It probably wasn't how Google's CEO-founder Eric Schmidt (of "Don't Be Evil" fame) envisioned things. Earlier this month protesters converged on the Google campus to protest the Google-Verizon joint proposal to keep the internet neutral. Called a "joint policy proposal for an open internet," it was innocuous-sounding enough, but to many it is being seen, above all, as a sellout of the wireless internet where Google itself is keen to play. There's some thoughtful consideration given in the proposal to treating all content equally -- but only where "wireline networks" are concerned. There is also a call for "network transparency," but since decades into internet build-out there is still little network transparency, this seems more like a wish than a policy suggestion.
The issue of net neutrality is awash with a jumble of technology, politics, and business. Way back in 1978 Rob Kling wrote in Telecommunications Policy:
"Proposals which focus on changing the kind or quality of data available to public policy makers assume that ‘rationality' is inherent in the data or techniques used to generate it. Yet the evidence seems to indicate that whatever ‘rationality' may be found in policy-making is as much a feature of the policy-making process as of the data that informs it."
This seems to be true of the latest Google-Verizon proposal. Its pronouncements are assumed to be self-evident. Little glimpse is offered of the mountain of information the two firms together could marshall to strengthen their arguments. Like many aspects of the intersection between technology and public policy, extended discourse about complex proposals is readily sideswiped by vested interests, political calculations, and guesses hazarded about future technologies.
Background: Flaws and calculations
NPR's Tom Cole sees several flaws in the proposal, and his views are typical. The FCC is to have no real enforcement power, instead relying upon a yet-to-be-identified advisory group operating through a "complaint-driven," case-by-case oversight. On the other hand, the new Wild West is wireless broadband, whose providers are free to create "additional, differentiated online services" within their monopolies.
Cable TV was once one of those "differentiated services." Seen from one perspective, it grew and made possible the wired Internet speeds many now enjoy. From a different perspective, it created a tangled bundle of services including "free" and "pay" TV, local and long distance voice, and broadband - with typically only one or two providers in a market. The palette of services is deep compared to years past, but those who want to limit costs may have a hard time untangling the service bundles. For example, in the case of Verizon, bundles offering wireless, broadband and TV services are least expensive when purchased together. Further, the bundled services are each delivered on a single fiber infrastructure -- no picking and choosing infrastructure, so make room for that now-mandatory battery backup.
The Electronic Frontier Foundation's Cindy Cohn applauds the proposal for avoiding direct FCC control of content, and considers the role of outside standards bodies potentially helpful, though not without risk. (Some such bodies, e.g., IEEE standards committees requiring 75% concurrence and debates lasting years, can be closed to public scrutiny; consider the case of Ultra-wideband below.) But she worries that the proposal sidesteps already exposed issues of censorship, content control, and adds her voice to the outcry over ceding the wireless Internet to purely mercantile interests -- the portion "currently most lacking in openness and neutrality."
Despite the lofty goals of net neutrality, as Harish Vadad points out, "not all packets are created equal and not all applications will get the priority". QoS requirements for Web content and email are different from voice and video. Traffic shaping with QoS mechanisms already come into play, as well as protocol differences (e.g., TCP vs. UDP). Service-Level failures can occur through intrinsic factors, not just the oft-mentioned BitTorrent "abuse," such as broadband HDTV users may experience around 10 p.m. on weeknights. My contract with Verizon gives me only an "up to" guarantee, and for downloads the service often exceeds that, but uploads are another story (see Figure A).
It doesn't take much imagination to see why Google would be concerned with QoS issues. Now that Google Voice has morphed into a long distance voice communications provider , one can imagine why peering and generally playing nicely with Big Telecom might make sense to Google. Verizon, a firm that in my geographical area, based solely on my own experience, is executing well with its fiber infrastructure, may want to head off a capabilities end run by the pure Internet players.
Political bandwidth: Limitless
Anyone assuming that net neutrality is a purely technical discussion should have their connectors cleaned. An analysis by the Sunlight Foundation recounts that just as there was Congressional opposition to plans as disruptive as a la carte cable service (imagine paying only for desired channels), there was opposition to net neutrality. No fewer than 74 Democratic and 171 Republican members of the House wrote the FCC in separate letters. Their gentle reminder? That Congressional direction is required before acting on net neutrality. Only last June, two of the five members of the FCC voted against a public hearing on overhauling the nation's broadband regulations and addressing net neutrality. The political rationale varies. Some claim that the FCC doesn't have the authority to reclassify broadband to Title II Telecommunication services, or that it would be challenged in court. Skeptics have a more cynical interpretation. In 2006, the last time Congress considered telecommunications legislation, the industry poured $59 billion into lobbying (source: Wall Street Journal via CNET).
