Further thoughts on why the end of net neutrality will be fine

If a cable company really is a monopolist, still they (mostly) maximize profit by giving customers (cost-constrained) what they want.  When the de Beers cartel had a monopoly on diamonds, did they also make you buy their favorite soda brand?  No, that would lower the overall value of the package and thus lower profits.

The main exception to this argument is that the monopolist may favor its own content.  Monopolizing instances of that practice still would be regulated under standard antitrust law, and also transparency requirements, and most of the critical discussion seems to ignore this.  Furthermore, it is harder to make a profit this way than you might think.  If Comcast promotes “the stupider Comcast version of CNN,” a lot of people just won’t be interested.  Most of these websites aren’t that valuable — look at the recent revenue results for Buzzfeed.  Nor do I think Comcast can get away with denying its customers say Google or Skype, either legally or economically.  That said, advocates of removing “neutrality” need to face up to the reality that they will be relying on discretionary regulation to a greater degree in some regards.  Read p.1 of the actual proposal:

Restore the Federal Trade Commission’s ability to protect consumers online from any unfair, deceptive, and anticompetitive practices without burdensome regulations, achieving comparable benefits at lower cost.

In the current debate, there is a common presumption that paying for slots hurts “the little guy.”  During the payola debates for radio, it turned out that payola favored the independent labels over the majors; see my book In Praise of Commercial Culture.  It doesn’t have to work out that way, but refusing to price scarce resources often helps the big established players, who can invest $$ to get what they want through bigger brand names or other means.   Note:

Pai says that one of the major mistakes of Net Neutrality is its pre-emptive nature. Rather than allowing different practices to develop and then having regulators intervene when problems or harms to customer arise, Net Neutrality is prescriptive and thus likely to serve the interests of existing companies in maintaining a status quo that’s good for them.

Furthermore, are there external benefits from small web upstarts?  Or are the external benefits from the big superstar internet companies?  If you are a Progressive who loves stable jobs and decent wages, you might think the more significant externalities are from the superstar companies.  Yet when it comes to net neutrality, all of a sudden the smaller companies are glorified and we need an ecosystem to foster them.  Overall, I don’t trust the regulators to make these decisions well, so I would rather take my chances with the market, even with some monopoly power at the cable end.

As Megan McArdle points out, over the last ten years consumers have opted overwhelmingly for the non-neutral private garden of Facebook.  That’s the real “threat” to net neutrality.  Personally, both as internet writer and user, I much preferred the older, semi-open, more neutral architecture of RSS and related systems.  The masses have spoken, however, and quite decisively in favor of less open systems and apps.  Nonetheless Alex and I still can do our thing on MR and in fact the project is thriving, and I would be shocked if it did not survive the new FCC decision.  That said, people want non-neutralities and they will introduce them to internet systems one way or the other, and suppliers will have to find ways to cope or perhaps even benefit.  To believe it could be any other way is a kind of wishful thinking, yes I want those old usenet groups back too.  All things considered, “net neutrality” is a biasing term, because the 2015-2017 period was by no means neutral either.  The notion represents a kind of undeserving “victory by language,” as who would wish to favor “bias” over “neutrality”?

Perhaps this point is misused a bit to make extrapolations, but still it is worth noting:

Pai…noted that today’s proposed changes, which are expected to pass full FCC review in mid-December, return the Internet to the light-touch regulatory regime that governed it from the mid-1990s until 2015.

More generally, I don’t see anything intrinsically morally wrong with a person deciding to “buy only one third of the internet.”  How many net neutrality supporters also favor or maybe even insist upon a’la carte pricing for cable TV?  What percentage of the public library do they take home over the course of their lifetime?

Or think of the whole issue in terms of a regulatory principal-agent problem.  Let’s say the water company has “too much” market power, and the public regulator doesn’t have the will or the resources to constrain the company properly.  Said company refuses to let Perrier flow through the pipes as an alternative option to plain tap water, for fear too much of the profit would go to France.  That somewhat mirrors potential problems from net non-neutrality.  But is it likely that a zero price for water is close to the correct solution?  I do get that alternative solutions might in some ways involve greater faith in outside regulators, such as antitrust authorities, but zero price is an awfully blunt instrument for a rapidly changing setting such as data flow.  It certainly hasn’t worked well for water, in a wide range of settings.

Finally, Viking notes in the MR comments:

The real benefits of net non neutrality would be applications that require a guaranteed minimum latency. Non net neutrality would allow some market participants to pay more for reduced latency, which could benefit video conferencing, virtual reality, remote surgeries, VOIP (already part of video conferencing) and other possibly new applications, say remote monitoring and control various kinds.

Are the defenders of net neutrality considering those opportunity costs in their assessments?  I don’t see it.

To be sure, net neutrality really might be better.  You might have a high opinion of the net neutrality regulator and a low opinion of all the other regulators of unjust or inefficient conduct.  You might think bandwidth won’t become scarce anytime soon, and that new, alternative uses for greater bandwidth just aren’t that promising.  You might think that access auctions disadvantage “the little guy,” and furthermore the positive externalities are on the side of the little guy, and thus we should stifle price-based access auctions.  You might think that rationing on a quantity/access basis will be more fair or efficient than rationing by price.  All that is possible.  But it seems hard to know those claims might be true.  Instead, those comparisons would seem to suggest a fair degree of agnosticism.  But when I read proponents of net neutrality, I am more likely to see a harsh excoriation of commercial incentives, or cable companies, than a balanced weighing of those considerations.

Neutrality ain’t neutral, it’s time to get over that myth.

Here is Tom Hazlett on the topic, here is my earlier column.


An arguments in favor of net neutrality that wasn't fully addressed in this piece is that you do not want to let monopolies set up a profit maximizing price discrimination systems that work by taking away functionality from lower tier services.

Even though possibly abuse is worrisome, anyone concerned with separation of powers should look askance at the idea that the FCC is being encouraged to reinterpret at 1930's era law instead of letting the legislature address the issue today.

