Net Neutrality

What is Net Neutrality?

The short version: Network Neutrality (or Net Neutrality) is the idea that Internet Service Providers (ISPs) should not treat your data differently based on its destination. They should instead treat all data “neutrally”, all endpoints the same.

That’s really it.

The push to ensure Net Neutrality has taken different forms over the Internet’s lifetime (and attacks on Net Neutrality have, too). The current push involves the reclassification of ISPs to be “common carriers” under Title II of the Communications Act of 1934 (and, where relevant, the Telecommunications Act of 1996).

Most technologists were hoping the FCC would classify ISPs as Title II “common carriers” when the 1996 law was passed, but that didn’t happen. Instead, the FCC put much weaker protections in place. While there were early signs of trouble, they held up reasonably well for a decade, when they started to really crack. The FCC tried to prevent further damage with some rule changes in 2010, which held up well until the courts struck them down in January of 2014. The primary arguments made was that the FCC was trying to regulate ISPs using authority it only had over Title II entities, without actually classifying them as Title II. Now the FCC is finally trying to correct that fundamental mistake.

Why is this important?

One of the key points that has allowed the Internet to grow from a tiny research project into one of the most transformative forces in modern civilization is that everyone had an equal shot. If you can put a web page up, your ideas have as much voice as everyone else’s, without consideration of your sex, race, political views, or how well connected you are. You didn’t need the blessing of the service providers to start your business, share your ideas, or publish your collection of cat videos—all you needed was an Internet connection and an idea people liked.

(There are, of course, many important caveats there: hosting a web site costs money, access is not uniformly distributed, especially geographically, and so on. But the general idea holds.)

It might be impossible to overstate the impact this form of Internet has had on the world. Everything from email to FaceTime, Twitter to Instagram, eBay to Etsy, developed under this model. It has built new economies, helped overthrow governments, and given billions of people access to resources unimaginable just a few decades ago.

Unfortunately, this model has started to fall apart. There were isolated cases early on, but starting around 2007 we saw a large increase in ISPs deciding which of their customer’s traffic was worth more. These “violations” of Net Neutrality have only intensified over time, in the market, the courts, and in the technical fabric of how the Internet is built.

Some common objections

Net Neutrality is a really good idea, supported by a large and diverse group of people. Outside of telecom companies and the politicians backing them, support for it is pretty overwhelming. But for folks not already familiar with the issues, parts of the topic can be confusing or misleading.

We shouldn’t create a bunch of new regulations for the ‘net.

Net Neutrality doesn’t create any new class of regulation. The current proposal for how to best ensure Net Neutrality is based on moving ISPs from one class of entity to another in a law from 1934 (revised in 1996). In the ‘90s, the FCC had to decide whether ISPs were “common carriers” under the Communications Act of 1934, and decided they weren’t. That was a mistake, and correcting that classification is what the current push is about.

Don’t fix what isn’t broken.

Unfortunately, it’s already broken.

The Internet became hugely successful exactly because it “grew up” under conditions very similar to what Net Neutrality advocates are calling for. This has led some critics to argue that the reclassification of ISPs is therefore unnecessary. But we’re no longer operating under those conditions, and haven’t been for some time.

Starting in the later half of the last decade, ISPs started getting more and more invasive about traffic filtering. For several years, we’ve seen ISPs blocking peer-to-peer traffic based on the type of endpoint (favoring commercial users over individuals). The abuses have just been getting worse. 2014 was the worst year yet, with a high-profile court case involving Verizon and a public fight between Comcast and Netflix. Both went in ways that are bad for the consumer. The Verizon case, in January of 2014, essentially got all existing protections of Net Neutrality (weak though they were) thrown out; the Comcast vs. Netflix fight resulted in Netflix paying what amounts to extortion money to Comcast to get Comcast not to intentionally slow down Netflix traffic.

The market will sort it out.

Unfortunately, the market for Internet service is pretty messed up. Most people in America live in places where broadband service is either a monopoly or duopoly—it’s not a free market. There’s no good competitive pressure on ISPs to respond to market demands.

What’s more, one of the biggest dangers of a non-neutral Internet is that new services wouldn’t even be seen. So when an upstart video service comes along to challenge Netflix, or the next “Uber for waffles” comes along, it would have a much harder time even getting off the ground. If people can’t get these new services in the first place, they don’t know what they’re missing, and thus can’t put market pressure on ISPs to provide it better.

But some traffic should be treated differently!

Sure! There are really good reasons why, for example, I want my Voice-over-IP traffic (e.g. Skype), streaming video (e.g. Netflix), or real-time multiplayer game data to be given priority over less time-sensitive data like general web page loads. The technical term for that is “Quality of Service”, or QoS (“Quality” there doesn’t mean quite the same thing as in normal conversation; QoS really just means priority levels). But Net Neutrality doesn’t prevent ISPs from implementing QoS to prioritize streaming video over general web pages loads. ISPs remain free to filter based on the types of service, but they have to treat all users the same. So an ISP could prioritize streaming video over generic web traffic, but couldn’t use QoS to speed up Hulu while slowing down Netflix.

The technology moves too fast for the government to keep up.

Enforcing Net Neutrality wouldn’t require the government to “keep up” with anything. Neither the FCC nor any other government body (or anyone else, really) would have to set QoS levels for new service or decide which services were more time-sensitive than others. ISPs, who everyone agrees are in the best position to know about the technical needs of their own networks, remain free to implement QoS or not, or do it for certain services and not others. If one ISP decides not to bother, and treat all traffic the same, while another thinks their customers want to prioritize streaming video, but not voice data, and a third thinks voice data is the most important, that’s just fine. The FCC isn’t involved in that decision, and isn’t looking to get involved. And when a brand new type of service comes along that nobody’s heard of before, the ISPs remain free to decide how (or if) to prioritize it, just like they are today. They just have to treat all providers of that service equally.

