A recent paper discusses the use of spectrum set-asides to achieve policy objectives. Its conclusion merits careful review.
In Telecommunications Policy, Bronwyn Howell wrote “Using spectrum set-asides to address distributional objectives: Lessons from Canada, New Zealand, South Africa and the United States”, concluding that Canada’s use of set-asides has raised costs and has failed to foster competition effectively.
The last time I looked at set-asides (in December 2023), I wrote that “ISED got it right” by using spectrum caps (instead of set-asides) as a means to ensure smaller service providers had equitable access to the 3800 MHz band. The average cost per Mhz-pop was $0.29, as contrasted with the 3500 MHz auction held two years earlier with an average cost of $3.28 per MHz-pop. At the time, I noted that the lower spectrum cost meant carriers would be in a better financial position to invest in physical infrastructure.
The Telecom Policy article notes that set-asides “are not economically neutral.” Set-asides distort spectrum auctions by reducing the total spectrum available, creating an artificial scarcity. This has been shown to result in higher auction prices.
In Canada, set-asides have generally been used to help “new-entrant” wireless service providers with market entry. The article also looks at setting aside spectrum rights for “economically disadvantaged communities” in specific geographies to redress economic, social and digital divides.
The rationale is that the spectrum can be used to attract service provision, confer profits from network ownership and upskill individuals in network engineering skills within disadvantaged communities. However, telecommunications network operation requires substantial additional capital as well as spectrum holdings. It is not clear where already financially-disadvantaged communities will raise this additional capital to deploy networks with any spectrum they may own.
Looking at data from examples in the US and New Zealand, the author found “it is not axiomatic that spectrum ownership is necessary for these communities to participate in the digital economy or that spectrum ownership alone will close digital or skills divides.”
Looking at Canada, we read that, despite what might be the best of intentions, Canada’s spectrum set-asides “failed dismally at inducing new competition”.
A 2023 paper by Kent Fellows and Jeffrey Church at University of Calgary looked at Canada’s AWS-1 auction concluding “there was a substantial opportunity cost associated with the set aside of spectrum.” They estimated that the increase in consumer surplus from the auction would have been 6 to 8 times greater, rising from $932M to a range of $5.9B to $8.2B. Further, they believe mobile penetration would have up to a million subscribers higher. Finally, the economists found that “the set-aside policy had a modest impact in reducing quality-adjusted prices, from one to two percent, whereas if the spectrum was allocated to the three incumbents, quality-adjusted prices would have fallen by four to six percent.”
The Telecom Policy article concludes appropriately:
While set-asides can address certain short-term equity concerns or sovereignty goals, realising sustainable gains in connectivity and economic development requires complementary resources — most critically, network capital, skills, and appropriate governance structures. By illuminating these trade-offs and complexities, we aim to inform more nuanced policy-making that balances equity and efficiency in future spectrum policy.
I couldn’t say it better.