Tyler did note that conventional antitrust law would still apply.

And many see price discrimination as a natural way to optimize social welfare and it may not trigger conventional antitrust law. Despite being usually legal, the combination of price discrimination + monopoly is responsible for some of the worst aspects of pseudo-markets, with healthcare in the US being one of the most broken examples. The absence of net neutrality in markets with monopolists creates a new playing field for that combination.

There is a broad category of economic thought that is so fond for making sure investments get made and markets clear that they don't care as much as they should about what happens to the consumer surplus.

"Refusing to price scarce resources often helps the big established players." Go back over the part where internet bandwidth is a "scarce resource." Why is the bandwidth customers use to access (say) my own small-biz website a scarcer resource than the bandwidth they use to access the cable company's preferred sites?

Or do you mean customers are the "scarce resource"? The cable company will make me pay extra so my website will be available to its customers; is that because its customers have suddenly become a "scarce resource"? If so, why are the cable companies doing fine right now, when I can reach all their customers the same as any other online business can?

If Comcast promotes “the stupider Comcast version of CNN,” a lot of people just won’t be interested.

They won't have a choice if Comcast is the only provider in their area offering wired broadband, or one of two in a duopoly of providers both doing the same tactics.

Comcast is the only cable TV provider in many localities. Are there instances of it using that position to disadvantage CNN (or some other channel) in favor of some Comcast-preferred alternative?

But cable is already terrible so if you're argument is "it will be like cable" that's a pretty bad one.

Errr yea maybe not CNN but all the time I'm seeing commercials on TV stations asking Comcast or Dish or whoever customers to call because they may loose the channel if negotiations don't go their way. Cable companies are clearly an example of what you get with a non-neutral net.

What? You mean people will use the Comcast news page instead of Facebook?

Are you kidding? If wired service sucks people will use their phones more.

The only people who are in a position to dictate are the ones who have a large number of users who want to use their services.

Comcast would find out very soon how irrelevant they are if they raise their prices enough to encourage another technology that puts them out of business.

I actually find Tyler's arguments pretty convincing, but you raise a great point that is I think broader than net neutrality.

Why is Comcast the only provider? In my city we have like 6 at least.

In many places across America, internet providers have locally-granted monopolies.

And there's your problem.

And the local granted entry barriers are still a big deal today and deserve to be the center of disucssion:


That said, it has further perked my interests in the federal preemption of local entry barriers staring in the mid 19th century until the early 20th century.


Internet access is more of a necessity than diamonds are. The de Beers analogy is a bad one here.

It's also bad because we have a closer analogy. CableTV services don't sell me what I want so I dropped cable. Many did. My internet provider is a cable company and they hard sold me on an internet/tv bundle for ages. Hopefully they will leave me the option to purchase straight up internet going forward without too much price rise. Luckily there is now competition in my area (there wasn't 2 years ago) so I'm probably safe. My suspicion is that this will be a good thing and will open up better high end services to people who want to pay for them. Kinda like Singapore and healthcare.

I completely agree, Bob.

deBeers wouldn't force the hypothetical diamond/soda package because it doesn't work to maximize their total revenue. Diamonds are luxury goods and therefore have a far higher price elasticity of demand than something as necessary or ubiquitous as the internet. ISPs are far more likely to have internet "packages" because they're far more able to - people will still pay for internet, regardless of what "packages" ISPs force on them.

Studying telecom engineering in college on the 90s, the future was QoS ("quality of service") segregation: users could pay more for low latency for video calls, or pay less and be pushed to the back of the queue, which is ok for say nightly batch file uploads. But it never happened: it is technically and, above all, organisationally difficult to implement, and the Internet got fast and thick enough (people don't realise those are separate things!) for the services people want.
You can see what net neutrality looks like to the user in places like Peru. Cell phone companies offer "the basic standards": Facebook, WhatsApp, Twitter, and maybe a couple more, for "free". That clearly favours established companies over newcomers: if it had happened a few years earlier, it would have been Microsoft's chat (but I have forgotten the name!) and WhatsApp would not have had a chance.
In a larger, more mature and complex market like US, many other things would happen, but I can imagine they would be on the same direction: pay-to-play, bilateral deals that put a big company closer to the client over the rest. But that is already happening: you can pay a CDN (content delivery network) to host your content on a network of data centers, closer to the users so it arrives faster and looks more responsive, which is critical for things like Google Maps.

But it never happened: it is technically and, above all, organisationally difficult to implement, and the Internet got fast and thick enough (people don’t realise those are separate things!) for the services people want.

This is correct. The late 90s were the golden age of quality-of-service and bandwidth management protocols and startups. Is there anyone who can fail to be impressed by IntServ, the RSVP protocol, and those beautiful equations to compute upper bounds on latency? But it was easier just to deploy bigger and bigger pipes.

And today the 5G cellular people and the NGPON2 people are dying to sell their equipment to stingy telcos - but the bandwidth is simply not scarce enough to justify to capex.

That said, bandwidth scarcity is not the only or primary driver for non-net neutralilty.

QoS across networks is difficult but it's being done (slowly) because the needs are so great. The 3GPP LTE standards have several classes of priority in order to encourage real-time IP services. (Verizon announced this year that a majority of their voice traffic is VoLTE, which is pretty impressive considering VoLTE is used across carriers.) The current FCC rules bias new services towards cache-able services. Great for Netflix; not so great for interactive, real-time services. "Live" streaming today is often at a 15-45 second delay and this is simply unacceptable for, say, interactive remote classrooms, teleconferencing, real-time narration for the blind, etc.

Further, QoS is used frequently for on-net traffic in order to improve quality of experience. For instance, to make their IPTV service (U-Verse) viable, AT&T used DiffServ to segregate their pay-TV content from Web Internet traffic. And T-Mobile and YouTube a few years ago were doing some useful experiments with TCP proxy bypass that improved YouTube streaming and improved users' phone battery life. They immediately stopped when the 2015 rules were announced because no one (including, as he admitted, the then-FCC chairman) knows which network management practices the "Internet conduct standard" will reach.

