Net Neutrality Reexamined

Who wants anything less than free and unfettered access to anything they want on the Internet? Well, nobody, of course. Everyone agrees that we want the fastest, freest Internet we can get. The problem is we don't all agree on how to get there, and we don't all understand the implications of solving the technical challenges required. The central question is whether government regulations should require service providers to give consumers access to the Internet with nothing blocked out, and nothing slowed down. Today we're going to log in to net neutrality, and take a deep dive into the issues driving one of the hottest topics around right now.

This take is going to be a bit different from others you may have heard. Rather than look at the history of net neutrality, current events driving it, and the laws, we're going to organize the issues visually to better understand the overall picture. The way I propose to think of net neutrality is to group the issues into a 2 × 2 grid. The left column is the arguments in favor of regulated neutrality; the right column is the arguments opposed to regulated neutrality. The top row contains the ideological concerns, and the bottom row contains the technical issues.

Let's look at those top two squares. Our ideology row is, unfortunately, a dichotomy of straight-up political party lines. Liberals tend to favor neutrality through regulation, and conservatives tend to favor neutrality through competition. What this debate is really about is government regulation versus the free market, and you can blame these two most explosive of all political terms for why net neutrality has garnered so much public attention. For a conservative, there is no more antagonizing red flag to a bull than the term regulation; just as to a liberal, there is no more antagonizing term than free market. Advocates of regulation believe that customers are best served when the government requires equal services be offered to everyone; and advocates of the free market believe that service providers competing for customers will naturally bring prices down and service quality up. And in the United States, that's about as close to a 50/50 split of the population as you'll find anywhere.

We have one big stumbling block when it comes to evaluating the effectiveness of the free market in providing unrestricted Internet access. Free market solutions are dependent upon the consumers' ability to choose, to vote with their wallets. The problem here is that many people do not have two or more broadband providers to choose from. This is a simple fact of limited infrastructure. If there was fiber running to every house in the world, great. There's not. And in a lot of places where consumers do have a choice of broadband providers, often one or more of them are terrible with slow speeds or poor reliability.

How many people in the United States have fewer than two good broadband providers to choose from? The answer is "Ha ha ha." It's probably somewhere between one quarter and two thirds, but good luck finding a trustworthily reported number. Just about every published number you can find has been wildly spun to fit some narrative. Reports from the left claim as many as two thirds of people don't have a meaningful choice; reports from the right claim only as few as a quarter don't. And there's everything in between. But regardless, it's a very big non-zero. Even the lowest estimates represent a lot of people — far too many people for whom we could cripple an important service.

Which raises the question: How important is the Internet? Utilities like water and power can be argued to be life necessities. Even the telephone, since we need to be able to dial 9-1-1 (in the US, apologies to all you other listeners with other emergency numbers). Television, on the other hand, not so much. Many people choose to have no television, unlike necessities like water. What about the Internet? Is it a necessity? It sure wasn't 25 years ago. But the world is shifting. Banking is being done online. Medical and other important services rely increasingly upon online scheduling. Email is a work necessity and telecommuting is sometimes becoming mandatory. Many of those who have given up television and/or telephones have done so because the Internet can replace them. We cannot treat the Internet the same as we treat water, because it's not as essential, but it is becoming increasingly so. It truly is a unique service, and the direction it's going is in the direction of an essential service.

So who wins on our top row, the left or the right? It seems clear that at the current state of infrastructure, where it's not possible for many users to choose between providers, our green checkmark must be placed on the left. Regulations need to prevent Internet providers from blocking or hampering certain services. However, if in the future, infrastructure is improved to the point that everyone can have choices, then it will make sense to have the conversation of how to balance regulation and the free market. Today it does not, and by the time it does, the Internet will be even more of a life necessity.

Now we turn to the bottom row of our grid: the technical issues. On the left, technical arguments for unbridled neutrality; on the right, technical arguments for limited neutrality.

Let's look at the basic example of Internet on a plane. You buy the service, and you find out that streaming video is blocked. No YouTube or Netflix on a plane. Why not? Because you would suck up all the bandwidth, and nobody else on the plane would be able to check their email or work on a Google doc. It's necessary for the airline provider to impose this restriction to keep the service more useful for more people. It's an example of an absolutely necessary fundamental violation of net neutrality.

The airline example is a microcosm of every other Internet environment. Your neighborhood broadband provider has to do this too. They generally do slow abusive users, peer-to-peer file sharing networks, denial of service attacks, even streaming video. It's called traffic shaping. It's done on a dynamic basis all day long, and it's the basic principle for maintaining network quality of service. Your home broadband service would be terrible if your provider was not, right now, prioritizing some traffic, and delaying other traffic.

