Scattered Thoughts on Net Neutrality

I have a few scattered thoughts on net neutrality. It seems to have left the news cycle after a brief flare of outrage. That’s fine, by the way. I don’t mind commenting on “old news.” It’s more useful to thoughtfully research something and try to actually learn something useful than to shout slogans and emote at the news. My impression is that 95% of the pro-neutrality people on my Facebook feed have no clue what they’re talking about, and the few knowledgeable techies who know how the internet works don’t understand how markets work. This will not be an essay, just some scattered thoughts. 

Zero-rating is fine and customers seem to love it. But strictly speaking it violates net neutrality. If a service provider’s data plan allows you to stream infinite Youtube videos, or play unlimited songs on a music app, or doesn’t count your Facebook use against your data plan, it is treating some data different than other data. But customers seem to love this. A friend was telling me how awesome his new data plan was because it zero-rated his music streaming app. He loved this feature. He didn’t even realize that it was connected to the “net neutrality” issue. I suspect he has a lot of company. I suspect that if you polled some random people and told them “X, Y, and Z won’t count against your data plan,” they’d say, “Great!” Most of them would not say, “This opens the door to an internet dystopia of walled gardens where a few media moguls decide what you can and cannot do online.”

Walled gardens tend to get broken down by the market. Theoretically a monopoly internet service provider could tell you, “You get this service for e-mail, this service for video streaming, this service for social media…” and could simply pick its own services or the services of a partner company. But people want to pick and choose a little. AOL tried to be a walled garden but failed. Thomas Hazlett, in his book The Fallacy of Net Neutrality, describes a similar process happening in Japan’s internet service provider market. Supposing a walled garden persists and is not broken down by market forces, that probably means that internet users don't mind them so much and ISPs are basically giving their customers what they want. 

If you’re seeking a knowledgeable critic of net neutrality , Thomas Hazlett is the guy. He’s a former economist at the FCC. His above-mentioned book is only 53 pages, and he has a longer book covering the same topics that I haven’t yet read. Here he is on Econtalk. I’m sure for technical, knowledgeable advocates of net neutrality, the knee-jerk libertarian position (that the government is generally bad at regulating things) is pretty unsatisfying. I think Hazlett does a good job of specifically addressing the concerns of neutrality advocates, not simply starting from a libertarian premise and pretending all policy implications fall out from there.

Without monopoly power, the case for net neutrality falls apart. If there is more than one ISP trying to sell you an internet connection, neither of them can simply call the shots and have their way with you. If there’s even a limited demand for a neutral internet, one ISP can steal a lot of customers from the other by providing it. “But what if all ISPs do the same wrong things?” Ok, sure, it’s possible, but not likely. Everyone is foregoing an obvious profit opportunity? And for a service that people are loudly announcing that they want (namely a "neutral" internet)? Sure, it's possible to get stuck in a bad equilibrium, but you need some explanation for why the equilibrium persists when any party could benefit by unilaterally breaking free of it.

I’ll submit that ISP’s don’t really have monopoly power. You can get a pretty big data plan on your phone. It’s not quite the same as having broadband through cable, but if you just read articles all day and listen to a few podcasts, this probably meets your needs. If even a small percent of internet users meet that description, that’s powerful competition that the cable-based providers have to contend with. Alleged “monopolies” are always too narrowly defined. There are less-than-obvious competitors to your cable company. You can rent or buy DVDs rather than streaming movies, sometimes getting them free from your local library. You can get your news from newspapers and magazines rather than the internet. Even if the cable company completely corners the market for internet service in your town, you can get most of what you get from the internet in other ways, some only slightly crappier than the internet versions of those things. This is a point I am stealing from Thomas Sowell, who pointed out that perhaps plastic or other metals can sometimes substitute for aluminum. This potential substitution limited the so-called “monopoly power” held by ALCOA, for which it was sued under dubious legal reasoning. At any rate, mobile internet will rapidly improve in the coming years and decades. A bad regulatory regime, such as “net neutrality done wrong”, could lock existing technology into place.

On that note, my understanding is that most of the bad behaviors neutrality advocates warn us about are already prohibited under existing anti-trust and deceptive practices law and regulations. If an ISP says, "Stream as much high definition video as you want, but if you want quality you'll need to purchase a fast lane," that's not deceptive. But if the ISP clandestinely throttles Netflix so as to make it unusable in a scheme to make you purchase more cable channels and Watch Instantly movies, that's deceptive. It's not actually providing the service that you paid for and is probably breaking the law. (I'm trusting the opinions of people who know the law here, like Ajit Pai and Thomas Hazlett. If someone knows a good contrary source on this point, please share.) That's not to say I like the idea of a government that aggressively pursues monopolies. The history of anti-trust action by the US government is kind of a joke.