As for the Peru and Portugal scare stories of a walled-garden Internet, they're a distraction. The 2015 "net neutrality" Order permits walled-garden Internet and ISP blocking. The Obama FCC was quite clear that filtered Internet is permitted. In fact, filtered Internet is less-regulated than conventional Internet access (this asymmetry is one reason the Trump FCC wants to repeal the current rules).


QoS across networks is difficult but it’s being done (slowly) because the needs are so great. The 3GPP LTE standards have several classes of priority in order to encourage real-time IP services.

Prioritization is not new and is not QoS. MPLS FECs are closer, and they have been around for a long time.

For instance, to make their IPTV service (U-Verse) viable, AT&T used DiffServ to segregate their pay-TV content from Web Internet traffic.

If they really used Diffserv I will eat my Portishead CDs. Probably they do prioritization according to the DSCP bits. No PHBs.

AT&T says they used DiffServ for U-Verse:

"AT&T combines Layer 3 DiffServ functionality with Layer 2 mechanisms to separate its U-verse 'triple play' platform into logically discrete voice, video, and Internet access streams and provide each service the network performance it needs to meet customer expectations."


From Page 51 it's apparent that by DiffServ they just mean "differentiated services" in the IP layer i.e. using the DSCPs like 802.1p bits.

Then in footnote 83 they claim that Diffserv is something else by citing a Nortel (!) marketing (!) document from 2003 (!).

But really there is no way that they could be using Diffserv in the sense of RFC 2474, which regards prioritization as the "legacy" use of the TOS field (see 4.2.2)

What an amazing ramble. You jump from one facetious point to the next without completing any of them, and the worst part is it doesn't sound like you really understand how the internet works at all. Comcast doesn't pay google for the right to "carry google." Comparisons to cable companies show an incredible misunderstanding of the issue.

Internet access is the sale of bandwidth, not content. Cable companies are fighting to make internet access the same as cable as their existing monopolies crumble under consumer choice.

With you 100%. Tyler is usually on the ball, but he seems to be way out of his depth here.

Lots of anger, no substantive response. Experience with "natural monopolies" elsewhere absolutely is relevant, no one is saying "the internet works just like those."

This has nothing to do with 'natural monopolies.' The Internet (in a simplified view) currently treats all data packets equally, and all Internet users are free to request and send data to any IP address. Yes, China is an exception, but certainly it should be no hurdle for everyone to agree that China's Great Firewall is not a desirable direction for the future of the Internet. It might help to understand what the Internet actually is - a way for any connected digital device that uses shared protocols to request and receive data from any other connected device that uses those protocols. This is why a protocol like Bittorent can take off, as it provides a more efficient way to share data, without requiring a new network to be constructed.

The Internet is an idea of how to transfer data packets, no more and no less. At least until recently, it seems. Though the truly insightful economist would be much more interested in asking why some people seem interested in trying to create an unnatural monopoly in place of the current Internet framework, which treats all data packets equally..

Looks like Tyler has been deleting my posts again. Happy thanksgiving.

I think by Tyler was referring to the possibility that Comcast demands fees, Google refuses to pay, and so Comcast does not carry Google (hypothetically). If Comcast can demand fees then probably they will carry less overall, since some will not pay.

You wrote: "Internet access is the sale of bandwidth, not content." Maybe that is literally true, but what does it mean? The whole reason people want bandwidth is to access content.

Facebook/Youtube/social etc are not unlike bandwidth. At their core, they are just a platform to host/access content. Obviously, these companies are not so rich because of the platforms themselves... They are in the content selling business--people make it, FB/YT takes it, bundles it and sells it as they sees fit (for ad money instead of a monthly fee).

Few people complain about social media the same way they do ISPs, but I think they're not that different. Facebook has made it almost impossible to see anything on their website without making an account. They are selling access to content.

...by *no Google* Tyler was referring...

I mean seriously. What kind of word salad is that Perrier analogy? Water companies don't transport discrete 'units' of a product, why would the company care where profits were going, and no one is suggesting anyone transport anything at no cost.

Come now, these are unnecessary observations to make at the best satire site on the web.

Looks like Mercatus getting funding from the telecom industry these days.

I would suggest checking the center's 990s, but a lot of the money that flows to Mercatus comes from the GMU Foundation in a lump, making the specific mix difficult to determine. Undoubtedly only by coincidence, of course.


It's easy to imagine a business model for an Amazon-only or Facebook-only ISP.

Does it really make sense to make that kind of ISP illegal??

This already exists in Portugal.


There is reddit meme showing some Portuguese ad for internet with add on fees for websites. It's prime example for NN supporters.

'If a cable company really is a monopolist'

Best satire site on the web, bar none.

Really prior? I use three network carriers in my daily activities.

Any one of them could be shut off. They know that and are acting in ways to make me not want to do that.

Since when did a network provider become a cable company with a monopoly granted by a locality?

As was the case in the past, though these days, competition as overseen by the local government is proceeding apace - 'There are three franchised cable service providers in Fairfax County. Comcast provides cable service in portions of the Reston area, while Cox provides cable service in all areas of the County other than the Reston area. Verizon is currently providing competitive cable service in many areas of the County and will provide service throughout the County by 2012.' Of course, considering that Bell Atlantic is now Verizon, and that your telephone company was the other corporation previously granted a monoploy to provide a wire leading to your location, it is not precisely competition that the same people who held the monopoly on the two wires into your house are competing with each other.

And of course, Cox advertises and provides free calling services, like a German cable company/ISP, right?

So change the subject then. I don't care how the signal gets to me, if it is strung on poles or wireless or copper pairs or glass fibers underground or a line of sight wireless.

