Ronald Coase and the radio spectrum

[Commentary] Back in the 1950s economist Ronald Coase first proposed spectrum auctions. Wireless licenses were auctioned in 1989 in New Zealand, 1991 in India, and 1994 in the US. Competitive bidding is now a standard policy tool in more than 30 countries. Over $52bn has been raised in America, more than twice that elsewhere. But that is barely the tip of the Coasian iceberg. The far more important reforms extend from a sharp liberalization of the rights granted wireless licensees. Traditional broadcasting licenses specify exactly what firms may do, fixing services, applications, technologies, and business models. Mobile phone and other modern licenses, however, convey airwave rights tantamount to spectrum ownership. They define band contours and then delegate choices about usage to market players. The result has been spectacularly successful. Some are asking policymakers to abandon Coase's spectrum property vision and move to shared spectrum space. The question is not whether "multiple parties occupy the same spectral space," but how we organize the sharing arrangements. Government does set aside unlicensed bands, but they have proven ineffective for the most valued wireless applications. In local uses where they attach to phone or cable networks built using privately owned "spectrum in a tube," wi-fi radios and cordless phones work. But the complexity of these plug-ins pales in comparison to the wide-area networks customers deem most productive. To provide those services, mobile carriers stack millions of "multiple parties" into the same spaces - 4.6bn subscribers at last global count. The most intensively shared wireless bandwidth is found exactly here, in spaces allocated to what regulators call "exclusive use" spectrum. That, too, is a most interesting error. But one, unlike Coase's, we should aspire to correct.