AT&T, Time Warner, and the Need for Neutrality
stratechery.com
stratechery.com
This piece addresses the philosophy of the potential issues better than some, but this net neutrality hysteria in general seems to me like a lot of well intentioned folks are trying to solve a problem which does not exist, despite having had decades to develop (while the associated technology is still rapidly transforming, the Internet is no longer such an immature, radical new thing). Adding controversial regulations to the most innovative technology sector in the US to solve a problem which might develop (and could in all likelihood be undone very quickly and without significant infrastructure changes, were this to begin to develop into a problem) seems ridiculous to me. I imagine there are many better trees to bark up.
But to entertain the idea, of the three final 'key principles' the author lists, the first two are reasonable and unlikely to break things (if it was possible for Congress to write a one-page bill that contains only the language necessary to implement these rules, and nothing else). The third one is a slap in the face of innovation and new services in my eyes, although the author acknowledges it may be extreme.
The only sweeping public policy that I could support to address a problem in this space would be for one which already DOES exist: the lack of competition in the ISP space for the last mile. A possible public solution could include cities implementing a basic cabling system that cover all residences in urban areas (independent from any other private cables, which would likely be superior) that any ISP can connect to and serve via a city hub (much closer to the backbone, with an access point being outside the urban area, allowing many competitors to build infrastructure around it).
The problem doesn't exist yet because a) the spirit of the early Internet is one of academic collaboration, b) early ISPs did not have much market power and could grow rapidly without squeezing their customers, c) existing telcos were bound by common carrier laws, and d) there has been substantial pushback from citizens, companies, and politicians during previous attempts to overturn the status quo.
This is the sort of excuse that was used in the run-up to the 2008 crash. Sometimes things are fine until they aren't.
> could in all likelihood be undone very quickly
Oh, really? There's some reason to think the US government would quickly act to reduce the income of large corporations by billions? Perhaps you can give an example of previous legislation that you think demonstrates your case?
> competition in the ISP space for the last mile
I am entirely in favor of this, but I believe it's in practice unachievable. I happen to be in one of the few places where I get competitive access. It's great! But for most places in the US, that ship sailed 15 years ago. Existing telcos have been very good at extinguishing competition and preventing significant new rivals from emerging.
Markets can be very good regulators. But when there is no market, regulation is a reasonable and necessary alternative.
https://arstechnica.com/information-technology/2017/06/50-mi...
Most people live in areas where people do want to run broadband. But they still don't have a competitive market; only 5% of people have at least 3 options, which is the point at which you get real competition.
There are multiple instances of internet service providers blocking VOIP in order to promote their telephone services, among other violations of net neutrality. They mostly got shut down by the FCC, until the FCC was sued claiming they didn't have the power to do this as internet wasn't title 2.
Basically the problem didn't exist only because it was regulated against. Now it is no longer regulated against.
Netflix didn't just pay. They battled it for a huge amount of time.
Who cares if last mile infrastructure is redundant? Redundancy is good most other places, and having more than one fiber line running past my house doesn't sound like a bad thing at all. Even if it was definitively inefficient, it would still be better for consumers than any monopoly on such infrastructure, regulated or not.
People need to just let ISPs run whatever cables they want wherever they want, whether that be attaching them to overhead power lines or under ground.
We should have competition at every stage of the internet, from backbone to search engines to last mile connectivity.
Much like the debate around affordable housing, the answer is to "let them build!" There are companies like Google/Alphabet willing to build out infrastructure if only they were allowed.
For an example, see Piteå's: https://portal.piteenergi.se
The problem is that when ISPs are running their cables, they also have a distressing tendency to cut other people's cables. See https://arstechnica.com/tech-policy/2017/06/lawsuit-comcast-... for an example where Comcast did that. In the discussion about it at https://news.ycombinator.com/item?id=14613986 you will find many other anecdotal stories of similar behavior.
Anyways it doesn't matter because it really is a natural monopoly. http://www.dtc.umn.edu/publications/reports/2003_11.pdf is old, but the economic principles that it cites for why broadband is a natural monopoly remain true today.
You think malice, I think incompetence. Cuts happen. ATT cut the Comcast cable in my neighborhood while laying fiber. They got it fixed and that was that.
But regardless, there is a very legitimate reason for the people who already have cables installed to object to the next company to try. This naturally leads to increased regulatory costs in a dense urban environment.
This is the case to be made, and I don't see where you're making it here. You're just assuming it to be true, then repeating it. The question is: why would the waste caused by the inefficiency be less costly than regulation? Or: why would regulation be necessarily less effective for customers than arbitrary redundancy at every level? Or come up with another question, and answer it. "Unregulated competition is better than regulated monopoly" is just a religion.