Cable networks are valueless unless someone buys the product that they carry. The video feeds that were the origin of the business are getting less valuable. Providing internet is getting more valuable. I said I have three providers including the local cable network as a source of internet access. They all have to compete for my business.

Everyone investing in providing bandwidth makes the experience better for everyone. Net neutrality doesn't encourage that, so for all the nice thoughts it will end with a worse experience. Something like the lofty nonsense that communism spouts, which you have fallen for in spite of experience.

'So change the subject then'

How? My comment regarded a cable company being a natural monopoly, or at least a government granted monopoly as was the case throughout the last several decades - at no point did I mention Internet services, though you did.

' I said I have three providers including the local cable network as a source of internet access. '

And below, you mention that at least one of them is clearly engaging in the sort of deceptive marketing that is forbidden in Germany.

'Net neutrality doesn’t encourage that, so for all the nice thoughts it will end with a worse experience.'

This would be a lot more effective argument if North America actually had the sort of high speed Internet access taken for granted in places like South Korea or Scandinavia or Switzerland or the Netherlands or Ireland, places where debating 'net neutrality' in terms of Internet access being provided as a basic utility is not a discussion.

'Something like the lofty nonsense that communism spouts'

What does your ISP providing the service it actually deceptively marketed have to do with communism?

Really? I want to dump VZW overy net neutrality, but what is my option? AT&T? Sprint and T-Mobile used to support net neutrality (kind of), but they've jumped that shark. Four options and none, is what I'm left with.

'consumers have opted overwhelmingly for the non-neutral private garden of Facebook' - by using an ISP who is currently not allowed to restrict or slow down access to any site that an Internet user want to receive data from (and at the contractually promised data rate, but deceptive practices in the U.S. is another subject entirely).

'That’s the real “threat” to net neutrality. ' - You really seem to have absolutely no idea of what you are talking about when it comes to changing the underlying theory and structure of how the Internet currently works. Net neutrality has nothing to do with any Internet site's walled garden.

'preferred the older, semi-open, more neutral architecture of RSS' - Absolutely no clue, even while apparently advocating for a certain policy framework.

'That said, people want non-neutralities and they will introduce them to internet systems' - Just like people demanded Digital Restrictions Management systems, until the MP3 format mysteriously triumphed anyways, after being tarred for years as an illegal format. And to stay on topic, for a number of years that audio format was also not allowed by various Internet filter systems - mainly corporate - since MP3 was considered to be only used for music piracy.

In agreement with comment by Brett, really an unfortunate half-baked set of "further thoughts" that need yet a bit more thought before sharing. The water pipes analogy was reminiscent of senator Stevens description of the Internet as "a series of tubes". Even toll roads would have been a better analogy. Tyler, you can do better than this!

Sadly, no, he really cannot, which is why treating this as satire is the most charitable Straussian reading possible.

Seriously, what is your god damned problem?

Further thoughts? Further thinkin ain't gonna help here. He needs a first-year course in how the internet works.

Is the Internet broken? Does it need to be fixed?

This is the key question, especially for those who can't understand simple analogies. Antitrust laws already handle the one true concern where an ISP would favor one content provider over another. The rest is through a misunderstanding of how the internet even works. The quote yesterday from the Cisco engineer that full implementation of net neutrality would set the internet back 20 years is a step in understanding. This regulation by people who like to regulate. Quality of service is your friend. There are some aspects of it in place now, but Viking had it right yesterday.

Net Neutrality doesn’t mean there can’t be QoS. This is one of the common misconceptions that I think the providers love to push.

Tyler argues that eliminating net neutrality is going to be good for small/independent firms and bad large established companies. I’m interested in hearing why he thinks they’re such strong advocates for eliminating these regulations if they actually benefit them?

QoS prioritizes certain packets over others, in violation of net neutrality.

If your concern is that a content provider gets throttled, Tyler covered that with his antitrust comment.

Some argue for a Net Neutrality (TM) with no latency cues, but we don't have to ask for that kind of net neutrality.

Still, I think I see where the are coming from. If someone says "make my packets highest priority" and a whole slew of other people say "no, me" who do you choose?

Note that if you give the last mile provider the ability to choose, that isn't and end to end solution. The last mile is only 1 of N carriers.

The last mile provider has a choke point more than an end to end architecture.

Your concern is valid, but I would not put it on the content provider to require lower latency, but jst route packets according to their headers. Router know what kind of traffic they are passing. Voip first, video (which can buffer a bit) and then other traffic. Even gaming can be recognized.

That's the question we should be asking. I'm undecided on the issue due to my limited knowledge of the technical side but I judge the pro neutrality side to have the better arguments. Tyler's argument essentially amounts to "let's give them the power to step on us because they probably won't because of this apples to oranges comparison."

In a peer to peer network, latency can only be a "hint" for the routers. Because sure, good actors will say their live video should be "fast" and their advertising video may be "slow."

But bad actors will switch those and others.

Ultimately bandwidth sufficiently in excess of requirements is the only thing that can deliver "on time" content you consider valuable.

But imagine that someone discovered a way to reduce those bandwidth requirements, and ever so coincidentally just happened to reduce the bandwidth used to transfer data from addresses that the person in charge of filtering/throttling considers unimportant. You know, a person at Comcast deciding that Netflix uses too much bandwidth, and instead of Comcast building their network in a way that actually fulfills their marketing promise of 1/10/100 mps data transfers to all customers, actually simply reduces bandwidth from sources Comcast considers wasteful.

To argue the provider side for a moment, their promise of X speed for you was always probabilistic. It assumed that Y percent of subscribers would request X at a time. It allowed a conservative trunk architecture.

Netflix surprised them because suddenly many people were requesting a lot of bandwidth. They had to build out to keep up, even though they were still just promising X and delivering less than X for all those "bad" Netflix users.

I feel for them, but yes they were promising X, and "but not video" was not in their customer contract.

