The last several years must have been very frustrating for the technology policy counselors at groups like the New America Foundation. One of the main theses animating their continued recommendations for greater government regulation of the Internet is that the privately-owned IP networks that constitute the Internet will choose to starve it of transmission capacity in hopes of creating artificial scarcity and driving up its price. But as each year has passed, more customers have chosen to connect to the Internet; customers have increased greatly their monthly usage of Internet bandwidth; and broadband providers have expanded the capacity of their IP networks to meet these burgeoning demands.
While most people would look at this and say “wow, the Internet market is working just as it should,” you might be less enthusiastic if your public policy advocacy for increased government regulation was based on a prediction that the opposite should be occurring.
So perhaps it was this frustration that caused James Losey of the New America Foundation to complain about Cisco’s announcement of a new router (the CRS‑3) with triple the capacity of its current flagship model (the CRS‑1). Rather than celebrating this advancement in IP transmission technology, Losey was reported to have had a sour grapes reaction that this development was simply “Cisco…looking for ways a service provider can sell the same amount of bandwidth to more consumers, and often at a higher cost.”
Aside from concerns that Cisco’s innovation was inconvenient for pro-regulation advocates, Losey’s statement is also puzzling as a matter of economics. There is little point in introducing higher capacity infrastructure if there is no intention of using the extra capacity that it offers. Further, when a new technology arrives that expands productive capacity, it generally has the effect of reducing per-unit prices – even though total expenditures may rise because of induced increases in quantity consumed.
Of course it is also possible that what Losey was really objecting to was not the transmission capacity improvements available from the CRS‑3, but the fact that this router has increased software intelligence that enables it to provide many unprecedented cloud-computing or VPN services in addition to basic IP transmission capacity increases. But if so, Losey’s statement becomes even more audacious. On what basis should he conclude that Cisco’s innovations in infrastructure intelligence should be throttled in preference for today’s limited capacity “dumbness”?
In any event, Losey should not worry for the future of “open” best-effort Internet transmission. The way all successful carriers engineer IP networks for quality-needy advanced services like video-conferencing or mission-critical cloud computing is by expanding total transmission capacity by so much that these services never need worry that they will face significant network congestion. But because these advanced services will, on average, use only a small fraction of this increased capacity, router queuing algorithms ensure that this unused incremental bandwidth is fully available for use by best-effort services. Thus, the introduction of advanced services generally enhances the quality and capacity of best-effort traffic. It does not decrease it.
So what is the moral? Innovations in IP infrastructure, whether in transmission or intelligence, should be celebrated, not censured. And even if you have a personal preference only for transmission innovation, you should not despair at innovations in intelligence – they will redound to your preferences, too.