Pending public rage, the FCC's flawed net neutrality rules will kick in later this summer. And yet, despite the impassioned coverage in the media, it's obvious that most people are still missing the point. As Wired points out today, a type of internet fast lane already exists, and nobody's talking about how to shut them down.
Fast lanes don't require slow lanes…
There are actually two different types of internet fast lanes. The one you're used to hearing about involves paid prioritization, or, the practice of companies paying internet service providers (ISPs) for preferential treatment. These fast lanes actually work like toll booths, where paying companies get to go through the gate first when traffic is congested. It's a bad idea, and one that the new FCC rules would allow. It also applies to the so-called last mile of connectivity, meaning the consumer doesn't have another option.
The other type of fast lane, though, already exists. It's been happening for years, and the new FCC rules don't even address it. (As Free Press's Derek Turner pointed out to us in an email, the agency agreed back in 2010 that it's not a concern.) As Wired explains in the first of a three-part series on net neutrality, plenty of large companies like Netflix and Google already enjoy the benefits of a type of fast lane that's created when they pay for the ability to connect directly to ISPs like Comcast and Verizon through special "peering connections" or "content delivery services."
Peering connections and content delivery services are two different ways of solving the same problem. When a company's traffic gets big enough, it's going to start clogging the regular pipes that carry traffic from the company's servers to the ISP. This is commonly known as the internet backbone.
Think of it like a jammed up interstate that connects two cities. An obvious way around this congestion is simply to build a new highway that connects the servers to the ISP directly. This is known as a peering connection. Alternatively, a company can ensure even faster connections by building servers inside the ISP, establishing a content delivery service. The FCC's current rules don't address this type of fast lane at all, and according to Chairman Wheeler, the agency isn't interested in addressing them any time soon, either.
…But that doesn't make these fast lanes harmless
This second kind of fast lane is how Netflix has improved streaming quality for customers on Comcast and Verizon's networks—albeit at a cost to consumers—in recently months. They're not paying the ISPs to prioritize traffic, per se. They're actually paying for new infrastructure that only they can use. Google was the first company to set up these kinds of agreements with ISPs, and now it's pretty common with all the major players, from Apple to Facebook. Everyone benefits, because customers enjoy faster connections to sites they visit frequently without slowing down connections for others. That's why peering agreements don't get the same negative hype; they don't come with the same consequences as the types of prioritization the FCC would allow.
Of course, that doesn't mean that the internet's existing fast lanes are necessarily a good thing. These arrangements have traditionally been made without money exchanging hands and for the benefit of the consumer, but it's easy to see it's all starting to go wrong. Now that Netflix is brokering high dollar deals for peering, you can imagine how competing video services (i.e. the ones that can't afford to write big checks to Comcast) are at a disadvantage. Even though the fast lane applies to a different part of the internet's infrastructure, it's still a fast lane that compromises the extent to which the internet is free and open.
That said, the damage is perhaps negligible. "Paid-priority, or router-level prioritization, is a zero-sum game, where all non-prioritized data is artificially slowed down in order to create priority for paying traffic. This is unjust and unreasonable discrimination," Free Press's Derek Turner told Gizmodo. "[Content delivery services] do not in any way harm or slow down the bits of any other content owner. Their faster delivery is achieved through geography and physics."
So what can we do about it?
That's a tough question. On one hand, the internet's been working pretty well so far without much oversight from the FCC or anybody else. Those Netflix deals in mind, things are starting to get a little bit shady, but even Tim Wu, the guy who coined the phrase "net neutrality," thinks it would be a mistake to over-regulate the internet in an attempt to keep things fair. This is not just a question of commercial fairness either. There are much bigger issues like free speech and the political process at play.
On the other hand, there's a solid argument to made about how trying to save the internet by banning fast lanes is like trying to fix a flat tire with duct tape. Even if the patch keeps the car on the road for a little while, it's still a temporary fix for a larger problem. The guys who run the internet's back bone all seem to agree that the solution is promoting competition between ISPs. After all, if a third of American households don't have a choice when it comes to their internet, of course the ISPs can do whatever they want behind the scenes. And when it's virtually impossible for small ISPs to face off against the big guys, we can't expect much progress.
Something needs to be done. While the consolidation of power among a small group of ISPs is a big problem, the smaller problems still need to be addressed. The FCC's new rules allow paid prioritization, and most net neutrality advocates would agree that this a bad idea. The second type of fast lanes—the peering connections and content delivery networks—aren't necessarily a good idea. Banning them directly would bear bigger implications on the architecture of the internet.
But hey, since everybody's suddenly very interested in how the internet works, why not start that conversation? Why not start coming up with ways to increase competetion between ISPs? Why not give that little video start up a fair chance? [Wired, Vox, Ars Technica]