'their promise of X speed for you was always probabilistic'

True, but in Germany, that was declared to be essentially fraud, and if an ISP says they will provide 10/100 mps service, they are not allowed to fall under that by arguing that their ads were deceptive, and that they were actually counting on customers using considerably less capacity than they had available.

By the way, if you ever hear providers complaining about "bandwidth hogs" those probably are just other customers making fair use of their X bandwidth.

They are just increasing Y, and of course the provider hates that.

This is the practical nub of the net neutrality debate: who pays the costs of building out the system to accommodate the demands of Netflix uses? We've never been in the position of one behemoth requiring so much bandwidth before. Net neutrality advocates insist that everyone pay the same (i.e., that non-Netflix users subsidize those costs), whole opponents would permit the imposition of those costs on the particular users.

"while," not "whole." Sorry.

I think it was kind of a one-time event. Newer (and "updated") ISP contracts specify both a bandwidth limit, and some kind of data limit as well. The data limit may trigger a pricing tier or down-throttling. I think either are fine, and by my lights "neutral."

Now that those are in place, I think it is much harder for any last mile provider to claim that new uses are unexpected.

"Net neutrality advocates insist that everyone pay the same (i.e., that non-Netflix users subsidize those costs), whole opponents would permit the imposition of those costs on the particular users."

There is nothing preventing internet providers from offering capped data plans for home internet the same way they do so for cell phone data. They don't do so because those plans are very unpopular.

Providers can, and do, impose data caps.

And your point? Netflix has customers because of the choice and price. They are losing customers here because of lack of choice, not because of some nefarious prior pipe dream.

Comcast is losing their fast, reliable feed of video market because their customers don't want it. If they really wanted a monopoly they would not offer internet, it cuts into their cable customer base. Let's see how long before their management and board is employed after making such a stupid decision.

Poor network latency is a design consideration for anyone coming up with a service. If someone has specific service demands, in the real world they need to either wire it themselves or get someone to wire it up for them. That already is the reality. Net Neutrality is about preventing such arrangements to be made.

One can be for net neutrality without being for Net Neutrality(TM).

As I say, I am fine with latency hints, and even collocation.

Put differently/redundantly, in a peer to peer network no one can police all packets to see that latency requests/marking is correct.

Indeed it may be explicitly marked falsely, to speed an attack of one sort or another.

Thus an honorable, and by my lights neutral, node would try to pass on as many low latency requests as soon as possible, but possibly doing triage, especially under suspected attack.

I am a big fan of this site, and in general of your thinking, but in this case, yesterday's Bloomberg column and today's post add up to an amazing pile of ideological nonsense.

The bit about users opting for the "Walled Garden" of Facebook as an argument for non-net neutrality takes the cake. It reads like you are on the payroll of the broadband carriers, or are confused about the actual architecture of the internet,

Clockwork_prior seems to get it : " It might help to understand what the Internet actually is – a way for any connected digital device that uses shared protocols to request and receive data from any other connected device that uses those protocols. This is why a protocol like Bittorent can take off, as it provides a more efficient way to share data, without requiring a new network to be constructed."

And what in the world does that even mean? prior is proving himself the idiot that we knew he is.

An example. The network I use at home slows down on sunday morning. People get up, watch a movie, check out facebook, whatever. Every weekend it is the same thing.

Say I need a reliable connection on Sunday morning for some specific purpose. It has to work. It is time critical.

Prior would say some count on the magic of network protocols to solve my problem.

In the real world I would make some arrangement with my carrier, in exchange for money, to get the service level that I need. Or the source of what I need would design their application to work in awful network situations, and the costs would be reflected in what I pay. Or maybe the application designers would, to serve their customers effectively, make arrangements with the carriers to provide the level of service that is required. In exchange for money.

The communist prior would prefer to take the price signal out of the equation in some desire to replicate the successes of the Soviet Union I suppose, but in the real world solutions to problems are bought by those needing the solution.

Hi derek,

I am a telecoms and physics engineer, so I will try to explain prior's comment. You may "know" he is an idiot, but I am not so privileged. His sentence is perfectly clear, cogent, and correct.

The internet is a protocol for sharing information (files) amongst connected devices who use this protocol. This simplicity allows engineers to leverage this and find new and creative new technologies such peer-to-peer networks. (An application of which is torrents.) Do I need to point out that such innovations would be impossible if the Comcast model was distributed widely under non neutrality?

As a consumer, I definitely prefer a world where I can get my internet straight up, thank you very much. Your point about guaranteed bandwidth is besides the actual point. If I have such a critical need for guaranteed bandwidth, I might very well engineer a specific solution, but this seems like a typical straw-man argument in this context.

I fully understand how it works. Well funded and well designed networks failed and went out of existence because of the characteristics you describe.

This discussion is something akin to demanding that ethernet be mandated as a network topology because otherwise someone would come up with something better and that would be awful.

We need a law to make Compuserve illegal. Otherwise we will lose bittorrent.

Utter nonsense. The internet protocol serves the needs of the marketplace, and for that reason is dominant. If it no longer for some reason that is meaningful to a large enough market segment, then something will replace it.

By the way the characteristics you describe mean that we all share the network with millions of devices that are essentially a huge numbers of attack vectors. This year amazon's hosting went down for the better part of a day as the result of an attack. The open nature of the network no longer exists due to the use of the open network as a means of doing harm. That characteristic broke the monopoly of Microsoft, and raises the costs and risks of new entrants into the marketplace.

The idea that any of this would be solved by cementing in stone a utopian version of the 'internet' that existed a decade ago is ridiculous.

So, what you are suggesting, is that since the internet is not perfect, and does not guarantee us against attacks vectors from the internet of things, we should scrap net neutrality? Holy straw-men, Batman!

Also, you wrote : "This discussion is something akin to demanding that ethernet be mandated as a network topology because otherwise someone would come up with something better and that would be awful."