While most people would agree that traffic shaping is a good thing, many believe broadband providers shouldn't be anti-competitive by prioritizing their own services and throttling traffic that their competitors deliver — such as the infamous examples where certain providers who also offered telephone service were found to be throttling various rival services like voice-over-IP companies.

But because bandwidth management is a full-time, round-the-clock job, there is no clear line that can be drawn between when hampering a competitive service is OK or not. Sometimes the competitor's service has higher usage and it must be throttled. Sometimes it has lower usage and can be prioritized. No legislation can effectively allow best practices in network management, while simultaneously preventing service providers from throttling down their competition. The realities change by the minute. It's really almost an honor system.

Here is an interesting question. Some percentage of Internet traffic is illegal, mainly illegal file sharing, people stealing movies. When data centers employ traffic shaping, should they put a lower priority on illegal file sharing, to benefit all legal traffic? Most people would agree — even many of those who support unfettered net neutrality. But doing so would be incredibly hard. We can't always tell which peer-to-peer file sharing is illegal. And even if we could, what's illegal? Hate speech? Who decides where to draw that line? Pandora's box goes ever deeper.

But furthermore, from an infrastructure perspective, there is no such thing as a "fair" Internet. The biggest content providers — Google, Facebook, Netflix, Akamai, Amazon, and others — are directly connected to Tier 1 networks, the highest-level networks over which data flows worldwide without having to go through anyone else's routing, negotiation, or settlement. You and I don't have that. Some content providers are in countries with poor connectivity. By its very physical nature, the Internet is not, and cannot practically ever be, a fair place — at least not without a lot of traffic shaping.

So who wins on our bottom row, the left or the right? From a purely technical perspective, our green checkmark has to go on the right. The ability for service providers and data centers to freely throttle traffic at their own discretion is super important. If they lose that ability, then no other rules matter.

Cisco is an example of a tech company that advocates for both boxes in the right-hand column of our grid, generally supporting the arguments of market competition and limited neutrality — keeping in mind that Cisco is a company whose business is network management hardware. I'll cite two of their principles listed in their online position paper. For our top row — ideology — they state:

Consumers are entitled to competition among network providers, application and service providers, and content providers.

...thus falling into the "free market" category, and placing them into the top-right square of our grid. For the bottom row, they advocate for:

The ability of broadband Internet access service providers to engage in pro-competitive network management techniques to alleviate congestion, ameliorate capacity constraints, and enable new services.

Again, talking about facilitating competition — the free market — through network health. In other words, Cisco feels that if traffic shaping technology can be employed to keep everything fast, then it's a level playing field for the service providers to compete by offering customers all sorts of packages, unhampered by poor performance.

And, quite predictably, the companies trying to deliver content over Cisco's equipment generally support unfettered net neutrality — landing in the left-hand column. Facebook is an obvious example. In a letter to the world, they wrote in the top-left corner of our grid:

...Net neutrality... ensure[s] that anyone with an internet connection has a fair shot at turning an idea into something that can change the world. That would change if internet providers were allowed to decide what content its customers could access, or charge customers more to access the websites and services of their choice.

Obviously, very much against a "free market" type of solution. And in the bottom left corner of our grid, Facebook gives its technical thoughts:

Net neutrality... ensures that internet service providers are not allowed to block or throttle internet traffic or discriminate against certain content.

I would like to assume that Facebook was taking for granted the fact that traffic shaping is an essential tool for network health, but they sure didn't say so. We have to keep in mind what Facebook does — just as we did for Cisco — and remember that Facebook is already enjoying "king of the hill" status as a Tier 1 service. Theirs is an easy position to take from where they sit.

So here's where I have to try and compress all of this into a coherent conclusion. It is ferociously complicated, and there is no simple solution. Either a purely liberal or a purely conservative version of net neutrality would both be terrible for a majority of users. The best of all practical solutions — were we to live in some utopian universe where such a thing could actually be crafted — would still irk some percentage of users. The irony is that we all want the same thing: freedom to provide and obtain whatever we want with great performance over the Internet. Why can't we have that? Because at the end of the day, it costs a lot of money and resources to provide it, and all the myriad parties involved all have different ideologies and conflicting notions for how best to provide it. That's the reality. What we're likely to end up with is probably little different from what we've always had: a certain amount of regulation, a certain amount of competition, and everybody wishing some component of their service was a little better.

By Brian Dunning
Follow @BrianDunning