5G can’t live on spectrum alone, says Eric Schmidt

• Schmidt is unimpressed by Pai’s “record breaking” spectrum auction
• In fact he thinks it probably represents a “digital setback”
• To beat back China, the US needs to build physical 5G infrastructure with government assistance

Ajit Pai (b. 1973) was a fresh-faced young lawyer when the UK’s mobile industry committed auction madness around the turn of the millenium, splashing £20 billion for 3G spectrum and helping to precipitate a massive telecoms financial crash which took years to iron out. 

But surely young Ajit took note. Surely that’s a mistake unlikely to be repeated by any country in the 2020s... isn’t it?

In fact Ajit Pai’s last act as FCC chair was to lever $81 billion out of the coffers of the big US telcos for Auction 107, which issued 280MHz in valuable “C-Band” spectrum for 5G. He’s been playing this up as some sort of tremendous triumph ever since. According to Pai the auction was “a record-breaking success” with all 5,684 spectrum blocks that were up for bid secured and gross proceeds exceeded $80.9 billion, “shattering the prior FCC auction record of $44.9bn.” 

“These results represent a strong endorsement by the private sector of the service rules and transition plan put in place by the FCC to quickly make the C-band a critical part of 5G rollout in the United States,” Pai maintained. 

Big money paid for spectrum is self-defeating

But sometimes big dollar numbers indicate failure rather than success and you don’t have to dig down very far to realise that $81 billion transferred to the US treasury means $81 billion less the telcos will have to spend on 5G infrastructure, and $81 billion that the 5G-using public and industry must ultimately fund That’s the problem with auctions. 

The 5G auction problem has been noticed by many, including former chief executive and chairman of Google, Eric Schmidt, who now runs a consultancy - Schmidt Futures. Schmidt, writing in the Financial Times, says he considers that investing in 5G mobile networks should be an urgent priority for the US and its allies to fend off the main geostrategic rival, China, and he views Pai’s ‘win for 5G” to be an “actual digital setback that America and its allies can ill-afford.” 

The big mistake, he says, was to maximise the 5G spectrum revenue without imposing meaningful requirements on the tecos to build the necessary network infrastructure to use it. That will probably result in the disinvestment and downsizing of the networks. And all for $81bn, which is “merely a trifle on a government scale, equivalent to less than a month of US debt issuance.” 

As a result, “Americans will face higher prices and weaker digital services — yesterday’s internet tomorrow. That is what happened when European telecoms companies paid over the odds during the 3G auctions... and Europe is still recovering from its lost digital decade.”

On the other hand, he says, China will soon have a national network generating speeds of 1 gigabit a second, giving it a huge head start and resulting in the next generation of technology giants not being European or American but Chinese. “America’s digital economy will become an also-ran.”

Is Schmidt being alarmist? Of course he is. I particularly don’t recognise his ‘lost decade’ in Europe throwaway line. On many measures (fixed broadband for instance) the UK was actually ahead of the US through the first part of this century and probably still is. 

So should spectrum auctions be going, going, gone? 

The Schmidt solution is to use the Auction 107 money for a special data infrastructure fund to provide direct aid to US states that build physical 5G infrastructure. If the US auctions more spectrum, he says, it must insist on getting infrastructure to go with it, and a true 5G network will require much more spectrum: Japan, China, South Korea, the UK and Canada, he points out, will assign an average of 660MHz of mid-band spectrum each for 5G by 2023. 

Future auctions must set stringent build requirements, with penalties for underperformance. And ideally the US should pursue alternatives to auctions.

He suggests pursuing the defence department’s proposal to share government-controlled spectrum with commercial providers if they build infrastructure quickly. This sort of government intervention is usually viewed with disdain in the US, but the incoming Biden administration may well be more accepting of it.