No, it's not. The internet is not a "product" sold by Comcast and their ilk. They choose to get into the business of selling bandwidth via their leveraged infrastructure. They don't sell any kind of network topology. If they had something interesting to offer that is superior to what we have, I am sure that consumers everywhere would be lining up.

Your whole spiel is pure obfuscation.

"So, what you are suggesting, is that since the internet is not perfect, and does not guarantee us against attacks vectors from the internet of things, we should scrap net neutrality? Holy straw-men, Batman!"

Jean, the burden of proof is on you. There is no reason to suggest that Net Neutrality has done any good. You should require proof before regulation. You claim to be a telecom engineer so you should know that packet priority is a necessity for a modern network. To claim otherwise is a fundamental misunderstanding of networking. So far, we have been able to build big enough pipes that this is only a trivial nuisance, but traffic will increase and necessitate prioritization more so.

There may be imperfect analogies in Tyler's post, but it should be easy enough to figure out what he means. Some people here are characterizing economic analogies as technical ones. If there is one that does not make sense to you then look at it as economic rather than technical and it will fall in line.

Derek, just as a note - I am a big fan of the Piratenpartei, and think the 'communists' are generally a bunch of totalitarian scum.

Bit torrent didn't take off because of some technical brilliance.

What did it provide? Free movies and music.

Wow so if I don't charge for tacos you'd be blown away at people lining up at lunch?

The technical brilliance is the peer-to-peer technology which underlies it.

But just like with MP3, remember that the real problem is piracy.

The fact that bittorrent is now routinely used as a way to distribute data (do you use Steam?) has nothing to do with piracy, by the way - it is simply is the most efficient way to distribute large amounts of data in coordinated fashion, being able to build temporary networks of enormous effective bandwidth. And to yet again keep this discussion relevant, bittorrent, both clients and various ways to recognize and then block bittorrent use were also part of standard filtering - in large part because bittorrent was able to utilize unused capacity. The Netflix/video surge is pretty much just in America - the bittorrent surge was global.

The main difference is that with bittorrent, the ISPs could only blame their users for actually using the Internet in the fashion they paid for, which is not really an effective strategy in most places (America does seem exceptional, however). In this case, the ISPs can now blame someone other than their customers, at least in public.But it remains the same fraud - sell customers something that is not actually delivered.

One critical element that everyone seems to miss in talking about net neutrality is that both end users *and* web hosts are already paying for a service.

You're arguing, here, in favor of permitting major service providers, without which the Web hosts cannot exist at all, and whose services have already been paid for at both ends, to unilaterally slow, or stop, those services - which I'm going to emphasize again were already paid for - in favor of whatever Web host is affiliated with that particular service provider.

"But - it's *my* service! " is the exact same argument Microsoft used to try to defend their bundling practices with Internet Explorer, and it wasn't ok there; what makes it ok here, especially when *the services are already paid for on both ends* ?

And your argument about Facebook is riddled with logical fallacies. People didn't choose Facebook because it's more closed off; they chose Facebook because at the time it provided a range of services you couldn't get anywhere else, and this far is still trying to do that. The average end user doesn't know, and doesn't care, how closed off it is. The two are unrelated, and you're trying to tie them together for the sake of your argument, but it remains a false equivalence.

"Already paid for" is meaningless, because the unexpected additional build-out costs well still have to be borne by someone. The question is whether Netflix users bear those costs essentially as user fees, or if the whole internet gets "taxed."

Right, but it is opaque to us how costs change over time. My first Cable ISP was about $20 for 1mbps. Now I pay $50 for 50mbps (which is amazingly the base tier).

In the background we have had


If the costs are "unexpected" that just means the company in question is managed incompetently. Any decent ISP will be monitoring their network utilization, and available excess capacity is always necessary, generally for easily predictable times (like the latest iOS release or Monday Night Football stream). If they're not already planning for stuff like this, they deserve to go out of business.

The bigger problem here is the local monopoly/duopoly status. Net Neutrality would not have to be a thing at all, as there would be actual competition to take care of the issue.

As a libertarian, I dislike unnecessary regulation. The _vastly better_ solution here would be to end the last-mile restrictions (I'd like to point out that I live in Austin, a very high-tech city, and have a choice of AT&T or SuddenLink. Period. They've got government-granted monopoly status. And since I'm not in an area Google Fiber -- they've got the money to buy exceptions to those grants -- is planned to build out, AT&T does not offer their GigaPower service in my area. Three blocks away, they do, because Google plans to be there.), but that's obviously not going to happen.

The least-problematic solution that's actually feasible is a return to the line-sharing requirements of the Telecommunications Act of 1996. While those were in effect, we had great competition, and I was able to easily get what was at the time great bandwidth for quite reasonable prices.

Another possible solution, although one that isn't politically feasible everywhere, is having the actual cable/fiber/copper plant run as a public utility, and letting anyone who wants to start a business offer services over it. The current telecoms are rabidly against this option, of course, and part of the regulations that they're trying to get implemented would block this.

Third possibility (low I grant you) is the reenactment of the 1866 Post Roads Act, but have it apply to all telecom providers this time and not just the telegraph industry.

The 1866 Post Roads act preempted state and local franchise regulations by granting a de facto national franchise. The act was particularly powerful because it granted companies the ability to build and operate as a right. If a state and local law prevented them from doing so, the courts allowed the companies to ignore the laws.

You can learn more in the dissertation linked below.


I don't know about you, but I pay for my internet service one month at a time. I guess I'd be angry if Comcast rejiggered the terms of my service in the middle of a month, but if they did it with a few months notice, what would I really have to complain about?

This is a smart comeback. I'd argue that a monopoly service provider has to tier their service in a content-neutral way. I, for instance, don't have a serious problem with T-Mobile throttling their "unlimited" plan, provided the cap is content neutral (it is, if course, deceptive advertising) (and, by my understanding, they don't actually throttle in a content neutral way).

God! This, this, a thousand times this! I ALREADY PAY FOR THE SERVICE. Comcast has no reasonable right to double charge me. Putting it in the fine print just shows that they're a monopolist who needs to be broken up.

