Not all Internet Traffic is Created Equal
This really gets to the core of the debate. The Oatmeal and other net neutrality hardliners fear that ISPs may use their unique position to discriminate among different sources of Internet traffic, favoring well-heeled and/or affiliated edge providers and ultimately resulting in significant (and externalized) harm to small businesses and free speech. The solution, as they see it, is to mandate bright-line rules that will prohibit ISPs from engaging in such discrimination. But, while simple, bright-line rules have a certain appeal, in this case — considering the inherent complexity and diversity of the issues at hand — such a broad-sweeping rule against discrimination would do great harm, because it fails to account for a key facet of the Internet: some bit traffic is qualitatively different than others, because some is bursty (like web-browsing) and some requires sustained bit streams (like VoIP calls). Thus, while “neutrality” of treatment would perhaps be warranted when comparing like services (i.e., apples to apples), when comparing unlike services (i.e., apples to oranges), mandating neutral or equal treatment could do significant harm to both. In recognition of the fundamental differences between different forms of bit traffic, Internet engineers developed, among other tools, the Differentiated Services (DiffServ) set of protocols, which were designed to enable differential treatment of bit traffic in real-time. And by allowing ISPs and transit providers the capability to treat some bits differently than others, the Internet itself can work smarter, faster, and better for all parties in the online ecosystem.
Also contrary to popular belief, these are not recent developments. The 8-bit DiffServ field has been built into all packet-headers since IPv4 was implemented (back in the 1980s), and it was retained in the updated IPv6 standards, so any claim that “equality” is embedded into the Internet’s foundation is just patently false. And these are not just hypothetical applications — ISPs already utilize the DiffServ protocols to prioritize voice-over-IP (VoIP) packets, such as Skype, on their networks. They do this because they know that two-way voice services are very latency-sensitive, so ensuring a seamless delivery of VoIP packets makes those services work better and their customers more happy.
Most everyone can agree that certain forms of Internet traffic ought to be prioritized over others (e.g., 9-1-1 calls undeniably should be prioritized over non-emergency traffic). But, more often than not, outside of the realm of emergency services, people struggle to think of prioritization uses that would be beneficial to consumers, competition, and society writ large. The nefarious examples of anticompetitive prioritization (e.g., An ISP prioritizing its affiliated traffic over its unaffiliated traffic, resulting in skewing of the competitive Internet landscape) are typically all that come to mind. However, putting potential harms aside for the moment, it is easy to envision a number of beneficial uses of prioritization via the DiffServ protocols:
Microsoft, Sony, Nintendo, Steam, and Blizzard, to name a few online gaming companies, would willingly offer to pay extra — on top of the hosting and transit fees they already pay — to get their users’ bits prioritized, because they know even minor latency (or “lag” as it’s typically referred to by gamers) can ruin the online gaming experiences that they’re trying to cultivate.
ESPN, Fox, NBC and others that own broadcast rights to live sporting events, and offer online streaming capabilities (e.g., WatchESPN, FoxSportsGO, NBCSportsLiveExtra), would willingly offer to pay extra to get their bits prioritized, because doing so would make their consumers happier — through higher-quality bit streams — and would allow them to finally outperform the illicit websites that host sports streams in violation of copyright law, convincing more consumers to subscribe to their legitimate online streaming services (which likely will soon be offered a la carte).
And on the flip-side, although many net neutrality supporters rail against the idea of getting “stuck in the slow lane,” there are innumerable web-based services that would likely jump at the chance to get their traffic de-prioritized, because they know their services are time-insensitive (a.k.a. latency-robust), so slower delivery of bits wouldn’t matter much to their users, and — this is key — the cost of delivering those bits would be lower than if they were in a faster lane of Internet traffic. For email hosts, software developers, or any startup company (not trying to live-stream video) looking to cut their transit costs, such an option would surely have at least some appeal (depending, of course, on the particular price points).
These are just a few examples of how Internet traffic prioritization could be hugely beneficial for consumers, but there are surely plenty more. To be fair, there are also numerous examples of how this sort of prioritization could be harmful — by allowing powerful companies to crowd out smaller firms, or by allowing ISPs to surreptitiously favor their affiliated services — which is why some reasonable rules ought to be put into place to guard against such harms. Whether those rules ought to be based on the legal authority of Section 706, Title II, or a new piece of legislation from Congress, is a debate for another time. The point here, simply, is that any hardline “net neutrality” rule that bans outright any and all forms of commercial “prioritization” would stifle innumerable pro-consumer and pro-competitive business models in their cradle, perhaps forever stulting the growth of the World Wide Web.
