Tuesday
Mar062018

The Best Way to Issue 5G Spectrum—Lessons from China?

Did mega spectrum auctions delay the deployment of 4G?

When spectrum auctions were first introduced they triggered a debate.  Would the high prices slow down network deployment by eating into available capex?  No, the auction proponents contended, they would speed up the deployment of new networks and capacity.  Reason One—the most ready operators and investors would win the auctions.  Reason Two—the incentive to recoup the spectrum prices fast would be higher than if the spectrum was allocated through beauty contests

We checked this logic in the early days of 3G spectrum auctions in Europe and, sure enough, the networks that used auction-based spectrum generally rolled out faster.  Yet now a different story is starting to appear.  According to one study of 47 countries, “auctions impose selection and debt effects on the after-market competition that could harm consumers.”  This in turn is reflected in the “3G mobile phone penetration rates among auctioning countries [being] 1.04–8.95% lower” than in their beauty contest peers, according to a recent academic study (Kuroda and Baquero, Telecommunications Policy, 2017).

Of course, the country peers examined in the study may not be entirely peers.  Countries that have used beauty contests have generally had fewer operators than those using auctions.  The presence of fewer operators—say three instead of four or five—generally results in more subscribers per capita even unadjusted for the greater number of SIMs per user in a market with more operators.  This, we think, is due to less market fragmentation and less churn, leaving more resources for network build-outs and pursuing new subscribers, including in rural areas. 

Still there is another argument being made for considering alternatives to auctions.  In a word it is China.  China uses neither auctions nor beauty contests but essentially administrative fiat in allocating licenses and spectrum.  The result is that China has deployed 4G networks faster by far than any other large country.  Overall, it has achieved 1 billion 4G subscriber connections to date, what the rest of the world has collectively achieved.  Bottom line—a government that doesn’t spend several years preparing the next major spectrum auction, as has the U.S., for example, with its Incentive Auction, can fast-track 4G deployment.

So what to do with 5G spectrum?  The quick lesson is that the holding of multi-band mega-auctions, the main trend in recent years, may in fact have slowed down technology and service deployment.  While using administrative diktat is not an option, easier ways to allocate spectrum should be found.  Possibly a regime closer to spectrum trading, allowing smaller, more routine assignments and exchanges of spectrum is the way to go—or even larger 5G auctions, though not a single mega auction, that do not re-invent auction theory each and every time. 

PrintView Printer Friendly Version

« From A (ARPANET) to Z (Zettabytes) | How Will AT&T/Time Warner Play in Asia and Africa? »