By the way, some years ago our very smart VP went to ICANN to sell them something. He came back with his mind blown and 5 or 6 ways to improve our product.

If Tyler could do a conversation with a backbone architect that would be great, and probably cut across most PR from FAANG.

Tyler quoted a cisco VP and engineer in the last post. Basically this is useless at best and goes down from there. 'Puts the internet back 20 years.' All of which is an honest answer to the technology behind the internet. A lot of people posting true things about the internet and networking that have nothing to do with network neutrality.

The real benefits of net non neutrality would be applications that require a guaranteed minimum latency. Non net neutrality would allow some market participants to pay more for reduced latency, which could benefit video conferencing, virtual reality, remote surgeries, VOIP (already part of video conferencing) and other possibly new applications, say remote monitoring and control various kinds.
People have been doing this since the beginning of internet.

I think this is an interesting argument, but if the market is large, they'd be better off building a parallel system than they would crippling the Internet as we know it.

Why go to the incredible expense of a separate internet (although there is one now, not available to you). A little quality of service goes a long way it you don't ignorantly make it illegal.

The de Beers analogy was not very good. De Beers had a monopoly on a commodity, not a platform. Therefore they had no way to impose their preferred soda on anybody because they didn't have a monopoly on a system that delivered soda. The major ISPs have local monopolies of a platform and can use that monopoly power to push their preference of applications on the user. We see this actually happen in places like portugal (http://www.businessinsider.com/net-neutrality-portugal-how-american-internet-could-look-fcc-2017-11).

I think Tyler's post made two decent points though:
1. "More generally, I don’t see anything intrinsically morally wrong with a person deciding to “buy only one third of the internet.” How many net neutrality supporters also favor or maybe even insist upon a’la carte pricing for cable TV? What percentage of the public library do they take home over the course of their lifetime?"
But without competition why would any monopoly ISP create lower cost alternative packages for people to access only certain sites? I think it's much more likely that they would simply raise normal rates and offer these "discounted" packages as an alternative.

2. "Finally, Viking notes in the MR comments:

The real benefits of net non neutrality would be applications that require a guaranteed minimum latency. Non net neutrality would allow some market participants to pay more for reduced latency, which could benefit video conferencing, virtual reality, remote surgeries, VOIP (already part of video conferencing) and other possibly new applications, say remote monitoring and control various kinds.

Are the defenders of net neutrality considering those opportunity costs in their assessments? I don’t see it."
Again, this sounds totally reasonable. This just ends up highlighting the problem of having telecom oligopoly though, since the real solution would be to invest in upgraded infrastructure to ensure that bandwidth and latency are at the levels that consumers demand. If there were open competition, new upstarts would create infrastructure to meet this demand instead of requiring regulations on the existing infrastructure.

The problem with giving up net neutrality is that it leaves a critical element of a contractual arrangement between the parties open ended for the ISP. When I pay for my access, the ISP would not be required to make any commitment of performance to me. I would know the numerator, but not the denominator.

All ISP's are granted public accommodation which allows them to stand up their networks. We don't have to let them create arrangements that are better for them than us. It's really that simple.

So you want to expropriate their property and remove pricing signals from the marketplace.

That should work really well.

No, we want transparent pricing. It's not actually that complicated to anyone who wasn't a Verizon lawyer.

The equipment may be their property, but the means of egress is a public good. This is not a natural private service. It only exists with public accommodation. We are granting large oligopolies in exchange for public consideration. Such an arrangement can only work when those oligopoly rights are 100% constrained by terms. Granting them an undefined level of control over the flow of service is stupid.

I think most people are over-generalizing this. Text and static image content is trivial bandwidth for the net now, and Comcast/etc. have absolutely no incentive to limit it. This whole debate is really about television.

The cable TV networks had a nice situation -- monopoly provider of video to the household. They could charge for channel packages, charge content providers to get on the system, and sell advertising. None of that works in the internet world.

If everyone is just going to Netflix or YouTube, the cable company is reduced to a commodity provider of bandwidth. They lose any leverage and any ability to sell ads or access. Net Neutrality just enforces this.

So they hate NN. They want to rebuild the internet as channel packages they can sell, charge Netflix/YouTube for bandwidth, and perhaps stick ads in there somewhere as well. Recreate what they are losing, in other words.

There are no political/censorship effects. They can let all the text/image/audio traffic through and not care a bit. That stuff doesn't impact their existing business model, and the bandwidth requirements are trivial.

In the long run, I think they are dead anyway. No one wants channels and bundles and separate fees for YouTube, etc. They will constantly be vulnerable to another ISP that comes along and gives unmetered service.

This is not some "death of the internet". It's the death throes of television networks.

Good comment.

Very good comment, although government intervention, a la the current FCC proposal, could prevent the market from working. Seen that way, it looks like Ma Bell in the '70s.

The comments on this topic are unusually good and interesting. It's been a long time (mid80's) since I had any real technical involvement in telecom, but I can follow the arguments.

It makes me wonder how many other topics are being pronounced on by economists who don't understand the basic technical / physical realities of a topic. That would make an interesting list / post / discussion topic for Tyler. Autonomous vehicles ? Bio-engineering? Energy systems? Cyber security? It's pretty clear most of the politicians they are advising don't have the first clue.

TC's arguments don't require a technical understanding of the OSI model for communications,and are technology neutral.

Bonus trivia: notice how often the comments here hover at 64 or 63 or 65 comments (sixty-four being the number of squares on a chessboard). Some sort of behavioral economics phenomena at work. This comment will be the 64th+1 comment.

