Some Power Plants Are Paid to Exist Rather Than to Generate
Capacity markets pay generators for being available on the worst day of the year. They exist because an electricity market that pays only for energy will not fund enough plant to keep the lights on.
The Physical Constraint
Electricity supply must match demand continuously. There is limited storage in most systems, so the grid needs enough generating capacity to meet the highest demand that will ever occur, not the average.
Peak demand happens rarely, perhaps a few dozen hours a year during extreme weather. The plants that meet it sit idle almost all the time.
The system needs plants that almost never run. Nothing about an energy only market pays for that, which is the entire problem.
The Missing Money
In a market paying only for energy delivered, a plant running fifty hours a year must recover its entire fixed cost in those fifty hours. That implies extraordinarily high prices during scarcity.
Regulators cap prices, for reasons that are politically inevitable and economically consequential. Once capped, the scarcity revenue is insufficient to fund the capacity, and the plant is not built. This shortfall is known as the missing money problem.
What a Capacity Market Does
A capacity market pays generators for committing to be available during a defined future period, separately from any payment for energy produced.
| Payment | For |
|---|---|
| Energy market | Electricity actually delivered |
| Capacity market | Being available if called |
| Ancillary services | Stability, reserves, frequency response |
The system operator estimates how much capacity is needed to meet a reliability standard, then procures it through an auction. Generators bid the payment they require to remain available, and the clearing price is paid to everyone who clears.
The Design Difficulties
These markets are administratively determined rather than emergent, which means the design choices are the market.
The quantity is set by a regulator judgment about acceptable reliability, not by demand. The definition of what qualifies as capacity determines which technologies can participate. And how intermittent resources are credited is genuinely contested, since a wind farm contributes to reliability but not dependably at the moment of peak.
Storage complicates it further. A battery can deliver at peak and only for a limited duration, so crediting it requires a judgment about how long the peak lasts.
The Argument Against Them
Critics argue the problem is the price cap rather than the market design, and that allowing prices to rise freely during scarcity would fund capacity without a separate mechanism.
That is theoretically coherent. It requires tolerating electricity prices that spike enormously during emergencies, which is politically unsustainable and falls hardest on those least able to absorb it. Jurisdictions that relied on scarcity pricing have experienced exactly that and faced severe political consequences.
The other criticism is that capacity payments can subsidise old plants to stay open when the system would be better served by investing in flexibility or demand response.
Demand Side Participation
A useful development is allowing consumers to bid reductions into the same market. A factory willing to shut down during a peak provides the same reliability service as a plant that starts up, frequently more cheaply.
Treating demand reduction as equivalent to generation is one of the clearer efficiency improvements available, and its uptake is limited mainly by measurement and by how much industrial load can genuinely be interrupted.
The Bottom Line
Electricity systems need plants that run only during rare peaks, and an energy only market with capped prices will not pay for them. Capacity markets pay for availability directly, which solves the funding problem and replaces it with a set of administrative judgments about how much capacity is needed and what counts as providing it.