Tales of the Sausage Factory: Unlicensed Spectrum Access Part I

In 2003, “wifi” went from geek toy to mainstream. But WiFi is only part of a much larger revolution in how people access and use the electromagnetic spectrum. Now, numerous competing and ill-fiting anaologies, “property,” “public commons,” “public trust” battle it out among Washington regulators. What’s at stake? While it sounds hyperbolic, this regulatory battle ground holds the key to the next stage of evolution of information technology. This is a background piece. I will post the current developments piece later.

Wireless provides a fun example of the power of innovation and how regulation can serve the public, serve incumbent industries, and how obscure policy issues in Washington impact your life.

To fully appreciate this requires some heavy duty background. Even if you think you know it, you get some insights into me by how I summarize it, so scroll through anyway.

Under U.S. law any transmission of electromagnetic energy needs permission from the the Federal Communications Commission. (47 U.S.C. 301) That includes devices that radiate energy as an incidental byproduct of functioning, so called “unintentional radiators” like computers or microwave ovens. (47 USC 302)

So for a very long time, we had very little innovation in wireless technology, because any time you wanted to develop a new technology, you had to go to the FCC to ask permission even to experiment with it. Then you had to get a license to offer the service. The license usually defined the exact nature of the service. So if you had a license to offer taxi dispatch service, that’s all you could offer.

Worse, the way the FCC alloted licenses was to take an entire band of spectrum and allocate it for a specific use nationwide. Then they gave free licenses based on comparative hearings (referred to in the industry as called “beauty contests”), where applicants tried to prove that a grant of a license would best serve the “public interest, convenience and necessity.” (As required by 47 U.S.C. 309)

Now this system made good sense in 1927, when Congress passed the precursor to the current Communications Act. Transmitters were mostly omnidirectional and used almost exclusively for AM radio. (Yeah, I know hams go back further and there was significant maritime use, but these were niche concerns that got their own carve outs in 1927.) Receivers were totally passive.

But even in 1934, when the Federal Radio Commission became the Federal Communications Commission (when Congress added jurisdiction over telephones and all other modesof communications), technology was advancing. Point-to-point was becoming feasible and, after WWII, spread spectrum had the potential to fundamentally alter the way we regulated spectrum use. But the regulatory regime remained unchanged. New uses got shoehorned into the old system and the “spectrum chart” (of what was allowed where) became this incredibly complex mossaic of individualized uses.

Part of the problem was that, by that time, we had an industry of incumbents who very much liked things as they were, thanks very much. Even when the FCC began serious regulation to promote the public interest in the 1940s, it was still the devil everyone knew.

In 1959, an economist named Ronald Coase decided that private markets would work much better than central planning. So he proposed that the FCC auction licenses as property and impose no other conditions or restrictions. Licensees, he argued, would have the incentive to best develop wireless technologies and the public would be better served by a dynamic market able to respond to new technologies. Indeed, he argued that because new technologies took so long to get to market through the FCC, no one bothered to develop them.

Coase’s theories were treated with scorn by the FCC and public policy types initially, but he gained adherents. The 1980s saw the rise of the “market school” and a steady drumbeat began to change spectrum into a form of property.

I should point out that, despite what many licensees seem to think, spectrum is _not_ a form of property. For one thing, there is no such thing as “spectrum.” What we are talking about is a right to emit electromagnetic energy at certain power limits cycling through particular frequencies of the electromagnetic spectrum.

Similarly, there is no such concrete thing as “interference.” “Interference” describes a condition where a receiver cannot properly interpret a signal directed at it.

To economists, these points are irrelevant. _Anything_ can be tradable property, whether it is a real thing or simply some concrete excludable right (e.g., the right ot emit pollutants into the air). But these points are critical both to technologists (because the concepts of “spectrum” and “interference” can change with technology) and First Amendment theorists (since we do not restrict speech unless we must).

But to resume our story. In 1993, the auction folks got their big break. The new Clinton Administration was desperately casting around for new revenue streams. So it siezed upon auctions. The first PCS auction in 1994 raised billions of dollars (although very little was actually collected because most of the companies went bankrupt and the Supreme Court ruled that, under U.S. bankruptcy law, the licenses were assets that could be resold. So Nextel and a bunch of other companies paid the bankruptcy creditors pennies on the dollar for licenses, and the merest fraction of what the licenses would raise for the Treasury if reauctioned).

Now auctions did not make spectrum “property.” In fact, the Communications Act is quite explicit that (a) no one can have a property interest in spectrum (47 USC 301), and (b) any property right you might think you acquire you waive as a condition of getting a license (47 USC 304) and that paying for this at auction doesn’t make it property (47 USC 309(j)).

So you’d think it would be pretty settled that spectrum is not property, right. Hah hah. Everyone who has a license — especially folks like broadcasters who have them for free — want to be able to treat it like property. They also want to exclude anyone else from getting a license, since the value of a license as property depends on scarcity.

And a core of libertarian/dereg Repulicans love property. Leaving aside the natural Libertarian inclination to believe that private property is the solution to all economic ills, addresses all social issues, and secures you a place in the world to come, it removed the FCC out of the picture. Except, of course, as a cop to keep anyone else out. But just as the Social Darwinists of the late 19th and early 20th century believed a “free market” included government muscle to stop labor from organizing and engaging in collective bargaining, Libertarians convinced themselves that a regime of government regulated monopolies was a free market.