The major problem that you haven't even considered is the effect on political discourse over the Internet. Major services such as Facebook and Twitter have defined all viewpoints to the right of Stalin as "hate speech" and will ban you for expressing them. Youtube is in the process of doing the same, and the rest of Google and Yahoo won't be far behind. This was not a problem when the Internet was wide open, but recently Gab (a Twitter competitor) nearly lost its domain name because of a similar policy by its upstream provider. At this rate the whole Internet will be for leftists only within a year unless something is done to prevent it. And I don't trust FCC or FTC to use such a power fairly even were it given to them.

About all I can suggest doing is to actively resist any further mergers among communications multinationals, while fostering and encouraging new competition with all of them to the extent there still is any new competition.

Marginal Revolution is to the right of Stalin and I don't see anyone trying to shut it down.

An example of taking literally something that should be taken seriously.

Tim Ferris recently moved to Austin after 17 years in Silicon Valley. One of his reasons was that even as a social liberal the Valley is becoming an uncomfortable PC monoculture where people are afraid to voice unorthodox viewpoints.

The point about dogmatic authoritarians controlling access to information is a valid one.

Why is there no mention of franchise agreements in this post?

Why does "net neutrality" continue to be discussed without reference to the competition restricting activities of municipalities through franchising?

Because that's not an exciting anti market meme that can virtue signal.

If you want to talk about unfair regulations,then let's talk about laws that prevent municipalities from providing internet access for their residents. If we got rid of these laws maybe rural areas would get internet access. Of course the large ISPs don't care about that. So instead of having a conversation about increasing rural access we talk about net neutrality.

Sounds bogus. Usually rural areas are not profitable and laws are passed so city dwellers subsidize rural dwellers.

It is a specific thing


Turkey in the smoker, smelling pretty good.

If you have bad actors in a global peer to peer(*) network, is there a guarantee that a startup can appear to route around malfeasance?

I think not. Thus government action is required.

* - still in spirit, if limited by practicalities, and still a worthy goal, as technology advances

I think Mr. Viking has not heard about symmetrical internet connections. If you wan to do serious videoconf or surgery, it's possible today to have those premium services. No need to change laws.

Been around forever. Not really relevant to NN.

Here's a different take on this reality:

It's about people fearing that they will no longer be able to get other people's attention more or less randomly. It's really not about limits to publishing or being heard (though there is some censorship afoot) - it's about how terribly difficult it is to get anybody's attention. Google and facebook have become elephants largely by providing something the mass market wants, and then finding out how to monitize it. Not at all different from MSM publishing high emotion click-bait to drive ad sales.

You see this in the MSM complaints about what facebook has done to ad revenues, and bloggers longing for "the good old days before facebook".

Why did people flock to facebook? They didn't actually care about blogs, they want to see cute cat pictures from their remote friends.

If you want to control censorship, change the DNS protocols to be fully distributed so people don't buy DNS entries from top domains.

If you want to control ISPs, phone companies, etc., impose revenue purity rules - saying that ATT can ONLY earn revenue from selling bandwidth to providers and consumers. No revenue from content, no revenue from advertising, no reveue from selling user data - or all of those taxed at 1000%. Force internet connections of all sorts to be pure commodities....

I take some issue with your water analogy. I don't disagree with your argument in the case of water, but I do disagree that it is as applicable to the issue of net neutrality as you claim it is. The Internet is different from water - whatever country you get water from doesn't change the information you have, or the opinions you hold. Water is water, no matter where it's from. Whereas access to certain websites has a huge influence on people's thoughts and beliefs. To get rid of net neutrality would be to give ISPs the ability to control (although indirectly) the information that people can get.

This is one of the biggest gripes people have (or at least one of the biggest gripes I have) with lack of net neutrality. It's scary giving ISPs more control over the content we can consume, and whether or not they're likely to abuse that influence in the near future is immaterial. Simply giving ISPs that power opens the door to the potential to abuse that influence in the future.

My town signed a contract with McDonald's to be the only restaurant. I'm now worried that they will raise prices or offer bad food.

My solution is to have the FDA regulate McDonald's to insure quality and set prices.

That's pretty much how US regulated utilities (electricity, water) work; is it different in India?

Roughly 70% of my web site consumption is via RSS, 25% search and 5% links from Twitter

Pretty much 100% of my web site consumption is this site, a few others like Wikipedia, e-books from Piratebay, chess online. But we are not the target audience Tom Davies. The target audience is the 80-100 IQ crowd, that's who is swayed by adverts, gate keepers and social media shapers.

Ray, I think I've seen you over on seekingalpha also - if it's the same Ray Lopez?

Yes, that's me.

Tyler is posting some deeply reprehensible shit here, and you really have to wonder why he is selling it so hard and being so deceptive about it.

Diamond companies forcing you to buy soda? AYFKM?

He's telling you that now that the world has the greatest content delivery system in its history, it's totally fine if a handful of people at enormous government-connected companies control what you can see on it.

If the Trump admin gets some leverage on Comcast, and leans on them until they slow down traffic to CNN.com to 1% of normal speeds? And maybe offers a $500/month package to double that speed? Totally fine with Tyler.

Or if Al Franken's next job is running Comcast, and he does the same thing with Fox News? Totally fine with Tyler.

Pretty sick.

You're worried that reducing the US government's regulation of internet content is going to increase the government's influence over internet content?

Most of my thoughts on net neutrality were influenced by reading Lawrence Lessig about 18 years ago. If you are looking for common sense arguments, they are compelling.

That said, I completely agree with Tyler that this debate, like many others, is clouded by emotion and dystopian scenarios.

I also believe that the current U.S. administration´s scorched earth policy (roll back anything done by the previous administration) is at least partially responsible, which means this is not about net neutrality at all. That is the current dystopian reality we live in.

https://www.technologyreview.com/s/421384/qa-lawrence-lessig/ for some of the compelling arguments.

I would like to throw this article in here concerning antitrust laws and net neutrality:


> What percentage of the public library do they take home over the course of their lifetime?

This is a really bad analogy. The point is that the public library doesn't discriminate what content is provided to the public. It doesn't matter if any one individual doesn't consume everything that is available.

Comments for this post are closed