Economist Chris Doyle has published a paper arguing that spectrum licence fees built on static opportunity-cost snapshots — the approach UK regulator Ofcom pioneered in its 2004 framework — systematically misprice the resource because they ignore demand swings, coordination costs and market concentration.

Doyle, an economist specialising in competition policy regulation with an emphasis on spectrum management, sets out the case in a paper published in Telecommunications Policy (via ScienceDirect). He proposes incorporating dynamic elements such as stochastic demand and coordination externalities into spectrum valuation — variables that most pricing frameworks currently leave out.

"Spectrum scarcity isn't disappearing," he told PolicyTracker. "But the way we manage it is changing. Increasingly, frequencies regulators would like to make available are already being used. Clearing incumbents can be expensive, slow or impractical. At the same time, improvements in radio technology and automated coordination make coexistence increasingly feasible."

That shift toward coexistence raises, in Doyle's words, an "interesting" economic problem: "What is the opportunity cost of a spectrum right when that right is conditional?"

What is wrong with AIP today?

Doyle takes a critical look at Administered Incentive Pricing (AIP), the key principle behind Ofcom's 2004 spectrum pricing framework. AIP was designed to set licence fees based on the economic opportunity cost of spectrum rather than merely recovering administrative costs. Doyle has written extensively on its theoretical foundations alongside experts such as Martin Cave.

"Administered incentive pricing, which Ofcom pioneered but others also applied, is used for pragmatic reasons," he said. "It takes snapshots of given moments in time and uses information from several sources — whether it's auctions or simulations — to determine what should inform the opportunity-cost measure."

He stressed he did not want to criticise the frameworks regulators have adopted. His goal, he said, was to explore what a dynamic framework for calculating the value of the resource would look like, and which variables it would involve. "The paper takes basic conceptual analysis and reminds us that we're not in a static world," he said.

Which three factors should shape dynamic pricing?

The paper identifies three specific influences that dynamic spectrum pricing would need to capture:

• Spectrum trading. Secondary-market trading is intended to migrate spectrum to higher-valued uses. In practice, trading does not happen in all markets — particularly where "spectrum ownership is concentrated among a small number of operators," the paper says. Where a few big players dominate, they hold back usage and prices fail to reflect true scarcity. Prices only move toward the "right" level when enough buyers and sellers compete freely.
• Band refarming and migration. These depend heavily on device compatibility, equipment availability and harmonisation processes. Their efficiency is tied to broader coordinated ecosystem migration, which Doyle argues should be factored into the price of spectrum.
• Congestion. The busier a network gets relative to available capacity, the more spectrum should cost at that moment. Price should track congestion in real time. A flat, unchanging fee becomes increasingly wrong — and costly — the more demand swings up and down.

Could regulators actually implement it?

Doyle is candid about the limits of his proposal. "In a nutshell, the paper is conceptual," he said. "In a fictional world, yes, you would do it this way. We're not in a fictional world and, therefore, dynamic spectrum pricing would potentially be resource-intensive and create many difficulties."

He frames the trade-off bluntly. "A static calculation will never get the best outcome in theory," he said. "But in practice, and this is the crux of the matter, it may be the best way to do it, because moving to a more sophisticated dynamic pricing schedule — although it looks correct from a theoretical perspective — may be so demanding in terms of information and so concerning to those paying the fees that it would almost certainly lead to more arguments about what the relevant factors are."

The paper also flags further elements that could feed a dynamic pricing regime: investment incentives, spatial dimensions and downstream-competition effects. Doyle calls these a "natural agenda" for future work but did not tackle them himself, because proper evaluation would require expertise beyond economics.

"My strength is economics rather than radio engineering," he said. "When you start to drill down into the type of usage and factors that may be very relevant to understanding the technology and how it may be evolving, I think that could be very helpful. Building potential simulations to compute prices would be quite an interesting exercise. It will most likely happen outside of a university context."

His analysis stops short of recommending that regulators deploy a layered dynamic framework. But he sees possible targeted applications. "There may be, in some frequency bands, a case where dynamic adjustments should be considered," he said. "Does it matter today? Maybe. Might it matter more in the future? And might AI and other tools that enable us to have more sophisticated pricing systems help? Maybe."

For spectrum managers at operators and regulators alike, that leaves an open timeline: no near-term change to fee-setting methodology, but a research agenda — and a possible role for AI-driven pricing tools — that could reshape how conditional, shared spectrum rights are valued as coexistence becomes the norm.