To celebrate the 30th anniversary of the first spectrum auction in 1994, TPI hosted a winter spectrum series, starting with a conversation with Nobel Laureate Paul Milgrom and FCC economist Evan Kwerel – two of the most instrumental people in making auctions happen. The introduction of spectrum auctions and tradeable property rights in place of administrative proceedings (so-called beauty contests) was perhaps the most important public policy achievement for the development of wireless telecommunications, which has been of enormous economic value. Most economists (and many others) agree that marketable property rights are the best way to allocate scarce resources to their highest valued uses, and spectrum is no exception.
Ironically, the subsequent TPI panel discussion on spectrum pipelines that followed only a few days later suggested that market-oriented spectrum policy may have seen its day. One of the participants, Penn Law scholar Gus Hurwitz, observed, “We are celebrating 30 years of spectrum auction authority with a bit of a funeral maybe.” Congress has thus far failed to renew the FCC’s auction authority and, in any event, there is no spectrum in the pipeline waiting to be auctioned.
If the market approach is in decline, the biggest culprit may be the ongoing lack of a good mechanism to allocate spectrum between government and private uses. Government agencies, the largest holders of spectrum, do not have to purchase the spectrum they use. They therefore do not face the opportunity costs of the spectrum they control and have a limited incentive to free up spectrum for private uses. There is, however, no intrinsic reason to shelter government agencies from the opportunity cost of spectrum, since these agencies, including those responsible for national security, purchase virtually every other input they use.
Incentivizing government agencies to use their spectrum more efficiently used to be a topic for discussion in spectrum policy circles, with proposals for user fees and similar mechanisms. NYU Sloan School economist Larry White and I proposed one such mechanism – a Government Spectrum Ownership Corporation (GSOC),
modeled on the GSA for real property. But these discussions are few and far between these days.
While government spectrum remains off-limits (with some notable exceptions), bands accommodating commercial uses are increasingly stepping on each other’s toes. Different governance models make their coexistence difficult, even if their uses are complementary.
The popularity of WiFi, for example, has led to continuing tensions between licensed and unlicensed models. The use of unlicensed spectrum has predictably mushroomed, because it is “free,” leading to demands for more unlicensed allocations.
Finally, the licensed regime hasn’t found a good way to accommodate new uses that may cause interference problems with adjacent licensed bands. This has proven to be a barrier to entry of new uses with interminable competing engineering studies – see, for example, recent well-publicized disputes involving interference with GPS and aircraft altimeters.
The buzzword of the day (as reflected in the panel discussion) seems to be “coexistence” among competing federal and non-federal uses. But figuring out the specifics of that “coexistence” is what markets generally do best – especially when property rights are reasonably well defined (as with flexibly licensed spectrum). Markets sort out competing uses and coordinate the activities of complementary uses. Without some sort of market mechanism to allocate spectrum among licensed, unlicensed, and other models we are left with competing technical reports.
Consider Citizens Broadband Radio Service (CBRS). A study by Recon Analytics undertaken for CTIA concluded that the CBRS is underused, with a high opportunity cost relative to an exclusive licensed model. An NTIA report, contrarily, argued that CBRS use was growing quickly.
During the spectrum pipeline panel discussion, the moderator, TPI Senior Fellow Sarah Oh Lam asked, “Does that mean we’re dialing back to beauty contests? Is it command and control again, industrial policy? Is that where we are? Or are there market mechanisms for figuring out what uses are valuable in a time when China and other countries are doing command and control? How do we know our government is picking the right use cases?”
That is the question, and of course we don’t know, but that seems to be the road we’re headed down. There was support on the panel for two new processes – the Partnering to Advance Trusted and Holistic Spectrum Solutions (PATHSS) Task Group and Emerging Mid-Band Radar Spectrum Study (EMBRSS). These are both committees of various interests – industry as well as the Defense Department – discussing ways to divvy up and share the spectrum. There is no evidence that resource allocation by committee backed up at some point by regulatory authority will yield efficient spectrum use.
Numerous studies show that the costs associated with inefficient utilization of the spectrum under the “command-and-control” system (see, for example, a seminal 1997 paper by MIT economist Jerry Hausman) were enormous. The beauty contest system was particularly ill-suited to handle the explosive demand for spectrum for new wireless products and technologies. So, it is concerning, to say the least, that we may be reverting to that legacy system.
