooling systems use large volumes of water, particularly during hotter periods when supplies may already be under pressure.

Cities should require operators to disclose expected annual consumption, peak seasonal demand, the proposed water source and the measures used to reduce potable-water use. The assessment should also consider future climate conditions rather than relying solely on historical availability.

Alternative cooling technologies may reduce water demand but increase electricity use or costs. The correct solution will depend on the location. Municipal authorities need enough information to understand these trade-offs rather than accepting a single sustainability claim in isolation.

In water-stressed areas, approval may require closed-loop cooling, recycled water or strict consumption limits. A project that depends on abundant low-cost water should not be evaluated as though that resource were unlimited.

Prioritise Brownfield Sites

The land question can determine whether a data centre fits the city at all. Brownfield sites, former industrial zones and underused commercial areas are usually preferable to greenfield development. They may already have suitable grid connections, transport access and separation from residential neighbourhoods. Redevelopment can also remove contamination or return abandoned land to productive use.

Not every former industrial site will work. Data centres require security, resilience and reliable infrastructure. Even so, cities should require developers to demonstrate why a greenfield site is necessary before allowing large areas of undeveloped land to be used.

Location also affects whether waste heat can be reused and whether new fibre or electricity infrastructure benefits surrounding districts. A site selected only because land is available may perform poorly against the city’s broader objectives.

Planning policy should therefore identify suitable zones in advance rather than assessing proposals one by one without a strategic framework.

Negotiate Digital Benefits, Not Only Physical Ones

A data centre’s contribution need not end at the property boundary.

Operators can support local fibre deployment, improve connectivity for public institutions or contribute to secure digital infrastructure used by schools, hospitals and municipal services. These commitments should address genuine local needs rather than functioning as small promotional programmes beside a much larger development.

Training can also form part of the agreement. Although the facility itself may not create large numbers of jobs, it can support apprenticeships, technical education and partnerships with local universities. The objective should be to build skills relevant to the wider digital economy rather than promising employment numbers the facility is unlikely to deliver.

Cities should avoid overstating these benefits. A training fund does not compensate automatically for weak tax revenue or excessive infrastructure demand. It is one component of a broader agreement.

Set Standards For Noise, Backup Power And Neighbourhood Impact

Data centres operate continuously. Cooling systems, generators and electrical equipment can affect nearby communities even when the facility appears visually unobtrusive.

Noise standards should reflect round-the-clock operation and cumulative development. A single facility may comply individually, while a cluster of sites creates a wider problem. Testing should include night-time conditions and low-frequency noise, which residents may experience differently from average measurements.

Backup generators also require scrutiny. They are essential for resilience but may produce emissions during testing and outages. Cities should examine the fuel used, testing schedule, emissions controls and plans for lower-carbon alternatives.

Building design, lighting, security fencing and traffic during construction should be addressed through ordinary planning controls. Strategic importance should not exempt a project from the standards expected of other large developments.

Measure Benefits After Approval

Promises made during planning should remain visible once the facility is operating.

A local value agreement needs reporting obligations covering employment, energy use, water consumption, waste heat, tax contributions and infrastructure commitments. Commercial confidentiality may limit some disclosure, but it should not prevent the municipality from verifying whether agreed benefits have been delivered.

Performance requirements can also be phased. A developer may receive planning approval subject to later milestones for grid investment, heat-network connection or environmental performance. Failure to meet those milestones should have consequences.

This protects the city from a familiar development problem: ambitious commitments are used to secure approval, then weakened once the investment has become difficult to reverse.

The Impact: A Better Bargain For Digital Growth

Cities do not need to choose between rejecting data centres and accepting them on almost any terms. They need a better method for deciding where these facilities belong and what obligations should accompany them.

The right project can strengthen digital infrastructure, expand the tax base, support energy investment and contribute to urban services. The wrong project can reserve scarce land and power while delivering little local value.

The distinction lies in the agreement. A data centre should be approved when its contribution is proportionate to the resources it uses and the opportunity costs it creates. That contribution may include tax revenue, grid reinforcement, waste-heat recovery, brownfield regeneration, water safeguards and digital infrastructure for the wider community.

National governments will continue to compete for data-centre investment. Municipalities should not be treated as administrative obstacles within that competition. They provide the land, infrastructure and political consent on which the strategy depends.

The strongest cities will use that position to negotiate deliberately. Hosting the infrastructure of the digital economy should create visible value for the place where that infrastructure is built.