Access rules could be tweaked but any way you look at it connecting every residence in the US with wires is very slow and expensive process and it basically fits the definition of natural monopoly.
Easements already exist for this purpose.
This is true, and I agree that enabling infrastructure development is an important aspect.
> any monopoly on such infrastructure, regulated or not.
I would never propose that the state monopolizes any aspect of the infrastructure, but rather that it offers an alternative while also enabling private infrastructure development (which as I mentioned, would likely have the most advanced infrastructure). I think there is a case to be made that the externalities of multi-layered private cabling in densely populated areas are sufficiently prohibitive to warrant a public alternative.
It will take a while for cable and telecom companies to take advantage of the lack of net neutrality, but I am sure they are already considering it. The only thing keeping them back nowadays, is the obvious weakness of the current administration and the uncertainty as to what the next administration will bring.
So the reason the "problem does not exist" is because the problem behavior has been banned until recently and may be banned again soon.
You may as well say "why ban slavery, it is clearly a problem that does not exist in the US, therefore the thirteenth amendment is useless."
Your later argument is a clever misdirection. "Let us all instead of arguing for this problematic net neutrality argue for something else that sounds better on the face of it, but is unlikely to ever happen." Sure I would prefer it if there was more competition at the local level, most people that are not connected to the telecom industry would. But that would be so incredibly expensive that it is unlikely to ever happen. Best case scenario it may happen in 20 years if we are lucky. And if you smother our internet industry in a monopoly for 20 years, it is unlikely to ever recover.
But hey, if we get real competition for local internet access, I am very much willing to give up net neutrality. But meanwhile, for those 20 or more years, while we are waiting for that competition to emerge, we must have net neutrality.
Net neutrality regulations were not created in the US until 2014. It is ignorant to think these stopped companies before 2014.
Such regulations were in place before that. Companies started questioning the authority which resulted in new laws needing to be made.
Saying the law is recent, unneeded, etc has been mentioned and refuted many times in previous discussions.
It hasn't developed because the FCC has previously been actively trying to prevent it from happening, resulting in a few different policy statements, rules, and court cases. We've had some level of net neutrality rules on the books for over a decade. It wasn't until recently they were removed. You're ignoring history by making this claim.
http://www.techlawjournal.com/topstories/2005/20050805.asp https://docs.fcc.gov/public/attachments/FCC-05-151A1.pdf
It very much does exist. https://www.freepress.net/our-response/expert-analysis/expla...
That's what "Software Defined Networking" is for; it is pure evil in an open source wrapper, why else would telecoms be contributing to it?
Digging trechnes is expensive, boring and critical path for kickstarting diversity of ISP competition.
No, we do not NEED Net Neutrality. The internet can/has/will survived for the short ~2ish years of really weak NN that was easily circumnavigable by corporate lawyers. The last form of NN we had allows Comcasts of the USA to legalize their local monopolies.
The solution is multifaceted. First, if we want to break free of the stranglehold, we need to lower the barrier of entry into the market for new ISPs, not raise it. Second, the FTC needs to examine the practice of "up to 50mbps" advertising and "internet service" vs "information service". Strict definitions of these would help us along. Third, vague definitions of "service abusers" needs to be eliminated or defined. If they are selling you "unlimited internet", then there is in fact, no limit. If you hit a limit by an ISP, they need to be reported in your bill when and where the policy violation took place, plus allow for a dispute process. Finally, there are a lot more dirty things ISPs are doing, like Verizon injecting HTTP headers that identify an account to advertisers.
The 2015 NN laws were too weak to accomplish what they sought out to do. I'm happy they're gone so we can possibly get some effective regulations in.
Running the last mile cabling is inherently naturally monopolistic. How many companies run power lines to your house? How many companies run sewer lines to your house? Water mains? Ultimately, we either need to have highly regulated last-mile providers (strong net neutrality/common carrier/title II) or have municipal last mile providers.
Personally I'd love to see a system like ERCOT[0] implemented. Last mile gets contracted out to some company a little over cost with regulations about ensuring capacity growth is planned for, while power delivery is handled by retail electric providers. Your home gets connected to some exchange a few miles away. You want to go through Comcast's network? Just give them the info off the ONT. Want to switch providers? Log into the other provider's website, give them your information, a routing change gets pushed out, and suddenly your packets are going out through CenturyLink or anyone else at the exchange. In reality its the last mile that's highly monopolistic, in most metro areas of the US there are probably at least half a dozen transit providers out there. They just don't want to spend the money digging trenches to everyone's homes to get 10% of the population of the area.