The rules are secret.

This one’s half true. The current push is to get ISPs reclassified under Title II of the Communications act of 1934, a law which has been public since 1934. Also relevant is the Telecommunications Act of 1996, also public since 1996. We have tons and tons of existing FCC regulations, legislation, and case law covering many different types of entities in these categories, all public, and this serves as the body of what’s being proposed.

It is true that there are some FCC regulations, mostly (but not exclusively) covering the transition into Title II status, which are not public. This is very common for FCC regulation. I don’t like that, and would rather it all be public, but it’s not at all unusual and to suggest otherwise is incorrect.

Specific arguments

More to come. There’s lots to pick from. In particular, the WSJ has decided not to let me read that L. Gordon Crovitz article. I’ll try again later.

Jeffrey Dorfman in Forbes

Jeffrey Dorfman published a article in Forbes arguing against Net Neutrality. It’s quite bad. Here’s a few reasons why.

Bad analogies are Bad, so here’s a bad analogy!

It’s impossible to ignore the fact that Dorfman starts off his article by complaining about bad analogies, then immediately launches into a really tortured one. Wouldn’t the variety of services you can get access to—which is what Net Neutrality protects—be a better fit for the “flavors” in his analogy? But I agree bad analogies are bad, so let’s just skip over that.

The Internet isn’t the energy or auto markets…

The paragraph about Obama “picking winners and losers” in the energy and auto markets is a total non-sequitur. The Internet isn’t similar to either of those in any significant way. I can’t help think this line of reasoning is simply about whipping up particular political support, without addressing the actual issue.

…but Net Neutrality doesn’t pick winners/losers, anyway.

Dorfman claims that Net Neutrality “stops any possible innovation”, without any support for that claim and, more importantly, in contrast to the actual history of the Internet. The ‘net we have today developed, until relatively recently under more-or-less neutral conditions, and I don’t think anyone could argue with a straight face that it hasn’t been innovative.

Can we stop with the name calling?

“Rabid”. Right, everyone who disagrees with him must be foaming at the mouth. Couldn’t it be that the vast hoard of supporters—including, overwhelmingly, people who are really, really well-versed in this stuff—simply have a reasoned opinion different from him?

He ignores market realities.

"If an ISP blocks Netflix  because of the bandwidth it
requires, consumers who want Netflix will take their
business elsewhere."

Most Americans live in a place with 1-2 options for broadband service. The market forces here are very weak. And over the past decade or so—especially the last few years—we’ve only seen further consolidation. At my house, my choices are AT&T or Time Warner, both of whom, historically, have behaved very similarly in these matters. If they both slow down Netflix, what should I do then?

He ignores actual Net Neutrality violations.

This is not theoretical. He says:

"...the companies that control the Internet backbone
 infrastructure that knits everything together do not
have the power to pick winners and losers either."

Perhaps not in an absolute sense, but in practice they very much do. You don’t have to look any farther than last year’s very public Comcast vs. Netflix fight. Comcast was shown to be badly throttling Netflix’s traffic—not high-volume streaming video generally, but Netflix in particular—until Netflix agreed to pay the extortion money Comcast was demanding. This is the largest such violation in the Internet’s history, and Netflix caved because they knew that Comcast, as the ISP running wires into the homes of millions of Americans, very much does have the power to pick winners and losers today.

He’s wrong about how existing services emerged.

It isn’t true that Facebook, Instagram, Twitter, and friends “emerged without the benefit of net neutrality”, as Dorfman claims. First, the FCC has had different levels of such protections in place for years. They’ve changed over time, but they’ve been there from when the Internet went public until January of 2014, when a particularly bad court case eliminated basically all of them.

What’s worse, Dorfman is conflating the existence of Net Neutrality with the government protection of it. When the Internet got started, it was entirely neutral. The un-leveling of the field has been a comparatively recent thing (since about 2007), increasing over time. Government action here is aiming to recreate that neutrality, under which all those services Dorfman cites flourished.

He misrepresents how the regulation works.

"The problem with government regulation of the
Internet is that by the time the government studies
how it works and what is needed, technology has
moved on."

This is just plain false. Net Neutrality doesn’t require any government body to study any new format, protocol, whatever. It does not seek to regulate any of that, and leaves such decisions up to the ISPs. Net Neutrality is about treating the endpoints equally. No new protocol, format, service, or other technical development will change the value of that.

That last paragraph…

So much wrong with this one. Let’s just walk through…

"As long as the government enforces the antitrust laws
and ensures that consumers can choose among methods
and providers for how they connect to the Internet..."

I’m willing to bet anyone who’s been watching the developments in Internet service laughed out loud at that one. Antitrust enforcement in this area is awful, and has only been getting worse for the last dozen years or so. Even if Dorfman’s condition here were sufficient (it’s not), it isn’t even close to fulfilled.

"...consumers can pick winners and losers by voting with
their time, their eyeballs, and their dollars."

See the “The market will sort it out” point above for why this doesn’t work, even if consumers had sufficient choices in broadband providers to exert useful market pressure.

"No government needed, thank you very much."

Wait, so as long as the government maintains some conditions, we don’t need the government? Again, this is just pandering to his like-minded political cohort.