Benkler on Markets and Telecom Policy

Here’s Yochai Benkler’s take on the evolution of the telecom sector, from page 152 of The Weatlh of Networks:

Throughout most of the 1990s and currently, communications and information policy around the globe was guided by a wish to “let the private sector lead,” interpreted in large measure to mean that the various property and property-like regulatory frameworks should be strengthened, while various regulatory constraints on property-like rights should be eased. The drive toward proprietary, market-based provisioning of communications and information came from disillusionment with regulatory systems and state-owned communications networks… In the United States, this model translated into efforts to shift telecommuniations from the regulated monopoly model it followed throughout most of the twentieth century to a competitive market, and to shift Internet development from being primarily a government-funded exercise, as it had been from the late 1960s to the mid 1990s, to being purely private property, market based. This model was declared in the Clinton administration’s 1993 National Information Infrastructure: Agenda for Action, which pushed for privatization of Internet deployment and dvelopment. It was the basis of that administrations’s 1995 White Paper on Intellectual Property, which mapped the most aggresssive agenda ever put forward by any American administration in favor of perfect enclosure of the public domain; and it was in those years when the Federal Communications Commission (FCC) first implemented spectrum auctions aimed at more thorough privatization of wireless communications in the United States…

The result of the push toward private provisioning and deregulation has led to the emergence of a near-monopolistic market structure for wired physical broadband services. By the end of 2003, more than 96 percent of homes and small offices in the United States that had any kind of “high-speed” Internet services received their service from either their incumbent cable operator or their incumbent local telephone company. If one focuses on the subset of these homes and offices that get service that provides more substantial room for autonomous communicative action–that is, those that have upstream service at high-speed, enabling them to publish and participate in online production efforts and not simply to receive information at high speeds–the picture is even more dismal. Less than 2 percent of homes and small offices receive their broadband connectivity from someone other than their cable carrier or incumbent telephone carrier. More than 83 percent of these users get their access from their cable operator…

The alternative of building some portions of our telecommunications and information production and exchange systems as commons was not understood in the mid-1990s, when the policy that resulted in this market structure for communications was developed. As we saw in chapter 3, however, wireless communications technology has progressed to the point where it is now possible for users to own equipment that cooperates in mesh networks to form a “last-mile” infrastructure that no one other than the users own.

This strikes me as a strangely skewed retelling of the last decade of telecom policy. In 1993, the Internet was available to a tiny fraction of the population mostly in Universities and corporations, with the fastest residential Internet access running at 14.4 kbps. Local phone and cable service was a monopoly almost everywhere, and long distance phone service was only weakly competitive.

Today, the majority of American households have Internet access, and it’s a hundred times faster than it was 15 years ago. Most households have two options for wired phone service, half a dozen choices for wireless phone service (made possible by those spectrum auctions in the 1990s), and almost unlimited options for Internet-based voice communications. The long distance market has become so competitive that it essentially price-cut its way into economic irrelevance.

The one telecom sector that hasn’t been significantly deregulated is cable television, and it still largely consists of local monopolies (although satellite TV does provide an important alternative). The people complaining about the lack of broadband have been largely silent about the franchise reform issue.

Perhaps most importantly, every part of the Internet other than broadband is fiercely competitive. Benkler complains about slow upstream broadband connections, neglecting to mention that a wide variety of services–Flickr, YouTube, Blogger–that will host your content for free. Or that the price of full-featured web hosting has been falling rapidly year after year. Control of the Internet backbone is widely dispersed among private companies, with none having near enough market share to raise concerns of monopolization. The market for Internet connectivity to medium and large organizations has plenty of competition, and there are no shortage of co-location facilities, which will put your server on their racks at prices that’s within the reach of your average computer geek.

So it seems to me that the 1990s experiment in more market-oriented telecom policies has been a great success overall, with residential broadband (a market that didn’t even exist in 1996) a possible exception. But it should be stressed that even there, the concern is only hypothetical. No significant network neutrality problems have arisen to date, and the few that have arisen have been dealt with quickly under existing laws.

Perhaps the strangest part of Benkler’s argument is his conclusion about mesh networks. If mesh networks are technically feasible, a system of property rights in spectrum would allow a company like Intel to purchase a block of spectrum and license for a reasonable fee to any manufacturer that wanted to make mesh networking hardware according to Intel’s specification. Indeed, if spectrum were truly non-rivalrous, Intel could license a variety of competing mesh network standards, all of which could share the same physical spectrum. Hence, if mesh networks are technologically feasible, the blame for their non-adoption rests with the propertization of spectrum being too slow, not too fast.

Moreover, we have ample precedent for spectrum auctions bringing greater telecom competition: just look at the cell phone industry. Those cell phone companies are using spectrum they bought from the federal government during the spectrum auctions of the 1990s. Partly as a result, there’s no serious concern about the lack of competition in wired phone service–true, the Baby Bells are still regional monopolies, but anyone who doesn’t like their service has plenty of options from wireless phone companies. It seems to me that the best way to promote competition for the broadband incumbents would be to auction off more spectrum so that companies could deploy next-generation wireless data networks. But Benkler hardly mentions this possibility.

SHARE: 10 comments