A large building is not an economic strategy
When a developer announces a billion-dollar data centre, headlines often treat that amount as a benefit to the host country. It is an investment commitment, not a measure of local prosperity. Construction contracts may go to outside firms. Imported servers represent spending, but not necessarily a domestic supply chain. Meanwhile, the centre can consume local electricity, water and grid capacity for decades.
Follow the money and the resources
The public should distinguish one-off construction jobs from long-term jobs and ask how much of the capital spending stays locally. Tax concessions, public land discounts and utility upgrades need to be accounted for alongside projected tax receipts. A centre buying electricity at a favorable rate might still be a good investment, but the underlying arrangement should be public and evaluated against what other customers pay.
What the country actually gains
The strongest case comes when investment improves local digital capability: reliable computing for domestic enterprises, workforce training, research partnerships, local supplier contracts, waste-heat reuse where practical and clear long-term tax contributions. None of these benefits should be inferred merely from the number of megawatts announced.
Measure outcomes, not launch announcements
Require public reporting on actual electricity and water use, permanent local payroll, benefits delivered, incentive costs, network upgrade allocation and environmental performance. Compare the figures against the forecasts used to secure planning consent. The useful accounting begins when the ribbon-cutting photographs are forgotten and the construction workforce has left.
These explainers separate announced plans, recorded outcomes and unresolved questions. Find individual facility sources and submit corrections in the Observatory.
General background: IEA, Energy and AI. For site-specific claims, consult each facility's source records.