> the FTC needs to examine the practice of "up to 50mbps"
I'd really hope not, at least not too aggressively. I greatly prefer my best-effort 100Mbps fiber internet for $30 a month compared to the hundreds of dollars a month for SLA-backed 100Mbit fiber at my office. If they had to guarantee service, you'd see much slower average speeds offered to consumers at higher prices. Oversubscription of capacity is how residential internet gives us such high speeds when we're downloading a 100GB game randomly in the evening for not a lot of money. If you want an SLA, pay for the SLA. There are plenty of us that are fine with all the speed most of the time, with sometimes being a bit slower.
[0] https://en.wikipedia.org/wiki/Electric_Reliability_Council_o...
Oh for sure! Maybe I overstated my point. Reasonable best efforts play a huge part in making affordable residential service. But right now, there's no way for a consumer to compare what's realistic, driving down competition, because they "assume they're all the same"
Therefore a strong regulation can follow deregulation.
Of course the opposite can happen. Deregulation can shift the Overton window to the point where regulations that once seemed reasonable are off the table. As an example, even most of the people complaining about growing inequality between rich and poor aren't openly suggesting a return to pre-Reagan top tax rates of 70%.
Net neutrality was in place much longer than this. Furthermore, there were put in place because companies continuously started to abuse them.
> need to lower the barrier of entry into the market for new ISPs, not raise it
That has nothing to do with net neutrality. Treating everything pretty much the same is way easier than building complicated billing systems.
There is nothing wrong with a company providing all these services, so long as they are not able to abuse their position.
Now, net neutrality is _a_ not very effective way of stopping dominance. But, its almost not what you think or want it to be. What you actually want is competition, so that when a company does something stupid (ie slow netflix to a crawl) you can up sticks and move to a better service. (no I don't mean a duopoly)
Now, you've just seen me say that NN isn't what you want. Let me unpack that rather controversial statement. Firstly NN is the equal treatment of _all_ traffic regardless of source, protocol or destination.
You really don't want that, as it means that real time stuff will get hampered by bulk traffic, and bulk traffic knobs (you know the people that are more than willing to spend $25 a month on a torrent box, $500 on local storage each quarter, but not pay for content.) Will get a free pass.
On the flip side, it means offers like http://www.three.co.uk/go-binge/shows (watching netflix doesn't count towards your mobile internet cap) can't happen. Now, judging by whats happening to Three, this is what customers actually want.
So, what America needs is bulk transit ripped away from the big three. Where there is ISP dominance in one area, Regulated wholesale prices, allowing virtual operators to provide internet using rented last mile infrastructure.
This is not perfect, as unless you fully strip those providing last mile infrastructure from those providing ISP functions, there will be a reluctance to roll out new technologies in less profitable areas.
I disagree. Net neutrality is not a problem. See [0] for a previous take on the misguided notion that QoS and net neutrality are at odds.
It might be an ideal solution if we could nationalize network infrastructure and guarantee local loop unbundling, as I think you are suggesting. But we are so far away from that possibility in the US that you seem to be advocating against net neutrality in favor of effectively the status quo.
Let's not throw support for net neutrality out the window in favor of a far away pipe dream or for unproved, supposed benefits from getting rid of it. Let's get back net neutrality first, in legislation this time, then see if we still think we need to nationalize the network infrastructure.
If I am misunderstanding your proposal then maybe you could clarify.
What the US has now is a free and open market, well the logical conclusion of it. (the carnivorous rats on the island model) The lack of regulation to stop consolidation is a big problem. The need right now is for AT&T to not merge with time warner.
I don't think nationalisation is what is required, I think separation of last mile infrastructure from client ISPs. This means that the Motivation for the infrastructure company is to provide stable reliable last mile connections. Because they are going to be used regardless of who is supplying connectivity, it doesn't affect investment. All of this can be achieved with private companies (indeed thats how it works in most EU countries)
This also lowers the cost of ISP startup drastically.
As for QoS, I love QoS, it means that my ADSL connection can still have low latency stuff (audio, video games) even though I share the same uplink with torrenting arseholes. But, I don't see how, with the weakness that is the FCC, they are going to legally distinguish between QoS to regulate service (ie bulk downloading via torrenting) and bulk downloading of a video on demand service.
Both have the same strain, both are video services, both can be throttled to mask the lack of investment in backhaul infrastructure. both need to be "shaped" to allow customers to use the ISP's "services" evenly.
I'd also like to point out that NN has only been a thing for <2 years.
Regarding QoS, why couldn't you as the client choose which packets are to be considered high-priority? Is it that you want to be able to saturate your connection with high priority packets that supercede your neighbor's packets if you're on a video call? Maybe a NN-friendly compromise is that everyone gets a % of their connection that get to be high-priority packets (you choose which) and you can pay more for a bigger pipe. Therefore if you want 1M up at high priority you need a 5M pipe. Your neighbor can keep their 1M pipe but only 200k of their packets can be high priority. (to throw out some numbers). It seems fair and gives you QoS options while not violating NN, which I think is overall more important than QoS anyway.