In defense of the propertizers, however, the defenders of the old regime did not have much intellectual oomph behind their arguments. For one thing, everyone agreed that the old system really sucked. The attempt to impose explicit public interest obligations usually resulted in the licensees making all kinds of excuses and never providing the service. Compartaive hearings, where licensees and would be licensees competed to show they would best serve the public, became avenues for rivals to exact greenmail or special favors. And, while defenders of public interest allocation were probably right that valuable social goals would go unmet and distribution of services would favor the wealthy, by the 1990s not too many people cared. Besides, the propertizers argued, the poor could hardly do worse under the current regime, which held back everyone. So while a “rising tide” approach in a free market wouldn’t be a big boost to the poor at first, everyone would benefit to some degree over time.

So when Michael Powell, famed Libertarian Republican became chair of the FCC in january 2001, it seemed like property was the way we would go. Sure the statute made this impossible, but the FCC would try it darndest and, with the expected cooperation of the D.C. Circuit, might pull it off.

But then a funny thing happened in a modest little corner of the FCC’s rules called Part 15.

In 1989, the FCC decided to allow certain low power devices to operate without an exclusive license. Since 1938, the FCC had allowed manufacture of devices that used very low power electromagnetic signals. But it had required manufacturers to come in with a new application for every new use. If you wanted to make a wireless remote for your TV, you had to come in with an application and technical specs and a place on the spectrum chart. If you wanted a cordless phone, you needed a new application.

So in 1989, the FCC changed the rules to make life simpler for everyone. It declared that it would establish that uses at low power in a wide variety of frequency ranges would be a priori permissible. Anyone could come in, get a device certified that it would only operate in the permitted bands at the permitted power, and then use it however they wanted. Just meet the cert requirements and you can do anything without anyone’s permission. But you have to agree to take any interference that comes your way without complaint, and you have to shut down if you interfere with someone protected.

The FCC has allowable “unlicensed” powers at a variety of frequencies. The most powerful is in the 2.4 GHz band, because that is shared with industrial scientific and medical equipment (ISM) and was deemed least likely to suffer interference.

Now add Moore’s Law, which, loosely, is that processor power doubles approximately every 18 months. Add also the devlopment of networking technology generally and the Internet specifically.

By the mid-1990s, interest emerged in using wireless technologiues for networking. In 1997, two things happened. First, the FCC opened up a new block of spectrum above 5 GHz with networking specifically in mind. Second, a law professor named Yochai Benkler wrote a paradigm-shifting piece called “Overcoming Agoraphobia.” In it, he proposed that wireless networks could operate without the need for exclusive licensing, and that this could serve as a tool for genuinely empowering individuals to speak and innovate via the electromagnetic spectrum. Benkler’s work in wireless became part of a broader movement known as the “information commons movement” (many may be familiar with Larry Lessig’s work in this in the area of copyright).

Unlike property, which relies on scarcity, the commons movement argues that value in many instances is better derived by ubiquity — particularly where there is no problem of allocation of physical resources to address. We do not grant exclusive use of words and images because words are scarce, for example, but because granting copyrights provides incentive to create. If you are an “open source” fan, you probably can quote by heart the argument for open source. Whether one measures value in terms of increased economic activity, increased personal autonomy, or increased individual creativity, many situations lend themseves to management as a common resource rather than a market for exclusive and scarce resources. (Caveat: the commons approach is often misrepresented as being anti-private property. This is a false characterization.)

The electromagnetic spectrum, argued Benkler and others, represents a situation where management as a common resource provides greater value than artificially created exclusive licenses traded as commodities. Benkler, joined by David Reed (one of the “fathers” of the Internet and one of the developers of the “end-to-end” principle) and others, argued that the technological limitations that required exclusive licensing had either been overcome or would be overcome in the forseeable future. As a result, exclusive licensing was counterproductive because it unnecessarily restricted free expression and freedom to innovate.

In 1999, the IEEE developed the 802.11 protocols for wireless networking in the 2.4 GHz and above 5 GHz bands. One protocol in particular, 802.11b, became widely deployed. Wireless ISPs (WISPs) even began using it and other protocols to provide broadband Internet access, particularly in places ignored by cable and DSL providers.

In 2001, “WiFi” became the one area in the tech sector experiencing growth. By 2002, the market for WiFi and other unlicensed wireless networking devices had grown into a multibillion dollar market. Meanwhile, developing a system of spectrum property turned out to be harder than its supporters thought. Worse, as companies became bankrupt, the exclusive licenses got mired in nasty fights as bankruptcy assets, making it harder to showcase their value as property.

So, in 2002, when the FCC organized its Spectrum Task Force to study how to manage spectrum, it did not end up neatly in property as everyone thought it would. Commons ahd a lot going for it. It was clearly producing boatloads of innovation and economic activity. Some Libertarians thought it was more free market than exclusive licenses, because it got the government out of management and left making the rules to private orgs like the IEEE. Many of the anti-propertizers liked it because it (a) wasn’t property, which many Liberals feel is the root of all evil, social decay and probably gets you a place in Hell; (b) empowered all kinds of nifty social networks; (c) wasn’t property; (d) reverberated with lots of good ideas about freedom of expression and the public good; (e) wasn’t property; and (f) had an actual chance in Hell of working in this administration.

And within the agency, a great division arose between those who supported property (which relies on exclusivity, and is therefore inimical to unlicensed) and those who supported commons (which relies on ubiquity, and therefore is inimical to exclusive licensing).

And Chairman Powell . . . . likes both.

Next time, 2004: the year of the paradigm wars.

Stay tuned . . .