On the other hand, a smaller voice consisting of four members from the House Energy and Commerce Subcommittee wrote the FCC chairman to critique the Google-Verizon proposal, identifying the segregation of wireless and overly broad description of "managed services." In their letter they wrote, "Rather than expansion upon a proposal by two large communications companies with a vested financial interest in the outcome, formal FCC action is needed. The public interest is served by a free and open Internet that continues to be an indispensable platform for innovation, investment, entrepreneurship and free speech." Senator Al Franken has referred to this as "the First Amendment issue of our time."
Unpredictable bedfellows: Engineers and lawyers
Lawyer Mitchell Lazarus wrote in IEEE Spectrum last year that net neutrality is just wishful thinking. In 2002, the FCC said that Internet providers were not required to open facilities to other ISPs. Monopolies with vested interests moved to protect themselves. For instance, wired network provider Madison River Communications blocked access by Vonage. Comcast blocked content that might compete with its pay-per-view service. As a result, and more importantly, smaller entrepreneurs whose inventions might interfere with entrenched paradigms must worry that they will not only be drawn into long court battles, but more likely, frozen out of courtrooms because they can't afford to litigate against wealthy adversaries. This makes raising capital even more difficult.
"Somebody is always regulating your channels. Will it be Comcast, Verizon, or do we have a rule through government that specifies minimal interference?" asked Google skeptic Siva Vaidhyanathan on a recent WNYC call-in show.
Some defensive actions ISPs might take could be illegal. Even more insidious are the actions they can take which are perfectly legal. They can demand expensive long term, or volume-based commitments to gain access to broadband service categories which are beyond the reach of startups and smaller firms. Perfectly legal. Expect that the cloud in cloud computing will need to be closer to the ground.
Lazarus bolsters his argument by reviewing the 12-year-old debate over Ultra-wideband. Here was an emerging technology supported by start-ups and radar companies, but opposed by just about every other existing corporate user of the spectrum. Not only did the fight drag on for more than a decade, despite what Lazarus argues was clear evidence that the new technology would not interfere with existing services, but the IEEE was unable to agree between two competing standards (search for "MB-OFDM" and "DS-UWB" for the sad history of competing standards) and disbanded its standards committee in 2006. The history of Ultra-wideband is an object lesson for anyone holding out hope for a straightforward role of professional associations in the Google-Verizon proposal.
Biggest benefits for the already big
The single biggest impact of the Google-Verizon proposal would be felt in one the hottest areas for investment in a weak economy. It would free the wireless Internet to engage in more price-based segmentation of services. Already in this space, and controlling not only the current revenue model but the existing infrastructure, big firms will stand to gain handsomely.
The Google-Verizon proposal tests our understanding of the distinction between large enterprises with government-sanctioned near-monopolies, and regulated public utilities accountable to a broader set of societal guidelines. Utilities can be privatized. Companies can act in the public interest and, if there is widespread adoption, economies of scale can result. But should the wealthy be allowed to buy passes for the HOV lane?
Seven questions for the new Internet rule-makers
1. Media and Telecom company size and the cost of litigation may exert undue pressure on fairness and policy formation. Is there a guarantee that profit-driven investment attracted to wireless enhanced services will benefit small and medium sized entrepreneurs as much as it will benefit the monopolies?
2. While file sharing of video is singled out as a resource hog, in fact the data to support this is not public. Can you show us the data about BitTorrent, streaming TV, and the like?
3. Cisco would like to upgrade both the wired and the wireless internet to such an extent that there's plenty of headroom for all to play. Wouldn't this offer a provider-neutral environment, or is the wireless genie out of the bottle?
4. Does this proposal institutionalize the disparity between rural have-nots and their wealthier urban/suburban counterparts? Will it be even worse for wireless?5. Figure A shows an account for 35mps down/20mps up with Verizon FIOS. The Verizon-recommended speedtest shows reduced upload speeds (though the FCC-sponsored tests of small file transfers say otherwise). Where is the transparency? Will the ordinary consumer with a modest at-home network be able to monitor an ISP's service level? Will they need to?
6. Trust in telecommunications providers is not strengthened by AT&T's collaboration with the NSA to wiretap and analyze domestic U.S. communications. Should we place additional trust in Big Telecom to segregate and price content for the wireless internet while handling more and more privacy data?
7. Is it still impossible to envision partnerships between small businesses and the current crop of broadband brokers?
- The original joint Google/Verizon post (http://googlepublicpolicy.blogspot.com/2010/08/facts-about-our-network-neutrality.html) by Alan Davidson (Google) and Tom Tauke (Verizon)
- The lay public is confused. A Reuters story points out that there is no consensus over the role of government in broadband adoption policy.
- H.R. 3458 http://markey.house.gov/images/PDFs/netneutralitybill.pdf).
Mark Underwood ("knowlengr") works for a small, agile R&D firm. He thinly spreads interests (network manageability, AI, BI, psychoacoustics, poetry, cognition, software quality, literary fiction, transparency) and activations (www.knowlengr.com) from a large island near NYC.