This is not true. The most recent law was in place for 2 years. It existed in other forms for a long time before that.
> You really don't want that, as it means that real time stuff will get hampered by bulk traffic, and bulk traffic knobs (you know the people that are more than willing to spend $25 a month on a torrent box, $500 on local storage each quarter, but not pay for content.) Will get a free pass.
No, I do want this. Just meter your service, stop selling it as unlimited, and either have the capacity to serve the customers that you've acquired, drop customers, or add capacity. Don't discriminate between a $25 torrent box and the company Netflix because that's what you think the people want.
Netflix is the example that works best, as the "man on the Clapham omnibus" can easily understand what it is, and why throttling it is bad for the customer.
Metering doesn't get round this, as that doesn't do anything much to cap peak load, there is still no reason for an ISP to invest in transit, or indeed allowing more edge caching boxes. (the customer isn't going to go away, so why put money into expanding edge footprint)
The only way capacity is going to be built is if there is an incentive (be it legal, cash or other) Why would an ISP drop paying customers _and_ invest in extra infrastructure?
So if we can't adopt some rule that would treat privately-owned cables, let's say on the last mile, as common carriers, what are some other options? Deny further vertical mergers affecting telecom infrastructure? This doesn't solve the current entrenchment. Let localities take the issue into their own hands and build out publicly-owned networks that would support open access for providers running on top? This is a sensible idea, that's currently difficult to pursue in almost half of the states [5], and most legislative efforts to make this difficult have come from Republican lawmakers [6][7], despite citizens of all political leanings in favor of localities being allowed to do this [8][9].
[1] https://medium.com/@TebbaVonMathenstien/network-neutrality-a... [2] https://en.wikipedia.org/wiki/FCC_Open_Internet_Order_2010 [3] https://en.wikipedia.org/wiki/Verizon_Communications_Inc._v.... [4] https://arstechnica.com/information-technology/2015/02/fcc-v... [5] https://broadbandnow.com/report/municipal-broadband-roadbloc... [6] https://arstechnica.com/tech-policy/2017/10/another-state-la... [7] https://www.fischer.senate.gov/public/_cache/files/81c82846-... [8] https://muninetworks.org/content/pew-survey-reveals-overwhel... [9] http://www.pewresearch.org/fact-tank/2017/04/10/americans-ha...
both are legal and both are harming competition
This administration's objections to the AT&T/TWC merger was probably rooted in television because the administration's opinions about net neutrality are rather well-known; it's doubtful they'd see the eventual outcome, if they even envisioned it, as problematic or anticompetitive -- and that's just the charitable reading that omits the Predient's own upset over the future of CNN, which too reinforces approaching the subject with a television lens.
That being said, it is somewhat ironic that zero-rating was never considered outright problematic by Wheeler's FCC [1] (or the EFF [2]) and the FCC resolved to investigate zero-rating practices on a case-by-case basis. They let some of them fly, but admonished several others, even in the last few days of Wheeler's term [3]. Then, Pai predictably stopped all such investigations when he took office [2].
In this favorable regulatory climate, it's no wonder that zero-rating looks like a valuable tool that businesses can leverage to make their offerings seem more attractive versus their competitors, making past predictions about a future of vertically-integrated content silos and delivery networks [4][5][6][7][8] seem a bit less like quackery.
As an added complication, it is much more difficult to explain to laypeople why zero-rating is problematic, than other topics that can fall under the wider umbrella of 'net neutrality'. Zero-rating is perhaps the most popular facet of delivery that doesn't adhere to net neutrality, attracting a wide range of supporters [9], including many that support missions that are helped or enabled by no-cost access to information. Once zero-rating is commonplace, efforts to undo it will be more politically challenging than the previous net neutrality debate.
[1] https://transition.fcc.gov/Daily_Releases/Daily_Business/201... [2] https://www.eff.org/deeplinks/2017/02/fcc-abandons-zero-rati... [3] https://www.eff.org/files/2017/02/09/fcc-zerorating.pdf [4] https://news.ycombinator.com/item?id=17305508 [5] https://news.ycombinator.com/item?id=12350087 [6] https://news.ycombinator.com/item?id=14229742 [7] https://news.ycombinator.com/item?id=15092525 [8] https://news.ycombinator.com/item?id=13557318 [9] https://en.wikipedia.org/wiki/Net_neutrality#Arguments_again...
Only while anti-trust remains as crippled as it is now. In practice zero rating kind of violations should be blasted with classic anti-trust, not with specially crafted net neutrality rules. So may be fixing anti-trust should be a priority.
To be clear, I'm in support of strict net neutrality rules, but they are often a fix for a symptom, while the root cause is actually anti-competitive abuse that runs amok because anti-trust is doing nothing.