ISPs are very high fixed-cost, very low marginal-cost entities. Even more so than utilities, like water or electricity. You have to lay a lot of expensive infrastructure, but then you can basically pipe out the service almost for free. Economic theory tells you that this gets you a "natural monopoly," because it doesn't make sense for a second provider to build all that infrastructure knowing that the marginal cost (=price) is going to be zero with even one competitor. So the notion that ISPs are some kind of monopoly isn't totally crazy, by any means. I just think that the policy implications are not so obvoius.

Internet fast lanes are a perfectly reasonable idea. Suppose you have a robot surgeon in Indiana being manipulated by a physician in New York. You’d want a dedicated, high-bandwidth connection. Less importantly but no less obviously, a high-powered gamer might want a similar service and be willing to shell out for it. Frankly, if there are bandwidth hogs who are slowing down my internet connection, I think I want my ISP to identify them and throttle them or otherwise get them to purchase a fast-lane. If these people are imposing a cost on other internet users, it’s perfectly fair to treat them “non-neutrally.”

Some people think that non-neutrality and fast-lanes means that ISPs will throttle your existing connection until you buy a “fast-lane”, which will only get you back up to the speed you’re used to but at a higher price. Or they think it means they’ll charge everyone more for the same service, but offer the former price for a slower service (that has been deliberately throttled). Somehow non-neutrality means things can only get worse: pay more for the same service or pay the same for less service. I think this is nonsense. Even assuming ISPs hold “market power,” this isn’t even an accurate description of how a monopolist tries to maximize its profits. Non-neutrality would probably mean more tiers of service. That’s a good thing! Scenarios where this makes everyone worse, or even scenarios that make the typical customer worse off, make little sense.

I think it’s funny that everyone was mad at cable companies for not allowing you to purchase channels a la carte, and now people are worried about the internet being sold a la carte. Maybe it’s different people, but I smell a deep inconsistency here.  Speaking of which, cable companies in effect “dedicate bandwidth” to your television and telephone service, which is non-neutral. I don’t see the big problem in going the extra step and treating some apps or websites different from others. Cable TV is just an app for streaming video. Phone service is just an app for making calls. Maybe I want dedicated bandwidth for streaming audiobooks, too. 

I also think it's kind of funny that people were cheering for Go Daddy and Google when they effectively kicked The Daily Stormer off the internet. Maybe my saying this will earn me a pedantic lecture about how a registrar is different from an ISP, but this would be to miss the point. Should internet choke-points be used to determine content, or shouldn't they? Why insist on "dumb pipes" but not "dumb registries?" Isn't this kind of editorial decision-making exactly the kind of thing neutrality advocates are worried about? That's what supposedly makes a neutral internet so useful: tiny niche groups can still get together and communicate without being overwhelmed. With a neutral  internet you get 100 million tiny "channels," rather than just ABC, NBC, and CBS. And wouldn't it be a shame if the big players stifled this diversity and creativity? Maybe it's okay if this kind of awesome power is used for good but not evil? (I hope everyone feels deeply uncomfortable about making that kind of argument.)

My cable company sells me movies for a pretty hefty rental price, about $5. It is, in a very obvious way, a competitor with Netflix. It would sell me more movies if I didn’t have the Netflix option, or if they throttled Netflix so much that it became unusable. So far they have declined to exploit this opportunity, despite being a “monopoly” in the narrowly-defined market of streaming video.

Many big internet companies are pro-neutrality, particularly the big bandwidth hogs Netflix and Google. I see “net neutrality” as mostly a fight over who is going to pay to upgrade the infrastructure. Netflix wants Comcast to upgrade its own damn pipes, and Comcast wants Netflix to shell out the cash for some of this since they are the big bandwidth hogs who are forcing the issue. Netflix and ISPs actually do work together to solve some of the bandwidth issues; Netflix installs a caching appliance at local ISPs. I understand none of the details about how this works, but the take-away is that both parties realize there is a problem and are working to solve it. Netflix wants its customers to have an enjoyable viewing experience or it will lose them as customers. Comcast realizes that a huge fraction of internet usage (and thus customer value) is thanks to Netflix, so they want the same thing. I have no dog in this fight. I don't really care whether Comcast or Netflix "pays for" the infrastructure. I feel like the net neutrality people are telling me I should care without really telling me why.

Some of the best information I've gotten on this issue came from the comments section of a Coyoteblog post, probably this one. I think the guy basically has his economics and his political economy/public choice theory right. But maybe he flubbed some of the technical details? A bunch of techies show up in the comments to "correct" him. Although often the technical details don't actually refute the economic arguments, which revolve around monopoly power. This lesson generalizes pretty well: if you have a good filter and cross-checking habits, you can learn a lot from reading the comments. If not, don't bother, because you'll just get frustrated and give up. Anyway, this kind of back-and-forth is often instructive. On this topic, I see the occasional pro-neutrality techie sneering that "You just don't understand how the internet works." I'd like to sneer back, "You just don't understand how markets work," but I doubt that would be very productive.

I wish I could offer you a neat, flowing essay tying this all together, but instead I've written another long-ish post filled with random thoughts. Bless you if you've read this far. Feel free to offer any useful information in the comments.