Who Pays for the Electric Century — and Who Owns It?
The electric build-out needs vast capital. Ownership and loss allocation will decide who benefits and whether society keeps investing.
Electrical infrastructure is built for years before it creates value and operated for decades before that value is fully known. This makes it capital-hungry and political.
Power plants, grids, storage, factories and data centres demand high upfront investment. Their future revenues depend on demand, regulation, interest rates, technology and social acceptance. Those financing an asset today are betting that tomorrow’s order will still honour its assumptions.
The first distinction is:
private return on capital ≠ economy-wide system return
A grid can be indispensable to an economy while offering an unattractive return to its owner if tariffs are politically constrained. A railway, power station or chip plant may have strategic value despite weak direct cash flows. Conversely, a company may earn high profits from controlling a toll gate without making the wider system more productive.
This gap justifies state intervention — and creates its dangers.
Governments can carry long-term risk, coordinate networks, fund research and enable strategic capacity rejected too early by private markets. They can also hide costs, protect politically favoured projects from correction and shift losses to taxpayers or savers. The market has no complete view of social value. The state has no automatic access to truth.
The second distinction is:
financing ≠ viability
A project is not economic merely because somebody will lend to it. Low interest rates can support poor assumptions for a long time. High equity valuations can mobilise capital without guaranteeing future cash flows. Public guarantees may reduce financing costs, but they do not eliminate risk. They move it.
The AI age makes this mechanism unusually visible. Hyperscalers, utilities and infrastructure investors are committing capital to data centres, chips, power plants and grids because they expect enormous future demand. Some of that wager will pay. Some will produce excess capacity, impairments and changes of ownership.
The possible correction follows a familiar chain:
doubt about returns → lower valuations → dearer capital → cancellation or sale → new owners
The physical asset does not disappear. A data centre, fibre line or power plant may continue operating after the failure of its first owner — often with a lower capital cost for the buyer. A technological revolution can advance while its earliest investors lose money.
That is neither a reason for euphoria nor mockery. It requires three questions to be kept separate:
- Is the technology useful?
- Is the specific investment sensibly priced and financed?
- Who carries the loss when the assumptions prove wrong?
The third question leads to ownership. Those who own grids, power plants, data centres, data and platforms receive more than cash flows. They control access, priorities and often information. Infrastructure funds, technology companies, utilities, governments and sovereign wealth funds therefore compete for the same nodes.
Private owners can build efficiently and accelerate innovation. But when one actor controls generation, data centre, cloud, model and customer access, vertical power grows. Public ownership can protect strategic goals, but may strengthen political control and reduce transparency. Broad and regulated ownership can distribute power, but requires capable institutions.
Households are owners and financiers too, often without noticing. They pay through electricity tariffs, taxes, pension funds and the prices of energy-intensive goods. When investments fail, they absorb losses as investors, taxpayers, workers or consumers. “The state will pay” is usually a longer way of describing a distributional choice.
A fair electric century must make costs visible without making investment impossible. It needs stable rules, but not protection from every entrepreneurial loss. It needs public planning, but no immunity for poor results. And it needs returns capable of attracting capital without allowing unavoidable bottlenecks to generate unlimited rents.
The political economy of electricity can ultimately be compressed into one question:
Who may collect when the system works — and who must pay when it fails?
The answer determines more than wealth. It determines whether society will consent to build the next system at all.